principles of magnetic testing ce betz

Few men are blessed with a character that combines a high order of mentality, friendli- ness for people, and a personal magnetism which draws friends to him. Carl E. Betz was one of those rare individuals.

Inventor, administrator, chemist, salesman, teacher, engineer, traveler, gourmet, father, grandfather, farmer, author, iecturer-all of these describe him.

Among all the men who have contributed to MAGNAFLUX, there are few, if any, who have done more than l l r . Betz. More than an engineer, he helped to lead the way into a new reaim of nondestructive testing.

It takes a special kind of courage and fore- sight to leave a solid job and join a new and untried company because of faith in a new concept. This is what Carl Betz did. The concept was nondestructive testing, the company, the recently established MAGNAFLUX Cor- poration. The time, the Great Depression (1935).

Mr. Betz was n native of Kansas City, Mis- sour], and received his degree a t the Univer- sity of Missouri a t Columbia. He graduated in 1913 with the degree of Chemlcal Engineer. He joined MAGNAFLUX Corporation in 1935 a s Technical Director after having spent 2Z years with Pittsburgh Testing Laboratory, the last 13 as Chief Chemist.

PRINCIPLES

of

M A G N E T I C PARTICLE T E S T I N G

by CARL E. BElZ

Chemrral Engrncrr Vicc President and Director (Reitred), M~gnailuv Corporatxon

Honorxy Life Member. Sooety tor Nondertructtve Tertlng hlember

Amerrcan Soctety for Metals Ameiicnn Sorsctr iar Tcrttng and Matertals

Amcncan Arroctatron for the Advancement of Soenre

First Edition

Published by Magnaflsx" A Divbr$on oTlllinotsToo1 \Vorks

Hniwood HclghIr. Illinois Febmav I , 1967

COPYRIGHT 2000 MAGNAFLUX"

A DIVISION OF ILLINOIS TOOL WORKS INC. HARWOOD HEIGHTS, ILLINOIS

ALL RIGHTS RESERVED INCLUDING THE RIGHT OF REPRODUCTION

IN WHOLE OR IN PART IN ANY FORM.

EIGHTEENTH PRINTING - NOVEMBER, 2000

PUBLISHED BY MAGNAFLUX'

A DIVISION OF ILLINOIS TOOL WORKS INC. HARWOOD HEIGHTS, ILLINOIS

MANUFACTURED IN THE UNITED STATES BY

MOBILE PRINT INC. MOUNT PROSPECT, ILLINOIS 60056

LIBRARY OF CONGRESS CATALOG CARD NUMBER: 66-29699

DEDICATION

T o ihe rr,m met, 1~ ,h0 h ~ d iibe r~isroti

lo ree lhe frrtlrre of i, trerc rden

and the corrrate and fitr/h

lo devore iherr lives

10 tnnhiog /hi, v,vos

become a redliry.

Alfred Victor de Forest 1888-1945

Foster Baird Doane 1893-1963

PRINCIPLES OF MAGNETIC PARTICLE TESTING

Page TITLEPAGE ............................................ 1 DEDICATION ............................................ 3 LISTOF CHAPTERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

" TABLE OF CONTENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 LIST OF ILLUSTRATIONS .................................. 23 FRONTISPIECE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

LIST OF CHAPTERS

CHAPTER TITLE Page

. . . . . . . . . . . . . . 1 . Histoly of the Magnetic Particle Method 47 . . . . . . . . . . . . . . . . 2 . Fundamental Concepts of the Method 60

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Sources of Defects 70 . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . How and Why Metais Fail 96

5 . Definitions of Some Terms Used in . . . . . . . . . . . . . . . . . . . . . . . . Magnetic Particle Testing 114

. . . . . . . . . . . . . . . . . . . 6 . Characteristics of Magnetic Fields 130 . . . . . 7 Methods and Means for Generating Magnetic Fields 141

. . . . . . 8 . Determination of Field Strength and Distribution 165 9 . Fieid Strength and Distribution in Symmetrical

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Objects 178 10 . Field Distribution in Large or Irregular-Shaped

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bodies 198 . . . . . . 11 . Magnetic Particles-Their Nature and Properties 209

. . . . . . . . . . . . . . . . . . . . . . 12 . Basic Variations in Technique 229 . . . . . . . . 13 . The Dry Method-Materials and Techniques 244

. . . . . . . . . . 14 . The Wet Method-Mater~ais and Techniques 255 15 . Fluorescent Magnetic Pafticles-Their Nature

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . and Use 274 .... 16 . Black Light-Its Nature, Sources and Requirements 290

7 . Demagnetization .................................. 306 . . . . . . . . . . . . . 1 8 . Equipment for Magnetic Particle Testing 325

19 . Automatic and Special Magnetic Particle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Testing Equipment 333

20 . Detectable Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366 21 . Non-Relevant Indications . . . . . . . . . . . . . . . . . . . . . . . . . . 382 22 . Interpretation, Evaluation and Recording of Results . . . . 392 23 . Industrial Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419 24 . Testing of Weldments. Large Castings and Forgings . . . . 440 25 . Standards and Specifications for Magnetic Particle

Testing 455 26 . Tests fo r Evaluation and Control of Equipment

and Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466 APPENDIX . . . . . . . . . . . . . . . . . . . . . . . t. . . . . . . . . . . . . . . . . . . . 475 BIBLIOGRAPHY ...................................... .. 491 INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504

TABLE OF CONTENTS

CHAPTER 1 HISTORY OF THE MAGHETIC PARTICLE METHOD

SECTION Page . . . . . . . . . . . . . . . . . . . . . . 1 Early Testing Methods 47

. . . . . . . . . . 2 Beginnings of Industrial Radiography 48 . . . . . . 3 The 3Iagnetic Particie Method Development 48

. . . . . . . . . . . . . . . 4 Progress Bet\veen 1930 and 1940 51 . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Early Equlpmcnt 51

. . . . . . . . . . . . . . . . 6 . Developments o f the 1940's 55 7 . Problems Due to Rapid Expansion of Use . . . . . . . . 56

. . . . . . . . . . . . . . . . . . . 8 Post-War Developn~ents 57 9 . Nuclear and Space-Age Requ~rements . . . . . . . . . . 59

10 . Future of Magnetic Particle Testing . . . . . . . . 59

CHAPTER 2 FUNDAMENTAL CONCEPTS OF TI-IE METHOD

. . . . . . . . . . . 1 What is the Magnetic Particie Method? 60 . . . . . . . . . . . . . . 2 How Does the Method Work? 61

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Magnetization 61 4 . Applying the Ferromagnetic Particles . . . . . . . . . . . . . . . 63 5 . Esamination of the Surface for Magnetic Particle

Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 6 . What Can the Magnetic Particie Method Find? . . . . . . . 65 '7 . On What Kinds of Materiais Does it Work? . . . . . . . . . 66 8 . What are the Advantages of the Method? . . . . . . . . . . 66 9 . What are the General Limitations of the Method? ... 67

10 . Comparison with Other Methods . . . . . . . . . . . . 68

CHAPTER 3 SOURCES OF DEFECTS

1 . General . . . . . . . . . . . . . . . . . . . . . . . . . 70 2 . Some Definit~ons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 3 . What is a Defect? . . . . . . . . . . . . . . . . . . . . . . . . . . 71

SECTION Page . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Magnetic Discontinuities 72 .......................... . 5 Classes of Discontinuities 75

. . . . . . . . . . . . . . . . . . . 6 Conventional Classification System 75 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Inherent Discontinuities 75

8 . Prlmary Process~ng Discontinuities . . . . . . . . . . . . . . . . . . 80 9 . Secondary Process~ng or Finishing Discontinuities . . . . . 89

10 . Service Cracks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 11 . Other Sources of Defects . . . . . . . . . . . . . . . . . . . . . . . . . . 95

CHAPTER 4

HOW AND WHY METALS FAIL 1 . General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 2 . Metal Failure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 3 . Early Attempts to Avoid Failure . . . . . . . . . . . . . . . . . . 96

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Metallurgy .. 97 5 . Strength vs . Failure in Metals . . . . . . . . . . . . . . . . . . . . . . 98 6 . How Metals Fail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 7 . Conditions Leading to Failure . . . . . . . . . . . . . . . . . . . . . . 100 8 . Overstressing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 9 . Impact . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

10 . Effect of Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 11 . Fatigue Failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 12 . Stress Raisers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 13 . Design Stress Raisers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 14 . Fatigue of Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 15 . Fatigue Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 16 . Fatigue Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 17 . Des~gning for Fatigue ............................. 106 18 . Inspecting for Fatigue Cracks . . . . . . . . . . . . . . . . . . . . . . . 107 19 . Fatigue in Torsion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 20 . Expenmental Stress Analysis . . . . . . . . . . . . . . . . . . . . . . . 110 21 . Rate of Propagation of Fatigue Cracks . . . . . . . . . . . . . 112 22 . Salvage of Parts Showing Fatigue Cracks . . . . . . . . . . . . 112 23 . Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 24 . Creep . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 25 . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113

8

CHAPTER 5

DEFINlTIONS OF SOME TERMS USED I N MAGNETIC PARTICLE TESTING

SECTION 1 . Need for Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Page 114

2 . Groups of Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 3 . Terms Relating to Magnetism . . . . . . . . . . . . . . . . . 114 4 . Terms Relating to Electricity . . . . . . . . . . . . . . . . . . . . 120 5 . Terms Relating to Electromagnetism . . . . . . . . . . . . . . . . . 124 6 . Terms Relating to Magnetic Particle Testing . . . . . . . . . . 125 ? . General Comments . . . . . . . . . . . . . . . . . . . . . . . 128

CHAPTER 6

CHARACTERISTICS OF MAGNETIC FIELDS 1 . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 2 . Magnetic Field Around a Bar Magnet . . . . . . . . . . . . . . . 131 3 . Poles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132 4 . Magnetic Attraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 5 . A Cracked Bar Magnet . . . . . . . . . . . . . . . . . . . . . . 134 6 . Effect of Flux Direction . . . . . . . . . . . . . . . . . . . . . . . . . I34 7 . Circular Magnetization . . . . . . . . . . . . . . . . . . . . . . . . 135 8 . Circular Magnetization and Cracks . . . . . . . . . . . . . . . 136 9 . Difficulties of Establishing Proper Fields . . . . . . . . . . . . . . 137

10 . Distorted Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 11 . Parallel Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 12 . Production of Suitable Fieids . . . . . . . . . . . . . . . . . . . . . . 140

CHAPTER 7

METHODS AND MEANS FOR GENERATING MAGNETIC FIELDS

1 . The Earth's Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141 2 . Magnetization with Permanent Magnets . . . . . . . . . . . . . . 141 3 . Electric Currents for Magnetization . . . . . . . . . . . . . . 143 4 . Field In and Around a Conductor . . . . . . . . . . . . . . . . 143 5 . Loop . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 6 . Solenoid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 ? . Yoices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 8 . Solenoids for Magnetization . . . . . . . . . . . . . . . . . . . . . 145

MOHAMED

Highlight

Page I SECTION Page SECTION i . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I 9 Effect of Coil Diameter 145 ! 8 Calculation of Field Distribution 172

. . . . . . . . . . . . . . . . . . . . . . 10 Effect of Coil Length 146 i . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Transformation Methods 172 . I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Analog Nethods 175 11 Circular Magnetization 147 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

I 11 Field Plotting 175

12 . Effect of Placement of Centrai Conductor . . . . . . . . . . . . . 147 . . . . . . . . . . . I 1 3 Circuiar Fields in Irreguiar-Shaped Pa r t s 148

. . . . . . . . . . . . . . . . . . . 14 Magnetization with Prod Contacts 149 CHAPTER 9 . . . . . . . . . . . . . . . 15 Effect of Type of Magnetizing Current 151 FIELD STRENGTH AND DISTRIBUTION

1 6 . Direct Current vs . Alternating Current . . . . . . . . . . . . . . . 151 I N SYMMETRICAL OBJECTS 17 . Sources of Direct Current f o r Magnetizing 1 . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178

! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Purposes 152 2 . Electro-magnetic Fields . . . . . . . . . . . . . . . . . . . . . . . 178 . . . . . . . . . . . . . . 18 . Motor Generators and Rectifiers a s Sources of D.C . . . . . 152 3 Permeability of Magnetic Materials 179

. . . . . . . . . . . . . . . . . . . . . . . . . 19 . Storage Batteries as a Source of D.C. . . . . . . . . . . . . . . . . 153 4 Material Permeability 179 . . . . . . . . . . . . . . . . . . . . . . . . 20 . D.C. f rom Rectified A.C. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 5 Effective Permeability 180

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1 Half Wave Rectified Single Phase A.C. 154 6 Initial Pernleability 180 . . . . . . . . . . . . . . . . . . . . . . . . . ‘ . . . ‘ . . . . . 22 Full Wave Rectified Single Phase A.C. 155 7 Self-Demagnetizing Effect 181 . . . . . . . . . . . . . . . .

23 . The Surge Method of Magnetization . . . . . . . . . . . . . . . . . 155 8 . Rule for Determining Ampere Turns for . . . . . . . . . . . . . . . . . . . . . . . Longitudinal Magnetizing 181 24 . Three Phase Rectified A.C. . . . . . . . . . . . . . . . . . . . . . . . . . 156 9 . Minimum Permeability Required for Magnetic

25 . Sources of Alternating Current . . . . . . . . . . . . . . . . . . . . . . 157 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Particle Testing 182 26 . Permanent Magnetization with A.C. . . . . . . . . . . . . . . . . . 157 10 . Minimum Permeability fo r Coil Magnetization . . . . . . . 183 27 . Skin Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 11 . Minimum Permeability fo r Circular iltagnetization . . . . . 183 28 . Magnetizing with Transient Currents . . . . . . . . . . . . . . . 159 12 . Currents Requlred . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 29 . Induced Current Magnetization . . . . . . . . . . . . . . . . . . . . . 160 1 3 . Effect of Shape on Fieid Direction . . . . . . . . . . . . . 184 30 . Flash Magnetization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 14 . Longitudinal ilfagnetization . . . . . . . . . . . . . . . . . . 184 31 . Suitable Field Strengths fo r Magnetic Particle 15 . Distortion of Fieid Due to Shape . . . . . . . . . . . . . . . . . . . 185

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163 16 Circular Fields 186 17 . Field Around a Conductor . . . . . . . . . . . . . . . . . . . . . . . 186

CHAPTER 8 18 . Field In and Around a Solid Non-magnetic . . . . . . . . . . . . . . . . . . . . DETERMINATION O F FIELD STRENGTH Conductor Carrying D.C. 187

AND DISTRIBUTION 19 . Field In and Around a Hollow Non-magnetic Conductor Carrying D.C. . . . . . . . . . . . . . . . 187

1 . Importance of Knowing Field Strength and 20 . The Case of a Solid Magnetic Conductor Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 Carrying Direct Current . . . . . . . . . . . . . . . . . 189

2 . Measurement of Field Inside a Pa r t . . . . . . . . . . . . . . . 165 21 . The Case of a Hollo\v Magnetic Conductor 3 . Experimental Field Measuring Techniques . . . . . . . . . . 165 Carrying Direct Current . . . . . . . . . . . . . . . . . . . 189 4 . Flux Meters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166 22 . Fieid Inside a Conductor-The General Case . . . . . . . . 190 5 . Magnetic Fieid Meters . . . . . . . . . . . . . . . . . . . . . 167 23 . The Case of a Cylinuer of Magnetic Material 6 . Magnetographs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1G9 with Direct Current F l o w ~ n g Through a Central

. . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 . Flux Shunting Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Conductor 192

10 11

SECTION Page ............... 24 . Magnetizing with Alternating Current 193

25 . The Case of a Solid Conductor Made of Magnetic . . . . . . . . . . . . Material. Carrying Alternating Current 194

26 . The Case of a Hollow Conductor Made of Magnetic Mater~al: Carrying Alternating Current . . . . . . . . . . 195

27 . Field Inside a Solid Conductor Carrying Alternating C u r r e n t T h e General Case . . . . . . . . . . . 196

CHAPTER 10 FIELD DISTRIBUTION I N LARGE OR IRREGULAR-SHAPED BODIES

1 . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 2 . The Case of a Square Bar. Circularly

Magnetized with D.C. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 3 . The Case of a Rectangular Bar. Circularly

Magnetized with D.C. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 4 . Nonuniform Cross.Sections. Circularly

Magnetized with D.C. . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 5 . The Case of an I-Shaped Cross-Section; Circularly

Magnetized with D.C. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201 6 . The Case of Prod Contacts on Large Objects . . . . . . . . . . . 203 7 . Laboratory Tests for Field Strength with Prod

Magnetization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203 8 . Correlation of the Tests with Practice . . . . . . . . . . . . . . . 205 9 . Prod Inspection U s ~ n g Half Wave Current . . . . . . . . . . . . 207

10 . Over-all Magnetization of Large Objects . . . . . . . . . . . . . 207 11 . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208

CHAPTER 11 MAGNETIC PARTICLES-

THEIR NATURE AND PROPERTIES 1 . General Description . . . . . . . . . . . . . . . . . . . . . . . . . . 2 . Effect of Size ................................ 3 . Effect of Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . Effeet of Shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 . Permeability ................... ... ...... . . . 6 . Coerclve Force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 . Hysteresis Curves . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 . Mobility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

SECTION Page 9 . Visibility and Contrast . . . . . . . . . . . . . . . . . . . . . . . . . . 221

10 . Wet Metliod Mater~als . . . . . . . . . . . . . . . . . . . . . . . . 223 11 . Pre-mixed Baths . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 12 . The Suspending Liquids . . . . . . . . . . . . . . . . . . . . . . . . . . . 224 I 3 . Available Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . 225

CHAPTER 12 BASIC VARIATIONS IN TECHNIQUE

1 . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229 2 . Llst of I7arlations in Technique . . . . . . . . . . . . . . . . . . . 229 3 . Charaeter~stics of Defects and Parts wh~ch Influence

the Proper Choice Among the Severai Varlabies . . . . . 230 4 . P r ~ m a r y Method Cho~ces . . . . . . . . . . . . . . . . . . . . . . . . . 231 5 . Choice of Type of Current-A.C. vs . D.C. . . . . . . . . . . 231 6 . Cho~ce of Type of Magnetic Particles-

. Dry vs Wet Methods . . . . . . . . . . . . . . . . . . . . . 235 7 . First Operating Decision : Residual vs .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Continuous Method 237 8 . Second Operating Dec~sion : Circular vs .

Longitudinal Magnetization . . . . . . . . . . . . . . . . . 240 9 . Third Operating Dec~sion: Amount of Current

Requ~red . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 241 10 . Fourth Operating Decis~on: Equipment . . . . . . . . . . . . 242 I1 . Other Operating Decisions . . . . . . . . . . . . . . . . . . . . . 243

CHAPTER 13 THE DRY METHOD-

MATERIALS AND TECHNIQUES 1 . History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244 2 . Advantages and Disadvantages of the Dry Method . . . 245 3 . Mater~als . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 245 4 . Steps In Applying the Dry Method . . . . . . . . . . . . . . . . . 2.16 5 . Surface Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . 246 6 . Magnetization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247 " I . Circular Magnetization . . . . . . . . . . . . . . . . . . . . . . . . 247 8 . Longitudinal Magnetization . . . . . . . . . . . . . . . . . . 249 9 . Application of the Powder . . . . . . . . . . . . . . . . . . . . . 250

10 . Inspection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253

13

CHAPTER 14 T H E WET METHOD-

MATERIALS AND TECHNIQUES

SECTION Page . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 History 255

. . . . . . . . . . . . . . . . . . . . . 2 Good Points of the Wet Method 256 . . . . . . . . . . . . . . . . . . . . . . 3 Less Favorable Characteristics 257

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Bath Constituents 257 . . . . . . . . . . . . . . . . . . . . . . . 5 Oil as a Suspending Medium 257

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Water a s a Suspensoid 259 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 The Magnetic Particles 260

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Strength of the Bath 261

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Making up the Bath 263 . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Maintenance of the Bath 265

. . . . . . . . . . . 11 Steps in the Application of the Wet Method 266 . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 . Preparation of the Surface 266

13 . The Continuous Wet Method . Magnetization and . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bath Application 266

. . . . . . . . . . . . . . . . . . . . . . . . . . . 14 The Residual Method 269 . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 5 . Cleaning After Testing 270

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 Rust Prevention 271

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 . Skin Protection 271 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 . Prepared Bath 272

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 . Pressurized Cans 272 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 . Lacquer Method 273

CHAPTER 15 FLUORESCENT MAGNETIC PARTICLES-

THEIR NATURE AND USE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 History 274

. . . . . . . . . . . . . . . . . . . . . . . . 2 Principle of Fluorescence 274 . . . . . . . . 3 Advantages of the Fluorescent Particle Method 275

. . . . . . . . . . . . 4 . Disadvantages of Fluorescent Pal-ticles 276 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Materials 276

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Strength of Bath 277

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Mixing the Bath 278 . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Maintenance of the Bath 278

. . . . . . . . . . . . . . . . . . . . . . . . 9 Steps in the Testing Process 279

. . . . . . . . . . . . . . . . . . . . . . . . 10 Examination for Indications 279 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 The Inspection Area 279

SECTION Page . . . . . . . . . . . . . . . . . . 12 Curtamed Inspection Booths 280

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Small Dark Cabinets 281 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Inspecting in the Open 281 . . . . . . . . . . . . . . . . . . . . . . . . . 1 5 Inspecting Large Parts 282

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 The Black Light 283 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Intensity Required 284

. . . . . . . . . . 18 Operating Characteristics of Black Lights 284 . . . . . . . . . . . . . . . . . . . . . . . . . . 19 . Black Light Filters 285

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 TheInspector 285 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Dark Adaptation 285

. . . . . . . . . . . . . . 22 Health Hazards of Mercury Vapor Arcs 288 23 . Avoidance of Operator Discomfort . . . . . . . . . . . . . . . . . . 288

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Eye Fatigue 288 . . . . . . . . . . . . . . . . . . . . . . . . 25 Post-Inspection Measures 289

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Skin Protection 289 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Prepared Bath 289

CI~APTER 16

BLACK LIGIIT-ITS NATURE. SOURCES AND REQUIREMENTS

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Definitions 290 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Ultrav~olet Light 290

. . . . . . . . . . . . . 3 Sunlight as a Source of Ultraviolet L ~ g h t 291 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Fluorescent Dyes 292

. . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Sources of Black Light 292 . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Filters for Black Light 292

. . . . . . . . . . . . . . . . . . . 7 Fluorescent Emission from Dyes 293 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Black Light Lamps 295

. . . . . . . . . . . . . . . . . . . 9 . Tubular Black Light Lamps 295 . . . . . . . . . . . . . . . . . . . . . . 10 Incandescent Black Lights 296

. . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Mercury Vapor Lamps 296 . . . - . . . . . . . 12 Lamp Output . . . . . . . . . . . . . . 297

13 . Commercially Available Black Lights . . . . . . . . . . . . 297 14 . Intensity Requirements of Black Light . . . . . . . . . . . 299 15 . Measurement of Black Light Intensity . . . . . . . 301

. . . . . . . . . 16 Causes of Vanations In Black Light Intensity 302 17 . Black L ~ g i ~ t Operating Characteristics . . . . . . . . . . . . 303 18 . Achieving Adequate Black L ~ c h t Intensity . . . . . . . . . 301

. . . 19 . Eyeball Fluorescence . . . . . . . . . . . . . . . 3011

CHAPTER 17 DEMAGNETIZATION

SECTION Page 1 . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306 2 . Reasons fo r Demagnetizing . . . . . . . . . . . . . . . . . . . . . . . . . 306 3 . When Demagnetization 1s not Necessary . . . . . . . . . . . . . 307 4 . Limits of Demagnetization . . . . . . . . . . . . . . . . . . . . . . . . 307 5 . Apparent Demagnetization . . . . . . . . . . . . . . . . . . . . . 311 6 . How Demagnetization 1s Accomplished . . . . . . . . . . . . 312 1 . Removing Longitudinal and Circular Fields . . . . . . . . . . . 313 . . . . . . . . . . . . . . . . . . . . . . . . 8 Demagnetizing with A.C. 314

9 . Demagnetizing with D.C. . . . . . . . . . . . . . . . . . . . . . . . . . 316 10 . Demagnetizing with Oscillating Current . . . . . . . . . . . . . 317 11 . Yoke Demagnetization . . . . . . . . . . . . . . . . 317 12 . Demagnetizing with A.C. Loops . . . . . . . . . . . . . . . . . . . 319 13 . Some Helpfui Hints for Demagnetizing . . . . . . . . . . . . . 319 14 . Checks f o r the Degree of Demagnetization . . . . . . . 320 15 . Choice of Demagnetiz~ng Methods . . . . . . . . . . . . . . . 321 16 . Demagnetizing wit11 Magnetizing Equipment . . . . . . . . 322 17 . Effect of Vibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324

CHAPTER 18 EQUIPillENT FOR MAGNETIC PARTICLE TESTING

1 . History . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 325 2 . Need for Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . 325 3 . Simple Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . 326 4 . Large Portable Equipment . . . . . . . . . . . . . . . . . . . . . . 327 5 . Stationary Magnetizing Equipment . . . . . . . . . . . . . . . 329 6 . Large, Heavy Duty D.C. Equipment . . . . . . . . . . . . . . . . . . 331 7 . Unit Variations . . . . . . . . . . . . . . . . . . . . . . . . 332 8 . Demagnetizing Equipment . . . . . . . . . . . . . . . . . . . . . 332

CHAPTER 19 AUTOMATIC AND SPECIAL MAGNETIC PARTICLE

TESTING EQUIPMENT 1 . Introduction . . _ . . . . . . . . . . . . . . . . . . . . . . 333 2 . Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333 3 . History of Spec~ai Unit Deveiopment . . . . . . . . . . . 334 4 . Factors Dictating the Need for Special Units . . . . . . . . . 338

SECTION Page 5 . Single-Purpose and General-Purpose Units . . . . . . . 339 6 . Automatic Equ~pment . . . . . . . . . . . . . . . . . . . . 342 7 . Advantages of Automatic Equipment . . . . . . . . . . 343 8 . Steps Leading to the Design of a n Automatic or

. . . . . . . . . . . . . . . . . . . . . . . . Speciai-Purpose Unit 344 . 9 . The Firs t Step Analys~s of the Problem . . . . . . . . . . . . . 344

10 . The Second Step . Consideration of Method Factors . . 345 . 11 . The Third Step Final Design Specification . . . 348

12 . Examples of Special Unit Applications . . . . . . . . . . . 349 . 1 3 Testing of Bearmg Balls and Races . . . . . . . . . . 349

14 . Billet Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353 . 15 Seam Depth Discrimination . . . . . . . . . . . . . . . . . . . 354

16 . Design Considerations, Other than Method Factors. fo r Billet Testing . . . . . . . . . . . . . . . . . . . . . 356

17 . Inspection of Large Castings . . . . . . . . . . . . . . . . . 357 18 . Inspection of Welded Steel Missile Motor Cases . . . . . 359 19 . Islultiple Test Systems . . . . . . . . . . . . . . . . . . . . 364 20 . Future Trends . . . . . . . . . . . . . . . . . . . . . . 365

CHAPTER 20 DETECTABLE DEFECTS

1 . The Magnetic Particie Testing Function . . . . . . . . . . 366 2 . Defects Classified . . . . . . . . . . . . . . . . . . . . . . . . 367 3 . Important CI~aracter~sCics of Discontinuities . . 367 4 . Surface Cracks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 369 5 . Detection of Surface Cracks . . . . . . . . . . . . . . . . . . . . . 370 6 . Discontinuities Lying Wholly Below the Surface . . . . 373 7 . Detection of Defects Lying Wholly Below

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . the Surface 373 8 . Two Groups of Sub-surface Discontinuities . . . . . . . . 373 9 . Deep-lying Defects . . . . . . . . . . . . . . . . . . . . . . . . 374

10 . Definition of Terms . . . . . . . . . . . . . . . . . . . . . . . . . 374 11 . Concept of Depth . . . . . . . . . . . . . . . . . . . . . . . . . . . 375 12 . Spread of Emergent Field . . . . . . . . . . . . . . . . . . . . . . 376 1 3 . Effect of Width . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 371 14 . Effect of I I e~gh t and Length . . . . . . . . . . . . . . . . . . . 377 15 . Effect of Shape . . . . . . . . . . . . . . . . . . . . . . . . . . 379 16 . Effect of Orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . 379 17 . Most Favorable Defect for Detection . . . . . . . . . . . . . 379

17

SECTION Page . . . . . . . . . . . . . . . . . . . 18 Effect of Afetliod of Magnetization 380

. . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Effect of Permeability 380 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 . Other Factors 380

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 . The O ~ e r a t o r 381

CHAPTER 21 NON-RELEVANT INDICATIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Definition 382

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 . False Indications 382 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 . External Poles 382

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 . AIl-over Pat terns 383 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 . Edgcof Scale 383

. . . . . . . . . . . . . . . . . . . 6 . Constriction in the Metal Path 384 . . . . . . . . . . . . . . . . . . . . . 7 . Sharp Fillets and Thread Roots 385

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 . Magnetic Writing 386 . . . . . . . . . . . . . . . . . . . . . . . 9 . External Magnetic Fields 387

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SO . Cold Wor'king 387 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 . Luders Lines 389

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 . Gram Boundaries 390 . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3 . Boundary Zoncs in Welds 390

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1 . Flow L ~ n c s 390 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 . Brazed Joints 390

16 . Joint Between Diss~milar Magnetic Materials . . . . . . . 391 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 7 . Forced Fits 391

CHAPTER 22 INTERPRETATION. EVALUATION

AND RECORDING O F RESULTS 1 . Essential Steps . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . 2 . Definitions . . . . . . . . . . 3 Tlic Problem of Interpreting

. . . . . . . . . . . . . . 4 . Outs~de I<nowiedge Required . . . . . . . . . . . . . . . . . . . . . . . . . . 5 The Operator

6 . Sources of I'inowledge and Exper~ence . . . . . . . . . . . . . . . . . . . . . . . . . . 7 . Supplemcntal Tests

. . . . . . . . . . . . . . . . . . . . . 8 . Simple Tests 9 . Binocular Rlicroscope . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . 10 . Filing

SECTION Page . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Grinding 398 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Chipping 400

. . . . . . . . . . . . . . . . . . . 1 3 Chipping for R e p a ~ r and Salvage 400 . . . . . . . . . . . . . . . . . . . 14 Destructive Methods-Fracturing 401

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Sectioning by Sawlng 403 . . . . . . . . . . . . . . . . . . . . . 16 Exam~i~a t i on of the Cut Surface 403

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Etchlng 404 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 E t c h ~ n g Cracks 405

. . . . . . . . . . . . . . . . . . . . . . . . . 19 A~icroscopic Examination 406

. . . . . . . . . . . . . . . . . . . . . . . . 20 Inspection Lighting 406

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Records 407

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Fixing a n Indication 407

. . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Lifting an Indication 407 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 Pliotography 409

25 . Black Light Photography . . . . . . . . . . . . . . . . . . . . . . . . 409 . . . . . . . . . . . . . . . . . . . . . . . . . 26 Polaroid Film Technique 411 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Color photograph)^ 412

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 . Summary 412 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 . Evaluatioli 413

30 . Tlie Probiem of Evaluation . . . . . . . . . . . . . . . . . . 413 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 . Design 414

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 . Stress-raisers 414 33 . Strength of Alaterials . . . . . . . . . . . . . . . . . . . . . . . . . . . 415

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 . Over-lnspcction 416 35 . General Evaluation Rules . . . . . . . . . . . . . . . . . . . . . . . 416 36 . Process Specifications . . . . . . . . . . . . . . . . . . . . 417

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 Summary 418

CHAPTER 23

INDUSTRIAL APP1,ICATIONS

. . . . . . . . 1 Indus t r~a l Uses of Magnetic Particle Testing 419 2 . Classification of Magnetic Particle Testing

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applicatioiis 419 . . . . . 3 Magnetic Particie Testing for Final Inspection 419

. . . . . . . . . . . . . . . . . . . . . . . 4 Receiving Inspection 421 5 . In-process Inspection . . . . . . . . . . . . . . . . 422 6 . Maintenance and Overhaul in the Transportation

Industries . . . . . . . . . . . . . . . . . . . . . . 425 7 . Plant and Machinery kla~ntenance . . . . . . . . . . . . . . . 431

SECTION Page 8 . Testing of Large and Heavy Articles and

Components . . . . . . . . . . . . . . . . . . . . . . . . . . 431 . . . . . . . . . . . . . . . . . 9 Some unusual Special Applications 435

CHAPTER 24 TESTING O F WELDMENTS.

LARGE CASTINGS AND FORGINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Introduction 440

2 . Weld Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 440 3 . Magnetiz~ng Techniques for Weld Testing . . . . . . . . . . . 442 4 . Other Nondestructive Test Methods for Weld Inspection 443 5 . Prod Magnetization and Dry Powder Tecilnlque . . . . 443 6 . Yoke Magnetization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444 " . t Type of Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444 8 . Examples of Weid Inspection with Magnetic Particle

Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 446 9 . Steel Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 450

10 . Nondestructive Testing a s a Design Tooi . . . . . . . . . . . . 451 11 . Gray Iron Castings . . . . . . . . . . . . . . . . . . . . . . . . . . . 451 12 . Forglngs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453 13 . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453

CXAPTER 25 STANDARDS AND SPECIFICATIONS FOR

MAGNETIC PARTICLE TESTING 1 . General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455 2 . Specifications for Magnetic Particle Testing . . . . . . . . . . 455 3 . Types of Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . 456 4 . Broad Procedural Gu~des . . . . . . . . . . . . . . . . . . . . . . . . . . 457 5 . Company Procedural Guides . . . . . . . . . . . . . . . . . . . . 457 6 . Product o r Industry Specifications . . . . . . . . . . . . . *. . . . . 458 7 . Process Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458 8 . Maintenance o r Overhaul Inspection . . . . . . . . . . . . . . 459 9 . Certification of Operators . . . . . . . . . . . . . . . . . . . . . . 460

10 . Standards for Acceptance or Rejection . . . . . . . . . . . . . . 460 11 . Repair Station Requirements . . . . . . . . . . . . . . . . . . . . . . 461

. 12 . Equ~pment Specifications . . . . . . . . . . . . . . . . . . . . . . . . 461 . . . . . . . . . . . . . . . . . . . . . . . .

I 13 Operating Instructions 461

SECTION . Government Specifications . . . . . . . . . . . . . . . . . . . . . . . 462 15 . Other Specifications of Interest . . . . . . . . . . . . . . . . . . . 464 16 . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 464

CHAPTER 26 TESTS FOR EVALUATION AND CONTROI. O F

EQUIPMENT AND PROCESSES 1 . General . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 466 2 . &lalfunctioning of E q u ~ p n ~ e n t . . . . . . . . . . . . . . 466 3 . Proper Magnetic Particles and Bath L~quid . . . . . . . 467 4 . Bath Concentration Incorrect . . . . . . . . . . . . . . . . 468 5 . Detachment of Fluorescent Pigment . . . . . . . . . . . . 468 6. Specifications for Suitabie Petroleum Base

Liquids fo r Oil Type Wet Bath . . . . . . . . . . . . . . . 468 7 . Settling. Test fo r Bath Strength . . . . . . . . . . . . . . . 469 8 . Other Bath Strength Tests . . . . . . . . . . . . . . . . . . . . . . 472 9 . Test f o r Black L ~ g h t Intensity . . . . . . . . . . . . . . 473

Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 475

Defects and the Strength of Materials by A . V . de Forest . . . . . . . . . . . . . . . . . . . . . . . 476

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491

INDEX . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504

LIST O F TABLES

TABLES Page

I ASTM Recommended Prod Spaclngs and Current Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (Chapter 10) 205

J I Magnetizing Force in Oersteds fo r l!arious Prod Spaclngs Uslng ASTM Recommended Currents

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . (Chapter 10) 205

111 Properties of Oils Recommended for Magnetic Particle . . . . . . . . . . . . . . . . . . Wet Method Bath (Chapter 14) 258

iV Bath Strength Chart (Chapter 14) . 263

V Bath Strength CharGFluorescen t Particles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . fChapte r15) 277

VI C u r ~ e P o ~ n t for Some Ferromagnetic Materials (Chapter 17) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310

. . . . . VII G u ~ d e t o Demagnetizing Methods (Chapter 17) 323

VIII Standardization Guide for Settling Test (Chapter 26) . 471

LIST OF ILLUSTR.4TIONS

Front~sp~ecf-Test~ng of Steel B~l le t s fo r Seams, Usmg Fluorescent htagnetic Par t~cles . The Process Discr lnl~r~ates Ee- t\\-een Shallo\\, and Deep Seams. (Courtesg Youngs- toicn Sheef and Tube Co?z~pang.)

CHAPTER 1 Page

Fig.

Fig. Fig. Fig.

Fig. Fig. Fig.

Fig.

Fig.

Fig.

1. Dr. H. H. Lester's Pioneer X-Ray Laboratory . . . . . . . . . . . . . . a t Watertown Arsenal. 1922 49

. . . . . . . . . . . . . . . . . 2. Professor A. V, de Forest 50 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. F . B . D o a n e . . 51

4. Early Esperlntental Equ~pment Used by . . . . . . . . . . . . . . . . . . F B. Doane. 1930 52

5. "Tree of Growth" of Nondestructive Testing . . . 53 6. Early A. C. kfagnet iz~ng Assembly. 1933 . . . . . 54 7. Storage Battery Unit Used in the Aircraft

. . . . . . . . . . . . . . . . . . . . Industry. 1932-1940 55 8. Automatic Unit for Testing Armor-Pierclng

. . . . . . . . . . . . . . . . . . . . . . . . Projectiles. 1943 56 9. Automatic Unit, Inciuding P a r t Rotation, for Testing

Automotive Connecting Rods, Uslng Fluorescent . . . . . . . . . . . . . . . . . . . . Magnetic Particles. 1948 57

10. Automatic Unit fo r Testing Steei Billets fo r Seams. 1956. (Coz~rtesy A?nerzca.n Steei and Wire

. . . . . . . Divlsion, U.S. Steel Corporation) 58

CHAPTER 2 Page

Fig. 11. Typ~ca l Dry Powder Pattern. This is the Orlginai Demonstration Piece Used By A. V. de Forest,

. . . . . . . . . . . . . . . . . and Iater by F. B. Doane 60 Fig. 12. Field Distortion a t a Discontinuity Lying

. . . . . . . . . . . . . . Wholly Below the Surface 61 Fig. 13. Field Distortion a t a Discontinuity Which

. . . . . . . . . . . . . . . . . . IS Open to the Surface 62 . . . . . . . . . . Fig. 14. Field Distortion a t a Surface Scratch 63

Fig. 15. Typ~ca l Discontinuity Pattern a s Indicated by . . . . . . the \Vet Method (Seamy Wrlst Pin) 63

Fig. Fig.

Fig.

Fig.

Fig. Fig. Fig.

Fig.

Fig. Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

Fig.

CHAPTER 3 Page

16. Typ~ca l Magnetic Particle Indication of Craclis . . 72 17. Magnetic Particle Indication of a Forced F i t

a ) White Light View . . . . . . . . . b) Fluorescent Particle Indication 73

18. Magnetic Particle Indication a t the Weld . . . . . . Between a Soft and a Hard Steel Rod 73

19. Magnetic Particle Indication of the Braze Llne of a Brazed Tool Bit . . . . . . . . . . . . . . . 74

20. hlagnetic Particle Indications of Segregations. . . . 74

21. Cross Section of Ingot Showing Shrink Cavity . . . . 76 22. Magnetic Particle Indication of a Sub-surface

. . . . . . . . . . Stringer of Non-Metallic Inciusions 78

23. Scabs on the Surface of a Rolled Bloom. . . . . . . . . (Courtesy of U. S. Steel Corporation) 79

24. Seam on a Bar Sholvn by Magnetic Particles . . . 80 25. Surface of a Steel Billet Showing a Lap.

(Courtesy of U. S. Steel Corporation) . . . . . . 81

26. How Laps and Seams a r e Produced by the Rolls. Over-fills and Under-fills . . . . . . . . . . . . . . . 82

27. Magnetic Particle Indications of Laminations Shown on Flame-Cut Edge of Thiclt Steel Plate. . 83

28. a ) Fluorescent Magnetic Particle Indications of Severe Cupping in Drawn Spring Stock, Ground Into the Ruptures a t One End

b) Section Through Severe Cupping In a . . . . . . . . . . . . . . . . . . . . . . . 1% Inch Bar 84

29. Magnetic Particle Indications of Cooling Craciis In an Alloy Steel Bar a ) Surface Indications b) Cross-Sections Showing Depths . . . . . . . 8.1

30. Magnetic Particle Indications of Flakes in the Bore of a Large Hollow Shaft . . . . . . . . . . . 85

31. Magnetic Particle Indications of Forging Cracks , o r Bursts in an Upset Section. Severe Case. . . . 86

32. Cross-Section of a F o r g ~ n g Lap. Magnified 100 X . . 86

Fig.

Fig.

Fig.

Fig.

Fig.

Fig. Fig.

Fig.

Page 33. Magnetic Particle Indication of Flash Line Tear

in an Automotive Sprndle Forging. Partially Machined . . . . . . . . . . . . . . . . 87

34. Magnetic Particle Indications of Defects in Castings a ) Surface Indication of an Internal S h r ~ n k

Cavity h) Handling Crack In Gray Iron Casting. . . . 88

35. Magnetic Particle Indications of Quench~ng Cracks, Shown with Dry Powder . . . . . . . . . . . 90

36. Fluorescent Magnetic Particle Indications of Typical Grinding Cracks . . . . . . . . . . . . . . . . 91

37. Magnelic Particie Indications of Grinding Cracks In a Stress-Sensitive, Hardened Surface . . . 92

38. Magnetic Pai-ticle Indications of Plating Cracks. . 92

39. Magnetic Particle Indication of a Typical Fatigue Crack . . . . . . . . . . . . . . . . . . . . 93

40. Fluorescent Magnetic Particle Indications of Cracks in Cranltshaft of Small A ~ r c r a f t Englne, Damaged in Plane Acc~dent . . . . . . . . . 94

CHAPTER 4 Fig. 41. Examples of Silky and Crystalline Fractures

a ) Ductile or Silky b) Brittle or Cleavage . . . . . . . . . . . . . . 99

Fig. 42. Brittle Fracture Which Occurred in a Tank Holding Gas a t Low Temperature . . . . . . . 101

Fig. 43. Face of a Typical Fatigue Crack. (Courtesy Sidney IVallcer, ARO E?~gz?teer?.?~g Conzpany) . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Fig. 44. Stress Concentration a t a Surface Notch or Crack . . . . . . . . . . . . . . . . . . . . . . . . . . 103

Fig. 45. Stress Concentration a t a Hole in a Surface Stressed In Tension, Shown by the Brittle Coating Method . . . . . . . . . . . . . . . 104

Fig. 46. Stress Concentration Pattern on a Ribbed Casting, Shown by the Brittle Coating Method. 104

Page

Fig. 47. The R. R. Moore Rotating Beam Fatigue Testing Machine. (Conrtesy Wiede??la?tn Divisio?~,

. . . . . . . . . . . Warner and Smase!~ Contpa71y) 105

Fig. 48. Fatigue Cracks Starting Out of a Sharp Fillet . . . . . . . . . . on a Racing Stocli Car Safety Huh 108

Fig. 49. Fracture Througll a Small Fatigue Crack Showing Circular Shape of the Smooth Portion. (Courtes!~ Sidney H'alker, ARO Engtneert?~y Conzpany) . 108

Fig. 50. Fatigue Cracks a t 45' to the Axis of the Wire . . . . . . . . . . . . . . . . . . . . . . on a Helical Spring 109

. . . . . . . . . . . . Fig. 51. Brittle Coating Stress Analysis Kit 110

Fig. 52. Stress Pattern Sliown in Small Ring Made of a Photo-elastic Resin. The Ring 1s Stressed in Com- pression and Photograplled with Transmitted Polarized Light. (Courtesy of Professor Wnl. H.

. . . . . . . . . . . . . . . . . . . . . . . . ~ a y I I T ) 111

Fig. 53. Progressive Improvement of Casting Design Resulting from Use of Brittle Coating Stress

. . . . . . . . . . . . . . . . . . . . . . . . . . . Analysis 111

Fig. 64. Corrosion Fatigue Cracks on an Oil Well . . . . . . . . . . . . . . . . . . . . . . Suclcer Rod 112

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fig. 55. The Hall Effcct 115 . . . . . . . . . . . . . . . . . . . . . . . Fig. 56. Hysteresis Curve 116

. . . . Fig. 57. Magnetograph of the Field Around a Magnet 118

Fig. 58. Wave Form of Full-Wave Rectified Single . . . . . . . . . . . . . . . . . . . . . . . Phase A. C. 121

. . . . . . Fig. 69. Wave Form of Three Phase Rectified A.C. 121

Fig. 60. Wave Form of Half Wave Rectified Singie . . . . . . . . . . . . . . . . . . . . . . . . . . . Phase A. C. 122

Fig. 61. Wave Form of Single Phase Alternating Current 1 123

Fig 62. Set of Wave Forms of Three Phase Alternating . . . . . . . . . . . . . . . . . . . . . . . . . Currents 123

CHAPTER 6 Page . . . . . . . . . . . . . . . . . . . Fig. 63. Fieid Around a Bar Magnet 131

. . . . . . . . . . . . . Fig. 64. Consequent Poles on a Bar Magnet 132 Fig. 65. Bridging of the Air Gap a t a Crack

a) Leakage Field . . . . . . . . . . . b) Magnetic Particle Indication 133

Fig. 66. Effect of Crack-Orientation in a Longitudinally . . . . . . . . . . . . . . . . . . . . . . . . . Magnetized Bar 135

Fig. 67. Circular Magnetization a ) Incomplete Ring with a Small Air Gap b) Complete Ring with No Air Gap, but with a

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Crack 136

Fig. 68. Effect of Crack Orientation in a Circularly . . . . . . . . . . . . . . . . . . . . . . . . . . Magnetized Bar 136

Fig. 69. Magnetic Flux Passing Diametrically Through a Disc, Instead of from Face to Face, when the Disc is Placed a t an Angle of hrearly 90' to a

. . . . . . . . . . . . . . . . . Longitudinal Field in a Coil 138

Fig. 70. Distortion of the Circular Field in a Bar Caused by a ) Contact with a Piece of Soft Iron

. . . . . . . . . . . . . . . . . . . b) The Shape of the Part 139 . . Fig. 71. Field Produced in a Bar by a "Parallel" Current. 139

CHAPTER 7 Fig. 72. Magnetization with a Permanent Magnet

a) Magnetizing a Bar by Placing One Pole of a Permanent Magnet a t One End

b) Magnetizing a Plate by Setting a Permanent Bar Magnet on End on the Plate

c) A Permanent Horseshoe Magnet, or Yoke, Placed on the Surface of a Plate or Other Part 142

Fig. 73. Field Around a Conductor Carrying Direct Current, and Its Direction . . . . . . . . . . . . . . . . . . . 144

Fig. 74. Field in and Around a Loop Carrying Direct . . . . . . . . . . . . . . . . . . Current, Showing Polarity 144

Fig. 75. Field in and Around a Solenoid Carrying Direct . . . . . . . . . . . . . . . . . Current, and Its Direction 145

Page Fig. 76. Flux Pa th for a Long Ba r in a Short Coil . . . . . . . 146

Fig. 77. Magnetographs of Field Showing a Cross Section of a Tube Magnetized a ) With a Centrai Conductor Centrally

Located, Carrying Direct Current b) With the Conductor Located Adjacent

to the Wall of the Tube . . . . . . . . . . . . . . . . . . 148

Fig. 78. P a r t Being Magnetized Circularly by Clamping I t Between the Head Contacts of a Magnetiz~ng Unit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

Fig. 79. iitagnetograph of the Field Around and Between Prod Contacts . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

Fig. 80. Effect of a Short Surge of High Current Followed by a Drop to a Steady Current of Lower Value 156

Fig. 81. Sine Wave Form of Single Phase Alternating Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158

Fig. 82. Oscillograph of the Voltage Across the Switch Polnts When a Single Phase A.C. Circuit 1s Broken, Showlng the Arc Quenched a t the Point of Zero Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . 158

Fig. 83. Residual Field Inside a Ba r Generated by a ) Slow Break b) "Quick Break" Translent Current . . . . . . . 160

Fig. 84. Direct Contact Method of Xtagnetiz~ng Ring- Shaped Par t s to Locate Circumferential Defects. 161

Fig. 85. Induced Current Method of Magnetizing Ring- Shaped Par t s to Locate Circumferentiai Defects. 161

CHAPTER 8 Fig. 86. Flux Meter Arranged to iileasure the Flux in a

Bearing Race . . . . . . . . . . . . . . . . . . . . . . . 168 Fig. 87. Diagram of a Soft Iron Vane Type of Magnetic

Field Meter . . . . . . . . . . . . . . . . . . . . . . . . . . 170 Fig: 88. Magnetograph of the Field Around an Iron Piece

Placed Adjacent to a Conductor Carrylng Direct ' Current . . . . . . . . . . . . . . . . . . . . . . . 171

Page Fig. 89. Calculated Field Distribution in and Around

a Magnetic Sphere and an Elipsord in a Uniform Magnetic Field . . . . . . . . . . . . . . . . . . . . . . . . . 173

Fig. 90. a ) Calculated Field In and Around a Ferromagnetic Tube of Moderate Eccentr~city, Carrying Direct Current

b) Calculated Field in and Around a Ferromagnetic Tube of Extreme Eccentrrcity, Carrying Direct Current . . . . . . . . . . . . . . . . 174

Fig. 91. Example of Conformal Mapplng Teclinlque for Analyzing Field Distribution . . . . . . . . . 175

Fig. 92. Sketches of Fields for

a ) An Iron Bar Having a Change of Cross Section

b ) A Rectangular Bar Having a Notch (Sketch only Partially Complete) . . , 176

CHAPTER 9

Fig. 93. a ) Magnetization Curve for a Stecl Having Relatively Low Material Permeability

b ) Plot of the Material Permeability of the Steel Relative to 13 . . . . 180

Fig. 94. Diagram of the Field Around a Bar Magnet . . . 184

Fig. 95. Behavior of a Longitudinal Field in the Upset Portion of a Magnetized Bar . . . . . . 185

Fig. 96. Distribution of the Field in and Around a Solid Non-magnetic Conductor Carrying Direct Current . . . . . . . . . . . . . . . I87

Fig. 97. Distribution of the Field in and Around a Hollow Nonmagnetic Conductor Carrying Direct Current . . . . . . . . . . . . 188

Fig. 98. Distrihution of the Field in and Around a Solid Conductor of Magnetic Materral Carrying

. . . . . . . . . . . . . . . Direct Current 189

Page

Fig. 99. ~ i i t r i b u t i o n of the Field in and Around a Hollow Conductor of Magnetic Material Carrying

. . . . . . . . . . . . . . . . . . . . . . . . . . Direct Current 190

Fig. 100. Graph Showing the Var~a t ion of the Field Inside . . . . . Any Conductor Carrying Direct Current 191

Fig. 101. Distribution of the Fieid in and Around a Hollow Magnetic Cylinder with Direct Current Flomrng

. . . . . . . . . . . Through a Central Conductor 192

Fig. 102. Distribution of the Fieid in and Around a Solid Conductor of Magnetic Material Carrying

. . . . . . . . . . . . . . . . . . . . . . . . Alternating Current 193

Fig. 103. Distribution of the Field in and Around a Hollow Conductor of Magnetic 3latertal Carryrng

. . . . . . . . . . . . . . . . . . . . . Alternating Current 194

Fig. 104. Graph Showing the Effect of Conductivity, Permeability and Frequency on the S k ~ n

. . . . . . . . Effect of Alternating Current 195

Fig. 105. Graph Illustrating the Skin Effect of . . . . . . . . . . . . . . Alternating Current 196

Fig. 106. Fieid Distribution in a Square Bar, Circularly . . . . . . . . . . . Magnetized with Direct Current 199

Fig. 107. Field Distribution in a Rectangular Bar, Circularly . . . . . . . . . . . . Magnetized with Direct Current 200

Fig. 108. Fieid Distribution in a n I-Shaped Ear , Circularly . . . . . . . . . . . Magnetized with Direct Current 201

Fig. 109. Field Strength. TI, Plotted Against Prod S p a c ~ n g for Varlous Direct Current Values, on One

. . . . . . . . . . . . . . . . . . . . . . . Inch Plate 204

Fig. 110. Comparison of the Effectiveness of Direct Current and Half Wave Current when Used with Prod Magnetization . . . . . . . . . . . . . . . . . 204

CIIAI'TER 11 Page Fig. 111. I-Iystercsrs Curves for Typ~ca l Dry Powders.

a ) When I!,,,,, was 500 Oerstetls . . . . . b) \Vhen I1 ,,,;,, 1v:rs 50 Oe~s t eds 218

Fig. 112. I Iystcres~s Ct11.vcs 101' \Vel 3Iethod Materials. a ) \Vhen H,,,.,, \\,as 500 Oersteds I,) When H ,,,,,, was 50 Ocrsteds . . 2 l 9

CIIAPTEL~ 12

F i g 113. Comparison of I ~ ~ d j c a t i o l ~ s of Surface Cracks on ;I P a r t klagnctizetl with AC, S t r a ~ g h t DC and 3 Phase Rectified AC Respectiveiy . . 232

Fig. 114. Dranr~ng of the Tool S t ~ l Ring Specrlnen with Sub-Surface Defects . . . . 333

Fig. 115. Comparison of Sens~t ivi ty of AC, DC, DC with Surge, and Half \Va\~e, for Locatrng Defects

. . . . \\!liolly Below the Surface . . 234 Fig. 116. Cornparlson of the Dry ancl the \Vet DC hlethod

for Defects Lyrng Ecio~v tile Surface . . . . . . 236 Fig. 117. Cornpat-rson of the D I . ~ ant1 the Wet h,Iethoils for

Defects Lylng Belo\\, the Surface. when AC and DC are used for Magnetizatiorl; . . . . 237

Fig. 118. EKect of the Hardness of a Specimen on tile Current Required to Locate Defects Lying Below the Surface . . . . . . . . . . . . . . . . . . 238

Fig. 119. BIapnetic leech Contacts , . 248 Fig. 120. Pre-formed Split Coils of I.'lesible Cable. 249 Fig. 121. a ) Squeeze-Bottle -4pglicator fo r Dry Powder

b) Air Operated Powder Gun . . 251 Fig. 122. Inspection of n Weld, Using One Leech Contact

and One Prod, with a Po\vder Gun . . 252

CIIAPTEIL 14

Fig. 123. Making tile Settling Test fo r \\let Bath Concentration 262

Page

Fig. 124. Mixing the Bath Using Dry Concentrate . . . . . 264 Fig. 125. Simpie Wet Method Unit Setup for Circular

Magnetizing . - . . . . . . . . . . . . . . . . . . . . . . . . , . 268

Fig. 126. Wet Method Unit Setup for Longitudinal Magnetizing . . . . . . . . . . . . . . . . . . . . . . . . . 269

CHAPTER 15

Fig. 127. Simple Inspection Unit for Fluorescent Magnetic Particle Testing . . . . . . . t . . . . . . . . . . . . 280

Fig. 128. Tabie-Top Inspection Cabinet \vith Built-in Black Light . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

Fig. 129. Effect of Intensity of Black Light on the Brightness of Indications. Same Indication is Shown Under Weak (30 Foot Candles) and Intense (120 Foot Candles) Black Light . . . . . . . . . . . . . . . . . 283

Fig. 130. Chart of Perception of the Human Eye . . . . . . . 286

Fig. 131. Chart of Color Response of the Human Eye . . 287

Fig. 132. Spectrum of Light Through the Visibie, Blaclc Light and Ultraviolet Wave Lengths . . . . .

Fig. 133. Spectrum of the Output of the High Pressure Mercury Arc . . . . . . . . . . . . . . . . . . . . . . .

Fig. 134. Light Transmission Curve of Black Llght Filter Glass (I<opp 41) . . . . . . . . . . . ~ .

Fig. 135. Ligllt Emisston Spectrum of Yellow-Green Fluorescent Dye . . . . . . . . . . . . . . . . . . .

Fig. 136. Construction Drawing of the High Pressure Mercury Arc Lamp . . . . . . , . . . . . . . . . ~ . .

Fig. 137. The 100-Watt Black Light Mounted on a Fixture on a Magnetizing Unit . . . . . . . . . . . .

Fig. 188. The 400-Watt Black Light . . . . . . . . . . . Fig. 139. Distribution of Black Light from 100 Watt and

from 400 Watt Black Light Lamps, Heid 15 Inches from the Worlc Table. . . . . . . . .

Page Fig. 140. Transmission Curve for Black Light Filter Glass

(Iiopp 41) and Response Curve of the Photo- Voltaic Cell . . . . . . . . . . . . . . . . . . . . . , 301

Fig. 141. Output Variations with Varying Voltages. 100 Watt "Spot" Black Light. . . . . . . . . . , . , , 303

CHAPTER 17 Fig. 142. Effect of Temperature on the Magnetic

Properties of Iron . . . . . . . . . . . . . . . . . , . . 309 Fig. 143. Flux Curve During Demagnetization, Projected

from the Hysteresis Loop . . . . . . . . . . . . . . , 312 Fig. 144. Demagnetization Units fo r Operation on

60 Cycle A.C. a ) For Intermittent Operation b) Feed-Through Type for Continuous

Operation . . . . . . . . . . . . . . . . . . . 315 Fig. 145. Magnetizing Unit with Automatic Current

Reduction for Demagnetizing, Using a 30 Point Step-Down Sxvitch . . . . . . . , . , , . . . 316

Fig. 146. Yoke for Demagnetization with A.C.. . . . . . . . . . 318

CHAPTER 18 Fig. 147. Magnetizing Yoke and Coil in a Portable Kit . Fig. 148. Small Magnetizing Kit, Operating from 120

Volts A.C.; Mounted on a Carry-All True$. Fig. 149. A Portable Magnetizing Unit Being Used

for Inspecting Truck Components . . . . . . Fig. 150. A Modern Portable Slagnetizing Unit , . . Fig. 1 5 . A Modern Bench-Type Magnetizing Unit . . . . Fig. 152. The Largest Bench-Type Magnetizing Unit,

for Large Diesel Cranlrshafts . . . . . . , , . .

Fig. 153. 20,000 Ampere D.C. Power Unit for Overall Multi-Directional Magnetization , .

CHAPTER 19 Fig. 154. Automatic Unit for Testing Bolts and Similarly

Shaped Par t s . . . . . . . . . , . . . . . Fig. 155. Spectal-Purpose Unit fo r Testing Steel

Propeller Blades . . . . , . . . , . . . .

Page .. Fig. 156. Special-Purpose Unit f o r Railroad Car Ase ls . . 336

Fig. 157. Magnetic Particle Equipme~it for Diesel . . . . . . . . . . . . . . . . . . . . . . Crankshaft Testing 337

Fig. 158. Automatic Unit fo r Testing Oil Field Pipe - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~ o u p l ~ n g s 337 Fig. 159. Special Unit for Crankshaft Inspection, of the

. . . . . . . . . . . . . . Fixtured Standard Type 339 . . . Fig. 160. Crankshaft Testing Unit, Specrai Throughout. 340

Fig. 161. Single Purpose Unit for Testing B e a r ~ n g Rollers. . 341 Fig. 162. General-Purpose Unit fo r Testing Automotive

. . . . . . . . . . . . . . . . . . . . . . . S t e e r ~ n g Pa r t s 342 Fig. 163. Model Showing The X, Y and Z Axes

. . . . . . . . . . . . . . . . . . . . of a Bearing Ball 350 Fig. 164. Special-Purpose Unit fo r the Inspection

of Bcanng Balls. Inset: Close-up Showing the . . . . . . . . . . . . . . . Three Testing Stat io~is 351

Fig. 165. Field and Current Distribution in a Bearing . . . . . Race Being Magnetized with a Head Shot 352

Fig. 166. Current and Field Distribution in a Bearing Race Being kIagnetized by the Induced

. . . . . . . . . . . . . . . . . . . . . Current Method 352 Fig. 167. Special-Purpose Unit f o r the Inspection of

B e a r ~ n g Races by the Central Conductor and . . . . . . . . . . . . . . . t h e Induced Current Method 353

Fig. 168. Fluorescent %Iagnetic Particie Indications of Seams of Varlous Depths, Shown by the

. . . . . . Res~duai and the Continuous Xlethods 355 Fig. 169. A Typzcai Large Billet Testing IJnit

Inset: Close-up of Inspection Station Sliowrng Billet 011 Cham Sling Billet Tu r~ i e r . (Coz~rtesy Yoz~ngstozun Sheet a?ld Tube Conzpa?i.1~) . . . . . . . . . . . . . . . . . 357

Fig. 170. Ileavy Duty Equ~pmen t f o r Applylng the Overall . . . . Magnetization Method to a Large Casting 358

Fig. 171. The Overall System Applied to a ilIiscella~~eous Assortment of Castings . . . . . . . . . . . . . . . . . 359

Fig. 172. Schematic D r a w ~ n g of Conductor Arrangement for Circtllar Magnetization of Cylindrical

. . . . . . . . . . . . . . hIissile Motor Cases 361

34

Page Fig. 173. Schematic Draxvlng of Arrangement fo r

Longitudinal 3lagnetiz:ltion of Cylintlrical . . . . . . . . . . . . . . hlissile Motor Cases 362

Fig. 174. lnduccd Current Fis ture for Missile . . . . . . . . . Motor Case Parts . . . 363

Fig. 175. Charac te r~s t~cs of Surface Defects 365

Fig. 176. Characteristics of Defects Lying Wholly Eelox' the Surface 369 . . . . . . . .

Fig. 177. Fluorescent Magnetic Particle 1ndic:ktion . . . . . . . . . . . . . . of a Forging Lap 372

Fig. 178. Concept of Depth of a Deep-Lying Defect . . 375

Fig. 179. "Spread" of the Emergent Field nt a Defect. , 376

Fig. 180. Effect of Interruption to the Flow of Water ( o r iIi;ignet~c Flux) Due to Shape and Orientation . . . . . . . . 378

Fig. 181. Effect of Shape and Or~e~ i t a t i on on Detectability 378

Fig. 152. Magnetograph Showing the Esterliar Fieid Pattern Produced \%'hen Prorls are Used to

. . . Magnetize Steel P1:ite . . . . 383

Fig. 183. Gear and Shaft Sl ios:~ng Nonrdevant Indications Due to Internal Splines . . . . . . . 381

Fig. 184. Fluorcscent 11idic;ltions of a Craclc a t the Root of a Eolt Thread . . . . . . . . . . 355

Fig. 185. Exaniplcs of Blagnetic \\'ritrng . . 386

Fig. 186. Roller Bearing Assembly \\!it11 Nonrclevant hfagnetlc Particle Patterns Procluced hy n Magnetic Chuck . . . . . . . 388

Fig. 187. Magnetic Particle Pattern Pro<ruced by Cold \Vorku~g Due to Machin~ng . . 389

CHAPTER 22 Page

Fig. 188. File Cut Applied to Cracks in a Ba r with iifagnetic Particles Re-applied . . . . . . . . . . . 398

Fig. 189. a ) Small Forgtng, Ground a t a Lap with Defect Not Completely Removed

b) Expioring Crack Indications on Line Pipe with Small Hand Grtnder . . . . . . . . . . . 399

Fig. 190. Fracture Through a Fatigue Crack That has been Blued by Hcating

a ) The Ear , Showing the Fatigue Crack b) The Fracture ................ 402

Fig. 191. Deep-Etched Section of an Automotive Steering Arm Showing Forglng Fold and Flow Ltnes In a High Stress Area . . . . . . . . . . . . . . . 404

Fig. 192. Etching Cracks on a Twist Drill . . . . . . . . . . 405

Fig. 193. Diagram for Setup for Blacli Light Photography. . 410

Fig. 194. Semt-Automatic Unit for the Final Inspection of Automotive Steering I<nuckles . . . . . . . . . 420

Fig. 195. Special Automatic Unit fo r Testing Rod Ends . . . . 423 Fig. 196. Semi-Automatic Unit for the Inspection of Small

Castings and F o r g ~ n g s . . . . . . . . . . 423

Fig. 197. Fluorescent Illagnctic Particle Indications of Grtncling Cracks . . . . . . . . . . . . . . 424

Fig. 198. Quenchung of Conveyor Wheels; Follo\ved By Magnetic Particle Testing for Quench~ng Cracks . . . . . . . . . . . . . . . . . . . . . . . . . 425

Fig. 199. Magnetic Particle Testing of Seamless Tubtng on an Automatic Unit

a ) Loading Side b ) Inspection Station c) Fluorescent Magnetic Particle

Indications of Spiral Seams . . . 426 Fig. 200. Sixteen Foot Long Magnetizing Unit for Diesel

Crankshaft, Installed ~n Overhaul Shop . . . . . 427

Fig. 201. Typical Fatigue Cracks in Gears . . . . . . . .

Fig. 202. Special Magnetizing Unit for Jet Engine Compressor Discs, Employtng the Induced Current Method . . . . . . . . . . .

Fig. 203. Spec~al Magnetiz~ng Unit fo r Aircraft Gas Turbine Compressor Blading-Assernble(~.

Fig. 204. Testing Landing Gear Components of a Helicopter on the Field, U s ~ n g a Special Portabie Unit. (Cozcvtes?, Jaeh-sonville Naval

. . . . . . . . . . . . ,417. Stntzon)

Fig. 205. Trucli Engine Connecting Rods and Caps with Fatigue Cracks . . . . . . . . . . .

Fig. 206. Testing Trucli Front Wheel Spindle with Portable Coil Unit , . . . . . . .

Fig. 207. Testing the Lifting Fingers of a Forli Truck. Inset: Close-Up of Fatigue Crack Indie a t ' 1011.

Fig. 208. Prepnratlons for T e s t ~ n g Steatn Turbine Spindle and Blading for Craclcs

Fig. 209. Spectal Unit Destgned for the Inspectton of Solla Fuel Miss~le M o t o ~ Cases

Fig. 210. Spectal U n ~ t for Inspectton of Do~nes and Nozzles

Fig. 211. Thrust Units of a L iqu~d Fuel Rocliet Eng~ne , Mounted on the Test-Fir~ng Stand. (Coz~tes?/ Rocketdwte DU1s10?%, Novth Ame~zcan Avzntion Covm~~n t io?~ ) . . . . . . . .

Fig. 212. Magnettc Particle Testing of Butt Welds on a Chamber of a L~quid Fuei Rocket Engine, Us~i ig Permanent Magnet Yoke. (Coz17.tesy Eocket(1yne Divrszo?~, h'ovth Ante?-zcati. Avratio?~ Cornovation) . . . . .

Fig. 213. The "500" Race a t Indianapolis Speedway, S ta r t of 1965 Race . . . . . . . . .

Fig. 214. Close-Uo of Inspection of 24 Inch Line Pipe for Three-Mile Miss~ssippi River Crossing

Page 427

428

429

429

430

430

432

432

433

434

434

435

436

437

Page Fig. 215. Testing of I'rotectivc Steei Collars for the

First Trans-Pacific Telephcmc Cable. Collars a r e BIagi~etizcrl on a Large Diameter

. . . . . . . . . . . . . Central Conductor

CHAPTER 24

. . . . . . . Fig. 216. Nomenclature of a Fillet Well-l

Fig. 217. Typical \Veld Defects . . . . . . . . . .

Fig. 218. Field Produced by a Yl?l<e Across a Tree Butt \ilcld . . . . . . . . . . . . . . . .

Fig. 219. Indication oE Non-relevant Open Root in Fillet . . . . Weld. lCourtcsy .lames W'. Otrlens)

Fig. 220. iifagiletic Particle Indications of Transverse Craclis in Mil 260 Wold, Sho\\w with YoAe. (Cozt~.t.eu?l N e . ~ ~ p o ? ' t i\'ezos Slc~~rlrz~.iirli?z~/ coiil7l~il!/) . . . .

Fig. 221. I n s i ~ e c t ~ n g \\?elds \!'it11 Magnetic Particles on tlic 32 Story hlicliigan Conso1id:itcil Gas Compaiiy Fuilrling in Detroit, hlichignn .

Fig. 222. Testing \Veld at the Base of a 285 Foot Television Tolver on tile Roof of One of the T\\-\.rn Towers of &Ial.ina City In Chicago

Fig. 223. Inspection a t the Wcld Stand of Resistance- Welded Pipe . . . . . . . . . . . . .

Fig. 22.1. Testing a \Veltled Oil Stor:ige Tank . . . .

Fig. 225. Some Typical FIa~idling Oacil-s i i ~ Gray Iron Castings . . . . . . . . . . . .

Fig. 226. Magnetic Particle Test of 18 Inch i\i;rr~nc Tail Shaft, Slio\ving Pntigue Cracks in the IZey\\,ay

Fig. 227. a ) Test Flock for Measuring Bath Strength

b) Test Elocl< 111 Use Bet\veeil H c a k of Unit c ) Test Block 111 Use in Coil . . . . . .

P R E F A C E

\VIien the first edition of the book "Principies of Magnafiux" was publislied in 1939, tlie magnetic particie method was still relatively unknown in industry. Certainly i t was unappreciated. This was ten years a f t e r tlie inception of t h e method in 1929. Those were ten years of intensive work on the par t of A.V. de Forest a n d F. R. Doane-and later of the staff of Magnaflux Corporation-to educate potential users in tlie value and usefulness of th is testing method. To us, of Magnaflux Corporation, tlie need fo r the method appeared self-evident, but the natural resistance of human nature to t ry ing out new ideas or any departure f rom time-honored customs, made potential users reluctant to adopt th is novei testing proredurc, and progress was siow indeed. A t times our efforts took on the aspect of a parent patiently forclng a child to take unpleasant medicine, bc- cause we ( the parents) knew i t would be good for the recipients. The book. "Pi~iiiciples of illagnaiiux", was writ ten by F. B. Doane In 1939 with the purpose to make available the knowledge and experience in the method accumulated up to that time, to aid in th is educational program.

Re-reading this first edition now-aimost thlrty years later-one is ,mpuessed by t\vo points: first, how little the fuildamcntal principles and gu~delines of the method have changed; and second, how great has been tile advance in the techniques of applying these puinciples and how \videly t h e method h a s come to oe used in a reas not even visualized in 1939.

\Vhen the second revis~on of the book was issued in 1918 the growtll in technique variations and expansion of fieids of use which had occuri,ed duriug World W a r I1 were so g rea t t h a t tlic book was almost doubled in size and contamed a great deal of new mat ter relating to new materials, new knowledge and new understanding of the conditions !\chieh controlled performance, and many details of new applications.

Now, a f t e r eighteen more years, the 1948 version of the book is

compietely out-dated, and the author feels keen regret that th is

J 1

I'REFACE

present &fort could not have been presented five or more years earlier. As a compensating factor, ho\rever, much has occurred in those five years-such a s the steel billet t e s t ~ n g developn?erit, and applications in tlie inspect~on of rocliets and aero-space rehicles- \vliich couid not liave been included m a 1958 revision.

The present new voliime, "Principles of 3Iagiietic Particle Test- Ing", is completely remritten. The format and organization follo\\,s tliat of i ts companion volume, "Principles of Penetrants". A considerable amount of nelv material, nluch of i t heretofore ulipu13- lished, has been incorporated. The earlier books \\,ere orienteil toward the e ra of steani ioconiotives and piston aircraf t engines- two uses wh~ch , major a t the time, have passed out of the light of first importance in industrial uses in the United States ior are about to ) . Instead, thls volume 1s oriented to\rard diesel iocomotires, automotive and rolling mill applicatioiis, and space age testing problems.

There is much new detail in the information given regarding ma te r~a l s and equipment. A whole chapter (19) is dcvote(1 to the design and use of automatic and speeiai-purpose units-a field in wliich there has bcen a tremendous advance in the past ten years. Chapters 8, 9 and 10 have expanded the discussion of field tlistribu- tion in \,a,-ions shaped objects. These latter three chapters ~nciurie some original work by the research laboratories of Magnaflux Cor- poration, anti were prepared by Dr. G. 0. McClurg, Research Director. Chapter 11, on materials, mas preparecl by Bruce Graham, Assistant Director of Mater~a ls Engineering. This department is constantly worlcing to improve materials used in magnetic particle testing. A. E. Cliristensen, Manager of Engineering, is responsible fo r the preparation of Chapter 19 on speciai and aulolnat~c eyulp- ment. The important deveioprnent in t lus approach to more efi'ect~ve and economical testing is the result of \vorlc during the past lifteen years by Nr. Christensen and h ~ s staff of experienced ilesign and development engineering specialists.

In this connection a tribute is due to the late A. I<. (Ding) Saltis, \vho pioneered tlie \\,hole idea of magnetic p:~rticle testing of billets; and, af ter some five years of effort, succeedctl in convincing tlie steel industry of the value of this important and no\\' acceptt?d steel mill practice.

I n presenting a book on a subject a s fluid and gro\\'nig :is is the

l'RIXCII'I,ES 01.' R1:lGXETIC PARTICLE TES'I'IXG

magnetic pirrticle method, the author IS allrare that ~ r l i a t has been written regarding techniques and niaterials will prcibably no longer be accurate some few gears hence. The rcader is aslrerf to appreciate this po~n t , and regard the statements made here a s being an ex- pression of our knowledge and experience a s of the present date, recognlzmg that in a Ee\rs years ne\s7 developments may make some present practices obsolete.

If the bool< is considered as a whole, the rcader xvill find a considerable number of repetitions of various facts, discussions and illustrations. These repetitions are intentional, in order to present a topic completely in its own context, and to avold r equ~r ing the reader to be constaiitly thumbing pages to locate cross-references. A coi?s~derablc number of cross-references are, of course, inevitably mcluded.

In addition to those persoils mentioned above as having niadc contributions to the text, the aotlior wishes to thank tlie many who hare a ~ d e d in thls uXorli in one way o r another. Especially, thanks go to D. T. O'Connor, Dr. G. 0. IvlcClurg, and A. E. Christensen, mho have patiently read and re-rcad the text. F. S. Catlin, Ronald ;\larlcs, Kenneth Schroedcr and Bruce Tyier have also checlied and re- checker1 the entire manuscript. These men have given mliat a t times seemed merciless criticism of statements made and plirases used by the author. For this the author is grateful, since i t lias iiiade possible the ~ubi is l i ing of this volume with confidence tha t i t is tech- nically correct m all details, and that grammar and word usage Ilave not bcen flagl.mitly abused. The corrections and suggestions made have added materially to the accuracy and style of the work. Arthur Debb has been helpful in the preparation of the sections on steel mill testing and practices.

F. S. Catlin and Ronald Marks, \vitIi tlie hell) of thew able staff m the Market Devclopmenl Department, have spearlleaded tlie search for and the selection of illustrations, and kave assumed the burden of the niany details tliat a r e a par t of the publication of any boolc. Mr. Catlin is also res1)onsibie fo r tlie design of the jacket cover.

The linc drawings, ~ 1 ~ 1 t h a few exceptions, are the work of the Ed Adanis Studio of Indianapolis, Indiana. The photographs a r e to a large extent by Gordon Carson Studio, Franklin Park, Illinois.

Lastl), my thanks go to iilr. W. E . Thomas, ~ v h o has in good-

PREFACE

natured patience borne all the f rustrat ing deiays in the completion of this work, and has contributed never-failing suggestions, en- couragement and suppoit.

Carl E. Betz. Greentree Farm, Martinsville, Indiana. September 1,1966.

I N T R O D U C T I O N

T h ~ s 1s a book which covers in considerable detail the many techn~cal and practical aspects of Magnetic Particle Inspection. While i t follo\r~s and replaces the various editions of "Principles of illagnaflux" originally written by the iate Mr. F, B. Doane and later rev~sed and up-dated by Mr. Carl E. Betz, i t is a con~pletely new work both in its organization and its detail, and illustrative material.

As covered by the author in the first chapter, this nondestructive method was first conceived by Major Wm. E. Hoke, but mas developed into a practical industrial tooi by Professor A. V. de Forest in 1929. Following the formation of a partnership between him and Mr. F. B. Doane in that year iwhicli partnership was the predecessor to Magnaflux Corporation) they both contributed ex- tenslveiy to i ts initial development. A few years later Mr. Betz joined them and over the past 31 years has made numerous significant personal contributions.

The method is inherently so readily applied and its results so essential to the proving of the Integrity of structural materials and parts, that i t grew in use to a point where it is probably accurate to say that i t is the most widely-used nondestructive testing method and is employed on more par ts than any other method in use today.

Nondestructive testing is a term used to describe a variety of methods which may determine either the physical soundness o r other characteristics of materials without injury to them in any \rfay. T h ~ s is the point of separation between such methods and the many other tests for strength or soundness w h ~ c h a r e used throughout industry, but mhich destroy or damage the part beyond its point of usefulness. The oldest of these comnlonly used "scientific" NDT

INTRODUCTION

metliods IS radiography, and of course i ts appiications are many and its utilization IS very widespread. In point of time, the magnetic

test was the second to come lnto general use, and in the past 35 years we have seen a number of other test methods developed and a great variety of instrumentation designed and used.

The result of these nondestructive tests has been to help make possible the manufacture and use of lighter and stronger equipment of all kinds, and the use of tha t equipment with a iiigii degree of safety to its operators and to the general public. All of us in this field of work feei a sense of personal satisfaction in Iiaving contributed to the fine safety record of railroads, aircraft , automobiles and trucks, turbines, bridges and structures of all kinds. Yet safety alone is not the sole reason for the employment of these tests. They a r e po\verful factors in econom~cal manufacture a s well. The fab- r~ca t ion of defective materiais is very expensive in both time and money and the assembly of defective components into a machine or structure usually requires expensive rework and replacement.

Magnaflux Corporation's policies have been based upon the high technical standards of men such a s de Forest, Doane and Betz, a s \veil a s their personai and business integrity. I t s people have considered tha t a primary par t of their responsibility i s t o insure t ha t the magnetic particie method be thoroughly understood and intelligently and econoni~caily used. To thls end, a great many educational activities have been sponsored and conducted by the Company, and in this work Cari Betz has iiad a leading part.

I t is a great pleasure fo r me to acknolr-ledge his contributions to the technlcai development of tlie subject of the hook and his personal leadership in the education and tralning of literally many thousands of peopie lnvolved in its specification, use, and evaluation, fo r the benefit of all.

W. E. Thomas President Magnaflux Corporation

CHAPTER 1

HISTORY OF THE MAGNETIC P:iRTICLE METHOD

1. EARLY TESTING METHODS. In the years prior to 1920 the term "nondestructi\~e testing" had not yet acquired a nieanlng o r found a place in the language of Engineering. There were many testing methods, of course, but nondestructive testing as such, \vas not In existence a t that time. None of the methods so important today had even been thought of. Although radiography had been born, i ts industrial application had not yet been visualized. Testing of materials was considered necessary to provide data fo r use in design, or to check the overall physical properties of the materials that were actually used. But such tests were pretty much limited to destructive tests of samples selected from large lots. These tests included chem~cal analysis; tensile, compressive or impact tests of steei and other metals; and similar efforts intended to reassure the engineer that he \\,as actually getting the properties that he had assumed in his des~gn. Proof tests of completed par ts or assemblies gave assurance of initial strength, but gave no information a s to what might be expected later in service.

But, a s a matter of fact, the need for nondestructive testing had not yet become very urgent. 31aciiinery was for the niost par t heavy and reiativeiy slow-niovlng, and large factors of safety were the rule. Of course, fatigue failures did occur In spite of tlie efforts to design strength Into moving par ts by making them bigger than they really needed t o he. Large crankshafts, railroad axies and similar stressed members broke all too frequently. But no concerted attack had yet bcen made on the reai cause of such failures or-on specific methods to avoid them. The par t played by small notch-like defects in initiating fatigue failures was not yet fully recognized-so the urge to find and eliminate such flaws a s a means to reduce failures in service had not yet come into being.

In the days follow~ng the end of World War I, two developments marked the birth of niodern nondestructive testing. Work a t the Bureau of Standards and a t tlie General Electrlc Company had shown tliat X-rays couid be used to obtaln pictures of metallic articles, revealing internal conditions: and the alert observation of

PRINCIPLES O F MAGNETIC PAltTlCLE TESTING

a man-William E. Hoke-also a t the Bureau of Standards had discovered the principle of the use of magnetic fieids and ferromagnetic particles to locate surface cracks in magnetic metals. He found that the metallic grindings from hard steel par ts which were being ground while held on a magnettc chuck, often formed patterns on the face of the par ts wl~ich corresponded to cracks in that surface. This irsas tlie essential basis of longitutlinal magnetization f o r the location of transverse cracks by the magnetic particle method.

I t is true that other roots or seeds of modern nondestructive testing also existed a t tha t time. The "oil and wliiting" lor oil and chalk) method was in use, in the railroad field particularly, to locate cracks in heavy locomotive par ts ; and in the casting tndustry to locate gross cracks. This was the forerunner of penetrant testing, to be developed twenty years later.* Ringing methods to test castings, ceramics and chlnaware for soundness were in use, and couid be considered a s foreshadowing sonic and ultrasonic methods for finding flaws. Magnetic and electrical methods were in use to determine certain physical prope~.ties, though they \\,ere not used for flaw detection a t that time.

2. BEGINNINGS OF INDUSTRIAL RADIOGRAPHY. In the decade from 1920 to 1930 modern nondestructive testing. began to show stirrings of life. In 1922 Dr. H. H. Lester set up in the Watertown Arsenal laboratory the X-ray equipment that started industrial radiography on its way. His work demonstrated that penetration of thtck sections of steel was practical, and that the result revealed defects such a s cavities, cracks; porosity and non-metallic inclusions where the defect thickness was a s small a s 2% of the metal thickness. This demonstration fired the imagination of engineers and sctentists. The idea of nondestructively finding flaws in the interior of solid metal by literally looking through it was an exciting prospect. From that time on, deveiopments and improvements have been continuous up to the present day, and have made radiography one of the nutstand- ing nondestructive test methods.

3. THE MAGNETIC PARTICLE METHOD DEVELOPMENT. Toward the end of this decade in 1928 and 1929, A. V. de Forest did his initial work which eventually resulted in tlie magnetic particle method a s we now know it. It is reported that some abortive efforts were made to appiy Hoke's discovery in the midale 1920's, but the method did not "catch on". I t remained for de Forest to envision the great - *See "Prlncipies of Penetrants" by C. E. Betz, Mapaflux Corporation, 1963.

48

CH,~I 'TEIL I

HISTOIIY OF T H E MACNETIC I'ARIIC1.E hll<TtiOD

Fig. 1-Dr. H. H. Lester's Pioneer X-ray Laboratory a1 the Watertown Arsenal. 1922.

possibilities which the method held. Ile realized that if cracks in any direction were to be located reliably, the direct~on of the magnetic field in the par t could not be left to chance, nor be only long]- tudinal in direction. Rather, i t was necessary to 11:lve some means for generating a field in a?iy desired direction. This could not be done successfully in all cases with external msgnets or coils carrying current-the oniy magnetizing means then availabic. To solve t h ~ s need, he proposed magnetizing by passing a sufictently large current through the ptece heing magnctized. Thts \v;is tlic first usc of circular magnetization, the method now so \v~dclg empioyc~l. He also iiisistcd on the use of magnetic powder of controlled slze, shape and magnctic properties, ;IS being essential to consistent an11 reliable results. These contributions ~~rovtried the tmpelus for the ricvelop- ment of a truly practical and useful nondestructive test.

%$agnaNux Corporation was formed by A. V. de Forest and F G. Donne, first a s a par lners l~ip kno\vn as A. V de Forest Associates


Fig. 2-Prot. A. V. de Forest.

and subsequently ~ncorporatetl in December 1934 a s Magnaflux Corporation. F E. Doane was associated with Pit tsburgh Testing Laboratory in Pit tsburgh, Pennsylvan~a a t tha t time and did some of the early work on t h e development of magnetic powders there. late^, much of the work of tieveloping techn~ques and the use of alternating current directly f o r magnet iz~na, was done \vith a. "breadboard" set-up of equipment In Doane's basement a t h ~ s home.

CIrarren 1

1:ISTOIIY 01' TIIE MAGNBI'IC l'~\ll'rICl,E hlE1'1101J

Fig. 3-F. B. Doane.

4. PROGRESS BETWEEN 1930 AND 1940. The decade f rom 1930 to 1940 saw large developments In r:tdiogr:~phy and m:ignetic particit: tcsting, and also saw the st;u.t of otlier t io~irlcst~~uctive testing methods. Figiue 5 is a dra\vlng whlc11 shows :it a g1;inre the nrogress and s u b s e q i ~ e ~ ~ t growth of v a r ~ o u s methods up nilo the 1960-79'70 decade. The reai proliferation of new mct.ho~is rlitl not st:iri t i ~ i t i l

the beginning of World War TI, rjilt has conlinoed a1 a n ;icceler:it,inl:

I C ~ A J ' T E R 1 FRINCII'LES 01%' RIAGNETIC 1'AILTICI.R TESTING

~- - HISTOIIY 01: TI IE ntnGxlil'tc I'ARTICI~E METROII ~ ---- ~ ~-

Fig. 4-Early Experimental Equ~pment used by F. 0. Doane. 1930

ra te f rom tha t time down to the present writing. There is every reason to believe tha t in the fu tu re many new methods will be devised to solve a s yet unknown testing problems.

Rut dur ing the 1930's the progress seems, from our present view- point, to have becn painfully slow. For the mag~iet ic particle method, teelinlques hail to be dcvelopcd and explored; equipment dev~sed and materials im~rovecl. Above all, users hail to be found to give In- dustrlal life to t lus promising testing procedure. I t was during lhis p e r ~ o d that the author became actively associated with the Cor- poration 11935), and contribute6 to the work of developing ina- terlais, eqtllpmcnt anil techniques.

$2

Fig. +Early A.C. Magnetlzcng Assembly. 1933

5. EAIII~Y GQUEPIIIENT. Ear ly :~ppi ica t ro~is were o .ude and early cqutj)metit s imple and nut yet pr~diict ioti miticled. Figure G shows an ;~ssenibly of elcctrieal compc,t1ctits-tri1ns1'c)rmer, switch, meter. ctc.-such a s \\.:ks pu t logelller 111 tile plant f o r early users. Such an tis stall at ion emnloyed ;~itcrii;iting citrrciit fo r magnetizing, anil \\,as first used f o r the test ing of tool-steel bars. Bu t most of the equipment :it titis time omploycrl rlirect ctiri.rnl f o r mngtieliz~tig, clerlvctl from s t o ~ i g o balterics. I'igure 7 sho\\zs :l ttnit e .utens~\~eig used in the :iit.cr;ift tnduslry i t i thc 1:rtc 1'J::O's. \vlilcR nro\~ide<l both ctrcol;~r a n d inngit~rdin;i~ m a g ~ i c t t z ; i t i o ~ ~ f rom storage i ~ a t l e r y i>o\\'er. T h e ~.;~ilroucis :,is0 were e:rrly in t l ie i~ . use of t h e new methoti, ni;riniy f o r tho Ioc;tti<~i~ o l fat igue craclcs ti1 a s l i ~ s , atid motion pzirts of steam loconioti\res. T h e aulomolt\~<! i~ i#ius t ry , too; rapirily becaiiic 111- terested atld by 1940 nl;iiiy titiits !set.e 11, itsc, i)olli f o r t h ~ t n s l l ~ c i i o i ~ o i trew t,;iris such as 1-rntit \\,heel s r ~ t ~ ~ ~ l l e s , cl'atlksilafls ancl cotiilcct- 11% i.nds, atici i r ~ r the 1o~ati01i riuritlg ~\,erlliiiil, of faligite cracks on bus a n d triicl< ctigine p:irts.

Duriiifi Ihcse yea r s r ; i r i i~ ,~ r : ip i~y alsc] mark tremet,<ious s t r ides ; and some fo rms of erlcly-curret~t tcst lng i rere pttt iiilu S I I C C C S S ~ U ~

use, nota l~iy f o r tiic i i i uc t~ t lon of steel bars for sc;inls

6. DE\'ELOPMENTS OF THI- 19~10's. The allvent of 'VI:urld \\'a,. I1 brought a rapid Increase 111 the ~ l e m a n d fo r all nonrlestrttctt\~e test-

Flg. 7-Storage Battciy U n i t Used by the Aircratt Industry. 1932.1940~

ing methods. The m:iglietic particle ine t l~od &%.as rcaciy to hand and found tnnumerable ne\v applications. The ability of nondestructive test ing techniques to assure the ouality and t~il~:firity of all corn- ponents of a n assembly, r a v e confidt!nce t h a t equipmetit, a rmament and ammunition would perform reliabiy III tlic lieid. The need to speed prorluction of w a r m;itei.i;ils tnt t i ;~ted the destgn and use of automattc tes t ing units. Figitre 8 sho\\'s ;in automatic unit fo r test ing a r m o r - p i e r c i n ~ projectiles. srhtcli \WE. pu t itito service i n

1943.

I'IZINCIPLES OF RlAGNUSIC I'AWTlC1.E 'VESTING

I~'iuuresccnt p e n i ~ t r i ~ n t methorls appcareri in iniil-1942 aiid foiitid t n r r n e ~ l i ~ i ~ e ;it~pIic;itioti. I'er~ctratil mi~tiioiis rrerc bailly nceried, iu dil i'ur iion-l'eri'ous niatci.i;ils \vhat tlic rliapiictic ~ ~ i i i ~ t t ~ l e metliod was iluiiig 111 the testing uf il.o!i and steel parts . Fluorcsceiit penel ranis \\,ere pu t t r ~ i11st:~nt UJC. Flui~rescellt ninglictie p;irt~cle.i, too, \ \ w e tiitri~~lticcci al)oul tlic same lime. They foiiticl many a ~ o l i c a t ~ o i ~ s nl oiice, althougli their p rea t u:ilue \\*:is not rcnlizeil iititil the Sttar \<.as ,>Yet'.

iJlli.;isoiiic fla\v rlelect~i)ti, 110th i ~ y tlic resiui;!iiee iinrl tllc iliilsii- c,ciii> mcI.liods, \sr:i-: \ru~.lierl oiit nnii used to sotlie c s t c i ~ t duriilg the \v;ir, Ll~~>iigli tile 1;irpc (lfv~!lopniciit of this iisrliil iccliiiioue r l i i i 1101 Lakc 1il;ice until a f t e r the iva~..

7. PIIOI$I,ITMS 1)1113 TO ILr\l'll> F>xI'ANSION 01' [ISIC. ?'lie r;ipliI c s - t>; t t is ic>~i IJS l l i~ t use of m:igtictic p;t~.Licle lestilig iliiriiig Llic \V;II, :it lirsl crc;llecl iii;in? pri~l)lcms, mostly iluc to i;icl< ol' esl)erience \villi Llic iiictlii~il. Crrvur~iiiietit t .ci i~t~~.emeiits fr~i. ~iispcctiuti I I Y me;!ns of m;igiietic p;irLiclcs orteil st!tniied iilireason:i\)lr!, i)ui \vcre applietl \rirlcly. Pl;!11)8 m;~iiufaclitrci.s \vfio had never lie;it.il <]I' tire niethi,ri rr!scnt~:il tlic ;irltlr!il t r<~ul) lc axid cspetisc tniposerl iipoti lhcin. In- ~ L r l ~ o t . ~ c ~ ~ c e i l Co \~r rnmcnt I I I S I I ! ~ ~ U I rcjecletl ; ! I L l i ;~! sho\\,ctl

Cii:,l"ri?ii s

ISISTOI~Y 01: T I IX nl:\(;sl?ric I ~ A I ~ ~ ~ I ~ L I : ~II.:TII(II) -

intlicalions, wluch r ~ j c c t i o n s L:iLer cxnwiencf )lrijvcil lo haoc l ~ e e t ~ un]itstilietl. I t took 111:lily yc;irs ti) ncvitmiil:ltc i!nr~iixli esnerlence atiil d:it:r to iiitcrt>ret mngneiie p;irticlc inrIic;itiotis coi.rccL1)-, so that tills landeticv to\raril i~vel'-ii!s~~ectioii cliri i>ot caits? \\.:isti, of g ~ o i l matt.~.ial.

The Air Force st;iiiiiiirdizcrl I J I ~ llic \Vet iiiclliorl n.itIi 0ir.cct ciirrent, nitllough 111 many cases tliii {It.!. ni~lhocl with ;iltr>rnalinp eu r rcn t was Lettcr ariapl.ed fo r ;I p;irtieiii:ir ~ i i s p ~ c t i ~ i i prolilen~. Later 111 t h e ivar tliis position \\,;is moclilic~rl anil itsc of A.C. fot. mngnetiziiig \\$as approveil by tl1c A t r Forel! uri(lcr certain circiim- stances. Rut in n niimbrr of cnscs it \!'as foiinrl tha t l l n \ i ~ iintnrn to exist \\,ere liot 1niiic;itetl by ctther metliorl aiid tnipi.ovrr1 irch- niques hail t o he d c ~ i s e d to iiisiire comt~lcti! rcliahilily fo r tlic tests.

8. POST-Walt D E V E L O I ~ ? N T S . 111 the years since the close of World \Val I1 a raptdly pl.o\uing list of iiew test mellioris has come into being a s the rcsiilt of test ing problems. Tlic rnagilel.ic

Fig. 9-Automatcc Unit Including Part Rotation tor Testrng Automotive Connecting Rods, U%ng Fluorescent Magnets< Paittcles. 1948.

57

1'RINCII'LI;S 01' RI.AGNI:TIC I'AI(T1CLE TESTING -- ~- ~

__ - -

p,artlcje method non(lesrructi\.e test ing has , howrver, r r t a i ~ i ~ i l its pre.emlnent position fo r t h e tletection ilf surface aiicl near-surfzicc flaws in ferromagnetic materials. Const:lnt lml~rolrcnlent of ma- terlals an11 technitlues has kept the mcthori abreas t of t h e dcnlallils maill? bp advallces in metals and alloys, a110 thei r utilization in new des lp~ i s and new applic:ltrons.

Outstanding has been the 1n1pro\vn1ent in the magnetic onrticles tliemsei\~es, slid especially the full realiz;~tron of thc value of the Hitorescent method fo r r;ipld and reliable inspection. Of etlual Im- portance has been the application of the methurl in lie\!, fields, m:ide possible by the recent introduction of ne\r ~ n a g n f t i z l n g technirlues. ill111 accumulated experience 111 d e s ~ g ~ i l l i g I;irger and more automated eouipment. Among these de\relopments a r e multi-dii'ectional magnetization; the rnajinctiz:itio~i of ia rge castings arid other pa r t s by tlic use o i very high amperage currents ( t h e "overz~ll" method) ; water-borne suspensions ol' maglietic part icles; an(!, in the a rea of equipment, the construction of large automatic equipment for the ~nsnect ion of steel tube rounds, billets and blooms of cons~derable

ICourtesy Arncrr~dn Steel m d Wire Dirrr~on. U.S. Steel Cori,orr%ionl Fig. l0-Autornat!~ Unit for Testing Steel Billets for Seams. 1956.

58

(:H,\I.TEII i

IIISTOI(Y 01: Till.: >l:\<;.UI:I'lC I'AII1'ICl.E RII<TIIOD ~ ~~ ~- ~. - ~

9. XL'CLEAR ANI) SI~:XCE-.~GZ: I:EQI!IK~~IENTS. The need f o ~ . absolute reli:ibiiity and fziiiiire-frcc perform;ince of missiles and rockets, aiirl nuclear poxrer re:ictors, has placer1 :I new responsibility on all nondestriictivr iestiiig pruceriures. R'lan!: rre\v n~e thods , a n d ne\v techniques fo r olrl methorls. h;ive :i~ip6!:lr~'rl to solve Sonic of the new test ing probli.ois ar is ing f rom the use of new m;tter~:ils and riew construction ideas.

T h e magnetic narticle method has C o u ~ ~ d i l ic r~?as~ng usefulness i r i the solutron of m r r o o s of these n t w tes t ing probltvns. No otlier no~idrs t rucl ive test ing method 1s i ts equai fo r the detection of very fine end very siiallow surface cracks iii ferl.omagnetic materials. Penet rant iiiethods have thls same pre-eminence f o r non-ferrous metals and other non-magnetic mater ia ls such as plastics, ceramics. ctc. -4lthough both methods have certain Iiniitations, and a r e ex- celled rn c e r t a ~ n applications by other k ~ i i d s of nolulestructive tests. these two methods h a r e continually enhanced t h e ~ r pusit io~i of b a n g mor r \\'idel? used than any other methods.

10. FUTURE OF MAGNETIC PARTICLE TESTING. The acceptzince uf nondestructive test ing by Indust ry is today practic:~lly universal. I ts value, not oniy in assur ing flat\,-free pro~lucts , but also in eKcct- ing material cost savings in the manufacture of almost every so5.t of product, has piaced it in t h e position of being an essential tool in the manufactur ing process. There is no question a s t o t h e continued usefulness of the magnetic part icle method in such a climate. Experience with the method is now s o extensive t h a t confidence in i t s indications and in the ability of experienced Inspectors to In- terpre t them, 1s the rule in those i n d u s t r ~ e s \\,here the method 1s

applicable. And the record already established of adaptabil i ty to t h e sojution of nextu tes t ing ])I-oblems leaves little doubt t h a t \ ~ l l c n new situations ar ise in i t s field, new problems will be solved.

CIIAI>TER 2

I~'Uh'l).ZMISNrl'~\I. ( 'ONCEPI" 01' THE iVll':'I'HOI)

1. ~ ~ T I $ A T IS T H E MAGNETIC PABTICI~E METIIOD? The m;ignetic t,articlc metlii~il of ~ r t ~ n r l ~ ~ s t r u c t i ~ ~ c tvslilig is ;I mi~thocl fo r loc;~t t~lx surface ;und subsitrf;ice ilisco~it.itiuilies in ferromagnetic ni;ilerial. i t rlepends For i t s oi~eratioli on Llic inct t h a t \'hell the n1;tterial or pa r t iinrler test is iiiagnctizcil, discoiiliiiuitics which lie in a direction geiiei.ally ~ I . ~ ~ I S V E Y S C to 1111' (iii.e~Ltoii of the niagnetic field n:ill cause ;I lealiage field to lle formet1 ;it ;liirl :ibove the slirfnce of the ilart. Tlie presence of tlits leakuge field, and therefore the ni'esenci.> of the cliscoutii~uity, is rtetecictl by the iise of linely divirled fe r romngne t~c ~ ~ a r t ~ c l e s :i~~piierl over tile s~ii.fitce, some of these l>articles Ibeinp gallie~.ecl 2nd liclcl I)). lllc leakage field. TIils niagnettcnlly licld col- lection of particles foi.liis a n outline ol' the cliscontin~irly a ~ i d in- r1ic:ktes its location, size; sh ;~pc and extent.

Flg. Il-Typical Dry Powder Pa t te rn Thss is the Orlglnal Demonstratton Piece Used by A. V. de Forest, acid Later by F. B. Doane.

Tliere ;ire niaiiy f:lctors thnt :~ffcct t l i ~ fill-mnt.ioti ;itid appeslrancti of lliis po\\rrlet' 1~;ittci.n. Soinc of tlicse ;Ire:

( a ) Direc t i i~ i~ i i ~ ~ r i sti.i~iictIi of the m a x ~ ~ e t ~ c iitkicl.

GO

( b ) filetllod of magnetization employed.

( c ) Size, shape anri (lirection of tlic discontiiii~ity.

([I) Ciial.acter of the niagtletic powdcl. and tile inelhod of apl]lyiiig it.

l e ) h1agnetic cllar:iclerist~cs of the 1~ai.L being tested.

( f ) Shape of the part , mhrch alfecls the distribution of tlie n iagnet~c field.

( g ) Character oi' the surface of t h r part-whether smootli ,,I. rough, or plaled o r na~nt r i l , etc.

The etYect of tliese various f a c t o i . ~ on the obtarliing of ootinium results in the detection of disconti~ioities will be discussed in detail in subseqoent cliapters.

2. 1-Iow DOES THE METHOD WORK? The nlctiiorl involves t1il.e~ essential steps :

( a ) Magnetizing the material or par t under test .

( b ) Applying the ferromagnetic particles o \ w tlie surface.

( c ) Examining tile surface fo r pomder patterns o r indications.

3. I~ACNETIZATION. Tlie magnetic fieid inside tlie pa r t must r u ~ i in a directioil such tha t 11 is lnterceptecl by the discontinuity. A number of techniques a re in use to accomplish this result, anri will be discusser1 in later chapters.

Fig. 12-Field Distortlor' at a Discontanulty Lylng Wholly Below the Suriace

iil

I'BIXCIPLES 01" I~III(;XE'FIC I 'AWICLE 1'ICSl'l.i'(; _ .- - .. _ __-._-.--_______-..-

\4'iieii tlie field is intercri~teii 113, t h e ~l iscnnt i i~ui l ) - . aurnr: of tlie field 1s forced out into tile a i r :tbove t h e ~i iseoi i l i i~ui l~: , Lo foi.111 a ,.lc;ikagc lielil" Figtire 1'7 i l~ i l s t ra les tile uath of flus lines liltel.- cepteii by a crack-like iliscontinuiiy lying \vl~olly l ~ i - l o \ ~ the surf;ie~!. In tills case tile l l~s foIlo\!:s th1.i-e :lItern:ltlve patl is:

( a ) Around thc tIiscnnl.inult)~ oil botli edges, c;ilisln$! ;ti1 i i~o 'ease 111 flus ilcnsity in these areas , in tlic interior [if Lhc material.

) Across tile d i ~ c o ~ i l i i ~ i i i t y . 'Vhis n:lth olrc'rs a high re I~cLa i~ce o r I-cstsiancc to the pitssage of the flus. siilce the l ~ e r m c : ~ h i l i t j ~ of the gap-usually alr-lillecl-IS 111tic11 i~\vf:r t han t h a t of tlie sor1-nunding met;ii.

i c ) Through tlic ;11r f iuvc t h e ~Iiscoiitiiiuity, c;iusiog a lcal<:igc field. The s t r e n g t l ~ ;knit l i~ieiisi ty of tliis leakage fielri is lie- ternlined to 21 1;lrgc esterit by tlie ;imouiit of mctai between the discontinuity niiii tlic siirface, the s t rength of t h e iielii i i l s~de tile metal, :i11i1 t h ~ ! rlinic!nsions of the discontinuity in a ~lirectioir a t r i ~ h t :tnglcs to thc surface.

I 1

Fig. 15-Field Oistortlon at a D~scontinutty which is Open to the Surtace

Tlie cioser the discontinuit.!, 1s to Lhc slirlkce, tile s t ronger allti more concentrate(1 \\rill be tile l ~ i i k i i ~ e iieitl. If tlie i l i ~ c o ~ ~ t i n ~ i i t ~ ~ actoally bre:iiis tile surf;~cc, t1;1t11 (;I) bot\rri?cn the discontiiiiiity ; in~i the stlrface ria lon~c! . eslsts . ail11 [lie le;ik;ige ffox jomps tiic g;iii

created 1)). tile riisco~itiiii~ity. In this case llic field a t the sul.f;~cc will be s t rong ;incl s11;irply coiicci~tl.ntc~l, i f thc rliscoiitiiiiiil~~ is !i:iri.ow and cr:lck-like. If it is sh;illow, ;in11 morc l i l x ;I scratcli, LIIEI.P

may IJe IIG leakage field :it all. Tile Illis \ \ - i l l siilinly s~rennl- l ine t)el~iw tlic scratcli in the i~orly of tlic inet;~l . 1:igur.e 1-1 i l lus t i .a t~s LIiis ease.

I I

I I Fsg. 14-Field Distort~on at a Surlace Scratch.

4. AI'I'LYING THE FERROhlAGNETiC PARTICI~ES. Af te r tile n;irt under test h a s been properly magnetizeil, tlio Serroinagiietic par- f.icles a r e applied. These may be dusted on a s ;i d r y po\\.iler, 01. flowed on ns n suspe~ision in some liould. Tliose pai.ticles \\'liicl~ c o n ~ e under

Ftg 15-Typical Discontanulty Pattern a s lndlcated by the Wet Method (Seamy Wrtst Pin)

the ~ ~ i f l u e n c e of the Imltage lirlrl prorluccd by n (liscontinii~ty r~fFe,- a io\r.cr reliictanee path to Lhc magnetic f111s ~ R ; I I I rlocs t11e art., an11 they a r e c o n s e q ~ ~ r n t l y ilr;i\vn ;iii(l heid by the ir!;rl<:~ge iicirl. Tl~escs m;ignetically i h f l i i p;irticles t l ~ e n u~.o\'~cli! the v~s ib lc evlilciice of thn presence of a 1e:ikapc lii,lil, :rnd t l i e r e f ~ ~ r v tile i ,~ ' r s t ,~~cc of siilne ltrnil of discont~nuity.

Since Ical<;~pe fields c:in i ~ e nroiluceil by co~iclitions other 111;111 rnetallic discontinuitics. :I m a g n c t ~ c uarticie p;ittcrn 1s I I U ~ i~r.iiiin incir e\.lcience of the presence of :I "defect" Conclitions such a s t l~ose caused by two components of ;III :~ssemhly in contact, or a sh ;~sp change I l l section \%'ill also pi.oiluce Iwkagc fields anil patterns of m;lgnelic particles. Sharp changes I I I permeability f rom ally caitsc \\rill do l ikew~sc. These circumsl:~nces \vill be ~lescsibed 111 detirii I I I

later chapters. Strength and concentration of the 1eak:ige Aelil 1s ;i controlling

factor in d e t e r m ~ n i n g the ch:1r:icter anrl appenr:lnce [IS the 111dic;i- tlon. It must be s t rong enougli to ovescomc g r a ~ ~ t y , sl~rl'iice ~ r r e g u - I / ' a l ~ t i c s , blouzi~ig olf of excess i~owt lc~ . o r ill.alnn,g (IF rxct~ss iiotl~d suspcnslon, ant1 other forces lendinp to p r f v ~ i ~ l the ~a t l i e r i r lp anit holding of par.ticies to foi'rn ;In r~~c l i ca t~on .

5. EXAXINATION OF TIIE SURFACE FOR ~ I A C N E T I C PARTICLE PATTEI~NS. \'isu;il 111speclio11 of the surf:rce of the i ~ a r t a f t e r 11 has been magnetized anti rnagnetic p a r l ~ c i e s applieil requwes o n i ~ gotxi light, good e y e s ~ g l ~ l , ant1 close attention : ~ n d alertness on the pa r t of the ~nspecto!. Thc magnetic pnrticies :ire availabie 111 several colors to iiicrease tile v~sibil i ty of ~ u d i c a t r o ~ ~ s under any give^^ conilit~ons. Fluorescent magnetic parlicies 1\4iich glow i n darkness o r din1 l ig i~ t \!'hen exposed to ,>ear-ultraviolet ratlintion ("biacli l ight"), oflei. the ultimate 111 contrast and visibility, and a r e i ~ o w \,el.?. ~ v ~ d e l v used in many applications.

The ~nspec to r usi~ally m:~rits the ioc;~tio!l of the indic;i t~oi~s xvh~ch lie sees, aild may o r may not, depending on 111s expel.lence anti the nature and importance of the insl~ection, ;~cccpt o r relect the pa1.t at. tha t time. In some cases iie ma)- segregate those ]i:irts shn\ving nldicalions, anti Iroltl then1 fov later r l isposit io~~. Sometimes t111s may require t h e jotlgment of quality conlr.oi nr cleslpn i~ersoniiel. With Incrc;istng expericllce in w u ~ ~ k i ~ ~ g with the methotl hr~\ve\~cr, tlie inspector hin~self actjolres, I I I most cases, tlie 1!11lgme11t n c c c s s : ~ ~ . ~ fo r m:lkirip the ;~ccept-i .e~ect dec~sioll ih~nlself. 111 Ii~glliy r.epctiti\.e tests, as 111 mass i>ri~iluctirul. all tlw \.:rrlai~lcs ~ I E t l ~ e test c;in be

ClrAi.Ter f!

I~'l!Nl):~hlRNTAL COS('l<lVfS OF 'THE Z1ETHt)D -. .~~~~ -. .. - ~ -. . - .

conlrolleii; so tlrat o?ily those ~iel'ccts consirlered importa i~t ;1l.r ~ndicated.

6. \VHAT CAN Tl lE I\.I:\tiNETIC PARTICLIC METHOU FIN!)'! The method oper;iles to rnr1ic;ite liif l3resence of surface le:lkage firl(is, and any such lielris from wl~atevcr cause will he ~ntIic;ited, provrrie(l tha t :

( a ) Tibe field 1s strong e i ~ o u g l ~ to hoid the particles ;~pplicd. Le:~l<- age fields a t vtLry s l~al lo\ \~ tliscontin~iities niay be very weak, and the licitis a t t l ~ e sui.face Fornietl by discont~~iui t ies ncliolly belo~v the surface may he \reak and riiffuse.

( b ) The magnetic p;irticlct; used ;~i.e s u ~ t a b l c for tlie purpose-- tha t is, a1.e fine cnoupli and have luph rnougll permeability to be held.

Untier proper c~rcunlstances-that is, nritli proper po\vrler aiid proper n~api~ctiz;ition-excceclingl!: fine discontinuities can be dc- lectetl. Even gri~ii l boulidar~es 111 steel and the oiitlines of magnetic dumarns can be srlo\r7n by using sl~eciai teciln~ques. The magnetic particle method 1s the ?>lost .seirsili?~e ir~crins o~~a i l r~Ole for loca t~ng vr1.y f ine and tre1.y .siinllorii surface cracks in ferromagnetic ma- terlals. At the other end of the scale, indicatioi~s may be produced a t cracks that a re large e11ougli to be seen by Ll~e naltetl eye. 111 this case magnetic particle testing 1s still worth while; because tile presence of a prominent and easily seen powder pattern makes fo r more r a p ~ d inspection, and assures tha t the c~,ack \\'ill 1101 be nllssctl by the inspector. Exceetlingly w ~ d e surface cracks will not urocluce a powder pattern a t all if the siirface openlng is too \\>ltle fo r the particles to bridge.

Discon t~nu i t~cs \ \ fh~cl i do not ;lcto;~lly hrealc tlirougl~ llic surface a r c also indicated in man), instances l ~ y tlits metborl, though e e r t : ~ ~ ~ ~ limitations in th is case most be recognlzctl anti unrlerstoocl.

If the discontini~ity IS fine and s i ~ a ~ , p and close to the surface-as, fo r example, n long s t r ~ n g c r of non-metallic ~nclus~ons-;l pootl s h a r p indicatio~i can be l~rotlucerl. As tile discontinuity lies deeper, the nlrlicat~on becomes f a ~ n t c r . P u t ano t l~er way, tlie deenc~. th r discontinuity lies belo\\2 the surface, the larger it must i ~ c to give ;I reatlnble ~ndic:~l ioi~. (See the disciission of Detectable Defects, 111

Chapter 20.)

The general statement can tlrercfore be 1i1;1dc that tlii. n1;ipnetic

particie method of nondestructire Lc:stilig is nru-enllneiit for lilirling all sizes and depths of cracks tha t 1ire;ili the surf;ice, biit runs into more aiid more difficulty in detecting interiial <liscoiitiiiuities a s they lie deeper and deeper belo\\; the sliri'nce.

7. ON \VH?\T KINDS OF h1.4TEIII.41. I ~ E S I T \vOllK? 'rkc m~tll0Cl can be used on ally ferromagnetic materi:iI, tliongil not in ;ill cases

equal effectiveness. I t \vorlis !lest on steels :lnd nlioys th;lt l1;lVc a h ~ g h permeability. Discontinnitirs lying ~rilolly below 1:hc siirfact~ a r e more likely to be located in soft steels h:i\ri~ig Iiigli pei~n~eahilit) ' , than in hardened steels and alloys \ishlcb in licurly ;1II c:lscs have a lomer pernleabilit)t. T h ~ s difference IS lcss critical if siil.f:icc liefects only a re b e ~ n g sought.

In the case of gray o r n~allcxble iron castings, su~.face cracks a re easily located. Tile method :~ l so \rol.Iis riuitc \$'ell oil metallic ~ ~ i c l i e l and cobalt. On the other hancl, s t a ~ n i e s s stcci :inil o ther ;il!oys n ~ l i ~ c h a re 111 the aiistenitic state cannot be tcsted with magnetic particli!s a t all, since iron in this s ta te is non-magnetic.

8. WHAT AXE THE ~ U V A N T A G E S OT THE %fE'rIIOI)? The nlaglletlc particie method has :I number of outstanding a d v a ~ ~ t a g c s within its field of usefulness-that is, on ferromajinetic materials. Some of these a r e the following:

( a ) I t 1s the best and most i.cliable metliod ;ivailal~Ic for finding surface cracks, especi;llly \,fry fin,! ;lnd shallo\\, ones.

l b ) I t is rap10 and s!mple to operate.

Ic) The indic;ltions a r e produced rlircctly on tlie siii'face of llio art, and a re a inagnelic pictore of the actuai discont~nuity. There is no e lec t r~c circuitry o r clcctronlc I'L!:L~~-oLI~ to be calibrated or kept in proper operating condition.

I Operators can iearn the n~ethorl easily, \vithout lenatliy o r highly t e c h i ~ ~ c a l t~.ainrng.

l e ) There is little o r no limitation clue to size o r sli;ioe of the par t being tested.

( f ) I t mill delect clacks filleri \ \~ith f o r c ~ g n n ~ a t e r ~ a l .

( 6 ) No elanor;$~t: 1~t.c-cic;ir1111g IS oi.~li~iai.ily necessary.

l h ) I t ~irill work \\,ell through thin coatings of pii~ril, o r ot.h<?r non-magnetic coverings such a s plating.

CilarTEn L

l ~ ' l i ~ l J : ~ ~ 1 i ~ ~ ~ ' ~ : t I ~ <'f~l?i<'I<l'TS 01: THE >l i? r l i i l l J ~ ~ ~-

( i ) Sliillecl i~ilc~.;itors can Juiipe crack cleptii uuitc ;icciir;itcly \sit11 sui ta l~ie pon.rir~.s anrl pro11r? terli~irquc.

(j) I t lends ~iscl i ' \yell to automatiun.

11) It utilizes elect~.o-mccl~;iii~c;il eouii~ment t1i;it C;III t ~ e i.uggcdl\. biiilt fo r plant environment, ;ind :ideqi~;itel\: nia~nt:iinc,I by erist inji 1x;int pri'sonnel in niost ~ n s ~ a n c e s .

9. \$'HAT AilE THE GESERAL LI&IIT:\TIONS OF 'i'llE iVIETII0D" Althougil Llle method has m:iny ilesirnblr ;uid attr:tcli\.e ailv;intngcs, i t has, a s docs every nx!tlio(I, certain I io~i tnt io~is . These the opwntoi. must be aware of, aiid take into a rc r~un t hy obscr\-111g the prec;iti- tions \vh~ch they dictate. Some of these :Ire:

( a ) I t ill n.orli only on ferromag~iet ic nla to .~aIs

b ) It is nut 111 all e;ises l.eli:iblc for ioc;itlllg r i i scor~ t~r~u i t~es !t'Iiicii lie \\,holly belo!\' the s~irf:ice.

( c ) The m;igiielic lie111 riiii,sl lie in a 11irect1011 th:it will intercci~t the princlp;il plane of the ( I i sco~i t~n i i~ tp , a ~.coulrement 1l1;it the nper;itor il~irsl Re ;r!v;ire of nnil kiii~iv lie\\' to nieet. Sonif,. tinlcs tills I.eoiiircs t\vo o r n1oi.c ~ e a o e l ~ t i a l inspectioiis \r.!t11 ililfereiit ma~i i r t i za t io~ i s .

id) A second s t e ~ ) - i i e n ~ a g i ~ e t ~ z ; i t i o ~ i - i ' ~ > I I ~ ~ \ v ~ I I I I I C ~ I ~ I I 1s often necessary.

l e ) I-'ost-cleailri~g, to i.eniove 1~en1ii;ints i ~ f tlir rn:ijiiiet~e p ; i r t ~ c l ~ s clinging to the sorrnce, may somctlmcs 1 1 ~ rcr!iiii.cil :iSter testing anrl riemagnetiz~ng.

(E) Orld-sba~etl p;ii.ts sometimes i~rescii t ;i 11rcrblcm srith respect lo llo\\. lo ap])l? the magnetizin!: forcc to ~ ~ r o ~ l i i c c ;I iirlcl i i i

the proper rlirectio~i.

) I:xcectli~igly iioa\*y cui.rents a r e sometinies rcqilii.erl lor tlie testing of vel.y 1:irgc c ;~s t i i~gs ;uiri f o r g ~ n g s .

(11) Care is i.eouired to ilvoicl local lie;iti~i!: anii I I I I I . I~ I I~I : (11 i l i ~ i i l \ .

finishetl 1 ~ 1 i . t ~ o r surfaces :it llic potnis ol e1ccti.ic;il co~it;ict.,

l i ) ~ n c ~ i v i d l i ; ~ ~ h:~nllliilg of par ts lor m;i!:~lctiz;itioii is iisu;\llo iiecesr:iry, n cIisa<loa~it;igc ~~;irLiculai.lr wit11 I;ii.!:r3 ~ i i t n ~ l ~ c r s of small n:irts.

PRINCIPLES OF RIAGNETIC PjiIITICLE TESTING

( j ) Altliough the indications a r e easii~r seen, experience and skill in in terpret~i ig the significance CIS the indication is sometimes needed. Comnareri to sonic other methods such a s radiog- l.epliy, ultrasonics alitl eddy-currents, iiowe\~er. interpretation of magnetic particle indications is f a r simplei in most cases.

lo. CO~IPARISON WITH OTHER METIIOLIS. There a1.e four other m;ijor nondestructive testing methods commo~iiy in use for the dctectlon of surface and subsurface discontiiiuitics in ferromagnetic rnatcrlals. In determining \vhlch of tliem to use in any given casc, the inspector should Rnow the reiative merits :ind limitations of the several methods. Thc following brief comparison be liclpful In ;lrriving a t a propcr choice when such a selection is required:

( a ) Rndiog~ouIr!l is super io~. to niagnetic particle testing in most cases fo r the location of disconti~iuitics which lie wholly be- l o , the surface. l t is not nearly so effective, however, fo r locating sorfacc cracks, in \vlirch ,111nqartic ~inl'ticles r?.z'eel iiII other ~irt!lltoris, on magnetic materiais. The magnetic particle method is usually fas ter and less cdstly to apply than is radiogral,hg, est~ecr:illy when 100% ~nspection of numerous articles is required, and the results a re immediate. I t is also less hampered by the shape of the par t than 1s r:~diography.

( b ) Pc?zetr.rntts ;ire the equal of magnetic particles for the location of surface cracl(s, proviileti the cracks a re ciean and a re not already iillcd ~ v i t h foreign substances. Because of this 1imit;ilion in the case of the 1helieli.ant method, and tlie fact tha t it is often slo\vcr, the magnetic particie melhotl, ~vliich prorluces immediate results, is to be preferred on magnetic materi;il fo r locating surface rlisconti~iuities in most c;ises. For loc;itinp discoiitinuities oil the interior of the metal, pciietr:ints a r e of course conipletely uiisuitablc, wilereas on magnetic matei.i;lls, rnagnctrc particles can do a remarii;ibly good job Illi(ler favoi.able circumstances anil controllerl conditioiis.

i c ) lilfl~nsorr?lrl is inferior to the magncilc particle method 1'01. the location of esceeilingly shallo!\r surface ci.acl(s. There is ?lo li~roli~ii 11111111r111111 iii1111 to tlie rlel)tl1 of criicks magnetic particles will detect, i e l proper n?atei,ials and tech-

68

nigues a r e used. lJlt~~:isouii(l, oil the other Iiaii(1, ~ ( ? ( I L I I I . E S :I

certain m~iirniiim cie],Lh-:iIt!ioiigh this is very siiia1l.-111 ordo. to get soun(i \ ~ i t \ ~ e reflection. I r i the c:ist, of brlom-tli~!. sni.f;lce disco~itinuitics ulti.asoiin(l is i i i general s i~pcrror , a s it will locaic iinr1i.l. hvor :~ble c~~nciititilis, tliscoiitinoities a t irrcat depths l>clo\tz the surf:ice. I-lo\\rt!ver, magiictic i,:irticles a1.e less l1ant1ic;ippeti by irrcgul:ir shapes than is ultrasounri.

Id ) Edd~~-crii .r .~~,ris a r e sub,lect to limitatioiis s~mi ia i . to ultrasound for the location of sui.face c i w k s , silicc a iinite riet~tli is requneri to iinbalance tlie iiilperlance bri(lire :mri secure readable indications. Erlrly current nietliorls in gcnrriil ;ire IbeLte~. adnptctl to tlic testing of iirin-magnetic thaii rn;igill~tic materials. Etld)' ciii.reiits, in conjunction \\.ith niagiictic sa turat ing bias ficlcls, ;il.c soiiietinics bcttei. 1'01. llie lociitioii of below-the-surface discoiitinuities tliaii m;ignctic i,ai.ticles, but tiler a re more limited in the iliv~tli they can peiieti.;it,b. They can us~ially indicate oiil!. ~liscoiitiiiiiitics lying elosr to tlie surface. i\lagnctic particles a r c less liam~)ercti bv size ; i iz<l

shape than a r e eddy ciii.rents.

Because ihc rnngnctic l~asticlc incthoti excels all others Si,r ii~irl- ing \'el.!: siiallo\\' surface cr;icl<s I I I ferroniagiietic rnatci.i;iis, il IS

t/lc Oesi 111rtIlod to use fo r location o f Satipuc cracks in ]>arts n,:icir. from such inateri:ils. Fatigue cracks, \ollrcli occur 111 1,ai.t~ iiighl!. stressed and subject to stress variations in service, almost i i i \~ariabls s t a r t a t the surface aiiil propag;ite iii\sar(is t h ~ ~ o u g h tlie seeiion. If tliese fatigoe c ~ l c l ( s c:in be loc;ite(l eiirly in 1:lieir ])rogress, i t hi.e;iI<- Oo\~~ii can be avoitle(l ;uiil the $)art can often lhe salrnjirti. T:ecailst> of its tremendous alliiity in loc; i t in~ socli l i i i v sli;illo\\ s t i r f : ic~~ S;itigiie cracl<s, the r n a g ~ i e t ~ c 1h;u.ticle nietliod is used i i i i)i.eSo.eiicc to ; i l l

others for ove~,haul o r ~i~aiiiteiiiincc inspection 01' niachiiicr.~~ of nil kinds.

SO(:'RCES OF UEFISCTS

1. GENEI~,\L. Noniiestriictive testlng processes protlucc inilica- tioiis of conditions 111 niet;tls : ~ u ~ i other m;~ter i ;~Is hy indirect nie:ins. atnl such inr1ic:ations niost l,e i r~ te r j~ re tcd to riclcrnuuc the coii~li i io~i ~vliich is c;liislng thcnl. The problem ol'tlle tcsling cnglnecr is, in this respect, son ic t l i~~ ig like tlint of the pllysici:in i i i mctlic:~l di;ignosis. He must decitle coi.i.ectly f rom the s y n i ~ t ~ ~ m s wi1:it the flisease iconrlition) is that fits the symptoms. To rlo {his wit11 ct:~.tariily, thc piiysicl:un o r su~~geo i l most be Lhorc~ughly f:imili;ir wttll the l1om;lll body, wit11 its processes :ind wit11 the rliseases it is s u b ~ c c t to ; ;in11 lie must :iiso kiiow \vh;iL tlic symytonis :ire tliat these rliscases ~rotl i ice.

Siniilarly, tlie noo(lcstructi\~e testing engineer miist kno\xr the nature of tnet:tls (or. other m;~ter i ;~ls l ~ e may I)c called l i~or i to test) r lio\\~ they a r e made, ho\ts tliep a re \\.o~.keri into liii~slieti form, and the discontinuities tliat ma); be i~rorilice~I 1))' eac11 step 111 tlieir nlanu- I'acture. Such l ; i~o \~ le~ lge anli f i~mil iar i ty t i manuf;icturing processes g ~ v e s tlic ~ ~ ~ s w e c t o r , 111 a t I \ ~ i ~ n c ~ , i~iformation ~ e r y heipfui in alert ing Iiini a s to n;h:lt sor t of defects m;ly o r may not be present, and \%,here the!, a r e liki?ly to occor. As a n abd lo successful insuection by nondestruct~\*e testing means, there is IIO substitlite fo r tile kno\rlecige o i "\vli:tt to look for and wllere to look fo r i t " %Iapiletlc particle testing IS no excepttoti.

Tliis and tiif follo\ring cbaptei. a r c tlevoterl to the siihlect of rlelects in iron and stecl and iio\v tliey a re produced; kind to a discussioii o l how mctliis faii anti how such llefects contribute to failure. I t seems proper to present this informalion a s a p r e l i m i n a ~ y to tlic description of ho\v defects :ire actually found \sith ni:~gnetic partlcie testing.

2. S o s l ~ I)EFINITIONS. There a re several synonyms for the nrord "defect", ;inri 111 i ioodestr~icti \~e testing tlic~se \r.ords a r c not a t all ~nlerciiangeable. On the contrnry they lievc specific con11otations that must be recognized by those \vorlii~ig in this fieiil. Tlrese words a re "bleniisli", "flaw" aiirl "defect" Tile diction:uy definitions a r e :

Cli:,l~TU, 3 SOURCES O F I)ISFI<CTS -- - -

C l e m ~ s t ~ . Imuerfection. I-:lemi.ili suggests somethung superficial sucll a s :I l~lot o r a stain.

I*'iazu. .A small defect 111 eoiitinuity o r cohesion, such a s a crack or I)rml;-or a d e ~ a r t o r e f rom perfection.

Deieci . Tile i:~ck o r Iraiit inot always visible) of sometlung essential to completeness o r perfection.

In nondesti~uclrve testing these definitions actjutre some spccific meanmgs not cle:irly conveyed by t11c dietlonary 1angu:ige.

A par t may Rave a blemish o r a flam in it and still not be "dcfec- live" in tlie sense titat its usefuliiess fo r i ts intended purpose ~vill be iml~:~ired. A spot o r stain \xroulcl have no effect on the service performance of a connecting roil of a n alitomobile engine ailil \\,ouid certainly not make it rlcfectivc; but on a mirror, fo r example, such a spot niigfit malie the mwror unusral>le and would then unquestion- ably make the mirror defective.

3. \VHAT IS A DEFECT? In nondestructive testing ianguage the word "defect" is correctly applied oniy to a condition \rl11ch mill interfere with the safe o r satisfactory service of the pa~.ticular par t in question.

T o avoid confusion, \ve use a foiirtli mord-"discontinuity" This r t r~~rd covers the conelition before t i is determined whether i t is a defect or not. The cause of magnetic particle indications is 1n all cases a discontinuitj'-wliethcr pliysical o r magnetic. And if we exclude those discontinuities tha t a r e present by d e s ~ g n anri consider only those present in the metal by accident o r a s the result of some n~ani i factur ing process, these may still not in all cases m a l e the pa r t defective In the sense tliat its service performance mill be affected unfavor;ibIy.

We come, therefore, to the conclusion tha t a discoiitinuity is not necessarily a defect. I t 1s a defect oniy $\,hen it \\;ill interfere with the performance of the pa r t o r material in i ts intended service. And a discontinuity wlucli may make one par t defective may be entirely harmless in anotller pa r t des~gned fo r a differeilt service. Fur ther - a s in the case of the nilrror - ~ v l ~ a t \xrouid be only a blemish in most cases may become a defect in a product in xvhicl~ appearance is a major factor in acceptable service use.

So we should be careful to refer to a discontinuity a s a defect only when it nlakes the specific par t in ~ h l c h i t occurs un.suit;xble f o r the purpc~se for \\~iiicIi i t was desigi~ed and manufact,ured.

PRINCIP1,ES OF MAGNETIC PARTIC1,E TESTING

I t 1s difficult t o adllnre stl-ictlr t o th i s precept, but i t s sigiiificance shoiilrl i ~ c v e r be lost sigiit of'. "Discoiitinuity" is :I long 1vorc1 a t ~ d diflicult to say-\rIlrrcas "riefect" 1s sllort ;in11 easy. Even I!) th is book "rlc.i'ect" may be iiscil rn a loose sense oil rrcc;lsio!is when "(lis- continnity" t\,oulrl I)E tile correct tern?.

But the nondesti 'ncti\~e test ing inspector should always under- s ta i~r l tiiat a "discoi~tiiiuity is a defect only \\#hen it in ter feres w ~ t i i the service fo r \ \ , I~IcII tllc paif was inteniled"

4. ~\:IACNETIC DISCONTINUITIES. M a g n e t ~ c i~nr t ic le indications a r e produced by ~riag?zctic tliscontiuuities i n t h e metal which IS being ex;lnllnc(l. T l ~ e s e (Iiscoiltinuities may llot aiways be actual ijhysical breaks in thc continuity of the metal. hfagnetic tliscoi~tinuities may be produced by causes other tlian nct i~al breaks o r fla\iTs in the metai. Therefo1.e a magnetic particle indication does :lot necessarily show the presence of a defect. ~Vlngnctrc discontinuities may be caused by:

( a ) An actuai ??retnl/ic discontinuity o r r i~p t i i r e a t o r nea r t h e surface of a part , \vl~ich may have been present in the o~.iginal metal, or may have bccii nrotluced by si~bseqiient forming, heating o r iinlsliing pr.ocesses.

( b ) Actual lr!ernllic discoil t~nuit ics which are , Iio\\.evcr, present i)y design-as fo r example. a forced fit b e t w e t i two members of an assembly.

Fig. 16-Typncal Magnetic Particle Indication of Cracks.

72

C H A ~ 3

SOURCES OF DEFECTS

Fig. 17-Magnetic Particle lndication of a Forced Fit. a) White Light View. b) Fluorescent Particle Indication.

Ic) The junction between t\vo dissimilar ferromagnetic metals having diKerent permeabilities; 01. between a ferromagnetic metal and a non-magnetic material . Indications may be nrorlucecl at such a point even t h o u ~ h the joint 1s perfectly sound. Such a n indication may be produced in a i"siction o r flash weld of two dissimilar metals.

Fig. 18-Magnetic Particle Indication at the Weld Between a Soft and a Hard Steel Rod.

Id ) The junction between two ferromagnetic metals by means of non-magnetic bonding materials, a s in a brazed joint. An ~ndicat ion will be produced though the joiiit itself may be perfectly sound.

Fig. 19-Magnetic Particle Indication ot the Braze Line 01 a Brazed Tool Bit.

-̂.-. ,y<=

i e ) Segregation of tlie constituents of the metal, where these have different permeabilities-as, fo r example, low carbon

<

PRIKCIPLES OF MAGNETIC PARTICLE TESTING

Fig. 20-Magnetic Particle Indications ot Segregations

.-

z . Cl111PTl31 3

SoGIICl:.S 01: UEFEC'rS -~ ~.

areas in :I lrigh carbon steel, o r a r m s of ferri te, wliich is magnctic, in a m a t r i s of :;l:unless steel wll~cli is austenitic and therefore iron-r1ragnctic.

5. CLASSES OF DISCONTINUITIES. There a re ;I numLIer of possibie \rays of classirying discontitiililies tha t occur in ferroni:rgnctic materials :ind parts. One bi.oatl grouping wli~clr is useful in I'cne- t r a n t t c s t i~ ig is tli;~t based on location-whether surface o r sub- surface. (See Challter 20 fo r a full discussioii). Tliis grouping is also useful 111 magnetic particle testing, since the ability of tills method lo F i i ~ c l iricni1)ers of these two gruuns varies sIial.piy. But beyolid this, the cinssification 1s too broad to be very useful.

Another ~rossi'ulc spstein is to classify discontinuities by the processes \\411ci1 proituce them. Although such a system 1s too specific to bc! suitable fo r all pilrposes, it 1s use0 extensively. \I1e speak of Eorprng defects, welding defects, heat-treating cracks, gr~iidii ig cracks, etc. Pr:icticnllp every process, from tile ortginal production of' metal from its ore, do\vn to the last finishiirg operation: can :ind does introtluce discontinuities \vlricR magnetic particle testing can find. I t IS therefore important tha t the nondestructive testing englnccr or iiisj~cctor bc a\r;ire of all of these potential sources of defects.

6. CONVENTIOXAI. CI,ASSII?IC.%TI~N SYSTEM. For many years it has heeo custoni:iry t (1 classify ~liscontin~rit les f o r ilie purpose of nragnrtic particle testing ( a s tr~cll a s fo r other nondestructive test- inx nrcthocls) ;iccorrling to t h c ~ r source o r origin in the various st:lgcs of pro~luctioii of the metal, its fabrication and its use.

Broadly these a re rcfcrrecl to a s follows:

l a ) Inherelit-I'ro(luceil tluring solidification from the l iqu~d st:1tc.

Ib ) Process~ng-Pr~i i rar j~ .

Ic) Pro~e~~i~~g-Seco~i t l : i ry , or finisl,ti~g.

t d ) Ser\ricc.

7. ih'llE1IEN~ ~ ' ~ ~ S C O X ' I ' I N U I T I E S . TIIIS group of discoiitiiiuities IS

present r i i met:il a s tlie result of i ts initial solidification from the nloltcli state, bcforc : i r i j8 of tiri? nt)er:~t~ons to forge o r roll i t into ~iseitii S I Z ~ S arrtl siiapes iiauc Ibegtiii. The names of these inherent ~ I i scon t in~~i t i c s :ire gi \vn : ~ n d t l i e ~ r sources described belo\\,.

PRINCIPLES OF DlACNETlC PARTICLE TESTING

Fig. 21-Cross~section of Ingot Showlng Shr~nk Cavjp.

( a ) pipe. A s tlie molten steel wl~ ich has been poured into tlie ingot mold cools, i t solidifies first a t the bottom and walls of the mold. Soliclification progresses gradually upward and Inward. Tlie solidified metal occupies a somewhat smaller volunie than the liquitl, so tha t there 1s a progressive s h r ~ n k - age of \.oltinie a s solidification goes on. The last metal to solidify 1s a t tlie top of the mold, but due to shrinkage tliere is not enougli metal to fill the mold completely, ancl a depression o r cavity is formed. This niay extend quite deeply into the ingot. Sce Fig. 21. After early breakdown of the ingot into a bloom; t h ~ s s h r ~ n l i cavity 1s cut away o r cropped. If th ls is not done completely before final rolling o r f o r g ~ n g into shape, the unsooii~l metal xvill sliow up a s v o ~ d s calle[l " p i ~ e " in the f i ~ i ~ s i ~ e d product. Such internal discontinuities, o r pipe, a re obviously unllesirabie for most. uses anti con- stitilte it triie defect. Special devices ("hot tops") and special handling of the ~ n g o t rIu~.ing pouring anil sol idi fy~ng can, to 11 large tiegree, cont~.ol the formation of these shr ink cavities.

ClrAaar 3 SOURCES OF DEFECTS

(b) Rlo?oAoies. As the molten met:~l in thc ingot mold solidifics thcre 1s an evolution of various gases. Tliesc gas bubhles rrse througli the liquid and mxny escape. %fan)', however. a re trapped a s the metal freezes. Sonic, usually small, \rill ap- p m r near the surface of the ingot; and some, oitcn large, \vill be deeper in tile metal, csl~ecrally near tlie to11 of the ingot.

Nany of tliesc blowholes a r c clean o ~ i tile rnlerior and a re weldccl shut agaln into sound metal t i u r ~ n g tlie first rolling o r forging of the ingot; but some, near the surface, may have become oxidized and do not weld. These may appear a s seams in the rolled PI-otluct. Tliose deeper in the interlor, if not n~elded shut in the rolliiig, niay :lppear ns laminations.

(c) Scgrcgntio?~. Another action that takes place rlurrng the solir1ification of thc molten metal is the tendoicy fo r certain elements in the ~ n e t a l to concent~.atc in the last-to-solidify liqiiid, resulting in an uneven dis t r ibut~on of sonic of tlie chemical constituents of the steel a s bet\vecn the outside and the center of the ingot.

Various means have been developed to minlniize thls tendency, but if fo r any reason s e w r e segregation does occur, the difl'erence in permeability of the segregated :ireas may produce magnetlc particle inclications. Unless severe; such segregat~on is generally not deleter~ous.

t d ) Non-?rielaliic I?zel?~srons. All steei c o n t a ~ n s more o r less mat ter of a non-metallic nature. The orlgln of such matter is criicfly the rle-oxidiz~ng m a t e r ~ a l s added to the nioltcn stecl in tlie furnace, tile ladle o r the ingot mold. Tliese additions a re easiiy ox~dizable metals sucli a s aluminum, silicon, manganese and others. The oxldes and sulphldes of tlicse additions constitute the bulk of thc non-metallic iriclos~ons.

When finely div~ded anti oniformly distributed,.sucli non- metallic matter does not usually in jure tlie steel. I-Iowever, i t sometrnies gathers into large clumps \rhlcli, when rollcd out: become long "str~ngers." Tliese stringers in some cases a re objectionable. When sucli a s t r ing occurs a t the surface o r just under the surface of a highly stressed par t 01. a bearing surface, it may lead to fatigue cracking. 111 general, non-metallic inclusions in steel selriom constitute a rcal de-

I'RINCIPLES OF MAGNETIC P.AllT1CLE TESTING ___ .___.-___-_- -_

feet, tllougll tlley are olten indic:lted with magrietic gart iclr .~. xon.metallic inciusions a r e sometimes added t o steel 111-

telltlon:tlly. The atldition of ieacl o r sulphur i i ~ stecls f o r the pul.posc of lmprovllig then. machinability is coninion I~x ic - tice. suckl ill slion. escessirc amounts of non-meiallic lnclusloils, \v~ucli serve to break u p the cliips wlien t h e nietal

turl le( j u r otiier\\,lse m:ich~neil. hiacii ini~ig i ime is rerIuct?rl .,,d tnni iife is Ienptheiied.

Fig. 22- magnet^ Pari~cle lndficatson ot a Sub-surtace Stilngel of Non-Metallic Inclusions.

Such steels, if tested \vitIi magnetic particles, map sho\xr a larming Loolii~ig patterns, \\~hicIi have no signilicance as defects. The magnetic particle operator must be familiar with th i s type of steel. Thougii servlng a useful purpose in their proper field, LRese steels sliould never lbe used fo r critical o r higtiiy stressed parts , o r i>nrts subject to fa t igue in service . , ' -p v- -

ke) 1 / 1 1 i.'is,srrl,r+s. Recause of the stresses set u p in the Ingot as tlie resiilt of slirinliage dur ing cooling, lliternal rup tu res may occur \\4iicii may be quite 1:irge. Since no a l r norm;illy reaches the surfaces of tiifsc internal bursts, they

m;>v be coni~te te ly \\'l~~lrled stlut rliiring roll in^ o r other \rrork- log, :lnd 1c;ivc iiu rliscoiitiiiiiity. if Lliere 1s an opcliing f rom tiie f issure tv the surface; hon,i:ver, :11r will enter ant1 oxidize the surfaces. 111 such :I case \\.elding does not occur, alid they will renialn 111 llie finisl~cil prrrdocl a s discontinuities.

( f ) Scti11s. \','lien l ioui~l stecl is first pilurecl into tlie ingot mold there is consliterable spl:ishing o r spatterlr ig up anrl agains t tile cool \\,:ills of the mold. These spl:islies solidify a t once and become oxidized. As the nloiten steel rises and t h e mold becomes filled, these splaslies \\.ill be reabsorbcd to a large e s t cn t into tlie metal. Bo t in some cases thcy will remaln a s scabs of osiilized metai adlicring t o t h e siirface of the ingot. These map remaln and appear oil t h e surface of the rollcd product. If they d o not go deenly mto the surface they may not coiistilute a defect, since they may be r e m ~ \ , e d on macliining. F igure 23 illustrates th i s condition on a rolled bloom.

,Coaric.:i)- 1i.S. Stcl.1 cur,loii,l,,l,,l Fig. 23-Scabs on the Surtace of a Rolled Bloom.

( g ) lngoi. finclis. Surface crnciinig of ingots occurs clue to surface stresses generated dur ing cooli~ig of tile iiigot. Tliey may be either longiturtin;~l o r transvei-se, o r both types may occur together. As thc ingot is worked iiiio billets by rolling. these cl.:icks form l o ~ i g seams. Iiispection of qu:~lity-product billets f o r scams of th is type ivith rnagiictic particles is now

PRINCIPLES 01' RYAGNETIC P.4RTICI.E TESTING - -- --

common practice in modern mills. Tills permits removal of the seams by Hame scarfing, chipping o r gr inding ~ v i t h o u t waste of good metal. If not removed before fu r the r rolling these dceper seams appear , greatly elongated, on finished ba r s and shapes, often making them unsuitable f o r marly purposes.

Fig. 24--Seam on a Bar Shown by Magnetic Particles.

a. PRIMARY PROCESSING DISCONTINUITIES. When steel ingots a r e worked down into usable s ~ z e s a n d shapes such a s billets and f o r g ~ n g blanks, some of tlie above described inherent defects may appear . Bu t the rolling and fo rg ing ixocesses mag themselves introduce discontinuities \vhich in mimy cases coiistitute (lcfects. Pi.imary processes as iiere coi~s~dere i i , a r e those \\'liich work the metal do\\:n, by either hat o r colil deformation, into useful fo rms such a s bars, rod ancl wire, forged shapes. Casting is another process usually includcri in th is group since, though it s t a r t s with molten metal, it results in a semi-tin~slietl product. Welding is . s~miiar ly includeti fo r similar reasons.

A ilescriptiou of tlie d isconl inu~t les ~ii troduccd by these primary

processes Sollows : (a ) Senrlls. Seams In rolletl bars o r ilrawil wire a r e usually

lllghlp objcctionable, aiid often do\ \s~i -gradc the product and makc i t unusable fo r iirst qoality purposes. As h a s just been

80

Cll.\rT1.21 3

SOlIRCES 01.' DEFECTS ~

.~~ ~~

described, severe seams may oi'igiri;~te fi.<lni ~ i igo t ci.;icl(s. but by proper clc:inlng u p (c i~i i r i i t i~~i l ing) of the billets l:y scilrfiiig, grin!lillg or chipping thesr! can be rlinilii:rterl before final rolling. Cunrlilio~uiig is I I ~ \ Y ~ist~ii l ly iiitle~l ihy use of magnetic particle test ing to ii~llicnie llie leiigtli :i11d se\,erity oC tile seams. See Cl~ai:tr~r 19. IS propei.ly c o i ~ ~ l i t i ~ ~ i i e ~ l ;it the billel stage, seams from llrrs source ner+ll not appezn. in tlie final rolled product.

Gut seams can be ii~tro~liicerl l ~ y tlic rolling o r rlr;kwrng processes themselves. Laps c;i~i occur I I I the rolling of the Ingot illto billets as the result of over-filling of tlie rolls. Tlits l>rodiices pl.ojecting l i i~s, \\>li~ch on subsequent passes a r e 1.oilcrl into the surface of the biilet 01. bar. \\'lien severe, tile billet often cannot be salvagetl axid is downgrailed.

C C ~ # , , , ~ , < ~ > ,,.s, sL<.et ~ : ~ I C , , S , ~ , ~ ,,,,, 1 Fag. ZS-Surtace ot a Steel Blliet Sl,ow~ng a Lap.

Siniilnrl\r, ctScli Iviien i~il lets \\~hlcli lii\,e ileen coi~dit ioile~i, anti itre free o l sc:lnls. n1.e rollvrl I I I ~ I I I~ill-s o r 1.011. I:I]IS resulting fruni over-fillo(l rolls can #rcciii., proililcing long ;lnrl oftt!i~ I~<!~J , <lee{> se:iiiis 111 the Ii11ishe11 ~ ) r o < l ~ ~ c t . 111 siiliilar ~ : ISII I<II I , iintl~>r-iills i i i i.11~. rolliiij: j>i.r~orss nl:ty, nil siil~sr~iiicirt )):rsscs ill! S~liIl?FX't~ f<ll'iil ;i S ~ ~ : l i l I , \ ~ ! l l ~ l l t)l%!li l~li l lS 1 1 1 ~ f l l l l lellgt~l of' l he II:I~. Se;inw ilvri\~r!l f l . o ~ ~ i I ; I~ IS noill iisl~iilly emerge lil

tlic sur lacc of tlic ba r :it ail i ici~le ; i i i~ Ic . Sc;~nis ilci.~vr'rl rrglm the fo i~ls !:ro1li1~6?<! lhy :II I t ~ ~ i ~ l ~ ~ i ~ - l i l l ~ ~ ~ l [ii~ss :ire likt,iy io I I < % niore iie:irly nt)i,ni;il 111 l,l~e s i i i ~ I ' ~ ~ c ~ ~ of tliv 11;lr. S ~ i ~ n i s or $lie

PRISCIPLES OP RIA(;KETIC PAIITICLE TESTING

marks may also be int~orlilcecl in tile ~ir:i\rliig nrocess dilv to defective (lies. Sucll seams may or map llot make the protluct defective. For some ]lUrpOSeS, such ;IS snrlligs o r b;lrs fur 11ea\~y upsetting, the most mmute surface ~mperfect ions ( o r discontinuities) a re cause fo r re~ect ion. For others, \vhere fo r example m;lch~nlng operations a r e expecleil to remove the outer layers of metai, seams \\~ii~cli a r e not too deep \\.ill be machined ;1\iaay.

I

Fig. 26-How Laps and Seams are Produced by the Rolls. 0ver.fills and Under-f8lIs.

( b ) Lumz?~nlioss. 1,aminations In rolletl plate o r s t r ip a r e formc[i when blowholes o r internal fissures a r e not \\,elded t ight dur ing rolling, bot a r e enlarger1 2nd flattened ~ n t o sometimes quite Large areas of liorizonlal d iscont~nurt~es . Lam~na t ions may be detectetl by rn;ijinetic p ; ~ r t ~ c l e testing on t h e cut edges of pl;itc, but (lo not give indications on 11l:lte o r strip surfaces, srnce these iIisco~itin?~ities 11re lnterlial and lie in ;I plane parallel to tile sorfaee. Ultr:~soii~c m;lr,pirlg technlgiies a r e used to define them.

Clmn'in :; SOIIILCES (IF 1)EI"IICTS

Fig. 27-Magnet~c Partacle Indications 01 Lammations Shown on Flame Cut Edge ot T h c k Steel Plate.

( c ) i Tlils IS a condition created when, in dr :~iv~ng or cxtrudir~g a bar o r shape, the m t e r ~ o r of the metai does not flo\v a s ;:ip~dly a s the surface. Segregation In the ccnter of the ba r usually contributes to the occiii,rence. Thc result IS a series of internal ruptures \\3ili~Ii a re se\rel.e defcct .~ \\~i.hcne\,er they occur. They may be indicated with magnetic particles, but only if the ruptures a r e large and approach tile surface of the bar. The cupplng problem can he minlm~zed by changlng die angles.

Id) Cooi i?~g C?.nclis. When alloy ant1 tool steel bars :,re rolled and subsequently run out onto a bed o r table for cooling. stresses may be set up due to uneven cooling \\,l~ich can be severe enough to crack the bars. Such cracks a re g e ~ ~ e r a l l y longiludinni, bu t not necessarily straight. 'l'hcy may he quite long, and usually vary in deptli nlong thelr length. Figure 29 shows the magnetic particle indications of sucir a crack. and also sections through the crack a t three points where the depth of the crack 1s different. The magnetic p:irt~clc indication \r:iries In intensity, being heavier nt points wllere the crack IS deepest:.

Ctml-m? 3 PRINCIPLES O F lMAGNETIC PARTICLE TESTING

- - -- %

SOURCES O F DEFECTS -

- Fig. 2%-a) Fluorescent Magnetic Particle Indications of Severe Cuppnng in Drawn

Sprlne Stock, Ground Into the Ruptures a t One End. b) Section Through Severe Cupping in 1%-tnch Bar.

29-Magnetlc Particle lndrcat~ons of Cooling Cracks in an AIIOY steel B~~ a) Surlace lndlcat~ons b) Cross Sectaons showing Depth

Fig. 30-Magnetlc Parttcle Indications 01 Flakes in the Bare of a Large Hollow Shatt.

( c ) Flakes. Flakes a r e internal rup tu res tha t may occur In steel a s the result of too rapid cooling. I t 1s believed tha t the re- lease of dissolved l ly~lrogen gas dur ing the cooling process causes these ruptures, and tliat controlled slow cooling a f t e r forglng o r o t h e r w ~ s e hot-\rrorktng the metal will reduce t h e ~ r occurancc. Flak111g usually occurs in f a ~ r l y lieavy sections and certaln alloys a r e more susceptible than others. I'' ' tgure 30 shonrs magnetic particle indications of flakes w h ~ c h have been exposer1 011 a niach~netl surface. Sincc these ~ w p t u r e s a r e deep In the metal-usually half way and more from the surface to thc center o r the section-they will not bc shown by magnctlc particle test ing on the ol.lg!n:ll surface of the part .

( f ) F o ~ g z ~ t g ~ . I L ? . s ~ s . When steel 8s \vorlied a t improper tcmpcra- tiires i t 1s subject to cracking o r ~ u p l u r i r l g . Too rap111 or too severe a reduction of section can also c;iusc blirsts o r cracks. Sucli rilpt11l.e~ may be 1ntern;ll l)ursLs, or thcv may bc cracks

PRINCIPLES O F RIACNETIC PAIITIC1.E TESTING

Fig. 31-Magnetic Particle lndicatmns ot Forging Cracks or Burst5 In an Upset Section. Severe Case.

on the surface. When on the surface they :lye rcadily found by magnctic {,article testing. If interior, they a r e usually not shown except \\.hen they ilave been exposed by machinrng.

cg) Fovgtng I,a?,s. As the name implies, f o r g ~ n g laps o r folds a r c formed when, in the forging operation, improper hati- dling of the blank in the die causes the metal to flow so a s to form ;I lap whicli is later squeezeil tight. Since it is on the surface and is oxidized, th is lap docs not weld shut. This

Flg. 32-Cross Section 01 a Forg~ng Lap. Magnified l O O X

86

CHAI'TW :: SCIUIICES OF DEFECTS - - -- - - -

type of ilisconiinoity is s o n ~ e t ~ i n c s diliiclilt to locate, bFcause if. niay be uixn a t tile sllrfacc anrl fairly shallo!~, mld often nl;ijr lie ;it only a very sliglit angle Lo the su~.l"ace. I11 some unusu;tl cases it ;llso may be solidly Riled uvith magnetic I ~ s I ~ ~ ~ s .

I ) I Overlieatiiig of f ~ r g ~ i i g s , Lo the !,oi!it <IS rncipient f u s ~ o n , results ill a coii~li t io~i 1~1ucf1 renders tlir i o r g ~ n g lin- usable in most cases, aiiil 1s referred to as borliing. Actual oxirlation is ho\t'ever nut the real source of the damage, but rntlicr the p;lrti;ll liqu~datioii duc to he;it, of matfrial a t the grain botind;uics. liuriiiiig is 2% serious defect but is not goi- erally shown by magnetic particle testing.

( i ) Flnsir i,l?ic 'I'cnrs. Cracks o r tears along the flash line of forgmgs a r e usunlly caused by improper tr imming of the flash. If sliallo\v, they may "clcan up" if mach~ned, and do not make the par1 defective; or, they nlay i)e too deep to cican up and in such cases the forging cannot be s:livaged. Such craclis o r tears can easily he found by magnetic particies.

Fig. 33-Magnettc Particle Indication ot Flash Line Tear in an Automotlve Spindle Foig~ng. Partsally Machined.

87

I'R1NCIPLES OF MAGh'lSTIC PAILTICLE TESTING

(j) Cnstings. Steel and iron castings a r e subject to a number of defects rvhich magnetic particle testing can easily detect. Surface discontinuities are formed in castings due to stresses resulting iron1 cooling, and arc often associated with changes in the cross section of tlie ]]art. These may be hot tears or they may be shrinkage craclts wluch occur a s the metal coois dolvn. Sand from the molti, ti-appecl by the hot metal, may form sand inclusions on o r near the surface of castings. Gray

Fig. 3-l-Magneilc Particle Indications ot Detects ~n Castings. ar Surtace lndicat~ons ot an lrltcrnal Sl,rlnk Cavlly. b) Handling Crack ~n Gray Iron Casttng.

CHAPTER 3 SOURCES OF DEFECTS --

iron castings may be quite brittle, and are often cracked- usually a t thin sections-dur~ng the "sl~ake-out" o r by rouglr handling during sorting.

(k) I.i'cld?lae??ts. A number of kmds of discontinuities may be formed during welding of both thin and heavy sections. Some a r e a t the surface and some are in the interior of the metal. Some of the defects peculiar to weldments are lack of pene- t r a t~on , lack of fusion, undercutting, cracks in tlie weld metal, crater craclts, craclts in the heat affccted zone, etc. These defects and their detection mill be discussed in detail, along with castings, in Chapter 24.

9. SECONDARY PROCESSING OR FINISHING DISCONTINUITIES. In this group are those discontinuities associated with the various finishing operations, after the par t has been rough-iormed by rolling, forging, casting or welding. Discontinuities may be introduceti by machining, heat treating, grinding and similar processes. These a r e described below :

l a ) Maci~mtxg Tears. These a r e caused by dragging of the metal under the tool when it is not cutting- cleanly. Soft and ductile low carbon steels are more susceptible to this kind of damage than a r e the harder, higher carbon or alloy types. Machining tears a r e surfacc discontinuitics and are readily found with magiletic particles.

(b ) Heal Treating C~aclis. When steels are heated and quenched to hardell them, 'or arc otherwise heat treated to produce des~red properties for strength or wear, cracking may occur if the operation is not correctly suited to the material and the shape of the part. Most common a r e quench cracks, caused \vlicii par ts a r c heated to liigh temperatures and then suddenly cooled by immersing them in some cool medium, \\~hich may be water, oil or cven air. Such craclts often occur a t locations where the par t changes section- from light to heavy-or a t fillets or notches in the part . The edges of keyways aiid tlie roots of splines or thrcads a r e likely spots to n.atc11 for quenching cracks. Craclts may also result from too raptd heating of the part, wbich may cause uneven expansion a t changes of cross-section, or a t corners \\There heat 1s absorbed-from tliree s~des-more raoidly than in the body of the piece. Corner cracking may also occur durtng quenching, hecause of more rapid heat loss a t such 1ocati1)ns.

PRINCII'LES OF DIAGNETIC PARTICLE TESTING

Fig. 3SMagnetgc Partlcle lndicat~ons ot Quenchtng Cracks. Shown with Dry Powder.

1Ie:lt treating cycles can be desrgned to ~ n t n ~ m l z e o r elimt- nate such c rack~ng , but fo r critical parts, testing with magnetic pitrticies is a safety measure usually applied, since such cracks ;ire scnous and thetr detection presents no difiiculty.

Ic) SL.ratghlolzng C ~ n c l i s . The process of heat t r e a t ~ n g often c:luses some w:trplng of the pzr t due to slight unevenness In Llir cooling diirtng qucnchlng. A hardened shaft , fo r example, may come frum the heat tl.e:~t operatiot~ 11ot quite stratgilt. 111 many cases thrse can be s t ra tgi~tened in a press, but if the amount of bend required IS too great, o r if the shaf t 1s very i ~ a r d , craclts may be formed. These, again, a r e very reaclily found with n~ :~gne t i c particles.

t d ) G~.til(/i?~g CI.RC/CS. Surface cracktng o i llartlencd par t s a s tile result of improper grtnding 1s frequently a source of trouble. Grinding cracks a re essentially thermal craelcs and a re related to quenclut~p cracks in more \xrays than one. The>, a re caused by stresses se t up by loc;ii heating utrcler the g r ~ n d i t ~ g wheel. TRej' a re in nearly all cases avorrlable if proper \\2heeis, proper cuts altd proper coolants a r e used, and if heels a r e properly dressed ailen en r cqu~red . Gut stnee proper g r ~ r ~ d i n g requires constant attention and care \vhich IS not always provlrlecl 111 practice, thesf dei'ects do occur. Since tiley a r e s h a r p sur~l':ice craclts they a r c easily locatetl with magnetlc

CHAPTER 3 SOUIlCES OF DEFECTS

~larticles, even if shalloxv, and locattrrp them 1s usu:ril~- of vltal lmnortance. Salvage of the cracked par t 1s seldom possil~le since gl.lntling is usually a final preclsron fir~ishrnp operation. Thc best use of maglretic p:irtlcle lestiilg in this ease IS to conclucl san~pi inp tests to m o t ~ ~ t o r thc grincling oprra t io t~, atld thereby conhol i t to avord the formation of grinding cracks.

Fig. 36-Fluorescent Magnetic Particle Indications ot Typical Grlndlng Cracks.

I-l:trdened surfaces often r c t a ~ n inte~.ttal stresses f rom tiif qlleilching opeletion whtch a r e not severe enoirgh to c:~use cracklng a t tile tune of nuenehrng. D o r ~ n g prlnding however, tire relali\~cly small increment of stress se t ~ t p by local he:iting u~irler gr~ndir lg milcel may c;lusc rupture \\,ilcn addell to the I.csi(lual stress alrc:~dy pl.esent. Such suriaces ~rso:iliy cr:iclt se\'erely arlrl e x t e ~ i s ~ ~ e I y , a s illustr;~tcd II I Fig. 37.

i e ) Elclit~r!) n i~r l . P i r ~ l i l i ~ ~ ~ j Ct,!~cks. I-lartlened surfaces \ \~h~ci l corr- tarn t~e~ l i lua i stt.esscs may be cracked i f l i ~ c g :ire prcl(led o r etchetl 111 acld. Attacli by tlre acid of thc su r fac r fibres of the metal grves tile 1nter.nai stress a c l r i~t~ce to b(! relie\'ed by tile form:ition of a cr:ick. Bef0r.e th is action \\,as fully u~rrler- stoorl, the heal tre:itmenl of the par t \v;ls often blnmerl for the crncklng, \ \r l~en sucir cracl<inp acl~rally occirrretl rlurlng actcl clcalii~rg S!II. pl:~ting o r fo r otllcr purposes. The hcat t r ea t operatioll did, i~o\ \~eucr , deser\'e some of the! blame, by le:ivrnp tile p : ~ r t nsttIi locicetl 11p I I I L V ~ I I ~ I I I . I ~ S I ~ I I I : I I sLr(~sses.


Fig. 37-Magnetic Particle Indications ot Granding Cracks in a Stress.Sensitive, Hardened Surface.

Fig. 38- -~agneGc Particle Indications of Plating Cracks.

92

CHAPTEE 3 SOURCES OF DEFECTS

( f ) Plating Cracks. When hardened surfaces a r e to be electro- piated, care must be taiien to ensure that nickling (o r other cleanlng operations preparatory to plating) does not produce cracks. Sometimes cracks are formed during the plating operation itself.

Residual stresses leading to etching or plating cracks may aiso be the result of cold work. Sniral sprlngs, cold wound, then plcliled for plating or hot galvanizing, have also shown such craclis. The hot galvanizing process itself may also produce cracks in s~irfaces contain~ng residual stresses. Penetration of the hot zlnc betwcen the grain boundaries d u r ~ n g the 1lot dip process provides points for rclief of such stresses by tlie formation of cracks. Copper penetration during brazing may resu l t in similar craclcing if the par ts contain residual stresses. Molten alloys from the bearing of a railroad axle journal durlng a "hot box" will penetrate the surface of the heated journal and provide tlie starting point fo r a fatigue crack and axie failure.

lo. SERVICE CRACI(S. The fourth major classification of discontinuities comprises those which are formed or produced af ter ail fabrication has been completed and the par t has gone lnto service. The objective of magnetic particle testing to locate and eliminate discontinuities d u r ~ n g fabrication, is to put the par t into service free from defects. However, even when this is fully accomplished,

Fig. 39-Magnetic Particle Indication of a Typtcal Fatigue Cracl:

03

I'RINCIPLES OF MAGNETIC I'ARTICLE TESTING A-

failures in service st ill occur a s a resillt of craclting caused by service conditions.

t a ) Fatigl,e L.rae/is. Tlie subject of f;~tigiie cracking \\'ill be thorougiily discussetl in t.lie folloli~ing Chapter, 4. As a source of discnntinuities, tile phcnomciion of fatigue is a prolific oils. RIctals wluch a re sui~jccted to alternating o r fluctuating stresses al,ovc a certain critic:ri level ( t h e fatigiie s l r c n g t l ~ ) usill eventually develop cracks, and finally fracture. Fntigtie craciis, even very siiallo\v ones, can rearlily he found with magnetic particics, and tlie p a i t ofteu can be salvaged.

( b ) Co~~.osio,z. Pa r t s whicl~ a re under tensinn stress in service and a r e a t tlie s:~me time exposed to C I ) ~ ~ O S I O I I from 1\~11;11e\~er cause, niay develop cracks a t the surface, referred to a s s t ress coyrosion crncliing. Sucli cl':~cks, under continuing corrosion rind stress i\vIicther reversllig, o r fluctuating o r not) will progrcs.5 tlirough the sectioii until failure occurs. When corrosion is adiletl to fiitiguc-pro~lucmg scr \~ice conditions, th is type of service fai1ui.e is c;kIled corrosioii fatigiie.

( c ) O~~!el~st?.c,s.sin{]. P a r t s of :in assembly 111 service tha t a r c slrcssetl i~eyoiirl the level for mliicll they were rlcslgneti a re \#cry likely to crack o r brcali. Such ovcr-stressiiig may occur 21s Llie result of a n :icc16lent.; o r a pa r t may become over-

Fig. 40-Fluorescent Magnetic Paiticle indications of Cracks in Crankshaft of Small Aircraft Engcne. Damaged hn Plane Accident.

94

CIJ.II,T~ 3 SOOIlCES OF DEFECTS

loadt~d due ti] some unus~ i :~ l o r emergelicy condition i ~ o t an- ticipated by tile clc'sipncr: o r :I pa r t may be loacted beyond its streiigth because 01 the failure of some rel:~tcri member of the s t r ~ ~ c t i i r e .

After complete failure Iias occurred magnetic particle testing ol~viously has 110 application a s reg;irds the fractured parl . But othcr par ts of the assembly \i~liich may aDpear ~in(l:uu:~getl may ii;ive becn overstressed dur ing the accident o r nverloadinp from other causes. Examination by magnetic particle t e s t i~ ig is iisiially carried out in sucli cases to determine ~vhe ther any crnclts have actually been forined. With this prec;~ulion salvagc of good par ts a f t e r the assembly Failure is oftfri possible.

1 OTHER SOURCES OF I~EFECTS. In tills cliapter a n :~ t t empt has been made to fami1i;irize the reader with most of the common sources of defects I discontinuities) \\'hich occur in iron and steel. Actiially the list a s given here is nowhere ne;ir complete. But the i n s ~ ~ e c t o r w o r k ~ n g \vith magnetic particle testing will eiicounter these discontinuities which have been described more frequentl!, than those from less common or more obscure contiitioi~s. Ne will often liave the n~etalliirgicai 1;iboratory of the plant available fo r consultation, and the met;~llurgist \itill usually be able to assign a cause to an indicated discontinuity :ind assess i ts ininnrtancc.

CHAPTER 4

HOW AND WHY METALS FAIL

1. GENERAL. One of the principal reasons for tlie existence of nondestructive testing is the need to find and eliminate defects in par ts o r assemblies so as to prevent their failure af ter they have been iabricatecl and put into service. In order to apply these methods effectively, the ilondestructive testing engineer or the inspector should know ho\v and wily metals fail. Those must accept and reject parts that show magnetic particie iiidications of discontinuities ( o r indications by any other nondestructive testing method) would be handicapped if they did not know tlie metallic weaknesses tha t can lead to the several types of metai failure.

This knowledge is essential to an accurate appraisal of the effect which the detected condition is likeiy to have on the performance of the part in service. Two other factors a r c equally essential-an accurate knowledge of what the service requirements are going to be, and a n accurate interpretation of the actual condition present, which the nonilestructive test had indicated.

2. METAL FAILURE. Ever since metals \\,ere discovered and useful implenients made from them they have been subject to failure. Metal axes broke under impact, swords broke off a t the hilt in combat, and when metal axles were used for wheels instead of wood, they too often broke under heavy loads. Naturally such failures x\~ould ;il\\,ays occult a t the most inconvenient. times-that is, when a little extra load was being imposed during some critical operation or service incident. A sword that broke in battle was likeiy to result in serious consequences for the wielder!

I t has been the problem of the ages to t ry to make metai par ts that ~vould not break in service; and in every age, the advances in design have been pusliing the frontiers of knowledge for this purpose.

3. EARLY ATTEMPTS TO AVOID FAII~URE. The natural and ObviouS remedy, \\,hen a par t broke, was to conclude that the par t nuist be weal< and must be "beefed up"-made bigger and therefore ( i t was assumed) stronger. Someone then invented tlle term "safety factor"

96

C i i A P r n 4 HOW AND WHY METALS FAIL

.-

-more recently called with some justification the "factor of igno- rance"-to describe thls beefing-up process.

Iinowledge of the distribution of the stresses xvithin a part , both dynamic and static, was not very extensrve o r even accurate. The means for s t udy~ng the effects xvhicli stress distribution had on the service performance of a par t had not pet been devised. Today \ve have several highly accurate nietliods of experimental stress analysis, and these have helped immeasurably in studies tha t have iead to an understanding of metai failure. The most important bit of knowiedge commg out of these studies was the realization that "bigger" is not necessarily "stronger".

Today the factor of safety has acquired a new significance. The designer can now use a much higher proportion of the ultimate strength of metals with assurance, since by using the methods of stress analysis, he can know in advance what and where the niaxlmum stresses in the pa r t will be. In his design he can then avoid concentrations above those tha t his seiected f ic to r of safety %vill permit. Today, factors of less than two a r e often employed instead of factors of 3, 4 and 5 which were formerly the rule. Even factors of less than one have been used when load cycles a r e very few in number. Furthernlore, this low safety allo\\rance is as much a cushion against the possibility of unusual over-stresses in the service life of the structure or part , a s i t is an allowance for unknown metallic defects that may occur. This is because nondestrtictive testing methods, if intelligently applied and properly mterpreted can almost ciitirely eliminate the danger of defective components being allowed to go into service.

4. METALLURGY. Metallurgy in the early days was in the ficld of the Alchemist, and his rituals and strange formulae for maklng metals strong were shrouded in mystery. But the early fabricator of metals certainly liad learned a thing or two about metallurgy, even if he did not l<noiv what he was dorng. Witness the fine steel that was forged into tlie weapons and armor of the middle ages- heat treated, hardened and tempered to exactly the proper point.

Tlie addition of the metllods of metallography ( the study 01 tlie structure of metals with the aid of the microscope) srxty or more years ago ,gave metallurgists the tool they needed t o iearn about and undcrstand the Dart played by metallic microstructure in the strength and working characteristics of metals. This technique was

97

CliAmn? 4

- AOW AND WHY DlEThLS FAIL

PRINCIPLES OF YAGNFTIC PAR'TlT1,E TESTING -- -

first brought to a high state of usr?fulness 113' Dr. Albert Sativeur, working a t Harvard University, about 1910. RIany, if not most, of the famous names appearing in the story of the intciisive expansion of our knowledge of metnis w l ~ ~ c h has occurred during the succeed- Ing years, are those of men who studied under Dr. Sanreur.

5. STRENGTH VERSUS FAILURE I N NETALS. Studies of the internal construetio~i of steels and other metals quickiy brought in-

derstanding of the basic factors iiivolved in strengtli or weakness, ductility o r brittleness, softness o r hardness. Atid with t l i ~ s undcr- standing, a new approach to the probiem of how-and !%ul~y-n~etais fail becaine possibie. I t was quiclcly realizeti that strength under dynamic loading was quite different from strengtli under staiic conditions. Discontinuities o r other flaws t1i:tt mere apparently harmless in static service were found to be major factors leading to failure under servlce conditions that involved stress v:rr~atlons. Merely making metals stronger by heat treatment or by addition of alloying clen~eiits did not avoid service fa i lu~es , and in fact often had the opposite effect.

Toclay, the study of the various modes of failure and thc conditions leading to their occurrence has given us the understanding tha t 1s needed to des~gn parts that will give failnve-free performance in service. Thls result is not only possible, but realizabie, provided the importance of tlie service mar r a~ i t s lllc cost il~volved to accomplisl~ it. Required is the choice of the riglit rnctal or alloy and i ts proper heat treatment for the expected service conditions; complete stress analysis studies to ~nal ie certain that no undue stress concentrations occur in any par t if unexpected loatis occur: aild full use of the methods of nondestructive testing to give assurance that each par t 1s free of defects which would introduce weaknesses and stress conceiilrations not allowed for by carefui design.

6. I-low MET,ALS FAIL. Two major modes of fracture of steel a r e recognized. Tliese are :

(a lC&a~,ageypes , usually called ibrittle fl.;tctul.es, i n whicli failure occur~&it l~out plastic Ao~v.

(b ) Slrear types, usually referred to a s ductile fractures, in \t,hich plastic flow occurs before fractllre.

In both these types of fr;icture the failure line c t ~ t s across grain bounclarics-that IS. follo\%'s sliu olanes within the crrains of the , . - metal. I-Iowever, these slip planes are on rlift'erent axes of the crystal in tlie cases of the tnro types of fracttrre.

98

Fig. 41-Examples of Silky and Crystalline Fractures a) Ductile or Silky. b) Brittle or Cleavage.

9Y

PRINCIPLES O F MAGNETIC PARTICLE TESTING

The appearance of the surface of the fracture in these two modes is entirely different. The fracture surface of a shear or ductile fracture is fine-grained o r "silky", whereas in the case of cleavage fractures the surface is crystalline in appearance. Fracture by propagation along grain boundaries also occurs but is less common and involves rather special conditions. (See Section 23 and 24, this Chapter.)

7. CONDITIONS LEADING TO FAILURE. The follow~ng list enumer- ates the most common of the recognized causes of failure in iron and steel :

( a ) Simple over-stressmg, due to overloading. (b ) Impact, induced by sudden and rapid stress rise. ( c ) Fatigue, induced by stress variations below the eiastic limit. i d ) Corros~on, induced by corrosion of stressed metal. te) Creep, induced by proionged exposure to high temperature

under static loading.

These will be discussed in the follow~ng sections.

8. OVERSTRESSING. I t is self-evident that ~f steel is stressed beyond i ts strength i t will pull apar t and break. Overloads sufficient to produce this result may occur in service from several causes.

Unusual conditions-such a s an aircraf t suddenly encountering excessively violent clear a i r turbulance--can induce stresses which are severe enough to pull the plane's structural members apart. Or, one member of an assembly may break; due t o fatigue or some other cause, and throw additional load onto other members, creating stress too high for the latter to withstand. Sudden stoppage of a vehicle, as in an accident, may also overstress some member of the frame o r the engine beyond its strength, and cause parts to break.

Failures caused by overstressing a r e usually of the ductile type, although brittle fractures can also occur from simple over-stressing in the case of very hard metals, or a t low temperatures.

Tlie point a t w h ~ c h the actual fracture originates niay be determined by the presence of some local stress-raiser. such a s a niclr or a scratch in the surface of the part, even though this condition is not the primary cause of the failure.

9. I~IPACT. Impact is actually another form of over-stressiiig. Tlie difference is principally in the rate a t whicli the stress is

C l i A r n 4

IfO\\' AND \VSfY METALS FAIL

applied. -4 sudden blow or impact can impose a very high stress at an exceedingly rapid rate, and failure occurs in the brittle mode. Such a brittle fracture may occur even in norninily ductile nlaterlals, if the ra te of rise of stress is fast enough.

10. EFFECT OF TEMPERATURE. Brittle fractures a t ordinary temperatures are most likely to occur only in hard mater~a ls which normally break with little 01. no plastic flow preceding fracture. However, temperature nas a great deal to do ~ v i t h thc niode of fracture of nearly all steels and alloys.

Fig. 42-Brittle Fracture Which Occurred in a Tank Holding Gas at Low Temperature.

At low temperatures soft steels wiiich are normally ductile will fail in the brittle mode, even without fast-rising stress application. And a t certain elevated temperatures, such steels may become "hot sliort" and fail in the brittle mode on over-stressing. 111 both cases, however, failure is more apt to occur \\,hen the stress is applied rapidly, as by impact.

Tlie composition of steels and alloys has a great influence on their behavior under low temperature loading conditions. The degree of "notch sensitivity" of the steel is a major factor. Notch sensitivity is definer1 a s "a measure of the reduction in strength of a metal caused by the presence of stress concentration."" Steels hav~i ig a low notch sensitivity will witlistand conditions favorabie for brittle failure better than those having high notch sensitivity. Alloys - 'A.S.M. Aletsls Handbook. 8th Edit*un. \'ol. I , Page 26.


normally !n the austenitic state a t ordinary temperatures or temperatures helow norm:ti-stainless steels of many types a r e austenitic-have a very low notch sensltlvity and are not a s likely to fail in the brittle mode, even a t very ion, temperatures, a s a r e ordinary carbon or alloy steels.

11. FATIGUE FAILURES. In the early efforts to understand why metals broke in servlce, fatigue failures were very puzzling. I t was observed that a par t in service when subjected to reversing or \vldely varying stresses often broke suddenly, even though these stresses were \veil below the elastic limit of the metal. Examination of the surface of the fracture showed a portion of i t to be smooth, with peculiar concentric marklngs, while the remainder was crystalline. This immediately led to the assumption t ha t when metal became "tired" af ter long serv~ce, i t "crystallized" and then broke with a brittle type fracture.

After metallographic methods were employed, studies under the microscope showed tha t there \<,as no readily discernable difference between the structure of the metal a t the break and of metal some distance away. And it was not long before the t rue explanation of this partly crystalline fracture was arrlved at-that IS, the

Fig. 43-face of a Typlcal Fatigue Crack.

CHAPTCR 4

H O V AND W H Y XETlLS FAIL --

fatigue crack first started and spread very slov:ly. During this p r l o d tlie faces of the crack "worked" against each other to produce the smooth par t of the fracture. Progress of the crack came tn small increments, each advance leaving its circular mark to glve the "oyster shell" appearance to the surface. When the remaining sound par t of the section finally broke i t broke suddenly, producing the crystalline portion.

12. STRESS-RAISERS. Examanation of the smooth portion of the fracture face almost always sho~ved these rlng-like marklngs, ~ v l i ~ c h appeared to have a common center a t some point a t the orlginai surface of the part. This point was cleariy the point a t which the fatigue crack actually started its lusually) slow progress across the section. I t was fur ther observable tliat this. polnt of orrgln could in most cases be assoc~ated with some small surface defect: such a s a nick, scratch or tool mark.

Experimental stress analysis methods showed that the tenslon stress across such a surface mark could be several times greater than the nominal average stress over the surface. Even though the defect was small, this multiplication of the local stress ievel across the defect was h ~ g i i enough to s t a r t the crack. F o r example, a s a loaded railroad car axle rotates, the surface stress 1s reversed with each turn, from highly positive to hlghly negative. Presence of a nick o r circuinferentiai scratch will furnish the stress concentration necessary to s t a r t a f a t ~ g u e crack. The stress reversals a s the axle rotates causes the faces of a startlng fatigue crack to work agalnst each other a s the crach- advances. Eventually the remaining sound portion of the axle breaks.

Fig. 44-Stress Concentration at a Surface Notch or Crack.

13. DESIGE*' STRESS-RAISERS. Further stress analysis studies showed tliat various features of de s~gn could also ac t a s stress- ralsers, producing stress concentrations f a r above the aserage for the surface. Stresses a t a hole, for example, can be three tlmes the average tenslon stress for the surface St i f fen~ng members intended

PRINCIPLES OF MAGNETIC PARTICLE TESTING -

Fig. 4%-Stress Concentration at a Hole in a Surface Stressed In Tenston -

Shown by the Brrtlte Coateng Method

to strengthen the par t often turn out to be stress-raisers of high magnitude, defeating their purpose entirely. Fatigue cracks were often observed starting a t such locations, due to these stress concentrations.

Fig. 46-Stress Concentration Pattern on a Ribbed Casting, Shown by the Brittle Coating Method.

104

14. FATIGUE OF METALS. Research and experience thus gradually explained the "ho~s" of fatigue. The "why" is still not altogether clear. but new methods of s t udy~ng what happen4 tnside the gratns of steel under fluctuating or reversing stress is today by \vay of giving this answer also. From the point of vie\\, of the non- destructtve testing engineer, however. the "holv" IS of more immediate importance than the "why"

Tremendous numbers of tests were made to produce failures in the laboratory under controlled conditions. Testing machines were devised which permitted test specimens to be subjected t o large numbers of stress reversals under various ieveis of maximum stress. I t was learned that for many metals there was a definite relation between stress level and the number of cycies before a fatigue crack was formed (or before fracture occurred). The R. R. Moore rotat- lng-beam type of fatigue testing maihtne was at first the most common means used for such testing, since with it the number of stress reversais couid easily be run into the millions in a relatively short period of time. Also, i t reproduced the conditions of shafts and axles which so frequently were the victims of fatigue failures. Other types of testing machines !vilich produced tenston stress vartations

1Co"rter~ \\oiedemnn i,i",iiO". Warner B S w i l r ~ r GO.,

Fig. 47-The R. R. Moore Rotating-Beam Fatigue Test~ng Machine.


o r in other ways, often on actual structural components, gave added information.

15. FATIGUE STRENGTH. From this mass of data, certain rela- tionships emerged \$;11ieh gave experimental means for predicting the fatigue life of various types of steeis and alloys. These relation- ships a r e called "Fatigue Strength", "Fatigue or Endurance Limit", "Fatigue Ratio" and "Fatigue Life".

Fatigue Strength* IS defined a s "the maximum stress that can be sustained for a specified number of cycles \vithout failure, the stress being completely reversed within each cycle, unless otherwise stated".

Fatigue o r Endurance Limit is "the maxlmum stress below which a materiai ean presumabiy endure a n infinite number of stress cycles".

Fatigue Life 1s "the number of stress cycles that can be sustained prior to failure for a stated test condition".

Fatigue Ratio is "the ratio of the fatigue limit fo r cycles of reversed flexural stress, to the tensile strength"

16. FATIGUE TESTING. TO determine the values of fatigue strength and fatigue limit for any given steel requires running many tests, often going to many millions of cycles. To get the t rue values fo r the steel under test, i t is necessary that the test specimen be f ree f rom flaws tha t would act as stress raisers and thereby would reduce the fatigue values belour what they xvould be for a flaw--free specimen. Even the corrosive effect of the water vapor or oxygen in the a i r can affect the results, and some tests liave heen run in a vacuum or an atmosphere of inert gas. Acceptable values for the fatigue properties had, therefore, to be the akrerage of many runs made on test specimens a s free from all stress raisers or flaws a s careful selection made possible.

Today, fatigue limits for most of the ordinary steels and alloys a r e readily available to designers.

17. DESIGNING FOR FATIGUE. \%'ith Iiuo\vn values fo r fa t ig~ ie strength and fatigue limits i t is ,theoretically possible to produce a par t for a given service tha t will not Pail in fatigue. Iinowing the number of cycles expected during the service life of the par t , the --- 'These definltionr a3.r quoted f rom Lhe A.S.AI. illetais Handbooli, Voi. I . 8 t h

,i Edit:on, Page 16.

C H r n 4

XIOW AND WHY METALS PAIL

designer has merely to set a maximum stress limit which is brio\\# the fatigue limit of the steel he ~n t ends to use.

As a practical matter, however, this is extremely dificult to achieve. Even thougfi the use of esperimentai stress analysis methods ~ndicates tha t no design stress-raisers a r e present in the part a s designed, some of the various types of discontinuities discussed in Chapter 3 a r e almost certaln to be present. Cracks, scratches, nicks, seams, near-to-surface stringers of non-metallic ~nciusions or large Interior inclusions-all these a r e stress raising hazards which cannot be satisfaciorily allolved for in advance, even by the use of a iarge factor of safety. This is because the size, location and stress-raising severity of these probably-occurring defects cannot be known in advance.

The methods of nondestructive testiiig a r e the means by whiin such stress-raising conditions can be eliminated. This requires the careful inspection of each par t by one or more nondestructive tests suitable for the conditions involved. Fo r iron, steel and other ferromagnetic materials, the magnetic particle method is the simplest and most reliable for locating and evaluating most of these discontinuities which would act as damaging stress-raisers.

18. INSPECTING FOR FATIGUE CRACICS. Fatigue cracks almost invariably s ta r t a t the surface, induced by the presence of some sor t of stress-raiser. They a r e most likely to occur in areas of high tension stress. Such areas include holes, notches, sharp changes of sectlon, stiffening members, etc. The most likely piaces to look for fatigue cracks, therefore, a r e a t sharp ciianges of section such as fillets, key\vays, splines, roots of screw threads and gear teeth, etc. On crankshafts, the most common point for fatigue cracks to appear i s a t the fillets where the main o r t h r o ~ ~ , bearings join the cheeks of the crank.

Fatigue cracks just s tar t ing may be very shallow, and very short, but such indications must not be ignored, since in most cases the conditions which caused the crack to s t a r t will cause i t to propagate to failure. Depth of an isolated crack can often be estimated by its length, since propagation tends to be uniform in all directions from the point of origin. The "oyster shell" markings on a fatigue fracture usually consist of uniform circular concentr~c lines and clearly indicate the point of origin. iltultiple fatigue cracks are often encountered on shafts and railroad axles. In these cases, many shallow cracks have started and later "joined up", sometimes making a

I07

C H A ~ X 4 PRINCIPLES OF MAGNETIC PARTIC1,E TESTING HOW AND WHY METALS FAIL

Fig. 48-Fatigue Cracks Starting Out ot a Sharp Fillet on a Rac~ng Stock Car Safety Hub.

(Courtesy sydnnr walker. ~ t l O Encmcermh. con^.)

Fig. 49-Fracture Through a Small Fatigue Crack Showlng Circular Shape ot the smooth Portion

108

r lng of cracks completely around the axle or shaft. In thrs case the apparent length of the crack obviously does not rndicate the depth.

19. FATIGUE I N TORSION. When a par t is b e ~ n g stressed In torston, the maximum tension stresses occur in a direction 4 j o to the torsional axis. The most common machine par t in this category is the helical wlre sprlng. Fatigue cracks in such springs will have a direction 45' to the a r l s of the zuzre. Lon~itudinal seams in the wire from w h ~ c h the spr ing was wound will therefore lie a t 4 5 O to the direction of nzaxtmzcm tension stress, and invartably form a nucleus for the s t a r t of fatigue cracks. Wire ior the rod from w h ~ c h the wire is drawn) intended for coiling Into sprtngs, such a s valve springs, the service load variations of whtch are severe and rapid, is usually inspected with extreme care to eliminate w ~ r e with even very fine surface seams, pits or other stress raisers.

Fig. 50-Fatigue Cracks at 45% to the AXIS ot the Wire on a Helical Spr~ng.


20. EXPERIMENTAL STRESS ANALYSIS. The methods of experimental stress analys~s have been repeatedly referred to in the preceding discussions. There are severai useful procedures in this field. A brittle coating (Stresscoat') applied to the surface hefore sstress- ing glves, when the part is stressed, a pattern of stress location and distribution, and the direction of maximum stress. Photo-elastic methods employ models made from certa~n plastics which, when stressed and examined with poiarized light, show stress distribution and concentrations visually. Electric resistance strain gauges provide means for measuring surface stresses a t points which the shape of the part and the nature of the servlce may have indicated. The maximum stresses, the location of which brittle coating or photoelastic tests have determined, can then also be measured with great accuracy by use of these strain gauges.

Fig. 51-Brittle Coating Stress Analysis Kit.

By making a proto-type of the first design of a new casting and subjecting it to stress analysis by the brittle coating method, the areas of stress contentration can be mapped and changes made in the design until the stress distribution approaches uniformity, or a t least is free of serious eoncenirations. Figure 53 shows an example of redesign by this means. The final design, a t right, is lighter, stronger, and much easier to cast. -- 'Stresscoat. Registered in the United States Patent Otliee. Property of Magna- flux Corpo~.ution.

CHAPTER 4 HOW AND WHY METALS FAIL

(Courtesy Pro,. Wm. H. hlurmy, M.I.T.) Fig. 52-Stress Pattern Shown an Small Ring Made of a Photo-Elastic Ressn. The

Ring IS Stressed m Compresston and Photographed wlth Transmitted Polarbzed L~ght.

Fig. 53-Progresswe Improvement ot Casting Desbgn Resulting from Use ot Brittle Coating Stress Analysis.


21. RATE OF PROPAGATION OF FATIGUE CRACKS. Once a fatigue crack has started, i t will usually continue to progress through the section until failure occurs. Foitunateiy, the r a t e of propagation through steel is usually relatively slow, especially if stress levels do not rise f a r above design limits. It is therefore usually possible to detect such cracks in their early stages by inspection a t intervals t ha t may be separated by several months o r even longer. Failure in service can thereby be prevented.

Fatigue cracks in general propagate much more rapidly in non- ferrous metals such a s aluminum and bronze, than they do in most steels.

22. SALVAGE OF PARTS SHOWING FATIGUE CRACKS. Regular In- spection of machine or other par ts liable to fatigue failure makes possible the early detection of cracks when still very shallow; a s has just been said. In such cases the cracks can often be ground or '%lended" out and the par t safely returned to service. The anlount of metal whlch i t is necessary to remove to eliminate the crack completely must not, of course, go beyond the minimum dimension permitted for the section involved.

23. CORROSION. When metal is under tension stress and is a t the same time subject to corrosion, cracks develop and progress through the section until failure occurs. The cracking is due to corrosive attack on the grain boundary material, so tha t propagation of this type of cracking 1s zster-granula$as opposed to fatigue cracking, which 1s trans-granular.

Stress corrosion cracking will proceed in stressed members even under static loading conditions. The stresses favoring such corrosive

Fig. 54-Corrosaon Fatigue Cracks on an Oil Well Sucker Rod

112

Cnarrnl 4

HOW AND WHY METALS FAK

attack need not be externally applied, but may be residual stresses from some previous operation such a s cold-forming, heat treating or welding.

The rate of propagation of stress corrosion cracking may be increased if the stress is variable. Under such circumstances the heno omen on is called corrosion fatigue. Par t s operating under high reversing stresses and subject t o corrosion a t the same time will fail much more rapidly than under simple fatigue. All values for fatigue limits a r e useless when corrosron enters the plcture. There IS no predictable endurance limit in the case of corrosion fatigue.

Magnetic particle testing will readil:' locate the presence of cor- roslon fatigue cracks.

24. CREEP. Steel under tensile load ant1 a t lilgh temperatures undergoes slow. changes that may eventually end in failure. The metal elongates in the direction of the tensile stresses. This phenomenon 1s called "creep". The greater the load and the hlgher the temperature, the greater IS the amount and rate of creep.

Creep data is available for many steels for design puTposes, ai- though the time required for making a serles of creep tests a t various stress levels and a t various temperatures can be extremely long. Progress of creep is very slow indeed, especially a t the lower temperatures and stresses. Such tests have shown, however, that there is no "creep limit" corres'Qonrjing to the fatigue endurance limit, and in design, stresses and temperatures must be adjusted so that failure will not occur durlng the service life of the par t or structure.

When craclctng occurs under creep conditions the cracks progress along the grain boundaries a s in corrosion cracklng. Cracks due to creep can be readily detected with magnetic particle testing.

25. SUMRIARY. The discussion in this chapter is not presented a s a complete coverage of the subject of metal failure. I t is intended merely to give the magiletic particle testing operator o r i n s~cc to r some knowledge of the more common ways in \\ihicli metals fail; to aid il!m in h is eramrnation of par ts that have been In serv~ee, and to increase his al\rareiless of the impoi.tancc of ~lefect detection by i iondestructiv~ testing means. The infol.mation given in this an(! the o~eceding chapter will also bc of lhelp in the pi-oblrm of "\rli;lt to look for and whe1.e to look for it"

CHAPTER 5

DEFINITIONS OF SOME TERMS USED IN MAGNETIC PARTICLE TESTING

1. NEED FOR DEFINITIONS. For the individual who undertakes to apply magnetic particle testing for the detection of fla\vs; certaln terms comn~only used in connection with the method should be defined, and be understood by hlm. I-Ie need not understand the vast amount of theory, nor be familiar in detail with all the areas of phys~cs and electrical engineering with which many of these terms a r e concerned. But he must understand their significance a s they apply to, and a r e used in connection with, magnetic partlcle testing.

The definitions a s given a r e not put in hlghlp technical terms, but a r e worded with particular reference to thew use in tile language of magnetic par t~c le testing.

2. GROUPS OF TERMS. The terms defined fall into four groups. These are:

l a ) Terms relating to magnet~sm. f b ) Terms relating to electricity. ic) Terms relating to electro-magnetism. i d ) Terms relating specifically to magnetic particle testing.

The definitions are arranged.in alphabetical order under each group.

3. TERMS RELATING TO MAGNETISM.

11) d l a g l ~ e t i s i ~ ~ . Magnetism is a property of matter that \\'as kno\iln to the ancients. The iron oxrde mineral, magnetite, i o r lodestone) exhibited, in its natural state, the property of attractrng bits of iron to itself. Fragments also had the ability of a l i gn~ng themselves in a north and south direction, and were the first compasses used f o r navigation. Later scientific ~nves t~ga to r s showed that all matter, including l iqu~ds and gasses, was affected by magnetism in one way or another, but to widely varylng degrees. Some materials such a s iron \\,ere strongly attracted to a magnet-others much less so. The ability of matter to a t t ract other matter t o itself in tills manner is called magnetism.

CHAPTR 5 DEFISITIOSS OF SOllE T61$nIS

(2) Coe~cice Forcc. See Hysteres~s.

(3) De?nagnelizatio?z. The process of renlovlng the magnetism existing in a part .

14) Fe~romag?zetic ??late~iuls. With most metals and other materials, the attraction o r repulsion when under the influence of a magnet is very slight. X few mater~als , notably iron and steel, and cobalt and nrckel, a r e attracted strongly. Tiiese materials a r e s a ~ d to be ferromagnetic. I n magnetic particle testing we a r e concerned only with ferromagnetic materials.

( 5 ) Flux Deasity. By t h ~ s term is meant the number of flux lines per unit of area, taken a t r ight angies to the direction of the flux. I t is tile measure of magnetic field strength.

(6) Gauss. This is the unit of flus density or induction. The strength of field induced in a ferromagnetic body is described a s being so many Gausses. I t is usually designated by the letter "B". Numer~cally, one Gauss is one line of flux per square centimeter of area.

f:) Hall Efect. An effect used in the measurement of nlagnetic fields. Assume an electric current being passed through a thin rectangular metal plate. The lines of flow are parallel to the edges of the rectangle. Then pairs of symmetrical points, one on each of these edges, will be a t the same potential. (See Fig. 55) Now iet a uniform magnetic field be applied

HALL EFFECT

POINTS o PIND b ARE AT T H E HALLEFFECT

SAME POTENTI&L CURRENT

I I

Fig. 5 S T h e Hall Effect.

PRINCIPLES OF nlAGNETlC PARTICLE TESTING

a t right angles to the piane of the plate. These pairs of poxnts will no longer be a t the same potential, and if any palr be connected tllrough a galvanometer, a transverse current will flow. The amount and direction of current so set up is determined by the strength and direction of the applied magnetic field. Transverse difference of potential is directly proportional to the longitudinal current and to the magnetic field, and xnversely to the thickness of the plate.

By controlling all the other variables, the strength and direction of the magnetic field can be measured with accuracy.

(8) Norse-shoe Magnet. A bar magnet, bent into the shape of a horse-shoe so tha t the two poles a r e adjacent. Usually the term applies to a permanent magnet.

(9) Hysteresis. When an unmagnetized piece of iron is exposed to a gradually increasing magnetizing force, and the strength of the induced field in the iron, "B", is plotted against the magnetizing force, "H", a curve like that shown in Fig. 56

B I

Fig. 56-Hysteresis Curve

116

CHAPTER 5 DEFINITIONS OF SORlE TERhtS

is produced. As H is increased, 16 also increases along the line "oa" to a point a t which fur ther increase in the niagnetizing force, 13, produces no fur ther increase in the flux density other than that w h c h urould occur in non-magnetic material o r air. This point, "a", on the curve is called the Saturation Point and the Iron is said to be magnetically saturated a t this point.

As the magnetizing force is then reduced, the flux density does not decrease to zero by the same curve, but lags behind, so that when H reaches zero there is still some flux in the piece of iron. T h ~ s field is called the Residual Field, and on the curve of Fig. 56 is the value of B a t wh~cl l the curve crosses the B axis. T h ~ s is point "b" on the curve.

As the magnetizing force is fur ther decreased, now in the negative direction, the value of B decreases till i t crosses the H axls a t polnt "c". The value of negative or reverse magnetizing force necessary to brlilg the flux density back to zero af ter saturation, and demagnetize the piece-distance "co" on the H axis-is called the Coercive Force.

As the magnetizing force is fur ther decreased in the negative direction the iron reaches a point of negative saturation ("d"), and a s H 1s then Increased in the positive direction the curve passes the point "e" on the B asis, the negative residual field. The uolnt "f" on the H axis is equal and opposite to the coercive force "c". Further increase of H brings the field strength B back a g a ~ n t o the point of saturation, "a"; thus completing the curve.

The curve "abcdefa" is called the N1~ste~eszs Curve, or H?/ste?ests Loop. I t gives much information about the magnetic cl~aracteristics of a ferromagnetic material, and the terms just defined will be much used in this discuss~on.'

i 10 ) I?td?~ction. Magnetic induction is the magnetism induced in a ferromagnetic body by sonif outside magnetizing force.

i l l ) Lzzes of Force. When a piece of paper is laid over a magnet, and iron filings or other iron powder is sprinkled over the paper the powder arranges itself into a pattern as shown in Fig. 57. This pattern is called a magnetograph. I t appears to consist of a series of curved lines and suggests that the magnetic force of the field flows along these lines. Although

117

CHII'TER 5 PRINCIPLES OF MAGNETIC PAtiTICLE TESTING UEI.'lXlTIOSS Oi' SOXI? TBRBIS

. - -.

there is actually no known movement, i t is most convenient ? (16) ItJagnetzc Piile. Tile flux lines ha\ze direction, and a s their to thlnk of the fidd a s flowing tIlr0Ugh the magnet and path passes From the iron of the marnet into the air, or from througli the a i r around it, folloi\~ing the path of these lines. the ai r into the iron; rnagnutic poles :Ire set up. Frrro- They a r e called Lznes or Fovce. magnetic niateri:~ls w e attracted to these i,oles. The point

Fig. 57-Magnetograph ot the Field Around a Magnet.

(12) illagnet. Materials that show the power to a t t ract iron a n ~ i other substances to themselves, and that exhibit polarity, are called Magnets.

(13) Magnetic Fieid. The space arounh a magnet withill wl~rch ferromagnetic materials a r e attracted is called a ilfagaetic Fieid.

(14) il.Jagnetic Flztx. The concept that the magnetic fieid is flow- Ing aiong. the lines of force suggests tliat these lines are therefore "flux" lines, and they a r e called Magwetic Fl?cz. The strength of the field is defined by the number of flux lines crossing a unit area taken a t nglit angles t o the direction of the lines.

115) ~Mngnetiz~ng Fo?.ce. For the purpose of this discussion, niagnetizing force is considered to be the total forcc tending to set up a magnetic flux in a magnelic circuit. I t is usually designated by the letter "H" and the unit is the "Oersted".

where the flux leaves the magnet is caller1 the norlA pole, and where the Hux enters the magnet is called the s o ~ ~ l l l vole. These names a r e deriver1 from the compass. One end of the compass needle always points to the earth's north magnetic jmle, and is called the "north-seeking" !,ole. The other end mhich points south is called the "south-seeking" pole. For purposes of magnetic language the terms a r e snortcned to simply .Vo,.tlr Pole and Soutit Pole.

(17) Ilfag?zetogvau/r. A mngnetogr:lph is a picture of a magnetic field made by the use of iron po\v\'der under conditions that allow it to arrange itself into the paLtcrn of the field. See "Lines of Force" (11 ) .

(18) Oe~sfcd . This is the unit of field strength !vhich produces magnetic induction. I t is designated by the letter "11" The Oersted and the Gauss a r e numberically equal In a l r o r in a vacuum. Oersted (I-I) refers to the magnetizing force tend- Ing to magnetize an unn~agnetized body, and Gauss refers to the field (B) so induced in the body.

(19) Pnvamag~i.ct,ic anrl Diamap~et ic . 411 n1aterials a r e affected by magnetic fields. Those \\rli~ch a r e attracted are called Paramagnetic. Those \vhicli :1re repcllcd are called Dia- m a g n e t ~ ~ . The reaction to a magnetic field of these two classes of substances IS very slight indeed. The few maler~a is that a r e strongly attracted by rniignctic fields a r e called ferromagnetic.

(20) P r r~nanoz t dln,qnct. Alnlel.ials such a s Lodestone exhibit the property of magnetism a t all times and with little or no reduction 111 strength when measured against time. Such materials a1.e calle~i r ~ e ~ s i a n o z t atagnels. Hard stcels and many alloys, \\,hen magnetized, retaln much of thelr magnetism and become pcrmancnt magnets.

(21) P e ~ ~ ~ ~ r a b i i i t ] j . This is the term used to refcr to the ease with ~ ~ h r c h a magnetic field or flus can be set up in a inagnclic ci~.cuit. I1 is not a constant value for a give^, inater~al, but is a ratio. At any given value of mngnelizing force, permea-

1'RINCIPI.ES O F SIAGNETIC PARTICLE TESTING

bility is E/H, the ratio of Rux' density, E, to magnetizing force, H.

(22) Reizictnnce. Reluctance 1s the opposition to the establisiiment of magnetic flux in a magnetic clrcuit presentetl by the path through \vh~cii the flux must pass, and is t h a t which determines the magniluae of the Rux produced by a given magnetizing force. I t is analogous to resistance in a n electric circuit.

(23) Remancnl Magnelisnz. This is t h e term applied to the magnetism remaining 111 a magnetic circult a f t e r the magnetizmg force has been removed.

(24) lZeszdzio1 Field. This is the field left in a p ~ e c e of ferromagnetic m a t e r ~ a l when the magnetizing force has been reduced to zero. I t is represented by point "b" on the hysteresrs curve (F ig . 56) .

(25) Retent,ivttg. Retentivity IS the property of a given m a t e r ~ a l of retaining, to a greater o r lesser degree, some amount of residual magnetism.

(26) Salz~rat ion. This term refers to tha t degree of magnetization where a fu r the r increase in H produces no fu r the r increase In the field in a given materiai other than is produced in air. Tlus is point "a" on the hysteresis curve.

(27) T e n ~ i ~ o v a ~ g ~t iagnet . A temporary magnet is a plece of soft iron o r steel which has little o r no retentivity, so t h a t i t re- tu rns substantially to the unmagnetized s ta te when the magnetizing force is \icithdrawn. L

4. TERMS RELATING TO ELECTRICITY. The followmg iisted elec- Lrlcal t e rms a r e those most frequently used in connection with magnetic particle testing:

i l ) A l f e ~ n a t i n g C ? w ~ e n t o r A.C. Alternating current is current tha t reverses i ts direction of flow a t regular nltervals. Such current is frequently referred to a s A.C.

(2 ) Air771evr T h ~ s 1s the unit of electrical ctlrrent. One ampere 1s the current whlch flows througll a con(luctor havlng a resistance of one ohm, a t a potential of one volt.

(3 ) Condzicti~*~ty. Thls 1s tlie inverse of res~stance, and refers to the ability of ;i cr>iiiluctar to c;~ri.y cursent.

( 4 ) Decuy. As used in connection with electr~clty, decay is the falling ofi to zero of the current in a n e lect~ical c~rcu i t . Mag- nctlc fields and electrical potentials can also decay in a similar sense.

(5) D i ~ e c t Czt~vent OT D.C. As the name implies, this term refers to an electric current flolving continually in one direction through a conductor. Such current is frequently referred to a s D.C.

(6 ) Full IfIci~ie Rectified A.C.-Single Phase. This is ~.i.ctiiicrl alternating current fo r wliicli the rectifier is so connected t h a t the reverse half of the cycle 1s "turned around", and fed into tlie circuit f low~ng in the same directlon a s the first half

I I

Fig. 5&Wave Form of Full Wave Rectified Single Phase A.C.

of the c)rcle. Tills produces p u i s a t i n ~ D.C., but with no In- terval hetween the pulses. Such current 1s also referred to a s single phase full \\lave D.C.

(7 ) Fz~ l l Wave Rectifier1 T h i . ( < ~ Phnse A.C. \Vhcii thrce phase alternat ing current 1s rectified the full wave rectification sps- tem is used. T h r result is D.C. \rith very little pulsation-

- - - . - - - - MEAN D C . VOLTS

TIME

- I Fig. 59-Wave Form ot Three Phase Rectified A.C.

PRINCIPLES O F XAGNETIC PARTICLE TESTING

in fact only a ripple of varjring voltage distiriguil;i~es it f rom straight D.C. (Fig. 59).

( 8 ) Hali Wave Rectified A.C. IV11en a singie phase alternating current is rectified in the simplest manner, the reverse half of the cycie is blocked out entirei),. The result is a puisating uni-directional current with intervals when no current a t all is Rowing. This is often referred to a s "half wave" o r a s puisating direct current. This type of current is illustrated in Fig. 60.

Fig. 6 6 W a v e Form of Half Wave Rectified Single Phase A.C.

(9) impedance. This term 1s used to refer to the total opposition to the flow of current represented by the combined effect of resistance, inductance and capacitance of a circuit.

(10) Indl~ctive Reactance. This is the opposition, independent of resistance, of a coil to the flow of an alternating current.

(11) Ohm. The ohm is the unit of electrical resistance. I t is the value of a resistance tha t will pass on& ampere of current a t a potential of one volt.

(12) Recti.fied illlevnuling Cur~.ent. By means of a device called a rectifier, \vhich permits current to flow In one direction only, alternating current can be converted to direct o r uni- directional current. This differs from direct current in tha t the current value varies from a steady level. This variation may be extreme, as in the case of half ware rectified siiigie phase A.C. (sce definition $8) or slight, a s in the case of three phase rectified A.C. (see definition $7).

(13) Reszstance. Res~stance is thc opposition to the flow of an electrical current through a conductor. I ts unit is the ohm.

(14) Single Plrase Altcr?rating C u r ~ e n t . Tliis term refers to a

simple current, alternating in direction. Commercial single phase current fol lo~r~s a sine wave, illustrated in Fig. 61. Sucn a current requires only t\vo conductors for i ts circuit. Xlost common conimerc~al frequencies a r e 25, 50 and 60 cycles per second.

I I Fig. 61-Wave Form ot Single Phase Alternating Current.

( 1 5 ) Tiwee Phase Alternating Current. Commercial electr~city is comnionly transmitted a s three single phase currents- tha t IS, three separate currents follou~ing separate sine curves, each a t 60 cycles (o r other frequency) per second, but with the peaks of their individual curves one-th~rd of a cycle apart . This set of curves is illustrated in Fig. 62. At least three (sometimes four) conductors a r e required for three phase alternating current.

Fjg. 62-Set of Wave Forms ot Three Phase Alternatlng Currents

(16) Transtent C?~vve??.ts. These currents a r e of short duration, generated by sudden changes in the eiectrical or magnetic conditions existing in an electrical or magnetic circuit.


i l ? ) Volt. The volt is the unit of electromotive force which tends to cause an electric current to Roiv through a conductor.

5. TERMS RELATING TO ELECTROMAGNETISM. The following group of terms has to do with magnetism rnduced by the electric current:

(1) Anzpere Turns. This term refers to the product of the number of turns in a coil and the number of amperes of current flowing through it. This is a measure of the magnetizing or demagnetiz~ng strength of the coil. Fo r example: 800 amperes in a 6 turn coil = 800 X 6 = 4800 ampere turns.

(2) Electro-nzagnet. When ferromagnetic material is surrounded by a coil carrying current i t becomes magnetized and is called an electro-magnet.

(3) Indzrced Current. Two fundamental electrical prlncrples must be understood in order to define the term ~nduced current. f a ) When current flows through a conductor, i t sets up a

magnetic field a t right angles to the direction of current flow. The strength of thrs field varies directly with the current.

( b ) When a conductor in. the form of a closed loop moves through a magnetic field, a current will flow in the conductor. This is the basrs of transformer action.

Passing an alternating current through a conductor will set up a fluctuating magnetic field. If a second conductor in the form of a closed loop IS placed in this Yield, the action of the fluctuating field moving across the conductor will set up a second alternating current of the same frequency. This is an induced current.

When the moving field is produced by the sudden discon- tinuance of a direct current ( a "collapsing field") tlie induced current is of only very short duration. Induced currents such a s this are used a s a method of magnet iz~ng ring-shaped parts.

(4) Inductance. When the current in a circuit is varied, the change in the magnetic field surrounding the conductor generates an electromotive force (voltage) in the circuit itself.

124

CliaPTER 5

nEFlNITlONS OF SOME T611AIS

If another crrcuit is adjacent to the first the change ni the magnetic fieid of the first mill generate an electromotive force in tlie second. The phenomenon is ltnowil as inductance- either self-inductance or mutual inductance. Inductance is measured by the electroniotive force produced in a coilductor by unit rate of change of the current. The iunit of inductance is the Iienry which is that inductance in u~hich an electrc- motive force of one volt IS produced \\,hen the inducing current rs changed a t tlie rate of one ampere per seconil.

( 5 ) LOOP. This term refers to a single turn of \\,Ire 01. cable used to carry electrrc current. I1 is used for magnet iz~ng and demagnetizing purposes.

(6) Sore?zozd. A solenoid is a coil consisting of a iiumber of ioops of wire or cable to carry e1ecti.i~ current. I t may be used fo r both magnetizrng and denlagnetizrng purposes.

6. TERMS RELATING TO MAGNETIC PARTICLE TESTING. The following group consrsts of terms that are directly in reference lo magnetic particle testing.

( 1 ) AZT Gnu. Inillen a magnetic circuil contams a small gap \\~l~icli tlie magnetic Rux rilust cross, the space is referiwi to as an a??. g m ~ . Cracks produce small a i r gaps on the surface of a part.

(2 ) Black Lzglrt. Black light is near-ultraviolet light having a wave length of 3650 Angstrom Units. I t is used in fluorescent magnetic particle testing to cause the particles to fluoresce and give off v~sibie light.

( 3 ) Ce?ttrnl Conductor. A central conductor 1s a conductor that is passed througli the opening in a ring o r tube, or any hole in a part , for the purpose of creating a circular or circumferentiai field in the tube o r rrng, o r around the hole.

( 4 ) Ci~czilar Mag?zetizatio?z. Magnetization of a piece of ferromagnetic material in such a way that the flux path is completely contained within the article, IS called circular magnetization. There a r e usually no external poles. I t is usuallt. produced hy passing current directly through tlie piece.

(5) Coil Shot. A "shot" of magnetizing current passed through a solenoid or coil surrounding a part, fo r the purpose of

125

f

PRINCIPLES OF DIAGNETIC PARTICLE TESTING

lo~igitudinal magnetization is called a "coil shot." Duration of tile passage of the current is usually very short-often only a fraction of a second.

(6) Dcfeet. A condition, of whatever kind, that renders a par t unsuitable fo r its intended servlce.

(7 ) Disconti,~.?dt?,. This is the general term used to refer to a break in the metallic continuity of the par t b a n g tested-a break that may be a crack, a seam or other interruption t o the continuity of the material. I t is such a break tha t produces a magnetic particle indication.

(8) D i s to~ t ed Field. The direction of a magnetic field in a sym- met r~ca l object will be substantially uniform if produced by a uniformly applied magnetizing force, a s in the case of a bar magnetized in a solenoid. But if the piece being magnetized is irregular in shape, the field 1s distorted and does not follo\\r a straight path or have a uniform distribution.

(9 ) False Indication. Any patch or pattern of magnetic particles not caused o r held in place by a iealcage field of any klnd 1s called a false indication.

(10) Fluovescence. The light given off by certain materials while they a r e being irradiated with short wave ultrav~olet or near-ultraviolet light.

(11) Heads. The c iamp~ng contacts on a stationary magnetizlng unit.

(12) Head S l~ot . A "shot" of magnetizlng current passed through a par t o r a central conductor wtlilgciamped between the head contacts of a stationary magnetizing unit, for the purpose of circular nlagnetization of the par t 1s called a "head shot." Duration of the passage of the current is usually less than one second.

(13) Indication. This term refers to any magnetically-held magnetic particle pattern on the surface of a par t being tested.

(14) Leakage Field. Tius is the lield forced out into the a i r by the distortion of the field within a par t caused by the presence of a discontinuity.

(15) Lo?tgitudinal illagnetization. 3lagnetization of a materiai in sucii a way that the magnetic liux runs substantially parallel

-

to the iong axis of the par t , the flux path completing itself through the a i r outside the material; is called longitudinal magnetization. I t is sometimes called bl-polar magnetiz:itlon, because nt least two external poles exist in longitudinal magnetization.

(16) dlagnek~c Discont i?~~~i ty . This refers to a break 111 the ?nagnetic uniformity of the part-a sudden change in permeability. A magnetic discontinuily may not be related to any actuai pliys~cai break in the metai, but it may procluce a magnetic particle indication.

(17) Metallic Disco?ttz?z~~it~/. This terms refers to an actual break In the continuity of the metal of a part , and may be located on the surface-as for ~nstance, a crack; or deep in the interior of the part-as for example, a gas pocket.

(18) Mzclti-di~cctional Magnetizafion. T\rro separate fields. Iiaving different directions, cannot exist in a par t a t the same time. But two or more lields in cliffercnt directions can be ~mposed upon a part sequentially in rapid succession. )\:lien this is done magnetic particle indications a r e formed when discontinuities a r e located favorably with respect to the directions of each of the fields: and will persist a s long a s the rapid alternations of field direction contiouc. T h ~ s , .zn effecl, does constitute two or more fields in different directions a t the same time, and enabies the detection o l defects oriented in any direction i n one operailon. The method has aiso been called by the name of "Duovec" *

(19) Non-Relevant Indicntim~s. These a r e true indications produced by leakage fields. Ilowever the conditions causing them are present by design or accident, o r other features of the par t having no relation to the daniaglng fla~\'s being sought. The term s~gnilies that such a n indication has no relation to discontinuities that might constitute defects.

(20) P w t . Throughout discussions of the applicalioils of magnetic particle testing it is necessary to refer to the arlicle being tested. The term "part" is used quitc generally fo r this purpose, a s a quiclc and easy ~rsord to say. Often the article actually is a "parts-of an englne or other assembly-but

'Duovec. Tr:iderna~k rt.g~str~.cd ,n thr U.S. falrnl ORcrr. I ' I I I ~ P I . L Y of Magna- flux Corpol-aL!ol,.


the word "parl" implies only tha t i t is the piece of material, of whatever sort, tha t is being tested.

(21) Prods. Two hand-heid electrodes which are pressed against the surface of a par t to make contact fo r passing magnetizing current through the metal. The current passing between the tmo contacts creates a field suitable fo r finding defects with magnetic particles.

(22) Qt~ielc-Bvealc. Sometimes called "Fast Break" The sudden breaking of a direct current causes a transient current to be induced in the par t by the rapid collapse of the magnetic field. I n magnetic particle testing, fast breaking of the magnetizing current is used to generate a translent current in a part which is favorable for finding transverse defects at the ends of longitudinally magnetized bars. Such defects a r e often concealed by the strong polarity a t the bar ends. A t such locations the lines of force of the longitudinal fieid a r e leaving the bar in a direction normai to the surface, which prevents them from intercepting transverse defects in those areas. Tlie fieltl induced by the translent current does intercept such discontinuities. See Chapter 7, Fig. 53.

(23) Resz~ltant o~ Vector Field. When two or more magnetizing forces operating in different directions a r e simultaneously applied to a ferromagnetic material, a resultant field is produced, having a direction which is determined by the relative strengths and directions of the applied magnetizing forces. Such a fieid is also referred to a s a vector field.

If either or both of the applied magnetizing forces are themselves varying in direction or amount, the resultant field is moving or swinging in direction and strength. Such a moving resultant field is sometimes referred to a s a "sw~ng- ing field".

(24) Th7esliold. In reference to currents or magnetic fields, the nunimum strength necessary to create a looked-for effect is called the threshold value. For example, the nlinimum current necessary to pt'oduce a readable indication a t a glven defect, is the threshold value of current for that purpose.

7. GENERAL COMMENTS. The list of definitions that has been given in the foregoing sections is not a glossary of all the terms that

CHAP^ 5

DEFINlTIONS OF SOiilE TERMS {:

: a r e used in connection with magnetic particle testing. I t is intended

i to give a sufficient background of understanding so that the text of ? later chapters need not be interrupted by the need to define such I : terms; and avoids the possibility that the following discussions as- : sume too much previous knowledge on the par t of the reader. In i general other terms will be defined in the text when they occur.

CHAPTER 6

CHARACTERISTICS O F MAGNETIC FIELDS

1. INTRODUCTION. Thirty-seven years ago (1929) when itfag- netic Particle Testing was first put to use, remarkably little published information existed on those aspects of magnetism and magnetic fields which are basic to the understanding and proper application of this nondestructive testing method. The magnetic properties of iron and steel and various alloys were well knourn, of course, but the use of magnets and magnetic fields was pre- ponderantly in the area of power generation and use-generators, motors, transformers, switch gear, relays, etc. Little had been done to analyze the behavior of magnetic fields insid.e the ferromagnetic material, since all the above mentioned applications are mainly concerned with the paths of fields eztenzal to the magnets. Nor was there any very satisfactory theory of the nature of magnetism itself.

Since that time, our understanding of atomic structure has resulted in theories of magnetism which have a sound basis, and new interest in various uses and applications of magnetism have stimu- lated extensive studies. The powerful magnetic fields used in our mammoth "atom smashers" are on one end of this line of interest, and the very faint fields involved in magnetic guidance systems, and protection of vessels from magnetic mines and magnetic guided missiles are a t the other end. And there is a tremendous volume of literature available today on these areas of the subject. (References on this phase of magnetism are listed in the Bibliography Appendix.)

However, little of the knowledge resulting from all this research and development is of much direct vaiue in the uses to which magnetism is put for magnetic particle testing purposes, and there is no point in devoting space in this book to the general subject of magnetic theory. We are interested not so much in what magnetism is, but rather in how i t behaves with respect to defect-detection with magnetic particles. We are further interested in the strength and direction of fields zuitht?~ ferromagnetic materlal, and how to create fields of proper strength and direction. And we aiso need to know

130

CHASTW 6

CHARACTERISTICS OF XAGNETIC FIELDS

how fields distribute themselves, both as to strength and as to direction, inside parts of various sizes and irregular shapes.

Our interest stems from a srngle point of view-that 1s: to produce a leakage field a t a discontinuity, strong enough to form a readable indication with magnetlc particles. One guide in this mterest is the basic fact that the field must intersect the discontinuity a t some appreciable angle, preferabjy approaching 90'. Therefore direction of .Celrl must be determined and known for reliable inspection. Next in importance is the strength of the field a t the point where the discontinuity lies. The field obviousiy must be strong enough to produce a leakage field above the threshold strength for readable indications. I t can, on the other hand, be too strong and produce confusing side effects.

Therefore in this and the next four chapters we will conslder these several elements from the point of vlew of defect detection- that is, Iiow fields behave, how they are produced and how they distribute themselves in varlous ferromagnetic parts.

2. MAGNETIC FIELD AROUND A BAR MAGNET. If we cons~der a bar of ferromagnetic materlal whlch has been permanently magnetized, we concelve it to be surrounded by a field of force which we call a magnetic field. By laylng a piece of paper over the mag-

Fig. 63-Field Around a Bar Magnet.

131

PRINCIPLES OF BIAGNETIC PARTICLE TESTING -

netized bar and sprinkling "iron filings" on the paper, the fine iron particles will arrange themselves in what appear to he lines. Such a pattern is called a magnetograph.. It should be emphasized that this pattern represents only the field which lies outside the bar. Nor is the field of force around the magnet confined to the plane of the paper on which the magnetograph is produced. I t exists on all sides, surrounding the magnet completely. The magnetograph is a cross-section of the field in the plane of the magnet itself.

The appearance of the n?agnetog~+aph suggests that this field is made up of lines, which extend, for the most part, in curved paths from one end of the magnet to the other. Some of the lines ieave and re-enter the bar along its length; the number of such lines increasing toward the ends of the bar. If the bar is of uniform cross-section and composition throughout its length we can infer that the field passes through it with a practically uniform distribution, except near the ends. As the result of the appearance of the magnetograph, magnetic fields have long been thought of as bang made up of "lines of force", and certain properties have been observed regarding them. Furthermore it is easy to think of these lines of force, or magnetic flux, as flowing through the bar, out into the air: and circling back to re-enter the bar a t the other end. This concept of flow is most convenient in talklng about the behavior of magnetic fields.

3. POLES. Although the flux is not proveably flowing, it does have directional properties. The compass, a magnetized needle, will

i Fig. 64--Consequent Poles on a Bar Magnet.

i CHAPTER G CHARACTERISTICS OF MAGNETIC FIELDS

swing so that the same end always polnts to the north. Similarly, compass needles will always polnt to the same end of a bar magnet. The two ends a t which most of the fiux lines leave and re-enter the bar are called poles. "The "north" pole is that end a t whlch the lines are thought of as leaving the bar, and the "south" pole is that a t .it'hich they re-enter. If the north poles of two magnets are brought together the magnets repel each other. If a north and a south poie are brought together they will attract each other and tend to he drawn together. The rule is : "like poles repel, unlike poles attract".

Normally a bar magnet has only two poles, one north and one south, located a t opposite ends of the bar. However a magnet may have a number of poles, called "consequent poles"-some north and some south. Figure 64 gives some exampies of consequent poles in magnetized parts.

4. MAGNETIC ATTRACTION. The concept of flux lines includes some other charactenstics. Each line is considered to be a continuous loop, which is never broken, but must complete itself through

- FLUX LINES 1

a.

I <

....---

LEAKAGE FIELD JUMPING AIR GAP

.h$AGNETIC PARTICLE I INDICATION

Fig. 65--Bridgrng of the Air Gap at a Crack. a) Leakage Field. b) Magnetic Particle Indication


some path. The lines always leaye the magnet a t right angles to the surface. They tend always to seek the path of lowest reluctance in completing their ioop. A piece of soft iron, therefore, when placed in a magnetic field, will be drawn toward the magnet, so tha t more and more lines of force may traverse it, and tlie high reiuctance path through the a i r will be a s short a s possible. This is the action which causes niagnetic particles to gather a t leakage fields a t discontinuities. The leakage field is jumping across a relatively high reluctance a i r gap a t a crack. The magnetic particies offer a lower reiuctance path to the flux, and a re therefore drawn to, and bridge, the a i r gap.

5. A CRACKED BAR MAGNET. If a bar magnet having two poles, N. and S., a t opposite ends is broken in the center of i ts length, t ~ v o complete bar magnets will result, each having a N. and S. pole. This process of breaking can go on till there are four, eight o r any number of separate complete magnets.

If the magnet is cracked-not broken completely in two-a somewhat similar result occurs. A North and South pole will form a t opposite edges of the crack, just as though the break had been complete. The strength. of these poles will be different from that of the fully broken pieces, and .ivill be a function of the depth of the crack and the width of the air gap a t the surface.

The fields set up a t cracks o r other physical or magnetic discontinuities in, the surface a r e called lenltage fields. The strength of this leakage field detexmlnes the number of magnetic particles which will be gathered t o form mdications-strong indications a t strong fields, weak indications a t weak fieids.

1

6. EFFECT OF FLUX DIRECTION. In the cases we have been considering-that is, a straight bar magnet-the flux lines a re assumed to run lengthwise in the bar, from end to end. Such a bar i s said to be ioilgitudinally magnetized. A crack produced by partially breaking the bar transverseiy will lie a t right angles to the long axis of tlie bar, and consequently will also be a t 90" t o the flux lines traversing the bar. Such a crack-orientation offers the greatest interruption to the lines of force, and therefore forms a strong leakage field.

Assume, however, a straight crack-like discontinuity lying in a direction parallel to the axis of tlie bar, and therefore parallel to the lines of force. No lines would be cut and there would be no

CHAPTER 6

CHARACTERISTICS OF A'LAGNETIC FIELDS

ieakage field. A iundan~eiltal requirement, therefore, if a discontinuity is t o produce a leakage field and a readable magnetic particle pattern is tha t i t must intercept the lines of force a t some angle. The leakage field will be strongest if the angle is 90' and will become wealter a s the angle the discontinuity makes with the lines of force becomes smaller.

I t is, ho\vever, often true that a crack which has a general direction parallel to the axis of the bar, but which deviates somexvhat from a straight line, may give quite strong magnetic particle indications. In such a case the segments of the crack not exactly parallei

1 I PATH OF MAY BE

MAGNETIZING PERMANENT OR MAGNETIC LINES CURRENT FLEXIBJE CABLE OF FORCE

I 4 i. i \ \ 1 X S L \

LONGITUDINAL 45' CRACK TRANSVERSE CRACK WlLL WILL SHOW CRACK WlLL NOT SHOW SHOW

IRREGULAR CRACK MAY SHOW I

k I Fig. 6 G E f f e c t of Crack-Orientation in a Longitudinally Magnetized Bar.

to the axis do, locally, intercept the lines of force. Figure 66 illus- t ra tes these several situations.

7. CIRCULAR NAGNETIZATION. If the bar magnet we have been considering is bent into a circle so that the t1r.o ends a re brought ciose together (see Fig. 67) the small a i r gap will Rave a strong flux jumping across i t and will have a strong attraction for iron pieces or particies brought close to it. This is essentially the familiar horse-shoe magnet.

But if we think of this gap as being completely closed to form a r ing a s if by welding, so that there 1s no metallic discontinuity of any kind, there will be no leakage field or external poies. The field originally present in the bar will still exist in the ring although there will be no external evldence of its presence. I n fact, the field


in the ring will be stronger than tha t which was in the bar, since the all-metallic circuit has a low 1-eiuctance, with no h ~ g h reluctance air gap. Also, there is no self-demagnetizing effect which 1s pres- en t tn t he bar magnet and tends t o limit the field in it. The ring i s now said t o be circularly magnetized.

Fig. 67-Circular Magnetization. a) Incomplete Ring with a Small Air Gap. b ) Complete Ring with no Air Gap, but with a Crack.

8. CIRCULAR X~AGNETIZATION AND CIZACXS. If the ring just described is cracked on the outer cylindrical surface in a direction parallel to the axts of the ring, the internal circular field would be intercepted and local poles \vould a t once be formed. See Fig. 67. The leakage field would then form an indication of the crack if magnetic particles are applied. Here again, direction of the field with relation t o the direction of the crack is of prime importance.

Instead of the ring, let us suppose that a bar or tube is circuiarly magnetized. See Fig. 68. In this case it 1s the crack lying parallel to the long axls of the bar or tube which ihtercepts the flux lines a t 90'. Cracks a t n g h t angles t o the axis wouid be parallel to the

1 1 cuRwF \ .-

CURRENT

\ \ \ MAGNETIC LONGITUDINAL CRACK AT 45' TRANSVERSE CRACK

FIELD CRACK WILL WILL SHOW WILL NOT SHOW SHOW

IRREGULAR CRACKS MAY SHOW

1

Fig. 68--Effect at Crack Orlentation in a Circularly Magnetzed Bar.

136

CHAPTQ? 6 CH.4RACTERISTICS O F MAGNETIC FIELDS

field and would create no leakage fieid. Cracks a t intermediate angles would create leakage fields of strengths varying with the angle; and the crack, essentially transverse, which "wanders" from a stralght line would probably give a n indication.

9. DIFFICULTIES OF ESTABLISHING PROPER FIELDS. Since successful magnetic particle testing depends on establisiiing fields in parts in directions which will cut across the expected or possible cracks, some of these field characteristics must be thoroughly understood. For example, i t must be remembered that the flux path will always go the route of lowest reluctance. This means it will follow ferromagnetic paths when i t can, in preference to taking a path containing a i r gaps of h ~ g h reluctance.

Thus, if i t ts desired to inspect a long bar by longitudinal magnetization in a short coil, i t is not usually possible to secure a truly longitudinal field clear to the end of the bar when the coil is located a t the mid-point of the bar's length. The flux tends to leave the bar a t some distance from the coil, and return around the outside of the coil to the bar on the other side, leavtng the ends of the bar not properly magnetized. When the bar is not much longer than the coil, i t will be satisfactorily magnetized for its entire length. But if the bar is very short, so tha t its length 1s less than its diameter-as, fo r example a disc-shaped par t such a s a gear-a longitudinal field may not be set up a t all.

Unless the disc is held exactly in the plane of the coil, i t will be found that the field takes a path across a diameter, instead of passing through the disc from face to face. If such a piece is hand- held in the coil i t is practically ~mpossible to establish a longitudinal field in i t to locate circumferential cracks on the cylindrical surface by this procedure. See Fig. 69. The proper manner in which to magnetize a disc to locate circumferential cracks is to pass current diametrically through it while clamped between the heads of a magnet iz~ng unit in a manner similar to that used for ring-shaped par ts as shown in Fig. 84, Chapter 7. The disc should be tested twice, with current passed across the diameters a t 90' to each other.

Similarly, i t is difficult to set up a truly transverse field a t r ight angles to the axls of a bar. If the bar is piaced crosswise in a short coil, with the bar positioned a t nearly 90' to the coil's axis, the result will he a non-uniform hut essentially longitudinal fieid foilowing the length of the bar. Transverse fields can be set up in a bar by the use of yokes, when the yokes are positioned so that the

137


PATH OF MAGNETIZING

PATH OF MAGNETIZING

FIELD BAR OR DISC DIAMETER

GREATER THAN LENGTH

Fig. 69--Magnetic Flux Passing Diametrtcally Through a Disc Instead of from ace to ace,-when the Disc 1s Placed at an Angle of N-90' to a Longitudinal Field in a Coil.

two poles a r e diametrically opposite each other and close to the center of the bar.

10. DISTORTED F r n u s . We have been considering symmetrical objects for our magnetization discussion, in wh~cli the fields set up a re likely t o be symmetrical also, and approximately uniform. Fields may, however, be distorted and caused to depart from expected directions by various conditions, and by the,shape of the part.

Variations in permeability in different sections of the part, or In differelit layers of the metal of an otheru.lse symmetrical part, may cause the Internal field, whether c~rcu la r or longitudinal, to depart from the pa t t l i t ~ ~ ~ o u l d take in a homogeneous symmetrical object. Variations In liardness, g ram size, conlposition isuch a s carbon content) all couid cause considerable variation in permeability, and affect the distribution of the field.

A piece of un-magnetized soft iron in contact with the surface of a circularly magnetized part, will cause distortion of t h e circular field. See Fig. 70. Irregular shaped parts will also cause fields to be distorted and to follo\v paths which a re often difficult to predict.

Chapters 8,9, and 10 a re devoted to a discussion of this important subject of field distribution.

CHAPTER 6 CHARACTERISTICS OF MAGNETIC FIELDS

$OF, tion anr I

b. 0. <,Em i l l X O l i i O 9. S L i C i O ( t i tilm ?SCCE Fi&O 01biO1Ti i i O" 5"iiPI OF Pllili

Fig. 7%-Distortion of the Circular Field In a Part Caused by a) Contact with a Piece of Soft Iron. b) The Shape of the Part.

11. PARALLEL FIELDS. If a ferromagnetic bar is placed along- side, and parallel to, a conductor carrylng current, a field will be set up in tlie bar which is more transverse than c~rcular , and is of very little use for magnetic particle testing. Figure 71 shows the path of the field so produced. Operators have tried to use this method a s a substitute for the head shot for tlie purpose of producing circular magnetization. Not only is the field produced not circular, but its essentially transverse direction is effective for finding seams etc. over only a very limited area of the bar. Fur- thermore, tlie external field around the conductor and tlie bar attract magnetic particles and produce confusing backgrounds. This method of magnetization should not be used for magnetic particie testing purposes.

/ CIeE"L).R F l L O snua

l R O Y l D COlDUCiDl i

Fig. 71-Field Produced in a Bar by a "Parallel" Current.


12. PRODUCTION OF SUITABLE FIELDS. The characteristics of magnetic fields most important to the practice of magnetic particle testing have been discussed in this chapter, without reference to the methods of producing them. In the following chapter, methods and means for pk-odueing fields a r e discussed, together with the effect of the many variations in magnetizing methods on the results of the inspection.

METHODS AND MEANS FOR GENERATING MAGNETIC FIELDS

1. THE EARTH'S FIELD. One of the properties of permanent magnets is tha t they can magnetize other ferromagnetic parts. Thts may sometimes be done merely by putting the piece to be magnetized in the external field around a permanent magnet. A more effective way is to touch or stroke the piece to be magnetized with one pole of a permanent magnet. Under such conditions, soft iron will be temporarily magnetized while a piece of hard iron o r steel will become a permanent magnet.

The Ea r t l i i s in effect a huge permanent magnet having a north magnetic pole just north of Prince of Wales island in Northern Canada, and a south magnetic pole in t h e Antarctic Ocean south of Australia, lust off the coast of the Antarctic Continent a t Com- monwealth Bay. The flux lines of the Earth's fieid c~ rc l e the globe in a generally north and south direction and, though with many devlations on the way, concentrate a t these magnetic poles. The Earth's field is quite weak compared to the strength of fields used in magnetic particle testing. I-Iowever, long bars of iron or steel can be magnetized to an appreciable degree by placing them for a period of time in a north and south direction-that is, parallel to the lines of force of the Earth's field. Magnetization is hastened by striking the bars with a hammer to vibrate them. The natural lodestones derived their magnetism from their location in the Earth's field.

2. MAGNETIZATION WITH PERMANENT MAGNETS. For magnetic particle testing purposes, permanent magnets can be, and sometimes a r e used to magnetize parts. This method of magnetization has many limitations and is properly used only when these limitations do not prevent the formation of satisfactory leakage fields a t discontinuities.

Permanent magnets in general are capable of setting up only what a r e essentially longitudinal fields. Jnvariably poles are present on the par ts so magnetized, due to the contact of the poles of the permanent magnet used for magnetization. Thls results in confusing

PRINCIPLES O F DIAGNETIC PAIZTICLE TESTING

adherence of particles a t such esternai poles. Control of fieid direction is possible over only limited areas. A srngie pole-north or soutll-of a permanent bar magnet set on end on the surface of a steel piate creates a sort of radial field in tlie piate around the pole. The flux of this field leaves tile plate surface a t some distance f rom the point of contact to return to the pole a t the opposite end of the magnet. Cracks which happen t o be crossed by such a fieid pattern will usually be indicated, provided the field produced in the plate is sufficiently strong. When the poles of a horse-shoe magnet or a permanent magnet yoke a r e placed upon the surface of a steei plate or part, the fieid travels through the plate o r par t f rom one pole of the magnet to t he other. Along a straight line drawn between the poles, the flux will he relatively straight and will be strongest near the poles of the magnet o r yoke. FieId strength along this line will vary, and be wealcest a t the point midway between the poles. The actual strength a t any point mill

I ..

Rg. 72-Magnetrzatlan Wlth a Permanent Magnet. a) Magnetiz~ng a Bar by Placlng One Pole of a Permanent Magnet at One End. b) M a ~ n e t ~ z ~ n g a Plate by Setting a Permanent Bar Magnet on End on the Plate.

Permanent Horse-shoe Magnet, or Yoke. Placed on the Surface of a Plate or Other Part

CnApm 7 METHODS A K D AlE.49S FOR GEKERATING 3IAGNETIC FIELDS

depend on the strength of the magnet and the distance between the poles. Cracks a t right angles ( o r nearly so) to thls line will be indicated, provided field strength is adequate. Outside this limited area the field spreads out, and cracks favorably located with respect to field direction may or may not be shown, again depending on the place where they occur. Figure 72 illustrates the uses of permanent magnets for magnetization of parts. This method of magnetization should be used only by experienced operators who a r e aware of and understand the lim,itations of the method.

Some of the other drawbacks t o the use of permanent magnets are 1) that i t is not possible to vary the strength of the field a t will, nor 2) can large areas or masses be magnetized with enough strength to produce satisfactory crack indications, and 3) if the magnet is very strong, i t may be difficult to remove it from contact with the part.

3. ELECTRIC CURRENTS FOR A'IAGNETIZATION. Use of electric currents is by f a r the best means for magnetizing par ts for magnetic particle testing purposes. Either longitudinal or circular fields can be set up easily; strength of field can be readily vaned I and by use of the several types of current, useful variations in field strength and distribution can be accomplished. E l ec t r~c current in one or another of its types, is used for magnetization for the purposes of magnetlc particle testing in all but a very small percentage of cases.

4. FIELD IN A N D AROUND A CONDUCTOR. When an electric current is.passed through a conductor, such a s a rod or wire, a magnetic fieid is set up in the conductor and in the space surrounding it. If the conductor is uniform and s t r a~gh t , the density of the field, or the number of magnetic lines of force, will be uniform along its length, and will decrease with increasing distance out from the conductor. The strength of the field will be directly proportional to the strength of the current flowing ( that is, the number of amperes). The lines of force take concentr~c circular paths around the rod o r wire, so that the field is circular and a t 90° to the axls of the conductor.

The field has directional properties, which depend on the direction in which the current is flowing. A simple rule fo r determining the direction of the field is to grasp the conductor ( no t necessarily literally while current is flowing!) with the right hand so tha t the thumb points in the direction of current flow. The fingers will then point in the direction of the fieid. Another way is to b r ~ n g a compass

143


MAGNETIC FIELD WIRE

I 1 I

Fig. 73-Field Around a Conductor Carryrng Direct Current, and its Direction.

needle lnto the field. The north pole of the compass wrll point in the direction of t h e lines of force. Fig. 73 illustrates diagrammatically the field around a conductor and its directlon.

5. LOOP. If the conductor carrying current is bent into a single loop, the lines of force surrounding the conductor will pass through the ioop, all in one direction. The field within the loop then has definite direction, corresponding to the direction of the lines running

I

I S INGLE LOOP I Fog 74-Reid in and Around a Loop Carrying Otrect Current. Shawtng Polarlty

through it. One slde of the loop will be a north pole and the other a south pole. See Fig. 74.

6. SOLENOID. If instead of only one tu rn , the conductor carrying current is looped a number of times, the coil o r solenoid will similarly be longitudinally magnetized-one end will be of north poiarity and the other south. The s t rength of the field passrng through the interior of the solenoid will be proportional to the

product of the current in amperes and the number of turns of wire in the solenoid-that is, the ampere turns. Thus the magnetizing force of such a cot1 can be varied either by varying the crlrreitt o r the

INES OF FORCE S -

I CURRENT SOLENOID \ I

Fig. 75-Field In and Around a Solenold Carryeng Cirect Current and its Direction.

number of tu rns in the coil. The measure of tile magnetizing strength of a coil is thus expressed in terms of A?rwcr.e Tul.ns.

7. YOKES. ElecLromagnetic yokes a r e U-shaped cores of soft Iron with a coil nround around the base of the U. When direct current is passing througll tile coil the Lure ends of the core a r e magnetized with opposite polarity, and the conlblnation is an electromagnetic yoke, similar to a permanent horseshoe magnet. Yoki?s a re sometimes used fo r creating Iongitiidinal fields in thc same way a s the permanent magnet yokes. D.C. yolces Iraue soiiie advantage over the latter in tha t their strength cat1 he varied by varytng the current in the coil; and also that they can be put on and removed from the pa r t in the uiimarnetized condition when no current is Howtng. Yokes magnetizer1 with alternating current also have numerous applications, and have the additioiial advantage tha t they can be used f o r demagnetizing a s well iis magnetizing.

8. S ~ L E N ~ I D S FOR MAGNETIZATION. AS has been sald, solelloids carryrng current a re the prefcrrerl menns foi.seltirig up longitudinal fields tn fe~ , ron~agne t ic parts. F o r large par ts the solenold may be wound directly around the part , using scvcr.al 1it t .n~ of llexilrle c;tbic. For sn~a l l e r parts, coils \vouiid on iixcd f rames a r c most oftetl useti.

9. EFFECT OF COIL DIAAIETER. Di;lrneLer ol' Lhc coil n.1111 reiattoii to the ilimenstons and shape of tllc p:trl bcln!: magtlclizctl I S ;L Inr!:~, factor 111 securing pr.ouet. magnctizatior~. In i~a~.t~colii i . , Lhc l~~n!:lh

145


and diameter of the par t niust. be considered in relation to the iengtll and diameter of the coil, \%,hell deciding on the amonnt of current to use for proper magnetization for magnetic particle testing purposes.

The ratio of the cross-sectional area of the part being magnetized, to the cross-sectional area of the coil is called "fill factor" \\'hen the coil is short low fill factors a r e necessary to produce a uniform field over the cross-sectron of the par t when the par t is placed in the center of the coil. The ratio of the par t diameter to coil diameter in general should not be greater than one to ten. If the par t 1s placed adjacent to the wall of the coil the fill factor becomes of less im- porhnce. This phase of magnetizing procedure will be discussed in more detail in Chapter 9.

TO. EFFECT OF COIL LENGTH. \Ylien a solenoid is used for the magnetization of long parts, best results might appear to be obtarned when the coil is a s long a s the part , thus forcing the field to traverse the par t f rom end to end. \tihen flexible cabie is used to form the coil in place on the par t over its entire iength, this can readily be done, although the technique is not very practical. I t is not much used in actual practice because the surface of the par t is hidden by the coil, and i t is therefore inaccessible fo r inspection, unless either the coil or tlie part is removed.

Coils for magnetizing purposes a r e usually short, especially those wound on fixed frames. But when a long par t is being magnetized In such a coil, i t will not be magnetized properly over i ts entire length unless either the coil or the pail. 1s moved. The lines of force

Fig. 76-Flux Path tor a Long Bar in a Short Coil.

146

CHAPTER 7 \lr: 'rlrilD~ AXD MEANS FOR GENER-\TING fiIAGNET1C FIELDS

will leave tlierr iongitudinal directron a t some distance on either side of the coil and return through the air on the path of least reluctance, consrsting of optimum amounts of a i r and won rn the circuit. See Fig. 76. Long bars must be moved through the coil, or the coil moved aiong the bar, and suverai "shots" of current through the coil given along i ts iength, with rnspection f o l l o ~ r ~ n g each shot. A single shot with the coil midway between the ends of a short bar may be sufficient.

1 I. ClncriLaR MAGNETIZATION. Longitndinal magnetization is used when the nature of the par t and the orientation of the defects being sought are such that the latter lie in a direction transverse to the long axis of the part. But because of the polarity associated with longitudinal magnetization external leakage fields a t the poles sometimes cause interfering background patterns. In circular magnetization there are seldom any esternal leakage fields except a t a discontinuity. Therefore, all other things belng equai, where a choice esrsts between longitudinal and clrcular magnetization, the latter is preferable.

When current is carried by a non-magnetic conductor, a circular field exists in and around the conductor; but if the conductor 1s of ferromagnetic material such a s iron o r steel, a very much stronger field is formed ?wide the conductor, whereas the field surrou?tdi?lg the magnetic conductor is the same as for the non-magnetic one. The strong field generated inside the ferromagnetic conductor is due to the high permeability of ferromagnetic material a s compared with copper o r other nonmagnetic metals.

When a bar i s thus magnetized, the internal fieid will be zero a t the exact center of the bar and increase to n maximum a t the s~ r r - face. If the conductor is a tube, there will be no field on the inner surface of the tube and maximum fieid strength a t the outer surface. Circular magnetism can be better lnduced in a tube by passing the current through a ce?lt?-a1 conductor threaded through the insrde of the tube. In such a case, field strength will be a maximum a t the znside surface of the tube, and some\vRat less a t the outside surface. A full discussion of field distribution in symmetrical objects will be found rn Chapter 9.

12. EFFECT OF PLACEMENT OF CENTRAL CONDUCTOR. If a central conductor is so placed tha t i t traverses the exact center of the tubular cross section the field in the tube will be symmetrrcal around

147

I'RINCIFLES OF XIAGNETIC PARTICLE TESTING

i ts cylindrical wall. If , however, the central conductor is placed adjacent to one point on the inner c~rcumference of the tube, the field in the tube wall will be much stI'0ngcr a t this point, and weaker a t the diametrically opposlte point. Fig. 77 shows some magneto.

Fig. 77-Magnetographs of Field Showrng Cross Sectlon of a Tube Magnetized a) With a Central Conductor Centrally Located. b) With the Conductor Located Adjacent to the Wall of the Tube.

graphs illustrating this situation, and although only the flux lines of the exter?zal fields are shown, some rdea of the unsymmetrical field is given by the illustration.

In the case of small tubes i t is desirable tha t the conductor be centrally placed so t ha t uniform fieids fo r the detection of discontinuities will exist a t all points on the tube's surface. In the case of larger diameter tubes, rings o r pressure vessels, however, the current necessary in the centrally placed conductor to produce fieids of adequate strength for proper inspection over the entire circumference become excessiveiy large. By placing the conductor adjacent t o the wall and leaving the current on, or giving a number of current "shots" a s the tube is rotated about i ts axis, a field of sufficient strength 1s produced with much smaller currents a t the points in the tube o r r ing wall adjacent to the conductor.

13. CIRCULAR FIELDS IN IRREGULAR-SHAPED PARTS. Both the direct contact methorl and the centrai conductor method are used to generate c~rcu la r fieids in all varieties of shapes and sizes of parts. Fo r small parts having no openings through the interlor, circular fields a r e produced by making contact directly to the part. This is cione by means of clamping devices, generally on a bench-like

e CHAPTER 7 METHODS AND MEANS FOR GENERATING MAGNETIC FIELDS

I unlt which also incorporates the source of current. Fig. 78 illustrates ; such an arrangement.

Fig. 7&Part Beeng Magnetrzed Circularly by Clampang It Between the Head Contacts ot a Magnetizing Unbt.

The contact heads must be so constructed tha t the surfaces of the par t are not damaged, either physically by pressure, or structul.ally by heat produced by arcing or high resistance a t the polnts of contact. Possible damage to finished par ts f rom t h ~ s cause 1s an important consideration and must constantly be borne in m ~ n d . It can be especially damag~ng to hardened surfaces such a s bearlng races.

Points of contact may have to be made a t several places for compiete inspection of an lrreguiarly shaped part, in order to get fields in the proper directions a t all points on the surface. This often necessitates several separate magnetizations and inspections. Mod- ern techniques ha~re greatly minimized the need for such multiple magnetizations by the use 01 the so-called "o~~era l l " magnetization method, by the use of multi-directional magnetization, or use of the induced current method of magnetiz~ng. Sec Chapters 10, 19, nnd24.

14. A ~ A G N E T I ~ A T I O N WIT11 PROD CONTACTS. Rlagnetization by passing current directly through the part, or through ;L local por-

PRINCIPLES OF 31.4GNETIC P.\RTICLE TESTING - -.

tron of the par t , is often resorted to fo r t h e inspection of large and niasslve articles too bulky to be put into a unit having clnnlping contact heads. Such local contacts do not al\rays produce t rue clrcuiar fields, bui a r e very convenient and oractical fo r many purposes. Inspections of large castings and weldnlfnts a re frequently made in th i s manner. Fig. 79 is a magnetograph s!io\ving the gen-

Fig. 79-Magnetograph of the Field Around and Between Prod Contacts.

era1 direction of the field when contacts a re made on the surface of a steel plate by means of two hand-held prods. The field here is somewhat similar to tha t produced by yoke magnets, except tha t in thls case i t is the current t h a t passes bet\veen the contact points, and the field crosses the a rea betxveen the contacts a t 90' to the current. '3racks parallel to the line between tlie prods will be shown by this nietliod of magnetization.

The method is widely used and h a s many advantages. Easy portability malies i t the most convenient method for field use fo r the inspection of large tanks and welded structures. Sensitivity to defects lying wholly beiow tlie surface is greater with this method of magnetizing than with any other, especially \vhen half wave current i s used.

The use of prod contacts has some disadvantages of which the operator should be aware :

150

CIIIPTEX 7

\II:THODS AYLI \ I l i N S Fllli (~I1ZEI1ATIXG M:\GXETIC F11:LDS

(1) I t is iiccessiiry to scan the suriace or tlie par t beillg inspected in small scctions, since suitable fieids esist only betjveen and near the protl contact. Since the snaciiigs a re seldoili greater than 12 niches and tisoally milch less, t h ~ s means many seoarate contacts and a time consuming lob. illoiiern technlqlles such a s the "o\<er:tiifl n~etliod previously referl.ed to, have in a great many cases replaced inspection \\.it11 prods, and have shortened the time for such inspections by orry large uercentages, of 30% or more.

( 2 ) I n t e r f ~ r e n c e of the esteynal fieid t h a t esists hc.t\veen tlie prods sometimes makes obser\,ation o i pertinent indici~tions diificiilt. The strengtli of tlie current that can be used 1s limited by this effect.

( 3 ) Great rare must be used to avoid lburning of the par t under the contact points. Busniiig may be caused by dirty cont;icts, 111- sufficient contact pressure o r escessive currents. The chance of such damage is particulariy great on steel of 30 to 40 poirlts carbon o r over. The heat under tlie contact points produces local spots of uery iiard material that can interfere with later operat~oiis, sucti 2s maciiiiiing. Sometimes actual cracks a re produced by t lus lieat- ing effect. Contact heating is not so likely to be damaging to low carbon steel such a s has been used fo r structural purposes.

15. EFFECT OF TYPE OF MAGNETIZING CURREKT. There a re basically two types of electric current in common use, and both a re suitable f o r magnetizing purposes for magnetic particle testing. These a r e Dtrect CZLTTC~L~ (D.C.) and .'iite~?zat?ng C?~rrent (-4.C.). The strength, direction and clistribution of the fields a re greatly affected by the type of current empioyed f o r magnetization. Under- standing the magnetizing characteristics of these types of current and the various modifications in their use, is of great importance fo r proper application of niagnetic particle testing..

16. DIRECT CURRENT vs ALTERNATING CURRENT. Direct current a s here being considered is a constant current flowing in one direction only. Alternating current is here considered to be commercial A.C., \vliich is current reversing i ts direction coinpletely a t the rate of 50 or 60 cycles per second. Commercial A.C. 1s referred to a s being 50 cycle or 60 cycle. One cycle coiisists of t\rro coniplcte reversais. See Fig. 81. There a r e also a few commercial services of 25 cycle A.C. 50 cycle A.C. is quite common in many countries, ~ ~ l i i l e in the IJniteri States the 60 cycle current is almost universally used.

151

[... ,~

I'RINClI'L13S OF hlriCNETlC PARTICLE TESTING

The magnetic lields produced by direct and by alternating current diff'er in many characteristics. The difference ~ \~hic l i is of prlnme importance in magnetic particle testing is that fields produced by direct current generally penetrate tlie entire cross section of the part, whereas the fields producer1 by alternating current a r e con- filled t o tlie metal a t and near the surface of the part.

Tlic phenomenon which causes A.C. to tend to flo\ir along only the surface lagers of metal of a conductor is known a s "skln effect" In nonmagnetic conductors the effect is not very noticeable untii frequencies much higher than 60 cycles per second a r e reached. In magnetic materials, however, the effect is tremendously greater, so that a t the coinmerclai frequency of 60 cycles, alternating current pass~ i ig through a ferromagnetic coiiductol. will be pretty well confined to the surface layers of metal. Circular magnetic fields generated under these conditions \!rill be similarly confined to surface or near-surface layers, a s though the current \!,ere traveling through a tube sheathing the balance of the cross-section of the conductor.

Careful esperiments have shown that in a steel shaf t six inches In diameter, magnetized by passing heavy alternating current of 60 cycle frequency from end to end, there was little or no field detectable more than one and one half inches below the surface, aim for all practical purposes the fieiil was entirely confined in the outer half ~ n c h of metal. From this, i t i s obv~ous tha t when deep penetration of the field into a par t is required-as, for instance, in tlie search for discontinuities lying !\'holly below the surface and possibly deep in the interior of the part-D.C. and not A.C. must be used as the source of magnetizing force.

17. SOURCES OF DIRECT CURRENT FOR ~IACNETIZING PURPOSES. CommeI'crai po\\.er IS today very rarcly furnished as direct current and need not be considered here as a practical soul.ce of D.C. Other possible sources are motor-generators; storage batteries, rectified A.C. and some very spec~al types of D.C. po\vcr supplies providing a very siiurt rluration of current flow.

18. MOTOR-GENERATORS AND RECTIFIERS AS SOURCES OF D.C. Motor-generators for welding and electroplating coinmonly furnish dii.ect current for these pi'ocesses a t voltaxes from around 6 o r 12 volts for plating, u p to around 75 volts for welrling. Current outputs for welding a r e iiommally about 300 amperes, though some gen-

CHAPTER 7 &IETHODS AND 3EANS FOR CENERATlNG BIAGNETi(. F I ~ ; L D ~ -

erators \\.it11 larger outputs a r e availabie. Plating generators or rectifiers for Large lnstallatlons may deliver much higher amperages, depending on the size of the installation.

Neither of these sources of D.C. IS very well adapted to furnish- ing magnetizing Current for magnetic particle testing. The high voltage of the welding generators means a wasteful expenditure of polrer .riihen high currents a r e being drawn. This same voltage means that any slipping of prod contacts can cause severe arcing and such Rashes can bevery damaging to the eyes of the operator a s well as to the parts being tested. The sudden application of constitutes practically a dead short on a welding generator damage the generator windings in a short time. Furthermore, for many magnetic particle testing applications, the maximum current so delivered is entirely insufficient fo r the inspection of many parts or welds.

Plating generators o r rectifiers delivering several hundred amperes D.C. a t from six to twelve volts, a r e free from troubies because of the low voltage; but thew output of current IS seldom large enough for most magnetizing purposes. ~f large enough, the installation woulil be very inflexible, and much mose costly than the other sources of D.C. ~ a g n e t i z l i l g currents available today.

19. STORAGE BATTERIES AS A %WRCE OF D.C. In the days before World War 11, storage batteries were \videly used a s a source of direct current for magnetic particle testing. A group of volt or 12 volt storage batteries connected In parallel will deliver many thousands of amperes momentarily when shor t -c i rcu~t~d tlirough a coil or through a part, and in many ways prov~de a very satisfactory source of direct current fo r magnetic particle testing purposes. In order to magnetize parts with D.C., i t is necessary f o r the current to flow for only a fraction of a second, and the drain on the.batteries for each shot i s therefore reiativfly small. The energy , ~ r i t h d ~ ~ , ~ ~ ~ by a shot may be a t once replaced by a charger of sufficient during the interval between shots.

Storage batteries, however, a r e Practically never used today for magnetic particle testing because of a number of undesirable char- acteslstics they possess. Storage batteries deteriorate if not care. fully maintained, and when equipment is used oniy a t irregujas

intervals, i t is often found that the batteries are "dolvn,, when occasion arises to use the unit. Also the fumes from the cliarglng

153

PR1NCIPI.GS OF R1:ttiSETIC F.kfI'I'ICLE TESTIXC

of the batteries a r e corrosive, a s is spilled L~attery acid, resulting In unsightly corrosion and damage to e q u ~ p n ~ e n t . One f l i r t i~e r disadvantage is that the high currents can be drawn only momentarily, whereas it 1s often necessary to have tlie current flo!~ fo r appreciabie ~ n t e r v a l s of time.

20. D.C. F R O h l ILECTIFIED A.C. Ey f a r the most satisfactory source of 13.C. is the rectification of alternating current. Both single phase and three phase A.C. a re furnlshetl c o m m e r c ~ ~ ~ l l y . (See Sec- tion 4 . Chanter 5 . ) I?? tlie use of rectifiers, the re\'erstng A.C. can be converted to u~u-directional current, and \\'hen three phase A.C. is so rectified, the delivered direct current is entirely the equi\salent of s t ra ight D.C. fo r magnetic particle testing purposes. The only difference between the two is a slight ripple in the value of the rectified current, amounting to only about 3% of the niaxlmum current value.

Advaiitages of such a source a re obvious. Maintenance problems and corrosion, which characterize the use of batteries a s a source of D.C., a r e elinunated. Equipment can be designed to deliver any desired current on any required work cycle. Variation of current value 1s relatively simple by the use. of tapped transformers, saturable-core reactors (po\\.er-stats) o r variable transformers. 111

the A.C. pa r t of the circuit. There a r e many variations in the application of different types of rectified A.C. \vluch a re highly useful, and \vhich could not be provided from any other source.

21. HALF WAVE RECTIFIED SINGLE PHASE A.C. Wheii single phase a l ternat ing current is passed througli a slmple rectifier, current is permitted to flow in one directioii only. The reverse half of each cycle is completely blocked out. The result is uni-directionai curi.ent which pulsates-that is, i t rises f rom zero to a maximum and drops baclc to zero. During the blocked-out reverse of the cycle no current flo~vs. Then the 11alf cycle forward pulse is repeated, and so on a t the ra te of GO puises per second.

When the half-rvave cui.rent Cfclivet~eti by S U C ~ I a circuit is

measured by a conventional D.C. meter, the meter reading \\#ill In- dicate the acerage current o i the ei2til.e c!~cl r . I-lornever, dur ing the non-conductive half cycle, the current is very nearly zero. I t is evident; then, tha t sircli a metcr is actually indicating only one half of the average ciirrent that flo~vs dlrrinfi the coniluctit~e half of each cycle. Tlus 1s not a trery realistic measurement of the effective

15.1

Cll iPTEn 7

RIET11ODS X U &IEANS FOli tiEXER:\I'ING -- \I.tGKETIC FIELDS -

magnetizing current level. Through the years it has been established t h a t the average current &!i.i!ig tlie c ~ ? i d l ( c t i ~ . e Aal i C ? I C / L ' IS a more representative measure of the eKectivc n~agnetization. The average current fo r tile conducti\.e half cycle is double that read on the n~e te r . -4 siinple solution is to doubie the meter scale, and continue to use the conventional D.C. meter. "hlagnetizing ampere" then replaces "ampere" a s the unit of meter calibration fo r half wave current. T l ~ i s is still less than tlie peak value of current dur ing t11e conductive cycle, wi~ ich 1s 3.11 I T ) times the average current. Since the meter scale has been doilbleit, the pcaii current of the conductive half cycle is 1.57 ( x / 2 ) times the "magnetizing ampere" reading.

There is a slight skin efTect due to the pulsations of the current. but th is i s not prono~inced enough to affect serio~isly tlie penetration of the field. The pulsation of the current is in fact useful, a s i t tends to impart some slight vibration to the magnetic pa~.ticirs, assisting them to arrange themselves to form indications.

The most predominant anplication f o r half-\\ya~e (H.W.) current is fo r \'eld inspection. In t lns appl icat~on, half \\3a\7e current is used with dry powder and prod magnetization, ~v11rcl1 combinat~on provides the highest sensitivity fo r discoi~tinuities \vhich lie \\,holly below the sul.face. In many eases suitabie amperages can be obtained f rom relativeiy small portable units. Half wave current is also \ v i d e i ~ used fo r casting ~nspection \\:hen sensitivity to siib-surface discontinuities is requrred. See Cl~ap te r s 13 and 2.1.

22. FULL WAVE RECTIFIED SINGLE PlIASE A.C. This SOUrCe of pulsating uni-directional current is sometimes used fo r magnetic particle testing purposes fo r certain special-purpose applications. In general it possesses no advantage over half wave, and is not a s satisfactory a s three phase rectified current \\,hen strai&!ht D.C. IS

required, because of its extreme "rippie" Also it clraws a higher current f rom the A.C. line than does half \rrave fo r a given magnetizing effect, which 1s a d i s t ~ n c t disodvantnge.

23. T H E SURGE METHOD OF &IAGNETIZATION. There is another way to get stronger fields in a par t when the current must be ieft on fo r more than a fraction of a seconci. A rectitier unit capable of rlelinerrng 400 amperes continuously c;in put out much higher currents fo r shol.t intervals. I t is therefore possible by suitable current control and switching de\~ices, to pass a very high current fo r a shor t period-less than a seconcl-and then reduce tile current,

155

PRINCIPLES O F MAGNETIC PARTICLE TESTING i

without interrupting it, to a much lower steady value. Figure 80 illustrates how this effect 1s produced. In the figure, the current value for both initial and steady currents a r e projected onto the hysteresis loop, and from it the values of the fields tha t result a r e indicated. The method is called the "surge" method of magnetization. The "surge" current carries the magnetization of the unmagnetized par t high on the hysteresis loop t o point "A" When the current drops back to a steady value, magnetization drops along the hysteresis curve to paint "M", resulting in a stronger fieid than the same steady current, without the initial surge, can produce. In the latter case the magnetization follows the "virgln" curve, and the field strength rises only to point "Nu.

Fig. 80-Effect ot a Short Surge of High Current Followed by a Drop to a Steady Current of Lower Value.

The method was formerly widely and usefully employed for weld inspection, although today the half-wave system has almost universally replaced it.

24. THREE PHASE RECTIFIED A.C. By f a r the most useful and most widely used source of direct current for magnetic particle testing 1s rectified three phase A.C. Three phase current is generally used for power equipment in most plants and 1s preferred over

CHAPTER 7 METHODS .%KD ME.iNS FOR GISNERATING .\IAGNETIC FIEI.DS

single phase current because of more favorable power transmission and line load characteristics. From the magnetic particle testing point of v ~ e w it is also preferred because it delivers, when rectified, current ~vhich for all practical purposes 1s direct current, and produces all the effects which a r e required when D.C. magnetization is indicated. (See Fig. 59, Chapter 5.)

From both des~gn and operational vielvpoints this system has many advantages. Switching and control devices a r e operated a t A.C. line soltage, which may be 230 or 460 volts or higher; befo?.e the voltage is stepped down to the low values a t which the rectified current must be delivered. By simply increasing the size of transformers and rectifiers, units delivering 20,000 amperes D.C. a t low voltage have been built. Much of the equipment that will be described and discussed later in this book will be of this type.

25. SOURCES OF ALTERNATING CURRENT. Alternating current, which must be slngle phase when used direct& for magnet iz~ng purposes, is of course taken from con~mercial power lines, and is usually 50 or 60 cycles per second in frequency. When used for magnetizing purposes, the line voltage, which may be 115; 230 o r 460 volts, is stepped down by means of transformers to the low voltages required. A t these low voltages, magnetizing currents up to several thousand amperes a r e often used. Switching and current control devices, as in the case of D.C. rectifier units, are operated a t line voltages before step-down.

26. PERMANENT MAGNETIZATION WITH A.C. One of the problems in the use of A.C. is the fact that, normally, the par t is not necessarily left with the level of residual magnetism which the peak current of the A.C. cycle generates. When the current is broken the remanent magnetism will depend on the par t of the cycle a t which the current 1s when interrupted.

Fig. 81 is a drawing of the sine wave form of a single phase alternating current. The field generated in a par t by such a current will rise durlng the first quarter of the cycle till the current reaches the peak a t point "b". As the current drops to zero a t point "c", the field in the par t drops to the ievel of residual magnetism whrch it can retain. As soon a s the current reverses, however, the magnetizing cycle also reverses and the field diminishes to zero; and then increases to a negative maximum when the current reaches point "d" on the reverse cycle. The problem then, if it is deslred to leave the par t permanently magnetized, is to hrealt the current on

157

PRINCIPLES O F kl.4GNETIC P:+RTICLE TESTING --

I hg. 81-Slne Wave Form ot Sxngle Phase Alternat~ng Current.

that portion of the cycle betx~een points "b" and "c" l o r between "d" and "a" on the reverse cycle). This may be accomplished by t he a r c which is drawn between the switch points when the current 1s broken. As the a r c hecomes long i t 1s quenched a t the point of

Fig. 82-Oscillograph at the Voltage Across the Switch Po~nts When a Single Phase A.C. Circuit Is Broken, Show~ng the Arc Quenched

at the Pmnt of Zero Voltage.

pllase A.C. break. Since the current lags beh~nd the voltage by one- quarter of a cycle, if the voltage i s interrupted a t the zero point of i ts cycle, the current will be broken a t the peak (point "b" or "d"),

thus leaving the masimum rcsldnal magnetism in t!ie part. Com- plete consistency in breiiking the cycle a t zero roltage is not obtained with ordinary stisitch gear. An a r c must be drawn between breaker points t o secure tliis result. S~vitcll gear is iisually designed to suppress such arclng mt!~er than to allou* the arc to be dralvn. Further , the char:lcter of the electric circuit affects the polnt a t \\shich the arc 1s quenched. As a result of these uncertainties, when the sr!s~dual method is being used, fnil-\'a\ae rectifitcl singlr phase A.C. is emnioyed. Th i s form of rectified A.C. is l<no<ran to give 10041, assurance of the niaximuni res~dual field in the part.

27. SICIN EFFECT. The skin effect when niagnetizing \irith A.C. has already been described (Section 16). For magnetic particle testing purposes this can he turned to adrantage if only surface cracks a r e being sought. Eecanse the po\r.er used and the heating effect on both par t and equipment a r e dptermincd by the "R.RI.S."* value of the voltage and current, equipment can be lighter and use less power than a D.C. unit of comparable magnetizing capacity. Jn man); cases surface cracks a r e the only interest, a s for instance In overhaui or nialntenance inspection where location of fatigue cracks i s the objective. Therefore fields confined Lo the surface layers of metal are entirely satisfactory and des~rable for this uuri)ose.

28. R ~ A G N E T I Z I N G WITH TRANSIENT CURRENTS. \Vhcn a direct current in a elrcuit is suddenly cut off, the field surrounding the conductors collapses, or falls rapidly to zero. The raprd change of field tends to generate a voltage (and current) \\,bich is oppos~te in direction to that \vli~ch had been established in the cll.cuit. When ferromagnetic mate r~a l IS under the influence of such a coilaps~ng field the eff'ect is mucii increased. Under certaln conditions the rapid collapse of the iield can generate very high ciirrents inside the ferromagnetic material, and the phenomcnoii can be niade useful In some magnetizmg ~ rob l ems .

Fo r example, when a bar is magnctizcd iongiludinally in a solenoirl, the two ends become poles and the flux lines leave and enter the bar a t 90" to the surface. The near-end portions of the bar ;ii.e therefore not truly longitudinally magnetized, and transverse cracks in this zone \\,ill probably not gi\.e reatlable indications.

'Alternating cui.i.cnt nleters nrr read in "R.3l.S." vnlucs. T h ~ s nieans "root of meall ssuarcs". and representi t l ~ e true measure of the po,rer ronsumptlon in the A.C. c i i c u i l .

159

PRINCIPLES OF iVlAGNETlC PARTICLE TESTING

When the current in the solenoid is interrupted with suffic~ent rapidity, however, a translent current is generated whrch flo~vs c~rcular ly Inside the bar, and generates a sor t of doughnut-shaped field close to the surface, whlch i s truiy longitudinal clear to the ends of the bar. The effect has been called "quick break" or "fast break", and is built Into quality D.C. magnet iz~ng units today. Figure 83 illustrates this effect diagramaticall$.

I

b.

Fig. 83-Residual Field lnsldc a Bar Generated by a) Slow Break, b) "Quick Break'' Translent Current.

29. INDUCED CURRENT MAGNETIZATION. An extremely useful application of this collaps~ng field method of magnetization has been developed for the magnetizing of r~ng-shaped par ts such as bearlng races, without the need to maire direct contact with the sur-

CHAPTER 7 3IETHODS A N D 'IIEANS FOR GENERATING DfAGNETlC FIELDS

1 CIRCULAR MAGNETIC FIELD 1

D E F E C T S ~ MAGNETIZING CURRENT

Fig. 84---Direct Contact Method ot Magnetiz~ng Ring.Shaped Parts to Locate Circumterential Defects.

PRIMARY

f l n

CURRENT CORE ~p

Fig. 8 L l n d u c e d Current Method ot Magnetizing Ring.Shaped Parts to Locate Circumferential Defects.


face of t he part. Figure 84 sho\lTs the direct contact metliod for producing circuiar fields in a ring to indica1.e circumferential cracks. T o obtain reliable testing of the entire cj'lindricai surface two magnetizations are required, because tlie areas near the points of contact where the current enters and leaves the ring a r e not properly magnetized for this purpose. The ring must be turned 90° and then retested.

By making the rlng the one-turn, short circuited secondary of a "D.C. transformer" a large current flowing circuniferentially around tlie ring can be induced. Figure 85 illustrates this effect. To accomplish this a standard magnetizing coil can be used. The r ing should be piaced inside the coil with i ts axis parallel to that of the coil. When the coil is energized this arrangement constitutes a n air-core transformer, the magnetizing coil being the primary and the ring-shaped par t a single-turn secondary. The total current induced i n the r ing is greatly Increased by inserting a laminated core of ferromagnetic materlai through the ring.

The choice of type of magnetizing current f o r this method is determined by the magnetic properties of the material to be tested. Fo r par ts which have high retentivity, D.C. with "quick-break" i s

i usually used and the par ts tested by the resldual method. When the D.C. field i s caused to collapse suddenly by tlie abrupt interruption 1 of the magnetizing current, the circular field generated by the

1 transient current so produced leaves the par t with a strong residual field. A bearing race is a good example of the type of par t that can be tested advantageously by this method.

For par ts tha t a r e soft, with little o r no retentivity, the continuous method must be used. The collapsing D.C. field method is not applicable in this case, since the magnetizing current must flow for a n appreciable length of time. By using A.C. or half-wave in the m a p e t i z i n g coil, however, the current may be left on, and a n A.C. o r half-wave current is induced in the ring, of the same frequency a s the magnetizing current. This current can bc allo\ved to flo\il long enough to produce indications by the continuotis metliod.

Regardless of whether D.C. and the residual metl~od, o r A.C. o r half-wave and the continuous method is employed, the induced current method 1s usually faster and much more satisfactory than the contact method. Only one operation is required ancl the possibility of damaging tlie par t due t o arclng is co~npletely eliminated, since no external contacts a r e made to the part .

CH,WTER I

JIETBODS AND SIE:%NS FOR GENERATING MAGNETIC FIELDS

30. FLASH MAGNETIZATION. There a r e several ways of generat- Ing high values of current fo r very short intervals tha t ha\.e some special applications. These methods have some interesting possibilities, although; to date, they have not been used very often in the field of magnetic particle testing. One example is the use of A.C. (or D.C.) a t line voltage, passed through a par t , o r a solenoid around the part. I n the circuit is an ordinary fuse and a switch. When the switch is closed the line is shorted through tlie par t or the coii and current rises t o a high value but is quickly cut off again when the fuse burns out. Nuch higher currents can thus be obtained from line voltages xsitliout the use of transformers than would otherwise be possible. Pa r t s a r e left \i,ith high residual magnetism, since the arc produced wiien the fuse burns out 1s one that causes the A.C. to be stopped a t the peak of the cycle. The method is not very practical, especially a t high line voltages, and this magnetizing technique is not generally recommended.

.4notlier example is the use of tlie discharge current from a large condenser. Such discharges may rise to several thousand amperes for durations of from one ten-thousandth to one one-hundred- thousandth of a second. The condenser is charged to a high voltage by means of a rectified A.C. circuit, and discharged when the voltage rises to a pre-determined point.

31. SUITABLE FIELD STRENGTHS FOR ~IAGNETIC PARTICLE TEST- I N G Over the years some rule of thumb values for suitable current strengths for magnetic particle testing have been in use. These have been based on experience rather than on any measured determination of what the requirements actually are. Fo r circular magnetization it has been customaiy to use the figure of 1000 amperes per inch of diameter of the par t being magnetized. If literally followed, this figure rluickly gets one into trouble, and must be radically modified in most applications. I t is ap t to call for currents too high for small diameter parts, and for large diameters is usually un- attainable and unnecessary. However, i t is pretty safe to say that if this amount of current is used, satisfactory field strengths a r e most likely to result.

No comparable figlire fo r coil magnetization had been determined until quite recently. 200 ampere-turns per inch of coil diameter has been suggested and used, and in many cases has been a fairly good guide.

The trouble with these rules of thumb 1s mainly that they take

PRINCIPLES OF BIAGNETIC PARTICLE TESTING

no account of varying shapes and magnetic properties of parts which are to be magnetized. Permeability of steels varles widely, and a soft steel of hlgh permeability requires a f a r lower magnetizing force to produce a suitable field than does a high cal-bon or alloy steel which may have a very low permeability.

In recent years much work has been done in the laboratory to establish the current requirements more realistically, based on actual conditions met when steel parts of varying composition, slzes and shapes are to be magnetized. The effect of the varying properties of steels, as well as the sizes and shapes of parts, on the strength and distribution of magnetic fields will be discussed in Chapter 9. The methods of arriving a t suitable current values to produce fieids for satisfactory magnetic particle testing under these varying conditions will also he given in that chapter.

DETERMINATION OF FIELD STRENGTH AND DISTRIBUTION

1 . IMPORTANCE OF IINOWING FIELD STRENGTH AND DISTRIBU- TION. It has been emphasized in the two foregoing chapters that field strengths must be adequate and field direction favorable, in relation to the size and direction of the discontinuity, in order that a good Indication be produced with magnetic particles. It is, therefore, obviously important that the operator applying the magnetic particle method know how strong the field is inside the part being tested; and also the direction which it has taken a t any point where a defect is being looked for. In other words, he should know the field distribution and strength, ins~de the part.

The need to know exists and is important; bu t the nzeans ior determining field strength and distribution inside a part with any degree of accuracy, are less than satisfactory.

2. MEASUREMENT OF FIELD INSIDE A PART. There is no generally applicable method known today which permits the exact measurement of field rntensity a t a given point zuitl~zn any given plece of magnetized iron or steel. In order to measure field strength i t 1s necessary to intercept or cut the flux lines, and this it 1s impossible to do without cutting the steel itself. Cutting the steel would a t once change the value of the field density ins~de the part.

Explorations of flux density inside a piece of magnetized steel have been carrled out by drilling a small hole into or through the plece, and inserting a sn~all rotatable coil or other measuring devlce. The field strength actually being measured by this procedure IS, however, not the field inside the metal, but only that which jumps the air zap of the drilled hole. Such measurements are useful in indicating the gradients of field intensity, as for Instance from surface to center of a part; but the actual densities obtamed are comparative only.

3. EXPERIMENTAL FIELD MEASURING TECHNIQUES. There are of course methods for measuring magnetic fields and these will be described. However, only one of these-the flux meter-measures the flux density tnszde a part; and this measurement must be made


on test speciincils of suitable specla1 shape. Tile procedure 1s not useable fo r a point to point measurement of field strength ~ns ide a par t aiready magnetized. I t Ilas some useful applications but is not an anslrrer to the general problem.

4. FLUX METERS. The flux meter is an instrument t ha t measures the total cl~a?i,ge of flzlz through a coil, independent of the ?'ate of change. I t consists of a coil connected to a ballistic galvanometer through suitable lonptime-constant circuitry. When the magnetic flux through tile coil changes, either by movlng the coil or the part , or otherwise varylng the field in the par t by some means; the needle of the galvanometer swings and indicates the amount of change. The change in flux may be in either the positive o r the negative direction.

If the flux is zero to begin \irith and is increased to a value "A" the total change will be the flus a t the value ".4". If the area of the coil 1s kno\irn, tlic aserageflux density within tlie coil-that is, tlie average flux per unit of area-can be readily caiculated.

This device measures the avenge flux, and consequently tlie flux oensity, througil the coil, provided the flux through the coil 1s zero either a t the s t a r t or a t the end of the measurement.

If an unmagnetized piece of ferromagnetic ma t e r~a l is magnetized ivhile in the coil, the total change of flux 1s mdicated by the meter. Or a piece of steel, already magnetized, may he inserted from outside the coil. The total change of flux, from zero to the flux in the par t mill be indicated. Or, the magnetized par t may be a t rest ~ns lde the coil anil then be withdrawn, reducing the flux through the coil to zero. A g a ~ n , the meter registers the change, which 1s the total flux in the nart.

The coil is separate from the meter, and is \\round by the iiser to suit tlie particuiar prc:iem a t hand. I t 1s connected to the flux meter by a p a ~ r of eiectrical leads. The flux measured is tha t parallel to the axis of, and within the coil. Therefore the direction of the field must also be known before a n1easurement can be madc. In an

C I l A P T 5 8 DETTEII)IINATION OF FIELD Sl'ILENGTEI A S D DISTIIII3C1'106

to the cross-section. Figure 8G illustrates this arrangement. Gen- erally the coil should be ~ n a d e to tit snugly; ho\vever. little error is introduced by coil fit ui~less tlie tit is yery poor, since, due t o the lugh permeability of the steel compared to alr , neariy all the flux will be in the steel ring.

The change in flux 1s obtamed by first setting tlie meter to zero with the magnetizmg current off, and then noting the reading xvhen the current is turned on. The value of the flus 1s obtained by the formula

I< x Deflection Flux =

Number of Turns

I< 1s a constant for the meter used, which in the example is 101 D 1s the deflection read on tlle galvanometer N IS the number of turns in the coil

The flux density 1s obta~ned by dividing the total flux by the cross-section of the coil

Flux Density = B = Coil Area

A IS the area of the coil. The result IS in terms of lines per unit of area. If the cross-section is expressed in square centimeters, the resulting vaiue of flux density will be in Gausses.

The instrument may be used to measure the flux density in a steel plate by drilling small holes through the plate so that ... a flux

Irregular shaped par t thc direction of the field is not easily de- I. meter coil can he mound through the lioles. The meter will measure termined. But with a s~mnle t r ica l part , such a s a benrinr: race. the flux a t right angles to a line between the holes and within the . whleh has uniform cross-section around its circumference, the direction of the field when magnetized with a central conductor is circumferentiai. In such a case the n~easu ren~en t of the flux is very accurate. The coil is wound around the race so a s to be perpendicular

coil when the plate is magnetized.

5. MAGNETIC FIELD METERS. Many field meters are available w h ~ c h measure tlie niagnetic field in air. As used in deterni~nlng field distribution in magnetic particle testing these meters almost


.

GALVANOMETER

Fig. 86--Flux Meter Arranged to Measure the Flux In a Bear~ng Race.

always measure " H (magnetic field strength) .rather than "B" lflux density), even though the meter may be calibrated in Gausses which is the unit of flux density. See the definitions of H and B in Chapter 5, Sections 3161, 319) and 3115). If the instrument is direction sensitive - f o r example, based on the Hall effect (Chapter 5, Section 317) ), then i t may be used to measure the tangential component of H a t the surface of tile part. The tangential component of H is tha t component of the field having direction parallel to the surface of the part. Since the value of the tangential component of H is tlle same on either side of the boundary hetween the steel and the alr, the value measured in a i r is also a measure of the field strength close to the surface inside the pai-t. If the permeability of the steel is known the flux density can be calculated.

168

CHAPTER 8

DETERDIINATION OF FIELD STRENGTH AND DISTRIBUTION

Instrumentation of this type is useful whether or not the permeability is known, because i t does give information on the direction of the field, and because i t can be used a s a con~parative measuring device. Thus, if the level of magnetization required for tlle satisfactory testing of a par t has been determined by trial, the tangential component of H can be measured for tha t field a t a given point on the surface of the part. By requiring tha t similar pieces when magnetized give the same reading on the meter a t tha t same polnt, assurance can be had that all pieces of that type will be tested a t the same level of magnetism, even when the test is made by different operators and on different equipment.

Meters \vhich measure H, and are sensitive to its direction, can be based on a variety of principles, of which the Hall effect is only one. One of the simplest meters is based on the force exerted on a permanent magnet placed in the magnetic field. The magnet will tend to align itself in the direction of the field-as, for example, the magnetic compass-and, if restrained by a spring, can yield quantitative information regarding the strength and direction of the field. Such magnets can be made very small, can be attached to a r ~ g l d shaf t and located some distance from the meter movement. Such devices have been used t o explore the field distribution Inside large parts, by drilling holes into the par t and slipping the magnet down the hole, taking readings a s a function of depth.

I t should be understood, however, tha t this technique of drilling holes into a sample and then inserting a field-sensitive device into the hoie, does not yield the true value of the field inside the par t , since the removal of the metal in the drilling of the hole causes a redistribution of the field in the same manner a s when caused by a (sought-for) discontinuity. The measurement is still made "outside" the part .

Another version of thts kind of field measuring device utilizes a soft iron vane which will, when placed in a magnetic field, tend to align itself with the direction of the field. This tendency can be made quantitative by restraining the motion of the vane with a spr ing or with the field of a permanent magnet, and transmitting the movement of the vane to a pointer and scale. Figure 87 is a sketch of such a n instrument.

6. IYIAGNETOGRAPHS. Magnetographs are made by placing a paper over a magnetized plecc of lron or steel. or around a conductor o r coil carrylng current, and sprinkling fine Iron particles over the

PRIKCIPLES OF MAGNETIC PARTICLE TESTING -

ELLIPTIC SHAPED SOFT IRON PIECE ATTACHED TO NEEOLE. 1

Fig. 87-Diagram ot Soft lron Vane Type of Magnetic Field Meter.

paper. Such magnetographs can be used to give some idea of how fields a r e distfibuted inside and around the par t or conductor. The patterns of the magnetogmpi?~, however, a r e merely pictures of the fieid distribution outside the part , and give oniy indirect information a s to how the fieid may be distributed znside. However, even though this is true, some inferences can be drawn from the mag- netograpli pattern, a s illustrated in Fig. 88. Here a piece of soft iron has been placed adjacent t o a conductor carrying direct .current. The pattern shows lines entering and leaving the iron, and the path can be pretty well inferred.

7. FLUX SHUNTING DEVICES. A varlet)' of devices have been developed for tile intended purpose of insuring that the field distribution in a particular par t b a n g tested 1s of proper magnitude and direction. These devices operate by attempting to "shunt" some of the field out through the surface into a n external test piece and back through the surface again. One of these devices 1s known a s the Berthold Field Gauge. Another, developed in Japan, is called

170

CHnmar a DETTEH~IINATIOK OF FIELD STRENGTII .4ND DISTRIBCTIOK

Fig. 8GMagnetograph ot the Field Around an lron Piece Placed Adiacent to a Conductor Carryvng Direct Current.

the "Magnetization Indicator", and is reported in the proceedings of the Third International Conference on Nondestructive Testing ~ ~ h i c i i was held in Tokio, Japan, in March of 1960. These indicators cons~s t of soft iron pieces into each of whlch have been machined an "artificial defect" in the form of a straight siot. I n use; the indicator is placed on the par t to be inspected so that tlie artificial defect is in the direction of the cracks that a r e expected to occur in the part. The par t 1s then magnetized and the magnetic particles applied in the usual manner. If the artificial defect in the indicator IS sliomn, then magnetizatioii 1s considered to be proper. The proper level of sensitivity for varlous sizes of defects is achieved by varying the width and depth of the artificial defect.

If properly used these devices can be very vaiuabie. I t should be born in mind however, that if the continuous method 1s being used, then the indicator is more truly indicating H, the magnetizing force through the indicator, than i t 1s the field strength, B, through the surface layers of tlie magnetized part. On a symmetrical par t the in-

171


dieation would be the same on copper, aluminum or other non-magnetic material as it would be on a magnetic part.

AS long as the operator understands that it is H that is being measured and that he is not actually measuring the field inside the part, the device can be useful for assuring that the magnetization of the part is of the order of strength required.

8. CALCULATION OF FIELD DISTRIBUTION. I t would be convenient if the field distribution within a part could be calculated from dimensional and magnetizing current data. However, no method is available for doing this on any broadly useful basis. In some cases it is possible to obtain a magnetic fieid distribution plot by calculation from theoretical considerations. In general, however, calculation of field distribution is limited to a few relatively s~mpie shapes. As a practical means of arriving a t field distribution inside a part for magnetic particle testing purposes, calculations of this sort are of very little value. Still, a brief review of mathematical approaches may be helpful in understanding what the problem really is.

The mathematical solutions of field distribution problems usually involves the use of Laplace's equation. This mathematical procedure is difficult and cumbersome, and not to be undertaken as an exercise by any but a skilled mathematician. For such, the calculations necessary to arrive a t the field distributions for example, for a magnetic sphere and a magnetic elipsoid in a uniform magnetic field, as plotted in Fig. 89, would still be laborious but perhaps not too difficult. Figure 89 shows the calculated field distribution both inside and outside the sphere and the elipsoid. In both cases the field inside is uniform. The set of conditions assumed for this case apply for coil or longitudinal magnetization. Details of the mathematical caicula- tions ulill be found in reference 8, cited a t the end of this chapter.

Mathematical methods are also applicable where parts are magnetized by the direct contact or central conductor method (that is, circular magnetization) provided the geometrical shape of the part is extremeiy simple. Figure 90 shows plots of the calculated fieid distribution in and around two magnetic tubes carrying direct current, having two different degrees of eccentricity.

9. TRAN~FORMATION METHODS. The transformation method is frequently called the conformal mapping method. It is based on the fact that there exists a class of compiex mathematical functions, known as analytical functions, whose real and imaginary parts

112

CHAPTER 8 DETERI)lINATION OF FIELD STRENGTH AND DlSTRIBUTlON

Fig. 89-Calculated Field Distribultion In and Around a Magnetic Sphere and an Elipsold in a Uniform Magnetic Field.

are separate solutions of LaPlace's equations, and that these functions can be transformed (hy changing the variable) while still retain~ng their analytical propeikies. This means that a solution of LaPlace's equation which satisfies a particular set of conditions, van be transformed into a solution which fits other sets of conditions, if only the transformation can be found which changes the conditions of the original probiem into the conditions of the problem the soiution of which is desired. A simple example is shown in Fig. 91, where the proper transformation changes the plot of a simple parallel field distribution just under the fiat surface of a piece of steel ( a ) into the flux distribution a t a right angle corner in a large piece of steel (b) . Many elaborate transformation functions have been worked out and are compiied in "dictionarles". The method applies in many other areas of engineering as well as magnetic field analysis.

Fig. 90-a) Calculated Field Distribution in and Around a Ferromagnetic Tube ot 2. Flux lines generally enter or leave a magnetic surface a t Moderate Eccentricity. Carry~ng Direct Current. rlglrt angles t o that surface.

b) Calculated Field Distribution In and Around a Ferromagnetic Tube ot Extreme 3. Flus and potential patterns should be drawn a s cursitinear Eccentrlcity. Carrymg Direct Current.

squares-that IS, four-s~ded figures with rlght angles a t the


corners and with equal average lengths of opposite sldes. These squares are drawn such that

{a ) The same potential difference exlsts across each square, (b) The same flux passes through each square.

Examples a re shown in Fig. 92, la) and (b) . Fig. 92-a g ~ v e s the sketch for an Iron bar having a change of cross-section, and Fig. 92-b gives the pai t~al ly finished sketch for a rectangular bar having a notch.

FLUX TUBES

1 b. AIR eo I Fig. 92-Sketches of Fields for

a) An Iran Bar Havlng a Change ~n Cross.Section. b) A Rectangular Bar Havlng a Notch (Sketch Only Partially Complete)

BIBLIOGRAPHY FOR CHAPTER 8 I. NEHARI, Z., "Confornaal Mappz?zg." McGraw-Hill 11952). 2. CHURCHILL, R. V., "Introdt~ction to Complex Variables." McGraw-Hill (1948).

DETERRllNATlON OF FIELD STRENGTH A N D DISTRIBUTION

3. KOBER. H., "Dictionary of Conformal Representations." Dover (1952).

4. KRAUS, JOHN D., "Eloctronzagnetzc~." McGraw-Hill (1953).

5. PAGE, L. AND ADAMS, Y. I.. "P+z?~czpLes ot Elect.rzczty." Van Nostrand i 1949).

6. WEBER, ERNST, "Electromagnetic Fields, Vol. I , Mappang o f Fields." John Wiley and Sons (1950).

7. SMYTHE, W. R., "Static and Dyna?nic Electricity." McGraw- Hill (1950).

8. MCCLURG: G. O., "Magnetic Field Distribution for a Sp l~ere and an Elipsoid." Amer. Jour. of Physics, 1701. 24, No. 7, pp 496-499. October, 1956. and Ibid: Vol. 25, No. 4, P 266. April, 1957.

9. MCCLURG, G. O., "Theonj and Application o f Coil Magaetiza- tion." Journal Soc. for Nondestructive Testing. Jan.-Feb., 1955, pp 23-25.

CHAPTER 9

FIELD STRENGTH AhTD DISTRIBUTIOX IN SYRlnlETRICAL OBJECTS

1. INTRODUCTION. In the previous chapter we have considered available methods for measuring the strength, direction and distribution of magnetic fieids in and around magnetized ferromagnetic materials without, however, taking into account how these fields were produced. We have also surveyed the possible methods of arriving a t field distribution by calculation or other theoretical procedures. I t is apparent tha t there a r e no quick and easy methods for accomplishing either objective for the purposes of magnetic particle testing. An approach is possible by inference and extrapo- lation from the information provided by the use of various field- measuring devices which give the strength and direction of the field in the a i r outside the magnetized part.

Fortunately, the requirements fo r fieid strength for magnetic particle testing a r e not particularly critical, and in most cases a considerable range is permissible. Experience has built up a large amount of "knolr.-how" f o r tlie proper magnetization of various types of material, and of sizes and shapes of parts, and for the use of available magnetizing means.

\2. ELECTRO-MAGNETIC FIELDS. Magnetization of par ts fo r magnetic particle testing is accomplished, with very few exceptions, by the use of electric current. The exceptions a r e those few cases where permaiient magnets a r e used-usoally when electric current is not available, as in some inspections in the fieid, o r where explosion hazards exist in a plant and use of electricity is excluded.

Magnetic fields in and around conductors carrying current foi- low certain iaws which a r e \veil understood. These Ia\vs a r e very lieipful in arriving a t approximate values for field strengths under various conditions, including the directions and distributions of these fields. In the following sections we will discuss the beliavior of fields in and around coniluctors carrying currents, and how this behavior affects proper application of magnetization for. magnetic particle testing purposes. The discussion will be c~nf ined to mag-

CHAPTEII 9

F1EI.D STRENGTH A N D DISTRIBCTION IN SYYlMETRlCAL OBJECTS

netic and non-magnetic conductors ~vliich are reiatively small in size and symmetrical in shape.

. PERMEABILITY OF h l ~ ~ ~ ~ ~ ~ ~ ~IXTERIALS. The term "permeability" is constantly used in niagnetization discussioiis in connection with magnetic particie testing. The coiicept of permeability is, hoxrs- ever, very often misunderstood. The question, "What is the permeability of a certain kind of stcel'?" cannot be answered in a slmple statement. There is no such thing as the perme:tbiiity of a piece of steel. Permeability is a ratio-B/I-I-and therefore varies a s 1-1 is

varied. Quite a few permeabilities have heen defined, such as material permeability, l i~asimum permeability, initial permeability, and apparsnt permeability among others.

4. MATERIAL PERMEABILITY. 111 magnetic particie inspection with CLI-czr1u.1. n~a.gnefizntion it is nzaterzal permeability whidi is of interest. Ey material permeability we mean the ratio of the fiux density, B, to the magnetizing force, 1-1, where tlie flux density and magnetizing force are measured when the flux path 1s entirely within the material. In such ex.periments one measures H, the magnetizing force, and B, the flux density produced by thpt magnetizing force, point by point for the entire magnetizat~on curve, tising a flux meter and a prepared specimen of tlie material. Figure 93-a sliows such a curve, plotted for a hypothetical steel having a relatively lour material permeability. On tlus figure the heavy curve is called the normal magnetization curve. Figure 93-b sl~o\\$s a point by point plot of the ratio of B to 13. This curve sliows how the material permeability, ,A, varies as a function of H. Notice that

varies from sonie initial value to a maximum and then tapers off to an eventual value of one a t saturation.

For magnetic particie testing, the level of magnetization is generally chosen to be just-belorn tile knee of the magnetization curve. As can be seen in Fikures 95, -a and -6, the niaximuiil permeabi!itJv occurs near this point. Coiisequentlg, fo r circular magnetization we a r e interested in the iliaz?nviL?lz i)iat.evlal evn icn l~ i l i t ?~ . For most engineering steels the maximum material permeability is in the range f 1 ~ o i l l ( ~ 0 ~ ~ G a u ~ s per Oersted.

In orcler to emphasize the tact that the material permeability is not constant for a given steel, the B vs 1-1 curves iiave been plotted in Fig. 93-a, u-hicR would result if it (lid remain constant and had thc values, respecti\~ely, of 1; 10, 100 and 1000. These ase tlie dotled lines on the chart.

179


I

/ p = PERMEABILITY

~1 'f 100 B =FLUX IINDUCTIONI U1

3 l O . 0 0 0 H = MAGNETIZING FORCE cn 3 0

(I. 62.8 125.6 H I N OERSTEDS ---r

04 . , , , , , , , , , , , b. 0 62.8 125.6

H I N OERSTEDS - Fig. 93-a) Magnetization Curve for a Steel Havtng Relatively

Low Materral Permeability. b) Plot of the Matertal Permeability of the Steel Relative to H.

5. EFFECTIVE PERMEABILITY. In the case of coil or longitudinai magnetization the permeability of prime importance and interest is quite different from the material permeability. The name for this permeability in common useage is "apparent permeability", but for magnetic particle testing purposes the term "effective permeability" is preferred as being more descriptive. The effective (apparent) permeability is defined as the ratio of B in the part to H, when H is measured a t the same polnt in absence of the part. The effective permeability 1s not solely a property of the material, but is largely governed by the shape of the plece.

6. INITIAL PER~IEABILITY. Initial permeability 1s the permea- bihty of an un-magnetized material as the magne t~z~ng force, H, is

C n . m 9 FIELD STRENGTH AND DISTRIBUTION IN SYWnlETRICAL OBJECTS

applied for the first time. With increasing H the field in the part increases along the "virgln" curve of the Hysteresis loop.

I t should be noted that when the term permeability is used without qualification, i t is the mas imzon material permeability that is being referred to.

7. SELF-DEMAGNETIZING EFFECT. When a part is piaced in the magnetic field of a coil, magnetic poles appear near the areas a t which the field enters and leaves the part. These poles produce in the part another magnetic field which is opposite in direction to the applied field of the coil, and consequently weakens the field in the part. This weakening is called "self-demagnetization", and the amount of i t depends on the distance between the poles and on their concentration. Thus the self-demagnetization depends on the geo- metric shape of the part. Obviously the effect also depends on the orientation of the part with respect to the applied field, slnce the distance between the induced poles will depend on whether the long or short axis of the part is held parallel to the applied field. In magnetic particle testing most parts magnetized in a coil are placed with their long axes parallel to the applied field.

The self-demagnetizing effect is responsible for the longitudinal field in the part being much less than might be expected from the strength of the magnetizing force applied. This weakening of the field in the part is taken into account by the effective permeability. The effect is discussed more fully in the paper entitled "Theory and Application of Coil Magnetization" which was published in the Jan.-Feb., 1955 issue of the Journal of the Soc~ety for Nondestruc- tive Testing-Val. 13, No. 1, PP 23-25. In that paper i t is shown that the effective permeability can be given by the equation

L L PC, , = 6 - -5, where - is the length-to-diameter ratio of the

D D ?

L part to be magnetized. This equation 1s valid providing the -

D ratio is not greater than 15, and the material permeability is greater than or equal to 500. Neither of these conditions is very restrictive, since the material permeability of most engineering steels 1s considerably greater than 500, and few parts magnetized in a coil have L/D ratios greater than 15.

8. RULE FOR DETERDIINING AMPERE TURNS FOR LONGITUDINAL MAGNETIZING. Since the effective permeability 1s independent of


m ~ g h t be impractical t o carry out the test (using clrcular magnetization) because of the heavy current requirements.

The same reasoning applies equally to long~tudinai or coil magnetization. Because of the self-magnetizing effect (see Section 7) the magnetizing force to produce a field of 70,000 lines per square Inch would be even greater than for circular magnetization. Ampere tu rns requlred of a coil under these circumstances m ~ g h t agaln be beyond the limits of readily available equipment. \

3 EFFECT OF SHAPE ON FIELD DIRECTION. AS long as the piece to be magnetized is of regular shape and moderate size, the direction a field will take under the application of various magnetizing forces i s quite readily predictabie. Bars, square or round, or tubes o r rlngs a r e exampies of such s ~ m p l e shapes. And in a qualitative way a t least; the relative Intensity of the field may also be predicted, given some knowledge of the hardness and the composition (and therefore the permeability) of the metal.

h4. LONGITUDINAL MAGNETIZATION. When a iength of bar or tube IS magnetized by means of a coil we have what seems to be a simple problem of distribution. Magnetographs of bars so magnetized Indicate a concentration of flux lines a t the ends, suggesting that the field traverses the bar from end to end, with some flux loss along the sides a s indicated in Fig. 94. If the bar IS relativelv

Fig. 94-Diagram of the Field Around a Bar Magnet

184

FIELD STRENGTH AND DISTRIBUTION IN SYhlMETRICAL OBJECTS

uniform a s to permeability, across i ts section and along its length, i t is a pretty safe assumption that the flux density 1s approximately uniform over the cross-section a t any polnt except a t each end of the bar.

. DISTORTION OF FIELD DUE TO SHAPE. If, however there IS an upset section along the iength of the bar, a s shown in Fig. 96, the field tends to Row out into the upset portion, but does not do so uniformly. The larger the relative diameter of the upset portion, the fa r ther will the field in t h ~ s section depart from a strictly longitudinal direction. External poles will tend t o form a t "C" and "Xui and the field direction will tend to become radial along the surface "BC". Such a field distribution is favorabie for finding circumferentiai cracks in the fillet a t "B", and in the surface "BC" -but may not be favorabie for locating clrcumferentiai cracks on the surface "CX". Thls wouid be espec~ally t rue if the diameter of the upset portion is iarge with respect to the distance "CX". This distribution of field can be verified by the indication of poles a t "C" and "X" by means of a small compass or field meter.

I I

Fig. 95--Behavior ot a Longitudinal Field In the Upset Portion of a Magnetized Bar.

When a t t emp t~ng to magnetize a par t of Irregular shape for the first tlme, some such analysls of the probable path of the field should be made. As the shape of the par t becomes more complex the problem becomes correspondingly more difficult. Sometimes separate

PRINCIPLES O F DfAGNETJC PARTICLE TESTING

coil magnetization must be applied to various projections of the part to insure proper field direction a t all locations. A satisfactory procedure can usually be worked out for the most complicated shapes after some experimentation.

In predicting the direction the field will take when magnetizing with a coil, i t is well to remember that flux lines ainrays must close upon themseives to form a compiete c~rcuit, and that they tend to follow the path of lowest total reluctance. Further, because of the very low reluctance of iron as compared to air, they will follo\v

e Iron path as far as possible. 9 . CIRCULAR FIELDS. In predicting or determining the distribution and intensity of fields produced by passing current through the part or through a eentral conductor threaded through an opening in the part, a different set of rules applies. Some of these rules are simple, and when applied to relat~vely s~mple shapes, enable the operator to carry out the inspections with the assurance that he has the correct fields. But often in complicated cross-sections, cut-and-try methods must still be resorted to.

17. FIELD AROUND A CONDUCTOR. The basic ruies regarding fields around a circuiar conductor carrying direct current are;

( I ) The fieid outside a conductor of uniform cross-section is uniform along the length of the conductor.

(2) The field is a t right angles to the path of the current through the conductor.

13) The field strength outside the conductor varies inversely with the radiai distance from the center of the conductor.

In passing current through a part to be magnetized we therefore know that in general the field set up will be 90' to the direction of the current path---a most ilseful rule to remenzber. Thls means that the current should always be passed in a direction parallel ( a s neariy as possible) to the direction of expected discontinuities, since the resulting field, being at r ~ g h t angles to the current, will then also be a t right angles to the discontinuities.

The length of a part being magnetized by passing current from one end to the other, does not affect the strength of circular field produced. The diameter, ho\vever, is very important, since the field strength a t the surface decreases \\,it11 an increase in diameter. A two inch bar carrylng 1000 amperes will have a field a t its surface twice as strong as would be present in a bar four inches in diameter

C H A P T ~ 9

FIELD STBENGTH AND DISTRIBUTIOK I N SYMhlETRICALOBJECTS -

carrying the same current-or half as strong as \vOuld be the case in a one inch bar.

18. FIELD IN AND AROUND A SOLID NON-MAGNETIC CONDUCTOR, CARRYING D.C. The distribution of the field ~nside a non-magnetic conductor, such as a copper bar, when ca,-r)iing direct current, is different from the field distribution en:ternai to the bar. -4t any pomt lnslde the bar the field is due onip to that portion of the current wh~ch is flownrg in the metai bet\veen the point and the centel. of the bar-that is, through a cylinder whose radius 1s the distance from the center of the bar to the point in question. Thus the field increases from zero a t the center along a s t ra~ght line, to a maxlmum at the surface. Outside the bar the field decreases along a curve as shown in Fig. 96. In cnlculating field strengths 01~tside the bar the current

I I Fig. 9b--Distribution ot the Field In and Around a Solid Non-Magnetic

Conductor Carrylng Direct Current.

must be considered as being concentrated at the center of the ba7.. If the radius of the bar is R, and the field at the surface IS F, then

F F the field a t a distance 2R from the center will be --, a t 3R, -, etc.

2 3 79. FIELD IN AND AROUND A HOLLOW NON-MAGNETIC CONDUC-

TOR CARRYING D.C. In the case of a hollow non-magnetic conductor


carrying direct current, somewhat different conditions exist. Obviously, there is no current flowlng between a point on the inside surface and the center of the tube. Since the field a t a given point in a conductor is due only to the current flowing between the point and the center of the conductor, it follows that there is no field a t the inside surface of a hollow conductor. At the outside surface, however, the field will be the same as in the case of a solid conductor of the same diameter carrying the same amount of current. Here, in calculating external fieid strength, distances must be taken from the center of the tuhe, not from its inside surface.

The gradient of field strength from inside to outside surface of the tubular conductor is steeper (depending on the wall thickness) than from the center to the outer surface of a solid bar of the same outside diameter.

Fig. 97-Distribution of the Field in and Around a Hollow Non-Magnetbc Conductor Carrying Direct Current.

The field external to the hollow conductor will be exactly the same as for the solid conductor, and will decrease along the same curve as in the case of the solid bar. This condition is shown in I-?;,, 07

C n a m 9

FIELD STRENGTH AND DlSTRIBUTION IN SYDlDlETRICAL OBJECTS

DIRECT CURRENT. If the conductol* carrying direct current is a solid bar of steel or other magnetic material, the same distribution of field will exist as would be the case in a similar non-magnetic conductor, but the strength of the field will be much greater.

Consider a conductor of the same diameter as that shown in Fig. 96. The field a t the center would again be zero, but the field a t the surface would be X F, where p is the material permeability of the magnetic material. The actual field, therefore, may be 1000 or 2000 times the field in the non-magnetic bar. Just outside the surface, however, the field strength drops to exactly the same value as for the non-magnetic conductor, and the falling off of fieid strength with increasing distance follows the same curve. See Figure 98.

I

I I Fig. 9GDistribution of the Field In and Around a Solid Conductor Of

Magnetic Matertal Carrylng Direct Current.

21. THE CASE OF A HOLLOW MAGNETIC CONDUCTOR CARRYING DIRECT CURRENT. A similar relationship exists in the case of the hollow conductor made of magnetic material as compared to the hollow non-magnetic conductor. The field again is zero a t the inside surface, and increases along an approximately straight line to a maximum a t the outside surface, but is again times the field - A&, " # .

strength in the non-magnetic tuhe. 20- T H E CASE OF A SOLID MAGNETIC CONDUCTOR CARRYING This distribution indicates an unfavorable field for the detection

PRINCIPLES O F RI.1GXETIC PARTICLE TESTING

of defects existing on the lnside surface of the tube, and this method of magnetizing should preferably not be used for the ~iispection of tubes for ~ns ide surface defects \vIiere maximum sensitivity IS

required.

Fig. 9%-Distribution ot the Field in and Around a Hollow Conductor of Magnetic Material Carrylng Direct Current.

Cracks, however, have depth, and a s can be seen from Fig. 99 the field strength curve rises steeply, progressing from the inside t o the outside surface. Thus, defects several thousandths of a n inch in depth will Intercept some fieid, and may be reliably detected. Before using this method for the detection of I.D. defects, experimental proof of its effectiveness should be obtained.

Outslde the tube, the field strength drops off with distance from the surface along the same curve a s with non-magnetic conductors. This situation 1s shown in Fig. 99.

22. FIELD INSIDE A CONDUCTOR-THE GENERAL CASE. Figure 100 is a plot of the field produced by direct current i n s~de a con- d u c t ~ ~ ; and 1s applicable to both magnetic and non-magnetic conductors, either solid or tubular. I n the plots, the field strength is glven a s a ratio, and the .dimens~ons a r e in "units" to which any numericai value can be assigned.

190

CHAPTER 9

FIELD STRENGTH AND DISTIIIBUTIOT IN STSl&lETRICAL OIiJECTS

The conductor is defineti as h:!ving a r ad i~ i s to the ou ts~de surface of "a", and a radius to the i n s ~ d e surface iif tubular) of "b" The case of b = zero IS the case of the solid conductor. The polnt "R" is the point ~ns ide the conductor for whlch information regarding field strength is desired. The field strength in the plot IS given a s a ratio of the field a t point R to the field a t the outer surface-that i s H, - In the curve, the vertical axis could be labelled B,/B,, since H.

Fig. lo&-Graph Showme the Variation ot the Field lnslde Any Conductor Cariylng Direct Current. -

PRINCIPLES OF UAGNETIC PARTICLE TESTING

B, = pH, and B, = pH, The ratio then, is independent of the permeability, and so can be applied to either non-magnetic or magnetic materials hy substituting the appropriate vaiues.

Similarly, since the dimensions are in "units", plotted for "a" (outside radius) = 10 units, any dimensions desired can be assigned to "a" and "b" and "R" by choosing the right units. For example, for a cylinder having outside and inside radii of respectively two Inches and one inch, the "unit" is 0.2 inch. Since on the plot "a" = 10 units, the outside radius wouid he 0.2 X 10 = 2 inches, and the inside radius "b" would be 5 units, or 0.2 X 5 = 1 inch. The horizontal axls of the plot is in terms of the position of the point R.

This plot is useful in that i t shokvs on a single figure how the field strength varies inside any conductor carrying direct current.

v

1 3 . THE CASE OF A CYLINDER OF MAGNETIC MATERIAL WITH DIRECT CURRENT FLOWING THROUGH A CENTRAL CONDUCTOR. A much better way to magnetize a tube when defects on the Inside surface are sought is to pass direct current through a conductor threaded through the interior of the tube. Thls case is shown in Fig. 101. Here the field due to the current in the centrai conductor is a maximum a t the surface of the conductor, and decreases along the same curve (for air outside the conductor) as before, through

Fig. 101-Dlstrlbutton of the Freld In and Around a Hollow Magnet~c Cyllnder with D~rect Current Flowing Through a Central Conductor.

192

CHAPTER 9

FIELD STRENGTH A N D DISTRIBUTION IN SYhlRlETRICAL OBJECTS

the space between the conductor and the inside surface of the tube. However, a t this surface the field is immediately ~ncreased by the permeability factor p, and then decreases to the outer surface. Here the field agaln drops to the same decreasing curve i t was following In the air space in s~de the tube.

This method, then, produces a maximum field a t the inside surface, and thus glves strong Indications of defects on the inside surface. Sometimes these indications may even show on the outside surface of the tube.

I t shouid be noted that, as i t affects the strength of the field in the tube, it makes no difference whether the central conductor is of magnetic or non-magnetic material, since i t 1s the fieid external to the conductor itself that constitutes the magnetizing force for T i n d e r .

. MAGNETIZING WITH ALTERNATING CURRENT. In the foregoing discussion, the magnetizing current has in all cases been considered to be D . C . Most of these rules do not hold when the magnetization is done with alternating current.

I t is a well known electrrcal fact that alternating current tends to flow only along the surface of a conductor. T h ~ s tendency is in part a function of the frequency of the current, and IS extremely

Fig. 102-Distribution of the Field In and Around a Solid Conductol 01 Magnetic Materbal Carty~ng Alternating Current.

193

PRINCIPLES OF h1AGNETIC PARTICLE TESTING

pronounced a t very high frequencies. Even a t con~merc~a l frequencies (60 cycies) the tendency is apprecrable, especially in magnetic materials. The phenomenon is referred to a s "sii~n effect".

From the principles already set for th in discussing field distribution when D.C. is used for magnetization, i t is obvious tha t if the current density is greater in the outer layers of the conductor, the field density will be correspondingly greater a t such locations. Experiments have indicated tha t in a steel shaf t six inches in diameter, magnetized by passing heavy alternating current through it f rom end t o cnd, there was little if any field detcctahle more than one and one-half inches below the surface, and that the field was almost entirely concentratetl in the outer half inch of metal.

25. THE CASE OF A SOLID CONDUCTOR MADE OF MAGNETIC MATERIAL, CARRYING ALTERNATING CURRENT. The field distribution in a solid magnetic conductor car lying alternating current is shown in Fig. 102. Outside the conductor the field strength a t any point 1s decreasing in exactly the same way a s when direct current is the magnetizing force, but i t must be remembered tha t while tlle A . C . is flowing, the field is constantly varying, both in strength and ,

I

Fig. 103-Distribut~on ot the Field In and Around a Hollow Conductor of Magnetic Mater~al.Cartylng Alternating Current.

-

FIELD STRENGTH A N D DISTRIBUTlOli 1N SYSlhlETRlf'AL OBJECTS

direction. lns tdc the conductor the field 1s zero a t tile exact center, and increases toward the ou ts~de surface, slowly a t first, then with Increasrng rapidity to reach a h ~ g h masimum a t the surface.

26. THE CASE OF A HOLLOIS~ CONDUCTOR XADE OF ~ I A G N E T I C MATERIAL. CARRYING ALTERNATING CURRENT. Similar differences in field distribution occtir in the case of the hollow conductor of magnetlc ma t e r~a l when A.C. is used for magnetization. Agarn there is no field a t the i n s ~ d e surface of the tube. Between the inslde and the outside surface, the field Increases a t an accelerating rate

Fig. 104-Graph Showjng the Effect ot Conductivity, Permeability and Frequency on the Skin Efiect ot Alternatkng Current.

195

pnrNCrPI.ES OF MAGNETIC PARTICLE TESTING

from zero on the rnside surface to reach a hlgh value near the outside surface with a maxlmum field a t that surface.

Figure 104 is a graph on wh~ch IS plotted the field distribution a t several points between center and surface of a solid conductor (magnetic or nonmagnetic) carrying alternating current. Dimen- sions are agarn in "units" so that any values can be assigned to them, as in the case of the plot for direct current (Fig. 100). The mag-

HI, netizrng force is agarn expressed as a ratio, -, and as before is

Ha

Rg. I O S G r a p h lllustrattng the Sktn Eiiect of Alternatsng Current

FIELD STRENGTH AND DlSTRtBUTlON IN SPRlMETRlCAL OBJECTS

Independent of the permeability of the mater~al composrng the conductor. The horrzontal axis represents the distance of the point R from the center in terms of fractions of the radius of the bar.

The several curves are plots of the field strength for various values of the term X,, which is a function of the conductivity and permeability of the conductor, and of the frequency of the alternating magnetizing current. The value of X, increases with an increase in any or all of the values for conductivity, permeability and frequency. The piots show at a glance the falling off of the field strength from the outside surface to the center of the conductor. The rate of this falling off IS more rapid if the conductivity is hlgh, or if the permeability is high, or if the frequency is high. The graph is a ciear illustration of the skin effect of alternating current.

Figure 105 also applies to the case of a solid conductor carrying alternating current, and is merely a re-plot of the data used for Fig. 104. In this case the graph of the ratio of the field rntensities HE - is plotted against the function X,, for various points along the .. H, radius from center to surface. T h ~ s plot shows particularly well how raprdly the field strength falls off as a function of depth for the case of alternating current in a conductor.

FIELD DISTRIBUTION IN LARGE OR IRREGULAR SHAPED BODIES

1. INTRODUCTION. The rules discussed in the foregoing two chapters fo r arriving a t suitable field strengths for magnetic particie testing, both for longitudinal and circular magnetization, a r e valid regardless of the size of the object b a n g magnetized. However, their application becomes imprncticai a s the size of the object becomes larger. Fo r example, although the rules regarding coil magnetization in terms of the L/D ratio mill remaln valid, coil diameter will become very large, since one of the restrictions of the rule of thumb for t h ~ s case is that the cross-sectional area o r the par t be not greater than one. tenth the cross-sectionalarea_f .t-I. For a six inch shaf t t h ~ s tvouid require a 19 inch coil, and for a par t 12 inches in diameter a 38 inch coil.

Similarly, the one-thousand-amperes-per-inch ruie for circular magnetization becomes impractical in most cases: since even a 12 inch par t would requlre 12,000 amperes. This amount of current is generally not available unless equipment for the purpose has been especially built. Further, the rules for field variation described in Chapter 9 apply, strictly speaking, only to uniform objects of cylindrical shape. \Vlien wc pass to irregular shapes and cross- sections it becomes more difiicult to predict current and field distribution. The difficulty becomes even greater a s these irregular shapes become quite large.

Two polnts, ho\vever, can be emphasized. First , if the rules w e followed the fields produced will be adeqriate for magnetic particle testing, and in most cases the fields so produced in large objects will be larger than required.

Secona, when liigh permeability steels are being tested, current values mucll io~ver than indicated by the rules a r e often adequate. When they a1.e not; local magnetization with prods or yokes is a means of securing higher slrength fields without the use of extremeiy high currents. Overall magnetizing of parts 1s often not necessary.

FIELD I)ISTRIRlITiON IN LARGE AND IRKEGGLAIC-SHAPED BODIES

2. THE CASE O F A S Q ~ ~ A R E EAR, CIRCULARLY ~ ~ A G N E T I Z ~ WlTH

D.C. The 1000 ampere-per-inch-of-diameter rule would require 2000 amperes t o magnetize a round bar 2 inches in diameter. If \vc take a square bar, 2 inches on a side, the longest diametricai dimension is the diagonal, or 2.82 inches. In thls cdse the current should be stepped up to 2820 amperes, according to tile rule.

i i i i s u r e n i ~ n - t ~ m ~ a a r i t h the Iiall Effect probe, of the tangentiai component of H (See Chapter 8, Section 5 ) on such a bar, indicate tha t the fieid strength a t the corners of the bar 1s about half that a t the center of a face. In other words, the field strength is not uniform over a square cross-section. Furthermore it does not foIlo\v the s ~ m p l e ruie of decrease with distance from the center, as in the case of a round bar. A round bar, having a diameter of 2.82 inches. carrying 2000 amperes, \vould have a field strength a t the surface of 6

-? or 0.71 of the surface fieid of a 2 Inch diameter round bar 2.82 carrying 2000 amperes. I n the case of the square har, however, the field falls off more rapidly a t the corners, where "diameter" or diagonal is 2.82 inches, and has a strength of o11iy half, or 0.50 that ,

of the field a t the center of a face (where "diameter" 1s 2 inches),. ' '.. . . ,

Figure 106 illustrates this case. .- . c ,

I SOUARE BAR, CIRCULARLY MAGNETIZED WlTH D.C.

FIELD STRENGTH , 0

AT CORNER s % / .

0 , .

FIELD STRENGTH AT CENTER OF FACE = 0 .

t.*? + zsa 4 I

Fig. 106Field Distribution in a Square Bar, Circularly Magnetized with Direct Current.

3. T I ~ E CASE OF A RECTANGULAR EAR, CIRCULARLY MAGNETIZED WITX D.C. Suppose it.e assume a rectangular bar, 2 Inches by 6

PRINCIPLES OF MAGNETIC PARTICLE TESTING i f

I IFiELO STRENGTH &T CEI1TER OF 6" FACE. 1

FlELD STRENGTH

0.T CORNER ' ' T-

FlELD STRENGTH . B AT CCENTER OF 2" F&CE

&oOO 6"-

p~

Fig. 107-Field Distribution in a Rectangular Bar. Circularly Magnetized with Direct Current.

inches in cross-section, and tha t we wish to produce in i t a field of the same strength a s we had in the 2 inch square bar, again using

-,direct current. This case is illustrated in Fig. 107. If we interpret the 1000 ampere-per-inch ruie a s applying to t he long dimension of the cross-section of such a bar, a current of 6000 amperes is Indicated.

When such a bar was magnetized in the laboratory with 6000 amperes D.C., and hy means of the Hall probe the tangential component of H measured and compared with the results obtained on the 2 inch square bar, the following values for field strength a t various points were found;

( a ) The field strengths a t the corners of the 2 by 6 inch bar and the 2 by 2 inch bar were approximately the same.

( b ) The field a t the center of the 6 inch face of the rectangular bar was 1.75 times the field a t the center of the 2 inch face of the 2 inch square bar.

I n this case i t can be assumed tha t the distribution of cun.e?zt is uniform over the 2 by 6 inch cross-section, hut the field distributio?l. is not uniform. However, this non-uniformity tends toward producing stronger fields than needed.

4. NON-UNIFORM CROSS-SECTIONS, CIRCULARLY MAGNETIZED WITH D.C. When the cross-section departs radically from a substantially uniform distribution of materrai, the above method of estimating probable field strength and distribution breaks down. Fundamental rules are :

( a ) Over a non-uniform cross-section cx~rrent will distribute itself evenly over the cross-section; so that the current flowing

i CHA~TER 10 i

FIELD DISTRIBUTION IN LARGE AND IRREGULAR-SHAPED BODIES ! !

in each element of the section is the same, provided contact 1s made evenly over the entire cross-section. If the piece is long enough the distribution of the current will be uniform even if contact is over a limited portion of the eross-section. Experiments have sl~oxvn tha t the current spreads out very quickly over the entire cross-section.

(b) If the cross-section varies along the length of the part, the current will spread out ( o r condense) t o occupy the changed section, and field strength will vary with current distribution, up or down. Field strength will be greater a t smaller diameters and weaker a t larger.

5. THE CASE OF AN I-SHAPED CROSS-SECTION, CIRCULARLY MAGNETIZED WITH D.C. Figure 108 sholvs t he cross-section of an I-shaped bar. The two sections a t the edges a r e two inches square, and they are joined by a one half inch by two inch web. This can be considered to be the two by six inch rectangular bar of Fig. 107 with the middle two inches machined away to ieave a one half Inch thrck web. Suppose that it i s desired t o magnetize such a bar clr- cularly, so that the field over the cross-section will be that called for by the 1000 ampere-per-inch rule. If we break the section down

Fig. 108-Field Distribution in an I-Shaped Bar, Circularly Magnetized with Direct Current.

rnto three parts-two 2 by 2 rnch bloclcs a t the outer edges, and a two by one half ~ n c h web between. we can say tha t each of the two by two inch edge areas would require 2000 amperes, o r 500 amperes


per square inch. Since the current distributes itself uniformly per unit of area, the center section, having an area of one square inch, mould take only 500 amperes for the same current density a s the end sections-a total of 4500 amperes for the entire section.

But if the central web area is broken down into four one half inch square elements, each of these would carry only 125 amperes-one fourth of the 500 ampere per square inch current density of the end sections. But by the 1000 ampere-per-inch rule each one half ~ n c h diameter element of the web \vould requlre 500 amperes, four times what i t would carry using the 4500 total amperes calculated above. And if the current is stepped up to give each one half inch square element 500 amperes, the web would take 2000 amperes total. The two-inch square edge sections would take the same current density, o r 8000 amperes each. With the 2000 amperes total through the web, 18,000 amperes altogether ~vould be needed.

Obviously this is an absurdly high value and experience has shown tha t i t need be nowhere near a s large a s this. On the basis of the six inch longest dimension of the I section, only 6000 amperes would be called for by the 1000 ampere rule, w h ~ c h on the rectangular bar of Fig. 107 gave nearly double the field in the center sections a s was found a t the edges. We can conclude therefore, that a t 6000 amperes for the I-shaped section, the field in the edge areas would be adequate, whereas the field a t the center of the web would be stronger than required.

This rather labored analysis has been included in this discussion to point up the fact that prediction of suitable current levels in irregular cross-sections becomes very complicated, .and a t best lacks much in accuracy. Results just a s accurate and with niuci~ less time and effort can be obtained by ail expei*ienced operator by cut-and- t ry methods, s tar t ing a t high currents and working do\vn, or in other cases, starting with low currents and working up.

One side comment should be made. With strong circular magnetization of the cl.oss-section of Fig. 108, local poles \vould appear a t the innel-corners of the two inch edge areas, marked a, b, c and d in the figure. Also a leakage field would develop a t the tnszde angles, e, f , g and h, between the edge areas and the web. To avoid confusing indications from these local leakage fields, it would probably be desirable to work with a s low a current value a s seems feasibie; with perhaps a still lower value for tlie iiiside fillets. In critical cases two Inspections may therefore be in order-one a t a lo\\, cur-

211"

C H A I T E R 10

FIELD DISTRIBUTION I N LARGE AND IRREGULAR-SHAI'EU BODIES

rent value for the web section, and another a t a h ~ g h e r current value for the square end sections.

6. THE CASE OF PROD C,ONTACTS ON LARGE OBJECTS. When magnetiz~ug a large par t locally by means of prods, using direct current, tlie proolem of determining actual fieid strengths and distribution is agaln very difiicult since many variables a re involved. The current passes ~ n t o the par t or plate a t one contact and leaves i t a t the other. I t \\sill spread out non-uniformly between the two contact points, both parallel to and normal to the plate surface. The pattern will depend on prod spacing, dimensions of the par t and its electrical conductivity. Obviously the over-all pattern of field density \\.ill depend on the pattern of current distribution, and obviously also, the current will be strongest on a direct line between tlie two p o ~ n t s of contact. Naturally, in malcing the inspection, the prods a r e so positioned tha t the area aiong this line and the area closely adjacent to i t is the area in which Indications of surface cracks a re expected t o occur.

The prod contact technique has a special ability to produce indications of discontinuities which lie wholly below the surface, often quite deep w i t h ~ n the part. Thus fa r , attempts to plot the current and field distribution deep within a relatively thlck section when prods a re applied t o one surface have ied to somewhat inconclusive results.

7. LABORATORY TESTS FOR FIELD STRENGTH WITH PROD NAG- NETIZATION. Laboratory tests have been made to give some information regarding field strength and current distribution, and directions, rn the surface layers of the metal by measuring the tangential con~ponent of "H" uslng the Hall Probe.

One series of tests was made t o determine the variation in field strength midway between the prods, \r.hen prod spacing and current (D.C.) were vaned. Prod spacing was vai-led from 2 tnches to 18 inches, and for each spacing, current was varied from 200 to over 2000 amperes. Figure 109 gives the curves for these data, with prod spacing plotted against field strengths, measured by the tangential component of H, each curlre representing a different current value. The curves sholvn a r e for one inch thick plate. Data for one quarter and one half inch plate a re altogether similar. I t is evident from these curves that field strength falls off 'apidly a s prod spacing increases, and that increases in current produce the greatest increase In field strength when the prod spacings are small. The

PRINCIPLES O F I\lAGNETIC PARTICLE TESTING I

Fig. 110--Cornparkson of the Effectiveness of Direct Current and Half Wave Current when Used with Prod Magnetization.

204

CHAPTER 10

FIELD DISTRIBUTION IN LARGE AND IRREGULAR-SHAPED BODIES

experiment of course gives no evidence whatever as to what goes on below the surface of the plate.

8. CORRELATION OF THE TESTS WITH PRACTICE. The values obtamed by these tests are not of much use, in themselves, for the purpose of determlnlng how much current and what prod spacing are best to use for any given testing problem. However, manyyears of experience in steel casting and weld inspection with prods has evolved a pattern of prod spacing and current values for various plate thicknesses. The A.S.T.M. in 1958 publis'hed tentative Standard E 109-57 T, givlng recommended prod spacings and currents for plate thicknesses under three quarters of an inch, and for thicknesses three quarters of an inch and over. These are quoted In Table I.

TABLE 1. A.S.T.M. Recommended Prod Spaclngs and Current Values.

-- ' I t svould seem t h a t the A.S.T.M. values for prod spucrng and current (Table 1 ) mlght se l l bc revlsce as to current specified, so as t o produce mure nearly 40 to 50 Oersteds a t all spaclnas.

205

Prod Spacing, Inches

2 to 4 Over 4 to less than 6 6 to 8

Uslng the recommended currents in these A.S.T.M. specifications and the field value for the mld-point between the two prods, taken from the plot for one inch plate thickness as shown in Figure 109. we can re-write Table I in terms of magnetizing force, H, expressed in Oersteds at the surface. The values so obtained are shown in Table 11.

TABLE II*

Magnetizing Force In Oersteds for Various Prod Spaclngs, Uslng A.S.T.M. Recommended Currents.

Prod Spacing, Inches

2 to 4 Over 4 to less than G 6 to 8

Section Thickness, Inches

Under %, lnch

200 to 300 Amperes 300 to 400 Amperes 400 to 600 A m ~ e r e s

Section Thickness, Inches

%. inch and over

300 to 400 Amperes 400 to 600 Amperes 600 to 800 A m ~ e r e s

Under yl, ~ n c h

40 to 25 Oersteds 35 to 20 Oersteds 30 to 20 Oersteds

%. Inch and over

50 to 45 Oersteds 45 to 40 Oersteds 40 to 35 Oersteds

PRINCIPLES OF MAGSETIC PARTICLE TESTING

Tlie magnetizing forces a s measured with the Hall probe a t the mid-point between prod contacts, can now be read for the A.S.T.M. recommendations. The Tabie sho%srs the magnetizing forces to vary from 20 oersteds for G to 8 inch spacing on plates iess than y,. inch thick: to 50 Oersteds fo r 2 to 4 inch spacing on :%. inch plate and thicker. It therefore becomes possible, by taliing Hall probe readings, to space the prods and adjust the current so tha t the minimum magnetizulg force of 20 Oersteds ( o r any othei- value tha t may be determined to be desirable) can bc obtained.

Also, since field strengths as measured in these tests a r e almost directly proportional to current, prod current magnitudes can be established in terms of amperes per inch of prod spacing, fu r various thicknesses of plate. I t is also possibie by means of the Hall probe measurement of H, to determine a t what other points, outside the prod line, tlie 20 Oersted minimum value of magnetizing force has been developed.

Fo r example, along tlie prod line or very near it, laboratory tests have shown tha t GO amperes per inch of prod spacing will produce the required 20 Oersteds minimum for plate thicknesses under s, inch; and 100 amperes per inch !sill produce the required 35 Oersteds minimum for thicknesses over :%. inch. As one moves away from the prod line a greater number of amperes per nich a r e required to produce these minimum values. Fo r example, a t a perpendicular distance out from the prod line of one half of tlie prod spaclng, 150 amperes per inch were required to produce the 20 Oersteds minimum for thiiknesses under ",$ inch. 250 to 350 amperes per inch were needed to produce the 35 Oersteds minimum for thicknesses over "/1 inch.

I t is evident that current vaiues rapidly rise above those permitted a t distances of oniy a few inches away from tlie prod line. ~ I a x i m u m current permitted is determined by the heating effect a t the prod contacts w h ~ c h can damage the steel; by the interference of the powder pattern produced by the field e z t e ~ n a l to the plate and around the prods themselves, which becomes pronounced a t high currents; by the limitations of the output of the current source; and finally by experience a s incorporated in the A.S:T.I\I. spec)- fication.

Experience has also shown that, except in special cases, practical prod spacing should be limited to 8 inches if maximum sensitivity

CHAPTER 10

FIELD DISTRIBUTION IN LARGE AND IRREGCLAR-SHAI'ED BODIES -

is required. Greatcr spacing requires higher currents and t he fields produced are "wasted" in spreading out into the plate beyond the prod line.

9. PROD INSPECTION USING HALF \VASE CURRENT. Similar experiments to those outlined above, but using half wave instead of direct current were also carried out. Curves comparing the magnetizing force produced on a !/* inch thick plate, with various current strengths, fo r direct and for half wave current a r e shown in Figure 110. I t is evident f rom these curves tha t field strength per ampere is somewhat higher fo r half wave than for D.C. a t the smaller prod spacmgs, but is almost the same for either direct current o r half wase a t prod spacings greater than 8 inches. Since half wave con- sumes less power and produces lower heating effects in equipment and a t the prod contact points, than does direct current, i t i s the preferred method when prod inspection is indicated. Also, better powder mobility is obtained with half-wave.

lo. OVER-ALL MAGNETIZATION OF LARGE OBJECTS. When the over-all method of inspecting large and complex bodies, such as castings, is used to replace inspection with prods, the matter of field direction and tlie amount of current to use becomes still more complex. Actually there is no method by which the points of contact or the current necessary can be determined except by experience, general analysis, and some cut-and-try tests.

For this type of inspection, large units, delivering f rom 6,000 to over 20,000 amperes D.C., have been used. In numerous instances, multiple circuits, each callable of delivering the maximum current, have been enlployed. Two or three such output circuits may be required. By a special s!%ritching circuit, current is passed i n quick successlon through each of the output circuits. Current i s carried to the castings or weidments to be inspected, by means of contacting jigs or by heavy flexible Cables, which a r e attached directly to tlie part. Or, they may be run a s central conductors through openings in the casting to form a coil, or wrapped around tlie casting for longitudinal magnetization.

Tlie exact arrangement of circuits must be determined experimentally fo r each size and shape of casting. The operation is re- warding, however, because more thorough inspection is achieved than is possible with prods and the man-hours of time for the inspection of large and complex-shaped castings can be reduced by

207


up to 80%. The over-all method will be discussed in greater detail In Chapters 19 and 24.

11. SUMMARY. Before leaving the entire subject of magnetizing means, field strength and distribution, current distribution and strength requirements, and going on to the actual techniques of appiying magnetic particle testing, i t is perhaps worthwhile to set down a few of the general principles and rules which an operator should remember :

(1) Fields should be a t 90' to the direction of defects.

(2 ) Fields generated by electric currents are a t 90' to the direction of current Row.

(3) When magnetizing with electric currents, pass the current zn a directzon parallel to the direction of expected discont~nur- ties. The field \\sill be a t r ight angles to the current and, therefore, also a t r ight angles to the discontinuity.

( 4 ) Circular magnetization has the advantage over longitudinal In tha t there a r e few, if any, local poles to cause confusion in particle patterns, and is to be preferred when a choice of methods is permissible.

(5) Fo r circular magnetization, use 1000 amperes per inch of diameter of the part. Thrs often produces too strong a field, but if used, the field so produced is sure to be as strong a s necessary.

(6) F o r coil magnetization use the formula given in Chapter 9 t o arr ive a t the ampere turns necessary for any given set of conditions. This formula is:

45,000 N I = - where NI is the ampere turns required, and

T . I ~ -, - L/D is the length t o diameter ratio of the part .

(7) F o r prod magnetization with direct current, a minimum of 60 amperes per Inch of prod spacrng will produce the mrnimum magnetizing force of 20 Oersteds a t the midpoint of the prod line for plate %. inch thick o r less. A safer figure to use, however, is 200 amperes per inch, unless this current strength produces interfering surface powder patterns. Prod spacing for practical inspection purposes is limited to about 8 inches maximum, except in special cases.

MAGNETIC PARTICLES-THEIR NATURE AND PROPERTIES

7 . GENERAL DESCRIPTION. There are two essential components of the magnetic particle testing process, each of equal importance for reliable results. The first is the proper magnetization of the par t to be tested, with fields of the right strength and in the right direction for the detection of the particuiar type of defect berng sought. The second ingredient of a successful test is the use of the proper type of magnetic particles to secure the best possible indications of these defects under the conditions prevailing in any grven case.

The particles used a r e in all cases finely divided ferromagnetic material. The properties of this material vary over a very wide range for different applications-including magnetic properties, size, shape, density, mobility and visibility or contrast. Varying requirements fo r varying conditions of test and varying properties of suitable materials have ied to the development of a large number of different types of materials now available on the market. The chorce of whrch one to use is an important one, since the appearance of the particle patterns a t discontinuities will be affected by this choice, even t o the point of whether or not a pattern is formed a t all.

Since the results of magnetic particle tests usually depend on the interpretation of the particle pattern by the inspector, the appearance of this pattern is of fundamental importance. For t lus same reason, the reproducibility of results by different operators in different locations depends on the use of the same type of particles by each operator, a s well as on the use of the same magnetizing procedure.

A vast amount of investigational worlc has been done with a view to determining what the desrrable properties of magnetic particles are. The values fo r the properties wl~ich were listed above have been determined for a iarge number of finely divided ferromagnetic mater~als . But knowrng these properties is not the same a s knowing what the effect of each of the individual characteristics is on the patterns produced. The very complex inter-relation of these properties, plus the influence of the numerous types and methods of

PRINCIPLES OF XIIGNETIC PAflTICLE TESTING

niagnetization, makes selecting the optimum properties of magnetic fo r any particular applicat~on a matter of experience.

Experience can and does lead to the choice of the most suitable of available particles in-any given case. The development of the various types of magnetic particles \\'as dictated by t h ~ s very need, and the formulations were worked out experin~enlally to produce the best results. Although esperience does not necessarily tell us zolw certain combinations perform better than others, demonstrabie performance comparisons in actuai use in finding defects is the most satisfactory way of determining which materlal is best for any given set of test conditions.

In the followtng sections we will discuss the various properties ~vhich a r e considered to affect the over-all results, but no attempt will be made to asslgn to each property, in any quantitative way, the role tha t i t plays, or its relative importance to the end results.

2. EFFECT OF SIZE. I t IS self-evident that size must play a n Im- portant par t in the behavior of magnetic particles when in a magnetic fieid, which ca?z be quite weak a t a discontinuity. A large Reavy particle is not likely to be arrested and held by a weak field when such particles a r e moving over a par t surface. On the other hand, very fine powders ?will be held by very weak fields, since their mass is very small. But extremely fine particles may aiso adhere to the surface where there are no discontinuities, especially if it is rough, and forin confusing backgrounds.

A. DT?J Pozudel-s. In general, for the dry poivders, sensitivity to very fine defects increases a s particle size decreases, but with definite limitations. If the particles a r e extremely small, of the order of a few microns, they behave lilte a dust-that is, they acc~imulate and adhere in depressions on even very smooth-loolcing surfaces. They will "develop" fingerprints and adhere a t any damp or slightly oily areas, whether or not leaitage fields cx~s t . These extremeiy fine powders, though undoubtedly sensitive to very wrali fields, are not desirable for general use, because they dir irave this heavy, dusty background.

In some special applications, p:n.tlcles of a specific size range are used. For instance, wliere i t is desned to detect only ra thr r large, coarse discontinuities, ant1 not produce indications of %.en7 fine ones, only large sized particles a r e used.

IIo\vever, mo,st dry ferroinagnctic po\vders used for rlerecting d i s c ~ i ~ ~ i n i i ~ t i ~ s are careful mixtures of iiartielcs o i all sizes. The sn?all:~r oiies ;idd sensitivity ant1 mobility, v'liile the larger oiies not only aid in locahng large defects, but by a sort of sweeping action, counteract to some extent the tendency of the iines to leave a rfusty hackground. Thus, by includin,o the entire size range, a balanced pcnvrlei. \\'it11 ~cns i t i r i ty over most of the range of sizes of discontinuities is produced.

6. IVr.t Mcf.trorl ~llnlei.7a1s. Wheri the ferromagnetic particles a r e applied as a suspension in some liquid medium, much finer particles can be used. The upper limit of particle size in most commercial \vet method matcri;~ls used for magnetic particle testing purposes, is in the range of 60 to 40 mlerons (about .0025 to .0015 inch). Particies larger than this are diificult to hold in suspension, anrl even the 40 to 60 mlcron s u e s scttle out of suspensio~i rather rapidly. Also, the large particles have another bad feature. When such a suspension is applied over a surface, the Iiquitl drains away, and a s llie film remain in^ over the surface becomcs thinner, the coarse particles are quickly stranded anrl immobilized. Such stranded particles often line up in \vl~at a r e called "drainage lines" to form a "111gh water mark" of particles tha t can be confused with indications of discrmtinuitics.

In the case of the finer partlcies, the stranding due to the dmining away of the liquid occurs much later, givlng the particles mobility iong enough to reach the influence of leakage fields and accumuiate to form true indications. The mznimzin. size limit for particles to be used a s liquid suspensions is in- determinate. Commercially available ferromagnetic materials commoniy used for tliis purpose include some exceedingly fine onrticles. The eiectron microscope has shown the preserice of a large proportion of particles a s sinall as one-eighth of a micron (0.000005, o r 5 millionths inch) in diameter.

In actual use, however, par t~cles of this size never act as individuals. Due to the fact tha t they are magnetized in use, they become actual tiiiy magnets, because the material used has appreciable retentivity. Under conditiol~s of quiet settling in a suspensionr these paiticlcs a r e ilra\sn together as a resull of their retained magnetism to form ciumps or :~ggregaLes of particles. These aggrcg:ttions of fine particles then tend to


ac t a s a unit \$,hen they a r e applied to the surface of par ts fo r magnetic particie testing. The speed and extent to which this process takes place increases with the rententivity of the particle material. Agitating the suspension breaks up the ag- regates, but they begin to form again as soon a s agitation ceases. This happens a s soon a s tlie suspension has been applied over the surface of tlie part. Tlierefore, since the particles act a s agglomerated units of varying slze, and not a s individual particles, there is no meaningful lower limit for the size of the individuai particles themselves fo r the wet method.

C. Advaatnges o f Agglontcration of Fi?te Particles. Thls ag- glomeratlon of fine particies into larger clumps is advantageous rather than otherwise, a s long a s the size of the aggregates does not become larger than the GO micron "limit" mentioned above. Individual particlcs of exceedingly small size, and therefore mass, move very slowiy through the liquid of tlie suspension under the Influence of leakage fields a t discontinuities. Unless special techniques a r e used, individual particles of the size indicated by the electron microscope a r e not particularly usefui for the location of exceedingly fine cracks until the process of agglomeration into somewhat larger units has taken place. In practical applications this process takes place while drainage of the suspension from the surface of tlie part is occurring. As the agglomeration proceeds, the clumps formed will vary in size, and, since these clumps act a s individuai units, the effect is that of a particle size range from very fine to relatively coarse.

D. Fluoresce?zt Particies. The above discussion applies primarily to magnetic particles that have not been treated with fluorescent pigments. Tile treated particlcs have fewer of the very fine particies, though when they nye present they tend to act in the sarne manner a s those not so treater! but with some\vhat less tendency to agglomerate. The same comment applies to all other types of wet method materials in which a bonding material is incorporated, such as the newer concentrates. Therefore, a meantngfui lower limit for "particle size" cannot be sct in these cases either.

3. EFFECT OF DENSITY. Most ferromagnetic materlais have fairly h i ~ i t densities-that is, they a r e heavi, in terms of weight per unit voiume. Thc ~ 1 e n s i t . i ~ ~ of the mntmials in common nsc vary

CHAPTER 11

MAGNETIC PARTICLES-THEIR NATURE AND PROPERTIES

f rom around 6.0 to nearly 8.0 times the Censity of \\rater. Such heavy materials tend to settle rapidly whether suspended in liquids or in air. This is the pr inc~ple reason for ruling out large-sized particles, since t he smaller slzes, having less mass per particle, as well a s more surface area in proportion to mass, settle out much more slowly. The density of many ferromagnetic powders i s lowered somewhat by compounding or coating them with pigments whose densities a r e lower than that of the particles. This i s true of both the dry, pigmented powders, and the fluorescent particies in liquid suspension.

4. EFFECT OF SHAPE. The shape of the magnetic particles used for magnetic particle testing has a strong bearing on their behavior in locating defects. When in a magnetic field the particies tend to align themselves along the lines of force, a s illustrated in a magnetograph. This tendency is much stronger with elongated or rod- like particles than with more compact or grobular shapes for the reason that the long shapes deveiop stronger polarity. Because of the attraction exhibited by opposite poies, the pronounced north and south poles of these tiny magnets arrange themseives into strings of particles, north pole t o south pole, much more readily than do globular shapes. The result is the formation of stronger patterns in weak leakage fields, a s these magnetically formed strings of particles bridge the discontinuity.

The superior effectiveness of the elongated shapes over the globular shapes is particlularly noticeable in tlie detection of wlde, shallow discontinuities, o r of those discontinuities which lie wholly below the surface. The leakage fields a t such defects a r e more diffuse, and the formation of s t r lngs due to the stronger polarity of the elongated shaped magnetic particles makes for stronger, readable patterns in such cases.

I n the case of the dry powders, there is another effect due to the shape of the particles which must be taken into account. Dry particles a r e applied to the surfaces of par ts by means of plastic powder bottles, or by rubber squeeze bulbs, or by the use of air-stream- operated powder guns. The ability to flow freely and to form uniformly dispersed clouds of powder that will spread evenly over a surface is a necessary characteristic fo r rapid and effective dry- powder testing. A powder composed only of e l on~a t ed shapes tends to "felt" together in a container, and to be ejected in uneven clumps. When a po\!~der behaves in this manner, the inspection becomes ex-

PRINCIPLES OF RIAGXETIC PARTICLE TESTING

tremely slow and difficult. Globular-shaped particles on the other hand, flow freely and sinoothly under similar conditions.

Clearly, then, a rlry powder must have free-Howing properties for casy application, yet have optimum shape for the greatest sensitivity for the format~on of strong ~ndications. These two opposing needs are inet by blending particles of di0'erent sllapes. A fair proportion of rod-like particles must be present for a sensitive blend. To llave powder that will flow \\.ell for easy and uniform application, a suffic~ent proportion of more con~pact shapes must also be present. Since there a r e considerable vari;ttions in the sliapes of the differelit types of po~vders a s received, production of a finished niagnetic particle testing po\rrder rea i~ i res close control to insure uniformity.

I n the case of particles for the d vet method of inspection, the individual particles a r e kept dispersed by "mechanical" agitation until they are flowed in suspension over the surface of the magnetized part. There is therefore no need to ~ncorporate unfavorable shapes merely for the purpose of improving the flo\v of the particles. Long, slender particles, with otherwise desirable characLeristics. could be used exclusively. Unfortunately, exclusive use of such particies would be cspecially costly, and the advantage to be gained from their use would not appear to war ran t the added expense to the user. Experience over the years has shoxvn tha t satisfactory sensitivity IS obtaiuable with particles of less than optimum shape In the wet version of the method.

Recause wet method particles a r e suspe~lded in a l iqu~d medium which 1s very much denser and more viscous than alr, they move in the leakage fields much more slowly than the dry powders, and they therefore accumuiate much more slo\vly a t discontinuities. In the vicinity of leakage fields, they can be seen to line up to form minute elongated aggregates. Even the unfavorable aggregate shapes, formed by simple agglomeration in suspension, \r.ill line up into magnetically-held, elongated aggregates under the influence of local, iolv-level leakage fields. This effect contributes to the really remarkable sensitivity of tllc fine particles comprising wet method materials, a sensitivity \ v l ~ ~ c h has not been equalled by "magnetic perturbation" detection instrument probes.

5. PERMEABILITY. In theory, magnetic particles used for magnetic particle testing should have a s high a permeability a s possible. This is for the reason tha t they must be readily magnetized by the

lo~v-level leakage fields t ha t occur in the Yicinity of a disc on ti nu it^. so that they will l ~ e dralsn by these fields lo the discontinuity itself to form a readable indication, even though these fields a r e sometimes very weak a t very fine discontinuities.

As a practical matter however, there 1s little connection between permeability and sensitivity for magnetic powders. For instance, the iron-based dry-n~etlioti polvders have permeabilities t ha t a r e l i~gher , by an order of magnitude or more, than those of the oxides used in tlie wet method. Yet a typical dry po~\scler has little value for detecting the extremely fine surface cracks tha t the \vet method par t~cles find so well.

One reason for t h ~ s seemlng paradox is t ha t permeability IS ]tist one factor in the desirable properties of a suitable po\17der, scarcely more important than size, shape or any other. Unless all other factors are in tlie proper range for the application a t hand, high permeability alone 1s of little value. Another reason 1s tha t published values of permeabilities are not adequate guides for judging the behavior of particles. Most such values a r e in lernls of nxozz- ?~zt~nz n~aterzai per?~ieoDilily, whereas in the low level fields a t discontinuities i t is the i?titial pernleabilit~r tha t \ ~ o u i d seein to be the determuling factor. (See the discussion of permeability in Chapter 9, Sections 3 to 5.) Of two po\\fders, A and B, even if powder A has a much higher inaterlal permeability than powder B, i t does not follow that the initial permeability of A will be higher than tha t of B. I t might just a s easily be the other way around.

6. COERCI\~E FORCE. AS a general principle, lo\\' coercive force and low retentivity a r e desirable in magnetic particie testing powders. If these values were high in a dry powder, for example, the particles would become magnetized during manufacture, or in first use, and thus become small, yet strong, permanent magnets. Once magnetized, their tendency to be controlled by the weak fields a t discontinuities \\~ould be over-shadowed by their tendency to stick magnetically to the test surface wherever they first touch it. T h ~ s acts to reduce mobility of the powder, and also t o form a Iilgh level of background \vhich not only reduces contrast and makes indications harder to see, but also wastes powder.

In tlie case of wet method particles that wouid becomc strongly magnetized because of l i ~ g h coercive force, this same objectionable background will form f o r the saine reason. In addition, such particles woulil stick to any Iron or steel in the tank or plumbing of


the unit, and cause heavy settling-out losses which would have to be made up by frequent additions of new particles to the bath. Another bad feature shown by highly retentive wet method particles is their tendency to clump together quickly in large aggregates on the tes t surface. As previously stated, excessively large clumps of material have lorn mobility and indications a r e distorted or obscured by the heavy, coarse-grained backgrounds. Therefore, particles having high coercive force and retentivity a r e not desirable for wet method use.

Experience and extensive tests have shown, however, tha t coercive force and retentivity of a low o r d e ~ a r e advantageous. In the case of dry powders, a iow order of residual magnetism appears to increase sensitivity, especially in the diffuse ieakage fields formed by defects lying wholly below the surface. This may be for the reason that the small amount of polarity already established in elongated particles aids in lining them up into strings when the weak leakage fields act upon them. The action is similar to that of the compass needle swinging in the very weak field of the earth.

Wet method particles a r e actually chosen to have higher than min imun~ values of retentivity and coercive force. In this case the reason is cleariy understood. As described in Section 2 above, these ultra-fine particles begin to collect a t discontinuities a s soon a s they a r e applied to the test surface, when the agitation which had been present in the bath, ceases. With insufficient retained magnetism the particles remain fine and migrate very slowly through the liquid, due t o their small magnetic moments and small mass, and the viscosity of the liquid suspending medium. The Indications of discontinuities will build up; of course, but very slowly, taking a s long as five to ten seconds. On the other hand, if magnetically "hard" particies a r e used, the test surface is covered with large immobile clumps a s soon a s the bath is applied.

Particies having intermediate magnetic properties collect into clumps more slowly while the indications a t e forming. The leakage field: strongest a t the actual discontinuity, draws particles toward it, while the particles themseives a r e constantly eniarging due to agglomeration. A t the same time they sweep up the ultra-fine particles as they move toward the defect. In this way all the magnetic fields present work together.

7. HYSTERESIS CURVES. The overall magnetic properties of the

1 various types of magnetic particles can best be shown by their hys-

CHAITER 11 MAGNETIC I'AIITICLES-THEIR NATURE AND PROPERTIES

teresis curves. In making these curves, the powders a r e compressed rnto tubular holders, either rod-shaped o r toroidal. The filled holder is then tested as a solid rod o r ring wouid be, making allowance for voids remaining in the packed specimens. Specrai operating procedures a r e necessary to make the test reprociucibie. The resulting hysteresis curves obtalned in this manner a r e those for the packed, shaped specimens, and not those for the riidividuai magnetic particles or small agglomerations of particles.

Some typical hysteresis curves a r e shown in Figures 111 and 112. Figure 111 shows the loops for a typical dry powder, ( a ) , when the maximum magnetizing force, H,,,,,, was 500 Oersteds, and (b ) when H,,,;,, was oniy 50 Oersteds. The shapes of these curves show that the material has a high permeability, shown by the height of the curve, B ,,,,,,, a iow coercive force, H,., and a low retentivity, B,-an excellent combination of properties fo r this type of powder.

Figure 112 gives curves fo r typical wet method materiais again a t 500 Oersteds maximum magnetizing force ( a ) , and a t 50 Oersteds maximum f b ) . In thrs case the much lower permeability is ciearly shown by the much iower height of the curve B,,,,.; and the much higher retentivity, B,, and coercive force, H,., i s shown by the way the curve spreads out. Also, the striking difference in the proportion of retentivity a s between the higll i500 Oersteds) and the low

B, (50 Oersteds) levels of magnetization is clearly indicated. -

B",,, B,

= 0.38 a t the high level, and - - - 0.057 a t the low level. No such B,,,,,

difference is shown by the two curves of Fig. 111 fo r the dry method 370

powder, the proportions being - - - 0.065 a t 500 Oersteds, and 5610

80 - = 0.115 a t the 50 Oersted level. 690

However, the actuai values for B,,,., and H, fo r the two types of powders a r e not by any means a measure of their relative sensitivi- ties fo r the location of defects, because these powders are used under vastly different conditions, and factors other than the magnetic properties shown by their hysteres~s curves a r e controlling. Never- theless, the magnetic properties do play an important par t under each of the conditions of use.

PRINCIPLES OF IAGKETIC PARTICLE TESTING

CHIPTER 11

MBGNFI'IC l'ARTIC1.ES-'fRI?IIL KATUIIE AND I'IIOPERTIKS


8. ~ ~ O B I L I T Y . It is obv~ous tha t when the magnetic particles a r e applied over the surface of a magnetized part , they must move and gather a t a discontinuity under the influence of the leakage field to form a readable indication. Any factor tha t interferes with this required movement of the particles will have a direct effect on the sensitivity of the powder and the test. Conditions promoting or interfering with mobility are different for dry and wet method materlais.

t a ) Drg Powders. Ideally dry powders should be applied in such a way that they reach the magnetized surface in a uniform cloud with a minlmum of motion. When this can be done, the particles come under the influence of the leakage fieids while suspended in air , and Rave three-dimensional mobility. This condition can be approximated when the magnetized surfaces a r e vertical o r overhead.

When the particles a r e applied on a horizontal or sloping surface they settle directly to the surface and do not have the same degree of mobility. Mobility can be acliieved In this case; however, by tap- ping or vibrating the part, which jars the powder loose from the surface and permits i t to move toward the lealtage fields.

When A.C., or half-wave rectified A.C. o r puisating D.C. a r e used for magnetization, the rapid variation in field strength while the current is on, imparts motion to the magnetic particles on the .surface of the part. The "dancing" of the particies gives them excellent mobility fo r the formation of indications.

The coatings applied to some of the dry method powders to give color to the indications serve a double purpose in tha t they also reduce friction between particles and the surface of the part , thus contributing another factor aiding mobility.

(b) Wet Method Materzals. The earliest device for imparting mobility to magnetic particies was tha t of suspension in a liquid. Hoke's lnitiai discovery of the magnetic particle principle was made when the fine metallic grlndings were washed over the surface of the magnetized steel blocks which were being ground, by the liquid coolant used for grinding.

The suspension of particies in a liquid, wliicl~ may be water or an organic liquid such as a petroleum distillate, allows mobility fo r the particles in two dimensions when the suspension is flowed over

h

CHAPTER 11 hlAGNETIC PARTICLES-THEIR NATURE AND PROPERTIES

the surface of the pa r t ; and in three dimensions when the magnetized par t is immersed in the suspension.

Wet method particies have a tendency to settle out of suspension, either in the tanks of the unit or somewhere on the test surface short of the defect. To be effective, the magnetic particies must move along mith the liquid and reach every surface tha t the liquid covers, without "runmng aground" or settling out sornex\~here along the way. According to Stoke's Settling Law, particles settle out of suspension a t a rate tha t is directly proportional to thelr size and their density-that is, the density in excess of that of the liquid medium-and inversely proportionai to the liquid's viscosity. From this one mlght conclude tha t to slow down settling and improve the mobility of the particies one need merely use low density particles that a r e a s small a s possible, and use a more viscous liquid for the suspending medium.

None of these devices i s very useful, either alone or in combination, to Improve o r control the behav~or of particles when used in suspension for wet method defect detection. None of the properties of particles or liqurds can be changed a s suggested above without sacrificing some other equally important property. Ferromagnetic particles a r e by their nature of high density. Particle density may be changed to a slight extent by the addition of lower density pigments, but this lowers the magnetic properties of the compound particles in direct ratio. Therefore, this scheme can be used only where the need for color and contrast is great enough so that a small magnetic loss can be tolerated. Particle size has aiready been discussed (Section 2) . Lo~i'erlng it too f a r results in slow indication build-up. Ralsing the liquid's viscosity produces similarly undesirable results-that is, reduces the speed of indication buildup by slowing the movement of particles through the liquid.

As a result of all th is the mobility of wet method particles is never ideal, nor can all possible means for improving it be used. But, while i t must be balanced off against many other properties, mobility certainly is one of the factors which a r e important to wet method results.

9. VlsIBlLITY AND CONTRAST. These are additional important properties that have a great deal t o do with maklng a magnetic powder suitable for its lntended purpose. Size, shape and magnetic properties of a particle may be in all respects favorable toward

221

pRIiYClPLES OF RIACINETIC PARTICLE TESTING -

making a perfect magnetic particle test medium, but if, having formed an indication, the inspector does not see it; the url~ole Inspec- tion fails.

Visibility and contrast a r e promoted by choosing coiors of par- t icles that are e:rsy to see against the color of the surface or the test part. The natural coior of the inetallic poxvders is silver-gray, and we are limited to the coiors black and red in the iron oxides commonly used a s the base for the wet method materials. Visibility against certain colors of test surfaces can be increased by coloring the powder particles in some way. By use of pigments the silver- grey iron particies a r e colored white, black, red or yellow, a11 with comparable magnetic properties. Oiie o r another of these colors gives good contrast against the surfaces of most of the par ts that a r e tested.

Among the dry powders, the gray-white powder loas the Erst to be used, and i t gives good contrast against the surfaces of many test parts. I t fails to give good visibility, however, against tile silver- g rey of a sand- o r grit-l~lasted surface, or against bright machined or ground surfaces. Choice of colors should be made by the operator to provide the best possible visibility against the surfaces of the test par t under the conditions of shop lighting that prevail. In similar f a sh~on , cho~ce of either the black o r the red wet-method materiai is made to suit particular lighting conditions.

In some cases i t has beeii found advantageous to coat the 7~al.t bewzg tes ted with a coior to improve contrast. Chalk or whiting, in alcohoi, has been so used in the past fo r the inspection of large castings and weldments, when lighting conditions were poor in the areas where the ~nspection \\.as beiilg made. Aluminum paint has been slmiiarly used. During World War 11 many small par ts were glven a flash coat of cadmium, principally fo r the purpose of making the black particle indications of the wet method more visible. The extra cost of coating the test surface was justified only when conditions were such that good contrast could not bc obtaiiied by use of the powders and pastes tha t were then available. The de\Ilce of coloi--contrasting the par t is rarely used today, because the Auores- cent materiais now available solve the prohlern in a niuch belter way.

The ultimate in visibility and contrast is achieved by coating the magnetic particles with a fluorescent pigment, and conducting the search for indications in total or semi-darliness, using black ligllt

t o activate the fluorescent dyes used. \$'hen indications glow in darhlless ~vitl i their o\Irn light, it is almost impossibic fo r an inspector not to see them. Magneticnlly, these fluorescent materials a r e theoretically less sensitive tlian uncoated particles, but this reduction in magnetic sensitivity is more than off-set hy the fact tha t patterns of particles can be readily seen even when only a few such particles go to make up the indication. A fluorescent indieation easily visible under Blaclt light is often ouite impossible to see when viewed in white l i ~ h t . The aitvantage in visibility and contrast of the fluorescent materials is so great tha t they a r e being used in a very high percentage of all applications. This 1s an example of magnetic particles in a particular use: in which one property super- sedes in importance not only the magnetic properties, but also some of the other properties which we have been discussing.

Fluorescent magnetic particles a r e usually available in wet method materiais only. I t is quite feasibie to make fluorescent dry particles, but the conditions under which dry particles are used a r e usually such tha t satisfactory visibility under test conditions 1s achieved without the expense of using the fluorescent version of the method. Today, however, in many applications for which dry powders mere formerly used, the fli~orescent wet method has been adopted. This is t rue for the inspection of large castings, steel billets and other products, for which the wet fluorescent method gives quicker coverage and better and quicker viewing for indications than the dry powder can do. Development of the over-all method of magnetization for such large objects has had much to do with maliing the wet method, with its qu~ck , complete coverage of large SUI-faces, attractive for such tests. This type of inspection was formerly conducted by the dry powder method, almost \~i t l iout exception.

10. WET NETHOD MATERIALS. Wet method magnetic particles a r e fundamentally s ~ m i l a r to each other, once they are dispersed in the suspending liquid. But before this step they appear in a variety of fo rn~s . In the past, the most common form, in tl~is country a t least, was a paste, made up of the po\17der and a liquid vehrcie, and having a consistency much like that of a paint pigment when ground in oil.

klany of the raw materials a s received a r e in a rather grainy or lumpy condition. When atlded t o an oil o r water bath in this condition they rio not disperse, thus making a mixture xvl~olly unsuited

PRINCIPLES O F HAGNETIC PARTICLE TESTING

for magnetic particle inspection purposes. As in the case of paint pigments, a pre-dispersion by grinding the r aw material into a paste with a small amount of oil or other iiquid, produces a materiai more readily dispersed by dilution with the liquid of the bath.

However, the thick paste produced by this process is difficult to handle and requires a special procedure before i t \%rill biend properly into the bath. More recently the orlginal pastes have been re- made a s d ry powder concentrates. These powders a r e much easier to use, a s they need merely to be measured out and added directly to the agitated bath. The agitation system of the modern magnetic particle units !\,ill pick up the.powder and quickly disperse it in the bath in the ordinary process of circuiation and agitation.

11. PRE-MIXED BATHS. To a small extent in this country, and to a great extent in England, pre-mixed, ready-to-use baths a r e sold and used. In England they a r e called "inks". These materials are manufactured in the usual manner and then made up into a bath of proper strength in the supplier's plant. Though more bulk must be shipped and paid for, the user is sure of getting a well propor- tioned bath of correct strength, provided it is properly mixed before using. This is especially attractive to the small user. Such premixed baths are often used on a completely expendable basis. They a r e sprayed onto the test surfaces from a pump-agitated drum and not collected for re-use a s they a r e in the ordinary testing unit. In many cases small pressurizeti aerosol cans containing the bath a r e furnished. These are particularly convenient for quick "spot" tests, such a s fo r maintenance inspection.

12. THE SUSPENDING LIQUIDS. Wet method particles may be suspended either in water or in a petroieum distiliate, o r in any other liquid having suitable properties. Water is initially cheaper, but additions must be made t o it before i t i s a suitable medium for suspending the wet magnetic particles. Wetting agents, anti-foamlng materials, corrosion inhibitors, suspeiiding and dispersing agents a r e all necessary. Dry material concentrates to be used for water suspension must contain all the extra lrlgredients necessary to make the finished suspension. Cost of the concentrates is comparable for water o r oil suspension. The cost advantage of \\rater-base baths lies in the difference in cost between water and the oil used. An- other advantage of water for operations requiring large vo~uines of bath is the freedom from fire hazard. This hazard exists in small installations also, but. the few gallons of distillate in the ordinary

CHAITER 11 I1IAGNETIC PARTICLES-THEIR NATURE .4ND PROPERTIES

testing unit Rave, over the years, not shown flammability of the bath as a serious problem.

The need to incorporate into thc powder all the special ingredients which are needed f o r water suspension or for oil suspension, makes necessary t ~ v o separate and distinct products. A water-suspendable poivder cannot be used in oil. This is because the various additives a r e insoluble in oil and will not disperse the particles in an oil bath. The additions made to the powders intended for oil suspension are, similarly, not soluble in water. However, with suitable water conditioners, some of the oil-suspendable powders c a ? ~ be used in water.

13. AVAILABLE MATERIALS. AS was stated earlier, the need to meet a variety of conditions for successfui magnetic particle testing has resulted in the development of a large number of different materials to accomplish this result. Below is a list of many of these materials, 1~1th the special characteristics of each. Their specific applications and methods of use will be discussed in detail in subsequent chapters, especially in Chapters 13, 14, and 15. I n order better to identify the several types listed, current designations used by Magnaflux Corporation a r e given.

Commercially available powders fo r the dry method a re :

#I. Grey Po?oder. This is a general-purpose, high contrast powder, by f a r the most widely used of the dry powders. I t is effective on dark surfaces, whether black. grey or rust colored.

#2. Yellozu Poluder. A pale yellow color, featuring low cost with fai r sensitivity, and good contrast on dark colored surfaces.

#3A. Blaclc P07l~der. Especially designed for use on light col- lored surfaces. This is also the cleanest powder to use, as i t is dust-free. I t is also the most sensitive of the dry po~vders. Its higher sensitivity is due to the fact that i t contains the highest proportion of magnetic inaterial of all the dry powders.

#8A. Red Poluder. This dark reddish powder is used on light colored surfaces a s is the biack powder. However, since the black powder on a silvery surface is sometimes hard to see, the red color often offers a better contrast, particularly under ~ncantlescent lighting where the red color Stands out.

I L 1 I\G.VE.rIC P ~RTICLES-'I'EIEIR NATURE A N D PROPERTIES

PRISCIPLES O F h1AGXETIC PARTICLE TESTING -~

! ~ 2 9 ~ ~ . Red Pozoder Concent~ate. Tliis item is similar in 311 re- X11. B h c k Pozrder. This po\lrder is similar in color to the one

F ~ ~ O I C in the ~ r r l nou.der 9C listed above, except that i t is Y',LI"" .- . --- r~ ~ - - - -

listed above, under $3, but features a coarse, insensitive compoundfd for rrater suspension, and is for use with particle size and regular particle shape. I t flows easily and

! water only. is used for application with automatic equipment in cir- cunistances where discrimination for oniy the larger de- B. Flzrorescc?~t filaterzals. fects is sought

Commercially available wet iilethod materials a re :

A. Non-Fluorescent &faterials.

#7C. Black Po7ader Coi~ce~ltrate. This is a n oil-suspendable dry powder. I t is especially suited for finding fine cracks on polished surfaces? such a s bearir~gs o r crankshafts. I t is the ~ ~ z o s t se7lsitiue of the non-fluorescent wet method powders for such applications, though indications mag be hard to see. This is similar t o the oider 'iB black paste, which was difficult to put into suspension. -4 very limited amount of this paste (7B) is still made for use in equipment with the older air-agitation systems. This powder, made for oil suspension, becomes \$rater-suspendable when used milh tlie pi-oper proportion of water-conditioners listed below a s \\rA-ZA, or WA-4, in section C.

#9C. Red Po?oder Co?zcentrate. This is a reddish brown oil suspendable powder. I t is fully the equivalent of the biack polvder, 'iC above, for all applications except the very fine cracks in polished surfaces. The red coior gives ini- proved contrast and visibility in situations where tlie black-on-silver contrast of the black powder is poor. Aiso, this color tends to be more visible than the black under incandescent light. This red powder is similar to the old 9B red paste, which shares the limitations of the black paste described above.

As in the case of the black po\$'der, 7C, this red oil-suspendable materiai becomes suspendable in water when used with the proper amount of water conditioners listed below a s WA-2.4, and WA-4, 111 Section C.

#27-A. Black Poluder Concent~ate . This item is s ~ m i l a r in all respects to the black po\l8der 'iC iisted above, except that it Is compounded for water suspension, and is for use wit11 water only.

X10-A. Flziorescent Porode~ Co?zcentrafe. A blend of fl~lorescent magnetic powders and suspending agents, this powder concentrate is for use with oil only. I t fluoresces a blue- green coior, and features high fluorescent brightness, for tests \\,here lighting conditions are not the best.

X12. FLz(.o,.escent Paste. This is an obsoiescent blue-green- fluorescing paste composed of lo\v fluorescent brightness magnetic particles and dispersing materials. I t is for use with water oniy. In those places where i t is still used it is favored because of its low tendency to foam, aiid its weak fluorescent background.

X14-A. Fl.?~o?.escent Poa:d.e~ Concentrate. This is the most widely used of the fluorescent materials. I t fluoresces a bright yello\r-green and features high sensitivity, high (but not estreme) fluorescent brightness, and easy handling properties. I t is designed f o r suspension in oil, but can be used for \\rater suspension \vhen combined with the proper amount of water conditioners iitsed belo\' a s WA-24, and WA-II in Section C. I t stays in suspensioil well and does not give objectionable fluorescent background.

X15. Fluoresce?tt Pozode? Co~~ce?z t~a te . This powder concentrate is identical in composition with 14-A above, but is made up entirely of larger individuai particles. I t is usually used as a water suspension but i t niay be used in oil also. Tliis material Finds its greatest application sus~ended in \vater fo r steel billet inspection, wilere only the larger defects a r e sought. I t i s aiso most effective for the inspection of steel castings and similar "toothy" surfaced parts, which would present a backgrountl problem with fine-sized particles. Because of its iarge particle s ~ z e i t 1s difficult to keep in suspension. The \\rater conditioner for use with it is estm-lo\\, foaming to prevent the bath from foamlng escesslvely during the more violent agitation required to lieep the particles suspended.

PRINCIPLES OF BlAGNETlC PART1CI.E TRSTING

#20-A. Fluorescent Pozudel- Concent?-ate. This powder is identical to 14-A above, but 1s compounded to include the proper amount of water conditioner. I t is for use wlth water only.

#24. F1,uorescent Poiadel- Concentrate. Thls is a special iron- base powder of extreme fluorescent brightness. I t fluoresces a bright yellow-green. I ts high density makes i t more difficult to keep in suspension. Some users prefer i t because they have found i t more sensitive in some applications, though its superiority over 14A above is hard to demonstrate. I t has the highest magnetic permeability of any of the fluorescent materials.

C. W a t e r Conditioners.

WA-2A. This material is a powder deslgned to help suspend the wet method niaterlals in water-particularly the 14A powder listed above, which is somewhat difficult to wet with water. This wetting agent features excellent wetting and spreading on oily surfaces, moderate to low foaming, good corrosion inhibition and a tolerance f o r oil con- tamlnation in water baths.

WA-3. This materiai is designed for use with the coarse particle fluorescent material listed a s #l5 above. In extremely well agitated inspection units, this conditioner features minimum foaming together with ability to suspend the difficult- to-wet coarse powder. I t has ?LO Corrosion inhibiting properties.

WA-4. The conditioner listed a s WA-2A above 1s slow to dissolve in water. Thls product (WA-4) is a liquid and was developed to eliminate that feature. Besides being rapldly soluble in water, thls product shoxvs strong corrosion inhibiting properties, although it foams slightly more than the WA-2A above. Like all water conditioners, this pi-od- uct 1s not affected by the hardness of the water used to make up the bath.

BASIC VARIATIONS IN TECHNIQUE

1. INTRODUCTION. In the application of magnetic particle testlng there are a considerable number of possible variations in procedure whlch critically affect the results obtained. These variations are made necessary by the many types of defects that are sought, and the many types of ferronlagnetic materials in whicli these defects must be located. The need to assure ad'equate sensitivity, and maximum reliability a t a n opti?nu?iz cost, has also caused many special techniques to be developed.

In order to make the proper choices among these variables for any given testing problem, the operator must Itnow what the possible varrations are, and how each affects the end result. A given technique may be either favorable or unfavorable, depending upon the specific case lie is considering.

2. LIST OF VARIATIONS IN TECHNIQUE. The following list gives the most important variations in the techniques used to apply magnetic particle testing:

( a ) T y p e of Current. Possible current types are A.C., D.C., half wave rectified slngle phase A.C., full wave rectified slngle phase A.C., and full \vave rectified three phase A.C.

(b ) Tvpe o f Pavtrcles and Method of Application. Magnetic particles may be applied dry or in suspension in a liquid-that is, the Dry ntethod or the Wet Method. Particles may be any of several colors, or may be fluorescent.

tc) Seqzcence o f Steps. The par t may be magnetized first and particles applied later af ter the magnetizing force has been turned off ( the reslduai method) ; o r the part may be covered with particles while the magnetizing force is acting ( the cantinuous method). With par ts havlng hlgh retentivity, a combination of these methods 1s sometimes used.

f d ) Divectio?~ of Field. Circular, longitudinal or other methods of magnetization a r e involved in this choice, as well as methods of malting contact o r applying a coil.

PRINCIPLES OF XACXETIC PABTICLE TESTING

ie) S c ? ~ s l i i v i f y Level . This choice has to do principally with tlie amount of current or other magnetizing force and the test medium, as well a s proper appiication by whatever method has been decided upon.

( f ) Eqt i ipme,~t . Stationary or portable units, use of prods o r yokes, o r special-purpose or automatic equipment may be selected, depending on the conditions invol\zed.

3. CHARACTERISTICS OF DEFECTS BND PARTS WHICH INFLUENCE THE PROPER CHOICE AMONG THE SEVERAL VARIABLES. 111 deciding Bow best t o proceed to apply magnetic particle testing and what ihoices to make among the possible technique variations, the characteristics of the sought-for or expected defect, and of the par ts in which they occur, must be surveyed in order to arr ive a t the optimum in sensitivity, reliability and cost for the test.

Following a r e a list of considerations with respect to the char- acteristlcs of defects and par ts mhicti determine the proper choice of procedure :

i a ) Are the defects open to tlie surface or wholly below the surface?

(b) Are they fine and sharp. o r a r e they wide open?

(c) Are they shallow o r do they go deep into the metal?

td) What is their physical size and shape?

(e ) If they have direction, are they longitudinai o r transverse to the axis of the pa r t ?

( f ) What is their location and direction with respect t o the stresses to whlch the par t will be subjected in servlce?

( g ) What is the service for wiiich the par t is intended-is i t critical as, fo r instance, a n aircraf t engine o r landing gear part-or would i t s failure involve no drastic consequences, a s f o r exampie, breakage of a tool?

(h) What 1s the size and shape of tlie par t in which the defects occur?

( i ) What a r e the magnetic characteristics of the p a r t in which they occur?

(j) What a r e the overall economic considerations? Test costs

vs. costs of failure, such as loss of good\vill, damage to customer s equipment, etc. l

4. PRIMARY ~IETHOD CHOICES. There a r e two b a s ~ c Ueclsions which n i~ is t be made before the details of operating procedure can be decided upon. These a r e :

( a ) T y p e o f Czu,-ent t o be Llsed. This choice is dictated by the location of the defects, %'lictlier they a r e open to tlie surface of the part, or whether they a r e located ~vlioliy below the surface. Clloice of current lies between A.C. and some form of D.C. If the defect is open to the surface, either A.C. or D.C. may be suitable and the choice is determined by other considerations. I f the aefect lies wholly below the surface A.C. cannot in general be used. Instead, D.C., either straight D.C. or some type of rectified A.C. i s indicated.

(h) T y p e o j dlagnetie P a ~ t i c l e s to be Used. This choice is primarily between the d ry and tlie wet method, and secondarilg among the various colors that a r e available, Including floures- cent particles. The dec~sion is influenced principally by the following considerations:

11) Whether the defect 1s on the szrrfnce o r roholl?~ below t h e sur face . F o r deep-lying defects below the surface the dry powder is usually more sensitive.

12) The szze of the defect if o n tire s ~ c ~ t a e e . The wet method is usually best for very fine and shallow defects.

i3 ) Convenience. Dry powder with a portable half wave unit for instance, is easy to use for occasional large par ts in the shop o r foundry, or for field inspection Work.

5. CHOICE OF TYPE OF CURRENT-A.C. VS. D.C. This is the first basic choice to be made, smce the skin effect o i A.C. a t 50 or 60 cycles per second frequency limits its use to the detection of defects which a r e open to the surface, o r only a few thousandths of an inch below it. However, the skin effect of A.C. is less a t lower frequencies, resulting in deeper penetration of the lines of force. A t 25 cycles the penetration is deinonstmbly deeper, and a t frequencies of 10 cps and less, the skln effect i s almost non-existent.


If the defects sought a r e a t the surface, A.C. has several advantages. The rapid reversal of the field imparts mobility to the particles, especially to the dry po!vders. The "dancing" of the powder helps i t to move to the area of leakage fields and t o form stronger indications. This effect 1s less pronounced in the !vet method.

Alternating current has another advantage in tha t the magnetizing effect, which is determ~ned by the value of the peak current a t the top of the sine wave of the cycle, is 1.41 times that of the current read on the meter. Alternating current meters read more neariy the average current fo r the cycle ra ther than the peak value. To get equivalent magnetizing effect f rom straight D.C., more power and heavier equipment is required. Thus A.C. equipment f o r a given output of magnetizz?zg force can be lighter and less costly, and better adapted f o r portability.

COLUMN I COLUMN 2 COLUMN : (Aci I0C:BATTERIES) IDC..RECTIFIEDI

420 Amp.

I95 Amp 595 Amp

Fig. 113--Campanson of Indications of Surface Cracks on a Part Magnetized with A.C., Stralght D.C. and Three Phase Rectified A.C.. Respectively.

D.C.; on the other hand, magnetizes the entire cross-section more o r less uniformy in the case of longitudinai magnetization, and

I with a straight line gradient of strength from a maximum a t the surface to zero a t the center of the bar in the case of direct contact (circular) magnetization. See Chapter 9, Sections 20 and 25, fo r

CHAPTER 12 BASIC VARIATIONS IN TECHNIQUE

field distribution in and around a magnetic bar carrying A.C. and D.C. respectiveiy. Figure 113 IS a comparison of indications of the same set of fine surface craclts on a ground and polished plston pin, obta~ned by the use, respectively, of 60 cycle A.C., D.C. from storage batteries (straight D.C.) and D.C. from rectified three phase 60 cycie A.C. Four values of current were used in each case, using a central conductor to magnetize the hollow pln. The indications p ~ o - duced with A.C. a r e heavier than the D.C. indications a t each current level, although the difference 1s most pronounced a t the lower current vaiues. Straight D.C. and rectified A.C. are comparable in all cases. The A.C. currents a r e meter (R.M.S.) values, so tha t peak-of-cycle currents, and therefore magnetizing forces, a r e 1.41 times the meter reading siiown.

Another cornpanson, which graphically shows the advantage of direct current fo r the detection of defects lying wholly belozu the surface, and the limitations of A.C. fo r this purpose, is shown in

1

i I Fig. 114-Drawtng ot a Tool Steel Ring S p e c m e n

with Artificaal Sub.Surtace Detects.

PRINCIPLES OF AlAGNETIC PARTICLE TESTING

Figs. 114 and 115. Figure 114 is a drawing of a r ing specimen of unhardened tool steel t.40 carbon), xi inch thick. Holes parallel to the cylindrical surface have been drilled, 0.07 inch i n diameter, a t increasing depths below the surface. The depths vary from 0.07 to 0.84 inch in 0.07 inch increments, f rom hole #1 to liole +12.

Figure 116 is the plot of the threshold values of currelit necessary to give a readable indication of holes in this ring, by the dry continuous method with central conductor, using 60 cycle A.C. and three forms of D.C. The three types of D.C. a r e straight D.C. f rom batteries, three phase rectified A.C. with surge, and half wave rectified single phase 60 cycle A.C. Currents a s read on the usual meters a r e varied from the minimum necessarv t o indicate hole #1

Fig. 1 1 j C o r n p a r l s o n of the Sensitivity of A.C.. D.C.. D.C. wlth Surge, and Half Wave, for Locating Defects Wholty Below the Surface.

234

C l i r r m 12

BASIC \',IRIATIDNS IN TECHNIQVE

in the case of each type of current, up to a maximum of over 1000 amperes. .4ltcrnating current required about 475 amperes to silow' hole #I and over 1000 amperes to shorn liole #2. Hole #3 could not he shown with A.C. a t any current value available. Hole #2 was shown with 475 amperes s t ra~gl i t D.C. , with 275 amperes D.C. preceded by a surge of double this amount, and by 175 amperes of half wave. Seven hundred and fifty amperes of half \irave showed hole #12, while 975 amperes were needed to show hoie ti10 with straight D.C.

F o r the inspection of finished parts, such as the macllilied and ground shafts and gears of fine machinery, dirett current is frequently used. Although A.C. is excellent for the iocatiun of fine cracks tha t actually break tlie surface, D.C. IS better for iocating very fine non-metallic stringers i y ~ n g lust under the surface. It is usually important to locate such stringers in parts of this type, since they can initiate fatigue failures.

These comparisons p o ~ n t up the importance of choosing the right current type to give the best indications possible, and show how the choice will vary, depending upon the nature and Iuc~t ion of the defects sougilt.

6. CHOICE OF TYPE OF MAGNETIC PARTICLES. DRY VS. WET M E ~ i i o D s . This is the second basic method choice to be made. The dry powder method is superior for locating defects lying \i'holly below the surface because of the high permeability and the favorable elongated shape of the particles. These form strings in a ieakage field and bridge the area over a defect. A.C. with dry powder IS excellent fo r surface cracks which a r e not exceedingly fine, but a s shown in the comparisons of Fig. 115, IS of little value f o r defects lying eJTen slightly below the surface.

Figure 116 is a comparison of the effectiveness of the dry method and the wet method for detecting defects lytng wholly beiow the surface, uslng the same unhardened tool steel rlng described in Figure 114. I t is clear from this comparison tha t the dry method is superior to the wet f o r this purpose a t any value of direct current used.

A fur ther comparison of the \vet and dry methods is shown in the graph of Fig. 117.

Ho\vever, when the problem is to find very fine surface cracks, there is no question a s to the superiority of the wet method, whatever form of magnetizlng'cwW6nt is used. In some cases, direct

PRINCIPLES OF nfAGNETIC PARTICLE TESTING

SUBSURFACE DEFECTS C0l. 1

Dry Powder Continuous Method

I ROO Amn D r I --- .....-. --

1600 Amp. DC

COI. 2 Wet Continuous Method

800 Amp. DC

Fig. 116-Cornpanson of the Dry and the Wet D.C. Method for Detects Lying Below the Surtace.

current is selected for use with the wet method so a s to get the advantage of better indications of discontinuities which lie just below the surface, especially on bearing surfaces and aircraf t parts. The wet method offers the advantage of easy complete coverage of the surface of par ts of all sizes and shapes. Dry powder is often used for very loeai mspections.

Selection of the color of particles to use is essentially a mat ter of securing the best possible contrast with the bacitground of the surface of the par t being inspected. The differences 111 vis~bility among the black, gray, red and yellow particles are considerable on backgrounds whlch may be dark or bright, and which may be

C H ~ P T E R 12

UlSlC YARIATIOiTS IN TECHNIQUE

TOOL STEEL RING - UNHARDENED

Fig. 117--Campanson of the Dry and the Wet Methods tor Detects Lymg Below the Surlace, when A.C. and D.C. are used tor Magnetization.

viewed in various kinds of light. Black stands out against most light-colored surfaces, gray against darli-colored ones, like those of castings. Red is more visible than either against silvery and polished surface%especially if the light by which the inspection is made i s f rom ~ncandescent lamps. The yellow-coiored powder is not used very estensively, but has some advantages under some certain conditions of lighting. If difficulty is had in seeing indications the operator should t ry some color of powder other than the one lie IS

using. In the case of the wet method, a s was stated in Chapter 11, the

ultimate in visibility and contrast is obtained by the use of fluorescent particles. The fluorescent wet method has been used in constantly increasing numbers of inspection applications for many years, principally because of the ease of seeing even the f a ~ n t e s t indications.

7. FIRST OPERATING DECISION: RESIDUAL VS. CONTINUOUS METHOD. This cholce between the rewdual and the continuous

P1IINCIPLES OF BlAGNETlC PARTICLE TESTING

method is a relatively easy one. I n the residual metiloll, par ts are magnetized and subsequently the magnetic particles a r e applied. The nietliod can be used only on par ts having sufIic!ent retentivity. The permanent field they retain must be sufficiently strong to produce leakage fields a t discontinuities which in turn will produce readable indications. The method in generai is reliable only for the detection of surface discontinuities.

Since hard materials which have high retentivity a r e usually low in permeability, higher-than-usual magnetizrng currents may be necessary to obtain a sufficrentiy high level of residual magnetism. Some idea of the effect \\,llrch the iower permeability of highly retentive steels has on necessary magnetizing current is shown in Fig. 118. The tooi steel ring of Fig. 114 was hardened to Rockwell C-63, and agalil tested with the dry continuous method using several forms of direct current. The curves mdicate that 4000 amperes D.C. was Iieeded to show hole #5. Even half wave a t 1500 amperes showed only hole #4.

MAGNETIZING CURRENT-AMPERES 0 1WO P O W 3 0 W

SURFACE

t, 2

w Y ' UC

CL

2 $9, xu ZLi q3*

E * * 9 .a

THRESHOLD INDICATIONS - DRY CONTINUOUS METHOD TOOL STEEL RING - HARDENED TO ROCKWELL C-63

Fig. 118-Effect of the Hardness ot a Spec~men on the Current Requ~red to Locate Defects Lymg Below the Surtace.

The difference in the behavior between hard steels and soft steels IS usually not very serrous if only surface discontinuities a r e sought.

Cl3 \PTER 12

R4SIC IAItI.41'IOXS IN TECHNIQUE

I t becomes important Ivhen illterest lies 111 defects belo\%, the surface. Then tlie figure of 500 for the nlmimunl value of material pernieability for a metal or alloy to be suitable for magnetic particie testrrig may become a lim~tation for sonie hard steels and alloys. (See Chapter 9, Sections 9 to 11.)

Either the dry or the wet methoti fo r particie application can be used in the residual nletllod. With the wet n~ethod, tlie magnetized parts may be immersed in an agitated bath of susl~encled magnetic particles, or they may be flooded \ritli bath by a "curtain spray". In these circurnstances a favorable factor, affecting the strength of indications, enters. Thrs is the t i~ i t c of immersron of tlie par t in the bath. By leavi~ig the magnetized par t in tlie bath or under the curtain spray for a considerable time; the ieakage fields, even a t fine discontinuities, can have trme to a t t ract and hold the maximum number of particles. This can mean an increase in sensitivity over the mere flo\ving on of the bath over the surface of the par t as it rs b a n g magnetized by the continuous method. I t should be noted, however, piat the location of tlie discontinuity on the part a s i t IS

immersed affects particle build-up. Build-up \\-ill be greatest on horizontal itu,,er surfaces, and less 0x1 verticni surfaces or horizontai lower surfaces. Also, raprd \r2ithdrawal from the bath or spray can wash OR rntlications held by extremely weak ieakage fields, aud care must he exercised in thrs par t of tlie process.

The residual niethod, either wet or drl', has many attractive features and finds many applications, even thougl? the continuous method has the znherent advantage of greater sensitivity.

This advantage for the continuous method is srniple, but basrc. \<hen the magnetizrng force is applied to a ferromagnetic par t , the field rises to a inaxrmum on tile hysteresrs curve, its value or In- tensity rlerivrng from the strength of the magnetizing force and the matenal permeability of the part . But \&en the magnetrzing force is removed, the ~eszdual ?nngwetisr~! in the art IS alzoal~s less than the field present while tlie magnetiziiig force \\,as acting. The amount of difference depends on the retentivity of tlie material. The continuous method, for a given value of magnetizing current, is therefore al?i:a?/s more sensitive than tlie residual; a t least so f a r a s sensitivity 1s determined by the strength of field in tlie part.

Aiso, techniques have been nrorl<ed out fo r the continuous niethod wilich make i t faster than the residual. The rntlicatron is produced

239


at the time of magnetization, whereas the residual method requires two steps-magnetization and application of particles-plus tlie added time f o r indications to build up if the immersion method is used. The choice therefore, a s between residual and continuous methods, should be f o r the continuous method unless special circumstances make the residual method more desirable. F o r example, the residual method is frequently used in special purpose and automatic equipment because the timing of magnetization and application of particies is not critical and the entire process is therefore more easily controlled.

The continuous method is, of course, the only possible one to use on low carbon steels o r iron having little o r no retentivity. I t IS

frequently used with A.C. on such materials because the alternating current field produces excellent mobility of t he particies. With the wet method the usual practice is to flood the surface of tlie par t with the bath, then simultaneously terminate bath application and close the magnetizing circuit switch momentarily. Thus the magnetizing force acts on the particles i n the film of bath a s they a r e draining over the surface. A "shot" of ?/, to of a second-either A.C. or D.C.-is sufficient time for the current to be on. Strength of the particie bath has been standardized to supply a sufficient number of particles in the film to produce good indications with this technique. I t should be noted, however, that the continuous method requires more attention and alertness on the par t of the operator than does the residual method. Careless handling of the bath-current sequence can interfere seriousiy with reliable results.

Probabiy the highest possible sensitivity obtainable f o r very fine defects is achieved by immersing the par t in the wet bath and passing the magnetizing current through the par t for a short time while imme~sed; and then leaving the current on while the par t is removed from the bath and while the bath drains from the surface. This technique is used today for the inspection of jet engine compressor blades and vanes, using fluorescent magnetic particles.

8. SECOND OPERATlNG DECISION. CIRCULAR VERSUS LONGITU- DINAL MAGNETIZATION. This declsion is determined by the shape and orientation of the defect in relation to the shape and principal axes of the part. The rule of thumb, that the current must be passed in a direction parallel to the defect, is controlling. This rule may require errcular magiietizat~on in some sections of the part , and longitudinal magnetization in others. Of course, if the principal direction of

CHAPTER 12 BASIC V.IRIATIONS IN TECHNIQUE

discontinuities is unknown both circular and longitudinal magnetization must be used, in order tha t all possibly-present discontinuities be located.

When feasible, circular magnetization is preferred, because i t produces a minimum of external poiarity that will attract and hold particles where no discontinuity exists. Particles so held may confuse or prevent the operator from seeing the actual indications of defects.

9. THIRD OPERATING DECISION. AMOUNT OF CURRENT REQUIRED. The amount of magnetizing current, or the number of ampere turns to use for opt imun~ results, is the final factor in determining the sensitivity level of the process. The types and ininimum dimensions of the defects that must be located-or the size and kind of defects that may be tolerated-are the major considerations in fixing current levels. The type of service for which the par t is intended, the importance of the par t in its assembly, and the effect on the continuing performance of the assembly if this par t fails, are further important factors in determining the sensitivity level required. For highly critical par ts maximum sensitivity would be called for in order that all minute discontinuities be located. Since the most critical defects a r e practically always on the surface, thls situation would probably call fo r the wet continuous method, with circular magnetization if possible, and the use of fluorescent particles.

I n some primary metal and industrial applications, sensitivity 1s deliberately limited and controlled so as to delect only gross defects and to ignore the small ones. This is true in the inspection of steel billets fo r seams. Here only defects over a certain depth a r e of interest, since sliallow ones .rt.ill scale off in sullsequent re-heating and rolling of the billets. Sensitivity here iscontrolled by the amount of current used, by the character of the magnetic particies, and by sequencing of magnetization and bath application.

The amount of current is initially determined by the rules of thumb given in Chapter 9. The formula for longitudinal magnetiza-

45,000 tion with a coil is: Ampere Turns = - , where L is the length

L/D and D the diameter of the part. (See Chapter 9, Section 8.) Fo r circular magnetization the rule is t o use 1000 amperes per ~ n c h of par t diameter. This rule cannot, however, be followed in all cases. For small diameters a somewhat smaller current is usually suilicient.

241

PRINCIPLES O F 3IAGXETIC PAICTICLE TESTING

Current must not be so high a s to heat the par t enough to affect its heat-treated structure, or to cause b i~rn ing due to the resrstance a t the points of contact. On the other hand, for large diameter par ts the level of current called for by the rule 1s estremeiy high and in many cases not available. Experience has shown that fo r most steels, fo r tlic location of surface cracks only, conslderably loxvcr current values than the rule calls fo r a r e sufficient. On large parts, if the purpose of tlie i~lspection is to locate defects lying w1ioliy belour the surface, either locai magnetization with prods must be used? or else high current for o\,erall magnetization must be provided.

If the current to be useti is A.C. or half wave, the above rules of thumb for current strengths \<'ill call for liiglier currents than needed, since tlie fields produccd by A.C. and half Israve tend to be stronger a t the surface than in the case of sinillar vai l~es of D.C. However, these rule-of-thumb current values a r e only g u ~ d e s to s t a r t from. Experleoce \\'it11 the results produced show quickly wliether the cllosen value is hlglier than required.

lo. THE FOURTH OPERATING DECISION. EQUIPI\IENT. This dc- cision depends on tlie slze, shape, number and variety of par ts to be tested. For production testing of numerous parts \\'Inch a r e relatively small, but not necessarily identical in shape. a bench type unit with clainping liead contacts for c~ rcu i a r magnetmation, and a built-in coil for longitudinal niagnetizatioii 1s commonly used. In such applications the wet continuous n ~ e t l ~ o d is g e ~ ~ e r a l l y selected, although some dry powder units of tills type have been in servlce.

If the par t or parts arc Iasge, portable ulllts using prods or C-clamp contacts and hand-\\,rapper1 coils may be ~i iost convenient.

Half-\\rave and dry po\r:dcr a r e often used with suc l~ portable equipment, a s in the inspection of \velds and large castings. The \vet method is also used \vith such portable ~ q u ~ o m e n t , in \\31i~ch case the bath usually IS not rccovcred and rc-used, but is allotved to drain alvay-the "expendable" technique. This system, usiilg fluorescent particles is being uscci incl.e:lsingly totias, especially \rritli the o ~ e r a l l metliorl of niagirctizatiun.

\Vhen the nuniber of den tical or closely similar par ts 1s large, as In m:tss pl~odi~ction opelxtioiis, s ingle-~~urpose magnetization and inspcction utiits or jixs on multi-purl:nse ui~i ts , usually automatic o r semi-;luton~:~t~c, ;11.c ofteii designed ;ind built. I n such cases the est1.a cosl. or tlic unit is jostifierl by greater oulpul per inspector,

CrInl'TFn 12

BASIC VAIII:LTIO.VS I N TECIINIQUE

and by niorc uniform and reiiabie inspection. Reproducibility of rcsolts is assured sillce each part receives exactly the same pre- selected treatment dnring niagl~etizalion and bath application. Automat~c units of this type invariably use the we1 melliod, because i t is easier to corer all surfaces of parts with a liquid suspension of particles than \\,it11 a spray or cloud of dry powder. (See Chapter 19.)

11. OTHER OPERATING DECISIONS. There arc a number of modifications and variations of the standard teclinlclues for magnetization and particle application that apply lo certain sizes and shapes of parts, and specific inspection conditions. \\'hen any of these variations a r e applicable, special tooling IS usually necessary. The choice of one of these special methods IS clictated by their ability to do the testing job more reliably and a t lower cost than can be done by more standard procedures. Some of these \rariations a re :

( a ) Multi-directional magnetization.

I b) The overall method.

( c ) The 111duced current method

( d ) The use of automatic units.

In some cases no power is available to operate magnetizing equipment in the area where the inspection must be ca r r~ed out, or the nature of the inspection does not juslify the purchase of more es- pens~ve equipment. I n such cases permanenz magnet yolres may be used; or electromagnet yokes operated either by A.C. o r rectified A.C. if some current source can be had. Yolies using automobile or other storage batteries a s a source of energizing current a r e another solution of the no-power problem. In explosive areas, pernia- nent magnet yokes fu rn~s l i a safe way to secure local magnetization for maintenance inspection. Inspection with these devices is usually limited to spot clieck~ng or occasional testing of miscellaneous parts.

CHAPTER 1 3

THE DRY DlETBOD-filATERIALS AND TECHNIQUES

1. HISTORY. \$'hen A. V. de Forest first applied magnetic particie testing to the drill pipe problem, he used the dry powder method. A t that time there was no record of performance for eitlier the dry o r the ~ ~ ' e t methods. In Hoite's original observation of inagnetic particic patterns on steel blocks, the fine grindings from the bloclis themselves provided the magnetic particles and they were carried by the grinding coolant tha t he wXas using. This was, therefore, the !\let method. But fo r de Forest's work some form of "iron filings" was doubtless readily a t hand-the mater,al may even have been mill scale, pientiful around a pipe mill. As it served his purpose, he continued to use the dry form of magnetic particles.

Having visualized the value of the new method in broader applications, he undertook t o improve the powder a s a means of securing reliabie and reproduceable results. He recognized that control of particle size and shape, a s well a s permeability, mere important fo r a satisfactory powder. Furthermore he saw the need for a colored powder to make a better contrast agalnst the surfaces of the pa r t being tested, and early devised a means for coating the particles to give them a gray-white color.

The material for the dry method was thus fully developed a t the outset. of commercial use of the magnetic particle testing method, and there has been little basic change since that time. Nuch more precise controi of particle size and shape, better coating methods and additionai colors have improved the performance of the dry powders, but essentially they a r e the same a s originally worked out by Doane and de Forest in 1929. The use of particles in liquid suspension, now more presalent than the dry method, was not begun until the middle 1930's.

The dry method was first use<! extensively by the railroads, the automotive industry and foundries, and during World War I1 fo r the inspection of welds and steel castings on all manner of war materiais from gun mounts to submarines. i t is still used somelvhat in similar fields today, although in many instances, even for these

CHAI'TER 13

THE DRY RIETHOD-M.4TERIALS AND TECHNIQUES

purposes, the wet niethod has taken over. T h ~ s appears to be due largely t o the development of fluorescent particles fo r the wet method, with their many advantages. Nevertheiess the dry method retains a field of importance of its own, a s in the inspection of welds and castings where the detection of defects lying wholly belol\' the surface is considered important.

2. ADVANTAGES AND DISADVANTAGES OF THE DRY METHOD. In tlte previous chapter (12) the good points and the drawbacks of the dry method were quile thoroughly discussed, and comparisons draxvn with the wet method. This discussion may be briefly summed up in the following list:

Advantages:

[ a ) Excellent for locating defects wholly below the surface and deeper than a few thousandths of an inch.

(b ) Easy to use for large objects with portable equipment.

(c ) Easy to use for field inspection with portable equipment

td) Good mobility when used with A.C. o r Half Wave

(e ) Not a s mess^," a s the wet method.

( f ) Equipinent may be less espenslve

( a ) Not a s sensitive a s the \vet method for very fine and shallow , cracks.

Not easy t o cover all surfaces properly, especially of irregular shaped or large parts.

& Slo~ver than the wet method for large numbers of small parts.

( d ) Not readily useable for the short, timed "shot" technique of the continuous method.

te) Difficult to adapt to a mechanized test system.

3. MATERIALS. Chapter 11 contams a description of the dry magnetic particles and discusses their various characteristics, contrasting tiiem with the wet method materials. Outstandiiig properties of the dry powders are their more favorable shape and higher permeability in conlparison %xrith the wet particles. It is these two characteristics more than any other tha t are responsible for the

I'RlKCIP1,ES OF %lAGKI.:TIC PAIiTICLE TESTING

good performance of these dry ponders within their field of al)- plication.

The a~a i l ab i l i ty of four colors-grey, red, blacic and yello\\;- increases the chance of gett ing good visibility under almost any conilitions of baci(ground or lighting.

4. STEPS I N AFPI,YING THE DRY METHCID. There a r e five essential steps in applylng tlie dry method of magnetic particle t e s t ~ n g .

These a r e :

( a ) Preparation of the surface.

(b ) Nagnetization of the part .

( c ) Application of the powder.

td ) Blo~rr-ofT o r removal of excess powder.

( e ) Inspection fo r indications.

In a fcm instances some cleaning of the par ts a f t e r testing is required to remove adhering powder.

5. SURF.ACE PREPARATION. I n general, the snioother the surface of tire pa r t to be tested and the more uniform its color, tlle inore favorable a re the conditions fo r the formation and tile observation of tile powder pattern. This statement applies particularly to inspections heing made on horizontal surfaces. For sloping and vertical surfaces, the dry ponrder may not be iicid on a very smooth surface by a weak leakage field. Tlie surface should be clean, a s f a r a s possible, arid dry and free of grease. The dry particles will stick to wet o r oily surfaces, and not be free to move around over the surface to form indicaiions. This masT completely prevent the detection of significant discontinuities. On surfaces tha t have been cleaned of grease by wiping with a r a g soaked in naphtha, a t h ~ n film of oil often rem:iins tha t is sullicie~it to interfere \\fit11 the f-;sEr movement of the powder. This film can be removed by dusting the surface \\,it11 chalk o r talc f rom a shaker can, and then w i p ~ n g tlle surface with a clean dry cloth. An initial application of the dry magnetic powder itself, followed by wiping, often \%,ill give a surface over which a second application of po\rlder \'ill movc readily. I f i t is feasihle to use it, vapor degveasing \%sill give a dry, oil-free surface.

Any loose dirt , ]~;iint; rus t o r sc:ile should be removed with a \!,ire bl.ush, o r hy shot o r g r i t blasliiig o r otller means. If cleaning 1s done \vith shot or g r i t blasting, there is a peening effect; especially on

softer steels, \vllich may close tip fine surface discontinuities. The etiect is more pronoi~ilced c.itl, shot than \\,it11 gr i t , hut if these clean in^ methods a r e used t l ~ e operator shotild hc aware of tlle danger of mlssing very fine cracl.rs. A thin, Ilard. unif0rm Coating of rozt o r scnlc \rill not usually rnterfere with the detcctioii of any bu t the smallest defects. The inspector sllouid know t h e srze of the smallest defect he 1s to consliler significnnt, in order to jtiilge ~she t l i e r o r not such a coating of rus t o r sc;iie should bc renlo\reU.

Paint o r plating or1 tlie surface of a pa r t has t h e effect of making a surface defect behaue like a sub-surface one. The relative thickness of tlie plating o r paint film and the size of the defects souglrt determine u.liet1ier o r riot the coatings sl~ould be s t r ~ p p e d . Tlle dry meihod 1s more elfcctive i n producing indications tllrougli sucll non- magnetic coatings than the wet methocl, but if fine cracks a re expected the surface should be strippet1 of tlie coatirig if its thiclcness exceeds 0.005 inci1. Most coatings of cadmium, nickel or chromium a r e usually t tdnner than thls, and t h e plating inakes a n excelleiit background fo r vlemilig lildications. IIot gal\'anized coatings, llow- ever, a re thicker, and in general should be removed before testing, unless only gross discontinuities a r e important. Broken o r patcliy layers of heavy scale also tend to interfere, by their tendency to hold poxrder around the edges of tlle breaks o r patches, and should be ren~oved if they a re extensive enough to interfere seriously with the detection of genuiile discontinuities.

6. MAGNETIZATION. All tlte usual metliotls of magnetization a r e applicable to the d r y po\vder method or testing. Although the residual method is sometimes used xvith dry powder, tlle continuous method is generally employed. This is especially t rue \sRen A.C. o r half wave 1s b e ~ n g used because these types of' current give mobility to the powder. The dry method 1s often used \\,hen large oblects a r e being inspected, and portable equipment is the power source.

7. CIRCULAR ~ ~ ~ A G N E T I Z A T I O N . Stationary magnetizing units, l i a v ~ n g a grill top and adjustable clamping co~i tac t heads, a re used fo r magnetizing small pai'ts. Tliese u r~ i t s a r e siinilar to those used fo r the wet method: but do not have tlie tank aiicl circulating pilmps. A pan to catch tire dry poltcler for re-use is uncler tltc grill table top. 'Fl~is pi,o\>rsion fo r catcllliig the riry po\vdcr for YE-USI! does $lot imnly that the same batcll of pow(1er shuulrl be used iii~leli~iitely. 111 time, xvith re~ieated <lusting over mngiietized surf;icc~s, the po\rder loses rnost o r ;ill of the fine p;irticlc sizcs, and t l irreforc 111sc.s sullsl-


tlvity t o fine defects. Color coatings of some of tlie powders also tend to he lost \vith prolonged use. Tlie powder should therefore be xvatclied and discarded and replaced with fresh materlal before this loss becomes serious.

For the testing of large objects such a s large castings and weldments, when portable units are used, contact fo r circular magnetization is usually made with hand-held prods, and powder applied while the current 1s on. Tlie current switch is built Into the handle of the prod. Fo r applications in which t he holding of tiie prod contacts by hand is difficult or t irlng f o r the operator, magnetic "Ieeclies" a r e available, to hold the contacts to the ~rrork magnetically. Figure 119 illustrates this device. The electrodes carrying the magnetizing current a r e held firmly to the work by s t rong permanent magnets. Both electrodes may be attached, or one leech and one hand-held prod can be used. The latter arrangement is particularly advantageous

Fig. 119-Magnetic Leech Contacts

in weld inspection. With hand held prods two inspectors are required, one to hold down the two prod contacts; and one to manipu- late the powder gun. With ieecli contacts one inspector can handle

CHAPTER I 3 THE DRY METHOD-hf.ATERl.4LS AND TECHNIQUES

one prod contact and the powder blower, thus making it a one nian operation.

Another system uses a double prod contact. The two prods a r e mounted on a common handle, spaced a t the proper distance. The operator can handle the prods with one hand and apply the powder with the other. The current control s w ~ t c h 1s in tlie handle on which the prods are mounted. This system is flexible and easy to operate, and requires less equipment than the leeches. However, holding the double prod properly in place requires more effort 011 the par t of the operator and is more tiring than the leech system.

Current values for the dry powder method of testing are tiie same a s Indicated by the ruies of thumb for other magnetic particle testing methods. These are, for clrcuiar magnetizntion, 1000 amperes per inch of diameter, and for prod contacts, 60 amperes ?nznimz~?.o,~ per inch of prod spacing. Somewliat larger current values are usually advisable, rather than the minimum. Up to some 200 amperes per inch of prod separation a r e frequently used.

8. LONGITUDINAL MAGNETIZATION. When tlie bench types of stationary units a r e used for the dry powder method, a fixed coil for longitudinal magnetizing is permanently mounted on the unit. When portable units a r e used, heavy flexible cable leads a r e usually wound into coils of three o r four turns to suit the slze andishape of

Fig. 120-Pre.tormed Split Colls ot Flexible Cable.

249

PR1XCIPI.ES OIZ MAGNETIC f'flllTICLE TESTING

the part . Fixed coils a r e seldom used with portable uni ts altliougli pre-iormed split coils a r e sometimes convenient, especially f o r tlireading t l in~ugh openings in p a l t s to forin ceiitrai eoi~ductors. Figure 120 illustrates a pre-formed sglit coil f o r th i s surpose. \Vhen alternating current o r half wave is used fo r nnagnet iz~~ig, the current is usu;llly turiied o ~ i 111 the coil, tlie ponder applied and tiic coil moved about to cause tlie particles to "dance" under tlie ell'ect of the alternating o r ]~ulsat ing field, and to nnove over the surface to form indications a t defects.

9. APPLICATION OF T I ~ E POWDER. A few rules fo r tlne application of the dry po~vrlers will, if followed, malie the process of testing easier and more effective. I t should be remeinbcred tha t Llie dry particies a r e heavier and lndiv~dually have a much g rea te r mass than the very fine particles of the \vet nietliod. If, therefore, they a r e applied to the surface of a p a r t with any appreciable velocity, the fields a t the discontinuities may not be able to s top thcm and hold them. This is especially t rue when vertical o r overliead surfaces a r e being esamined. The powder slioulrl reach the surface of pa r t s a s a thin cloud with practically zero velocity, dr i f t ing to the surface so tha t leakage fields i ~ a v e only to hold i t in place. This teclin~que has been refensred to a s the s t m e a s tha t "for gently salt ing a very good steak" For vertical and overhead surfaces, the fields must overcome the pull of gravi ty which tends to cause the particles to fall a\j3ay. Eut , since t h e dry particles have a \vide range of sizes, the lincr particles niill be lield under these conditions, uniess tile leakage fields a r e extremely weak.

On horizontal surfaces this problem is m i n ~ m ~ z e d . The usual mistalce, however, i s to apply too much powder. Since, once on the horizontai surface of a part , tlie powdel. Iras no mobility (unless A.C. or half \r2a\,e i s being used) too heavy a n application tends to obscure indications. If the p a r t can be lifted and tapped, tlnc excess powder urill fall away and indicalions be more readily visible; or; the excess poxvder can be gently blown away with a n a i r stream not s t rong enough to biow OR magnetically held particles forming a n indication.

Various devices have been used to make proper po\vrler application easy. TTVO of the most w ~ d e l y used a r e shown in Fig.. 121. The "squeeze bottle" IS l i g l ~ t and easy to use. With some practice, by a comb~nat ion of shaking a s \vith a salt siiaicer, and :I squeeze on the bottle, powder can be ejected with m ~ n i m u n i \viiocity. Practicing

CH:\ITEK 13

THE DILY 31ETBOD-3l..tTlilll:%LS .4ND TECIINIQfiES

with tlie hottle on a sheet of usIilte jia11cr \\rill ra ther qu~ck ly enable the operator to produce a n even. geiitle osera11 coverage. With the

Fig. 121-a) Squeeze-Bottle Applicator tor Dry Powder. b) Air Operated Powder Gun.

251

I'PINCIPLES OF RlAGNETlC PARTICLE TESTING

po\\rder gun or blower a better ~ o b of application can be done, especially on vertical and overhead surfaces. The powder gun throws a cloud of po\vder a t low velocity, much like a very t h ~ n paint spray. Held about a foot distant f rom the surface being inspected, a very light dusting of powder permits easy observation of the formation of indications. On horizontal surfaces the excess of powder is blo\vil aura), with a gentle a i r stream from the blower. Two push-button valx7es on the blower gun control the Aolv of powder or clean a x . Less powder is Used with the gun, which is a saving in cost, and better Inspection is accomplished. Figure 122 sho~vs the inspection

Fig. 122-Inspection ot a Weld, Usmg One Leech Contact and One Prod, with a Powder Gun.

of a structural weid in the shop. The operator is using one magnetic ieech contact and one hand held prod, \vhich he can move from point to point, operating the current on-and-off switch by means of a push-button in the handle of the prod. A t the same time, with the other hand, he operates the gun to apply powder.

As was said earlier, clean, smooth surfaces are usually best for successful dry powder testing, provided the surface is horizontal

CH.\ITER 13 THIZ DI<Y hlETH0I)-3lA'YERl:tLS A N D TECHNIQUES

tvhen tested. If the surface is rough, poxviler tends to gather arid be held mechanically by gravity in depressions or the rough surface. A stronger stream of azr than slioulrl normally be ~ s e d may be required to blow off this loose powder. Care must be talien in the inspection of such rougli iweas (as, for example, a rougli \veld bead) that weakly-hekt indications a r e riot also blown away. By watching tile area very carefully \vhile applying po~vder and \vhile L)lo\vlng off the excess, \veaIt indications can oftell be seen a s the powder shifts.

For very critical inspections, the weld bead is sometimes machined away. Indications of discontinuities w h ~ c h a r e wi'holly beiow the surface a r e more readily formed and seen by the inspector on the smooth machined surface of the weid. If the surface being tested is vertical o r even a t an angle to the horizontal, a n extremely smooth surface becomes a disadvantage, since the dry po\vder tends to slide off easily, ancl weak leakage fields may not be able to hold i t in place. Under these circumstances the rough surfaces give better results.

lo. INSPECTION. A good light and good eyesight a r e the principal requirements for observing the presence of indications on the surface of parts. Choice of the best color powder fo r contrast against the surface is an a ~ d to visibility. On the large discontinuities, powder build-up IS often very heavy, making indications stand out clearly from the surface. Fo r finer cracks the build-up is less, slnce only the smaller particles are heid by the leakage field in t h ~ s case. For esceedingiy fine cracks i t may sometimes be better to go to some form of the \vet method, which is more sensitive to very fine discontinuities.

In the case of discontinuities lylng wholly and deep below the surface, experience and skill, or carefully established, controlled practices fo r repetitive tests a r e required to secure the best results of which the method 1s capable. The depth below the surface and the size and shape of the discontinuity (see discussion in Chapter 20) determines the strength and spread of the leakage field. A really esperienced inspector will observe the surface as the powder IS allo\rred to dr i f t onto it, and can see faint but s~gnificant tenden- cies of the powder to gather. Often indications a r e seen under these conditions, but a r e no longer visible when more poxvder has been applied, the excess blown off, and the surface t i~e?z examined for indications. Standarilized techniques for careful and proper applica-

PRINCIPLES OF hlAGSETIC PARTICLE TESTING

tion of the powder can provide excelleni sensitivity where similar assemblies are repetitively tested.

In the case of steels having appreciable retentivity, indications a r e held a t the defect by the remanent field, making the inspection somewhat easier. In low carbon steels, such a s low to medium strength plate and structurai steels, the retentivity is very low. On these i t is important to make the inspection while the magnetizing current i s on and the powder i s being applied, since indications may not remain in place a f te r the current is turned off. This is particularly t rue on vertical and overhead surfaces, where gravity plays a par t in causing particles to fall away if loosely held. Harder new liigh strength steels a r e being used more and more for pipe, p r e s u r e vessels, and structural shapes. Smaller discontinuities become more important in this type of steel, since i t is used a t higher stress levels, often approaching the yield point. Although retentivity is high: great care must be taken in the inspection of these steels, especially fo r fine cracks open to t he surface.

C H ~ P T E R 14

THE W E T RIETIIOD-AlATERI.%LS AND TECHNIQUES

1. I-IISTORY. The original de Forest <levelopment of the fizagnetic Particle Testing filetlioti (1929) employcd the dry powder technique. Hoke's first observations of the magnetic particle pr~nciple back in 1915, however, \\,ere made ivith Liie maglietic particles (grindings) suspended in a liquid (grinding coolant). The actvantages and disadvantages of the two nlcthods of bringing thc particles to the leakage field a t defects were, a t that stage, neither realized or even considered.

I t was not until the middle 1930's that liquid suslwnstons were first used. .At tlje \Vrtglit Aeronautical Company, a t Patcrson, N.J., the blaclt magnetic oxide was first suspended in a light petroleum oil similar to iccrosene. And about the saine time, the General Elec- tric Co. a t Schenectady, N.Y., started usiiig fiiiely ground mill scale suspended in a similar light oil. The idea \<!as also qutcltly taken up a t \Yriglit Field, Dayton, Ohio, by 0. E. Stutsman, \rho was in charge of inspection of a i rcraf t engiiies mid parts in manufacturer's plants and in niodification centers.

Initially, troublc was experienced in properly dispersing the fine black magnetic oxide in the oil bath. This problem \\,as solved in the h,Iagnaflux Corporation laboratory by pre-dispersing the oxide in a l i g i ~ t oil by grinding i t a s paint pigment or ink is ground, \vitli a small amount of oil velucle. A suitable dispersing agent was incorporated in the misture to prevent excessive iloculation in the bath. The pre-disnersed oxtde prepared in thts fashioii resulted in a concentrate of paste consistency, and was referred to as "Paste" for the \vet method.

The paste was made into a bath of suilailie conccntratioil by diluting it with petroleum distillate. Thts \\,as rather a tedious and messy procedure, since i t involved iil.st diluting the paste with bath liquid to :a tlilii s1u1.r)~ before adding it to thc bath. Neveriheless, the pestc form of wet method particles: with oil fo r a bath, was used exclusively until 1956.

I-lowever, n fe\r7 incidents involving fires in cqutnment focuse(1 attcnlion on tilts ilazard. These wcrc duc to electrtc spa1.k~ a t con-

255


tact heads, which ignited tlie bath vapors. This led t o the development of particles tha t would be compatible ~vi t l i water a s a liquid for the particle bath. I t was some time before a really satisfactory formulation was achieved, since there mere problems in attempting to use water fo r a bath liquid. These had to do with wetting of oily surfaces, rusting, electrolysis in equipment, etc. Today, water is used extensively, both in tank-type units a n d in testing of large objects by the expendabie bath technique.

With the use of water a s a suspensoid, the paste form of particies became even more difficult to disperse, and this led to a reformuia- tion which produced the magnetic particles in the form of a dry powder concentrate. The powder incorporated the needed niaterials fo r dispersion, wetting, rust inhibiting, etc. Today, the dry con- ceiitrate has almost completely replaced the paste form for 'both oil and water suspensions. Use of water instead of oil fo r the batli is also increasing, especially fo r very large installations where many gallons of bath are required, a s in the large billet testing systems.

2. GOOD POINTS OF THE WET METHOD. AS IS true of every process, the wet mmhod has both good points as well as less favor-

i able characteristics. The more important of the good points of tile

I wet method, which constitute the reason for its extensive use, a r e t he folloxving :

i a ) I t is the most sensitive method for very fiae surface cracks. (b ) I t is tile most sensitive method for very shallozo and fine

surface cracks.

ic) I t quickly and thoroughly covers all surfaces of irregular shaped parts, large o r small, with magnetic particles.

t d ) I t is the fastest and most thorough method for testing large numbers of small parts.

le) The magnetic particles have excellent mobility in liquid suspension.

( f ) I t is easy to measure and control the concentration of particles in the bath, \\;hich makes for uniformity and accurate reproducibility of results.

ig ) I t is easy to recover and re-use tlie bath. (h ) I t is well adapted to the short, timed "shot" technique of

magnetization for the continuous metliod.

( i ) I t is readily adaptable to automatic unit operation.

C H a m n 14

THE \\'ET MFL'IIOD-&$ATERIALS A N D 'Ti%c'HNIQ~JES ! B

3. LESS FAVORABLE CHARACTERISTICS. Some of the less attractive characteristics of the \set method a r e the following.

l a ) I t is not usually capable of finding defects lying holly below tlic surface if more than a few thousandths deep.

(b) I t is "messy" to \vorK with, especially when used for the es- pendable technique, and in field testing.

( c ) When oil is used for a bath and direct contact for circular magnetization, it can present a potential fire hazard.

( d ) A fairly critically designed recirculating system is required to keep the particles in suslJension

( e ) Sometimes it presents a post-inspection cleaning problem to remove magnetic particles clinging to the surface.

4. BATH CONSTITUENTS. Tile magnetic particles used in the \stet method were discussed and described in cletail in Chapter 11. Briefly. their outstanding characteristic is their extremely small s~ze-down to one-eight11 of a micron. These very fine particles do not act a s individuals, ho\vever, but aggiomerate into groups wliich have diameters of some 5 microns or more. T\\,o colors of particles are available, red and black. In addition (and most widely used) fluorescent particies give maximum visibility (Chapter 15). The more modern dry concentrates a r e forn~uiated to include other Wath constituents, with the fine magnetic particles already bonded together in optimum sizes.

The bath liquid may be either a light petroleum distillate of specific properties, or water. Both require conditioners to maintain proper dispersion of the particles and to permit the particles freedom of action in forming indications on the surfaces of parts. These conditioners are usually incorporated in tile powders.

5. OIL AS A SUSPENDING LIQUID. Oil was a natural first choice as ;I bath liquid, since most machine pai.ts that a r e i n s~ec t ed tend to have an oily film on their surface. Gross amounts of oil o r grease should be removed of course, but any film remaining is readily \vetted an(] dissolved by the light oil of the bath.

The oil should liave very definite properties to he suitable for bath purposes. I t shouid be a well refined l i gh t petroleiim distillate of lo\\, viscosit)., odorless, with a lo\\, sulphur conteiil, with a lugh


flash point and a close-cut, fairly iirgh boiling range. The usual specifications f o r a suitable uil a r e given in Table 111.

TABLE I11 PROPERTIES OF OILS EECOMBIENDED FOR MAGRETIC PARTICLE

\VET METHOD EATH.

Of these properties, viscosity is probably the most ~mpor t an t from a functional standpoint. I t should not exceed three Centistokes a s tested a t 100n F., and must not exceed five Centistokes a t the point and temperature of use. High v~scosity slo\\~s the movement of partrcles under the influence of leakage fields. Laboratory tests have shown tha t a t v~scosities above five Centistokes the movement of magnetic particies in the bath 1s sufticiently retarded to have a deiinite effect in reducrng the build-up, and therefore the visibility, of an indication of a sniall discontinuity. Heavy oil from the surface of par ts tends to build up in the bat11 and increase its viscosity. This IS the n i a ~ n reason for pre-ciean~ng par ts to remove oil and grease.

Addition of light non-voiatile oils t o the bath a s an "automatic" rust preventive has been suggested. This must be done \\!it11 caution, if a t all, because tests have sholvn tha t even sniall additions can raise tlie bat11 viscosity above a safe unit. Tlie addition of ten percent of an S.A.E. #lo lubricating oil will raise the vlscositg of the bath froni three Centistokes to above five. If the bath already has a viscosity of four Centistokes; a s little a s two percent of $10 oil will raise i t above five.

Much lighter distillates \vould have a much lower viscosity tiian those usually used, but they ~vould have other properties unde-

Viscosity, Iiinematic a t 1003F (38' C.)

Flash Po~nt-Closed Cup

Initial Eoiling P o ~ n t

End Pomt

Color (Saybolt)

Sulphur-Low Available

CHAPTER 14 THE: U'ET METITOO-III1TERIALS AND TECHXl(1UES -

slrable m a magnetic particle bath. Lighter distillates \youid have an in~iral boiling point lo\vFr tiian that specified, and therefore a lolver flash point, nialciny- thcni a greater fire hazard. Also, elsapurn- tioii losses from the tank ~vould be greater with a lighter uii. Greatli- ing of unpleasant fumes from a light distillate leads to operator discomfort. Odors from distillate are obje~tionabie froni tlie operator 's standpoint. Odor of distillate 1s assoelated with color a11d sulphur content, ~vhicii 1s the reason for specifying that these properties be witiiin limits. The approved, li~glily refined oils a r e nearly odorless, ancl meet the color and sulphur speciiicat~ons.

6. WATER AS SUSPENSOID. Tlie great attractions of the use of water lnstead of oil for magnetic particle wet method baths a r e lo\ifer costs and the complete elimination of bath flammability. I t shouici be pointed out here that, although there have been some In- stances of oil fires in magnetic par t~c ie testing units, they have been relatively few over more than tlurty years of uslng oil for tills purpose. hi some of tlie earlier autoniatlc or senii-automatic units, the oil bath was sprayed onto the part , creating a mist of oil \irhich was more easily ignited than the liquid oil. Getter means of applying tlie bath-such a s flo\v~ng or "dumpingu-liave almost completely eliminated this source of troubie. But, as units became larger and the volume of oil needed for tlie bath became much greater, practical considerations Indicated a change froni an oil to a water bath as an o ~ e r a l l economy and safety measure.

The use of water as a suspensoid 1s quite extensive today. But the use of water is not a panacea for all the problems w h ~ c h the oil bath presents, because, although they are different from tliose of the oil bath, water has problems of its oxvn. The cost advantage of water base baths lies entirely in the cost of the oil, about 40 cents/ gallon, a s agalnst water, the cost of ~vliich is nomlnal. This 1s

because of the coniplex forniulation necessary in the way of wetting agents; dispersing agents, rust ~nhibitors, anti-foam :igents, etc., ~ i~ l i i ch run uo the cost of water-suspendable concentrates. Conse- quently, water-suspendable concentrates a r e comparable, or even slightly higher, In cost than those used for oil suspension.

Water baths must be used 111 shop areas where the temperature 1s above freezing. Use of anti-freeze liquids is not feasible because the quantities needed to be effective \roulcl raise the v~scosity of tlie bath above the max~nium allou~able. Because the use of detergents to assure the wetting of oily surfaces causes f o a m ~ n g of the bath,

259

3 Centistokes I max.)

135" F. (57' C.) Min.

390' F. 1199" C.) Blin.

500" F . (260' C.) Mas.

Plus 25.

Pass the Copper Test ASTM-D129-52.


circulation systems must be designed to avoid a i r entrapment o r other conditions t ha t produce foam. Anti-foaming agents a r e used to minimize this tendency, but a r e not 100% effective.

Since water is a conductor of electricity, units i n which i t is t o be used must be designed to isolate all high voltage circuits in such a way a s to avoid all possibility of a n operator receiving a shock, and tlie equipment must be thoroughly and positively grounded. Also, electrolysis of par ts of the units can occur if proper provision is not made t o avoid this. When water was first berng tried, some users simpiy cliaiiged the bath in their units f rom oil to water. Rapid corrosion and occasionai electric shock to the operator resulted. Consequently such conversions a r e not t o be recommended. Units desigqzed to be used with water a s a suspensoid are, howevei-, safe fo r the operator and minimize the corrosion probiem.

There is no critical requirement a s to the water tha t is used f o r the bath, a s there is in the case of oil. Ordinary tap water is suitable, and hardness is not a problem since the mineral content of the water does not interfere with the conditioning chemicals necessary to prepare the bath.

7. THE MAGNETIC PARTICLES. The second ingredient of the bath for app ly~ng the wet method is of course the magnetic particles. These have already been described. Three basic types of materials a r e available :

i a ) The oil base pastes, either black, red or fluorescent. These a r e primarily for use with oil suspensoids, though with suitabie water conditioners they can be used with water. Their use 1s rapidly being abandoned in favor of the newer dry concentrates.

(b) Dry concentrate powders, either black, red o r fluorescent. One type is used with water a s the suspensoid, and a second type for suspension in oil.

(c ) Water conditioners. These a r e used to suspend oil pastes in water, and t o make u p the quantity of conditioner in the water bath u~llen needed.

A s has been said, the dry powder concentrates are now preferred for both oil and water baths. The new self-dispersing particies a r e readily picked up by the pump circulating system, and a r e easily maintained in suspension. Some of the oldest wet method units used

C H A P l E R 14

- THE WET IETHOD-DI.XTERIAI,S AKD TECHNIQIIES

a i r for agitation of the bath. This form of agitation is not satisfactory for the dispers~on of the dry powder concentrates, or main- taining bath suspension, and has not been used in new equipment for a good many years.

8. STRENGTH OF T'IE BATH. I t should be obvious that the strength of the bath is a major factor in determining tlie quality of the indications obtained. Too heavy a concentration of particles gives a confusing background and excessive adherence of particles a t external poies, thus interfering ~ r i t l i ciean-cut indications of very fine discontinuities; so that there is danger of their being missed.

The gurde, therefore, to proper bath concentration, is the fornia- tion of satisfactory indications xvith tlie particular technique being used. For the wet continuous !/2-second-contact process used on most small stationary units, a different concentration is required from that which may be used in the tank for dipping parts for the residual method, where the exposure time of the magnetized part to the bath is much longer.

Experience has produced some bath concentration rules for the $$-second-contact system \vhich arc satisfactory for most applications. Table 11' lists these concentrations for the usual wet method particles. (Fo r fluorescent types, see Chapter 15.) But the best mcthod of assuring optimum bath concentration for any given conibination of equipment, bath application, type of par t and defects sought, is to make use of par ts with kno\vn defects. Bath strength can be adjustcd by cut-and-try methods until satisfactory indications a r e obtarned. This concentration of bath can then be adopted a s standard for those conditions.

I t IS iinportant tha t the proportion of magnetic particles in the bath, once the satisfactory concentration has been arrived at , be maintained uniform. If tlie concentration varies, the strength of the indicalion will also vary, and interpretation of the meaning of an indication may be erroneous. Fine indications may be missed entirely with a weak bath.

I t must be remembered that the important cousrderation in bath strength is the proportion of active ingredisnt - tha t is, the actual amount o i magnetic oxide per unit of volume. Since the proportion of oxide varies in the different pastes and powders, the amount of these materials required t o produce the necessary concentration of magnetic particles in the bath will vary also. To determine the

PRlSCIPLES OF BI.4GNETIC PARTICLE TICSTING

actual amount of o s ~ d e s in the bath, a measured sample of the bath is filtered, the oxicic purified from the v a r ~ o u s compounding materials, anrl the separated ositle then drred and xsselghed. This IS, however, a laboratory test ant1 not readily used by the operator a t the unit. Since thc bath strcngtli should be checlied no less often than a t the beginning of every shift, when the unit is in continual use, some quick and easy test nlcthod is requlred.

The settling test has served t h ~ s purpose very well, and is easily and quicltly performed a t the unit. I t 1s not as accurate a s is d e s ~ r - able, but is reasonably quantitative and 1s reproducibie. I t can be easily standardized svitli the material 111 use, and IS quite satisfactory a s a daily guide for the operator. I t is less satisfactory on some of the newer, more comples formulations than on the older types of bath mater~als . In these cases another form of tes t (nrhich will be described) can be used if preferred. (See Chapter 26, Sec- tion 9.)

i n the settling test. 100 ml. of well aeitated bath a r e run from - -. ~ ~~~

the bath applicator or taken from a well agitated tanli, mto a pear- shaped centrifuge tube, a s illustrated in Fig. 123. The volume of solid niater~al that settles out af ter a pre-cletcrm~ned interval (usu- allv 30 rn~nutes) is read on the graduated c u l i u d ~ ~ c a l Dart o f tl,e

Fig. 123-Making the Settling Test tor Wet Bath Concentratnon

262

CHIIPTE~ 14

TEE WET >IETHO&3IATERIALS AND TECRKIQUES

tube. Dir t in the bath will also settle out and usually shows a s a separate lager on top of the ox~de. The layer of dirt and lint IS

usually easily distinguishable, as i t is of a different color from the osidc. See Chapter 26, Section 8 for details of this test.

Table IT' shows the suggested amounts of the several concen- t ra tes to be added per gallon of bath, and the volume of solid material which settles out when the bath is made up with these amounts of concentrates. The black and red pastes, 7-X and 9-B a r e included in the Table, since they are still used t o a slight elrtent, although they a r e rapidly becom~ng obsolete.

Xagnaflux Corporation

Deslgnntion of Material

Oz. of Concen- t ra te per gallon of bath

Settling Test. * Volume Range, in mI.

Llquid to be used for the

9. ~ I A I ~ I N G UP THE BATH. When a new bath is to be made up, for a new unit or af ter dumping a dirty bath from a unit In use, i t is important first to make sure tha t the agitation system is clean, and not clogged by dried particles o r accumulated dirt such as lint or shop dust. Nest fill the tank with oil or water a s required, and operate the agitation system to make sure i t 1s functioning properly.

( a ) F o r Drv Pozoder Co?rce?ltrate. Measure out the required amount of powdered concentrate in the graduated cup and pour i t directly into the bath liquid in the tank. The agitation system shouid be runnlng and the concentrate poured in a t the pump intake, so that it will be quickly drawn into the pump and dispersed. See Figure 124. After 10 minutes of

7-C Black

Powder

bath

1%

1.3 to 1.5

Oil Water

9-C Red

Powder

'Settling time 3s 30 nrsnutcs, uslng a 100 mi centrifuge tube. "Regulrcs use of w\.atr?v conditioner WA-2A to disperse in water.

* *

1

1.7 to 1.9

Oil Water * *

27-A Black

Polvder

2%

1.1 to 1.2

Water

29-A Red

Powder

2%

1.4 to 1.5

Water

7-A Black Paste

9-B Red

Pnste

1

1.4 to 1.6

Oil

1%

1.3 to 1.6

Oil

PRINCIPLES OF AlAGNETIC PARTICLE TESTING

Fig. 124--Mixrng the Bath Using Dty Concentrate.

operation the bath strength should be checlced with a settling test. The amount of settled material shouid check approximately with the figures given in Table IV.; if the reconl- mended amounts of concentrate given in the table have been used. If the volume is less than called for add concentrate- if more, add liquid. A better plan for using the settling test a s a bath check IS to observe the actual amount of settling for a carefully made up bath of the desired strength, record it, and thenceforth using this figure fo r the standard of bath strength for subsequent tests.

The new pre-wet concent~.ates \\,ill disperse very quickly even through the large volume of bath in large units.

(b) F o , Oil Pastc a??d Oil Eath. The procedure 1s wmilar to tha t followed in the case of the dry powder concentrates, except that the paste must be \xselghed out ~ns tead of measured. I t is transferred to a mix i ig cup or ho~vl. Add balh liquid, a little a t a time, and mix, until a smookh thin slurry has been produced. This slurry is then poured into the liquid in the tank a t the point where the itgitation system will 111ck it up and disperse it. After a g l t a t i n ~ ten minutes the strength

2G/

CI~APW.R 14 THE WET hlETIiOD-AIATEKI.II.S AND TIICHNIQUES

should be checked by a settling test a s in the case of the dry powder concentrate.

lo. IIIAINTEXANCE OF THE BATH. AS the bath is used for testing it will undergo changes due to use. Some of these changes a re :

( a ) Drag-out of magnetic particles, both by mechanicai and magnetic adherence to parts, thus tending to reduce particle concentration in the bath.

( b ) Loss of liquid due to the film which adheres to the surface of parts.

I c ) Loss of liquid by evaporation, tending to increase particle concentration.

id) A gradual accumulation of shop dust, d i r t from parts not properly cleaned, lint from wiping rags, and oil from parts tha t carry a residual film of oil.

( e ) illiscellaneous scraps of foreign material w h ~ c h shop worlters sometimes toss into t he tanks!

I t is important, therefore, that the bath be frequently checked and corrections applied a s needed. Frequent settling or otherw~se suitable tests a r e desirable, if the unit is in constant use. Concen- t ra te should be added when the oxide concentration is low. Evap- oration or liquid drag-out should be watched, and volume main-

tained when the level drops appreciably. Since Ioss of liquid mag be either by drag-out o r by evaporation, and the corrective measures are different for these two types of loss, i t may be difficult to iieep the bath in balance over long periods. This is especially true of water baths, since water has a higher evaporation rate than the oil. To make up for evaporation loss, addition of oil o r water is required. Conditioning materials a r e not volatile. They \\,ill therefore build up IU the bath a s the liquid evaporates. Bath volume lost by drag- out, however, inciudes conditioner in solution, and proper correction sliould then include the right amount of conditioner.

I t is difficult to know what the cause of volume loss actually is

In any given case. However, fo r a unit in constant use it can be assumed that more than 50% of the Ioss is due to drag-out. Fo r a unit used only occas~onally, loss by evaporation i s lilcely to be tile maJor one. Actually the problem is not a serious one, because with constant use the accuniulation of dirt, scraps, lint, etc. requtres the dumping of the tank and a newr bath before loss of liquid becomes

265

PRINCIPLES O F DIAGNETIC PARTICLE TESTING

serious. But mag?~etic particle content is of most critical importance and should be carefully watched a t all tiines.

Dir t accuniulation in tlie bath can usually be easily determined, a s i t s hon~s up in the settling test fo r magnetic particles. When the settling test is run; tlie oxide, being heavy, settles out first. Dirt; lint, etc. is lighter and settles inore slo\x~ly. I t is seen as a second layer on top of the oxide. Fo r oxide determination, this layer of d i r t must be carefully excluded from the volume read. When the layer of dir t approaciies the voluine of oxide, formation of proper indications will be impeded, and the batli sliould be dumped and a new one made up. This niay occur a s often a s once a xveek when a unit is in constant use.

11. STEPS IN THE APPLICATION OF TIIE WET METHOD. Tlie steps in applying magnetic particle testing by the \vet nietliod consists of 1) surface preparation; 2) magnetization; 3) application of the magnetic particles; and 4) inspection for indications.

12. PREPARATION OF THE SURFACE. In general the same require-

ments apply for the wet method a s for the dry technique. Dirt, rust, loose scale and oil or grease sl~ould be removed. The oil bath will dissolve oil or grease, but this builds up the \,iscosity and fluorescence of the bath and shortens its useful life. With a water bath, oil on the surface of the par t makes wetting more difficult, although the conditioners in tlie batli a r e iisually sufficient to take care of a moderately oily surface. But excessive oil on par t surfaces contaminates even the water bath in time. Paint and plated coatings, if over 0.005 inch thick, should be stripped. Tests have shown tha t non-magnetic coatings of any kind, in excess of 0.005 inch in thickness, seriously interfere \vitli the formation of magnetic particie indications of small discontinuities.

13. THE CONTINUOUS WET METHOD. ~!IAGNETIZATION AND BATH APPLICATION. All the usuai methods of magnetization a r e used in the wet method-longitudinal; circular; multi-directional, induced current and over-all magnetization a r e employed, xvith either A.C. o r D.C.: a s a r e the continuous and residual methods of bath application.

The term."continuous method" implies that the magnetizing force is acting while the magnetic particles are applied. \VhiIe the current in on; maxinium flus density will be created in the par t for tlie magnetizing force being employed. I n some cases-usiinlly when

... CHAPTER 14

THE \Vl$T RIETHO~-~l.4TEIII~tLS .4XD TECHNIQUES -- -- A.C. o r half \vave D.C. IS the magnetizing current being used-the current is actually left on, sometimes for minutes a t a time. while the magnetic particles a r e applied. This is more often needed in dry method applications than in the wet.

To leare the current on for long durations of time is not practical in most instances, nor is it necessary. The heavy current required for proper magnetization can cause overheating of parts and contact burning if allowed to flow for any appreciable length of time. Over-heating of the polver equipment could also result. If such truly continuous currents a r e required, power supplies must be built niucli heavier and they become much more costly. I n practice the magnetizing current is norm all^^ on for only 1/? of a second a t a time. All that is required is that a sufficient number of magnetic particles a r e "in the zone" and free t o move while the magnetizing current Rows. The bath ingredients a r e so selected and formulated tha t the particies can and do move through the film of liquid on the surface of the par t and form strong, readabie indications in the time of one-lialf of a second. This is, of course, one reason why the viscosity of the bath and bath concentration a r e so important, since anything that tends to reduce the number of available particles o r to slo\v their movement tends to reduce the build-up of indications.

The procedure for this method of magnetization and bath application is simple and easy to perform, once the operator understands the importance of the sequence and timing of the two operations. Bath from the nozzle of the unit o r other applicator shouid be Rox\aed liberally over all surfaces on the part. The magnetizing current button should be pressed at the instant the source of bath flow is withdrawn. The timer in the circuit of D.C. units gives a q2 second "shot" of current of pre-selected strength. With other units, 1A.C.) the operator controls the time tlie current is on. There should be no interval of delay between the withdra\x~al of the flow of bat11 and the current "shot". I t is most important that the maximum film of bath be on tlie surface of the par t when the current Ro\vs. On the other hand, i t is of equai importance that the flow of bath from the nozzle or applicator bc stopped h e f o ~ e the current button is pushed, else indications rnay be washed away. I fany operators \rill give ~ W O or elZen three shots of current in quick succession to strengthen the indication some~~hat-but ?lot with any re-:ipplicalion of batli.

For small parts the usual equipment is a bench-type unit with

PRINCIPLES OF HAGNETIC PARTICLE TESTING

Fig. 12%-Simple Wet Method Unit Setup tor Circular Magnetizung

contact heads between which par t s a re clamped for circular magnetization. Under the table of the unit i s the bath tank, the control equipment, power supply and agitator pump. The bath in the tank is kept constantly agitated by pump circuiation. Bath 1s available from the pump system througll a hose and nozzle for flooding the part. Current values a r e determined basically by the 1000 ampere per inch of diameter rule, modified by exper~ence for very small or very large parts.

For longitudinal magnetization, the fixed coil, which can be seen in Fig. 126 1s used. The sequence of the magnetizing and bath application operations is the same a s for circular magnetization. The amount of current t o pass through the coil to produce the proper number of ampere-turns, is determined by the formula (See Chap-

te r 9) NI =--- 45'000 where N = the number of turns in the coil, I L/D,

the current, L the length of the par t and D its diameter.

The magnetic particles i n the bath could be either black or red, depending on wil~ch color gives the best contrast with the hackground of the surface of the particular par ts being inspected. Or, more often; fluorescent particles a re used for maximum visibity. (See Chapter 15.)

268

CHIPTF.~ 1'1

THE WET BIETHOD-MATER1AI.S AND TECHNIQUES

Fig 126-Wet Method Untt Setup tor Longltudenal Magnet~z~ng

14. THE RESIDUAL METHOD. In the residual method the parts a re first magnetized; and the bath is subseqz~ently applied, either by flowing-on or by immersion of tile par t in a tank containing the bath. Obviously, satisfactory inspection by the residual method can only be attained on parts having a relatively high retentivity. High carbon or alloy steels are usually Ilighly retentive of magnetic fields, even in the unllardrned condition. Hardened steels of this type have even higher retentivity. On the other hand the material permeability of such steels may be quite low.

With usual amounls of magnetizing current, residual fields in such steels can be established which very satisfactorily glve indications of surface cracks-even very shallow ones. Ilowever the location of defects ~vholly below the surface in such steels is not usually possible by either the continuous or the resldual methods, if the discontinuities a re more than a few thousandths below the surface. Although the res~duai method is not as ~ l d ~ i y used today a s the continuous method, i t does have some advantages wh~ch make it attractive in some clrcumstances.

269

PRINCIPLES O F MAGNETIC PARTIC1,E TESTING

The residual method is applied in two different ways: i

i ( a ) Magnetizing and appiying the bath a s a "curtain" ( no t a spray) which may be left on to flov- gently over the magnet-

/ ized par t long enough to develop good indications.

! I

(b ) Magnetizing and lmmerslng the par t in an agitated bath of magnetic particles for the length of time needed to produce

I i good indications. I 1 The "curtain flow" method is used largeiy on automatic o r special-

I purpose units. Pa r t s a r e usually mounted in jigs o r fixtures, given

I a shot of current and the curtain flow turned on. The system avoids

I the need for critical t iming of the magnetizing-bath application sequence.

I n the imiiiersion method a s usually practiced, mostly on relatively small parts, the par ts a r e magnetized rapidly, one a t a time, then placed in a t r ay and immersed in a tank containing a n agitated bath of magnetic particles. The par ts must be so disposed in the t r ay t ha t they do not touch one another, else non-relevant indica-

1 tions may be produced a t such points of contact. Haphazard loading i

I into a basket fo r immersion should not be permitted. Strength of

L bath and immersion time both have a n effect on the size of the indications produced. If the leakage fieid a t a shallow crack is weak: fo r example, prolonged iinmersion permits more particles to come into the influence of the field and makes the indication more prominent and visible.

f The residuai method is capable of close control, and of giving

i very uniform results-possibly t o a greater degree than the con-

i tinuous method a s usually practiced. The fact tha t i t is applicabie only t o par ts having relatively high retentivity 1s probably the reason, a s much as any other, tha t the method is not used more extensively.

I 15. CLEANING AFTER TESTING. After testing, a film of bath re-

i mains on the surface of parts, which may be objectionable. The bath liquid dries rather rapidly, whether i t is oil or water. I n the case

i I of oil, a film of oxide particles remains af ter drying, which may be

I difficult t o reniove. If removal is important the par ts should be

! cleaned a t once af ter inspecting. They shouid first be demagnetized if they contain rentanent field, then washed in solvent spray or by some other means. Where water is the bath liquid, a water spray

CHApTfR 1 4

THE WET \IETHOD-MATCRIALS AND TECHNIQIJES

af ter demagnetization does a good cleaning job, especially if the bath has not been allowed to d ry on the parts.

If par ts have been allo~ved to d ry before cleaning, removal of the bath residue may be more difficult. Mechanical seruhbing or detergent mashing may be necessary. Sometimes vapor de-greaslng may remove the residue, and in some cases the use of ultrasonic cleaning has been successfui.

Cleaning is in many cases not necessary, especially on unfinished parts which nfill be subsequently machined or otherxvisc processed. If fluorescent particies have been used, the residue left on the surface is very slight since particle concentration i s low in fluorescent particle baths. On the finished parts, however, when inspection is the final process, cleaning is very often, if not ai\va)rs, required. On highiy polished surfaces, resldual po~vder f rom the bath can be unsightly, and may in fact, contribute to early rusting.

16. RUST PREVENTION. After testing by the wet method using oil as the bath liquid, the surfaces of par ts are left vulnerable t o rusting. The bath oil is by specification free of any residual non- volatile material, and when i t dries i t leaves no protective film. Some sort of rust preventive should be applied soon after testing. A light coating of n t~nera l seal oil is often used a s a temporary preventive-or some other form of more permanent anti-rust compound can be applied. Addition of a small amount of mineral seal oil to the bath itself has been used by some operators. For the reasons discussed in Section 5 of this chapter, this procedure is ?tot recommended.

When water is the vehicle of the bath, the drled film on the surface of par ts consists of the various conditioners that have been used in the bath forntulation, in addition to residual magnetic particles. One of the conditioners is a rust inhibitor, so that some rust protection IS afforded by this inhibitor af ter testing. IIorvever, this is by no means permanent, and a rust preventive should be applied unless the par ts a r e soon to be processed further.

17. SKIN PROTECTION. Continuous exposure of the hands and a rms of the operator to the bath of the wet method may cause the skin to dry due to the removal of the natural oils of the skin by the bath liquid. This can result in drying and c rack~ng of the skin, \rrhich in turn may open the way to secondary infection. In most cases the liberal use of some good protective hand cream-prefer-

271


ably one that contains lanolin or s i l i c o n e s ~ v i l l mlnlniize such effects. Operators should avoid oil-soaked sleeves wiirch are constantly in contact with the skin, a s tbls intensifies the effect.

Cleanliness is, ho\rel,er, a nzz~sl if secondary nifection IS to be avo~ded. Tllorough waslling of the hands and a rms follomed by a tliorouglily rubbed-in application of hand cream mill usually provide protection for sonie iiours. After work a thorougii washing with soap and warm water tirill remove the hand compound and dirt, and a fresh application of crearii I prevent fur ther drylng and chapping.

The practice of adding liquid skrn protector to the bath IS not a good solution to the problem. In tlie first place such additions a r e Ineffective ni avoiding skin drying, and in the second place, additions of any ktnd to the bath may raise i ts viscosity and affect the proper Inspection of parts. Synthetic rubber gloves a r e soinetinies used, but these are clumsy and sloxv up the inspection. Synthetic rubber aprons are, ho~\~ever; recommended to keep the bath liquid from saturating clotl~ing. Ordinary rubber gloves or aprons a r e not suitable for use when the bath IS oil, because oil softens and destroys rubber. They a r e satisfactory, however, \vlien uzater IS the bath liqu~d.

In some cases rndiv~duals have a true allergy to\rsard petroleum oils or other ingredients of the bath, especially very light-complex~oned persons. If the usual preventive measures just described a r e ineffective to prevent a rash, there is usually no remedy but to transfer the person involved to other \eorli.

18. PREPARED BATH. In this country wet method magnetic particle baths a r e ux aimost all cases made up a s needed by adding the concentrate to the bath l iqu~d. In some operations abroad, i t is the usual practlce for the supplier to prepare the bath ready to use, and ship ~t in cans or drums t o the user. Fo r large users thrs In- creases sliipping costs, although the custom does have some attraction for small users.

19. PRESSURIZED' CANS. Prepared bath is widely sold in t h ~ s country in pressurized cans for spraying. Such cans. usually containing 011-based baths, a r e very convenient to use for spot-checli~ng or small area tests in the field. They a r e often furnished in kits, including a permanent magnet or electro-magnet yolte. whlch makes

CHnPTEn 11

THIS WET 1IETHOU-MATERIALS AND TECHNIQUES

an extremely portable package for small field testing jobs, or for nianitenance testing around the sliop.

20. LACQUER METHOD. An interesting variant of the usual wet method is a bath in which the magnetic particies a r e suspended in a thin quick-dry~ng lacquer, either clear or with a small amount of white pigment. Such a lacquer bath is sprayed o r painted onto the surface of a magnetized part. A t leakage fields, magnetic particies are drawn to form indications while the lacquer i s still liquid. When the lacquer dries, the indication IS fixed in place. If a strippable lacquer IS used, the film containing the indication can be peeled off. and transferred to a report or otlier record. Or i t can, if transparent, be projected or reproduced by photographic p r~nt ing . If the method is used, care niust be taken that tlie v~scosity of the lacquer bath does not exceed the maxlmum allo\ved for oil baths. If i t does, ioss of sensitivity for the inspection results. The method is not very practical fo r general use, but i t is one technique used to advantage if a removable transfer of a specific indication i s desired.

FLUORESCENT MAGSETIC PARTICLES--THEIR NATURE AND USE

1. HISTORY. Fluorescent magnetic particles were first offered t o industry In the summer of 19-12. The method of coating magnetic particles with fluorescent dye was dev~sed by R. C. and J. L. Switzer, \vho were ~ s sued a patent covering the idea late 111 1911. 1,Iagnaflux Corporation secured an esciusive license under this patent, and af ter an extensive program in cooperation with the Switzer Broth- e r s to refine the formulation of the new particles and prove their practical application a s a reai advance in the a r t of magnetic particie testing, marketed the fluorescent particles under the trade n a n ~ e "I!Iagnaglo"*.

When exposed to near-ultraviolet light-or "black light"-the magnetic particles so treated glow with thew o1r.n light, having a highly visible yellow-green color. Indications produced are easily seen, and the fluorescent particles give much stronger ~ndications of very small discontinuities than do the ordinary non-fluorescent magnetic particles. This fact caused the nel\r n1ater1al to be put t o use a t once, espec~ally in critical tests where very small defects were sought. Tests were faster and more reliable than ~ v i t h the older non-fluorescent particles.

Slnce the close of World II'ar 11. and espec~ally since the early 195O's, the use of fluorescent particles has grown a t a steadily increasing rate, and they are most trridely used today.

2. PRINCIPLE OF FLUORESCENCE. Many substances have the property of fluorescence. Such mater~a ls give off light \\,hen ener- g ~ z e d or exposed to light of a wave length shorter than that of the v~ole t of the v~sible spectrum. This "black" or invisible light energy is absorbed hy tile fluorescent material and re-emitted a s light of a longer wave 1cngth in the v~sibie range. Different substances fluoresce in a variety of different colors, from blue through green, yellow and red. Many minerals fluoresce brilliantly under black or

*Magn:,plo. Tr;!rlcmnrk registerecl ,n U.S. Pntcnt Oflice. Pioocr-ty oT hlagna- flux Col.et,rntion. Also i-eg~stcreil ,n Canndn, Grcnt Bi.ltnin, Italy, nrirl hleslco.

ultra~fiolet light, and prospecting for such o r e s - - ~ , lor example, ores of tungsten and vanadium-1s often done a t nlght w ~ t h a portable b1;icli or ultra~.iolet light.

Specifically, the dyes used to t reat the niagne!ic particles to make them fluoresce, glow \sit11 a greenish-yello\\' color when exposed t o black light. The b1;iclc light used for fluorescent magnetic particle testing is radiant energy Iiaving a wave length of 3650 angstrom units. True ultraviolet, of 3000 angstroms and below, also activates many substances to make them fluoresce. I I o ~ r e v e r ~ ultraviolet of this shorter wave length 1s i iarn~fui , causing burns and d a i n a g l ~ ~ g the eyes. The biack light of 3650 angstroms, on the other hand, is entirely, harmless, and is uscd in thousan~ls of iocations \ \ s l t ho~ t any ill effect \\$hatever on the inspector.

Except for a felt' necessary modifications, fluorescent magnetic particles a r e used In wet suspensions a s 111 the ordinary v e t method.

3. ADVANTAGES OF THE FLUORESCENT PARTICLE METHOD. Flu- orescent particies have one tremendoi~s advantage over the untreated o r "visible" particles. This is thew ahility to g ~ v e off a brilliant glow under Mack light. This brilliant glow serves three principal purposes :

(1) In semi- or complete darkness even very m ~ n u t e amounts of the fluorescent o x ~ d e a r e easily seen, having the cffect .of increasing the apparent sensitivity of the process trenien- dously, even though, niagnetically, the fluorescent particles a r e not superior to the untreated oxldes.

(2) Even on discont~nuities large enough to give good visible indications, fluorescent indications are so much more easily seen that the chance of the inspector missing an indication is greatly reduced even when tile speed of scannlng parts is ~ncreascd. To state i t simply, the fluorescence of indications takes "the looking out of seelng"

(3 ) Inside drilled holes or cavities, or in sharp corners such a s threads or key ways, the fluorescent 1ndic;ltions a r e clearly anti readily seen, x\~I~ile v~sible color ludications can be easily obscured.

The fluorescent particle method then, is faster, more reliable and more sensitive to very fine defects than the visible colored ~ a r t ~ c l e method ~n most applications. indications are harder to overlook,

i PRINCIPLES OF MAGNETIC PAIITICLE TESTING

especially in high volume testing. In addition, the fluorescent method has all the other advantages possessed by tlie liquid suspension technique.

4. DISADVANTAGES OF FLUORESCENT PARTICLES. Fluorescent magnetic particles used in suspension In liquids have the same unfavorable cliaracteristics which go with the usual wet method techniques; and there is the additional requirement for a source of black light, and an inspection area from which a t Least most of the white light can be excluded. Experience lias shown, ho\rrever, that these added special requirements a r e more than justified by the gains in reliability and sensitivitjr.

5. MATERIALS. There is no difference between tlie fluorescent and non-fluorescent materials so f a r a s the liquids and bath requirements a r e concerned. Oils must nieet the same specifications a s listed in Table 111; with one additional requirement. The oil itself must not fluoresce strongly, nor with any color other than the usual bluish- white of most petroleum products. Most of the commercial distillates approvecl for the regular wet method a r e also satisfactorj, for use with the fluorescent particles. The reason for limiting the fluorescence of the bath itself is obvious, since fluorescence of the film of oil on a par t would produce a confusing all-over background.

The particles for t lus method are magnetically the same a s the visible type, but they niust carry the fluorescent dye and the binding material that holds tlie dye and particle together a s a unit. This loading of the particles would tend to malce them less effect~ve in producing indications were it not for the fact that to be easily visible, a fluorescent particle Indication requires only a small faction of the particles needed for the non-fluorescent type. Tlius the overall effect is a large vructicul and effective increase in sensitivity.

The fluorescent particles are no\ir supplied primarily a s a clry concentrate incorporating all the ingredients necessary for dispersion, and-ln the case of concentrates fo r suspension in water-for water conditioning. Probably a iarger proportion of all fluorescent particles are used in water suspensron than is the case with the non- fluorescenl type. This is ~ a r t l y because the fluorescent particles a r e favored for large installatio~is, such a s for the inspection of steel billets i'or high volume testing, and for the ~nspection of large Ob~ects hy Lhe esperidable nlethod, in the founrlry and shop, and In

FLLIORESCENT SIAGKETIC PARTICLEGTHEIII KATURE AKD USE

the field. Tliese operations inherently consume large quantities of fluorescent particles.

I t is, of course, of great importance that the bond betvSeen the fluorescent dye and the magnetic particle be able to resist the vlg- orous agitation it receives In the pump circulalion. If dye and magnetic particle can become separated, the dye tends to clirlg to the surfaces of the part separately, independent of any magnetic attraction, causlng a meaningless and confusing bacliground. A t the same time the magnetic particles that are held magnetically a t indications have lost some or all of their fluorescing ability, which results in a net loss in sensitivity for any size of discontinuity.

6. STRENGTH OF BATH. The quantity of fluorescent particles used to make up a suitable bath is very much smaller than for the non-fluorescent type. I t is actually less than 1/6tli a s mucli. This is because fewer particles are required to form readable indications. Concentrations of the order of those required for the visible tyue particles >\rould result in an excessive background fluorescence from tlie particles in tlie surface film of bath. This ~vould make fine indications difficult to see.

Table V sholvs the recommended amount of particle concentrate to be added for tlie several types of powders and pastes for oil and water suspensions. See the discussion of bath strength in Cliapter 14, Section 8. The same requirements dictate the proper strength of fluorescent particie baths.

TABLE V.

BATH STRENGTH CHART-FLUORESCENT PI\RTICLES

~\lagaafiux ~ o ~ . p o l a t ~ o n / j Paste I D I S I I I I ~ ~ O I ~ o i Concec~trate. I lo-* I 14-A I 11.. I 20.1 I 4 ' I 10

02. of Concentrate 11 0 . 2 1 / 6 1 l / 6 1 l Q 1 2 . 0 per Gallon of Bath. 1 1 0.2

Settling Test' Volume Range in ml.

Liquid to be used far bath.

iScttling time !s 30 manutcs, usmg 100 ml ccntrifugc tuhe except as nnt,.d R ~ l r n . . . - . - - - -. - . . .

""Settling timc Is 5 mlnutes. ""Settling time is 45 nlanutes, using 1 Iiler Itnliaf sedimentation cone.

277

PRINCIPLISS O F MAGNETIC PARTICLE TESTIXG

Tile xzoiume of particles tllat settles out in the test for bath stren@l~ is larger for the Nuorescen!. particles in proportion to the qllantity of concentrate used, a s compared to the non-fluorescent particles (Table 1 V ) . T h ~ s is b~,caose the fluorescent dye and binder reduce tiie density of the particles so tllat they do not settle so mpidly, nor so compactly, in the standard settling time of 30 nlinutes. The amount of settled-out material is not unifornliy pro- portionate to the amotint of concentrate added, for tlie double reason tha t comi~osition of the several concentrates varies to include iieces- sa ry conditioning materials, and that the specific gravity of the several suspensoids aiso varies, partly a s the result of these additives. These factors aflect the rate of settling and the degree of compactness of the settled matei.ial in the time allowed for tiie settling test. The $15 polvder is a special large particle powder, which settles more rapidly ( 5 min.: the X24 1s a specla1 po\rrder \\~hich settles very rapidly and compactly. To get sufictently accurate readings a 1000 ml sample is used, in an imhof sedimentation cone, with a 45 minute settling time.

7. R I I X I N G THE BATH. Except f o r the difference in the quantity of coilcentrate used, the details of h;ith mixing a r e the same a s for the non-fluorescent particles. (Section 9, Chapter 14.) Old type oil pastes must be first reduced to a thin slurry by dilution with bath liquid before adding to the bath ; and the dry concentrates are measured out by volume, in a calibrated measure, and added directly to the taiilt a t the point where the pump suction line \\rill pick them up a t once.

8. nXAINTENANCE OF THE BATH. Here: again; the rules are identical with those dcscl.ibed in Sectiort 10, of the urevious chapter, for non-fluorescent p;irticles. There are three additionai sources of de1erior;ition in :I bath of fluorescent particles that must be watclled for, and that require discarding of the bath when the condition becomes severe. The first is the separation of the fluoresccnt pigment from the magnetic particles, \!*i~icl~ occiii.s in some of the older types of particles. Such separ;~tion causes a falling off of fluorescent brightiless of indic;itions, and an increase in the ol~erall fluorescence of the bacl<groiind. When this has occurred to a noticeabie degree, the bath sirould be dumped and a new oiic pi'eparetl. This condit~on cani~ot readily be dctccted in the settling test, but nlust be observed by the operalor in the way the bath i>erfurms.

A second source of de t c i~ io~~ ;~ t i o i~ of the i)ath of fluorescent par-

CHAPTEE 15

1~LLiOI:ESCEXT 31AGZETIC P.4R'PICLES-THEIR NATKTRE AXll USE

ticles \vhicll does not act in the case of the non-fluorescent type, is tlie accumulation of n~uyricfic dust o r dirt in the bath. In foundries and metai tvorking shops there is a cunsitlerabie amount of finely divided magnetic material 111 the dust carried by the air, and this material \\.ill accumulate in the bat11 along \vith other <lust and dirt. In a bath of non-fluorescent particles tliis does no special harm, until the accumulation of total dirt is excessive. In tile case of fluorescent particles, however, i t tends to decrease the brightness of the mdica- tion. This is because the fine magnetic material is attracted to indications along \\zit11 the fluorescent particles, and it talces very little of such non-fluorescent material to reduce significantly the fluorescent light emitted by the indication.

-4 third source of deterioration of the fluorescent particie bat11 is the accumuintion of fluorescent oils and greases from the surfaces of tested parts. This accumulation, in time, builds up the fluorescence of theliqutd of the bath to a point a t ml~ich it interferes with the viewing of fluorescent particle inrlications.

9. STEPS IN THE TESTING PROCESS. .Spplication of the test using fluorescent magnetic particles is identical in every way with the procedure described in detail in Sections 12 through 16 of the preceding chapter, except, of course, fo r the exam~nation for indications. One precaution in the preparation of the surface of parts before testing must be given special attention. This is the renloval of siirface oil and grease. Most petroieum distillates, iubricaling oils, and greases fluoresce with various degrees of brightness. Such materials must be kept out of the testing bath because of tlie increase in backgroiind fluorescence which they produce.

lo. EXAMINATION FOR INDICATIONS. After parts have been processed with fluorescent magnetic particles and a r e ready for cs- arn~nation for indications, the procedure is entirely different. The inspection for indications must be conducted under black light in an area from \viiich a t ieast most of the \rhite light has been excludeti. Under proper circumstances, with adequate black light intensity, the location of discontinuities is easy and rapid, suice the indications glow brightly xvith their own ligiit. But there are certain requirements that must be met for prooer inspection, and the test engineer and operator must know about these ;inil tiii~ierstand them.

I . THE INSPECTION AREA. Complete darkness is of course the optimunl for masimunn \,isibility aild contrast? slncc even minute points of light emissio~i ;!re readily seen in complete iiaricness. To

PRlNCIPI.ES OF RIAGXETIC PARTIC1,E TESTING

secure complete darkness means a light-proof room; but in most practical applications it is not necessary to go this f a r in tlie exclusion of white light. As a matter of fact, absolute dafkness is never achleved in any case, since the black light lamps give off some visible violet light. This sinall amount of vioiet light is actually not a disadvantage, a s i t malces i t possible to see tlie par t being handled. Indications, which glow with a br ight yellow-green, a r c in good contrast with the violet, and a r e readily seen.

Fig. 127-Simple Inspection Uni t tor Fluorescent Magnetic Particle Testing. ,

12. CURTAINED INSPECTION BOOTI.IS. Neariy all fluorescent magnetic particle inspection i s carrled out in a curtained enclosure around the magnetizing unit, with suitable black lights mounted inside, usually inciuding a t least one hand-held lamp. Tlie curtain black-out is compiete a t the ends and the back of the unit. Toward the

CI~APTER 15

V1,UORESTENT MAGNETIC PARTICLES-THEIR N.&TI!RE AND USE

f ront the cui"rains extend out a t the sides, overhead and across the front, allowing space for the operator to stand in the darkened area bcfore the unit. Usually these curtains do not extend to the floor or even much beiolv the height of the woriiing top of the unit. They do admit some white light, but access to the booth is easier and the ventilation inside is improved by this arrangment. I t normally does not admit enough light to interfere seriously with good inspection under the black light on the working top of the unit.

Painting the Aoor under the booth a dark color helps reduce the light admitted. Bright shop lights in the area of the unit are not desirable.

Since the biack light generates considerable heat, small ventiiat- ing fans in the booth a r e often necessary. Usually a white light is also provided to facilitate cleaning of the booth and making up baths, etc., but this light shouid not be turned on dunng inspection of parts. Even brief exposure to bright white light destroys the dark adaption of the inspector's eyes. See Section 21, this chapter.

13. SMALL DARK CABINETS. For occasional iiispection of small parts \vith fluorescent magnetic particles where complete facilities are perhaps not warranted, small table top viewing cabinets a r e sometimes used. These coiisist of a box-like enclosure in which a black light is mounted. Pa r t s map be held under the black light througii an openlng a t the table-top level, and are viewed by the inspector through a slot in the upper part of the cabinet. See Figure 128.

Tlie device is not very satisfactory for high sensitivity inspection, because full dark-adaption of the inspector's eyes cannot be attained, and very fine indications may no1 become visibie. How- ever, par ts can be held quite close to the black light bulb, and a high intensity of black light can be secured-much higher than neetled under standard conditions. This compensates to a considerable degree for the lack of full dark adaptation of the inspector's eyes.

When used, the cabinet should be located in a dimly lit area, and the inspector sliouirl put 111s eyes to the viewing slot and wait a short time before attempting to make decisions a s to the plcsence or absence of indications. Gross ~ndications, of course, are easily seen without any dark atlaption of the eyes.

14. INSPECTING IN TI IE OPEN. In some instances it is necessary to inspect objects or parts of a n assenlbly in the fieid. A super-brlght

PRINCIPLES OF MAGSETIC PARTICLE TESTING

Fig. 128-Table.Top inspection Cabinet, with Built-ln Black Light.

fluorescent particle is available which can be used more o r less i n t h e open, wi th only a malie-shift a r rangement t o exclude whi te light. With such a n ar rangement , and a I i ~ g h intensity of black l ight radiation, satisfactory tests f o r all bu t very fine discontinuities can be carried out.

Field liits a r e available f o r th is type of work, uslng portable magnetizing units such a s yolies, and uressure s p m y cans containing prepared !fluorescent particle bath. A high intensity black l ight is also inc~ubed.

15. INSPECTING LARGE PARTS. I.'requently ~nspect ion f o r cracks must be carried ou t on very large o b ~ e c t s or structures. I n such cases t h e lnspection is usually made in place, and only portable tes t ing equipment can be applied. Examples a r e large forgings a n d cast- i ngs ; t h e magnetic blades of s team turbines assernbled on t h e spindle; welds in prcssure vessels, brldges, buildings, etc.

282

Since small craclis a r e iisiially \-cry ~rnport:int in s t ic l~ inspect~ons, high intensity blaci; l ights should be used, and some degree of white l ight esciusion obt:iined. Often a makc-shift f rame, covered witli canvas o r blaclc plastic sheet ing serves excellently f o r the piispose.

16. T H E BLACK LIGHT. Blacli iiglits a r e a criticoi pa r t of tllc equipment used in Nuorcscent inagnetrc i>articlc testing. In tlif foilo\\.ing chapter (16 ) t h e conslructioii and operation of the black l ight \\rill be discussed in detail. A few pr:lcticai points should be nlcntioned liere, ho~\.ever, s ~ n c e ail understanding of the requirements at t l ~ c ~vl7ct 01 Z~ISI)~CIIO?I is iinpoi.tani if the oue1.aio1. is to secure good results.

The m t e n s i t y of black l ight o t tlic z~ocnt ot i?is~>cct.ron IS of the utmost importance. T h e most conlmo~i black l ights currently 111 use a r e 100 a n d 400 wat t quar tz mercury a r c lights. They a r e sealed illto a houslng ~ \ ~ h i c R uncludes reflectors, and which IS covered \\,ilh a special filter glass. The filter t ~ a n s m i t s black light, but csciudes nearly all visible light a n d is opaque to the shor t wave lcngth oltra- violet light. The 100 w a t t l ights a r e ful.nlshed in two t y i x s of 1.r-

Flg. 129-Effect of Intensity of Black Light on the Braghtncss 01 lndicatlons Same lndicat~on under Weak (30 Foot-Candles) and Intense

(120 Foot-Candles) Black Light.

fiectors, elthe!. "spot" 01. "flood", but only tile "spot" 1)~pe is rcc- ommended f o r insuection work. T h ~ s light concentrates most of its output ~n a SIX lnch d~an ie to - circle, and gives 21dcquate intensity

PRINCIPLES OF BIAGNETIC PARTICLE TESTING

over this area. I t therefore insures a high level of black light intensity a t the point of inspection. The 100 wat t flood light spreads i t s output so much that satisfactory intensity a t any point is not obtainable. The 400 wat t lamps, on the other hand, give satisfactory light intensity over a 16 inch diameter circle. Batteries of these lights a r e used in installations where a large area must be illuminated, a s in the inspection of large steel billets. (See Frontispiece.) The areas of satisfactory black light intensity cited above obtain when the filter glass of the lamp i s 1 5 inches from the worlc. Much higher intensities result when tlie lamp is held closer than 15 inches.

17. INTENSITY REQUIRED. The Intensity of the black light radiation used to energize a fluorescent material determines the amount of visible light which the materiai emits. Within limits, doubling the intensity of the Mack light a t an indication will double the brightness of the indication. Therefore black light availabie at the indicution must not fall below a safe minimum. Usually 90 to 100 foot- candles is adequate, but for critical work, i t should be greater than this (see Chapter 16, Section 15) .

This intensity (90 to 100 foot-candles) is obtainable with the 100 wat t spot light over a six Inch diameter circle when the lamp is held 15 inches from the surface of the part. When a hand-held portable light is being used, much greater intensities a r e obta~nable by holding the light closer to the work-or, a small par t can be held closer to the lamp. The operator should remember tha t very little black light-20 to 30 foot-candles-is sufficient to light up gross indications, but without sufficient intensity, a frne zndication 7eay go entirely unseen.

18. OPERATING CHARACTERISTICS OF BLACK LIGHTS. Mercury a r c black lights should be turned on five to ten minutes before starting to use them for inspection. This i s hecause the full output of ultraviolet light i s not developed until the mercury a r c gets hot, which requires a t least five minutes. A certain minimum voltage is required to maintain the arc-about 90 volts-so tha t a sudden drop in line voltage below this point will cause the light to go out. If this happens, the a r c will re-establish itself, but not until the whole lamp has cooled down considerably. This requires usually f rom five to ten minutes. Lights should therefore not be operated on shop lines subject t o severe voltage variations unless constant- voltage transformers are used to avoid such occurrences.

CHAPTER I5 FLUORESCENT hlAGNE1'IC I'AWICI,ES--THEIR KATLIRE AND LISE

The life of the mercury a r c capsule is shortened by frequent starting and stopping. I t is, therefore, usually more economicai to s tar t the light a t the beginning of inspection and not turn it off again until its use is ended for the day, even though there may be intervals when i t is not actually in use. The output of a black light falls off slowly over the life pe r~od of the mercury arc capsu'le. This life period may be a s much a s 1000 hours for the 100 watt lamps, and up to 4000 hours for the 400 wat t type, but varies greatly from lamp to lamp. As the lamp gets older, therefore, i t is most important to check its output a t rather frequent intervals, to make sure tha t the output has not fallen helow the minimum permissible intensity. See Chapter 16, Section 15 for the method of measuring black light intensity.

19. BLACK LIGHT FILTERS. The black light filters should be kept clean. Dust and dir t collecting on the filter glass will reduce the black light output significantly. Care shouid be taken against breakage of the filter. The glass gets very hot from the heat of the mercury a r c and even though the giass has been heat-treated t o minimize the effect of thermal shock, a splash of moisture o r contact with a cold object con crack it. A cracked filter should not be used. This i s because, even though the visible light tha t gets t1iroug.h a ciack may not be enough to interfere seriously with the fluorescence of indications, the unfiltered white light which may get through such a crack may affect the dark-adaption of the eyes of the inspector.

20. THE INSPECTOR. AS in any job of inspection, successful and reliable defect location depends on the inspector remaining alert even though there is a great tendency to get bored with tlie job when very few indications show up. If an inspector apprec~ates the importance of the work he is doing-in some applications lives will depend on his effectiveness-he is not likely to become perfunctory in his examination of parts. I l e must have good eyesight, of course, and if he wears glasses, they must focus on the surface of the par t he is examining. This may not aiways be the case with the two focal distances of bi-focal lenses. A special pair of glasses with an in-between focal distance may be necessary.

21. DARK ADAPTATION. The human eye is a peculiar and complex mechanism which has the power to change its ability to perceive objects and differences in light and colors, dcpeiiding on tlie

PRINCIPLES O F DlACKETIC I'ARTICLF; TESTING

general level of illum~nation that exists wlieii the observation is being made. In strong white light the salsit ivity of the eye to small differences in light intensity IS not particularly great, though its ability to differentiate colors and degrees of contrilst is a t . a maximum. When the illumination falls t!i a low level, a n entirely different perceptive mechanism euines into use. In the dim light, the ability to distinguish differences ln color and contrasts is poor. but the ability to see dimly-lighted ohjects and small light sources is tremendousljr increased. Figure 130 is a n attempt to show these two ranges of perceptive ability of the eye in graphic form. No scale is consenlently applicable in such a representation slnce the t\\.o areas of vision a r e entireiy different and operate by entireiy different mechan~sn~s. Erer~-one has experiencd the sensation of "not being able to see a thing" when passing from a brightly lighted room to a dark one, but finding af ter a short time tha t objects in the dark room become quite readily visible. Thls increased ability

Fig. 130-Chart of Perception of the Human Eye.

CHAPTER 15 FLUORESCENT filAGNETfC PARTICI,ES-THEIR NATURE AND U S E

This is a most important phenomenon in connection \sit11 flun- rescent magnetic particle testing. Fluorescent indications a r e viewed in darkness o r dim light where the more sensitive eye mechanism is in action. The dark adaptation of the eyes necessary to the seeing of Huorescent indications requlres from five to ten minutes tn the dim light before i t IS xve11 attained. Once the eyes are dark adapted, minute light sources, too small to be seen in a bright light envlron- ment, appear relatirely brilliant and easily seen. The ability to perceive small light sources such a s fluorescent indications is increased by the fact that the eye is dralsn to any source of light in a dark background.

The color of the Hriorescent light from the fluorescent magnetic particles is a yellow-green, and this color is a peculiarly appropriate one because of another characteristic of vision with the human eye. The eye is not equally sensitive to all colors of the spectrum, but varies quite widely, reachlng a maximum in the yellow-

WAVELENGTH IN I I M G S I R O W

to see in dim light is called "dark adaptation". When returning to the bright room, one is a t first dazzied until the eyes readjust to t h ~ s high light level.

Fig. 131-Chart of Color Response of the Human Eye.

Figure 131 shows the curve of the eye response to the varrous colors. Use of the yellouf-green color fo r the Huorescence of magnetic par-

PRINCIPLES OF RlACNETIC PARTICLE TESTING

ticies adds one more plus factor to the sensitivity of this method. ye ] l o~~-g reen light has the added advantage of being "out of con- test" a s to color; compared to the white or blue-white fluorescence of oils and distillates a s encountered in magnetic particle testing. Thus the perceptibilily of fluorescent indications is, by their nature and by the environment in which they a r e viewed, of a dis- tinctly higher order than those of the non-fluorescent type.

22. I-IEAI'TH I-IAZARDS OF I\IERcURY VAPOR ARCS. Black light is in no way Injurious to the operator. The wave length range, pealc~ng a t 3650 i , is below the shortest visible violet light in the spectrum, but is \\re11 above the wave length of ultraviolet. light which causes sunburn and other injurious effects-3000 1 or iess. The mercury a r c inside the lamp produces large amounts of this short-wave ultraviolet, but i t is compieteiy removed by the envelope and the black light filter combined. The clear glass envelope surrounding the mercury a r c capsule is not necessarily entirely impervious to the short wave ultraviolet, and some of i t may get through. The lamp should therefore never be turned on without the filter in place, and cracked filter glasses or bulbs should be replaced immediately.

23. AVOIDANCE OF OPERATOR DISCOMFORT. When working with black light the operator will quickly notice that many objects and materials have the property of fluorescence. The teeth and finger- nails fluoresce with a bluish-white light. Dyes in neckties or other clothing may become startlingly bright, usually not with the same color they have in white light. Men's white shlrts usually fluoresce with a quite bright blue-white coior, due to daylight fluorescent dyes incorporated in today's detergents to provide the "whiter than white" fiii~shes.

The human eyeball also fluoresces, and when black light is allowed to 'each the eyes directfy, an unpieasant effect is experienced when th i s fluorescence is seen, a s i t were; from the inside. This effect, though it rs unpleasant, is quite harmless. But black light should not be allowed to reach the inspector's eyes In sufficient amount to create this effect; because he cannot see normally while i t is on.

24. EYE FATIGUE. Althougli woriilng with black light creates no direct health hazard, any operation requiring close attention of the eyes creates Fatigue, and if this becomes acute, i t can interfere with epective inspection. "Breaks" sliould be permitted a t reasonable intervals. If eye fatigue is complained of chronically,

288

CHAPTER 15 FLUORESCENT IIIAGNETIC PARTICLES--THEIR NBTURE AND USE

yellow-green tinted glasses a r e excellent a s a means of reducing this effect. These glasses must be of the r ight composition to pass all the yellow-green light from the fluorescent indications, and to cut out the biack light and most of the visible violet that passes the biack light filter.

25. POST-INSPECTION MEASURES. After inspection the need to clean par ts of residual particles and drled bath ingredients is less than for the non-fluorescent types of particles since bath concentration is so much iess. If required, however, the same means a r e used for the purpose-that is, scrubbing with a detergent or with solvent, washing In a vapor degreaser, or using ultrasonic vibration while parts are Immersed in a bath of water or oil. Rust prevention measures must also be taken just a s in the non-fluorescent technique.

26. SKIN PROTECTION. The use of fluorescent magnetic particles in liquid suspension does not increase skin irritation o r similar hazards above the level existing in the non-fluorescent technique. Extensive medical laboratory skin-patch tests made on the fluorescent dyes have not shown tha t they have any tendency to affect the skin. But since the materials of the baths used in this testing method a r e solvents, these solvents do tend to dry the skin, and cause i t to chap or crack. Strict cleanliness should be observed around the testing unit to avoid the danger of secondary mfections. The use of hand creams t o help replace the oils which the solvents have extracted goes a long way toward preventing cracking of the skin. The operator should a v o ~ d allowing his clotl~ing t o become saturated with the bath liquid, since this accentuates any tendency to skin irritation.

27. PREPARED BATH. Pre-mixed bath, contaming the r ight proportions of fluorescent magnetic particles and conditioners, a r e available, but are not usually purchased In bulk for filling tanks of testing units. It is more economical to purchase the easily mixed concentrate and buy the oil in bulk. Fo r water based baths the liquid i s aiways a t hand for mixing with the concentrate.

Prepared bath is convenient, however, in connection with field inspection or other occasions when portable magnetizing equipment 1s being used. The bath in such applications IS furnished in pressurized spray cans for easy application to surfaces. The Itits fo r these inspections aiso ineiude spray cans with solvent fo r surface cleaning before and af ter inspection.

289

1;:: CrtarrEn 16

j _: BLACK LIGWT-ITS NATUKE, SOUIICES AXD HEQUIIIEMENTS --

BLACK LIGHT-ITS NATURE, SOURCES AND REQUIRE&fENTS

1. DEFINITIONS. Black Light, as the term is used for the purposes of fluorescent magnetic particle testing, i s radiant energy having a wave length in the band from 3200 to 4000 Angstrom units, peaklng a t 3650 A. T h ~ s is shorter than the shortest visible violet wave length, and longer than tlie t rue "liard" ultraviolet. I t is often called "near-ultraviolet" light. Black light has the property of causing many substances, such a s certain minerals and dyes, t o Auoresce. Though the radiation i s not v~sible to the eye and is therefore characterized a s being "black" light, i t i s produced by a high pressure mercury a rc a s a by-product of visible white light along with other sliort-wave radiation.

Fluorescence is defined a s the property of certain substances of "emitting radiation a s the result of, and oniy dttnng the absorption of radiation from some other source." The terms "phosphorescence" and "lum~nesceuce" both refer to other types of light eniis- sion. The former is caused by the absorption of other radiation, and continues t o glow for some time after the energrzing radiation has been removed. The latter is a self-generated light resulting from organic, chemical or electronic reactions within the substance itself. To produce flzco?.escent light, the energizing applied radiation must be acting d u r ~ n g the time light is emitted.

In the fluorescent magnetic particle testing process, dyes a re used which absorb the ~nvisible short-wave black light, and emit t h ~ s energy in longer wave lengths in the visible range, givmg a b r ~ g h t yellow-green light. Other substances may fluoresce in varrous colors, from blue tlirou@li green, yellow, orange and red.

2. ULTRAVIOLET LIGHT. Ultraviolet liglit is the term applied t o radiation of wave lengths shorter than the shortest visible violet wave length-from around 4000 Angstrom units down to 2000 A. I n general, tlie shorter the xvave length the more penetrating and active is the radi a t' ion.

However, the band between 4000 A and 3200 i is relatively inactive when compared to tlie band from 3200 1 do~vn to 2000 A.

'I!ltraviolet around 2600 A is very harmful to many fornis of life. It will kill bacteria, cause sunburn, generate ozone and can be very injurious to the human eye. The electric a rc a s used in weiding IS

rich in such s'llort >vare ultraviolet, and welders ( a s wcll a s watch- ers) must take special care to filter out these short waves before they react1 the eye, or else not look directly a t the arc.

\\'elding helmets a re fitted with very heavy dark blue or dark purple glass filters to safeguard the eyes. Filters are also used in the case of Huorescent magnetic particle testing, only 111 tliis case the filter is a special one, designed to pass the masimuln amount of wave lengths \dhich activate tlie fluorescent dyes, and to exclude all other wave lengths t o a s great a degree a s possibie, both visible light and short wave ultraviolet. The desired wave lengths are between 3500 :I and 3800 A.

3. SUNLIGHT AS A SOURCE OF ULTRAVIOLET LIGHT. Radiation from the sun is composed of energ?. of a very w d e range of wave lengths in addition to the visible light which the eye can see. Just below tlie v~sible light there is a strung invisible radiation in the ultraviolet. T l i ~ s energy will cause fluorescent dyes to glow, although the strong white light also present does not normally permit tliis to be apparent.

I I

I I ! ~ l i l ~ l ~ l i l ~ l ~ l ~ ! ~ J 2000 25110 3000 U O D .iOW ASW 1000 5SW 6- 6Ya 7 m o

WaYL LCNCTH IN IINGSTEmIS

Flg 132-Spectrum of Llght Through the V8stble. Black Llght and Ultrav!olet Wave Lengths

However the now familiar "daylight fluorescent" colors, p a ~ n t s and p r ~ n t i n g inks a re energized by the natural ultravrolet i n sunlight, to malte them super-bright even in full daylight. Some of these colors react to ~1101% !wave ultrav~olet a s well a s to black light. At dusk and dawn there is a difference in light intensity and light distribution in sunliglit, due to the longer distance the sun's rays must travel through the atmosphere a t those times of day. This results in different amounts of refraction for dill'erent wave lengths. There is greater refraction of the shorter wave lengths in sunlight, so tha t these a re bent to~vard tlie earth t o a greater degree than


are the longer visible rays. Thus the actinic energy s t r ik~ng daylight fluorescent materials a t dusk and a t dawn is proportionately higher than a t high noon. This effect causes daylight fluorescent objects to appear especially bright a t twilight and just after sun rise.

For the purposes of fluorescent magnetic particle testing, however, it would not be very practical to filter sunlight to secure proper black iight intensity. Therefore artificial sources must be utilized.

4. FLUORESCENT DYES. As has been stated, numerous fluorescent dyes are known, and a good many have been used in fluorescent magnetic particle and in penetrant testing. The best dyes for this purpose react strongly when energized with black light of 3650 A wave length, and the preferred ones emit light in the green-yellow range, the most favorable range for easy "seeing" with the eye. Dyes have been especially developed to give improved performance for the purposes of fluorescent inspection, and those in use today are f a r superior in fluorescent brilliance and color to those that mere available ten years ago.

These dyes are also remarkably stable in their fluorescent light output during prolonged exposure to blafk light-a property of importance for fluorescent particle testing, and one not possessed by very many of the fluorescent dyes available.

5. SOURCES OF BLACK LIGHT. Although sunlight contains a large amount of ultraviolet, i t is not very adaptable for the production of black light for the purposes of fluorescent particle testing. How- ever, the electric arc drawn between two metal or carbon electrodes, is especially rich in energy in the ultravtolet range. The enclosed high pressure mercury vapor arc lamp also offers a convenient source which is high in output of the desired biaclc light wave iength. With few exceptions, 'black light lamps used for fluorescent particle testing empioy the,mercury arc in one form or another. Figure 133 gives the spectrum of light given out by a high intensity mercury arc lamp, and shows the energy distribution over the range of ultraviolet and visible wave lengths. There are a number of pealcs, one of which is a t the desired wave-iength of 3650 A.

6. FILTERS FOR BLACK LIGHT. The glass filter almost un~versally used to separate out the 3650 A wave-length is a dark red-purple coior. I t is selected to remove effectiveiy practically all visible light from the energy given off by the mercury arc. At the same time,

CHAPTER 16 BLACK LIGHT-ITS NATURE, SOURCES AND REQUIREDlENTS

2200 3000 do00 W O 61100

W#VE LFNCIiH IN ANGSTMMS

Fig. 133-Spectrum ot the Output of the High Pressure Mercury Arc.

it also removes all radiation of wave length beiow 3000 A-that is, it eliminates all the harmful short-wave ultraviolet. I t passes near-ultraviolet radiation in the range from 4000 A (the lower edge of the range of visible violet) down to 3200 A. The radiation passed by the filter peaks a t 3650 A, the optimum for energizing most of the fluorescent dyes used for magnetic particle inspection. Figure 134 shows the transmiss~on curve of such a filter (Kopp 41 glass).

I t will be noted from the curve that a small amount of visible light is transmitted. This is not altogether undesirable, stnce it permits the inspector to discern the objects in the immediate vicinity of the black light source, and therefore facilitates handling of parts during inspection. The filter also passes infra-red radiation of wave length longer than the visible red, but thls is of no consequence from the point of view of fluorescent particle testing.

7. FLUORESCENT EMISSION FROM DYES. I t has been stated that fluorescent dyes are available that emit light in various wave lengths from red to blue. Tests have determined that the eye finds a yellow- green light more easily seen than reds or blues, so that the dyes

293


NPP .L %' TYIC<

WAVELENGTH IN ANGSTSDhlS

Fig. 134--Light Transmiss~on Curve ot Black Light Filter Glass (Kopp 41).

WAVELEWTH IN mGSTROMS

Fig. 1 3 S L i g h t Emlsslon Spectrum of Yellow-Green Fluorescent Dye

294

.~= ::.,: . ., ".' CHAPTFX 16 ::

RL.4CR LIGHT-ITS NATURE. SOURCES AKD REQVlREhIEXTS .- . .: ' most widely used for fluorescent particle inspection a re those which

give oft' this xvave iength of light. Figure 135 sho~vs the emission curve of a typical dye when energized xvith black light of 3660 .\ wave iength.

8. BLACK LIGHT L.A&IPS. There a re a number of types of black light lamps available commercially, though not all a r e satisfactory a s sources of radiation for fluorescent inspection, Some of these lamps do not give sufficient energy in the 3650 .A range t o meet the ntz.?az?mi?n intensity requirements for inspection purposes. All use colored filters of glass intended t o remove risible and short wave radiation, but not all are equally successful. Some commercial filters pass excessive amounts of white light.

9. TUBULAR BLACK LIGHT LAMPS. Tubular black iights a r e similar in construction and operation to the familiar tubular fluorescent lamps for general illumination. These emplo). a low pressure mercury vapor arc, and the inside of the tube is coated with a phosphor which fluoresces under the energy of mercury vapor discharge. Surrounding these tubes with larger tubes of red-purple filter glass (or using tubes of red-purple giass to manufacture the lamps) results in a black light source.

Small, battery operated tubular black lights and 6 watt tubes operated on 110 volts a re used for detecting and ~dentifying fluorescent minerals. They a r e sometimes used in fluorescent iuspec- tion, but only when i t is known tha t the sought-for defects, if present, mill produce strong indications. In such cases the lamp should be held very close to the work and the eye also should be brought close to the area being examined.

The larger size 40 and 60 watt tubular lamps h a w the same limitations a s the small ones-insufficient output. These larger tubes will give proportionately more light, but i t is produced over a larger area. The tubes are 18 to 36 inches long, whereas the 6 watt tubes and the battery lamps a re only 5 to 6 inches long. Since tile larger tubes a re not portable, a s the 6 watt tubes are, they have even less general utility for fluorescent magnetic particle inspection. If mounted in 'banks of 4 o r 6 they may give fairly good results if close enough to the inspection tabie (15 to 18 inches), espec~ally if oniy gross defects a re being sought. The black light intensity level in even a bank of tubular lights is likely to be too low to ensure the location of very fine cracks. Tubular black lights also put out relatively larger amounts of visible light than do the mercury a rc


types. One further drawback to the use of these tubular lights is that their output of 3650 A light drops off rather rapidly with age.

10. INCANDESCENT BLACK LIGHTS. A black light lamp has been on the market which is similar to an ordinary photoflood lamp, except that the envelope is blown out of the red-purple filter glass. This is another low output black light source of little or no value for fluorescent inspection work. The incandescent filament does not put out a great deal of ultraviolet, so that the amount of black light passing through the filter is not adequate. Lamp life is short and heat output great. Like the 6 watt tubular lamp, its only utility is in the intermittent energizing of gross fluorescent indications. I t cannot be relied upon for any important flaw-detection work.

11. MERCURY VAPOR LAMPS. The enclosed high pressure mercury vapor arc lamp is by far the most important black light source for fluorescent inspection and is almost universally used. The construction of this lamp IS shown in the drawing (Fig. 136). The

,v . . MERCURY VAPOR L A M P CONSTRUCTION

7 Fig. 136--Construction Drawbng of the High Pressure Mercury Arc Lamp.

CHAPTER 1 6

BLACK LIGHT-ITS NATURE, SOURCES AND REQUIREMENTS

mercury arc is drawn between electrodes enclosed in a small quartz tube, Q. The current-carrying electrodes are E, and E?, and E. is an auxiliary starting electrode or heater. The resistor R limits the current in the starting electrode. This arc cartridge is mounted and enclosed in an outer glass envelope, B, which serves to protect the quartz cartridge and also to focus the light emitted. The lamp is supplied current from a special current-regulating ballast transformer, which limits the current the arc can draw.

When current is first turned on, the mercury arc is not set up a t once. A small, iow-current arc through the gas in the cartr~dge is first started by the auxiliary electrode, bringing about suffic~ent vapor~zation of the mercury in the tube to start the arc between the main electrodes. This process takes about 5 mlnutes to build to full intensity.

12. LAMP OUTPUT. The quartz inner capsule passes energy of the higher ultraviolet wave lengths, as well as visible light. See Fig. 133. The character of this spectrum is controlled by the design and manufacture of the lamp, principally by selecting the vapor pressure inside the quartz capsule. At very h ~ g h pressures (100 atmospheres) the spectrum is practically continuous over a wide range, but at a somewhat lower pressure level (10 atmospheres), the output is largely in the visible and ultraviolet range. It is about equally distributed over that range, and includes the des~rable black light wave lengths. Lamps of thls type yield the maximum energy in the form of black light after filtering, and this is the type used for inspection purposes.

13. COMMERCIALLY AVAILABLE BLACK LIGHTS. The self-contained mercury arc iamps illustrated in Fig. 136 are available in several forms. The 100 watt reflectorized lamp is most commonly used in ordinary work. The 100 watt bulbs are made as either "spot" or "flood" lights. The latter gives a more even illumination over a larger area than the "spot", but is not nearly so desirable for critical inspection use. The illumination ievel is less than the minimum usually requlred unless the iamp is held extremely close to the work. The "spot" lamp, on the other hand, concentrates much of its energy in a relatively small area, thus giving maximum illumination on the locations at wh~ch the eye of the inspector 1s focus- ing in look~ng for indications. To be more specific, this type of lamp glves adequate intensity of black light for nearly all inspection purposes over a 6 inch diameter circle in a plane 15 inches from

297

PRINCIPLES OF MAGNETIC PAKTICLE TESTING

the black light filter. In the center of this circle the light intensity is about double the intensity a t the edge of the circle.

Fig. 137-The 100 watt Black Light Mounted on a Fixture on a Magnetirtng Unit.

Figure 137 illustrates a lamp of the spot type, mounted in a convenient fixture on a magnetizing unit. The inspector can lift i t easily and hold i t in his hand to direct the light to the area being examined. Figure 138 s h o ~ s the 400 watt mercury arc black light, similar in all respects to the 100 watt, except, of course, that its size i s unsuitable for portable mountings. T h ~ s light a t 15 inches gives a much larger circie of maximum intensity than the 100 watt. It may be mounted much farther from the work and still glve adequate illumination over an area ten times as large as that covered by the 100 watt light. Batteries of the 400 watt lamps are used

CHAPTER 16

BL.4CK LIGHT-ITS NATURS, SOURCES AND HEQIJIREhlEKTS

Fig. 1 3 L T h e 400 Wan Black Light

when, in larger installations, large areas must be covered adequately with black light. Currently even higher wattage lamps are becoming available, whlch will undoubtedly be used for large test-area illumination.

14. INTENSITY R E Q U I R E ~ ~ ~ E N T S OF ELACK LIGHT. Since the amount of visible light emitted from a fluorescent pigment is directly dependent on the amount of black light energy supplied to it, the intensity of black light at the point of inspection is of the h~ghest importance.

Three factors determine how perceptible an indication is: ( a ) the amount of dye a t the indication; (b) the awwunt of response of the fluorescent dye, in the form of emitted visible light in relation to the energy supplied by the black light; and ic) the actual a?jlou?zt of energy supplied to the dye by the black light radiation. Given an indication with a fixed amount of fluorescent dye, the brilliance

1 PRINCIPLES OF MAGNETIC PARTICLE TESTING !

of the indication varies with the intensity of incident Hack light. This consideration is critical when seeking extremely fine indications.

Experience has shown that 90 foot candles a t the point of inspec- t i on is the minimum black light intensity adequate for the location of fine cracics, such a s grinding cracks. Larger or more open cracks, which hold a larger amount of fluorescent particles, may require only 70 foot-candles, and 50 foot-candles may be sufficient for many gross defects. However, unless one kno5t.s in advance bow fine a crack i t may be important to detect, no permanent installation should be set up to provide less than 90 foot-candles a t the working polnt. These levels assume substantial exclus~on of white light from the inspection area.

As previously stated, this intensity is provided over a 6 inch diameter circle by a 100 watt spot lamp held 15 inches from the work. At the center of the spo t the intensity is a t least double t h ~ s amount. Figure 139 shows the distribution of blacic light intensity, r

RELATIVE AREAS OF MINIMUM' DESIRED INTENSITY (100 Fc.) AT

TYPICAL DISTRIBUTION CURVES OF

15 INCHES FROM SINGLE 400 BLACK LIGHTS FROM 100 AND

AND 100 WATT LAMPS RESPECTIVELY 400 WATT LAMPS ON A PLANE 15 INCHES FROM LAMP

Fig. 139-Distribution of Black Light from 100 Watt and from 400 Watt Black Light Lamps, Held 15 Inches from the Work Table.

s .irr

f#$ .:? pr?: BLACK LIGHT-ITS NATURE, SOURCES AND REQUIREMENTS

furnished by the 100 watt spot lamp and by the 400 matt lamp, when held 15 Inches from the work, from the center of the spot radially to the edges of the illuminated area. .. ~

15. R~EASUREMENT OF BLACX LIGHT INTENSITY. For the purposes of control of the inspection process, black light intensity should be checked a t regular intervals for reasons discussed in the following section. Exact photometric measurements are not required, although reproducible determinations of intensity with reasonable accuracy are essential. Fortunately such a check can be easily and quickly made with very simple equipment.

Standard light meters, consisting of photo-voltaic cells and indicating micro-ammeters, give reasonably accurate readings of light intensity directly in foot-candles. Standard meters of this type are calibrated for white light intensity readings. However, the meters have a reasonable response in the black light wave length range, as can. be seen from the curves in Fig. 140. In this figure the response of the photo-voltaic cell is superimposed upon the curve of light transmission by the filter glass. Readings taken under black light will not be numerically accurate, but are reproducible, and

I Fls i 4 ~ T r a n s m l s s r o n Curve for Black Llght Ftlter Glass (Kopp 41) and Response Curve of the Photo Voltalc Cell

PRINCIPLES OF I)fAGNETIC PARTICLE TESTING

therefore serve adequately as a measure of the re iu t i~e black light intensity.

The meters generally used are the 'Vi'eston Model 703 Type 3, Sight Light Meter, and the General Electric Meter, Model No. 8DW40Y16. The meters are calibrated from zero to 75, or zero to 100 foot candles with multiplier masks to multiply readings by 10 for higher intensities. In use, the meter 1s exposed to the black light, face up, a t the polnt the measurement is to be made, and foot-candle readings taken directly from the meter scale: using the multiplier mask when needed. "Viscor" or other filters which pass only visible light and filter out black light should of course not be used in maicing such measurements.

6 CAUSES OF VARIATIONS IN BLACK LIGHT INTENSITY. It has been stated that the intensity of the black light a t the point of inspection should be checked a t intervais, because a lower-than- optimum intensity may seriously affect inspection results. Some of the causes of intensity variations are the following:

( a ) Variations in bulbs. Black light output of mercury arc lamps of the same nominal wattage ratings may vary significantly in lamps from different manufacturers. Even bulbs from the same manufacturer cannot be relied upon to produce uniform amounts of blaclc light. Wattage ratings of bulbs do not, therefore, guarantee the amount of black light that individual bulbs will produce, even when new.

(b) Black light output of a given bulh varies almost directly with applied voltage, and a bulb that gives good intensity at, say, 120 volts will produce much less blaclc light a t 105 volts. Light meter checks are the best means for verifying the output on any given electric suppiy circuit.

ic) Black light output of any bulb falls off with age, and as a bulh nears the end of its life output may drop to as iow as 25% of what it was when the bulb uras new. The lives of these bulbs may vary widely. Nominal life expectancy is given by the manufacturer ilOOO hours for the 100 watt spot), but for varlous reasons the actual life of a bulb in inspection \rrork is less than this figure. The manufacturer's life tests are based on continuous, steady operation in a fixed and nrell ventilated position. Black lights used in inspection are subject to numerous starts and shut-offs, and to rough

CHAPTER 16

BLACK LIGHT-ITS XATURT;. SOGRCES AXII IZEQUIREMENTS

L I N E V O L T S

Fig. 141-Output Vargations with Varying Voltage. 100 Watt "spot" Black Light.

handling. In addition, due to the needs of filtering and portable housing, such lights may operate a t higher temperatures than are desirable. (See also Section 17 below.)

id) Accumulations of oils films or dust and dirt on the bulb and filter seriously reduce the black light output: sometimes by as much as 5070. This cause of variation can be avoided by keeping the filter clean, but an output check with the meter is the only safe way to make sure that the cleaning has been effective.

17. BLACK LIGHT OPERATING CHARACTERISTICS. AS has been stated, when the current to a black light bulb of the mercury arc type is first turned on, it takes five minutes or more for the bulb to warm up to its full output, and no inspection should be started until this time has elapsed. If, for any reason, the arc is extin- guished, as by interruption of the current, the bulbs will not respond immediately when the current is again turned on. Time must be allowed for the lamp to cool some\vhat and then for the arc to re-establish itself. This takes up to ten m~nutes. Customarily, a


black light, once in operation, is best left on even when not actually in use continuously.

Another reason to leave the bulb turned on is that each start affects the life of the bulb, possibly reducing it by as much as three hours per start. I t is therefore better, from the standpoint of bulb life, to leave the light on all day, even if it is in actual use for only a few hours.

Low voltage will extinguish the mercury arc: and where line voltage is subject to w ~ d e fluctuations, with low points a t 90 volts or less, the lamp may go out. Black lights should not be operated on such circuits if i t can be avoided, but if it cannot, special constant- voltage transformers for the black lights should be used.

Large line-voltage fluctuations or low voltages may cause annoy- ance and delay due to bulbs going out, and loss of bulb life due to unnecessary starts, but high voltage surges also decrease bulb life seriously. Line voltages above 130 volts may cause very early burn- outs, and the special constant voltage transformers should be used if line voltages are consistently high.

18. ACHIEVING ADEQUATE BLACK LIGHT INTENSITY. The 100 watt spot black light bulb in the hand fixture is a very convenient, reliable and flexible source of adequate black light. I t may be held closer to the work than 15 inches, though the amount of visible violet light in the spoC may cancel out some of the advantage of doing this. Another advantage is that the inspector's eye can be closer to the polnt of inspection when the black light is hand-maneu- vered to a favorable location.

In some installations, especially where large numbers of parts are being examined, this approach is not so convenient, and it IS

desirable to have adequate black light intensity over a much larger inspection area. To accomplish t h ~ s , clusters of black light spot lamps can be directed into the inspection area. In a very few cases banks of tubular black lights have been adequate. Where still larger areas are to be illuminated, a number of the larger 400 watt iamps are used.

19. EYEBALL FLUORESCENCE. Reference has been made several times to the fact that the eyeball fluoresces when black light 1s

directed a t it. Althougti the effect is harmless, i t is not only annoying, but interferes with vislon while it exists. In the location and manlpu-

BLACK LIGHT-ITS NATURE. SOURCES AND REQUIRE3IENTS

lation of black light fixtures, this occurrence must be avoided. Black light is reflected from shiny surfaces just as uihite light is, and if reflected into the eyes will cause such fluorescence to occur. Place- ment of black lights therefore must avoid reflections as well as direct radiation into the eyes.

When the hand held light is used, the surface of the part being examined should be a t an angle so that reflection into the eyes does not occur. Although it is sometimes necessary to brlng the eye close to the work when iooklng for fine indications, this must be done without getting into any direct or reflected black light.

Yellow glasses (such as "Wilsonite" sun glasses) may be worn by the inspector to cut out all black light from the eyes and in cases where the operator is bothered by the fluorescence of his eyes, this is an effective remedy. The lenses of these glasses must pass all of the green-yellow fluorescent light from the indication, but exclude Hack light. Yellow glasses also exclude most of the visible violet, and therefore many operators do not like to wear them, smce they give the effect of working in almost total darkness. Before adopting such protective glasses, the light transmission curves of the yellow glass should be checked against the light emission curve of the fluorescent dye.

CHAPTER 17.

DEMAGNETIZATION

I . INTRODUCTION. After having gone to a great deal of trouble and care to introduce a strong magnetic field of special klnd and direct~on into a pal+, the operator IS frequently faced xvith the problem of getting rld of this field again, af ter the test is completed. Dei~lagnetizing af ter ~nspection is sometimes as much of a prohiem a s mas proper magnetization in the first place.

All ferromagnetic materials, af ter havlng been magnetized, 11-ill retain a residual fieid to some degree. Thls field may be negligible in very soft materials; bu t in harder niateriais i t may be conlpa- rable t o the intense fields associated with the specla1 alloys used for permanent magnets. Almost any ferromagnetic materlal may, f o r one reason or another, be subjected t o magnetic particie inspection, and may have to be demagnetized aftertvards. The problem of demagnetization may be easy o r difficult depending on the type of materiai. &lateriais having a high coercive force a r e the most difficult to demagnetize. High retentivity is not necessarily related directly t o h ~ g h coercive force, so that the strength of the retained fieid IS not always a good guide a s to the probable ease or difficulty of demagnetizing.

I t 1s not ai\vays necessary t o demagnetize par ts af ter inspection, and since the process involves time and espense, there i s no need to apply it unless there is some good reason to do so. I n the earlier days of magnetic particle testing demagnetization was almost always carried out a s a mat ter of course and without really conslder- m g whether i t was actually necessary o r not. However, in many cases i t zs essential to demagnetize and t he operator should understand the reasons for this step; a s well as the problems liivolved and the available means for solving them.

2. REASONS FOR DERIAGNETIZING. There a r e many reasons for deniagnetizing a par t af ter i t has become magnetized from any cause. Demagnetizing before magnetic particle testing is sometimes necessary, in cases where highly retentive par ts come up to the test with s t rong res~dua l fields from some previous operation, such a s magnetic chucks o r cranes. See Fig. 186, Chapter 21.

306

Demagnet~zxtion is necessary when t l ~ e residuai field in a par t

t a ) may interfere with subsequent maciilning operations by causing cliips to adhere to the surface of the par t o r of the tool, the tip of which may become magnetized from contact with the magnetized part. Such chips can Interfere with smooth cutting by the tool, adversely atl'ecting both finish and tool life.

f b ) may interfere with electric a r c welding operat~oiis to he performed subsequentlp since strong residual fields may deflect the arc away from the p o ~ n t a t w h ~ c h i t should be applied.

tc) may interfere with the operation of instruments wliich are sensitive to magnetic fields, such a s the magnetic compass on an aeroplane, or may affect the accuracy of any form of instrumentation incorporated in an assembly whicl~ includes the magnetized part.

( d ) may Interfere with the functioning of the par t itself, af ter it is placed into servlce. Magnetized tools, such a s milling cutters, hobs, etc. will hold chlps and cause rough surfaces. and may even be broken by adherent chips a t the cutting edge.

te) might cause trouble on moving par ts by holding particles of metal or magnetic testing particles-as, for instance, on balls o r races of ball bearings, o r on gear teeth.

( f ) may prevent proper cleaning of the par t af ter Inspection by holding particles magnetically to the surfaces of a par t .

( g ) IS likely to interfere with the magnetization of a par t a t a lower level of field intensity, not sufficient to overcome the remanent field in the part .

( h ) may hold particles which Interfere with later applied coatings sucli a s plating or paint.

Although demagnetization is not necessary in practically all other cases, i t is often still practiced a s a routine step with no partrcular objective.

3. WHEN DEMAGNETIZATION IS NOT NECESSARY. Demagnetiza- tion is not necessary and is not usually carrred out when


(a) parts are of soft steel and have low retentivity. In this case the remanent field is low or disappears after the magnetizing force is no ionger acting. An example is low-carbon plate such as that used for low strength weldments, tanks, etc.

(b) the material in question conslsts of structural parts such as ~veldments, large castings, boilers, etc., even when made of high-strength alloys whlcb retain a considerable field. In such cases the presence of a resldual field would have no effect on the proper service performance of the part.

t c) if the part is to be subsequently processed or heat-treated and in the process will become heated above the Curie point, or about 770°C (about 1390°F). Above this temperature steels become nonmagnetic, and on cooling are completely demagnetized when they pass through the reverse transformation.

) the part will become magnetized anyway durlng a following process, as, for example, being held on a magnetic chuck.

ie) a part is to be subsequently re-magnetized in another direction to the same or higher level a t which i t was originally magnetized as, for example, between the steps of circuiar and longitudinal magnetizing, for magnetic particle testing purposes.

( f ) the magnetic field contained in a finlshed part is such that there are no external leakage fields measureable by ordinary means as, for example, the testing by circular magnetization of welded and seamless pipe.

The case cited under (e) above is sometimes a cause of confusion. The establishment of a longitudinal field after circular magnetization wipes out the circular field, since two fields in different directions cannot exlst in the same piece a t the same time. If the magnetizing force is not of sufficient strength to establish the longitudinal field i t should be increased, or other steps taken to insure that the longitudinal field actually has been established. For example, a large part having a large L/D ratio may requlre sexzeral longitudinal "shots" along its length to wipe out the circular field. But, once a trzllv longitudinal field is established in the part, the circular field no longer exists. The same is true in going from longitudinal to circular magnetization. If the two fields, longitudinal and circular, are applied s~multaneously a field will be established which is a vector combination of the two in both strength and direction.

CHAPTER 17 DEhlAGNETIZ.4TION

But, if the fields are impressed successively, the last field applied, if st.rong enozlgh to establish itself z?z the part, will wipe out the remanent field from the previous magnetization. If the magnetizing force last applied is not of the same or higher order of magnitude as the preceding one, the latter may remain as the dominant field.

4. LIhlrTs OF DEMAGNETIZATION. When a piece of unmagnetizcd steel is first magnetized, the field within the part increases from zero to the saturation point along the virgin magnetization curve shown as the dotted line on the hysteresis curve in Fig. 56, Chapter 5. Once having been magnetized, the field in the part cannot again be made to traverse this line unless the part is eomplete2~ demagnetized. All steels have a certain amount of coercive force, and it is extremely difficult to bring the steel back to the zero point by any magnetic manipulation. In fact, it is so difficult that for all practical purposes it may be said that the only way to demagnetize a piece of steel completely is to heat it red hot, and cool i t with its length directed east and west, to avotd re-magnetization by the earth's field.

When steel IS heated it passes through a transformation polnt, approximately a t 770°C., or about 1390°F. for soft steels) at which it becomes nonmagnetic and its permeability drops to I, the same as air. Above this point the steel becomes austenitic. Fig. 142 illustrates the change that takes place in magnetic properties when iron

I I

Fig. 142-Effect ot Temperature an the Magnet~c Propettles ot iron.


is iieated above tlie Curie Point. Table VI gives !he Curie Point f o r several materials. When the steel cools do\vn i t goes through the reverse transformation, and unless cooled under tlie influence of a magnetic field, will contam no residual magnetism.

TABLE VI

CURIE POINTS FOR SOME FERROaIAGNETIC MATERIALS

Other means of demagnetizmg almost always leave, fo r one reason or another, some residual field in the part. As a practical matter, therefore, the process of demagnetization results 111 either

t a ) the best possible job tha t existing means will produce, o r

Ib) tlie level of residual magnetisin considered permissible in the particular par t involved. "Complete" demagnetization is usually not necessary, even though i t is often specificti. Tlie principal reason for attempting to secure complete demagnetization is in cases where the remanent field may affect the operation of instruments sensitive to weak magnetic fields.

It must be remembered tha t the earth's field will always affect the magnetism in a ferromagnetic par t and mill often determine the lo~\ser limit of practical demagnetization. Long parts, o r assemblies of long parts, such a s welded tubular structures, a r e especially likely to remain magnetized, a t a level determined by the earth's field, in spite of the most careful demagnetizing technique.

Many articles and par ts become quite strongly magnetized f rom the earth's field alone. Handling of parts, such a s transporting from one iocation to another, may produce this effect. Long bars, cle- magnetized a t the point of testing, Rave been found magnetized when delivered a t tlie point of use. I t is not unusuai to find t ha t par ts of aircraft, o r autoinotivc englnes, railroad locomotive par ts or, in fact; any parts made from steel of f a i r retentivity, a r e quite

strongiy iix~gnetized af ter having been in service for some time, even though tliey may never 1i:ive been near any artificially produced magnetic field. Par ts also become niagiletized acc~dentally by being near electric po\srer lines carrying heavy currents, o r near some form of magnetic equipment.

Such parts have apparently operated perfectly s:ttisfactorily nl tliis Iolv level magnetized condition, which suggeiis that the need for extremely thorough demagiietization may often be over- emphasized.

I t seems therefore, tha t the limits of demagnetization may be considered to be either the maximum extent to \\.hich the par t can be demagnetized by available procedures, or the level to ~vhich the terrestrial field will permit i t to become demagnetized. These limits may be fur ther modified by the practical degree or limit of demagnetization which is actually desired or necessary.

Specifications for demagnetization shouid be examined to take these facts into account. I t is unqlrestionably true that specifications for demagnetization have called for levels of remanent iields lower than are practicable or even possibie of a t t a~nment . If demagnetization of parts is called for, the specification should state a limit of permissible residual field t ha t i t is reasonably possible to achieve. Unrealistic requirements shouid be modified in tlie light of what needs to be or \$,hat can be done. See Section 14, this chapter.

5 . APPARENT DEMAGNETIZATION. I n considering the problem of deniagnetization i t is important to remember tha t a par t may retain a strong residual field af ter Iiaving been circularly magnetized and yet exhibit little or no exter?zai evidence of such a condition. On the other hand, par ts which have been longitudinally magnetized possess externai poles which are easily detected.

Since it may be sonletirnes only the external evidence of a remanent field that 1s objectionabie, par ts may not require demagnetization after being circularly magnetized. Tlie c i ~ ~ u i a r step should be conducted after tlie longituciinal magnetization if tliis is lo be done. But circular fields, even if they show no external evidence of their presence a t the time of testing, may still be objectionable. As soon as such a par t is cut into, or even in some cases if i t merely comes in contact with an unmagnetizeri par t , an external pole will be developed. Thus, if a circularly niagnetized par t is maciilned, the tool t ip may become magnetized and hold chlps; or the cut being

PRINCIPLES O F DIAGNETIC PARTICLE TESTING

made may draw a field to the surface and set up an external pole that has tlte same effect. In such cases, if truly co?iiplete demagnetization is desired, it is often helpful if the longitudinal niagnetization be conducted last, slnce i t is easier to remove this field and to check the extent of its removal.

6. HOW DEMAGNETIZATION IS ACCOMPLISHED. There are a number of ways and means of accomplishing demagnetization, of van- ous degrees of effectiveness. A11 can be explained by means of the hysteresis loop, and nearly all are based on the principle of subjecting the part to a field continually reversing i ts direction and at the same time gradually decreasing in strength down to zero.

What happens is illustrated in Fig. 143. The sine wave or curve of a reversing current a t the hottom of the graph is used to generate the hysteresis loops. As the current dimin~siles in value with each reversal, the loop shrinks and traces a smaller and smaller path.

I Fig. 143-Flux Curve Durtng Demagnetoation. Prorected

from the Hysteres~s Loop.

The curve a t the upper right of the drawing represents the flux in the part as ~ndicated on the diminishing hysteresis loops. Both current and flux curves are piotted against time, and when the current reaches zero the remanent field in the par t will also have approached zero.

Precautions to be observed in the use of this principle are : first, to be certain that the magnetizing force is high enough at the s tar t to overcome the coercive force, and to reverse the residiial field initially in the part ; and second, that decrements between successive reductions of current are small enough so that the reverse magnetizing force will be able, on each cycle, to reverse the field remaining in the part from the last previous reversal.

Frequency of reversals is an important factor affecting the success of this method. With higll frequency of current reversals the fieid generated in the part does not penetrate deeply into the section, since penetration decreases as frequency increases. A t a frequency of perhaps one reversal per second penetration of even a large section IS probabiy near 1000/o although for moderate sized parts, the 60 or 50 cycle commercial frequencies of alternating current give quite satisfactory results.

7. REMOVING LONGITUDINAL A N D CIRCULAR FIELDS. AS has already been said, in most cases it is the external effect of a residual field that is objectionabie and that constitutes the reason for demagnetization. The external poles attract ferromagnetic chips and particles and influence the operation of magnetic instruments. A part wi~ich is longitudinally magnetized has external poles, and demagnetization aims to reduce these to the lowest possible value. The degree of success attained can be checked by means of a compass needle or field meter. See Section 14, this Chapter and Chapter 8, Section 5.

A part which is circularly magnetized may have no external poles a t all and there may be no apparent indication of the presence of an internal field. Such a field is more difficult to remove by any of the usual methods, and there is no easy way to check the success of the demagnetization. There may be local poles on a circularly magnetized piece a t projecting irregularities or changes of section, and these can be checked ~v i th a field meter. However, in undertak- ing to demagnetize a circular magnetized part, it is often better first to convert the circular field to a longitudinal fieid in a D.C.

I'RINCIPLES OF BIAGNETIC PARTICLE TESTING

solenoid. The longltuillnal field is more easily removed and the extent of removal can be easily checked.

As has been said, ;~ltliougll a clrcular fieitl may glve no evidence of i ts presence; i t can still cause trouble, and may reyuire removal. If, therefore, demagnetization is called for, the operator must know the best steps to be taken to secure a s thorough a lob a s possible, ~slietl ier the existing field be longitudinal o r circular.

8. DEMAGNETIZATION WITH A.C. The most common method of deniagnetizing small to moderate sized par ts is by passing them through a coil through which alternating current a t line frequency is passrng iusually 50 to 60 c.p.s.). Alternatively, the 60 cycle A.C. 1s passed through a coil with the par t inside the coil, and tlie current gradually reduced to zero. I n the first case the reduction of the strength of the reversing field is obtained by witlidrawal of the part , axially, from tlie coil ( o r the coil f rom the par t ) and for some distance beyond the end of the coil ior par t ) along that as ia l line. In the second case the gradual decay of the current in the coil ac- complishes tlie same result.

This is a simple, quick, easy way to demagnetize (\\,hen i t works), to produce an acceptably low level of residual field. Fo r best results par ts should be passed through the coil with their longest dimension parallel to the axis of the coil. The par t should be held close to one \vall of the coil, o r one corner if the coil is rectangular in shape, to take advantage of the stronger field a t such locations. Ideally the coil should be of a slze to be nearly filled by the part, and for this reason, demagnetizers of t h ~ s type a r e furnislied with various coil diameters.

Small par ts should not be piled into baskets and the baskets passed through the coil as a unit, since A.C. will not penetrate into such a mass of parts, and only the few parts on the outer edges ~vi l l ever be properiy demagnetized. Even these will often be only partially demagnetized because of their contact with other par ts deeper in the pile. Small par ts can be demagnetized in multiple lots oniy if they a r e piaced in a single layer on a t ray which holds them apar t and in a fixed position with their long axes parallel t o tlie axis of the coil.

Sixty cycle A.C. 1s not very effective for demagnetizing par ts of iarge or even moderate size because of i ts inability to penetrate. The difficulty can be overcome by using a iower frequency A.C. Twenty-five cycles-per-second current is f a r more effective than

C R A P ~ 17

DEIAGNETIZAT?ON

60 cycles. In some installations motor-generator sets delivering 25 cycle current have been used in order to achieve better demagniiza- tion and still retain the convenience of simple coil-type equipment.

Fig. 144---Demagnetiz~ng Units for Operation an 60 Cycle A.C. a) For intermittent operation. b) Feed-through type for continuous operation.

The decaylng current method has been made simple to operate and 1s widely used. Built-ln means for automatically reducing the alternating current to zero, by the use of step-down switches, vari-


able transformers, or the newer saturable core reactors, makes this method easy to apply. When this method is used the current may be passed directly through the part instead of through the coil. This procedure is more effective on long circularly magnetized parts than the coil method, but does not overcome the lack of penetration due to skin effect, unless much lower frequencies than 60 cps. are used.

Another method of achieving "decaying current" demagnetizing is to draw'A.C. from a motor-driven generator. By shutting off the motor, the voltage of the generator, and therefore the current in the demagnetizer, drops to zero a s the generator dies down to a stop.

Fig. 14SMagnetirtng Unit with Automatic Current Reduction for Dernagnetizbng. Using a 30-Pomt Step Switch.

9. DEMAGNETIZING WITH D.C. There are several methods of demagnetizing with direct current. Although more effective, they are essentially identical in principle to the A.C. methods just described. By using reversing and decreasing D.C., lower frequency

reversals are possible: with resulting more complete penetration of even large cross sections. hlechanicai switching makes possible automatic reversals a t as low a frequency as desired. One reversal per second is a frequency commonly used. Thts method requires specially built equipment, but is most effective. It is one of the more successful means of removing circular fields, especially when the current is passed directly through the part, and will demagnetize even large sections. When using a coil in conjunction with this method, the part remains in the coil for the duration of the entire cycle.

A "trick" method for accomplishing demagnetization with D.C. involves the so-called "single shot" method. Reference to the hysteresis curve of Fig. 55, Chapter 5, shows that the residual field of point (b) corresponds to a coercive force of tc) on the minus H axis. By applying exactly the correct field to overcome this coercive force, the residual field in the part will be brought to the zero point on the loop. The method is not a very practical one however, because the amount of coercive force required is not readily obtainable for the varying kinds of steel encountered. The amount of field which must be applied to counteract exactly the coercive force must be arrived at experimentally by the cut-and-try method. Demag- netization by this means has been useful occasionally on large objects when other means have not been available.

lo. DEMAGNETIZING WITH OSCILLATING CURRENTS. The use of oscillating circuits is an attractive means for securing a reversing decaying current for demagnetizing purposes. By connecting a large capacitance of the correct value across the demagnetizing solenoid, the solenoid hecomes part of an oscillatory circuit. The solenoid is energized with D.C., and when the source of current is cut off, the resonant resistance-inductance-capacitance circuit oscillates a t its own resonant frequency, and the current gradually dies down to zero. The system works well in special cases, but is not adaptable to general use. This is because the size and character of the part affects the value of the inductance component of the resonant circuit, and any considerable variation in this factor will prevent the circuit from oscillating. The system is in use in connection with D.C. yokes and a circuit designed for a specific part to be demagnetized tsee also Section 11).

11. YOKE DEMAGNETIZATION. Yokes, either A.C. or D.C. provide a portable means for demagnetizing when other methods are

317

PRINCIPLES OF blAGXETIC PABTIC1.E TESTING

impracticable fo r tlie circ~imstances involved. I n some cases yokes a r e more effective than coil-type demagnetizers, because the fieid of the yoke can be concentrated into a relativeiy small area. Tlius, even some parts with a hlgh coercive force can be demagnetized in this concentrated field.

Fig. 146--Yoke for Demagnetizatmn with A.C.

Yokes for demagnetizing a r e usually C-shaped. The space between t he poles of the C should be such tha t the parts to be demagnetized \%rill pass between them a s snugly a s possibie. With A.C. flowlng in t he coil of the yoke, par t s are passed between the poles and withdrawn. The method is very effective, but is limited to comparatively small parts. Sometimes the yoke i s used on large parts for iocal demagnetization, by placing the poles of a U-shaped yoke on the surface, moving them around the area and then withdrawing the yoke while i t is still energized. This type is sometimes called a "growler" because of the noise i t makes \\.lien in contact with the part, due to the vibration set up by the alternating field. A modification of this method is a single pole A.C. electromagnet used in the same way. In general the single pole magnet is less effective than the yoke.

D.C. yokes a re similar in appearance to A.C. yokes, but use low- frequency reversing D.C. instead of A.C. for demagnetiz~ng purposes. This system is much more effective in penetrating larger cross-sections, and has the advantage, as, has the A.C. type, of concentrating a strong field in a small area.

In some models the method of damped oscillations is incorporated

in the desfgz for the purpose of obtaining tlie reversing and dimjn- ishlng fieid. B y using a resonant "CIR" circuit (capacitance, ni- cluctance, resisrancej tlie current oscillates and dies to zero with frequency and time dependent on characteristics of tlie circuit. This means of demagnetizing is, when used, designed for a specific part, since the part itself constitutes a portion of the magnetic circuit ~vhich IS directly related to tlie inductance. Variations in inductance, which parts of different slzes, shapes and compositions would In- troduce into the circuit, could destroy the oscillating characteristics on which this system of demagnetization depends. This being t he case, the usefulness of this system is limited to applications in w h ~ c h the parts to be demagnetized a r e closely similar in character and size.

D.C. yokes, properly designed for a specific job, probably provide the deepest penetration fo r demagnetizing of any of the possible methods.

12. DEMAGNETIZING WITH A.C. LOOPS. Portable units for maintenance ~nspection around a shop provide a flexible means for demagnetizing. A coil of any desired slze and number of turns can be nonnd from heavy flexible cable-00 or 0000, o r even larger-to suit the part. A.C. can be passed through the coil, which may be with- dramti xsiiile the current is flo~ving. or inay be left in place while the current is reduced to zero by one of the devices tha t have been described.

13. SohlE I-LELPFUL HINTS FOR DEMAGNETIZING. There a re a few "tneks of the trade" tha t can be helpful for demagnetizing certain s ~ z e s and shapes of parts, or for achieving the minimum level of remanent field. Some of these are:

( a ) When a short par t i s being demagnetized in an A.C. coil by the method of withdrawing the par t along the line of tlie axis of the coil, i t 1s helpful to rotate the par t both around axes parailcl to and transverse to the coil's axis. This should be done while the part is in the coil and during the entire time of withdrawal. The procedure is also effective in dernagnetiz- ing hollow o r cylindrical parts.

(b ) As a v a ~ i a n t of the above procedure, r~ng-shaped parts may be rolled through the coil, which helps achieve a lower level of remanent magnetization.


(c) A short part with an L/D ratio of one or less can sometimes be better demagnetized by placing it between two "poie pieces" of soft iron of similar diameter but longer than the part. This combination is then passed through the coil as a unit. I t has the effect of increasing the L/D ratio and facilitates the removal of the field in the part.

id) For the demagnetization of ring-shaped parts an effective method is to pass a central conductor through the ring. The central co.iductor is energized with A.C., and the current caused to decay to zero by means of either a step-down switch or a stepless current control. The latter method of decay can he much more rapid (down to a few seconds) than the step-down switch, which requires about 30 seconds to complete its cycle.

ie) The method of id) above can also be used with reversing, decaying or step-down D.C., instead of A.C.

( f ) For long parts, an A.C. coil with current on may be moved along the length of the part, and the part then withdrawn from the coil's influence.

(g) As a variant of ( f l move the coil on "high" current along the length of the part, then step-down by one of the current decay methods.

(h) For iarge hollow parts, a central conductor with high A.C. current is passed through and close to the wall of the part, and the part rotated 360°; then the A.C. is caused to decay to zero.

14. CHECKS FOR THE DEGREE O F DEMAGNETIZATION. Since demagnetizing methods vary widely in effectiveness as used on parts of different shapes and magnetic characteristics-hardness and coercive force-it is often important to check the success of the operation In some way. There is no effective way to check the degree of removal of c~rcular magnetism without damaglng the part by making a saw-cut into it. It Is, however, relatively easy to check the effectiveness of demagnetization in the case of longitudinal fields, since external poies are always present. The simplest device is the small hand-held field meter described in Chapter 8, Section 5 (Fig. 87). In use, the meter is brought close to the location of suspected residual polarity with the meter in a position normal to the surface, and the dial and pointer farthest from it.

320

CHAPTER 17 DEiVlAGNETIZATiON

If there is no reslduai field the needle will remain stationary. If there is a residual field the needle will move in a plus or mlnus direction depending on the polarity of the field. The amount of movement indicates the strength of the fieid. Although the dial is not calibrated quantitatively, the field meters supplied by Magnaflux Corporation do glve approximately quantitative readings. Full scale deflection (ten divisions) is equivalent to appfoxlmately 12 oersteds or gauss. At the low end of the scale, the value is about 1.5 eel-steds or gauss per division.

Some process specifications use this devlce to specify the degree of demagnetization deslred. These will often call for a maximum of less than two divisions-three oersteds-for critical parts; and iess than five divisions for parts in whlch a somewhat higher residual field can be tolerated.

If a field meter is not available a simple substitute is a short length of tag wire, or any light iron wire. A piece of sucli wire, 4 or 5 inches iong, can be balanced on the end of the finger, and when brought near a part having a local pole, the end of the wire will deflect toward the magnetized area. If the wire is hard and will itself retain a magnetic field (tag w ~ r e is usually hard from cold drawing) i t should be demagnetized just before use, so that its own remanent fieid may not be the cause of deflection toward the part being tested. The method is not quantitative of course, but an experienced operator can use it to judge quite closeiy the amount of remanent field.

15. CHOICE OF DEMAGNETIZING METHODS. Following are some of the factors that must be considered in the selection of a demagnetizing method. The several methods are listed with their broad advantages and limitations.

(1) A.C. Coil

f a ) Suitable for reiatively small and soft parts.

(b) Handles large numbers of parts in high production.

I C ) Maximum part diameter for thorough demagnetization, 1 to 2 inches.

(2) A.C. TILTOZL~IL Current

( a ) Worlcs well on relativeiy large andhard parts.

ib) Best suited for reiatively low production.

321

PRINCIPLES OF 3lAGNETIC PAIlTlCLE TESTING

ic) Works well with 30-point step-do\irn, infinitely variable current reduction or reactor decay methods.

i d ) Maximum part diameter fo r thorough demagnetization, up to about 3 inches.

(3) Reverstng D.C. Coil ~oith SO-Point Step-Dozcn. I?tfinitely Variable o r React07 C z ~ r ~ e n t Decay il.letlrods.

( a ) Worlis well on small par ts in standard coil.

(b) Fo r larger par ts use full length cabie wrap.

(4) Rezicrsing D.C., Thl.oz~g/l Cz~rrent-30-Point Step-Down, Infinitely Variable o r Reactor Current Decay Methods.

b) Best for iarge hard parts.

b Best for diameters greater than 3 inches.

( c ) Best for par ts that a r e of difficult size, shape or composition.

( d ) Works well with central conductor when applicable.

Table VII is a ready reference chart for these several demagnetizing methods, and the conditions with respect to size, hardness and production rate for which each is or is not suitable. Like the list above, this table is given a s a guide only, since there a r e no clean-cut lines of division betxveen, fo r example, "small", "medium" and "large" slzed parts, etc. The ciassifications should be helpful, however, in selecting the most promising demagnetizing method to try, basing its success or failure on check tests \\,it11 the field meter. Fo r those circumstances fo r which the given method is listed as not applicable, the ciassification is based on either lack of effectiveness o r impracticability.

16. DE~~IAGNETIZING WITH MAGNETIZING EQUIPAIENT. I n many cases demagnetization is a separate step performed on separate equlpment af ter the magnetizing and inspection operation 1s completed. Often, however, i t is convenient and possible to demagnetize on the same unit used for magnetizing. This 1s in generai true of a11 A.C. equlpment.

In its simplest form i t \vould include provisions only for the coil method-withdrawal of the par t from the coil's field. Equzpn~ent: i s also available mhic'h provides for decaying A.C., wluch makes possl-

322

TABLE VII

1 a ) Simple A.C. coil method, the A.C. b a n g provided to the coil especiall)~ for demagnetizing purposes.

(b) Decaying A.C. current method. In this case A C. is connected to either the heads or the cod, a s deslrcd, for de- magnetizrng purposes.

( c ) Reversrng D.C. step-down methods, available to either the heads o r coil.

When considering methods and equlpment, the possibility of using one unit for both purposes should be looked into, a s such an arrange-


merit may constitute a savlng of both time and effort fo r the demagnetizing operation. Often parts, which are moderately large and heavy, a re inspected while still in place on the unit in the magnetizing position. It is then quick and convenient if, af ter inspection, the p a n can be demagnetized without removing i t from this position.

17. EFFECT OF VIBRATION. In some magnetiz~ng operations, especially in the case of fairly large complex welded structures, a stabilizing or "soaking in" effect can be produced on the magnetic field by hammering or otherxvlse vibrating the par t during and af ter the application of the field. Complex assemblies of t h ~ s sort a re often very difficult to demagnetize due to the fact tha t different components a re of varying size and section, and take individual directions in their long axes. Vibration during denzagnetizatio?~ of such parts seems also to be a help in removlng stubborn points of residual fields. Yoke demagnetizers sometimes help remove local polarity from such assemblies, though more often they merely move the residual field from one location t o another.

The earth's field plays a par t in the difficulty of demagnetizing such structures. The part to be demagnetized should be placed so that i ts principal axis, or the axis of i t s principal, o r its longest member, is in an east and west direction. A long par t lying in a north and south direction can never be demagnetized below tlie level of the earth's field by any method. Rotating the par t or structure on its east-west axis while demagnetizing often helps reduce the field in transverse members which a re not lying east and west. Vibration of the structure durlng the demagnetization process 1s also helpfui under these circumstances. Complete removal of all fields from such a compiex structure is virtually impossible-but agaln it should be pointed out, complete removal i s seldom necessary.

CHAPTER 18

EQUIPMENT FOR MAGNETIC PARTICLE TESTING

1. HISTORY. The many forms of automatic and special-purpose equipment for conducting magnetic particle testing which a re berng built today are the result of thirty-seven years of evolution and growth in technology and in the practical requirements of the method. I n the days prior to the mid 1930's. when experience with the method was quite limited and its potential applications only partially visualized, the equipment that was used was small and quite simple. I t was des~gned fo r all-purpose use-a unit should magnetize the greatest possible variety of parts.

This all-purpose philosophy dominated the design and development of larger and more versatile equipment until the early 1940's when the requirements created by World War I1 calling for many new applications a t much higher production speeds, dictated the construction of many new equipment types and vanations to fit special uses. The demand for speed t o keep up with mass production rates caused a considerable number of automatic or seml-automatic special-purpose units t o be built. Then the true age of automatic equipment began in the early 1950's and continues today, and automatic processing of parts, from nliscellaneous small parts to massive steel billets and blooms, is now common practice in many industries.

I t is not the Intent to give in this Chapter a complete catalog of available equipment. The following sections a re Intended to give only a quick survey of tlie so-called "standard" or general-purpose types of magnetic particle testing equipment, anrl a few examples of some of the modern equipment now being built and used. The following Chapter, 19, will describe the evoiution, use and scope of automatic and "special-purpose" equipment, the problems gov- ernlng its design, and its advantages in many speclal applications.

2. NEED FOR EQUIPMENT. Magnetic particle testing is a process whie'h does not depend for i ts successfui application upon any particuiar plece of equipment a s do the X-ray, uitrason~c and eddy current methods of nondestructive test~ng. The proper magnetization of most parts calz be accomplished by very simpie means, and

I'IZIXCI?I,ES OF I_LCh'ETiC 1'AETICI.E TESTING

a t il'e outset the possible need for the large :?lid complicated equip-

ment beiiig used today '\,as not even vis11aiized.

31agnetic particle testing equlpnient serves tlVo basic purposes. whici~ dictate the sizes, sliapes and functions of ~iiodern units.

These are :

(1) to provltle cuiiz:r!?ria:rt msins for acconiplisl~ing pwpe?. magnetization. P,ro,urir magnetization with respect to field strength and direction, is, as tlie prevrous cliaplers of this book ii:ive siio\in, of preeininent importance. Crinrerue?~t means, i?roviding sufficient power of tile right sort, suitable conlacts and coils, me:ins for applying the magnetic particles and \\jell-lighted space for careful esam~nation of the processed part for indications, can only be ach~cved with equ~p- men1 des~gned Lo meet t h e ~ e requlrcmeuts for various types of parts and the conditions under wliicli they must be tested.

( 2 ) to make possible rallid testing of parts a t Iviiaterer speeds are required. \?itti assurance that the results will he reliable and reproducible. \Vith suitable equipment all parts can be tested under idcnticai conditions of contacts, current lypes and values, and techniques of particle application, when used In different locations and by diiferent operators. Such control bf the proeess becomes more and more important as the slze of the defects sought becomes smaller and finer; or as it becomes desirabie to finti defects over a certain nllnlmuin size, ignoring the very fine ones.

3. SInrPLE E Q U I P ~ ~ E N T . For occasional testing of small castings o r n ~ a c h ~ n e parts for surface cracks, for n-liich small, easily portable equipment is most convenient, and for the inspection of \velds, magnetic yokes are often xtlequate and very easy to use. Tliey are nbie to put a strong iield n ~ t o that portion of the part that lies be- t~veen the poles of the yoke. Either dry polvder or wet particles may be used. Yokes are available for eitlier A.C. or D.C., and in one model perinanent magnets a re used. The latter permits inspections where no source of electr~c current 1s z~oailable, or ts l~ere its use is not perniissible because of the lire or explosion iiazard in tlie area where the inspection milst be made. For longitudinal niagnetizalion of shafts and sp~ndles and similar articles, the portable iiits a re fuxn~shed w i t l ~ a fised coil covered \vitli a heavy protective rubber coating. Figure 147 illustrates sucii a tesling liit.

CA,\~'EP. 18

EQI!!?\IENT FOR Jf:lGNETlC ?,\Il?'!CI.E TESTING --

Fig. 147-Magnetiztng Yoke and Coil in a Portable Kit,

4. URGE PORTP.BLE EQUIPMENT. Lal-er portable equipment 18

lade for use where larger power requirements o r heavier duty

I t Fig. 14&Small hlqnetirlng Kit. Orlrmting from 120 volls A C

: Mounted on a Carry al l TrucK.


cycles make the small kits inadequate. One of the smallest of thls series is illustrated in Figure 148. It operates from 120 volts A.C., and delivers up to 700 amperes, either A.C. or half wave D.C. Dual prods for direct contact are prov~ded, and flexible cable for maklng loops for longitudinal magnetization are included.

Large units with outputs of up to 6000 amperes, delivering either A.C. or half wave D.C. are used for testing of castings, forgmgs or weldments, where such heavy currents are required. The units are equipped with current controi switches and with the 30-po~nt step- down switch for demagnetization; or, more commonly today, wlth the saturable core reactor decay system. Figure 149 shows such a unit being used to test truck components as a part of a planned maintenance program. This unit IS equipped with the saturabie- core reactor which provides stepless current control, both for magnetizing and demagnetizing.

Fig. 149-A Portable Magnetizing Unit Being Used for Inspecting Truck Components.

The most modern of the series of portable units i s the power pack shown in Fig. 150. This type of unit delivers from 4,000 to 6,000 amperes output of magnetizing current, either A.C. or half wave, with infinitely variable current control. In addition it has a self-regulating current control feature. The current "Dial-Amp"

CalPTtx IS EQUIPXENT FOR &lAGNETlC PARTLC1.E TESTiNG

Fig. 150-A Modern Portable Magnetizing Unit

stem automatically delivers the selected amount of current to the external magnetizing e~rcuit, and automatically compensates for variations in load ~mpedance and line voltage. Thus, the pre-selected current is continually maintained throughout the magnetizing cycle. Further optional remote dial controls peimit setting the desired mperage from a point 110 ft. or more away from the equipment.

5. STATIONARY MAGNETIZING EQUIPMENT. A large variety of tatiunary, bench-type units are available, with various characteris- cs to fit all different testing requirements. The smaller sizes are ed for small parts easily transpoited and handled on the unit by

and. The larger ones are used fur such h e p y parts as long diesel gine crankshafts, where handling must be by crane. Such units e made to deliver A.C. or D.C. with various types of current

Figure 151 sho~vs one of the smaller units (52 inches) of this pe. This unit is for use with the wet method, either with coiored

C!1APmI! 18

PRIXCIPLES OF ?JAGSTTIC PARTICLE TESTITJC. EQlllP%iVWf FOR >IAGS'ETIC P:\ILTlC!.E TESTING -

cent particles are used. Demagneiizing can be accomplished by means of one of the A.C. decay systems.

6. LARGE I-IEAVY-DUTY D.C. EQUIPMENT. At the top of the list of heavy-duty stationary equipment are tliose direct current units designed for application of the "over-all" method of magnetizing. for the insl~ection of very iarge, co~nplicated castings. 111 size these units are not large in comgarisoii with the steel billet testing units (Chapter 19) but in current output they ;Ire by f a r tile most power-

magnetic particles or the fluorescent type. The portable black light ful built. Rectified three phase A.C. is delivered 1i~it.11 current values IS seen mounted a t the left. The hood may be let down during black running as h1gi1 ;is 10.000 amperes. Such high curreilts a t e needed light inspection to secure the necessary exclusion of white light. to magnelize, a t one time, a11 entire casting wh~ch may weigh many Direct current up to 4000 amperes, derived from full wave rectified tons. Another feature of these units is that they deliver current, three phase A.C., is delivered to the adjustable contact heads. 100 separatciy anti m rapid succession, through three circuits, makiig 1nc1i units of t h ~ s t ~ p e will deliver up to 6000 amperes. A built-in possible the location of cracks in anjr direction in one operation. The coil is provided for longitudinal magnetization. This unit is equipped system is lino\\.n a s multi-directional magnetization. By means of with the infinitely variabie current control by means of a saturat~le eiectronieclia~~ical smitch~ng, demagnetization can be accomplisfled core reactor, and also with the self-regulating current control. by utilizing one circuit in conjunction with reversing D.C. and n

Figure 152 shows the largest of the bench-type statianary uiiits. 30-point step-down switch. It will handle nriy small or long heavy parts, such a s Diesel crank- The cost of these units is easily justified by the savtng in lime shafts, up to 182 inches (16 feet) 111 length, and up to 28 inches in and labor over the older point-by-point Inspection with prod con- diameter or "swing" Alter~iatiiig current up to 5,000 amperes is tacts. 111 addition. the system is much more reliable in the location

PIlINCIPLES OF MAGNETIC PARTICLE TESTING

AUTOMATIC AND SPECIAL HAGNETIC PAKTICLE TESTING EQUIPMENT

1. INTRODUCTION. The need for equipment u, do special testing tasks grew out of the application of magnetic particle testing methods to an increasing number and variety of parts, both as to size and to shape. As the use of the method expanded it quickly becnme apparent that the standard unit, intended for testing a large variety of parts, was not adequate to serve the testing needs vhich were being called for. Units to perform specific tests on a imited variety of parts were being demanded by industry.

~ i ~ . 153--20,000 Ampere D.C. Power Unit for Overall This was a netv philosophy, since previously it was felt by users Multi-Directional Magnetization. of the process that to justify its existence equipment must be as - .

versatile as possible. A single unit had been expected to accom- of surface and near-surface defects than the prod method, and often odate almost any type, size or shape of part that was brought to shows indications of some serious defects which the latter method for testing. hns missed. The unit is most often used with fluorescent mametic

~ - particles by the expendable technique. In the present chapter it is the intention to glve a comprehensive

7. UNIT VARIATIONS. A great number of variations of these view of special and automatic units. Such units are being used in typlcnl magnetizing units is available. These variations are in size, large numbers today in many applications. We will also analyze in current output and kinds of current, in the methods of current considerations i~ivolved in their design and use. Some of the modern control, and in numerous types of fittings to expedite magnetization special-purpose equipment will be described and discussed.

of odd-shaped parts. In addition there are many accessories, such as contact pads, automatic bath applicators, contact clamps, leech

2. DEFINITIONS. Magnetic particle testing equipment available today may be considered to fall into two major groups-standard

contacts, steady-rests for heavy shafts, prod contacts, speciai shaped coils, powder guns, etc.

8. DE~~AGNETIZING EQUIPMENT. Demagnetizing methods and Sta~tdard units are thosc which are designed to handle a variety equipment have been described in the preceding chapter, and in of sizes and shapes of parts, and are, as the name implies, made in connection with the various magnetizing units in which demag- quantity to a standard design. They may be stationary or portable, netizing features are embodied. When separate demagnetizing units for use with either dry or wet methods, and are used with little or are required-usually for the demagnetizing of large numbers of no modification for the testing of widely differing types of parts, in small parts-the A.C. coil is used. A large size of this type demag- a large variety of industries. netize; is shown in Figure 144. Chapter 1:. Coiis of this type are Spec?a.l units are those ivh~ch have been especially designed to made in a number of sizes, since demagnetizing efficiency is higher take care of some sitriation out of the normal, which standard units if the parts nearly fill the coil opening. In the case of special units, cannot, for one reason or another, handle. They may be speciai as speclal deniagnetizers are often built into the unit to suit the size to the metliod of magnetizing or particle application, or be designed and operating cycles required for the patticiiiar application.

332 333

PRI~C~I 'LES OF IrIAGXETiC I'Ai<'T:CLE TESTING

to handle an unusual size, shape, or nuniher of pcirts. They may or may not be autoniatic.

Special uriits can be furthcr broken down into two groups :

{ij Special-purpose, wlilch may be a ) single-purpose or b) geu- eral-purpose.

Speciat-pl~rvose units are those which are built to do a single testing job. T h ~ s special job may be a variatiori III magnetizing technique, in the way the magnetic p;trticles are applied, or in the way parts are handled. They may be silrglc-p~opose, in xvhich case they are to test a single type of part and no others, possibly by a special processing technique. Or they may be ge?~er.al-p?~rposc, in which case they are designed to apply a special magnetizing or processing technique to a (perhaps liniitcd) variety of parts.

Autmnatic units, on the other hand, are those in which part or all of the handling and processing steps are handled automatically. Either single- or general-purpose units may be partly or entirely automatic. Even standard onits, by addition of standard accessories, may be made automatic in some of their functions. The princip:ll purpose of automatic units is to speed up the entire haiidling- process~ng-inspection cycle. Tl i~s is accomplished through automation of one or more of the important steps involved in any given testing operations.

5. HISTORY OF SPECIAL UNIT DEVELOPMENT. In the early days of the development of the niagnetic particle method, magnetizing units were designed and built as needs for equipment developed. The first desigiis of what became stan(lard units were made for some special testing purpose for which, at the time, there were no suitable ex~sting des~gns. Thus we had small units and large units, vertical units and i~orizontal units, stationary units and portable units. Some \%,ere for A.C. and some for D.C. But each one was designed wit11 the idea that i t would handle a s many applications as possible, similar to the initial onc for wii~cli it was built.

The building of special-purpose and automatic equipment did not enter the picture until World War 11 created the need for faster inspection of mass-produced parts. A number of automatic units were built during tlie early 1940's for testing such items as projectiles, bolts and nuts, aircraft engine valve springs, etc. Special-

Clihm'rR 19

AUTO3IATIC AND SI'ECJAL EQUIP>lEXT

purpose units for handling la~-grr nnd heavier parts sucn as steel propeller Slildes, propeller hubs, engine cylinders and engine mounts, also came into being, some of them ar least partly auto- inntic.

Fig. 1-Automatic Unit for Testing Bolts and Similarly Shaped Parts.

Following tlie war the demand for equipment shifted from war materials to ~ndustrial production for civilian use. The railroads undertoolr to test all axles of locomotives and passenger and freight ears, which parts had been severely tased by war service. Figure 156 stzows a specla1 unit built in the late 1910's for railroad ear axle testing.

When the diesel revolution came, additional special unit types \Irere developed to handle critical parts of these locomotives. Crank- shilfts, connecting rods, pistons and various gears and eastiiigs were tested during manufacture anu later a t scheduled overhnut periods. Many of these pnt.ts mere tested on spec~al-purpose units.

and just af ter the war. The special purpose and the auton~at ic units a r e a iarge factor 111 accomplish!ng this end.

4. F;ZCTO~ DICTATING TJIE NEED I:OK SPEGJM, UNITS. The justification for a special u111t in any given inspection situation i s created t.y the presence o i one o r more of the folio\\-lng factors:

'{a) The need to employ a processing technique not available with standard units.

\ b ) The sue , shape or ~veight of the par t or par ts to be tested, wtien they a r e sucn a s to be not readily processed on standard e<ju~pnient..

'(c) The necessity of insuring absolute reliability and reproducibility of results, a t selected high or 1ozo sensitivity ievels, which end can be achieved by automation of the processing cycie.

''\id) The need to increase the output per man-hour of the testing unit, whic l~ can be accomplished by autonlation of the test-

:1 loll. lng oper t~

The ~ndividuai importance of these factors will vary a s t he service and production requ~rements fo r the par t to be tested xi l l vary. When the safety of the pul~l ic or ope~a t i ng personnel-as in aircraf t operation-are dependent on the failure-free performance of tile part , the need for absolute reliability (factor c ) \vould he the controlling consideration. Or automation (factor d ) may be needed in order to keep pace with production xvlien 100% in-process In- spection IS called for.

One generalization tha t can be made is this ; When "accept- reject" standards a r e to be c~sed and the rate of ]>reduction i s h ~ g h enough to use the fllll rapacity of a standard unit, automation of a t least the processing cycle 1s indicated. By this means the reproducibility of results will be achieved. \\:hich will perinit the application of standards of acceptance :inrl re~ect ion \uithout con- s~de ra t i o r~ of e:~cii case by the inspector. If, in addition, the r a t e of production i s vcl.?] high, fur ther :iutomation can be proiitabiy em- ployeti. 13y controlling process ~ a r i a i ~ i e s by automation; the hazard of error due to lhf iiunlari f ;~ctor is :remaidously reducer].

The special unit need ~ i o t al\v;lys be ccimp1ic:ltett or espensive. Sometimes a siteciai ;icccssory call be I~uiit 1'01. 2% st;untiard itnit and do tile .lob arleyuatei?. I11 othcl. c;ises, of course, \Then all processln.:

rid handling is autom:ltically cnntt~ollecl and sequenced, the unit is ecessarily more co~iipies and m:,i.e cosily. Fo r example. Fig. 159 sho\rs a spucia! unit for cra~tk-;hait In- rction ~vh:cli is n standard unit ctlu~ppcd ivit11 special fix:u~,cs to ocess crmtkshafts aut.on~:;iic:ill!.. Thls unit 1s q i ~ i t r ine:ipcnsive

.lien compared to the one siio\\:n in Fig. 160. This 1s also a specl:ll nit for crankihafts, hut 1s completely spcc~al , including sutorn;ilic recessing, eonveyor~zed par t handling and demagnetizing facili- es. I t 1s necessarily more complcx and costly tiinn the modified

tandard unit of Figure 169.

Fig. 159-Special Unit for Crankshaft Inspection of the Fixtured Standard Type.

5. SINGLE-PURPOSE A N D GENER,ZI.-PURPOSE UNITS. Whether or not they include any au tomat~c features, spec~ai-purpose units may

e built to ~ n e c t two different obiectires. If the testing job to be one is. for exampie, the inspection of crankshafts in a nlass pro-

&ion operation, the unit is designed to hiuidle cmiikshafts and thing else. This i s a st?igle-]~u,pos~? special unit. Other examples this type of application a r e the testing of b e a r ~ n g balls, rollers

d races; j f t e~igine conipressor bladcss; m~ss i le motor cases; rail- ad car axles; siecl billets and blooms, ctc. In these cases, the nature of thc equipment design 1s sucli that,

o iiandle properly the par ts fo r vvli~ch i t is intended nial<es ihe de-

CIIAPTER 12 PRINCIPLES OF AIAGNETIC PARTICLE TESTING AUTO>IATIC A?iD SPECIAL EQUIPhIEiVT

Fig. 160--Crankshaft Testing Unit. Speclaf Throughout.

sign t~nszriiabie for other parts. The reason for unsuitability ma be the processing technique or the method of part handling, or botl .

A special processing techiuque may be developed for a specific application, resulting in a single-purpose special unit. However, i :I wide number of essentially similar parts can he tested ily thi same technique, the special unit may be designed to handle th variety of parts involved. This, then, 1s a geaerai--),rirl)ose specia unit. An example is the equipment deslgned and built t o test larg castings by the overall method. (See Section 17, this Chapter.) TI units are special, but they xvill test iarge castings of many sizes an shapes. Other examples of general-purpose special applications are the testing of forgings, small castings and miscellaneous Ilia chined parts such as nllght be produced in a joh shop type of operation. For these applications the special unit design IS usually dictated by the part-handling requirements-size, \veight or production rate-rather than by processing tecimique.

When a special unit 1s lndicated for a given testing job, there is usually little question as to wl~etlicr it will be single- or generai-

Fig. 161-Single-Purpose Unit for Testing Bearing Rollers,

purpose in design. The testing need itself dictates the choice. The need also indicates where the emphasis should be placed in the design of the unit. The end to be achieved may be one or more of the following :

( a ) Insurance of adequate controlled sensitivity. a Ion. (b) Uniformity of test applic t'

(c ) Test production output, to beep pace with the production rate for the part.

td) Avoidance of potential damage to parts during clrcolal. magnetization.

le) Minimizing reliance on the human factor.

Tlie final design of sucll a unit may xvell require a compromise among sonie of tlicse objectives if more than one is mvoived. Ho\v- ever, such compromise must not go so f a r as to interfere with the satisfactory performance of the unit in fulfilling its primary reason for existence. Any sl~eclal unit design nzzut be bcchi~tcall?~ cnwect

I'RINCII'LES O F MAGNETIC PAIITICLE TE!;TTXG

Fig. 162-General-Purpose Unit for Testing Automotive Steerlng Parts

jol- t h ~ ' g121e11 al~plicatiult, and any design ccnlpromise must not violate# this requirement. Figure 161 i s a n example of a single- purpose unit fo r testing criticai rollers for Bearings. Figure 162 shotvs a general-purpose type unit- fo r tesling of a variety of automotive steering parts.

6. AUTO~IATIC EQUII'&IEST. Use nf autolnatic equipment 1s

usually called for :IS a means of lncreaslng L11c ou tp i~ t o i the testing operiitioi~. Tl:is end call be xccomplisheti : \(a) By n ~ a l i ~ n g the process cycle autom:~tic. Thls relieves the

operator o i most duties and shortens processing time. \ b ) By developing specmi processnig teduiiques w h ~ c h may re-

duce proccss~ng and inspection time. k c ) BY d ~ s i g i l l n ~ spccini means for handling tile size, volume,

1veig11t and s11:lpe of the par ts to be tested. In siJme applications automatic handling i a s in the case o i very ileavy par ts such a s stcci billets) IS essential to permit effective and ecanomrcally practicai inspcc t in~~ of the par t at all.

Some of the desices for titcre:i.il]ia the spceii of testing \viilcii a r e used in ~ 3 i IOUS oper:itlons nre tile follu\~ing:

( a ) Ifandling the 11;irt on a r:o!lrc:i,r ti:rokigh one or more o i t t ~ c inspt?ction :;te:)s ~n\-ol\.cd in the eniirc "tinor to floor" testing cycle.

f b ) Autonixt~c rotation o r n~anipuiiiiiou of the par t for v i t?~s i~~r ; . by the inspector.

( c ) Jlultirlireciionnl il)uovec:') ~napnetizarinn of the par t Lo reduce the number of proce.is~n(r 2nd i~~spect ion steps.

id ) TJse of speci:11 fixtures to simpliiy the task of positioi~lng parts ;lcculately f o r tlic reqii~red rnagneliz~ng upemtion.

,e) Automatic segrepatitrn of the inspected ~ a r t s , upon slg11:iI

by tlic inspcctr~r. into such eategoi.~es a s "good", "repair" or "sc2.a))"

7. ADVANT,XI:ES OF AUTOAIATIC EQUIPhlENT. AS hils J U S ~ bee11 s a ~ d , autornatic equlpnient :~ciueves a p r ima r j objcctise by incrras- ing the per-m:in-tiour output of the testing operation. IInn.ever. there a r e oLher irrlvanta.~cs \ \ * j~ i i l~ tile use of automatic eqtl~pment accoml,lishes, \sIi~ch in some ~nsinnccs ni:ly l ~ e thc m;ijor reason for the application.

The xpl:ear:lllce of tlie indic:ition of' a glvcn discontinuitj in :I par t is ;~lfectetl by the I 'o l lo \~~ng factors:

(a ) Streilgtl~ oC the 1e:ikn~e hcc).

( b ) 1Itlr:ttion of the existence of tile Iealage lield.

ic) Tile nlrmllcr of m:ignetic particles :ivailai]ie in the near ~ i c i n i t y of Lhe icnknge lielri, and t l~e l r f~.eerlom to more and fo1.m all in~Iic:~tion.

( 0 ) Se(111encji1g (if ille rn:rg~~cii%~ilg c i~ r r en t s h ~ t itnd the app l i c~ - tic111 of tlie nlagiiciic pnrtlcles.

If the inspector 1s to ~ u d g e the severity of :I given ciiscontit~uity, he must do so o ~ i the 1x1~1s of' ilic cipl~c:irance of thca inclieation. Thcrefure i t 1s imliort:uit 111i11 the yruecsslng o l all p;irts he alike and that no rarl:~ti<ni in the appeal.>illce or i l ~ e indic:it~oiis Ltc lliie to vart;itiorl.s in groctLsslng. ~~i;l)cc~:~ll!. I I I respect to f;iciorr 11, c and d : a b o v ~ .

:;-I:{

I'BINCIPLES OF I1IAG.hETIC PARTICLE TESTING :%VTO>IATIC AND SPECIAL EQUIPXENT

Automatic eilulpment insures that the i~idications truly reflect d) Define locations and directions of suspected defects. the severity of the discontinuity in tile part, in that factors b, c, aild e) Define the pro~li~ction requiremenls-that is, the required d a r e automatically malniained uniform, and a r e not dependent on rate of testing-for the special unit, anticipating some dowil- the human factor. tirile for the unit anu some percentage of rejected parts.

One reason for favoring the res~dua l wet method i i ~ automatic iith this inforniation the designer is ready to proceed to a con- units on such materi:tls 1i~Iiic11 have sufiicient reteiitivity, is tha t the ration of actual details of design. two factors, b and e, a r e niore easily controlled an* d is riot a factor a t all. If the short-shot conliiiuous mcthod must be used because of 0. THE SECOND STEP. CONSIDEPATION OF METHOD FACTORS. the magnetic charnctcristics of the part, great care 111ust be ex- s a seconrl step the cho~ces among the various method variations ercised in the tiesign of t he control of the current-bath srLquence. st be considered, and dec~sions made a s to how the test is to be

formed. These considerations can be broken down into separate 8. STEPS t 2 E ~ D t N ~ TO THE DESXGN OF AN AUTOMATIC OR FPECIAL- enis and esamlned sepiirately.

PURPOSE UNIT. The automatic o r spec~al-putpose unit ?,~?isi be de- slgried t o be tecliii~ccilly correct in all aspects of tile processing a ~ ~ d . Method r,J App/icntioit of t h e illagitctic Pa~ticics . This is a first

handling of the par ts to be tested. This 1s assumed by tlie user, and consideration, and the elloice lies bchireen the wet and tlie d ry must be a first coi~s~derat ion of the desigiiel: Therefore, in under- taliing the ileslgn of a specla1 uiiit to solve a given inspection ],rob- The met metliod is used almost exciusively in special units lem, all the factors afl'ccting the pro~ected .design must be carefully fo r nunieroos reasons. Some of these a re : examiner1 and itnalyzed in advance. An orderly progresslorl tlirough ( 1 ) Ease of applic a t ' [on. these factors and f i ~ ~ d i n g satisfactory solutions for each olie ill tur11,

( 2 ) Speed of application. can result in a sgcciai unit rles~gn wlilch xsill produce tile restilts called for. (3) Assurance of coinplete coverage of the surface of tlie part .

( 4 ) Greater sensitivity for most sought-for surface defects. 9. T H E FIRST STEP, ANALYSIS OF THE PROBLEM. Tlie first step (.i) Easier sequencing of particle application In an automatic

i s a broad eram~natioii of the psobleni; to determii~e the o\,etnll arocessi~ig cycle. contiitions arid requirements to be nlet and achieved. To do this i t is ( G ) E a s ~ e r re-capture arid return of excess particles to the necessal'y to : circulating system for re-use.

ta) Untierst:rii(l thoroughly the purpose of tlie proposed testing program.

113) Defi~ie the part or par ts mvoivecl in te ims of : (1) Size. (2) Slial~e. (3) M'eiglit. ('1) S u ~ i a c e finish. f6) >fagnetic characteristics. (6) Heat treatment. (7) Operations prior to and follo\\:ing tiie proposcd

testing.

l c ) Define the scrvicc rc(juiremeilts for tile part.

All these considerations have to do with method reliability, :~nd the econom~cs of the process.

1,Vafer os Petro1euo~-Ease Ve111cle fo r the Wet Method Eatlrs. One of the reasons that water was first' prol~osed as a medium for the \vet method bath was to elinilnate the fire-hazard :issumetl to be ~iilierent with petroleuni-base liquids. While the ~ncidence of lire in equipment using petrolctim l iqu~ds has been very Iols over the yesirs, tlus i s one of tlie consideratioiis in m a k ~ n g a cI101ce.

For esampie, the inspection of billets and blooms is an area in which tlie fire hazard is very real if petroleom-based baths a r e uscil. In m;111y cases bloom and billet ends a r e covered with a licavy nuii-conductive scale. Specla1 magnctiznig circuitry

PRIIiCIPLES OF 3f:lGXETIC PAIWICLE TESTINC --

has bee11 devised to b u m through this scale to make eontsct, b deliberate arcnlg at the eontrtct points. This arclng could easil ignite a ilamniable bath, arid because of tllis viater is used exclus~vely in these units.

Rut there a r e other reasoils fo r the use of water a s a bath, w h ~ c h ~nvo l r e the econoniics of the process, especially where large \-oiomes of bath m e required. a s in billet testing. In a typical ~nsrallation of this sort, 50 to SO gnllons uf bath must be added per eight hour day to replace tha t lost by "carry-off Water 1s esseritially "free". xvhereas a n oil n~etliuiii ntay co -10e per gallon.

011 tlie otlier side of the economic ledger, ho\vever, i s the fac t ha t mater baths, iri a recovery system, tend to foam. T111s foaming is caused by additives to improve the ability of the bath to wet completely the surfaces of tlie par t t o be tested. In some cases i t i s necessary to IIBE other additions to control lhis foaming tendency. These additions increase the cost of particle concentrates fo r use \vith \rrater. I-lowever, costs of the partici concentrates fo r water and oil a r e comparable, so there is stil the advantnge of tile difference bet~veen the cost of oil and water.

A t first thouglit, the espendabie techniclue would appear to be ~sas te fu l and costly. I-lowever, fo r certain applications i t can well be ;I iess costly tcchn~que when the bath mediuni is mater. Fo r example. \?hen applied to the testing of large castillgs using the overall method, i ts use obsiates the need for a large bath reservoir and a filtering arid rccirculating recovery system. Water also presents less of a disl)osai problem than does , petroleum based bath.

C , C o l w of Pnrticles Useci. The choice here is based on the particular color of magnetic particie which prot~rdes the best average contrast with the surface of the parts. Eecause of their greater visibility, Ruorescent magnetic particles a r e f a r more ~ ~ t d e l y used in specirll units than a r e tile visible colors, xvet o r dry. In- dications delirieatect by flool.escent p:lrticies ate much more clearly and qu~ckly seen than a r e those i~idicatecl by daylight visible particles. Ho\xrevcr, there a r e sonic few cases \shere daylight-v~sible particles ulTcr a n advantage. This is particularly the ease \\-hen tlie soi~glit-fur defects give poorly defined or diffuse fields. This wouid occur a t sul>surfnce de f ec~s o r a t

Palriy shnlio\s and o p ~ i i sul.S:lce tlcfec;~. i n sucll cines ciayliglit ~ S ~ I ~ I I I X ; L L I O I I 1101 i~lilv prc~t~ides better til>st.rvation of tllc indication. LUL aiso j~erniits ii I T I I I I . ~ t l i ~ r i ~ u g h e\.ali~i~tion of tile tieit!tt t l ~ a r ~ can tic rn:i<le in 3 cli'rkencd rowm.

T i i 1 1 I . Thc c11o1cc of currcnis osu;illy lies among four types-alterilatii~p ci~r i .e~i t , reciifieri :1 piiase A.C. direct curri!nt, lralf \v; i~c rc:ctilie~l s~irglc 1di:isc A.C., and full xmr3e rectified single ph:is<> .\.(I

Since must :iulom:ltic :iiicl sper ia l -p~~rpnse eqoipiiie~>t ir~\~oivvs tlie detection o i surface iiefects, any of tlic four types of current a r e soitai~le. Tiiu ciioice Ina). t11creCnre be bascci on considerations like tile fol low~ng:

( a ) Consistent? with tlie type of n~agnetizing cuyrent used in prior or subsetluent testing of the same part .

(b ) Consistency with the type o i magnet iz~ng current use11 11,

testing similar parts.

( c ) Stand:~rdization \r.ith the type of niagnctizing curtent preferre11 in a particular company or plant.

(d ) L~mit i i ig the current drawn by the magnetizing circuit froni the power lines. A.C. and half ware D.C. operate from a single pliase A.C. line input. Tliese dm\\. more current per leg and tend to unt>alance the sliop distribution system. D.C. froni rectified 3 phase .4.C. does not preseni t111s difficnltj~.

( e ) D.C. is easipr to control for conveyorizcrl coil-maguetizzt- tion of parts. The timing of ~iiugnetizing c u r ~ e n t cut-off is less critical isitli D.C. tllaii with A.C. due to tlie demagnetizing effect of the latter.

( f l When :I "fast break" 1s re<run.e~l, full wave rectified A.C. IS the 11:ost. ~ R e c t i s ~ ! t y l ~ e of circuit to use.

I n the case of multi-directional magnctiz:ition, different types of magnetizing current may be used in tire diiferent circuits, for the same part. A.C. o r half ware D.C. for the head shot is often combined with s~ng i e phase itdl \vase rectified D.C. in the coil shot. The reason for t h ~ s is the critical requirement for "iast break" for tlie coil shot is better achieved mit11 D.C.

CIilPTER 19

PRINCIPLES OP WAGXETIC PARTICLE TESTING )MATIC AND SPECIAI, 12(IUII'.\IENT -

Directio?~ of Mag?~eti.:tng Cz~n.e?lt. Tliis choice is determined b, ie unit design inust be first one that provides the most efFective the most likely direction of the sourht-for defects. There a r nd sensitive technique for the defects sought. It must also lnsure -

a ) Circular magnetization. b) Longitudinal magnetization.

lc) Rlulti-directional magnetization.

t t h ~ s technique is properly applied. Rate of output of the unit Id in this case be of secondary importance. Alternatively, if

e application is an in-process testing to up-grade the quality of arts later to be sub~ected to final inspection, a co~npromise in nsitivity may be permissible in favor of maintaininr hirh pro- . - .

If both circular and longitudinal magzietization are calle for, the multidireclio~~al system shouid be considered and ma 12. E X A ~ ~ P L E S OF SPECIAL UNIT APPLICATIONS. From the pre- often be applicable. If usabie, i t will shorten totai lnspectio ceding discussion it might be assumed that special equipment is pri- time. I t often may not be useable however, because such factor marily cles~gned for testing parts that might be easily handled as shape, surface finish and magnetic characteristics, ind' manually, in order to insure better quality or increase production. vidually or in combination, may ruie it out. It is no doubt true that the majority of parts tested on special equip-

The manner of applying the magnetic fields selected 1 ent do fall into this category. However, there are many parts subject to numerous consideratioils dictated by sensitivi side this group that involve such large sizes, heavy weights: com- qulrements and part characteristics. - configurations or unique test requirements, that they could

be tested a t all without special equipment. F . Contin~sous 9 s . Restdual Rfethods. Since the residuai inetho

application IS easier to controi, i t is preferred where th n the remaining sections of this chapter we will describe five sensitivity requirements and the part characteristics per the use of special units was the only way to apply There are some cases where a combrnation of continuous iag7ietic particle testing on any practical or ecoilomically feasible residual methods is employed. This involves the continuous application of magnetic particles before, during and after e 13. TESTING BE~RING BALLS AND RACES. For many years posure of the part to the magnetizing force. Such a combinatio arclened steel bearing balls resisted a11 efforts to test them with favors s~mplicity in the design of equipment. The sensitivity lev agnetic particles to locate cracks and other metallic flaws. Dewsing of this technique is likely somewhere between that produced b a speciai magnetizing technique, and the subsequent design of an the continuous and residual methods separately. However, i tomatic unit for applying it, finally solvcd the problem. effectiveness in any particular case sl~ould be carefully esta lislied, as it inay be affected by such factors as surface finis The usuai two-step head shot and coil shot teci~nique did not manner of applying the bath and the magnctic propel.ties of th a isfactorily disclose all .defects in whatever direction, for two part.

(a) A ball's LID ratio is sucll that a coil shot is ineffective. 11. THIRD STEP. FINAL DESIGN SPECIFICATION. Having react)

decistons with respect to each of the above outlined consideratio tb) it'ithout marking each ball there is no way to orlent the ball the designer 1s ready to make up his final design specification. properly for the three head shots that would be necessary to frequently happens, however, that some of the o ~ t i m u n ~ desira disclose all defects. features are in conflict anrl compromises must be made. A nletliod was finally worked out wherein the ball was treated as

In making such compromises care must be taken that they are n cube, and sepavateiy processed and inspected, using circular made in the area of tecluiical accuracy or in insuring that the prin agnetization, through its X, Y and Z axes, respectively. Controlled ob~ective of the spec~al unit is accomplislled. For example, whe tation of the ball between each of the three processing and in- the ultimate in test sensitivity is required, as for aircraft part pection positions assured proper orientation and therefore a re-

C H A I ~ T ~ 19 PRINCIPLES OF 31AGP;ETIC I',\IIT?Cl.E TESTIKG .~I:TO:,I.%TIC .AYU SPECI;\L EC!JII'>IENT ---~-

Fig. 163-Model Showing the X, Y and Z axes of a Bearing Ball.

1i:ihle overall test. Suhsequenl de.ii&ps permitted magnetizntion without any eiectrical contact with thc ball fusing a variation of the induced current metl~od) t i ~ u s e1imin:iting :uiy possibie danger of damage to the ball -due io arcing a t the points of niagnetiz~rig cnrrent contact. Figure l G 1 illustl'aies this unit.

Another interesting esa~npic is the testing of bearmg races. Althnugh t h ~ s ring-shaped part Iind long Ibecn tested on standard etlulprnent, i t was a slow t)rocess, t l ~ r f e inspections being required. Also, the need to make contact on the li:lrdened, highly ~~olis l ied surface of tlic race, resulterl in a certnin amount of burnnig, mh~ch In some ~nst;itices runiecl the race.

Defects lying on :iny of tlte surf:icc.s i n a dircction transverse to the circumference of tlie ring can be easily located by circular magnetization using a eentrzil conductor. Circumferential defects, parallel to the ball path, a r c a different matter entirciy. Fieid in the ProI)cr direction can be induced by a head shot :>cross the diameter of the ring. But i n ttie area of tlie contact points the current splits and follows arilund to the olher contact in t\ru portions. A t the r~oints o i sep:ir:~tion of the cnrrent the field is clistorted so tha t

reliable inspection of those areas cannot be had. Consequently the ring must be turned 90" ant1 given another shot, and relnspecterl. In adtiition to the time consumed t h ~ s method ha s two major objections:

[ a ) The possibility of damage due to arcing o r contact resistance l~eat ing a t p o ~ n t s of eontilet.

( b ) The possibility of distortion of the race due to excessive c lamp~ng pressure.

Figiire 16.5 is a diagr:~n~ of the current ant1 field i n a bearing race when being magnetized circuinrly by giving it a head siiot.

The cont:~ct damage problem was solved by tlie use of the incluced current metlioci using :in auiom;itic spr?cial-purpose unit. Figure 1GG sho\vs tlic cul'rent and field ilistribution tr'ilen the race is mag- netizetl by the in~'iucetl current metl~od. This techn~que provides a complete :~nd reliable test mitliout mak:ng contael on tlic race, anri

CHAPTER 19 PRINCIPLES OF nlAGNETIC PARTICLE TESTIXG

rovtdes fo r magnetization by central conductor, inspection, mag- etization by induced current, ~nspection and, finally, demagnetiza- ion. A similar technique has also been applied to c e r t a ~ n criticai c t engine pa rk , both during manufacture and a t overllaui.

14. BILLM TESTING. Billet conditioning in the past has been onducted by visual examination (wluch usually required pickling)

f o r seams, wliich mere then ch~pped, ground or flame scarfed out. The system was f a r from satisfactory, because the inspectors mlssed many seams entirely, and removed much good metal where no seams or only very shallow seams ex~sted. Scaie I-emoval, by pickling or g r i t blasting, improvcs t t l iss i tuat ion somewhat, but is a costly procedure. The economics of billet testing and conditioning i s a complex subject, but i t 1s easy to appreciate the advantage of eliminating the pickling and of being able lo remove all tliose seams, and oniy those seams, which would be detrimental t o the finisheti

Nagnetic particle testing systems have, in tlle past ten years, been developed svhich accurateiy locate s~gr~if icant seams and indicate

their approx~mate depth. Properly designed eqtiipment, circtlla m:ignetization, the conlinuo~is methoci and xater-based fluorescen magnetic particles p rwide tlie means for ilolng tills. Billets m:iy 1 classifietl without. the need for scale renioval. and segregated, usual! into three groom nccording to their degree of seiiniiness. The.. groups are iisuaily selected w i t h ~ n the following ranges:

( a ) P?i?iie. No seams deeper than a pre-deierm~neii ievei, usual1 0.C121J t o 0.030 inch. These hillets a r e shipped o r fnrtiicr ~~ roce+s rd "as is". The very shallo~t. seaiiis \\-ill "scnle out" ~r "\\,ash out" on reheating.

( b ) Contlitio,~. Seams deeper than the abovc level, bnt siiallo\ve than a level establislied as too deep for economical remove call for surface conditioning. Seams of tills group may b from 0.050 l o 0.070 in mnximurn depth.

( c j Dorrvz--yraile. Billets with seams deeper than 0.070 w e di- verted and used for lower grade products, o r scrapped.

-- &+= ,: ,, ::;...s.* .y-*zz:w;.:T&*-s L2...J sky@:::.-.. :..> ? 3 Y , ?&$ $~$L$z;~&T2?2z<r~;;:@ ; j; - .,w,...%.. :.?s~::.-='+*.&: 'u&*

'5i-p~*-:.i . -- .. ?*> ',i, ,- .<.3- ..;. $:I. -=:..Gi* . ..: :t..i.:.,c3T&=I$ %.R<. -a%-- ..%- .. .A*, , = * -,. <<-. ., ,. . ~. ,* .* ?,$-".:.y .,, . ,37..< :,T,:..$ ..? :,.. :<>;:F;!t;:?$?j

,.T ;. 2 . . .3 ><, :-:a&>. .,. , ,. . . * ,; - .-*,. -;.<: . . &--.,., ;.- *.*::22:.,5

$,:.,>::.>;::--; .,:- .-.> Ly, &p*, *.:* * j , ~ . . ~ . . - . ; . : . , .-.-.... ;,., .g:$%;$

.?.. ,.*. -.:q-: :.?&q'.,. <.?< sap . p&$,$b-w,..i I?," - . +'*... ,@,+pjL$,;: J . . . , . FF .b+.-j::&?&*&*&&z& Ilr. a s ,

'g. 168-Fluorescent Magnetic Particle Indications of Seams of Various 15. SEAM DEPTH DIscRI5f1XATIox. Eillet testing with magnetic Depths, Shown by the Resldual and the Continuous Methods.

particles requires that the inspectors jiidgc the seam deptli so a s mark tlie billets properiy anri classify them a s to quality. This a satisfactory separation of the billets into the three levels of made possible by con t ro i i~~g the entire processing cycie, s., tha t the fluorescent Lrilliaiice of the seam ~rlrlicatiou reflects i ts ;tpprox~niate eam depth d i sc r~m~nat ion 1s made feasible @ the foiloa~ing: depth.

i a ) Specially developecl magnetic particles, deliberately insenst- IVitIi equipn~clit properly designed, operated anii ma~nla ined , i t tivc to weak leakage fields a t very fine seams.

is e;~siiy possible to discrimmate with a tugh degree of accuracy (b) Spec~ally developed ions fo;iming wetting agent, permitting between tlie intlic:~tions formed by seams varying ni dc l~ th by ap- practi'cal use of water a s tile suspending medium, elinliilating prox~m;tteiy a two to one ratio. This does not rneali thiit magnetic the problem of excessive foaming of the bath. particle testing can be em~iloyed for seam depth deter7nlnntio?~. In (c) Specially des~gned electl-ical magnetizing circuitry to pro- testing billets we a r e looicing a t indic:~tions of seams only on the vide a high enough voltage to a r c throogii the scaie on the basis or ~ela t . i z !e depth. Fo r example, ;ire they shallo\ver than 0.025 billet ends, then quickly drop back t o tha t requlred for the ~ n c h ; over 0.025 ~ n c h but under 0.070 incii; o r over 0.070 inch? We desired pre-set magnetizing-amperage lc\'cl. do not care what the nctircrl seain tlepth is-\ve care only about whicii

This arcing de\.icc is made necessary because three phase of the three classes i t falls into. Extrernc accuracy :it the dividing full wave D.C. is used for billet testing, often a t current line betweeri the clz~sses is not required. densities a s iow a s 200 amperes per inch of cross-section.

Control of the bath concentration and the ~ n a g n c t i z i ~ ~ g curre!it Thus the m;rgnetizing voltage reqiiired t o prodlice this cur- level can \ryeaiten intlic:%tions of si~allo\\~e!- sc;~rns so tha t they a r c rent is frequently too iow to break thr0ug.l~ the niill scale sc;ircely seen. The real determ~natioli to be made, tliercfore, is on hot-salved or hot-sheared billet ends in order to mdte whether a seam is untler o r over the 0.0'70 i~ ich level user1 in the example above. The method does result in classification of i~illets

355

t CHAPTER 19

I'RINCIPLES O F 3IAGNETIC I'AIiT1CI.E TliSTlh'G SPECIAL EQUIPJIENT 1 i

(d) Prec~se seqocnclng of bath applicatimi cut-off and magnetiz- termined auton~ntically based upon the poorest grade any one ing current cut-off. Both are "on" for long time periods mspector chooses. severai seconds-compared to the fractional seconds usua

( f ) Billet transfer beLween proccssi~~g and i~andling during the associated wilh wet inethod particle testing. v~ewing cycle must be such that indications are not smeared,

ie) Efrectii~e separation of n~il l scale from the bath susnensio nor false iiidicntions transferred onto the billet's surface. to preserve bath quality.

ff! Suitable black light type mld placement, to PI-ovicle ailequate black light i n t cns i l~ oil the l~illet surface, without. restrictitlg the rnspector's vlston or freedom to mark sexms.

In short, the method is de-sensilized, and the sensitivity level is then controllcd along with all the otlier method variables, so that reliribility ant1 repeatability of results are assured.

16. DESIGN CONSIDERATIONS, OTHER TAAN &IETIIOD FACTOR FOR BILLET TESTING. In additio~l to the testing method, design con siderations inrolvi~lg billet handling and mill operating conditions create problems not encountered with tlie average special unit. For erainple:

ia ) An~ltient temperatures in the billet testing and contlitioni areas may range from -20" to 120" F. T h ~ s calls for spec attention to operator and inspector comfort iand therefore effectiveness) and avo~ding of freezing of the water balh recil-cnlating system, and of the bath on the surface of the billet d u r ~ n g testing.

. - -. . . . .. . . . . . , ,. . . - . . - Fie. 163--A Tvnrcai Larae Billet Testinr! Unit. Inset: Ci0se.u~ of . . - -

tnspection Station Showtng Billet on Chain Sling Billet Turner. (b) The design must antic~pate the billets b a n g bowed or tnristed,

for they often are. 17. INSPECTION OF LARGE CASTINGS. The large castings which (c) Billel transfer from station to station must be rapid and justify the use of special equipment are those intended for critical

positive, yet cush~oned to avoid excessive punishment to the serv~ce conditions, and therefore require exeeptioiially thorough equipment. testing. They may weigh from a few hundred pounds up to severai

( d ) Delivery of a processed billet to, and tliscl~arge from the tons. Conventionally such castings, too large to be handled on ii

viewmg station must be accomplisl~eil rapidly so a s to allow standard unit, are tested using prods and expendable dry powder.

maximum vle\irtng time \vitIl maximum safety of the in- Although this technique can be very efTective there a re severai draw-

spectors. backs to its use in production. Some of these are:

f a ) Great care must be taken that prod contacts are made in-a (e l Since t\rro or more inspectors view and judge parls of the regular pattern to make sure that the entire area io be ~ n - same billet, each may judge his section t o be bette? or worse

than the sections of the other. Therefore, electrical circuitry specter1 has been properly covered.

is incorporated to permit each inspector to judge the billet (b) I t is time-consuming. a s he sees hrs section. Classification of the billet is then de- 1 c) Some casting deslgns do not permit proper prod magiietiza-

PRINCIPLES OF MAGXETIC P;IItT!CI;E TCSl'Ih'G

tion in critic;ll ereas due to difiiculty In making contact a t tlie right points because of t!le sliape of the casting.

td) Arc burns from prod coniacts a re sometimes damaging.

le) Usually prod inspect~on is a trvo man operalion.

The special equ~pment most frequently used for large casting inspection- is a unit composed of a f11g11 amperage, high duty cycle, 3-phase full wave D.C. multi-direc.tiona1 ponrer pack; a system for spray-gun appticalio~l of expendable fluorescent m a g ~ ~ e t i c particles; and the necessary cables, contact clamps and black lights. The enlire casting may be magnetized at one time. ?'his method 1s called the "overall" magnetizing method.

Fig. 1 7 G H e a v y Duty Equipment for Applying the Overall Magnetization Method to a Complex Casting.

The magnetizing current is sl\pitciied rapldly and sequentially into t\vo or three separate circuits. I-Ieavy cables a re wrapped around, threarled through or clamped to the casiing and a separate set of coils or contacts is used for each hlgh amperage circuit.

lvnscscent magnetic par?iclc hnth IS used. anrl is f i o ~ c ~ l 'iver. tlie siing while cnr~.i.n! 1;; "<in:" tovesing all areas to he 111spe~:t~d. his t.ech:iiau? overcomes. a t 1s:ist in large 11ie:1s11rt?, the five dr.:~\v- .eks to the prod a n d dry po\i~ies Lcci!ri~qu~? lisie(1 :ibove. The ad- antages achlevcrl tbr.oilFh tlie use of the overall method are so brqous that it is easy to sce the ji~stificiition for the cost of the in-

Ilntioii. Proccssi~ir. and nispection times linve often teen reduced one qunrler or. less of tlioutl s'qu~sed for the pr.ud magr~etization

and pomdrr method. In one :tpplication the time rvns cut to 10::; The over:111 metlioei ciin ::is0 be :~ppiied through the use o i spec1a1

magnetizlnx fixtures or i~ ; ind l i~~g devices. \\'iierc the ~nspectlon operation is such that the range of casting sizes afid shapes is not too great, a spec~ai magnetizi~ig unit and handling fisture using the overall method is often employed. Figures 170 and 171 each illus-

Fig. 171-Overall System ppllecl to a Miscellaneous Assortment of Castings. t 18. INSPECTION OF WELDED STEEL MISSILE B10TOtl CASES. hlissile

motor ease inspection poses a ~lnjque challenge to the design engineer, bec:iuse oftcn it is necessary to devlse unconventional magnetizing methoils to secure llic necessary sensitivity for this most critical ~nspection. Tlus 1s an excellent es;lmple t n illustrate the great ad:apt:thility of ii~ag~li'tic particle testing. In reading thts book thc impression may !lare been glven that 11ard and Enst ruies govern the proper use oi the method, but this is not necessarily

CHAPTER 19 PRINCIPLES OF MAGNETIC PAllTICLE TESTING AUTO&I.ATlC AND SPECIAL EQCIPSIENT

the case, as denionstmted by lhe soiutio~i of the missile motor case and coil magnetizing. techniques can sometimes be used for tlie smaller diameter cases of simple deslgn, but for tile larger sues and more complex configurations special techniques must he employed. The highest attainable strengtlt-to-\\,eight ratios are sought for

all tile materials and components used in the design and construc- Ordinary coil and central conductor tecliiiiques are not practicable tioii of missiles. Desi~mers are forced to use lower safety factors for several reasons: than they would for more ordinary products. To achteve t h ~ s result (a) The motor ease clin~ensions are such that unreasonably hip11

1

I they rely mainiy on three th~ngs: magnetizing currents \vould be requ~red.

f a ) Use of the best and most advanced materials for their pur. [b) Its configuration IS such that vital areas would be shielded pose. from tlie magnet~zing force, or tlie magnetic field would be

distolted, resulting in a loss of reliability, to an unknown f bi Closely controlled manufacturing metliocls. degree, for the location of all defects. Field direction and dis- (c) Use of all mailable methods of nondestructive testing. &lag- tribution in a complex object such as a missile motor case is

netic particle testing is only one of the methods involved. ~mpossihie to predict wit11 certa~nty. (See Chapter 10.)

ic) Longitudinal magnetization wit11 a coil, agaln because of the Motor cases are pressure vessels, subject to extremely high pres- size and shape of the motor case, would be ineffective. In sures when fired. They are also as thin-walled as a reasonably

minimum safety factor will permit. Therefore the most minute defect is cause for concern. And sznee they are stressed both longitudinally and circumferentially, defects in any direction are significant.

Some of the major design considerations for equ~pment to be used for testing missile motor cases are the follow~np:

(a) Fluorescent magnetic particles sliould i ~ e used, for maxlmum sensitivity to surface cracks.

fb) 100% surface inspection sliould be provided fol-both inside and outsid-for defects \vhich may occur in any direction.

(c) Magnetization should be accomplished without current contacts being made directly to the surface of the motor case. Arc burns from electrical contacts would themselves constitute defects, and cannot be tolerated.

t d) Automatic control of the entire processing cycle is deslrabie 1- to insure reliability and relwoduceability of test ~esults. q ~ : ~ . : tei Handling and processing equipment design should bc such ;. , . that i t insures against any form of damage to the motor case

Q I during the entire testing process. 8,. Figure 172 illustrates, schematically, a Polaris motor case and Bi A large number of units have been built for motor ease testing, tlie means used to provide effective circular magnetization. Thls

4 including sonie for the Titan 111. These sections were 10 feet in "central conductor", \vIiic11 is actually a shaped single turn coil, k, ! diameter and up Lo 11 feel in length. Stantlard centrai co~tductor closciy conforn~s to tile mside profile of the case. Due to its close c : 1; $ :! 360 561

,, J ,, .- a,.,

PEINCIPLES OF XAGNETIC PARTICLE TESTING

positio~~ing to the %%-z11 of the case, much stronger fields are generated in its immediate v~cinity than could be sccored a t reasonaI>\t,:e current leveis by a truly centrai conductor. Furthermore, tlie return conductor, also closely conforming to the outslde surf:ice. rc~nforces the magnetic field. Complete circul:tr magnetizatio~l of the eniire cylinder is acconiplislied by rotating the case one full revolution while 3-phase, full wave D.C. is flowing in the conductors.

PO PIE

. ENERGIZING COIL

Fig. 173-Schematic Drawing of Conductor Arrangement for Longitudinal Magnetization ot Cylindrical Missile Motor Cases. (See Fig. 209, Chapter 23.)

Figure 173 illustrntes schemaLically the same motor case with a yoke-like lnduced field fixture for providing tlie longitudinal magnetic field. The Iongitudi~iil lield is induced in the .4-E-C-O sections (see dram~ng) and complete magnetizalion is accomplisl~ed by rotating the case a full revoiution wilile 3-p11:lse f u l l \v;ive D.C. is

flowing in tlie fixture. Repositioning the pole pieces to pomts E and F and repezrting the process provides efiective magnetization i n the E-E and C-F sections. Before this techn~quc can be applied to any specific motor case, proof c~f its effectivericss must bc? obtarned by esperimentai means. If the cast \%we rolled out flat xvitli tlie magnetizing fixture in place, it can rc:tdily be seen that the flux

ies do not iollo\s a strnignt. line from pole Lo pole of the yoke. Tliey SpIYad out in an eiipticai pattern (sec Fig. 722, Chapter Tj re- ultincr in a \%,eakenetl field as i t progresses from the poles to the

midpoint bet\veen them. If the case in question were longer, tlte entire iength should he magnetized in separate stages through ections A-B, E-C and C-D: etc.: rather than singly from A to D.

Bath application is critical for both circuiar and longitudinal agnetizations. I t must be applied in the area close to the magnetiz- g sources, wlisre the Iicld is strongest. The flow of bath must cover

tlie surl'acc tliorougl~lp, but must ire gentle, without much velocity, so that patterns are not washed away by tlie flow of the bath.

Port areas, such as thrust-revel-ser stacks, call for additional specla1 cons~derarion. Circumferei~tial defects in the \veld cones, as well a s 111 the parent metal itself, must he looked for. The previously applied circular and longitudinal mametizations are not entireiy reliable in these areas because of field distortion around the

Fig. 174-Induced Current Fixture for Missile Motor Case Ports

PRINCfPLES OF MAGNETIC PARTICLE TESTmG

Figure 17.1 sho~vs how t h ~ s phase of the problem was liandled with a semi-portable A.C. ~nduced-current fixture. The core of the fixture IS inserted ~ n t o the openlng of the port and the current turned on to produce a toro~dal magnetic fieid around the w;rlls of the cylindrical section. While tile current 1s on, the bath of magnetic particles IS applied from a hand hose. After turning ofI the magnetizing current ( to avoid demagnetizing action by the A.C. if the fixlure is withdrawn vlrile the current is flo~ving) the fixture IS withdrawn and the inner and outer surfaces erani~ned for defects.

19. ~ ~ U L T I P L E TEST SYSTEPS. From the foregoing discussions i t should not be concluded that specla1 m:tgndic particie testing equipment 1s any k ~ n d of panacea, or that magnetic particie testing is the singie no~tdestructive test for any and all defects associated with ferrous parts. This is certa~nly not the case. Although magnetic particle testing meti~ods hold an outstanding place in the testing of parts made from magnetic materials, radiographic met.hods, eddy-currents and ~1ltl.as0111cs also have a definite place in this same area. There are numerous instances where two or more tests are applied to the same part. For example:

( a ) Magnetic particies and radiography for welded pipe, steel castings and forgings, and m~ssile motor cases.

(b) Xagnetic particles and ultrasonics for melded pipe, steel castings, general tveld inspection, forgings, and billets and blooms.

(c) Magnetic particles and eddy currents for welded plpe and pipe coupiings.

These multipie test applications have not usually been conducted s~multa~~eousiy or by use of a smgle station of equ~pment. They were applied separately with totally separated facilities. This is not surpris~ng, since the introduction of tlie various nondestructi\se testing niethods have followed each other over a long period years. Manufacturers "made do" with available test methods un other n~ethods were developed and oRered. In recent years, howeve planned multi-~netllod nondestructive testing installations iiav been piaced into service for such parts as :

( a ) Welded pipe-magnetic particles pius radiography.

(b) Welded pipe-magnetic particles pius eddy currents.

CHIPTER 19 AI!TOSI-\T1C AXD SPECIAL EQC:IP31ES'I'

(c) \Velded pip^-eddy cul.rents plus ultrason~cs

id) Billets and blooms-magnetic particles plus ultmson~cs.

The trend to~vard multi-method nondestructive testing systems mill undoubtedly continue, because there are very definite economic as well as techn~cal advantages to be ga~ned by combining more than one tesl into a single handling and testing ~nstallation.

20. FUTURE TRENDS. In the course of this chapter \ve i~ave discussed spec~ai equipment-autoniatic and spccial-purpose, single- purpose and multi-purpose-for testing parts varylng as w~delj. a s from one quarter inch heartng balls to ten foot diameter bear~ng races; 30 foot iong elln tubes, ten foot diameter m~ssile motor cases, five ton blooms and 30 ton steel castings. The trend toward special equipment of all types ant1 slzea has continued a t an accelerated rate during recent years. The reasons for t h ~ s trend are numerous. The wide demanll for higher quality by users of materials, brought about by modern d e s ~ g ~ ~ trends, aiong with the demand of nianufacturers for increased production rates, has forced the move toward special equipment. The above demands are, of course, the result of tech- iiologtcal advances In materials and fabr~cating teci~niques spurred by the need for higher strength-to-\%,eight ratios.

On the side of the nondestructive testing industry, a large factor in the increas~ng use of special equipment is the availability of improved special unit designs in such areas as mechanical handling, magnetizing systems, ~i-et method bath application methods and electr~cal eontroi circuits. A further a ~ d to the design of special applications is the development of greatly improved magnetic particles, to make the results obtainable from tlie use of specla1 methods much more effeetise.

Finally, as always, the tiemonstration by successful installations of tlic? advantages of such systems enlises more and more users to be attracted to them. There seems to be little doubt that the trend to~vard speciai equipment and methods will continue, stnce the underlying reasons for it are not lilcely to lessen in number or value.

\ . . i ,. A

%. /,; ~ -<- 7. %,.A. .. om tile poin: of rielv of tlie disconti~iuity itself. What are the -I ,=

acterlstics of the dcicr t i t se l j ~sh lch allo\if or pre~:ent i k detec- CHA~TER 20 by one or more of the procedures which hare been rvorked out?

DETECTABLEDEFECTS D c - ~ c ~ s CLASSIFIL?). In Chapter 3 defects were classified, described with respect to when and how they were caused dur-

1. THE 3IACNETIc PARTICLE TESTING FUNCTION. A s e a t den the production, or the service life of tlie part. From the p o ~ n t has been said in tile preceding chapters about defects and discon- view of detectability, hoxirever, we are not concerned ~ ~ i t h the tinuities, anif how they may he detected wit11 the magnetic pariicle w i l l of ihe discontinuity. Of lmporLqrlce only is its size, s l~ape, testing method. I t has been emplias~zed that all discoiiiiriuities are ntation and location, with respect to its ability to produce leak- not defects, and that a given discontinuity may or may not be a fields. The oniy classification which is pertinent to thts analysis, defect, depentiing on its location lsith respect to operating stresses erefore, is that \irh~ci~ separates discontinuities Into t ~ v o g.roups: 111 the part, and whether or not i t will interfere with tile perform- those that constitute a break i n the surface of the part, and 2: ante of the part in its intended servlce. e that lie entirely beioxv the surface. I t is t h ~ s consideration

I t is the function of magnctic particle testing to show the presence e than any other that detern~ines the selection of specific tech- of discontinuities, but i t is not par t of its function to determine ues for magnetizing and epplying particles. .shethey or not such discontinuities constitute defects. But sine i t a n n o t be known in advance how serious discontinuities may 3. I l i i r o n ~ a x ~ CHARACTERISTICS OF DISCOIZTIXUITIES. In con- or xvhether they will be harmful or not, the function of the met dering discontilluities broadly, and their cl~aracteristjcs ~yhlch is to find all discontinuities which its proper application is capabl r e a bearing on their detection with magnetic particles (as we]] of indicating. How liarmful to the service lire of the part s their possible role as defects) a number of points are of im- disconti~iuities may be is a matter to be decided by human ~ u d rtance, as listed below r ment, by those who know the service retluirements of the part, an A. Discontinuities OI)CII t o t h ~ S I I I . ~ ~ C [ ? . See Fig. 175. can appraise tile magnitude of the effect on these requirement which a given discontinuity may have. The problems of interpreta (1) Deptll, D. The distance from the surface to the farthest tion and evaiu;rtion of indications and discontinuities will be dis Point Of the defect, measured in a direction normal to

cussed in a follo~ving chapter (22).

The usefulness of magnetic particle testing in the search To (2) Lengih, L. The longest dimension measured a t the sur-

defects, and tlie controlling factor in the decision \vhether to face, in a direction pmallel to the surface.

the method or not, depends, therefore, on exactly xvhat disconti 3 ' I ' The longest dimension measured a t the sur- ties the method is capable of finding. In various places in the pre face, in a direction parallel to the surface, and at 900

ceding text, the proper means and procedures to use in magnetizi a part so as to provide the optimum conditions for finding ( 4 ) Shape. Sharpness a t the bottom-V or U. greatest number of defects have been fully discussed. Fiules Itair (5) Angle of pelletration with respect to the surface. been given aiso regarding what magnetic particles to choose an horn to apply them in order to g ~ v c the best possible indicati (GI Orientation of principal ilimension The many variations of techniques have been discussed a t lei ( a ) wit11 respect to the longitodinal axls of tile part, with respect to their effect on the defect-detecting abilities of t l (bl ir'ith respect to the trnnsverse asis of the part. method. It would seem worthn.l~ile, before ending these discu to look a t the problcin from the other side of the fence, a s i t were

(c ) \\-ittl respect to ihe surface.

('7) Ffeqliency. The number per unit of surface area. 3GG

31:T

CIII1'TP.K 80

nETECT;tRLE DEFECTS

( 6 ) Or~entation of planes of ltrriicipal dimensions ( a ) with respect to the surface, f b ) with respect to the longitudinal axis of the part, ic) with respect to the transverse 11x1s of the part.

(8) Interrelationsh~p. Grouping, alignment, etc. 17) Frequency. Rumber per unit area of cross-section.

(9) Relationship of all cl~asacteristics to service stresses (8) Interrelationship-grouping. the part, and to critical stress locations. (9) Reiationship of all characteristics to service stresses in

(10) Stress-mising effect from all considerations. the part and to critical stress iocations.

B. Discontinuities iying .zuholly below the su?fffce. See Fig. 176. f 10) Stress-raising effect from all considerations. (1) Length, L. Longest principal dimension, measured a t the Note: I n the above listings, the term "surface" in all cases refers

surface, and in a direction parallel to the surface. to the szcrface 011 101iich the inszjeciioii. is belng ainde.

From the standpoint of deteetabilitp. only items one through six e a bearing. The entire list, one through ten, bears on the ques-

Length. of the discontinuity as a potential defect. (3) Height, H. Dimensioii normal to the surface.

(4) Depth, D. Distance from the surface to the nearest par uities malte up by f a r the largest and most important group of the discontinuity, measured at 90' to the surface. Note filch magnetic particle testing is used to locate. This is true for that the dimension, D. has a different meaning as between ro principal reasons. Fit.st, the snrface cracii is the type most surface and sub-surface discontinuities. ffecticely located with magnetic p:~t.ticies; and, second, surface

(5) Shape. Globular, angular, flat, sharp-comered, etc. racks 3s a class are much more important and dangerous to the

365 340

OZTEC'I'..$RT.E IIEFE('T5 PRINCIPLES O F ~IAGNETIC PARTICLE TESTING -.

or successful detection in 3112. given rdse it must be possible to serv~ce lire of a part than are defects lying v:holly Ibelom thc surfac up a lieid of suRicient strength and in 8 generally iavorabic They are tlieri?fore more frequentl): the object of the inspectiu~ irection to pruciuce siwng lealrage fields. This 1s especially true if Fibre stresses are usually hgliest a t the surface of a part, and an disconti;iuities are small an11 fine. Assum~ng that this has been break ~n the surface constitutes a point of still hlgher concentratio n, the most favorable ci~aracteristics of a discontinuity itself of stress. A surface el-siclc by its nature is very sharp nt the botto. r its detection are and is the most severe Itind of stress-raiser. For the latter reaso surface discontinuities are looked for wit11 extreme care, if expecte ( a ) that its depth be a t right ang;les to the surface,

stresses are even moderateiy higii. (b) that its width a t the surface be small: so that the alr gap

I The surface discontinuities looked for ~ v i t i i mapletic particie test it creates IS small:

~ n g illciude all fatigue and service cracks, and such serious source I (e) that its li:ngtli at ihe surface be large u'itli respect to its of potientiai failure as seams, laps, quench~ng and grinding crack I width,

i as well as many surface rulbtures occurring in castings, forgillg and xveld~nents. td) that it be comparatively deep in proportion to its surface I opening.

5. DETECTION OF SURFACE CRACKS. I t has been stated repeated1 e field set up in the part then, should be ul the ilirection a t right that the magnetic particle testing inethod is the ?nost sensitive an les to the length of the defect. reliable methocl for locating sul-face cracks in ferromagnetic lnat rial. That t h ~ s is true is due largely to the fact thal in the sre ncip~ent fatigue cracks and fine grinding checks often have a majority of cases, no extremely critical conditions or technique h of less than 0.001" and a surface openlng of pefhaps one tenth retju~red for the detection of surface discontinuities by this me or less. Such cracks are readily located usnig some form of Magnetizing and particle al~plication methods may be critica I e wet method. The d e p t h of the crack has the least erect on its certain special i~istanccs (See Chapter 19), hut in the case of mo ectability, except tllai, up to a certain limit, the deeper the crack applications the retloirements are relatively easily met, becaus stronger mill he the indication for a given level of magnetiza- leakage fieitis tend to be strong and are highly localized. A few slnlpl This is because the ieakage n t ~ x is stronger as the crack bui in~portant principles are involved xx,Il~ch nlust be observed, bu mes deeper, due to the greater distortion of the field in the there 1s usually a fairly wide latitude 111 the selection of procedure Hovever, this effect is not particularly noticeable beyond and milieriais when surface crac1;s on/!/ are being sought. If th aps one quarter of an inch in depth. If the crack is not close- discontin~iities are of a size or character to be in the threshold are ,ed but is wide open a t the surface (has a large dimension W) of detectability special techniques may he necessary. It IS, Ilo~r'ever reluctance of the resulting longer air gap reduces the strength reiatioeiy easy to define the ciinmcterlstics of a surface disco~ttinuit. that mnke it favorable for detection.

I t is almost self-e\'ident then, that detectabilit3~ involves a rela- The mater~ai must, of course, I J ~ ferromagnetic, and should have onship between surface opening and fleptl~. A surface scratch,

;I n ~ a x ~ n ~ u ~ n n ~ a t c r ~ a l permeability of not less thnn 500. (See Chap- h may be as wide open a t the surface as its depth, does not ter 9, Sections 9. 10 and 11.) Five hundred 1s the material permea Ily gi\-e a magnetic particle pattern, although in some cases it bility of nickel, and surfiice cracks are quite easily foutld nl this do so at hlgh levels of magnetization. Because of many other metal. Cobalt and some types of 3Ionel metal ~ v h ~ c h hare a material iables that may enter, it is not possible to set. up any exact values pernieability of less th:ni 500 are still sufliciently magnetic t l ~ a ' this relationship, but in general a surface discontinuity \vliicll surface c~.i~cits m;tv I J ~ located, using high-sen3itivity techniqtles t least five times as deep as its opening a t the surface, will be i.e., strong direct corrents \ritll circulal. magnetization, and we magnetic 1,;lrticies.

371

There is some e~sidence of a limitation a t tlre other extreme, namely, a crack of some depth but with its surface opening so small that no indication is produced. If the faces of a crack are tightly forced together by compresslre stresses, the almost complete absence of air gap may produce so little ieakage field that no particle indication is formed. Shallolv crzlch-s procluced in grinding or heat treating and subsequently put lnto strong comuresslon by thermal or other stresses have been reported, which gave no ~ndications with magnetic particles. Sometimes, with careful, maximum sensitivity tech- rrlques, faint inclications have been produced in such cases. The operator should be alert to the possibility of this occurrence when dealing with a part the surface of which may 11ave residual con& pressr?!e stresses from any cause.

One other condition sometimes approaches the limit of detectability and should be mentionecl. This is the case of the lap. produced In forging or rolling whieh, though open t o tlie surface, emerges a t an acute angle. (See Fig. 32, Chapter 3.) Here :lie leakage field produced may be quite weak, because, due to the small angle of emergence, ilnii the relativeiy h ~ g h reluctance of the actual air gap which results, very iittle leaitage Rus takes the path out through the surface lip of the lap to jump this high reluctance gap. When laps are betng sought- usually alxvays when inspecting nemiy forged parts- h ~ g h sensitivity methods, generally with the use of fluorescent particles, a re des~rable. F i y r e 177 shows the faint

Fig. 177-Fluorescent Magnetic Particle Indication of a Forgmg Lap.

312

C I I A M ~ 20 1)hTECT.\llI.F: DEFECTS

indtcation of a forging lap, produced nsith Auorcscent particles, whlclt sectioning sho\tred to extenct, a t an angle, qulte deeply lnto the body of the part.

6. DISCONT~NUITIES L'iItiC WHOLLY BELOW THE SURFACE. The magnetic particle method is capable of finding many defects xchich do not break the sol-face of the Imrt in which they occur. This 1s an important ability, since there are circumstances when radiography and ultrasound, methods whose primary field IS locating such defects, cannot be used. These two methods are ~nherently better adapted to the location of interior discontinuities than magnetic particles, but sometimes the shape of the part, location of the defect, or the cost or availability of the methods and the equipment needed, makes the magnetic particle method the best one to use. As a group those discontinuities which lie wholly below tlie surfnce are less dangerous from the point of vlelv of potentiat failure than are surface cracks. This is because they are usually (though not always) more or less rounded In shape and, lying below the surface, are in an area of fibre stress below the maximum. They are, therefore, iess severe stress-rarsers than even a very small surface cmck.

The detection of such discontinuities with magnetic particles is nonetheless often important, and much work has been done to determine the optimtlm conttitions for success in this area.

7. DETECTION OF DEFECTS LYING WHOLLY BELOW THE SURFACE. Definition of the limiting conditions that determine whether or not a discontinuity beio!t7 the surface is likely to be found ~vi th magnetic particles, is not nearly so simple as is the case with surface craclcs. A iarge number of variables are factors, any one of whtch may be determining in a gtven case.

The question, often asked, "I-Iom deep bclo\v the surface can a defect be detected with magnetic particles?" has no answer in specific terms. But some of the factors and \.ar~ables that affect the detectability of deep-lying discontinuities cnn be defined ;in11 understood, so that an operator can be aware of what the problem really is.

8. Two GROUPS OF SUB-SURFACE DISCONTINUITIES. The sub- surface discontinuities ~ ~ ~ h i c h m;~grretic particles ~rzill locate may be put into two groups. The first of these comprises those small vords or non-metallic inclusions which lie close to and often i r ~ s f under the surface of the part. Xon-nletaltic inclusions are present in a11 steel products to a greater or lesser degree. They may occur

a s indiv~diiai enti:ics, or the,: may be aligned a s Iolig stringers. In the inspection of machined, higli!y finishcu parts, xv]lcre the \vet continuolis D.C. method using lilgh magnetizing

IS a comnlon techn~qut?, these stringers a r c r~f ten foulid. They a r e seldom significant unless they occur in excessive numbers or lie in a transverse direction in an area of high stress. Since they a r e usually very small they a r e not found unless they lie very close to the surface.

Because of the nature of these discontinuities they produce llighly localized but rather mealc leakage fields. Contrary to the rule earlier laid down, t ha t the dry method excels for finding deep lying defects? the met method i s thc best one to use for this inclusion type of discontinuity if its detection is important. This is because these fine, non-metallic strirlgers a r e not really "deep-lying", for the reason that, though sub-surface, they must be very close t o the surface to be found at all, by any method.

9. DEEP-LYING DEFECTS. The second and much more i m p o ~ k n t group of sub-surface discontinuities a r e those larger and more serious conditio~is which may be quite deep in heai'y sections-- perhaps one quarter inch t o two inches or more helolv the surface. These may be, in weldmcnts, lack of penetration, sub-surface lack of fusion, o r cracks in the beads belo~v the iast in heavy welds. Such deep discontinuities in castings may be internal shrinlcs, slag inclusions, o r gas pockets. Although i t is ~ n ~ p o r t a n t to know and define limiting conditions for the detection of such disconti~luities, t o do so presents a real problem. But because of the oflen-expressed interest t ha t e s ~ s t s in the problem of locating deep-lying discontinuities, i t i s ~vo r th while to consider i t in some detail.

10. DEFINITION OF TERIVIS. I n Section 3 of this chapter i refer to Fig. 176), the meanings of t he terms "Depth", "Height", "Length" and "Width" a s app ly~ng to discontinuities lying \rholly belox~ the surface were defined. The ~mpor t an t poiut to remember here is t ha t these dimensions a r e always related to tha t surface of the pa r t on. zoltich t l ~ e -n~agitetic pcti.ticlc i~rspection 1s bei:iig ?iutde. In some instances, a s in a rectangular cross-section, the discontinuity being sought may be coritiguous to txvo or three stirfaces of the body i n which i t i s loc:~tett. Detection may be more successful on one of these surfaces than on thc others because of shape o r orientation of the defect.

In the case of a tube, fo r esumple, the defect mag he "deep" with

Cf11rnB '30 IJE3'ECTRLE DEFECTS

-- - - -- -

respect to the outer surface, but may be q o ~ t e close to tirr surface of the Inner wall of the tube.

11. CONCEI'T OF DEPTH. The principal reason why one cannot set a iimit in terms of inenes or fractions to thc depth to nhich magnetic particle t c s t ~ n g can reacil in locating internal discontinuities, is tha t size and shape of the discontinuity itself ID re1;ltion to the sizc of the cross-section in ~vliich it occurs, is of controlling ~mporlance. T l i ~ s idea of relative dimensions can be easily expressed qualitatively by saying tha t "thc deeper the discontinuity lies in a section, the larger i t must be to be detected with magnetic particies."

This is, Iiomc\~er, not a very satisfactory way to dismiss the problem, and the concept of depth can be better r~sualizcd hy taking a concrete esample. Suppose me cons~der a round bar having a ratlius of ::{I inch and that. i t has a defect having a height, FI, of 0.025 inch, lyrng a t a depth, D, of 0.025 11ich below the cylin<lrica!

Fig. 178-Concept of Depth of a Deep-Lykng Defect.

surface. Lcl us fur ther assume that the width, W, is 0.02 ~ n c h and the length, L, 1s 0.025 inch. I n another bar suppose we co~~sitlcr that all these dimensions a r e doohled, and in a tlnrd bar that tiley a r c all multiplied by ten. The latter t\vo bars \vould have ratlii, respect~vely, of :\;, inch, and 334 inches.

The percentage relations of these dimensions 111 each of thu three c:ises a r e tile same, and wit11 steel of ordin:~rily fauolshic mag~letic properties, and by uslllg pl'opef fillx density and s~iitable niagnrtic particies, Ire would exprct no dilficulty in getting good tndications

'7 5

CHAPTER 20

of any of these three defects-~n fact, they would all s11o.r~ ab equally well. However, if we take the discontinuity of Bar I-esult, especially when the defect lies quite close to the surface. (H= 0.25 inch) and put it mto Bar #I (Diameter = :%. inc dimens~on of depth is, ho~i'euer, the dominant factor in deter-

mount of build-up over a discontinuity lymg wholly below the indication, a t least with circular magnetization. There would be no appreciabie field distortion. Nor \vould the discontinuity of Ear #1, if transferred to Bar #3, without change of size, but at nce from the test surface to the nearest edge or surface of the the same depth as the discontinuity of Bar #3, be expected to give a good (or any) indication.

13. EFFECT OF WIDTH. Width is defined as the longest dimens~on Stated another way, ihe "depth7' of a defect, as i t affects detecta- f the discontinuity measured in a direction parallel to the surface

bility with magnetic particies, is a relative term, and cannot be considered alone. On the contrary, depth must be considered in relation to the size of the defect, and the totai dimensions of the cross- section. It is relatively easier to find a defect of a given size and oriented 90° to the surface, and a t right angles to the flux lines. shape a t one half inch below the surface if it occurs in a four inch section, than if i t occurs in a section only I inch thick. In the latter case, i t rvould be a t the center of the section. (See also Section 18, this Chapter). ave assumed a crrcular sliapef are the dimensions of prinie

12. SPREAD OF EMERGENT FIELD. Though we have said that the defects in bars 1.2, and 3 discussed above wouid be found with equal facility, although they have widely differing depth, D, dimensions, the appearance of the indications would be quite different. The width he defect would be R L U C ~ less detectable than if this dimension were of the particle pattern on Bar #3 u~ould be probably as much a s % he 0.04 ~ n c h mentioned above. There must be an appreciable high- inch, whereas the paltern on Ear #I \vould be compact and only eluctm~ce gap to set up an obstruction in the path of the field, else some 20 thousandths of an inch wide. The reason for this difference o leakage flux will be crowded out to the surface. lies in the angle through which the leakage field spreads a s it leaves 14. EFFECT OF HEIGHT A N D LENGTH. These dimensions, in the the surface. This is a function of the actual depth, D. The more metal ease of the forging flake discussed above, are measured in a plane between the defect anrl the surface, the more "spread" of the leakage a t right angles to the flux lines, when we are constdering a dis-

continuity from the point of view of detectability. Thew magnitude determines the size of the obstruction presented to the lines of force, and consequently also the amount of crowding of flux to the surface

onzontal diameter to say one eighth or one sixteenth of an inch, the chance of detecting it would shrink to the vanishing point. The width, measured in :I direction parallel to the flux would enter into

Fig. 179--"Spread" of the Emergent Field at a Detect. the result not a t all.

3'76 377

op of a tin c:lii; on edge belolv the surface and at right angles to jf, instead of a lens-shaped flake me assume a discontinuity having

the surfacs and the flow, would reuse a sharp disturbance in tlie a sort of ttibular configuration such a s i1lustr:ited in Fig. 176, novenicnt of the n.a:er: hilt a st?algh.ht. rountl stick placect ; ~ t a

having a longest dimension, i: parallel to the flux lines, but present- lei to the direction of Aow, n.oul<j ila\ze \:cry

ing a substalltiai circular face a t right angles t g i the f l t l ~ path, the would be quite direrent. The distortion of the field would Le

less sharp, alld although there \rould be crowding between the dis- 15. EFFECT OF SHAPE. We liare beeii considerlnp a lens-shal)ed continuity and the surface, the leakage field, if any, mould be spread feet and the area it presents to ohstruct the liiles of forcc. A out along the distance corresponding to L, o r longer, and would her~cal gas or slag inclusion of the same diameter as the iens- not be expected to give a very re:idable indication. nuity wouid present tlie same projected

111 detectability of a discontinuity lying belo\5 the rea, but would he 2?r?!cfi less likely to be detected. In sucii a ease the

surface, a r e concerned pr~niar i ly with the projected area which ux lines would "streamline" around the sphere, and the disturbance is presented a s an obstruction to the lines of force, and the sharpness reated in the field wotiid be much less sharp. of the distortion of field produced. It is helpful to think of the field 16. EFFECT OF ORIENTATION. Another most importnnt Euctor in the specimen a s flo\ving through i t like a s t ream of water. The 'n detectability is the orientation of the defect. The lelis-shaped

bstruction disciissed aborc \isas eonsidcrect to he a t 90C to tile irection of the flus. If, however, the same defcct is i~iclined, either ertically o r horizontally, to an angle of only GO0 or '70" instead of OC, there \vouid be a noticeable difference In the amount of leakage eld, and therefore i n the strength of the indication. This ditferelleic ~oiild be due not oiily to the fact that the projected area of the iscontinuity \youid be reduced, but also to a "streamlining" effect s in the case of the sphere.

17. ?Y~OST FAVORABLE DEFECT FOR DETECTION. We find emerging TURBULANCE IN A STREAM om this analysis a better concept of the kind of defect are most MAGNETIC FLUX OR WATER ceiy to be able to find with magnetic particies wlien it lies wllolly

~ i ~ . i , g o - ~ f f ~ ~ t of lntarruption to the Flow of Water (or Magnetic Flux) the interior of the part. The most favorable discontinuit). is one Due t o Shape and Orlentation. ich has a height and length of the same order of mamitude, both

than tlie width or thickness. The !nost this discontintiity is to i t lie so that

e plane of its general direction is at. right angles to the direction

The least favorabie shape for detectioxi is a flat discontinuit?, n g pnrallcl to the stirface instead of 90"; and of discontinuities

rojected area to the flux path, the spherical

TO this summary shottld be added the thought-already expressed that the nearer the discontinuity lies to the surfaee, whatever its ape, the smaller i t can be and still be located; or, conversely, the eper i t lies, the larger i t must be to be detected.

Fig. 181-Effect of Shape and Orherltation on Detectability.

378 379

PRINCIP1,ES OF MAGNETIC PARTICLE 'I'ESTING -- III~XECTz\ULE - 1)EFECTS .- A

18. EFFECT OF &IETIIOD OF MAGNETIZATION. Otller factors than beyontl this, the smooiher the surface is, the more easily can tile

the size, shape and orlentation of defects play a par t in whether they powder move to arrange itself ~ n t o a pattern when leakage fields

will be detected or not. The direction and strength of the mabaetic are weak and diRuse. The surface on wlllch t11e inspection is

field IS obviousiy an important consideration. Direct current--or should be horizontal for best i.esults, so that. gravity docs not tend

better, half wave current-shouid be used for deep-lying defects, to cause the po~vder to fall awns. \Veld beads are often machined

and the dry powder ainiost alu-ays gives the best nldications. off to provide a smootl~ surface for this inspectioi~.

For yeally deep-lylng defects, such a s are sought in thick weid The powder should be applied a s a light cioud, w111ch is cspeclo]ly

ments, tlie prod method of magnetization IS superlor t o any other. \i5ell accomplished by the air-operated powder gun. Patterns can

D.C. yokes a r e elfectix,e if tile discontinuity 1s fairi). close to the be observed as the powder drifts to the surface. Such patterns often

surface. but when depth 1s of the order of one quarter ~ n c h to two disappear a s more po!irder IS applied. When the excess of pomdel.

inches o r more, prod magnetization is by f a r the best. The orerall is blown off. even with a very gentle a i r stream, s ~ c h patterns arc

mdliod now belng w~dely used instead of prods, on large castings often also blo\vi~ away, leavmg only those lleld by stronger alld

particuiarly, 1s superior to prods for finding all surface discon- more concentrated leakage fields. Thus, the limit of detectability

tinuities, but less effective than prods for finding defects iylng may often be dependent on the skill of the operator.

wholly below the surface. 21. THE OPERATOR. The detection of deep-lying defects With prod magnetization i t is possible to produce indications, on a greater degree of skill on the par t of the operator than an,, other

tile surface to n;llich the prods a r e applied, of discontinuities lying field in \i*hich masiletic particle testing is used. It is probably true,

a t the center of the section, o r even of those lying near to or even on tha t out of several inspectors \\rho migilt be given the same part

the opposite surface. Thls 1s commonly done in iocating lacic of contain~ng a small, deep sub-surface defect, not ail would find tile

penetration in a sin@e V butt aeid. Spaclng of prods is important, indication. I t is fo r tills reason that operzltors who a r e to engage and t h ~ s has been discussed in detail in Chapter 10, Section 6 and in the inspection of ~freldments with tile magnetic particie metilod,

follo~ving sections. a r e given qualification tests in o ~ d e r to assure tliat tlleq. are petent in the application of the method for this purpose.

19. EFFECT OF PERIIIE,\BIL~TY. I t is more or less self-evident tha t the permeability of the par t shouicl be high if deep-lyingdefects I t is extremely desirable that new operators betng trained for. a r e to be located in a given specimen. A maslmunt ma t e r~a i perme- this purpose be given the opportunity to see tlle c h ~ p p i n ~ - ~ ~ t

ability of a t least 500 has been mentioned in connection with the re-welding of some of these deep-lying defects, so that they can detection of surface cracks, but the value sl~ould be much higher see exactly \?;hat caused the indications they produced on the surface for deep-lying defects. Sirice the need to find mternal v o ~ d s IS most of the weldment. In this way tiles are able to confirm their work urgent in the case of weldments i t 1s fortuitous that the steel in- and build up experience ~ ~ h ~ c h makes them better abie to reeognr.e volved is very often lo\\, carbon steel in t l ~ e forin of plate o r shapes, signilicant patterns wl~en they occur.

~ v l ~ i c h l ~ a s a very h ~ g h mater~a i permeability. The gro\rrlng use of harder steels fo r pressure vessels and piplng 1s unfavorable to the location of deep defects m such materials. The effect of reduced permeability was s i~own by the tests oil the tool steel ring, unhardened and af ter hardening. See Figs. 115 and 118, Chapter 12

20. OTHER FACTORS. Other factors \<;hich may play a n important part in the detectability of any glven sub-surface discontinuity, are the condition of the surface and the method of applying the powder. Obviously the surface should be elem and dry: but

CIXAPTER 21

SON-KELEV:kXT IYDICATIONS

I . DEFIXITIOX. R~ference \\,as made in Chapter 3, Section 4, t o indicillions causcd by magnetic o r metallic discontinuities 1i;hich a r e present by design, o r by cooditions which hare no hcaring on the strengtli o r service uscftilness of a parl . Such patterns hnve no reintion t o potential defects, whether they a r e caused by design or accident.

Before an ouefittor has progressed very f a r In h ~ s experience with tnagneiic particle testing he ~irill h a r e become aware of this type of indication. T l~ey Can sometimes be very puzzling, since in many c;lses investigation to account fo r them a t first reveals no apparent reason for t l i e ~ r c,ccurrence.

Such nidic:~tions a r e t rue particie patterns, actually formed and held in place by ~iiagnetic leakage fieids; but the lealtage fields responsible a r e nut caused ijp conditions that are relevnnt to the strength o r useability of the part. Tlie name "non-relevant" tias been gi\.en to this type of pattern. Obv~ously the magnelic particie operator must be aeqt~aintecl wit11 these non-relevant intlic;?tions and be abie, ir-hen they occur, to recognize tlicm f o r what they are.

2. FALSE INDICATIONS. Tlie term "false" has sometimes been applied to ail i~on-relevant indicatio~rs, but the name is not a good one, s~ticc. such inctic;~tioiis a r e in nearly all cases magnetic in ongln. There is perhaps one truly false indication, and tllat is the case of particles lieid ~neclia~licallg o r by gravity in surface irregularities, with no relation whatever t o leakage fields. Similarly, \r-hen usung the met method, a "dra~nage line" of particles will often form. In such cases i t is only necessary to shake o r blow, os to rinse oli tlie particles to prove that they a r e not magnetically held.

3. EXTERNAL POLES.^ Particles ~vi l l adhere to local poles a t sharp corners, sharp ridges o r surface irregularities, but these efrects a r e not usually very confusing. Such patterns a r e most IiIteiy to be encountered when longitudinal magnetization is being used. Their occurrence is often a n indication tha t too s t rong a mag i~e t i z~ng force has been applied.

382

4. ALL-O\'ER PATTEI~XS. Sometimes suriacc patterns o r a1l:over patterns a r e produced \7,,licn n1;ignetiaing clrculariy. These are the ~ndications of flux lines of tlie exlei-nu1 Acld produced by the niagnetizing current. The lines of the pattern \vill always appear a t right angles to the direction of current floxv, and a r e usually pro- iiuced onip when ton much cnrrent is being used.

Fie. 182-Magnetograph Showlng the External Field Pattern Produced When Prods Are Used to Magnetize Steel Plate.

5. EDGE OF SCALE. A ratlier obviously non-relevant indication t l ~ a t wli~ch can appear at the edge of a patch of tightly adherent ill scale. It is usually a patch of very tlitck scaie that produces

uclt a n indication. Since the mill scale is magnetic but with a very mucii lo~ver permeability tlian the steel from trllicll i t was formed, this indication is really the result of a magnetic discontinuity. Thougli obvious, i t is surprlstng how frequently a careful examlna- tion i s requ~red t o eoiivince an operator tha t the indication is not

genume metallic surface discontinuity. Visual examination is in ost cases suliic~ent, but occasionally a more careful cleaning oi" th r lrface is found to be necessary. Such heavy scale patches a r e today ot very conln~on on rolled o r cast pi-oducts. The extent to \r-hich scaie patches, if they occur, may tend to

confuse tlic test, depends on the liind of clefects bclng sought. For exantpie, if lookrng for seams in hot-roiled bars or billets, an irrezu- l a r patcii of scale could not be confused with the straight-line indi-

of a seam. Furthermore, the irregular scale pattern is not i e web, and at the base of certain gear teeth that lic adjacent to ally way si~nilar to ally defect likely to be found in bars or billets

6. CONSTRICTION I N THE ~IETAL PATB. One of the tilost co Such patterns can usually be recogn~zed by the Pollolving

non-relevant indications is that caused by a constriction in the aracteristies: path through <hie11 the flux must pass. Such constrictions a (1) Giventhe sainc magnetizing technictile they will appear on all caused by the shape and construction of the part. The crowdin parts arid m the same location. They n,ill all have substan- of flux to produce a leakage field is exactly the same as that tially the same appearance-:t condition airnost never met produces indications over a sub-surface defect, only in the c where real sub-surface defects me invoired.

(2) They can always be related to some feature of construction Figure 183 shows a gear and splined shaft magnetized circula or cross-section which accounts for such a leakage field. by means of a central conductor, with particle patterns on the o side of the shaft showing the inside splines. The field is cro~vded ou (3) They seldom look, to an experienced operator, like indica-

by the reduced thickness of the metal a t the base of the spline tions of any sort of relevatit discontinuity.

Patterns also appear on the gear between certain of tlie holes f there is any question of the presence of a real defect a t such ocations, the- part should be demagnetized and then remagnetized

a lower level, reducing the magnetizing force until the non- levant pattern disappears. There \vill still be field enough to glve 1 indication if a real defect exists in such areas.

7. SHARP FILLETS AND THREAD ROOTS. Similar in character to he non-relevant indications just described, are those due to leakage elds a t sharp fillets or roots of threads. Here the flux lines tend to idge the air gap a t such points rather than to follo~v the metal path actly. Confusion is particularly likely to occur in locating actual

racks a t the roots of threads and in sharp fillets-l~laces xvliere racks are very likely to occur.

Fig. 183--Gear and Shaft Showcng Non-relevant Indications Due to Internal Splines.

Nagrietic \vriting seltlom yr!rcs :my trouble in interpretation be- use its appearance is characteristic, the pal,ticirs b e ~ n g loosely

eld and peculiarly fu::iy in outline. They seldom resembir the pat- d by all ,?ctu:il discontinuity. If the object i s t l~orougl~ly

and a t the non-relevant field. magnetized and reniapnetizcd a s usu:~l, and the inspection n~edium a m applisd~ magnetic w i t i n g \rill riot again uppear, n.hereas a

other \\riien either o r both a r e in a permallelltly ma I flaws will repeat its indication af'ie? any number of such cycles. ent fo r tlie occurrerlce of 1:lagnetic \vriting IS that the

i i n which i t occurs must ]lave a sufiiciently high retentivity, so it will retain an extremely local increase o r decrease in mag-

he phenon~enon is more apt to occur in circularly mag- than in those longitudinally magnetized. To prevent

such par ts should not be allo~ved to colne into contact with ferro- agnetic objects af ter being magnetizoil.

9. EXTERNAL MAGXETIC FIELDS. Non-relevant indications niay ,pear on hardened steel par ts due to residual local poles, which a:: be caused by proximity of magnetic fields f rom power lines,

186 is sho\vn a roller bearing assembly, having a quenclling t A. The patterns a t polnts B a r e local poles left by

magnetic ciluck. Den~agnet iz i~ig follo\ved by remag- the magnetic chuck indications, but the indicn-

the crack \\,ill, of course, reappear.

. COLD WOIIKING. The type of plastic deformation called cold ting produces a hardening in steel \vit11 a consequent ehatlge in eability. When the cold working is very local, the abrupt

hange is often st~tlicient to cause a particle ~at te l -n . produced is a t times similar m appearance to mag-

On demagnetizlng and remagnetizing, however, tile cold xvork reappears, whereas that due t o mametic

lnicroscope will usually show the typical grain distortion rep- enting cold work. Cold x o r k indications can be removed only

357

Fig. 1 8 ~ R o l l e r Bearlng Assembly with Non-Relevant Magnetic Particle Patterns Produced by a Magnetic Chuck.

by heating the part, usually to 800CF. or above, and allowing th part to cool slowly.

One example of such cold work indications was found in the case of some neiv helical springs which gave a pattern resembling dis- continuous seams. The indication on each turn of theihelis u-a immediately opposite the ~udicatious on adjacent turns. This occu rence was a most unlikely coiiicidence if the indications were cause by seams. Further investigation revealed that the springs we being compressed by complete ciosure ~irith a strong blow durii the acceptance tests, and cold working was o c c u r r i ~ ~ g a t the contact areas between turns.

Another example was found in the case of new aircraft bolts. These had a c~rcumfere~~t ia i or transverse indication which; in some instances, had the appearance of a fatigue crack---obviously impossible in new bolts which had lleitller been in service nor strained in any was. Examination under the microscope of additional specl-

CI~AI.TER 21 NUN-IZELEVztNT INDIC2~TIONS

mens indicated cold working-, u~hicli mas coniirnled, being found to hare been caused by ci~lipers or snap gaugcs used during the finai grinding operation. The laser of cold work was extrenreiy sl~:illo\~; since the bolls mere hardened, but the effect undoubtedly ,?as responsible for the non-relevant indications.

Figure 187 shows another exampie of cold work patterns. Thls is a piston pin on which a heavy rougl~ing cut was taken, possibly u~ i th a dull tool. The metal was cold worlied under the polnt of the tool to such a depth that subsequent finishing cuts and grinding did not completely remove it. The spiral pattern was brought out when the pin was longitudinally magnetized, and suggested the cause in machining.

Fig. 187-Magnetic Particle Pattern Produced by Cold Working Due to Machtning.

11. LUDEPS LINES. Planes of slippage, called Luder Lines, In materiais which have been stressed beyond the yieid point, are occasionally the cause of non-relevant magnetic particle indications.

Bent plates for assembly for tank members u;hich have bee11 formed coid by means of rolls, if magnetized before stress-relier~lng by passing current through the piate with xvidely spaced prods, will

i

CU~ZTER 21 PRlNCIP1,ES OP 3lAGh'ETIC A R T I C L E TESTING NON-IIELEVAXT IZD1C3$T10SS - -- - --- -

show. indications of Luders Lilies when dry powder is dusted over i6. JOINT BETWEEN ~ I S S I \ ~ I L A R II~AGNETIC I\IATERIAI,S. Some- the surface. They appear as a n nicomplete crlss-cross pattern of times a piece of hard stcei is butt-welded to a softer piece by any of lines running a t 4 5 O to the asis of curvatul.e. several niethods. A t the point of ir.eiding tllerc n.ill be a sharp

12. GRAIN BOUICDARIES. When grain size 1s very large, the ch:tnge of permeability, the soft steel iiavmg a high permeability

macrostructure s l ~ o ~ r ~ i n g grain outliiics may be found to he shown and tlic hard steel a inuci: !mrer one. If n ningneiic fieid be set up

by a magnetic particle pattern, even though no metallic discontinuity so as to flow across this j : m t there u~ill be a conc~iitrated lriakage

exists. The pattern is due to sharply different permeabiiity as be- field and coiiseque~~tly a niagnetic particle paitern. This pattern,

tween the gvain and the boundary material. however, does not give any information regarding the soundness of the .iseldrd joint..

13. BOUNDARY ZONES IN WELDS. In weld inspection an indica- Figure 18, Chapter 3 slio\vs an example of this condition. A piece tion is often obtained a t the boundary betx!*ecn the fused metal and of soft steel rod is butt-!r.elded to the end of a sinlilar rod of hardened the base metal. Other indications m the form of lines may appear a t stcei. The strong magnetic particle indication is caused by the sharp the edges of decarburized zones. These occurrences actually indicate change in permeability a t the xveld. Under such circumstances it is an abrupt change in per~iieability m the path of the magnetic flux, difficult to get a true magnetic particle indication of an aetu:?llg but a re not ?~ccessa r i l~ inclicative of an ohjectionabie condition. uiisound jojnt, even ut loxv levels of magnetization. Many sound welds ~ v i l l yield a po\\'der line a t the junction of base and weld metai. If in doubt, ~netallurgical evaluatioii should be 17. FORCED FITS. One other example of a non-re1ev:int indica- obtained. tion should be mentioned. When two members of an assembly are

very i i g h t l ? ~ fitted together, a s in a pressed fit between a shaft and 14. FLOW LINES. &Ian' steels, particula~.ly in forgings, \vill pinlon gear, a magnetic particle pattci-II of this Joint may be formed.

show a pattern as slio\\.n in Fig. 20, Chapter 3. Thls magnetic If tlie fit is tight E ? I O I L D ~ 110 indication may be produced. since the

i particle pattern is similar to tlie grain flow in forgings brought out air gap between the two members nlay be almost zero. If an indi- I by deep etching, and is, in fact, produced by these grain-flow lines. cation appears i t is never misleading, uniess the joint is hidden b The indications are brouglit out only by high sensitivity methods and 1 by paint o r rust and the operator does not know that the ~ o i n t i

a t levels of magnetization higher than ordinarily used for inspec- exists. Usually polishing with fine emery cioth \vill reveal the line i tion. The occurrence of such a pattern is not indicative of defective between the tmo members of the assembly. See Fig. 17, Chapter 3. i steel.

Dendritic segregation and segregation of carbides and other metallic constituents may also cause indications. Obtaiiiing such indications usually depends on some incrcase in sensitivity over norinal inspection practice.

15. BRAZED JOISTS. When two pieces of ferromaglietic material a re joined by brazing, the film of bvass forms a ?nng?tetic discontinuity, even though the jomt may be perfectly sound structurally. A magnetic particle pattern will be produced outlining the joint. Since the braze metal is not ferromagnetic, local pules at the edges of the magnetic material will be formed just as though there were no metallic junction a t all. This pattern has no sigliificance whatever from the point of view of the strength of the joint which must be evaluated by other tests. See Fig. 19, Chapter 3.

INTERPRETATION, EVALUATION AND RECORDING OF RESULTS

1. ESSENTIAL STEPS. In any nondestructive testing process there a re three essential steps. These are:

(a) Production of the indication, whether on film, a meter, oscilloscope, tape o r other indicating device, or directly on the surface of the part as in the case of penetrant and magnetic particle testing.

. -

( b j Interpretation of the indication as to what condition is present to have caused it.

(c) E\laluntion of the condition indicated, as to its effect and seriousness from the standpo~nt of useability of the part.

A. Prodl~ct io~l o f the indication. The first step obviously is to apply the nondestructive test properiy so a s to produce mdica- tioils of the discontinuities or other f l a w mhich the method is designed to detect and which i t is desired to locate in the particular situation. The previous chapters of this book have been devoted to t h ~ s step for tlie magnetic particle-method. The var~ous conditions and requirements for producing magnetic particle i~idications under all sorts of circumstances in all kmds of ferron~agnetic ~ n a t e f ~ a l s have been tl~orougiily discussed.

The first step of the testing process has thereidre been ach~eved when the test iias been properly and intelligently applied. Those parts on tvluch 110 indications have appeared niay tlten be presumed free of f law, and can be passetl directly to the nest process in t l i e~ r manufacture. Those parts on rdl~cii indications hazw appeared require further consideration before they can be used, repaired or scrapped.

B. I?z tc?~~.e tal ion of thf: indieafioxs. The indications of the presence of lla\t.s uhich tlie various methods of nondestruc-

CHATTER !!? INTF,RPRET:TATION. EV.II.UIZTION AXU RECORDIXG OF RESELTS -

tive testing produce are, literally, just that. They are indications that sometli~ng rn the part is not normai. But they do not, in themselves, tell exactly what the condition is mhicll has produced them. In magnetic particle testing, cvery pattern IS produced by a magnctic disturbance setting up a leakage field, but furtlier knowledge and information is needed t o enable us to say whether the pattern, and therefore the disturbance, 1s really s~gnificant or not.

Given, therefore. an indication of a magnetic disturbance evidenced by a magnetic particie pattern, someone must de- c ~ d e what the condition in the part is that has caused the pattern-in other urords, he must inter7,ret the indication in terms of its cause.

C. Evolxnfion. Before the part can be disposed of, however, a further and final decision remalns to be made. Once it is determ~ned what tlie condition in the part is, then the condition must be eut~inated in terms of its effect on the useability of the nart for i t s Intended service. After this decision the part may be accepted as satisfactory or rejected as scrap--or perhaps salvaged by reworking it in some manner. A further cost-saving analysis can often determine wliat control of processes can do to prevent future similar occurrences.

3. THE PROBLE~I OF INTEI~PRETING. We have just said that In- terpretation consists of dec~ding what 1s causing the particle pattern. Obviously, if our evaluatiori of llie condition 1s to be intelligent, it is not sufficient to say that there is a surface crack or that there IS a discontinuity under the surface. The exact nature of the condition, if known, \viIl determine whether rejection 1s warranted or lvllcther salvage is possible. Such kno\vledge niay, further, lead to the correction of processes to eliminate recurrence o f the condition in the future.

The magnetic particle pattern does not, in itself, tell us what condition is caus~iig it. Some of the typical characteristics of patterns producerl by surface and sub-surface discontinuities have been shown 111 foregoing chapters, and a great amount of information is ai~oays obta~~iabie from an observation of the indication itself. Zut to translate such observation into a thoroughly accul.ate identification of the rlefcct 1s not always easy. T h ~ s is the prolAem of interpretation.

PRIXCIPLES O F &lllGNETIC PARTICLE TESTIXG ISTERl'RI..XATIf)N; EV1LD:tTIOX AXD RECORDISG O F REStiLTS

4. OUTSIDE I<NO>~'LEDGE REQUIRED. The fully compete~it non An experiencecl inspeetor, \'.'el! equ~pperl with knowledge stlch as destructive testing engineer sliould be prepared to interpret indi. ut.lh~ed above, will !i!io\v a t the outset that certain defects :nay be cations wherever they arc encountered. I t IS true that in many esent and that oi'iers cannot be. His probiem of n~lcrpreiat~oii

appiicatiot~s spec~al-purpose units or controlled tests are set up to a t once narroxvcti to perhaps a very few possibilities, since the

find indications of specific defects, which, when found, need no story of the part ruies out the rest. Often he is actually look~ng

further interpretation. But to make a dec~sron as to tile identity or n specific kind of deicct, as. for instance, flakes in a large

of a defect without such built-in knomiedge calls for considerable orglng. On the other hand, in any given plant such as a foundry

experience and general iiiformation regarding the part. This infor. r forge shop the types of delects that can be present are usually

ination is not al~vaps available lo an otlierw~se perfectly competent learly understood, so tlial here agaln tlie problem is simplified.

inspector. To be able to producc. intlications properly does not in 5. T I ~ E OPERATOR. I t is very evident from the above considera- itself a t all qualify an inspector to rnterpret his results dependably. tions that the "operator", to xvhom frequent reference has been

In order to do this, a t least some, but preferabiy all, of tlie fol- ide in these pages, or the inspector or other person conducting low~ng points and sources of information should be possessed by, e tests, IS the Besr to the attainment of the greatest possible value or be available to the inspector : rom the use of magnetic particie testing. This is indeed equally

( a ) A knowiedge of the material from \!,h~ch the part is made e of any niethod of inspection where readings must be taken -\i.tiether liigli or low carbon steel, and what alloys are meters or mstruments, or pliotographic negatives must be ex-

present and in what arnount. To make use o i this loioxr4edge ~ned. The necessity for ~nterpreting and evaluating the observa-

he must already know sometliing of steeis and stcei-making, ns made by any nondestructive testing method necessarily intro-

and the character of the defects lilcely to occur in various duces the human element into the ~nspection results. Even in the types of steel. case of electronic metl~ods where the equipment is pre-calibrated

and set to give nutolnatic indication of a rejectable discontinuity, (b) A knowledge of the processing J11story of tlie part-w'hetiier setting the limits of acceptability involves human judgment, a s

made froni folled stoilc, or from a forging or casting; also does also monitoring the equipment to insure that i t continues to what machining operations-gr11idi11g, lapplng, etc. have operate properly. been applied to i t ; also the heat treatment it ilas received -wheLher hardened, carburlzed, ni t r~ded, etc., and he shotiid For the above reasons, the better equipped the inspector is with know enougii about sucil processes to be familiar ~ \ ~ i t l i the b:lckgrounri of ki~o\viedge and experience, the better mill be the defects each may introduce into the part. esults from the test. In a large plant, i t is not always possible to

Ic) A kno~vledge of how metals fail and \\;hat conditions are nd well qualified personnel to operate the magnetic particle units, likely to lead or contribute to such failure. nd if the training and esperience of an inspector is not adc4uate,

i d ) Past experience with similar parts. Such experience often e shouid not be charged will1 the responsibility of interpreting

indicates what to be on the lookout for m ~ d helps to ~deiitiip r evaiuating his results. I t IS not the intention to imply tliat all idifations must be referred to some higher authority in the eveiit

an indication \vllen i t is found. a t the inspector is not qualified to interpret what he finds. Where (e) Facilities for maicing destri~ctive tests on specin~ens con- any similar parts are b e ~ n g esam~ned, rules for acceptance and

taining indications of ~vlilch the cause is not clear. Cutting ejection can be laid rloxvn, since in such cases the types of defects up the part and eramining tlie defect in section often makes 'kely to be found can be predicted. But the indications that do not possible identification of siniilur ~ridiciitions \ohen they are fit sucli specifications must he individually uiterpreted. encountered.

( f ) A general knon.ledge of metallurgy IS extremely iieipful, if 111 most cases the operator of tlie test equipment is a member

not essential. of tlie pla~it inspection staff, in w111ch case the ass~stance of the

CI3AFrb-R 12 IPiTERI'RET:T:\'i'IOX, EVALU:YI'ION AND RECORDING OF KESliLTS

! i &fetallurg~cai Department and the Testing Laboratory should b But iurlher than this, accurate inte~pretation depends equally

I easily available to h ~ m to a s s ~ s t in the niterpretation o i indication on actual experience with typ~cal defects, and in order to gain such In some plants the responsibility for interpretation may be take experrence he must actually ~dentify a sufficient number of cases entirely out of the hands of the inspector n'lio makes the test, an so as to be familiar with the appearance of the indications the be delegated to the metallurgist, chief inspector or a qualifie various discontinuities re likely to produce. supervisor---except as pre-determined rules have been prepared €0

the inspector's use. 7. SUPPLE~~ENTAL TESTS. The absolute ~dentification of a given discontinuity involves the use of supplemental tests. and these tests

The test operator should always possess good v~sion, since he often i.equlre the actual cutting up of a specimen, or otller\vise must, in the first place, see the indications. As f a r as his mental probing Into i t to see \?hat thc discontinuity really is that lies a t equipment goes, he must he conscientious and have a temperament or beiow the surface to produce the indication. which can be relieu upon not to become careless nnder possible monotony. These qualities are usually found to a sufficient extent Many of these tests are simple and can be performed with a mini- in any good inspector in a plant malting quality steel or quality mum amount of lal~oratory or testing equipment. Others requile steel parts. Such a man will make a good magnetic particle inspector specialized equipment and experience involving the use of testing and, if sufficiently intelligent and alert, he should soon acquire the laborator~es and the metliocls of metallographical investigation. experience to ~nterpre t most indications. But he should not be

8. SIMPLE TESTS. To wipe off the magnetic particles forming an entrusted with tlie responsibi1it.y of working out new magnetizing indication is almost instinctive, and sometimes the defect, if a procedures for new and widely differing classes of parts made of surface crack, is quite readily seen, once the exact location has been different k ~ n d s of steel unless lie has or acquires addition;~l back- revealed by the indication. A low power hand glass is a most coii- ground along the lines of the subject matter of this book. To venient pocket tool to aid the eye in such a first check. If the surface this he must possess considerable native intelligence, initiative a is rough, or is covered with a light film of rust, polishing the area a strong sense of responsibility. where the indication appeared with fine emery cloth usually ren-

I i the inspector liimself is expected to pass on the acceptance ders the defect more r~s ib l e for study. and rejection of costly or important parts, then the best man available should be .used for the work. Another sinlple checli is to off the indication and again

apply the powder or liquid suspension of particles, t o see whether 6. SOURCE OF I<NOWLEDGE AND EXPERIENCE. In order to acquire the intlication will be reproduced by tlie residual field in the speci- the itnowledge and experience which is required ior adequate in- men. Since +he residual method is always less sensitive than the

terpretation of india~tions, various sources of ~nformation are continuous, i t is obvious that if the original indication was produced available to the inspector. Assuming that he is competent and \veil by the continuous method, the manner in which i t reappears by the ~nformed as to the niagnetiz~ng procedures and teclin~ques involved restdual method gives at. once some indication of i ts severity and in produc~ng indicalions, he should study the operations and proc- extent. As a matter of fact, this criterion of reproduction by the esses within the plant to becomc familiar with the defects typical residual field has in some eases been made tile basis of acceptance of those operations. I-Ie s'hould, further, read texts on such subjects or rejection of parts. There are, however, too many factors involved a s steel-mak~~ig, metallurgy, fatigue of metals and how metals fail for i t to be safe to conclude that if the indication does not reappear, in service. The plant metallurgist may have, or a t least can recom- no defect is present. mend, suitable books for this purpose. The neecssary kno\vledge, if not already in the inspector's experience, is not qu~ckly or easily Often times it is worth while, i11 order to confirm the indication, act~uired; and yet, the knowledge of what to look for and where to demagnetize the part and repeat the test from the star t to make to look for i t is indispensible to the ~ntelligent application of the sure that the ~nclication really does come back in the same form test and the interpretation of indications.

PRINCIfLES OF MAGNETIC PARTICLE TESTING . .-

9. B ~ x o c c ~ a n Mlcnoscom. This ~nstrument is an invaluabic aid in esamining surface discontinuities. It is vasy to use, moderate in cost, and does nor require a skilled technician. 3Iagnifications Trom 10 to 32 diameters are available with this instrument, and are adequate to resolve even e:cceedingiy fine cracks and other surface conditions.

A good light to illuminate t'ne surface just under the lens is essential, and usually cioanlng of the surface and polislung with very fine emery cloth is helpful. Sonietin~es light etching brings out the crack more distinclly.

If none of these surface examinations reveals a discontinuity which will :iccount for tlie magnetic particle indication, and the indication persists after the surface is smooth and clean, the pre- sumption is that the discontinuity is belo\\, the surface. Confirma- tion of this may he had iron1 the appearance of the powder pattern itself, under a giass or a microscope. The indication of a sub-surfack defect will appear diffuse rather than con~pdct, and is likely to be a bit fuzzy a t the edges.

10. FILING. USE of a file is perhaps the easlest way to determine the depth of a surface crack or seam. Either a flat tile used corner- wise or a triangular file mill readily cut to the bottom of a crack,

Fig. 1-File Cut Applied to Cracks in a Bar, with Magnetic Particles Re-applied.

398

or eise determine quickly that the crack is so deep that the specimen is unuse:ible. Reciieci<inp at the filed point \%,it11 mnpnetic particles is a mt?mls of knowing whether the file cut 11as gone to the bottom of the deiect or not. Such a test 1s shown in Fig. 188. One of tlie two cracks is obviously much deeper than Lhe other. T!le amount of particle surface-build-up had already 1ndic:ltcd this fact before the fi!e test xvas applied.

11. GRINDIKG. A slmilar test for determiiring the clepth and extent of surface discontinuities is to grlnd the surface of the part

Fig. 189-a) A Small Forg~ne. Ground at a Lap with Defect not Completely Removed. b) Cxplorlne Crack Indications on Line Pipe with Small Hand Gnnder.

399

CHAPTEIL 2? PRINCIPLES O F M-AGNETIC PARTICLE TESTING INTEI~I'RETATION, E~A1,UATIDS AND llECORDlSG 01' RI:SUI,TS - -

at the indication. A permanently niounterl grlnding wheel, or port- ases there is sufficient res~duai magnetism in the material to bring able wheel mounted on a flexible shaft may be used, depenciing on ut the indication ~vlien particles are reapplied, arid remagnetiz~ng whether tile par t 1s siiiall or large. 1s then not necessary.

Figure 18% sho\'s a small forging which has been ground a t a This removal of the defect can also be doue very rapidly by flame lap. I t is common practice a t many plants to investigate a11 indica- ougirig a s rn the case of conrlitioning steel billets and blooms. By tions of t h ~ s nature by such grinding. Shallow laps are usually suitable mnn~pulntion of the burning torch, metitl may be renioved ground out entireiy a t the time, so that the objectionable condition very rapidly in local areas a t the defect. By sratchrng the surface is a t once removed from tlie part. of the metal under tlie flame before actual biirn~ng commelices, it

Another very t~seful tool for mvestigating defects is the small is often possible to observe the presence of a surface craclt: and thus hand grrnder sho\vn in Fig. 189b. This tool is provlded with a num- see ~'heii it is all removed. However, rechecking with magnetic bcr of small grinding heads or burrs; which permit grinding a t particles is easy to do and is a better and safer precaution to niake the indication wvitli a niiziimun~ removal of metal beyond that strictly sure that the part is ready for repair or further processing. local to the crack,

Such grinding methods permit exploring a crack or other defect over a greater depth o r area than can be done by filing.

12. CHIPPING. A slill more rapid method for exploring cracks or other defects rn large objects is by chippnig. This is usually done ~ixith an air ch~sei, though sometimes a hand chisel and iiam- mer is more conveniently used.

Chipping methods are most commonly used on welds. castings and large forgings-not oi~ly that the ~nspector may determine the actual deptli of tile defect, but also t o remove the defect in preparation for re-melding or repair if this is permissible.

13. CHIPPING FOR REPAIR AND SALVAGE. Grinding and cl i i~ping methods, therefore, often serve tlie dual purpose of determining the extent of the defect, and a t the same time actually removing the defect from tlie part. As stated above, forgings are commo~ily ground to remove shallow discontinuities entireiy, if they do not penetrate below the dimensional tolerance of the part. Recheclting with inagnetie particles siiould be done to insure con~plete removal of such defects.

In chipping operations i t is useful to observe the nianner in \rhich the chip comes away from the chisel as the cut proceeds either along or across tlie crack. If the cracli extends eoml)letely through the ellip being removed, the c h ~ p will separate illto two parts. Wheii the cut extends below the bottom of the crack the c11ip ill not separate. A recheclt with magnetic particles IS still. however, easily applied and is a safeguard to make sure tint the bottom of the defect has bee11 renioved before welding reparr is beg~in. 111 most

14. DESTRUCTIVE ~ZETKODS-FRACTURING. In order t o derive the masimuni possible information regarding a questioned indieation a number of tests which invoive complete destruction of the part are most useful. Such tests cannot be regularly applied in the course of routine inspections, but they are invaluabie to l l ~ e inspector who is attempting to improve his knowledge. By observing the nature of the defects which occur in tile products for whieh he is responsibie, lie tlius increases l~is nbility to make reliable interpreta- tions in the future.

Sucli tests arc also frequently used after a part has been rejected because of an indication, a s a sort of autopsy to confirm, and thus improve, the judgement of the inspector \\rho made tlie rejection. This again lends to increased knowledge and confidence to aid future decisions.

One of the easiest of these destructive confirming tests IS to break tlie part or attempt to breait it, through the discontinuity. The part may be placed in a vise in such a manner that the metal OII one side of the cracli is held firmly, leaving the wst of the piece p~.ojectiiig from the vise so that, wile11 struck heavy blows xvitli a hammer it will tend to break a t the crack. Such procedure works best on iiard parts, but i t can olteii also be applied on relatively soft inaterials by usi~ig :I very heavy hammer blow.

Another way of conipleting the fr:~cture through a surface crack if the section is too large for the liammer and vise method, is to put the part in a large testing machine and appiy a load in sue11 :I fasl~ion that the pnrt tends to break a t the. cmclc. Such methods arc par- ticularlr effective fur studying tlie origin of a fatigue crack, or 111

Thts \rould ~ lo t 'he true of austentic steels, \sliich do not bnrome distinguishing between a heat treating and a grinding crack. The brittle at even very lo\%* temperatures. fares of n heat treating crack are likely to be oxidized \shereas a .. - ~ -~ .*

grinding crack will not sholv such osidation. 15. SECTIOXIRG BY SAWING. Very often it 1s desirable to examine Sometimes it is important to make sure that the fractur~ng oper- a crack in cross-section, and varlous methods for. sectio~iing a part

ation has not es%ended the depth of the orig~nal cracii. This cat1 be are used. If the material to be sectioneti is soft enough, the easiest done by heating the specimen containing the crack to a suflicientlp thing to do is to saw across the cracli. The depth and direction of high temperature (500° F.) that the surface becomes blued. Rrhen the c1.acli 1s slio~v~l 011 the section. TIIIS niay be done either with

Fie. 190--Fracture Through a Fatigue Crack that has been Blued by Heating a) The Bar Showrng the Fatigue Crack. b) The Fracture.

fractured at the crack, the original crack face \%-ill be blue, the rest of the fracture \sill not be. Fig. 1903 shows a fatigue test bar \%,it11 a number of small fatigue cracks. Fractured after blueing, the extent of the original crack is ciearly shonw. hTote the circular propagation of the craclc from the point of origm.

Another t r ~ c k that can be applied to facilitate this fracturing test, particuiariy on s~iiall parts, 1s to chill the part in dry ice and then im~iiediateiy fracture. A t these very cold temperatures, \\'ell beiow zero degrees F., most steels, even t.1iose \rIl~cii may be soft a t ordinar>- temperatures, \\.ill bcconie brittle ellough to break e:~sily.

a halid or poxver hack saw, or with a slotting saw on a milling machine, or with an abrasive cutting \\,heel.

Sometimes wheii large forgings, castings or other bulky parts are being investigated, a complete sectiol~ is difficult and costly to make, and it is simpler to renio\.e a p~ece of metal by one means or another so as to include a cross-section of the oack for its entire depth. Slotting sbws, chiseis, drills and various inge~iious devices :ire sometimes necessary to get such a plece out of the larger part.

One relatively easy technique is to use a core drill. This is a hol- iow cutter \vliich drills a core, usuallp about an i~ich or thereabouts in diameter, to as great a deptli as may be requ~red to reach the bottom of the craclr. This method of cutting out a section is called "trepann~ng", and is often used in the investigation of welds in heavy plate.

\i7hen it is necessary to make a cross-sectional cut on very hard materials the usuai san-ing niethods are either too slow or cannot be used a t all because the saws are less h a ~ d than the pelt itself. In this case sectioning is done by means of a high speed abras~re wheel or disc. In using this method great care must be taken to avoid heating of the part during the cutting operation. Heating a t the cut is very rapid, and even slight locai temperature increases may extend the depth of the crack be~ng investigated beyond its original estent. Also, such heat map change the structure locally, which may be n m:itter of importance. Frequently removing the specimen and cooling it under water IS necessary. &Iuch better IS to flow a stream of cold water over the saw and sspcclmen at the cut. Some wi~eeis actually operate under \v:iter, wiuc11 is untloubtcdly the best technique.

16. EXA~~IN.ITION OF THE Ctir SURFACE. So~iletinies these C I I ~

cross-sections can be examined n.itliout further treatment, eithcr directly with a hand giass or a l)i~~oeular microscope, or by the application of magneiic p;rriicles after a suitahlc mi~gnftizatlo~z. Aiore

C11AITER 22 1 INTEBPREXATION; EY.41.1iAl'lON ATD IIZCORDING OF RESI!LTS 1

often, however, it is nccessary to smooth the surface, usually done in studying the cross-section of a weld, and reveals the pattern by dl-a\v-filing follolved by some degree of polish~ng with emery made by the weld metal and the heat-affected zone in the base metal. clotli. Filing and polislt~ng sl~ould be in a direction parallel to 18. ETCHING CRACKS. When etching is used-especially deep the crack to avoiU dragging metal across thc discontinuity and thus etching-there 1s a danger that must he borne in inind. Hardened, obscum~g i t from observation. Foilowing such polishtng, magnetic quenched articles ~vliiclt have not been thoroughiy stress-relieved by particles may be applied or the surface examined under the binoc- tempering or drawing, but ~vliich are not cracked, may contain resi- ular microscope. with or without a light acid etch. duai stresses to such an estent that when the surface fibres are

17. ETCHING. Light etcliing with dilute n i t r ~ c or hydrochloric attaclied by etching reagents, these stresses may cause cracks to actd is often useful as a method for studying a defect either on appear in the surface as the stresses relieve themselves. Sucli crack- the original surface of the part or on a cut cross-section. Deeper ing, often spoken of as etch-crackiilg, appears on the etched surface etching brings out flow lines, which may be important (especially and is oftcn indistinguishable from cracks \~htcil may have been on foi-gings), and also reveals much information regarding a defect. present before etch~ng. Figure 192 slio~vs magnetic particle indica- ',3,$t<&,>> Figure 191 slioms how deep etching. on the section if an automotive ,:M.3$?.

Fig. 192-Etching Cracks on a Twlst Drill

Fig. 191-Deep-Etched Section of an Automotive Steerlng Arm Showtng Forglng Fold and Flow Lines in a High Stress Area.

steering arm shoxvs by the flow lines that the indication found with magnetic particles (arrow) was caused by a fold a t the upset boss on the original forging, and not by any subsequent machining or heat treating operation.

When differences in grain structure are also of interest, the mac- rostfucture may be brought out by etching the section with a soiution of arnmoniuin persulphate. This technique is sometimes Iielpfui

tions of cracks on the etched surface of a hardened twist drill, wli~ch were not present before etching.

Faiilire to recogntze this occurrence has led to false conclusions rcgarding the presence of cracks in such articles. If, therefore, the surface of :i highly hardened part is to be etched as a means of confirming the presenee of cracks or of investigating cracks revealed by magnetic particle testing, the specimen ~110uid first be stress- relieved by heating to temperatures which range from TOOo F. for hardened carbon steels up to as high as 1400' FI for some alloys, holding for a short time ancl then eoolii~g slowly.

19. &flcnoscoI>rc EXAMINATION. A great deal of information a s to the character and urtgin of a crack o r other defeci can be obtained by examining sections a t high magnifications by metallographic methods. Sucli examinations a r e usually beyond the skill of the inspector and a r e usually in the province of the metallurgical laboratory. The sections, af ter careful preparation, are exani~ned. either unetched, o r etched to bring out the gram structure, a t mag- nificztions \\,hich may range fro111 50 to 500 diameters or even lugher.

20. INSPECTION LIGHTING. I t should be obvious that satisfactory results froni a n inspection operation cannot be expected if the inspector is handicapped by poor lighting and cannot plainly see the indications he may be required to judge. Still, i t is surprising to note how many plants a1loi.r t h e ~ r ~nspectors t o work under lighting conditions which a r e unsatisfactory, despite tlie fact tha t results of the ~nspection may be of great importance, econoniienlly, to company operations.

Inspectors should not be requ~red to judge indications without fully adequate light. Among the requt re~ne~l t s of good inspection ligiiting are, first, tha t i t should be uniform from one ~nspection to another. a s fo r instance, between the inspection unit and the quality

and ourside sources of light whlch w o u ~ d interfere, can be minimized.

21. REcnRDs. f e r n ~ a ~ l e n t records of the appearance of rndiza- tions can be of great viiltle for a number of purposes. Records showing the i yp~ca l appearance of acceptable or rejectable indications of discontinuities a r c useful fo r the guidance of inspectors in the testing of iarge numbers of similar parts. A record of indications of discontinuities which $ire stibsequently investigated by sectioning o r other means is an essential port of such a report. Sometimes :I

pa r t 1s pu t back into servlce conta~ning a known discontinuity which experience has shown wili grow slol~dp. Comparison of the indication obta~ned a t the next inspection x ~ i t i ~ the records of the previous ones is a positive nteilns fo r checking such growth mte. And records may be useful in statistical studies of tine occurrence of different types of discontinuities.

22. FIXING AN II\'DICATION. Sometimes i t 1s desireti merely to fix a n indication on t h r surface of a par t to protect i t from smearing o r other damage during handling. T h ~ s can be done easily by the use of a clear lactluer. The indicirtion 1s developed on the part and a light coa t~ng of clcar lacquer applied. Spray cans of lacquer make t h ~ s a simple process. ...~. ~~~ ~ - ,

c o ~ ~ t r o l laboratory. Daylight varies from hour to hour, so tha t If the ~ndication is produced with dry po\vder, surpius polrder satisfactory lighting can only be obtained througliout the day should be carefully r e~~ loved so a s to get the indication itself in c i ~ r by artific~al l igi~t. Artificial l i g l~ t fo r magnetic particle purposes outline. A light lacquer spray can then be gently applied. shoold be white where possible fo r visibie magnetic particles. If colored, the color sl~ould be such tha t the maximum color-contr:ist If the ~ndication i s produced by the met method it IS necessary to of the magnetic particles will be b rougl~ t out by the lighting. For allow the bath liquid to d ry before applying the lacquer. example, when the lighting is by incandescent l igl~ts, the red mag- 23. LIFTING AN INDICATION. Very often i t 1s des~red t o transfer netic particles a r e more visible than the gray o r black. a n indication froni the surface of the par t to somc permanent record

Another Important reqoirement is that f~ighlights sllould be m~nimlzed a s much a s possible. Daylight, \\,heti strong, fulfills this requirement probably better than mly other, but unfortunately it i s not cons~ste~i t ly available in proper intensity. It is extremely difficult to prevent i~~gh l i gh t s entirely, particularly on par ts h a v ~ n g small diameters and cuwed surfaces, when artilicial light 1s being used. A properly [lesigned l igl~t ing system, l~omi ver, can do a very sntisfactory ]oh. \Vlten the mtlieations helng exmilined are made with fluorescent particies the same i.ecluircments apply, but are more easily met, s ~ n c e the light source (blaclt light) IS usually moveable,

place, such a s a report or a notebook. There are two good and simple lvays to do t h ~ s .

( 1 ) Use o! T I . ( L ? I S P ~ ? . F I L ~ Plastic T a ~ e YSeoleh Tnpe":. The three- quarter inch tvidth of clear colorless tape is prcferretl. If the indication is dry powder i t is necessary only to lay a s t r ~ p of the tape smootl~lp rlo\vn onto the indicution, nud press i t firmly along tile line of tlie 1nr1ic;ition. The powder 11oild-up should not be a l l o ~ ~ e t l to be Iicauy, a s better clelineation is obtained if the powder does not spread out when the tap- is pressed into place.

The tape can be immediately stripped off from the surface. and tlie indication will come away aiso, stuck to the tape. The tape can then be put onto the page of a notebook or a report. Duplicates, if wanted, can be made by repeating the process with a fresh powder indication, or by photographic printing of the transparent tape on sensitized photographic paper. Also, the tape car1 be mounted on glass and used as a slide for projection on a screen if desired.

If the iiidication is produced by the wet method i t is essential that no batli liquld remain in the particle build-up. If t h ~ s is not fully dry, the indication will "squash out" like an ink-spot, test and be of no value as a record. Two or three hours may be requ~red for the indication to dry thoroughiy. The drying may be hastened by carefolly applying a small amount of a volatile solvent sucli as naphtha or carbon tetm- chloride if the batli liquid was oil; or alcohol can be used if the bath liquid was water. Application and removal of the tape 1s then tile same as for dry polrder iiidications.

Dry powder records should be made from clean surfaces free of ponrder background. The lacquer is sprayed along the indication witli as little scatter a s possible, then built up in sufficient amount to glve a film thick enough to permit stripping off. Stripping can be done as soon as the lacquer is dry. The strip of removed lacquer can be trimmed to suitable length and width witli scissors or knife.

With wet metliod indications the bath liquid should be removed by a solvent or allo~ved to dry as with tlie tape transfers. The s t r~pping lacquer IS then applied as above.

The records made with stripping lacquer are likeiy to be better than those made with tape because the film is more transparent, and because the ~ndication has not been spread even a little bit by pressure, a s it may have been in the tape process. Also the particles are incorporated ~ n t o the lacquer film zind noL just stuck in the sticky surface, as in the case of the tape. This makes the record more durabie if it is handled very much, or for lonz-time records. -

A fevv articles wh~ch may be useful when applying tape 24. PNOTOGRAPIIY. Photography is one of the best ways to malte are a small pair of curve-bladed scissors for shaping the tape; permanent records of the appearance of indications, but takes more broad-end tweezers for handling the tape; and a wooden time and equipment than the tape or lacquer techniques. Also, if stick or lead pencil for introducing the tape onto interior precise delineatio~i of tlie contour of indications is important, the surfaces. These are, however, riot essentiai. tape or lacquer records lifted directly from the surface of the part

I t is frequently desirable to record not only the appearance can be more exact. A photograph, on tlle other hand, shotvs the indi- of the indications, but also their locations on a part. In this cation in its natural environment on the part in ~vhlch the diseon- case the tape sliould be cut long enough to cover not only the tinuit), occurs, anrl has this advantage over the lifted records. ~ndication, but also extend to a corner, hole, keyway or other Black and white photographs of parts contaln~ng indications change of section \shich may be used as a "bench mark". If sometimes require some ingenuity in "posing" the specimen and in so~ne of the inspection bath or dry powder is allowed to fall securing lighting tliat sets off the indication so tliat it gives a clear on this base p o ~ n t the tape \!.ill pick it up and thus reference picture and be a faithful reproduction of the indication on tlie the iocatiori of the indication on the part. resulting photograplf.

(2) Use of St?-ipublc lac we^. A clear lacquer is now availabie Use of tlie Polaroid eamera and film process is a most convenient in sprily cans which, when dry, can be stripped from the and qoiclc method of m a k ~ n g photograph~c recorils. The imniediatr. surface of a part, giving a lacquer strip similar in character availabilits of the picture nialtes it possible to make corrections in to the transparent plastic tape. This material 1s very easy to ligiltilig or positioning quickly. Several successive shots a11 he made use and glves a clearer film than does the plastic tape. This if duplicates are wanted. is a desirable character~stic when the record is to be duoli- cnted by photograpl~lc printing or nsed as 11 slirle for pro- 25. ELACIC LIGI-IT I'HOTOGRAPHY. CVlieii the fluorescent particle leetion. method has bee11 used, pl~otography of the results is the only way

-106 400

C a a m 22 INTERI'ILIfT.4'i'IO&, EVALFATlClN .kKD RECOItDING 01' IZESULTS

PRIBCIPLISS OF h1AGNETIC PARTICLE TESTING -- - he camera is essential to filter out the black light which viould

j affect the film. A fast pnnchromatic film IS preferred. The parts

I should be cieaned of all fluorescent bacicground, and set up in the dark room with the t\so 100 watt black spot lights placed so as to

1 some experience and cafe, good photographs of fluorescent indica- bring out the bfilliance of the fluorescent ~ndications :vith as little tions can be taken to make a s t r ik~ng record of t h e ~ r appearance, reflective high-lishting of the part as possible. A light coiored non- In either black and vhite or color. fluorescent background is usually deslrabie so that the black outline

i In addition to the technical skill required of the photographer, of the part sho~rs in silhouette against it. he is also under the obligation to see that the photogmpi~ neither exaggerates tlie size and brilliance of the indication, nor min~mizes

The exposure time varies greatly with the brilliance of the indication. With a C filter exposures vary from 20 minutes a t f/32 for

it. This requires good ob]ecti\ve judgment, since photography can in this iiistance easily give a false value to the appearance of the

heavy bright^ ~rldications up to an hour a t f/22 for fine indications of lorn intensity. If the thlriner K2 filter is used, the exposure time

indication, as compared to what the eye actually sees. is cut to about one half and the definition of the part is improved, In spite of the difficulties, photographs are often desirabie and hut undesirable highlights from reflected black light may come

the follo\ring outline, based on experience, will make the task easler through stronger. for the photographer who has not prev~ously attempted to take

1 pictures under black light. F i y r e 193 sho~vs diagramatically a To increase the definition of the part as a whole and separate i t suggested set-up for this purpose. A K2 or G filter over the lens of from the background, white ligiit may be used for a short time

g during the exposure, but the white light should he placed so that it

e i does not illuminate the indication areas either directly or by high-

!! & Since there is no practical means for pre-calibrating esposure, r: p . J

one or more test negatives should be made for each set-up. A g Polaroid camera is very useful for making these test negatives.

8 The negarive after normal developn~ent should show the indications li c i

solid black, but not spread wider than they appeared on the speci- 4i men. The rest of the negative should be thin, but with parts clearly :i. ci defined. It should be checked espec~alty for highlights mterfering ..,. ,. 3 with or resembling the indications. Such highlights can often be :,< ,,i moved, weakened or diffused, or eliminated entirely by re-position- .; i i :l

Printing can be handled normally, usually on medium contrast paper. The object is to produce a quite dark impression of the part as it wouid appear under black light, with clear white indications in the picture where fluorescent indications appear on the part. If color transparencies are to be reproduced later, the "color correction" techniques should be used, to hold the high contrast within reproducible bounds.

26. P O L A R O ~ FILM TECHNIQUE. Use of Polaroid Rlm has con- L siderable advantage in photographing fluorescent indications, since

Fig. 193-Diagram for Set.Up for Black Light Photography.

PRINCIPLES OF PAGNETIC PARTICLE TESTING

trial shots can be made and corrections applied in much less time than with ordinary film. There is much less experience beh~nd the technique for the use of the Polarold camera, but the following is a suggested procedure whtch has been successfully used.*

In a dark room two 100 watt black lights are positioned on either slue of the fiuorescent indication a t 24 inches distance and 45' angle from it. A 4" s 5" camera wvith a G lens filter IS used. Focus- ing is by ground glass camera back. Positioning of the camera is such as to secure a full size reproduction of the indication, although t l i ~ s may be altered depending on the size of the indication and the part. A 4" x 5" Polaroid film holder with Type 57, 3000 speed film is used. Exposure time is one to three minutes a t f/22 lens opemng.

The photog~aph is then copted using 55 PIN Polaroid film to secure a negative for making multiple contact prints.

27. COLOR PHOTOGRAPHY. Photography with color film is sometimes attractive, but esposure time and ligtiting are even more difficult to work out than with black and white film. The p'hotog- rapher must first have mastered the technique of blaclc arid white photography, after which the challenge of fluorescent pictures in color may be suffic~ent for him to make the attempt. I t is not, however, recommended as a method of ~ e c o r d i n g indications, unless fairly repetitive tests are involved, to permit standardized techniques.

28. SUL~IIIARY. After reading the forego~ng pages of this chapter the reader may have received the impression, perhaps a discouraging one, that the ~ntefpretation of magnetic particie indications is such a difficult and complicated matter as to be beyond the ability of the average inspector to master. I t has certainly not been the ~ntention to exaggerate the difficulties of interpretation, but rather only to emphasize the absolute necessity of his acquiring some experience before too much reliance can be placed on his judgment, and to point out tlie kind of experience required and the sources from which it can be obta~ned.

That satisfactory n~terpretation, for all practical purposes, is not too difficult of attainment IS borne out by the extensive use of the magnetic particle method after many years of evolution and development. It siiould aiso be repeated that the problems of interpretation and evaluation exist to ns great or a greater estent

'Reported by S. W. Gearhart, Birdsbar0 Corporation. Birdsbo~o, Penn. 1963.

112

C I l n f f E R 22 INTEIIPRLTATION. EVALUATION AND RECORDING OF IIESELTS -

in every other method of nondestructive testing, partly because service requirements are not always rsvell defined.

29. EVALUATION. \Ve have stated that "evaluation" of a defect means the determination, involv~ng a deosion by some person or persons \'ho are charged with that responsibility, as to whether the condition found by nlagnetic particles in a given part is cause for rejection, or whether the part may be either used as is or salvaged.

The decis~on is an entirely separate one from that invoived in interpretation, s~nce it is usually only where pre-determ~ned standards have been well established that parts are automatically accepted or rejected on the basis of the mere presence of a certa~n type of discontinuity. Furthermore a great many factors enter ~ n t o such a decision which are in no \my connected with the matter of locating, and then determining the nature of, a g~ven discontinuity.

30. THE PROBLEM OF EVALUATION. The problem of evaluation is one of wwrhether a g~ven discontinuity i s of such a magnitude, or shape, or is so located that it will endanger the satisfactory performance of the material or part in the service for wvhlch it has been manufactured or designed. Some cons~derations entering into the answer to this question are the folloufing:

in) What type of actuai service stresses 1s the part to be called upon to withstand, and rs!-liat is their magnitude, direction and duration-that is, \vill the stress be steady, or pulsating or reversing? Have these been positiveiy and experimentally determined, with reasonable accuracy?

(b) Where does the defect lie with respect to these stresses-is it in an area of high stress or iowv stress, and is it parallel to or a t some angle to the maximum stress direction?

(c) What is tlie nature of the defect and how\, severe a stress- raiser 1s it? Also, Itow is i t or~ented with respect to other stress-raisers that may be present due to design or construe. tion of the part?

Id) What is the importance of the part to the entire structure 01. assembly, and how serious wo111d the result be if the part failed?

(e) What is the history of experience with s~milar parts in s~milar service-have they a record of frequent or occas~onal

,113

C H ! . P ~ ~ I 32

IYTERPRFXATIOK. EVALGATION AND RECORDZSG OF RESUIXS

failurc, or do they never break? And, hare deslpn or fabric; the pre-analysis of which available methods may have been tion techniques to nnprove or deteriorate thl inadequate or impractical. history record?

(d) A loading applied for any reason, in service, difierent from It s&7rcely tile scope of this book to discuss a t grea that designcd for.

length these and complex factors, since they vary 1n n a t ~ r and importance ~,i:h each applicatiox~ and desimi. It is rather the (e) Vibration. pul.pose here merely to point out some of the more obvious consider- ( f ) Overstress due to accident. ations as they may lnfiuence the end result of inspection with mag netic particies. 33. STREI~GTH OF ~IATERIALS. The use of a "factor of safety"

in design is a device for alloxsing for unicno\vn or unpredictable DESIGX, ~t one of A. V. de Forest's favorite co strcss concentrations or operating conditions. When weight is not

nlents tllat many mctye failures of structures and machines are d a primary consideration the factor of safety may be high-the part or structure m::~ be made five or ten times as heavy as the best alculations the designer has been able to make indicates that it ceds to be. -4 large factor of safety increases costs, however, and

are ti~emsetves unsuitable for such service. still does not necessarily make sufficient allowance for stress-raisers. w h a t is meant is failure to evaluate with sufficient esactness th Fatigue cracks continue to occur in massive forgings and castings,

due to locally higii stresses resulting from the presence of stress-

Even with the extensive knoudedge now available of the i,ehavlor thus produced. of materials, and with the information obtainable by modcn~

>fodern of stress analysis make possible today's desi methods of stress analysis, a designer nt1w.t at some point assume by locating serious stress concentrations in ndx~ance; and design C n value for the strength and other properties of the materials he

decides to use. He must--or should-take into consideration the belo\\. tile danger point for the materials in question. 'cry difierent behavior of these materials under different types of

impact. And for services where weight as in the aircraft and missiie fields, he has been made of such stress analysis tools, the designer still c

only design for eaieulaterl or field-test-measured stresses, or ma as closely as he dares to this assumed strength, cutting his

reasonable allo\~ance for foreseeable overloads. Stresses in service f safety as low as possible. In the deslgn of missiles: utiliza-

are sometimes raised far above tilose for which the part is designed, of the strength of materials well up tolrard the ultimate

by the designer's controi or his ability to foresee is not uncom~non. I t is therefore of the utmost importance .engtl~ of the material actually entering iwto tfie st ,-uct~dr~

and provide for. as near as possible to the strength cnpability lte has usszcmed, Some of these conditions are listcd belolv: isers be not present

(a) ~h~ presence of notches such as cracks, scratche Magnetic particle testing, along wit11 other nondestructive testing etc., which act as stress-raisers. methods, is one of the links in this design chain, since it provides

(b) ~ ~ ~ i d ~ ~ l stl-esses, remaining from fabrication, heat treatin a means for insurance against stress-raising or other meakenmg onditions for ucliich the designer has riot or could not allow. I-Iom- or assembly processes. ver, intelligent evstloation of a condition revealed by niagnctic

tC) Unkllown distribution of stresses due to shape or size, fa articles must he preceded by knowledge and understanding of the

414 415

PRINCIPLES OF DIAGNETIC PAIITIC1.E TESTING

extent to which the designer has coz~nted o n freedoln of the finished port from such a condition.

34. OVIZR-INSPECTION. The purpose of magnetic particle testing as applied to materiais and parts is, therefore, to help make sure that the finished articles are as good as they need to be to do the job for which they are designed; but as there is in general 110 need for them to be any batter than such considerations demand, intelligent evaluation must be on guard asalnst over-~iispection.

A condition which is consldered objectionable in one part for a particuiar service may not be a t all objectionable in another part intended for a different service. Examples could be cited in which parts containing defects of a certaln size and shape have been placed in service without any detrimental results, while defects of a similar size and shape in other parts, more highly stressed perhaps, or in more critical locations in the same part, would have been consldered damaging enough to have warranted rejection. Evalua- tion: if intelligent, must therefore come down to a consideration of each individual case in the light of all possible knowledge as to the design of the part and the service for which i t is built.

Avoidance of over-emphasis on magnetic particle indications IS,

therefore, important-and such over-emphasis is a definite tendency where compiete information and sufficient experience are lacking. To "play safe", inexperience will cause the scrapping of many good parts, since to have the courage to accept parts containing discontinuities requires knowledge and the intelligent acceptance of the responsibility involved.

35. GENERAL EVALUATION RULES. Everything that has been said in this discussion thus fa r has emphasized the fact that general rules for evaluation of conditions revealed by magnetic particles cannot be laid down. And general rules are not necessary when the evaluator is equipped with adequate knowledge and experience. Nevertheless, inspectors are somet imesal l too frequently, in fact --called upon to make decis~olis regarding the seriousness of a defect when they lack such adequate information.

As a guide for inspectors, a few bas~c considerations-not rules- may he stated which, even though more or less obvious after a thoughtful anaiysis of the subject, may still be of some help when set forth as they are below:

416

IXTERPRFXATION. EVALUATIOX AND RECORDING OF RESULTS

\ (,a) A defect of any kind lying at the sttrface is more likely to be harmful than a defect of the same srze and shape which lies wholly below the surface. The deeper it lies below the surface the less harmful it is.

\b) Any defect having a principal dimension or a principai piane wliich lies a t right angles 01. a t a considerable angle to the direction of principal tension stress, whether the defect is surface or sub-surface, is more likely to be harmful than a defect of the same size, location and shape Iying parallel to

\ the stress.

& Any defect which occurs in an area of high tension stress must be more carefully considered than a defect of the same

%\

size and shape in an area where the tension stress is low.

Defects which are sharp a t the bottom, such as grinding ' "u raek , for example, are severe s t r e smisem and a,, t he rc

fore more likely to be harmful in any location than rounded \ defects such as scratches.

( ) Any defect which occurs in a location close to a keyway or B fillet or other design stress-raiser is likely to Increase the effect of the iatter and must be considered to be more harmful than a defect of the same slze and shape wliich occurs away from such a location.

36. PROCESS SPECIFICATIONS. The problem of evaluation has been solved in another way in certain large plants where many parts of the same design are manufactured, and where acceptance and rejection must be expedited in the interest of production output. Qualified engineers llave taken each part of a given design, considered the defects whlch mlght be encountered, and set up stund- ards for thnt part ieula~ part which can take into consideration once and for all for that part all engineering and service requirements. Such specifications should include an exact statement of how the part is to be magnetized, the sequence of magnetizations if more than one is required, the amount and kind of current to use, and any other necessary points to insure that the inspection is carrled out in identicai manner on all parts of the type, no matter by whom or on what kind of a unit. Rejectabie and acceptable conditions can then be spelled out and acceptance or rejection becomes "automatic." But if the deslgn of the part is changed, resulting perhaps in a

PRINCIPLES O F MAGNETIC PARTIC1,E TESTING

different stress pattern in the part, the test specification must be revised.

37. SUMMARY. T ~ I S discussion of the problems attending the evaiuation of the results obtained by use of magiieiic particle testing has sought above all to present ni propei. perspective the part which the method should play in the business of produc~ng better, safer and more reliable machines and structures. The concept that a "defect is only a defect when i t affects tikc useability of the particular pait in which it occurs," is fundamental.

This concept implies a further principle mh~cli should also be appreciated, namely, that this test method tas indeed any other) should be used only when the inform:ktion it is able to give is significant and critical from the point of vie\%- of determin~np the suit- abiiity of a particuiar part for its intended service.

One additional thought is also worth keeping well in mind, and that is that the most valuable use of magnetic particle testing, as well as any other nondestructive testing method, both in the inspection of new materials and or materials that have been in senlice, lies not in the rejection of unsuitable parts, but in helping to locate the reason for the occurrence of the defects which it detects: to the end that processes can be corrected or designs improved so that future occurrence of similar defects can be avoided.

i. INDUSTRIAL USES OF A ACNE TIC PARTICLE TESTING. Magnetic particle testing, with ~ncreasiiig usage over 35 years, has beco~ile a standard testing praciice 111 most sections of industry where iron or steel is made, fabricated or used for important end use. Its applications are so numerous and varied that it would be difficult to give esampies of a11 of its industrial uses.

I t is the Intention in the follo\v~ng sections to give a brief outline of the several principai purposes for which the method is being used, with a few esamples of interesting. or uii:isual applications.

2. CLASSIFICATION OF XIAGNETIC PARTICLE TESTING APPLICB- TIOSS. The follo\vlng is a list of the pr~nclpal industr~al use-areas for the method:

(a) Final Inspection.

(h) Reee~riiig Inspection.

(el In-Process Test and Quality Controi

id) Maintenance and Oserhaul in the Transportation Industries.

je) Plant and Machinery Ivla~ntenance.

( f ) Testing of Large Objects and Components

3. MAGNETIC PARTICLE TESTING FOR FINAL INSPECTION. His- toncally, when the method \\-as new, final inspection was the first area in which it was used. The test was applied to finished articles before they were shipped to the custonier, to make sure that the product shipped was not defective. Customer dissatisfaction, costly returns and adjustments. handling of replacements and reworlt were all reduced by this means. At the same time reputations for quality products could be built up. One man~ifactu~er of tool steel was so sure of the quality of his product after instal l i~~g the method that he publicly advertised tliat he would replace without cost any bar or billet in ~~:illch a seam was found. 411 this was--at tliat time -suGcient incentive to add the relatively inexpensive testing pro-

PRINCIPLES OF MAGNETIC PARTICLE TESTWG -

cedure of magnetic particie nondestructive tesling. Finai inspection \yas a necessity anyhow, and this new procedure was a refined tooi.

Today the empliasts has shifted toward in-process testing, so that defective material and parts can be culled a s early as possible, and before final inspection. In many instances final inspection is, hoxv- ever, still a must. Thls is especially t rue with parts on which a final operation may be the source of surface defects. The iargest group of products in t h ~ s class 1s pree~sion l~ardened and ground mach~ne parts, such as gears, splined shafts, crankshafts, and innumerable jet engine, aircraft and automotive parts as well as many other articles of this type.

One good example of final ~nspection with magnetic particles is the case of hand tools such a s hammers, axes, ch~sels, cutting and planing blades, etc. Heat treatment-hardening and tempering- followed by grinding of edged tools, is the fii~ai operation in their manufacture. Presence of grinding or quencl~ing cracks is highly objectionable for safety for the user, or early failure of the tool. Final inspection, usually with fluorescent particles. can guarantee absence of cracks in tools shipped to customers.

In the automotive and aircraft manufactur~ug fields, many hardened and ground precision parts are used. These include gears, valves, steermg Icnuckles, spindles, axles, crank- and cam-shafts and many others. Such parts are commonly given a final ~nspcction

Fig 194--Semi-autornatrc Unst for the Final Inspectton ot Automotwe Steerlng Knuckles.

before assembly. Figure 191 sho\vs the loading end of a sen?]- automatic unit used for 100::b final inspection of auton~otive steering knuckies. The parts are belng magnetized circularly between the heads of the unit shown. A stainless steel (non-magnetic) conveyor earries the magnetized parts under a curtain of fluorescent partielcs in water suspension, and then to a i100ded inspection booth where they are examined under blacl; light. Subsequently the parts go tl~rough a second station, are magnetized longitudinally and bath reapplied, snd again inspected. The parts a re hard and have good retentivity, lending themselves well to t l i ~ s version of the residual \vet method.

4. RECEIVING INSPFXTION. In piants where many finished or semi-fin~slied parts are purchased from other manufacturers, the function of receiving inspection is to make sure that such parts, when they are received from the vendor, are suitable for further processing or use. When large numbers of a given k ~ n d are invoived in a iot, the inspection tests a re often made on samples talten from the lot according to statistically prescribed rules. If the sample lot sho~ivs defects beyond a set tolerance limit, the lot may be inspected 10Of5. Extremely critical parts are inspected 100% in any case. Magnetic particle tests are in many cases part of the rece~ving inspection program, whether or not the specifications under ~ v h ~ c h the parts are purchased have required the manufacturer from wi~om the parts were bought to make such tests.

Rece~vmg inspection, l~owever, encompasses much more than the checii-testing of parts receiveti from outside sources. Testing of raw material \\:hen rece~ved, before any work is done on it, is a very profitable way in which to weed out any initially defective material. Such inspection of ~ n c o m ~ n g raw material is really Ule starting p o ~ n t of a program of in-process testing. Any work done on material or parts already defective is a costly waste of time and effort. Mag- netic particle testing is used extensively on incoming rod, bar, forg- Ing bianks, rough castings, etc.

As an example, in the case of coil spring manufacture, such a s valve sprlngs, bars or \rzire, ~od-ends are tested for seams with magnetic particies prior to drawing the spring mire. Seams are particulariy objectionable in \\,ire for t h ~ s purpose, since they iead to early fatigue of coil springs. Cold-heading stock 1s another esample m e r e seams lead to defective bolts, since the heads urill split on up-setting. In this case the rod is received in coils, and these

Fag. 1 9 S S p e c r a l Automatic Unbt lor Testvng Rod Ends

Fig. 196-Scml-autcmatic Unit tor the lrispcction of Srnai! Castings and Farglngs.

.I$+

grintirrs. Tcsis ;ire co~icc~itr:llrcl on I I ~ W set-~irrs. 11, m:iki siirc i l i ; ,~ tl;~? frc:rl : I I I ~ c ~ ~ i ~ l a ~ ~ t ~ >IIX; co~.r(ict., :!i;il the rirc~]i<!i. x r ~ t \si~~:,ji !s hc11,i: uscii, :inti i11:li tiie s c i - ~ i i ~ I.; !Jruttllcili.g I I O cracks. ' \ \ : I I~II !he timi' fiir iil.c.;~i~ig illc rrlleei II:IS c~rlllr! 1111 ;I ni:lcliil~i., l!if t ~ t s t i ~ i . ~ ilnit IS ag;1111 011 i i ~ e j n l ~ to ni::kc. su rc 111;11 "Irl:ii!c(l" \v!i<:ci;: :Ire no: c::iislny crziclis ;it ili;it st;lptS. S11cl1 :I j~i.trtrrilni rctluccs losses i'r0111 g1.111~1iiig cracks 10 a vcry I I I W pcr.ecli1:txi,. Fi r i i rc I!?: sfio\rs soi i~e craciien pe:irs, \\.itti ~u t i i~ i t i r ind proclncrd \vltll i l i ~ i ~ r i ~ s c e ~ i i n1:iglict.i~ nilrtlcles.

Fig. 197-Fluorescent Magnetic Particle Indications of Grrnding Cracks

F i ~ u r c 193 s1iou.s art iniit~e!ioi~ h;?rrielilng nprt.:itir~ii hi!ilig c:~rrii!d out oil cilnvcyor \vheers. Af te r liigli-frr!iju~?~icv i r ~ ~ l u u i ~ o n l~en!.irig. t h ~ y a r e q ~ e i l t l ~ e t l 111 water. Cracks sornctlmc~s i c s ~ ~ l t from this severe trc:~tnienL and m:igtu:tic: piuticlc tcsting is ap!)lictl 101V: :is tile i ~ e r t o])er;ition. The iliscrt si lo~vs cr':~cl;s liidicni~:i~ by fluorcscent magrletic p:~~.iic.lcs rrn a siniil:u'ly Iiarr11:neil part .

At the upper en11 01' ilie size range, FiE. 1!)!1 s11o\\.s t l i ~ '100'1 prcriluction i r~s)~t?ct ivi~ of se:imlcss tulles. r:inging from ;i 10 10 i~iclic.: ill di:inieter R I I I ~ IIJ) to '10 i'6:c<t I Q I I ~ . Tl~i: iltttorniillc illlit. maglit.1izes the tube Iiy riiis.xiiifi di~.ect current t!~rough it \vliiIe fIuoresct,~it ni;ic- nciic p:irt~clc? 1intI1-n xv;~ic? suspcnslon-1s : ipli i!I . insr~i:clii~n under I~!;lcli light reveais surf;lce sc;lms, ioars :unrl tither il:irn:+ginz c lef t~ts . This is r~;;tlly a fiilai r ~ i s p c c f i o ~ ~ . e t in l i~~fi ;it tlic em1 of tlic

CI(,IP~.TIL 23

PILINCIPLES 01: hlAGNETIC Pl\RTICLE TESTING INDI!SI'RIAI, AI'I'I.IC.1'L'IOKS -- -- -.

Fig. 199-Magnetic Pariicle Testing ot Seamless Tubing on an Automatic Unit. a) Loading Side. b) InsDection Station. c) Fluorescent Pallicle Indications

Fig. 200-Sixteen Foot Long Magnetinrig Unlt for Diesel Crankshafts, Installed on Overhaul Shop.

- . . or Spiral Seams. Fbg. 201-Typical I-atlgue Cracks I" Gears

PRINCIPLES 01' MAGNETIC PARTICLE TESTING

The safety of a i rcraf t operation today has become something generally taken for granted, and arrline aircraft acc~dents a r e practically never traceable to failure of structur;ll or cnglne components. This situation was not achieved, anri is not malntdlned, wit l~out the most intense and painstaking tests a t intervals dictated by ex. perience. Every available method of nondestructive testing 1s employed in the elaborate facilities fo r engine and aircraf t overhaul \i4ricl1 the airlines maintain. Tile intervals between overhauls is measured in hours of englne operation, and have been lengthened a s designs and testing procedures have been improved. Spot tests a t shorter intervals, without dismantling the aircraft, a r e made on critical components such a s landing gear parts.

Magnetic particle testing and fluorescent penetrants play a large par t in this program. Figure 202 shows the magnet iz~ng operation on a compressor disc of a l e t enmne, anri Fig. 203 tlre examination

Fig. 202-Specbal Magnetizung Unit for Jet Engine Compressor Discs. Employrng the Induced Current Method.

of a compressor rotor assembly for transverse defects in the blades. Both operations a r e employing magnet iz~ng units designed especially f o r these parts.

Figure 204 shows the use of a special portable magnetizing unit in making tests on axc ra f t par ts in tlre field. I t is sho~irn belng used to examlne a helicopter landing gear, with tlre dry powder method.

Fig. 203-Speclal Magnetiztng Unit tor Aircraft Gas Turbine Compressor Blading-Assembled.

(Courtesy Juckvinville Nuvni Ai r L o t i o n )

Fig. 204--Testing Landing Gear Components of a Helicopter on the Field, Using a Specral Portable Unit.

Automotive overllaul maintenance programs Ilave becn particu. l ady effective for keeping truck and bus fleets and OK-the-road equipment in safe operating condition. k1ang- operators of iarge fleets of vehicles have installed shops for tear-down, inspection and repair and re-assembly of englnes and veh~cle components, follow-

Ftg. 20jTruck Engme Connectlng Rods and Caps Wlth Fatlgue Cracks. 1

Fig. 20&-Testing Truck Front Wheel Sptndle with Portable Untt. 1

i ' 1 i d l ' ~ t . K 23 IYDI'S7'111Al. API'LIC:\TIONS - - .. .-

ing the plan used by the airlines and railroatls. The result has beeti a large reduction in s e r ~ i c e failures of equipment, and the cosily accidents and delays resulting from them. Figure 205 ~ 1 1 0 ~ ~ s t\\w truck englne connecting rods rv1i11 fatigue cracjis \\.ell at1v:inced in both rods and caps. Figlure 2116 sho\\'s the coil of a portabie magnet iz~ng unit belng usetl to test tlle f ront wheel spindle of a truck ~ y i t i ~ o u t removing it from the ;issemUly.

7. PLANT AND IYIACHINEKY &IAINTEN.~XCE. Planned inspection programs a r e also widelj, and profitably used in l ieep~ng heavy equ~pinent in operation \sit l~out breakdolwns \rshile 111 use, \\rhich can be very espeiisive and often unsafe. Punch press crankshafts and frames, because of the sudden and severe stress applications, a r e particularly vulnerable to fatigue failures. \Vitll small portable magnetizing units, checks can be made of known o r sus~~ec ted trouble spots; and f:ltig\ie cracks found early in t h e ~ r progress, nxithout dismantling the press. IVl~en the cracks are found in time, repair by welding is in many cases easily accomplishcri.

The plant safety program also makes good use of portable magnet iz~ng units. Crane h6oKs become \vork-hardened on the inside surface of tlie hook where concentrated lifting loads are applied. Fatigue cracks develop in thls area and often propagate quite rapidly. Testing all crane hoolis in a plant a t intervals of a few weeas or months, depending on the servlce they a r e subjected to, is a safety rcqulrement in many plants. The lifting fingers of fork troclts a r e also oergr vulllerable to fatigue failure. Figure 207 sliolvs the magnetic particle test being applied, and the fatigue cracks found a t tIic,i~iside bend ladius of tlie lifting member.

The n~agnet ic blading, shaft and case of steal11 turbi~les have long becn profitably esamlned for ~ncipieiit failure a t planned down-times. Xagnetic particle testing with portable equipment has played :in important par t in detecting fatigue cracliing in power- generating steam lu rb~nes over a period of 30 years or more. Figure 208 sho\t8s a trailer contrun~np magnet iz in~ equipment being lowered to the turbine room floor. The spindle h:is been lifted out of thc case and mounted on teniporary bearlngs so that i t can be rotated \vhile blading and shaft are tested with portabie magnetic particle equipment.

8. T~s'rixc Ol, L:II'~GE AND 'IEAVY AI~TICLES :\NU COEIIPONGNTS. One or tire cliaractcristics of magnetic particle t e s t~ng lohich makes

CHAPTER 23 PRINCIPLES OF MAGNETIC PARTICLE TESTING ISDI:Sl'IlIAL AI'I'LICATIONS - ~- --

- Fig. 207-Testing the Lifting Fingers of a Fork Truck. . .

Inset: Close-up of Fatigue Crack Indication.

Fig. 208--Preparations for Testing Steam Turbine Spindle and Blading for Cracks.

i t anplicablc to sucli a i:lrgc variety of industrial telitliig problems, IS the fact t11;it size and shape a r e in 110 case a deterrent to i ts use. Tiny nuts and bolts can be inspected equally a s well a s large ma- chi i~es and striictures. The :~utomatic testing of huge billets welgh- Ing a n u n ~ b c r of toils on masslve equipment ilas bcen described in Chapter 19.

Also described in Cliapter 19 was the al~plication of magnetic particle testing to the inspection ol' solid fue l roch-et engines and motor cases. Figures 209 ant1 210 i l l ~ ~ s t r a t e tlic cquipment used fo r the testing of the P o l a r ~ s motor cnscs. Figure 209 s i i o ~ ~ r s t h e unit for testing t h e case, and Fig. 210 shows t h e testing of the domes and nozzles.

Fig. 209-Speckat Unit Designed for the Inspection ot Solid Fuel Missile Motor Cases.

In the case of l iqu~d fuel rocltet engines, th rus t unils, fuel tanks and all pumps, piping and tubing a r e carefully inspected wit11 every possible kind of nondestructive test. The problcm here 1s not one of guarding a g a ~ n s t fatigue cracks to cnsure long life, slnce the rockets to date give only "one-shot" scrvlcc. R e r e the objcct of the inspection IS to find and e l i m i ~ ~ a t e any defect wl i~ch could cause the assembly to malfunction in any way whcn fired. Magnetic particle testing is used on all fcrromag~ict ic pa r t s lo make surc that no crack o r other defect is present which \vould iower t11c strength of the par t

PRINCIPLES OF MAGNETIC PARTiCLE TESTING - below that assumed b y the desrgn. Figure 211 is a vlcrr- of three th rus t units of a Ilquid fuel rocket engine, seen f rom below, mounted

Fig. 210-Specral Unit for inspection of Domes and Nozzles. !

(Cuuriess Rucli~vlrne i ~ i v ~ r i ~ n . North ~ i r n c n c n n .%rlncion Cor~~orn t ion) Fig. 211-Thrust Units ot a Liquid Fuel Rocket Eng~ne,

Mounted on t h e Test-Firlng Stand. - I

on the test,-fir~ng s t a ~ i d . .And Figure 212 s:~o\!'s t h e m a g n c l ~ c particle test being applied to lbutt \!'el<ls rn a cylin?~.rcoi lnot~il~clI:iml~er. Pernixiient magnet yokes ~viti i dry l , o \ ~ d ~ r arc? being i~scri. These m e most convenient since there :11.e 110 e l ~ c t r i c C ~ I J I C S 111. 1)0\1.(!1. j?:lrlis r e q u ~ r e d ; and egective bec:tose oiily a very sn1:ill scciioi~ of' n,c.ld 1s rnspected sit one time, and thc yoke glves a s t rong eolicentrnted iielil for th i s purpose.

(Cuurioy i lockeidyne Divlavon, NurlL Amsiican Ari i . l~on Cor!#orillsunl

Fig. 212-Magnetic Particle Testing ot Butt Welds on a Chamber ol a Liqund Fuel Rocket Engsne. Ustng Permanent Magnet Yoke.

9. SOME UNUSUAL SPECIAL APPLICATIONS. The versatility of the magnetic pilrticle method has foi~nil fo r it some interesting sinci out-of- the- t r rdi~~i i r~~ ul~lilic:~tic~ns. One of tliese is the testing of racing car eiiglnes arid c a r parts before tlic annual h lemor~a i 1);iy "500" race a t 1ndia11:tpolis. This year-19GG-marlis the :tot11 in wl i~ch cars lia\'e been tested wit11 magnetic particles a t the tr:lcli. Some critical parts, sucli a s the stecriiig parts, cranlish;~f'ts, co~ i~ icc i - 111g ~ o d s , etc., must 1);iss magnetic p i ~ r t ~ c l e I I I S I I C C ~ I U I I 1~ef0i.e tile cars a r c ~ ) c r n ~ i t t e d on tllc I IYI~K. 111 the e i ~ r l y days of Llio ;~j)l)lic;iti~tn of inagnetic particle tcs i i~ig , some i l r ~ v e r s olr~ecletl it) tlic icsi cvr!il though in;iily i ~ c c l d ~ i ~ i s Iiil~i J ~ ~ C \ , I U I I S I ~ resi111~!d f'r(1111 stt!el .~~ig 11:~i-i

fai1i11.e~. T i ~ e ~ r fcal. \\,as t I i ;~i c~, ;~cl is \ i r~~ul t l be f't1u11<1 Lli:ii \<'~11i1~1 l<t!aiJ them oui of tho race. 'l'111s ;~ttitudt! 118s cli;il~gvil e r~n i~~ie tc ly , ; ~ i i i l

the? Crec,loi~~ ~'I.OIII f i t i l i ~ v ~ of )i;iris ~ I I I ~ I I I K t l ~ c rirc<! ~ ' ~ I I L ' I I 1I1t. l(*si.?

,135

PRINClPLES OF IIIAGNETIC PAIITIC1,E TESTING -

Fig. 213-The " 5 0 0 Race at indianapoiis Speedway. Start of the 1965 Race

hcip to assure, with increased chances for completing the race, has now made such tests welconle.

In one recent race the record favoring such testing is inlpressive. Of 22 of the cars qualifying, only three had not had their engines and transmissions tested. Of these three, two dropped out of the rat+-one a t 66 laps because of drive shaf t troubles, and one a t 80 laps due to gear failure. Nine of the first ten cars t o finish liad had t h e ~ r engines and othcr critical par ts inspected, and in seven of these, craclced parts liad been found and ~eplaeed.

Another ~nterest ing application was the lnspection in the field of a 24 inch gas transmission pipe for a three mile crossing of the Mississippi River. Although the pipe had already been inspected a t themill, extraordinary steps \\,ere taken to make sure that no failure ~vould occur in this criticai servlce. In order that all sourccs of possible failurc \{,ere elim~nated, magnetic particle testing was used a t the site to inspect the entire outer surface of the pipe. Ey means of magnetic leecll contacts spaced ten feet apar t along the pipe, 1500 amperes of direct current were passed to magnetize the plpe circularly. Figure 214 s l~ows a close-up oC this operation. Dry powder was uscd, and all indications were ground out with a small hand

431;

gr~niler. See Figure 189-h, Chapter~22. If the discontinuity could not be removed \i,itliout golng deeper than .060 lnch (plpe wall was 0..500 inch thick) the sect1011 containing the defect \vas re~ected and cut out of the line. Defccts fouild consisted of cracks, laps, laminations o r other breaks in tlic surface. A total of 13 sections, from three Inches to twenty inches in length were found to contam rejectable defects. In addition t o magnetic particles, radiograph)' and ultrasonics were also used to cl~ecli the melds and the \\,all

Fig. 214--Ciose.up of inspection of 24 lnch Line Pipe, Three-Mile Miss~sslpp~ River Crossing.

Testing of steel billets for more accurate conditioning beforc they a r e rolled into fin~shed s ~ z e s and shapes has already been described. An unust~al application 111 one plant IS the testing of 2"xY billets of a nickel-~ron alloy, used for drawing into fine ivlre fo r the electronics industry. The wire, used for lead-.islres into enc;~psulated devices, must be free of scams to avold leakage of a i r into the capsules. The alloy was suHiciently magnetic that indications of seams cotild be obtained on the billets. By carefui removal a t the stage, the result-

ing \\,ire \r.l~en drawn do~r i i to very small diameter, \\-as f ree of any seams.

In the 1:lyirig of 21100 miles of telcpllone cable bct~veen the main- land and I-Iaxvaii, m ~ d aiso a trans-Atlantic cable, magnetic particle testing played a ]>art in insuring troublc-free performance. 111 this length of cabie, 114 repeater stations Toere required. Tllese are complex eiectromc units, incorporated into the cable, required in long transmissio~i lines to keep tlie s~gnn l s clear and separ:ite. The dev~ces a r e protected mitli numerous layers of ~nsulntioii, water- proofing and protection from physical damage. Structural strength without sacrifice of flexibility IS given by a serles of hardened steel collars. These rings nrere inspected with f luoresce~~t magnetic particles fo r seams, splits, suriace nicks o r scratches, and any minor surface irregularities. Defects were found in sis to eight percent of the rlngs !

Fg. 215--Testing of Protective Steel Collars on the First Trans.Pacific Telephone Cable. Collars are Magnettzed on a Large Diameter Central Conductor,

Figure 215 silows thr! method of magnetizing, uslng direct current. Four rings :at a time werc placed on the central cond~ictor for ni;igneliz~ng and ~nsl>cctlun \+fit11 fluorescent maglietic particles. At the right a r c slio!vii fotrr rrngs with rlidications of surface seams o r cr:iclts.

One npplicatioli involved the use of the method to detect rlefects in a iloii-magnetic ii1ater1:ll. Cast aluminum rotor l ~ a r s for electric ~iiotors were inspecter1 for nllernal voids using Ruorcscent magnetic particles. Direct current of moderate intensity \\,as passed tlrroogh the bars and the fluorescent magnetic particie suspension flowed over tile si~rface. Tlle magnetic particles arranged tlien~selves in a uniform pattern on all surfaces of the bars. The pattern consisted of circumferential lines, corresponding to the lines of force around the bar, a s in a magnetograph. Bars \iritllout voids shonwi an all- o w r uniform coating of fluorescent particles. The adherence uf particles was, of course, produced by the e.rtev11nl magiietic field due to the current the bars xvere c;krrying. When \'o~<ls \\*ere present, the pattern of particles nras distorted a t the location of the vo~ds , where the current density in the bars changecl because of Llle reduction in the conducti\re cross-section of the bars a t such 1oc:ltion. Presence of voids was thus clearly inclicated.

CHAPTER 24

TESTING OF IVELDMENTS, IARGE CASTINGS Ah'D FORGINGS

1. INTRODUCTION. The testing of large objects and structures presents some speclal problems for whtch special techniques have bccn derelopetl. Thc inspection of welds In motierate to heavy see- tions is incliidcd in tlus group, for thc reason tha t speclal procedures arc requlretl.

In the case of weld testing, the article itself may be large, as for esampie a pressure vessel, or a 60 story building, but t he ~nspection is normally confineti to tlie locai area of tlic ~ileld itself. T h ~ s has led to the use of yokes and prods, \i>htch a r c es!~ectallp suited to such local magnetizations. The use of prods and thc rci~u~remeii ts of current values and contact spactug have been thorouglllp discussed in Chapter 10: Sections 6 througli 9. Some additionai details of \velci tnspectiou tec i~n~ques will be outlined 111 the succeeding sections of this chapter.

Large castings and forgings, too large to l~andle o r be acconi- modatcd by standard, or fixturetl n~agi ie t i zn~g units, a r e aiso inspected will1 prods or yolies. IIowevcr, 111 tlus case the local nature of tllc fields produccd a r e a handicap, slncc, unlike welds wliere the location of soaght-for defects 1s itnown, the entire large surfaces of castings and forgings must be ~~atnstalcmgl!: gone over, potnt by potnt-a labortous and time-consumtng project. The introductton of h~g l i output multi-directional (Duovece) power packs, delivering up to 20,000 amperes or more, has made possible overall magnetization of medium large articles. Tilts system achieves an ~ncreased effectiveness in defect location; and a great rcdnclion in time requ~ red for the inspection. I t 1s described in detail in Chapter 19, Section 18.

2. \VELD DEFECTS. Figures 216 and 217 sho\v; respectively, tile terms used 111 relation to tillct weids, and thc defccls runt1 t i l e~ r designations whtch are commonly found 111 fillet and biitt ~velds. In acldi- tion cracinng may occur tn thc nreld metal a t thc base of the Vee of a multiple-pass VCC butt weid 111 a heavy section, cspcctally in the

.I40

Fig. 217-Typ~al Weld Detects

411

first t\vo o r three passes. These cracks, if not detected and ~liipped out, will often propagate into the subsequent iayers of xtzeld metal as they a r e iayed on. Other defects a r e slag mcluslons and voids. lack of penetration between weid and base metal, and cracks in the heat-affected zone in the base metal adjacent t o tire \veld.

Many of these defects a r e open t o the surface and a r e readily detected with magnetic particles, using prods o r yokes. For detection of siag inclusious and voids, and iack of penetration a t the root of the weld, wlllch a r e bclo\v the surface, prod magnetization is best, using half 1val.e D.C. and dry powder.

3. D~AGNETIZING TECHNIQUES FOR \VELD TESTING. The use of prods for p roduc~ng circular magnetization in the inspection of welds, using half wave D.C. and dry poxvder, has long been standard practice fo r t l l ~ s test by the magnetic particle method. Today yokes a r e used extens~veiy for the location of surfacc cracks, though they are not suited if discontinuities lying ~vI1011p belom the surface, suih a s slag inclusions, a r e being sought. Fo r this purpose prods should be used. Yoiies, using either A.C. o r D.C., or half wave D.C. -or even permanent magnet yokes--can introduce a strong field across o r along the surface of the \veld, and will readily locate all surface cracks. They a r e much more convenient to use than are prod contacts. In checking for cracks in tlie early beads of a mul-

7 MAGNETlZiNG

-PATH

PLATE

CORE

OF FLUX

I - - , CRACK in / WELD BEAD

- I Fig. 218-Field Produced by Yoke Across a Vee Butt Weld.

442

tiple-pass wrld in ;I t h ~ c k section, the yolic IS very efft~ctive.

4. OTIlER NOSDESTRUCTIVE TEST ;\II<THoDs FOR \\:El.D ISSPEC- TION. R:idiugral~Iry n-;I:; specified xnd usccl for \ ~ e l d tnsocctioii 1)efore the niagiietic particle ~ncthod was clc~ciopetl. Il:~tling,-;ipiiy FXCI'IS in locatn~g internal defects sucll :is sl;ig ~ ~ ~ c l u s r o ~ i s , gas uocltets, lack of penet l . i~t i~n or otlri'r vo~tls, and is still the niost etTrctirc method for tlus purpose. Radiography 1s not reliable for Iocatiiil: fine sorface cr~icks, iio\\.evrr. Today, ult?u. stlnic mcthotls ;ire used estmslvely for the locatloll of' surface ?racks, ;is n.ell ;Is sllcll su1,- surface discontinuities a s lack of fusion, lac& of penetration, anti craciis in the Ireat-affected zone.

Eddy Current niethods also 1ia1'e :ipl,iic:itions in this field, eitlier uy a probe scanning the weld zone o r by p:iss!ng tho 1velileri product through an enc~rcling coil. Other lnethods of r le tcct i~~g ieakage fields wliich a r e used to some extent for melrl inspection iother th:ur magnetic particles) employ probes for p~cking up the field disturbances and 1ndic:tting them on a meter, oscilloscope, or pen recorder, or sometimes the ~nagnet lc tape teclunique. The I;itter is rolled under pressure over the ni;~gnetized weld bead, and the le:tk;~ge fieids a r e transferred and fixed magnetically on the tape. The tape is subsequently scanned and read by electronic n1e:ins. These techn~ques a r e lirnited to applications \\;here tlie \veld surface is quite smooth, as on electr~c resistance ~velded plpe, for esample.

The magnetic particle methotl 1s 1,owcver still the most eflective in many circumstances, part~culariy In irregul:lr sections wlrere tlie contour of the par t or the configuratio~i of tire surface a t the \veldetl junction makes both radiogral~lry arid t~ltrasound inappiicable. Ultrasounrt has no aci\~;lntage over magne t~c particles in sensitivity fo r detecting suriacc craclts---in fact, the magnetic particle mctliod is the ? I I . O S ~ . sensitive for very shallow cracks, and is lnorc re:idily interprcterl because of the easy read-out of' surface crack patterns. Ultrasound, on the other hand, has advantages for hlgli speed In- spect~on on velds 111 regular-shaped parts, lending itself better to automation and recording.

5. PROD ~'IAGNETIZATION A N D DRY POWDER TECHNIQUE. The requirements for the proper application of (fry magne t~c particles using prod contacts for clrcular magnetization have been tlroroughly discussed in Chapter 13. I t IS suggested that the reader refer back to ihat chapter, espec~ally Sections 4 tlirough 10. Tlie requirements


fo r surface prcparatlon, the effect of positio~i-vcrtical o r overhead versus horlzontai-the importance of ho\v tllc powder is applied, etc., all apply specifically to weld inspection, and need not be repeated here. The important matter of prod spaclng and amount of current is detailed in Chapter 10, Sections 6 throufh 9, and should also be reviewed.

6. YOKE IllAGNETIZATION. If a yoke is to be used, the same re. quire~nents for surface preparation a s for prods apply. Positioiling of the yoke with respect to the direction of the defects sought, is, of course, different f rom tha t of the prods. Prods a r e spaced ato?~g the weld bead to locate cracks parallel to the bead ; tliey a r e placed on opposite sides of the uzeid to locate transverse cracks. Since the ,field traverses a path lbet~\reen tile poies of the yoke, these must be placed on opposite sides of tlie bead for parallel cracks, and along the weld for transverse cracks.

Powder application is slmilar to tha t nsed with prods. Ho\sever, because of the strong external fields associated with the poles of the poke, powder should l ~ e applied sparingly and be directed a t the area between the poles. Since discontinuities lyiiig whoily below the surface are not looked for when yokes a r e being nsed, thc powder application is not critical atid the tndications of surface craclis are easily seen. Such tests a r e widely uscd on hull welds in nuclear submarines, for example.

7. TYPE OF CURRENT. I t has been stated that if discontinuities lying wl~olly below the surface 01 a \rreldment a r e sought, the prod method using half wave D.C. 1s used. A.C. 1s satisfactory so long as surface cracks only are or interest. A.C. has no appreciable deptli penetrating pollrer, but produces a good field for surface cracks. In discussions of the location of discontinuities lylng wholly beion, the surface the question 1s often asked "how deep a defect cat1 be Iocated ~ v i t h magnetic particles?" The reader is referred to Chapter 15 for a discussion in great detail of the concept of depth in relatioti to detectability or deep-lylng defects with magnetic particle testing.

There is one set of ci~cnrnstances in which the penetrating power of half wave results in confusion. This is the case of a double-tee fillet weld in which conlplete pcnetr:ition is not specified and an open root 1s permissible and neariy always present. W l ~ e n half wave 1s used with prods, this open root will most probably be indicated on the weld surface a s shown in Fig. 219. Mr. d. W. Owens pointed

344

CHAPTER 24 TESTING OF WE1,DMENTS. LARGE CASTINGS AND FORGINGS

(Coarwry Ncwnon N e a r Shipbuilding Comilmnyl

Fig. 22GMagnetic Particle Indications ot Transverse Cracks ~n Mil 260 Fillet Weld, Shown with Yoke.

I'ILIKCII'1,I<S 01' RIA1~KISJ'IC !'AILTICI,E TESTIVG ~ -..-A -- -

out in 1944 thnt this confusion could be overcome by using A.C. 111ste;td of half on such \velds. A.C. will sllow surface cracks in the welrl metal, but will not peni:tralc iiecply enough to produce the conf~isirig ~odication of the open root, \sllicl~ in this case is there by design and therefore is not a weld defect. Portable units for \\,elcl insnection often can deliver either hulf xvave D.C. or A.C. so tli:~t each is available \\,hen needed in such a situ:ltion.

8. EX~\AIPI.ES OF WELD INSI').:C'I'ION WITH MAGNETIC PARTICLE TESTING. X few exi~mples of the application of magnetic particle testing to the inspection of \t-elcls will serve to illustrate some of the things that are being done in this area.

A. Insileefio?~ of St~?rct?cvnl I'Velds. Wit11 welded construction taking over the f;lbrieation of all sor ts of steel structures, the need [or shop and field inspectioii of welds to assure safety has bedome urgent, and today nearly all such construction calls for \veld inspection. Any o r all of the three standaril nondestructive tests for welds a r c used-radiographic, ultra- s o n ~ c and masnetic particle.

Fig. 221-Inspecting Welds with Magnetic Particles on the 32-Story Michigan Consolidated Gas Company Building tn Detroit, Michigan.

Radiographic inspectior~s of ~velds under (.he coitditions shown in Fig. 221 are diHicuit, but not ~mgo.isihlc, except where long ~ ~ e l [ l e d box heams give 110 opening to piece fihn properly. A major handicap is 111 radiation danger, requiring

the area to be cleared of people. Costs a r e ;ilso (11lite high. Ultrasound also has many ap~~l ica t ions 111 this field. It has tlic advantage of easy portability and freedom from riidiatlon hazard. Costs a r e in the medium range. Eut careful handling is required for a delicate instrument in precarious loc;~lions such a s slio~\.n m Fig. 221.

IIagnctic particie portable n~agnetizlng units can be placed on a construction floor or platform, and cables of considerable length usetl to reach the actual pomt of inspection. Only the prods, which include the current-control s \ \~~ t ch , and the powder anplicator need to Ile handled by lllc inspector. Light portable units xvit11 outputs up to 900 amperes were clesignctl espec~ally to facilitate these field ~nspections, and are in \\-idc use today. Remote current-control can also be had a t the point of inspection by means of a small dial on a light cal~lc. Thc inspector can operate the infinitely variable self-reguiating control as much a s 100 f t . from the magnet iz~ng unit. Uililer some circumstances, permanent mag~le t .yokes can he used if i t is not feasible to supply power to a standard portable unit.

Structorai weld inspection \vlth magnetic particles recelvea i ts first impetus in California, where the State Departmci~t of Education began to use \\~~?'ldcd, single story schooi boildings a s :I safeguard against earthqualie damage. The practice of all-\\.elded fabr~cat ion spreatl slowly a t first, but has bccoillr a stnndarii practice today. Figure 222 shows ailother application of n~i~gi ie t ic particie testing of a critic;ll \\fclil a t the base of the 285-foot televisioll mast being. erected on top of onc of the i\larina City To\vers In Chicago.

B. Electi-zc Resrstn?icc 'i'eldad Steel Pipe. The !\,elding of steel pipe 11y the electrlc resistance metl~otl provides an esamplc of the automatic production testing of welds. Figure 22:! shows an insiallation !vhicll is applylng the magnetic particle icst to the \veld of a 10 inch pipe, withiii a few feet of the weld stand. The powerful electromagnetic yokes shown arc user1 to produce a field transverse to the xveld. A special coarse dry poxvder is applied in escess, and the surpius IS rceovereci and

PRINCII'LES O F h1AGNETIC I'ARTICLE TESTING

Fig. -- .--....p....-.. ~~-

Roof of One ot the Twfn Towers of Marma City In Chicago. I the

returned to the storage hopper shown. The seam then passes under the eyes of the inspector, who marks the indication of cracks a s they go by-either by hand o r by paint-spray application which he controls.

Speeds a r e f rom 50 to 150 feet per minute, depending on pipe size, whlch may run a s large a s 36 inches in diameter. Different pole pieces fo r the electromagnets a r e provided to fit different slzes of pipe. The unit can be made fully automatic, althougb the cost of complete automation is difficult to justify. Defects found a r e cllipped o r ground out and repaired by hand. Nost ~n ipor tan t i s the ability t o control the weider, so that, a s soon a s defects occur, them cause can be eliminated. Failures on final hydrostatic tests have almost been eliniinated by this test.

C. Presszwe TJessels. The earliest application of radiographic nondestructive testing was on welded pressure vessels-both

448

CHAI.TEB 24

TESTING OF WEl.I>hlEKTS, LARGE CASTISGS AND FORGINGS

Fig. 223-Inspection at the Weld Stand of Res#stance.Welded Pipe.

hammer- and fusion-welded. Magnetic Particle testing was early used on many types of pressure vessels because of i ts great reliability for the location of szo:tcice craclcs-an area In which radiograpliy is less than fully dependable. Welds around nozzies and man-ways, whlcli could not be x-myeii becaiise of shape and contour, were also magnetic particle tested.

Such tests continue today, with a trend to\sard greater use of magnetic particle testing on pressure vessels lor missiles and nuclear cqn~pn~ent-where esery precaulion against any reduction in strength a t a weld must be taken. The extreme seiisitivit)~ of magnetic p;irticies Tor very fine and very sliallow cracks is a large factor i n i ts use fo r sucii purposes.


Fig. 224--Testing a Welded Oil Storage Tank.

9. STEEL CASTINGS. Magnetic particle testing is used extensively on castings of all kinds. Defects sought include all kinds of surface Aaws such as shrink cracks, hot tears, porosity, blowholes and slag pockets. For small castings, standard units are generally used, with either the dry or the wet method. Defective castings are segregated and saivaged, if possible, by grinding out the defect, provided it does not penetrate too deeply into the section. Larger castings are inspected with prods, again with either dry or wet magnetic particles. Defects are cut or ground out, and repair by weiding is often permissible.

The overall method of multi-directional magnetization already described in Chapter 19 is the most rapid and most satisfactory method for the testing of very large castings, some of which may weigh many tons. Although this application requires heavy and relatively expensive electr~cal equipment, sometimes with outputs of 20,000 amperes or more, the saving in inspection time per casting, and the Increased effectiveness in reliably detecting important surface cracks, can justify the expense of the installation on a purely

450

CHAPPPR 24

TESTING OF WELDMENTS, LARGE CASTINGS AND FORGINGS

economic basis. The savings are most pronounced when a large run of castings from the same general design are being made. Once the proper magnetizing pattern has been worked out, the process is purely repetitive. Inspection time per casting has usually been reduced to only 25% of that required for prod inspections, and in one application a t least, to only 10% of the former prod inspection time. Prod inspection of large and complicated castings frequently did not locate all defects on the first test, and additional cracks would show up on re-inspection after repalr. Because the all-over method usually finds all significant cracks on the first inspection, re-inspection time after repair is often cut in half.

With the overall method of magnetization, fluorescent magnetic particles in water suspension are invariably used, mainly because of the greater visibility of the indications. This advantage is especially important when cracks occur in holes or recesses difficult to illuminate and see into with white light, but for which the black light that can be projected into such places is adequate to make indications readable. Dry powder does not perform well at the high levels of D.C. magnetization usually employed for the overall method because it tends to form patterns of the external field, thus losing the mobility necessary to form indications of defects.

If defects lying wholly below the surface are cons~dered important, the prod method with dry powder 1s still the most effective. Locations where porosity or internal shrinks are likely to occur can usually be predicted, and the prod inspection confined to those few areas.

10. NONDESTRUCTIVE TESTING AS A DESIGN TOOL. One of the most useful applications of magnetic particle testing methods (as well as for all other nondestructive testing methods) is in the checking of pilot casting procedures and mold design on a new article. Testing of pilot castings can indicate locations subject to shrlnk cracking, porosity, etc., and point to remedial changes in the design of the casting itself, or to corrections in chills, or to the mold with respect to locations of gates and risers.

11. GRAY IRON CASTINGS. Gray Iron castings are Inspected for many crack-type surface defects with the magnetic particle method. These are the most common defects occurring in gray iron. Because of the distribution of the graphite flakes throughout the metal the magnetic field is badly distorted, making location of below-the-


surface defects difficult or impossible, especially on the lower grades. In the case of chilled iron which is extremely hard, the permeability IS low and this again defeats deep defect detection, although surface cracks are readily found. In the case of nodular and malleable iron, somewhat better results can be expected for deep-lying defects, although as a practical matter the greatest number of applications and needs for iron castings iie in the detection of surface cracks oniy.

One area of saving in the gray iron foundry by use of 100% magnetic particle inspection is the detection of handling cracks. Gray iron 1s very brittle, and rough handling in shake-out and in subsequent sorting operations often produces more cracks than are present when the casting leaves the mold. Figure 225 shows some

Fig. 225--Some Typlcal Handling Cracks tn Gray Iron Castings

typicai handling cracks indicated with magnetic particles. When the number of such cracks showing up on inspection becomes exces sive, special handling precautions can be set up and such production losses largely eliminated. Some shake-out machines have been completely rebuilt to correct this tendency to cause cracks in castings.

In the production of malleable castings, the white iron castings are commonly inspected before malleableizing. A cracked white iron casting takes up valuable space in the furnace and can oniy

CHAPTEB 24

TESTING OF WELDhlENTS, 1,AIIGE CASTINGS A N D FORGINGS

result in a defective malleable casting wh~ch must be scrapped. Magnetic particle testing in the X;hite iron stage assures that oniy sound castings go into the malleableizing furnace-a considerable economic savlng.

12. FORGINGS. Many of the practices in testing large castings are duplicated in the case of large forgings, too large to handle by normal magnetizing and inspection techniques. The prod method is normally used on such articles, althougil in some cases the overall method is applicable. Fig. 226 shows ~nspeetion on overhaul of an 18 111th marine tail shaft, with fatigue cracks shelving in the keyway. In such a shaft, the areas vulnerable to fatigue are a t the

Fig. 226-Magnetic Particle Test of 18 mch Marbne Tail Shaft, Showong Fatigue Cracks tn the Keyway.

keyways where the drlve gear and the propeller are attached, a t the outer end of the outboard bear~ng, and at polnts adjacent to couplings and line bearings.

13. SUIIIMARY. The successful lnspect~uii of weltln~ents and large cast~ngs and forgings requires the highest type of skill ant1 knowledge of the method.

I t is only in the case of volume-produced identrcal items and tlie s~mplest welded articles, or small castings and formngs, that pro-


cedures wh~ch can be carried out by the average ~nspector can be easily set up. For complex weldments and large parts of ~ n t r i c a t ~ shape, the inspector or nondestructive tes t~ng engineer must work out the proper magnetiz~ng techn~que h~mself, whxch may he different for each case. Success here depends on a full knowledge of the theory and practice of magnetic particle testing as has been detailed in the pages of this book, supplemented with practical experlenee.

CHAPTER 25

STANDARDS AND SPECIFICATIONS FOR MAGNETIC PARTICLE TESTING

1. GENERAL. Because of the very large number of var~ables that may affect the end result of magnetic particie testing it 1s highly desirable that the procedures for conducting any glven ~nspection be exactly specified and described. Also, such gu~des to rejectabie conditions as it may be feasible to establish should be set up to assure, as f a r as possible, uniform interpretation of the indications that are produced. The former objective is relatively easy to accomplish and 1s ~ndependent of the human factor, provided that specified procedures are faithfully followed by the operator. The second objective, however, must in most cases rely on human ~udgment based on someone's knowledge and experience with interpretation of various magnetic particle indications of defects.

Since the effect of a glven discontinuity on the usefulness of a part may differ vv~dely in different parts intended for different types of servlce, i t is obviously impossible to write generai standards for acceptance or rejection. I t is a curlous fact, however, that only a few years ago industry was clamoring for the sponsors of the method to tell the users where to draw the accept-reject line. Clearly it is the function of all nondestructive testing methods to locate and if possible ~dentify discontinuities. But the accept-reject decision must be made by the engineers who design and use the part being tested, for only they can judge accurately the in-serv~ce effect of a condition disclosed by nondestructive testing.

2. SPECIFICATIONS FOR MAGNETIC PARTICLE TESTING. For the purposes of t h ~ s discussion we are concerned with specifications of only one type-namely, those which iay down rules for the method for specific applications, to insure reliable, uniform and reproducible results, and not acceptance standards.

The tendency to spell out every detail of procedure for these purposes certainly has some desirable objectives, but there are also some results of it that are less desirable. By reduc~ng all activities with regard to the ~nspection either to automatic processes or to

455


processes governed by rules and specifications, the following ad. vantages are presumably acllieved:

ia) Increased over-all reliability of the inspection.

Ib) More uniform results.

ic) Reproducibility of results by different operators or inspectors.

(d) Inspectors can be used who do not have a high level of education or experience.

(e) Minimized training of operators and inspectors.

(f) Conservation of time of skilled and experienced inspectors.

Some less desirable results may be the following:

( a ) The tendency to relieve operators of the necessity for think- ing and developing skills, and feeling a personal responsl- biiity for results.

(b) De-emphasizing the importance of watching for abnormali- ties or unusual occurrences in materials and processes that may affect the inspection, or product usefulness.

tc) The tendency to make the ~nspection perfunctory-which it should never be.

Still, the balance of weight is definitely on the desirable side and standards and specifications are certainly important and necessary to insure reliable tests. Therefore much has been written and proposed to provide guides and controls in all phases of magnetic particle testing practice.

3. TYPES OF SPECIFICATIONS. Specifications for magnetic particle testing fall Into several groups, each with a somewhat different objective from the others. These groups are:

(a) Broad procedural guides-general.

(b) Company procedural guides.

ic) Procedural guldes for specific types of products, or for industries.

Id) Procedures for testing specific articles, specified by the purchaser or by the company internally-process specifications.

CHAPTER 25 STANDARDS AND SPECIFICATIONS

ie) Procedures specified for user overhaul inspection of a com- pang's products.

(f) Specifications for certification of operators.

(g) Limits for acceptance or rejection: set up by a buyer, or by a company itself for quality control.

01) Repair Station requirements-especially for aircraft. i i ) Equipment specifications.

(j) Instructions for operating specific types of equipment or individual special units.

4. BROAD PROCEDURAL GUIDES. A number of technical societies have prepared broad procedural guides for the use of their members who wish to learn about or use magnetic particle testing. These consist for the most part of a very general description of magnetic particle testing. They are lntended to be informative but not binding. They give the reader an idea of how the method works, and the steps by whlch it is properiy applied. Usually examples of applications are glven, often with some illustrations of types of indications of typical conditions.

The sections on magnetic particle testing in the Handbook of the Society for Nondestructive Testing constitute by far the most complete presentation of this type. Some of the other Societies whose handbooks include a section on magnetic particle testing are the American Welding Society, the American Society for Metals, the American Society of Mechanical Engineers, the Society of Auto- motive Engineers, the American Society of Tool and Manufacturing Engineers, and others. In addition, there are varlous books, published both here and abroad, dealing with nondestructive testing that include similar more or less complete descriptions of the method.

5. COMPANY PROCEDURAL GUIDES. Many large companies that make extensive use of magnetic particle testing have prepared their own versions of the procedures that are to be used in their operations. These are usually rather broad in their approach, except that they may be limited to those specific procedures applicable to their products. As an example, if the dry method is not used in their plant, no guide for the use of this technique need be mcluded. Also the procedure specified may be written around the particular type of equipment used in that plant.

457


Such manuals are used for tralning new operators, and for the guidance of those reguiarly using the method. Included in this type of specification or manual are those issued by various facilities of the Armed Services, prescribing the materials and techniques that are approved for use. Such Technical Orders (T.O.'s) are issued by the U.S. Air Force, for exampie, for the use of inspectors engaged in testing aircraft parts on overhaul. These T.0.k are usually very complete and cover all types and phases of magnetic particle testing which are used in their specific areas of application.

6. PRODUCT OR INDUSTRY SPECIFICATIONS. Some method specifications have been written from the point of view of a single industry or type of product. For example, the description of magnetic particle testing which appears in the Handbook of the American Weld- ing Society is directed entirely toward its use for the testing of welds and weldments. The U.S. Air Force Technical Orders are, of course, written for use in the testing of aircraft and aircraft engtne parts. A company that makes only castings would prepare a specification that would only cover the tests of castings and associated weldments, and this would be no broader than need be for this single purpose. The U.S. Air Force has issued a specification that is limited to tests of castings, as a class.

7. PROCESS SPECIFICATIONS. A much narrower type of specification is that drawn up to prescribe the exact method of testing of a single specific article. These are usually prepared by the quality control or engineering department, and in addition to calling for an exact testing procedure, usually also include a definition of indications which may be tolerated and of those for which the article must be rejected. Such process specifications may refer to: or make use of, the broad method description of a company-prepared or hand book type of procedural gutde, hut should be specific with regard to those factors u~hicli u~ill insure proper inspection of the particular part to whicl~ the process specification refers.

Some of the factors that should be covered are:

( a ) Cleaning or pre-inspection preparation (paint stripptng or removal of plated coattngs when required, degreastng, etc.).

(b) Specific method of magiietizatioii and particle application to be used. i l ) Wet or dry method.


(2) Circular or longitudinal magnetization or both. (3) A.C., D.C. or Half-Wave D.C. current. ( 4 ) Continuous or residual method. (5) Fluorescent or visible magnetic particles. (6) Amperes or ampere turns of magnetizing force.

(c) A specific type of equipment is often called for.

td) Procedures and limits for demagnetization if required.

( e ) Post-cleaning when required for following processes.

( f ) Post-inspection surface protection - corrosion prevention and packaging.

In addition, critical areas are usually defined and the indications described which may be accepted, and those which are immediate cause for rejection, so that disposition of what otherwise might be questionable parts can be expedited. Length, direction and frequency of indications and their location with regard to critical stress areas are usually the means for classifying acceptable and rejectable indications.

These process specificattons and acceptance limits are an excellent device for insuring good inspection. They must be prepared with care by competent engineers who are fully aware of theservice requ~rements of the part, the directions and magnitude of the stresses and the location of stress concentration areas. Those pre- paring the specification should also be thoroughly familiar with the various techntques of magnetic particle testing, and the degrees of sensitivity control obtainable.

Similar specifications and acceptance limits may be agreed upon in advance between buyer and manufacturer to cover the Inspection and the acceptance and rejection limits for a specific purchased part.

a. MAINTENANCE OR OVERHAUL INSPECTION. hlanufacturers of complex and critical products (as, for example, aircraft engines, diesel engines, etc.) frequently include a section in their maintenance manuals in which recommended testing procedures are given for use on overhaul. Such a guide is broad enough to cover general techniques and equipment, but is specific with respect to the tests of certain critical parts and components. Such guides are of great value, since they tend to assure in advance that the inspection for maintenance purposes that will be performed by the user will


be suitable and proper for the specific items involved. The chance is minimized that the inspection may go wrong because the inspector responsible may not be sufficiently experienced to perform the inspection properly. Advice is often included in such manuals as to where to looic for trouble and how to judge the significance of an indication when found; and, if possible, how to repair the part.

9. CERTIFICATION OF OPERATORS. The U.S. Air Force and other Federal and State Government departments that make regular use of these test methods long recognized the dnnger of poor results if operators and inspectors are not properly trained and experienced. For this reason a system of operator qualification was long in use, and specifications were issued under which operators were required to be qualified. Such qualification in some form was a pre-requisite for an operator who was going to conduct inspections-espec~ally of Government-purchased materials, or on Government-used equipment such as aircraft or parts on overhaul.

Usually such certification was limited to certain types of inspection on certain types of equipment. If the operator changed jobs he needed new certification tests in order to qualify for the new work. Operators and inspectors in vendor's plants were required to be so certified as were those directly employed by Government agencies.

Responsibility for certification of operators under current Gov- ernment rules now rests with the prime contractor for a given government project. Qualification tests are often given in accord- ance with Government specifications, and include written examinations on materials and techniques, as well as actual demonstration of ability to find and interpret defects in actual parts.

10. STANDARDS FOR ACCEPTANCE OR REJECTION. Over the years there has been an insistent demand for standards to govern acceptance and rejection of parts on the basis of indications produced by this (and other) nondestructive testing methods. Users have complained that they are told how to use a method of test but are not told what to do about indications of defects that are disclosed. A little reflection should make it readily apparent that any broad standards for this purpose are impossible.

Whether or not a given discontinuity warrants rejection of a given part depends entirely upon factors that have no relation whatever to the inspection method that produces the indication of its presence, nor to the wze or type of defect, per se. The service

460


requirements of the part are the controlling factors. The nature of the stresses in service, stress concentrations, notch sensitivity of the metal, the critical nature of the function of the part-even location and orientation of the defect on the specific part-must all enter into a decision as to acceptance or rejection. These considerations cannot be generalized, but must be applied by those possessing the information, to each specific case. A given defect in one part for one type of service may warrant rejection, but may have no significance a t all if present in another part for a different type of service.

The preparation of standards for acceptance or rejection is the responsibility of the designers of the part and of the quality control department, and are most effectively included along with the test process specifications. If such standards are not available and the person judging the results of the inspection does not possess the necessary knowledge, he must seek it from those who are responsible for the design and service peiformance of the part. He should not expect broad standards to be available. However, knowledge of the material and its intended service, as well as of defects in genera1 and their effect on strength and performance, often allow an intelligent, experienced inspector,.with the use of common sense, to arrive a t a logical decision regarding acceptance or rejection.

11. REPAIR STATION REQUIREMENTS. Another type of specification, bearing on the use of magnetic particle (and other) inspection is that setting forth the requirements for repair stations for aircraft structures and aircraft engines. The Federal Aeronautics Administration (F.A.A.), which has jurisdiction over the approval of shops or stations for repair or overhaul, specifies the extent and type of equipment and other facilities necessary to insure proper and satisfactory inspection of all parts before they may be re-used. Magnetic particle equipment is a requirement for a fully authorized repair and overhaul station for aircraft and aircraft engmes.

12. EQUIPMENT SPECIFICATIONS. Specifications for standard equipment for applying the magnetic particle inspection method have been prepared and are used for procurement of such equipment by the Air Force and by other Government agencies. Such equip ment or material specifications are not (necessarily) applicable to general manufacturers.

13. OPERATING INSTRUCTIONS. Manuals or operating instructions are prepared by responsible manufacturers of test equipment

461

PRINCIPLES OF BlAGNETIC PARTICLE TESTING -....

t o guide the purchaser in its proper use. These are not specification in the usual sense of the term: but may include method procedure and techniques as applied to the use of the specific piece of equip. ment.

For large pieces of equipment, where automatic handling is volved, these manuals become lengthy and complex. The functi of all controls-each button, switch and relay-must be explaine so as to be understood by the persons responsible for keeping the unit operating properly. Successful operation of one of the large steel billet testing units idescribed in Chapter 19) would be difficult if not impossible without the specific instructions furnished in such a manual.

14. GOVERNMENT SPECIFICATIONS. Following is a list of U.S Government specifications currently in use, which have to do with magnetic particle testzng:'

Mil-C-6021 Castings, classification and inspection of.

Mil-1-6870 USAF Inspection Requirements, Nondestructive, for Aircraft Materials and Parts.

Mil-Std-23 Nondestructive Test Symbols.

Mil-Std-271 Bu-Ships Military Standard Nondestructive Testing Requirements for Metals.

Navshzps 250-1500 Standard for Welding of Reactor Conduit Coolant and Associated Systems.

Mil-M-19698 Magnetic Particle Unit, Portabie.

Mil-M-11472 Magnetic Particle Inspection Process, for Ferromagnetic Materials.

Mil-M-11473 Magnetic Particle Inspection, Soundness Requirements for Weldments.

Mil-Std-410 Qualification of Inspection Personnei (Magnetic Particle and Penetrant).

Bu-Weps Mil-1-6868 Inspection Process, Magnetic Particle, (General Requirements for Magnetic Par- ticle Inspection)

Nav Aer 00-l5PC-503 Technical Inspection Manual, Volume 3, Section 4, Magnetic Particle Inspection.

462

CHAFTER 25 STANDARDS AND SPECIFICATIONS

USAF T033, E2-1-1 Inspection of Materzal, Magnetic Particle Method.

Bu-Weps Mil-1-18620 Inspection, Magnetic Particle, Require- ments for (Applicable to Overhaul of Air- frames, Engines, Propellers and Acees- sories) .

Mil-M-68678 Magnetic Particle Inspection Unit. Mil-M-23527 Magnetic Particle Inspection Unit-Light-

weight.

MS-17980 Magnetic Particle Indications on Steel Nuts.

Follo~ving is a list of some of the Technical Society Specifications which are currently in use:*

SNT-TC-1 NDT Personnel Qualification, Recommend- ed Practice.

ASTM E-109-63 Dry Powder, Magnetic Particle Inspection.

ASTM A456-61T Method and Specifications for Magnetic Particle Inspection of Large Crankshaft Forg~ngs.

ASTM E125-63 Reference Photographs for Magnetic Par- ticle Testing of Ferrous Castings.

ASTM A275-61 Steei Forgings, Heavy, Magnetic Particle Testing and Inspection of.

ASTM E138-63 Wet Magnetic Particle Inspection.

ASTM A340-61 Definition of Terms, Symbols and Conver- sion Factors Relating to Magnetic Testing.

AWS Fourth Edition Handbook 8.24 to 8.36.

SAE AMS-2300 Magnetic Particie Inspection, Premium Aircraft Quality Steel, Cleanliness.

SAE AMS-2301 Magnetic Particle Inspection, Aircraft Quality Steel, Cleanliness.

SAE AMS-2640 Magnetic Particle Inspection.

'Taken from Matertals Evaluation, March 1966, Vol. XXIV, Number 3, pp 158 and 159.

463


15. OTHER SPECIFICATIONS OF INTEREST. Among the m specifications on the subject of magnetic particle testing which been put Into use by industry and puhlic works offices, a few a interest as examples.

The State of California, Department of Public Works, Divrsio of Highways, Materials and Research Department, have adopte specifications covering the inspection of welds in steel structu such as bridges and buildings. The des~gnation of this specificat is "Test Method No. Calif. 601C, Jan. 1959-Test Method an Techniques for Control of Welding Procedures Performed During Fabrication of Welded Steel Structures". Par t VIII of this specifica. tion defines the procedures to be followed in the magnetic particle testing of w e l d s t h e currents to be used for different plate thicknesses per inch of prod spacings, precautions to he observed and method of reporting results. The State Road Commission of West Virginla has a similar specification, which grves a description of defects which are cause for rejection or which may be repalred.

The Interstate Commerce Commission has issued a n order, #59-A, (10/1/63) which requlres that ail cargo compressed gas tank trucks must be completely re-inspected with magnetic particle testing methods a t the rate of 10% of a total fleet every 6 months. All welds, internal and external must be tested.

An example of more highly specialized industry specifications is one Issued by the Manufacturers Standardization Society of the Valve and Fittings Industry. Their MSS Standard Practice SP-53, "Quality Standards for Steel Castings for Valves, Flanges and Fittings" covers dry powder magnetic particle testing techniques as approved for this area of inspection.

16. SUMMARY. It IS evident from this review that specifications which have been promulgated in the area of magnetic particle testing have come into being as the result of a need to make sure that proper and uniform procedures be followed in any given case of magnetic particle inspection, rather than to leave the selection of method and technique to different inspectors who may or may not be fully qualified to make the selection, and who are almost certain to have different ideas and come up with different method variations. In an operation such as magnetic particle testing in which t.here IS

reliance on the human factor, the careful consideration of method variations and selection of the most suitable one, by persons fully


qualified to do so, gwes the highest assurance that the greatest possible vaiue will be obtalned from the tests.

CHAPTER 26

TESTS FOR EVALUATION AND CONTROL OF EQUIPMENT AND PROCESSES

1. GENERAL. In applying magnetic particle testing (or any other non-destructive testing process) guides for the operation of the equipment and the details of the process assume that the operator and the equipment he is using will be capable of carrying out -and will faithfully carry out-the instructions and the procedures specified. However, operators and supervisors have need to verify that the critical requirements for magnetization and bath application are actually being met as the Inspection proceeds.

Examples of things that might go wrong and lead to completely abortive results are the following:

(a) Meter reading magnetizing amperes not registering the actwl current being passed through the coil or part.

(b) Wrong magnetic particles being used, or wrong suspending liquid for the bath.

(c) Bath concentration incorrect.

(d) For fluorescent particles, separation of fluorescent pigment from the magnetic particles during use.

(e) Insufficient biack light intensity for the reading of fluorescent particle indications.

2. MALFUNCTIONING OF EQUIPMENT. If an operator, following procedural specifications, reads the called-for current on the ammeter, he assumes that the current is doing the specified magnetizing job. This is a fair assumption with equipment that has been properly designed and constructed on quality lines, and been properly maintained. However, ammeters can and do go out of calibration, and the supervisor should check the ammeter for accuracy a t intervals.

Transformer coils can, with long use (or abuse) burn out, or connectors can loosen or corrode and break. When this happens the meter will usually not read a t all, and operators should make a

C H A ~ 26 TESTS FOR EVALUATION AND CONTROL

practice of watching the ammeter while magnetizing parts. In some installations a test is run a t the start of a shift by placing a one inch copper bar between the heads of the unit and passing the maximum current through it. The meter is checked to make sure that i t is reading up to the rated maximum output of the unit.

An example of an unusual failure of equipment was the case where the copper bar test appeared to pass the rated output current according to the meter reading. However, no current was flowing through the bar (or parts). It was found that a short circuit had developed between the head contact piate and the frame of the unit and the current was entirely hy-passing parts clamped between the heads. In addition to such an electrical check on the meter, therefore, some overall performance test is desirable. (See Section 8: this Chapter.)

3. PROPER MAGNETIC PARTICLES AND BATH LIQUID. When an operator requisitions magnetic particles for his process from the stock room, he may assume that he has been issued the proper material. This is not necessarily the case, since the stock room attendant may pick up a container of material that is intended for use elsewhere in the plant. The operator should, of course, verify that the container label is correct, for the material he is using. Normally, if the material is not correct, the fact will be detected by its appearance and its behavior in use. However, with the un; fortunate tendency to relieve the operator from the necessity of think~ng, by issuing detailed rules for him to follow, he may not observe that an error has been made.

An instance when such an error led to an absurd failure of the method came to light some years ago when the stock room issued red lead paste illstead of red magnetic particie paste, and the operator went ahead in carefree fashion with a nice red bath, until somebody noticed that for some time no defects had been found in parts tested on this particular unit. Fortunately such errors do not happen often, but the incident points up the need for alertness to make sure that all factors in the process are working properly.

In another case, in which an oil-base bath was being used, the stock room issued naphtha instead of the proper high-flash distillate. The result was a fire in the equipment when an arc a t the contact heads ignited the naphtha bath. Again, alertness on the part of the

467

PRINCIPLES OF MAGNETIC PARTICLE TESTING - operator or supervisor m ~ g h t have averted t h ~ s loss. Specifications for a suitable oil for wet bath purposes are given later in this Chapter f Section 6).

4. BATH CONCENTRATION INCORRECT. In the course of use a cer- tam amount of magnetic particle material is removed from the wet bath by clinging either magnetically or mechanically to the surfaces of parts. In time-depending on the number of parts tested-this results in a bath with insufficient magnetic particles and more testing material must be added. On the other hand, carry-out of bath liquid on the surface of parts, as well as evaporation of liquid when the unit is not in use, also acts to change bath concentration. Frequent tests for bath concentration are a most important requirement. A weak bath results in faint indications and thus tends to reduce the apparent severity of a discontinuity; or a faint indication which might he produced by a bath of proper concentration, might be missed entirely, if the bath is below strength. Sections 8 and 9 describe available bath-strength tests.

5. DETACHMENT OF FLUORESCENT PIGMENT. Modern fluorescent magnetic materials of the best quality are made by a process resulting in a dye-magnetic-particle combination which is permanent, and does not come apart even after prolonged, vigorous agitation. However, some types of fluorescent magnetic particles tend, in the course of use, to lose theif attached fluorescent p~gment under the action of the vigorous agitation in an inspection unit. If t h ~ s occurs, fluorescent indications will become weaker in proportion to the number of particles which have lost their fluorescent color. In such a bath, the free dye tends to adhere over the entire surface of the part, not by magnetic attraction, and creates a confusing background. The non-fluorescent magnetic particles are attracted to leakage fields along with those still holding their fluorescent pigments, and act further to dilute or obscure the fluorescence of an indication.

Such a situation is quite different from the weakening of indications due merely to a lack of magnetic particles. Again, alertness on the part of the operator enables him to sense that something is wrong with the appearance of indications, as well as to observe an Increase in fluorescent background.

6. ~PECIFICATIONS FOR SUITABLE PETROLEUM BASE LIQUIDS FOR OIL TYPE WET BATH. The important characteristics for petroleum

* F

CHAPTEE 26 i TESTS FOR EVALUATION AND CONTROL

distillate for the wet-bath suspenso~d for magnetic particle testing were discussed in Chapter 24, Section 5. The tests and how to make them are listed below :

ia) Viscosity-Kinematic a t 100°F (3g0C) 3 Centistokes, max. ASTM test method D-445-65

(b) Flash Point. Closed cup. 135OF (57'C) Min. ASTM test method D-93-62

(c) Initial Boiling Point. 390°F (199°C) Min. ASTM test method D-86-62

(d) End Point. 500°F (260°C) Max. ASTM test method D-86-62

ie) Color. Saybolt plus 25. ASTM test method D-156-64

( f ) Sulphur Available. Must pass copper test. ASTM test method D-130-65

7. SETTLING TEST FOR BATH STRENGTH. The settling test for bath strength was dev~sed over 30 years ago when the wet method of magnetic particle testing was first introduced. With some notable exceptions, i t is still In use today, in spite of the fact that it lacks much in accuracy, and there are several inherent sources of error. It has the advantage of being simple and easily applied by the operator a t the unit. I t also glves Information as to the condition of the bath from the point of view of contamination with dirt and lint. Other, newer methods are available f o ~ checking bath strength, although all have some draw-backs of their own. Some of these will also be described.

The settling test was discussed in general terms in Chapter 14, Section 8. Details of the test are glven below :

Equipment: ASTM pear-shaped centrifuge tube, Goetz, 1.5 ml stem, graduated in 0.1 ml.

Stand for supporting the tube in a vertical position.

Procedure :

(1) Let pump run for several m~nutes to agitate the suspension thoroughly.

469


(2) Flow the bath mixture through hose and nozzle to clear hose.

(3) Fill the centrifuge tube to the 100 ml line.

(4) Place the tube and stand in a location free from vibration and let stand for 30 minutes for the particles to settle out.

(5) Read the volume of the oxide portion of the sediment, and refer to the table for the correct volume for the type of magnetic particle concentrate being used. If reading is too high, add liquid to the bath. If too low, add concentrate.

The sediment in the tube will consist of two layers, and sometimes three. The upper layer consists of light dust and lint, which being the lightest takes longest to settle. Below i t is the oxide layer, and in taking the reading the upper layer should be disregarded. Its amount, however is a clue to the condition of the bath. If heavy iron dust from grinders or mill scale is present, this material, being heavier and of larger particie size, will settle out the fastest and appear as a layer under the oxide. I t should also be disregarded.

The settling test worked well enough on the orig~nal magnetic particle paste mater~als, but the introduction of new materials varying in specific gravity, has made the test less definitive for use with such materials. Other methods of bath checking are comlng into use.

Table VIII gives the settling readings for the several magnetic particle materials, for a new bath made up with the quantity of magnetic particle material recommended in the table.

As has been said, this method is not an absolute one, and the results may vary among different operators and with different bath liquids. Its main usefulness lies in day to day comparative readings for a given unit. The settling values in Table VIII should not be accepted as the standard for day to day checking. Rather, the settling volume for a new bath in a given unit should be determined and used for the standard for that unit. This elim~nates the variables of bath liquids and individual preferences for concentrations differing from those shown in the Table. Build-up of mill scale or heavy dust, and of light dust and lint, can be observed, and when it becomes excessive corrective action can be taken. Usually tkis means that the bath must be discarded and a new one made up. As was said above, the middle iayer of oxide should be read as the

470

CHAPTER 26 TESTS FOR EVALUATION AND CONTROL


measure of bath concentration of useful magnetic particles.

8. OTHER BATH STRENGTH TESTS. A number of other tests for checking bath strength have been proposed and some have been used. Most of these are based on measuring the amount of magnetic - teridl in the bath. This can be done by means of a coil around the hose or pipe through which the bath is flowing, and measuring continuously, with read-out on a meter, the magnetic properties of the stream. This can be converted into the weight of magnetic particles in the bath per gallon.

Such a device might have some value in convenience, but actually would probably be no more accurate or satisfactory than the settling test. It would make no distinction between magnetic dust (common in foundries and steel mills) and magnetic testing particles. ~t would, of course, have to be calibrated separately for every type of magnetic particle used.

Various test blocks containing a series of artificial defects, just below the surface and having graduated depths, have been proposed and some are in use. Some of these are permanently magnetized, and a series of milled slots of the same width but increasing depth are used as indicators. Others use a single slot of varying depth, from shallow to deep. One version of such a device, shown in Fig. 227a, consists of a piece of soft iron having a hole drilled

I *

C. TEST BLOW POIITIONED IH EOhL

Fig. 227. (a) Test Block for Measuring Bath Strength. fb) Test Block In Use Between Heads of Unit.

(c1 Test Block In Use an Coil.

472

CHAP^ 26 TESTS FOR EVALUATION AND CONTROL

through it, through which a copper rod is passed. A tapered slot, cut into the upper surface of the block and filled with a nonmagnetic material constitutes an artificial defect of varying depth. The block can be used either in a coil for longitudinal magnetism check, or by passing current through the copper rod passed through the hole. Figure 227 shows the block in use for both tests. Bath strength 1s indicated by the length of the indication when a current of a specified strength is passed.

Following is the test procedure :

Process the block by the wet continuous method. If the bath is good, and the circuits are operating correctly, the following results should be obtained :

'Head or Coil Current Amoeres-D.C.

Approximate Length of Indication

% of defect length - Zero

'Five turn coil.

Failure to obtain the above results could be due to one or more of the following:

(1) Quality of the bath is poor-too strong or too weak.

(2) Head and/or coil circuit is not operating correctly.

(3) Operator's "wet continuous" technique is incorrect.

This device gives only qualitative information regarding bath strength, but does give an overall check on the equipment and the operator's technique. Other variations of this test dev~ce are in use and glve similar results.

Development work is proceeding, and it is probable that a fully satisfactory quantitative bath strength testing device will be available in the near future.

9. TESTS FOR BLACK LIGHT INTENSITY. When fluorescent particles are being used frequent checking of black light intensity is of great importance, if reliable and reproducible results are to be obtained. A detailed discussion of the causes of black light variations is given in Chapter 15, Sections 16 to 19, and in Chapter 16, sections 14 to 18. Black light intensity is measured by exposing a photo-


voltaic cell light meter to the radiation of the black light source held a t a specified distance from the meter, with white light excluded.

Light meters used are:

Weston Model 703 Type 3, Sight Light Meter.

General Electric Light Meter, Model 8DW 40Y 16.

The meters are calibrated directly in foot-candles of white light. The reading is therefore not truly in foot-candles for black light. It is, however, quantitative and reproducible, and therefore gives reliable comparative results. Meter readings of not less than 90 foot-candles at the point of inspection are considered minimum for satisfactory br~ghtness of fine fluorescent indications.

APPENDIX

The article which is printed below is the last bit of technical writing produced by A. V. de Forest, Founder (with F. B. Doane) of Magnaflux Corporation, before h ~ s death in April of 1945. The paper was to have been giveil by him a t a Society meeting in Feb- ruary of that year. However, due to illness, he was unable to do so, and the author of this book read the paper in his stead. It has not been published to date, and is reproduced here in its entirety in A. V.'s original language.

The subject matter is basic to our interests in the detection of flaws and the causes of failure in metals. The discussion is clear and concise. It would be difficult to find a more comprehensive dis- sertation on this subject, which a t the same time is as interestingly written and as understandable. After reading it, the varlous phases in which defects affect the strength of materials-and in which the properties of metals in their various aspects influence the effect of defects-are much more clearly understood.

The prime reason, however, for the inclusion of this paper in a book such as this, is the fact that i t is a truly classical bit of technical exposition. The substance of the paper, written over twenty years ago, is still as pertinent today as i t was in 1945, notwithstand- ing the great increase in our knowledge in these areas.

We feel that reading of this paper will be well worth while for anyone interested in the subject matter of this book.

Carl E. Betz

DEFECTS AND THE STRENGTH OF MATERIALS

by Alfred Victor de Forest. 1945.

Before it is possible to discuss defects and their relation to strength, it is necessary to understand clearly what is meant by these two simple words, defects and strength. As discussed here, both are related solely and directly to use. In this sense: therefore, a pal* is defective if it will not perform the function for which it 1s intended, quite regardless as to whether or not i t meets an arbitrary set of specifications. Strength, likewise, will be used to mean ability to resist successfully the forces for which the part was designed under the circumstances under which it was intended to serve a useful purpose. We will have to limit the discussion to thin mm-

- - - -. . . paratively narrow field or we shall be led fa r from the known paths of engineering.

Strength in metals is measured in many terms 1 tensile: compression, shear, torsion, and in brittle materials, bending as well. All these manifestations of strength are likewise defined by the rate a t which loads are applied. If loads are applied for months and years, creep strength may be the all important factor. This is especially true when metals are used a t elevated temperatures as in steam turbines, fired pressure vessels and especially gas turbines. Rates of loading for test purposes which require a few mlnutes to a few hours are misnamed "static" tests, and are those most commonly referred to, again erroneousiy, as physical tests, while in truth they are mechanical tests devised and used by mechanical engineers, and not by physicists.

Tests made a t high rates of loading, where the metal is broken in less than a second, and sometimes in tenths of thousandths of seconds, are termed impact tests. The higher rates of loading when applied to armor and projectiles are also termed ballistic tests. The highest rates are those caused by the action of a detonating explosive applied directly to a metal as when a block of T.N.T. rests directly on the surface and is exploded by a pnmer. The rate of travel of the pressure wave in the expiosive is approximately 17,000 ft. per second, about the speed of sound in steel, and this rate of application of load is far higher than that reached in normal

476

DEFECTS AND THE STRENGTH OF MATERIALS

ballistic tests. Ordinary impact tests, such as the Charpy and Izod, apply a blow a t 15 ft. per second. Ballistic tests run from 800 ft. per second for revolver ammunition to 2500 ft. per second for rifles of normal variety. Velocities up to 5000 ft. per second are obtainable under special conditions. Explosions or detonations may operate up to 20,000 ft . per second.

The most important aspect of strength as applied to moving parts is termed endurance or fatigue strength. About one hundred years ago it became evident to engineers that metals did not stand un- limited repetitions of ioad even though this load was below the elastic limit. Very many investigators using a great diversity of testing machines have studied the effects of load reversals or load repetitions both from the point of view of learning the fatigue strength of the different metals and alloys and for the purpose of investigating the fatigue strength of full-sized parts. An excellent review of the subject is given in the book "Prevention of Fatigue of Metals" prepared by the Battelle Memorial Institute and published by John Wiley & Sons in 1941. As Magnaflux inspection is applied very frequently to moving parts, it is important to considel- the various aspects of fatigue strength, because the factors which influence this are by no means the same factors which influence static or impact strength.

In this discussion we will confine ourselves to the magnetic materials, that is, the ferritic steels.

The fatigue limit is the maximum stress a metal can withstand when the load is applied an indefinitely large number of times.

The fatigue strength of smooth specimens without notches or fillets is closely proportional to the ultimate tensile strength and under conditions of complete reversal of stress as produced in a rotating bend machine, the fatigue limit is found a t about 45% to 50% of the ultimate; meaning that a 100,000 lb. per square inch steel will have a fatigue limit between 45,000 and 50,000 lbs. per square inch, regardless of the position of the yieid point or the elastic limit. In extremely hard and strong steels containing mar- tensite and with tensile strengths above 250,000 lbs. per square inch, this proportion drops off and the fatigue limit lies around 35% to 40% of the ultimate.

PRINCIPLES OF MAGNETIC PARTICLE TESTING - The significant factor is that fatigue strength has no relation

whatever to ductility, for the growth of fatigue cracks takes place with extremely small amounts of local deformation, in most cases too small for observation. The fact is well known but has not been completely explained. The exact mechanism by which the fatigue crack is produced is as yet unknown and so far there is no known method of inspecting a partially fatigued specimen in which an actual fatigue crack has not yet begun to grow. Very small fatigue cracks have been observed, of a length of only five thousandths of an inch, and their microscopic appearance is exactly the same as that of large craclcs. There is undoubtedly a change in the metal a t the location a t which the fatigue crack will ultimately occur but what happens has as yet escaped positive proof.

The subject is complicated, as is shown by experiments where the specimen is loaded to a high fiber stress for a small number of repetitions of load and then subsequently tested a t loads close to the fatigue limit. I t is certain that in many cases the fatigue strength, as shown by the final test: is improved by not too many cycles of high stress, possibly by local yielding a t high stress areas. When the high stressed level 1s repeated sufficiently, often the final fatigue strength is iowered.

I t is also known that if the specimen is repeatedly loaded below the fatigue limit, the final endurance strength is in many cases raised, so that it would appear that the final fatigue limit is shifted one way or the other by previous strain histories. As there is no method of measuring directly what state the specimen is in, no engineering advantage can be taken of this situation and i t is not considered that an old crankshaft is better than a new one by virtue of its previous history. The possible improvement is, in most cases, small, less than lo%, and the same conditions which produced the improvement might also be detrimental if the number of cycles and the stress limits had been above the damage line. As fa r as ordinary engineering practice is concerned, a part which has long been in use but contains no fatigue cracks may be considered as sound a s it was originally.

All experiments show that all fatigue cracks grow slowly, especially so in steel. It may require an additional 10% to 40% repeti- tion of load to break a fatigue specimen after the appearance of the first fatigue crack. The rate of growth of these cracks has not been

478

APPENDIX DEFECTS AND THE STRENGTH OF MATERIALS

adequately investigated but there is some evidence to show that different steels have different rates of growth and that these rates of growth are independent of the actual fatigue limit. From theoretical considerations i t might be supposed that this rate of growth is in some way connected with the damping capacity or hysteresis loss in the metal, (i.e. the ability of the steel to absorb energy). At any rate, the materials which owe much of thew strength to cold work, that is, wire and cold-rolled products, generally have a higher damping capacity and a slower rate of growth of fatigue cracks. Materials whlch are highly elastic, such as fully heat treated steels, and precipitation hardened alloys, show a more rapid rate of growth of fatigue cracks but in this field experiments are very limited. Further discussion of this point, and much detailed information on fatigue can be found in the book "Prevention of Fatigue of Metals", (previously referred to). The low rate of growth 1s a very important factor for the engineer, for it has been found that on repeated inspection a t overhaul periods many small fatigue cracks are located before they have reached dangerous proportions.

One of the largest applications of Magnaflux inspection is in this field, where locomotive and railroad rolling stock, automotive and airplane parts are regularly inspected after a reasonable number of miles or hours of operation. It has been the universai experience on the railroads that such periodic inspection is well worth while, for in addition to preventing a possible accident, the small fatigue cracks when found may be machined, chipped or ground out and a smooth fillet provided instead of a sharp-bottom crack. The part is then returned to service and frequently operates for very long periods, if not indefinitely.

Many experiments indicate that the frequency of load application is unimportant, a t least until the part reaches some 40,000 repetitions of load per minute. At these very high rates of loading, most materials tend to become hot, even though below the fatigue limit, and the fatigue limit tends to increase. These high frequencies are not met with in large parts but are primarily interesting when an attempt is made to build very high-frequency fatigue testing equipment in order to get through the test in a shorter time. This may be done on small specimens and some testing equipment has been built to operate above 30,000 cycles a minute. As i t is the number of


cycles that counts, in many cases of slowly repeated loads the total fife of the structure may not be sufficient to reach the actual fatigue limit and in this case the permissible stress may be higher than would be the case in hlgh-speed machinery.

3. CYCLES TO REACH THE FATIGUE LIMIT

In the steels under discussion the fatigue limit is reached after ten million cycles, provided the part is kept free from corrosion. That is: for any given loading, failure will occur within ten million cycles if the loading is a t or above the fatigue limit for the specimen. This rather definite and limited number of cycles is of great advantage, and fatigue tests can be discontinued if the part has not failed after this number of loads. In the case of the non-ferrous materials, such as the aluminum and copper alloys, this limitation does not exist and the fatigue limit decreases continuously as the number of cycles is increased up to a billon or perhaps more. That is, loading, to be under the fatigue limit, must take into account the expected number of stress cycles the part may have to sustain throughout its life. The reason for this situation has not been explained.

It is the universal experience that individual specimens show far more variations among themselves when tested in fatigue than when tested under ordinary static conditions. There IS, therefore, more uncertainty a s to the life of a part under fatigue than can be explained by the extent to which the steel is not uniform as shown by static loading. The reason for this undoubtedly is that the fatigue crack is initiated by very small variations in surface conditions or local metallurgical circumstances, such as particularly oriented grain boundaries. It is, therefore, to be expected that if one hundred specimens are tested in exactly the same manner, the weakest specimen may easily be 10% below the average. The stress level between the weakest and the strongest in any set of tests is called the scatter band and, quite obviously, good engineering requires that the width of this scatter band be taken into consideration when designing a part. The width of this scatter band may be particularly great in complicated components, such as ball bearings in which fatigue may take place anywhere on the surface of any ball or on the mating surface of raceways. This occurs even though the ball

APPENDIX ! DEFECTS AND THE STRENGTH O F DIATERIALS

bearings are more carefully heat treated and sized and made from cleaner steels than any other engineering components.

Under almost all conditions of fatigue loading, cracks start from the surface of the parts unless some particular treatment has been used to raise the strength of the surface. Smooth specimens are stronger than rough ones but the degree of roughness is difficult to measure. I t appears probable that the sharpness of the notch in the surface is the most ~mportant factor. A ground surface may be smoother than a machined surface but the sharpness of the fine notches or scratcnes left by the abrasive grit may be greater than those left by the machine tool. The direction of the notch IS likewise important and rather coarse grinding which lies parallel to the direction of stress is f a r less detrimental than a smooth appearing grind in which the grlnding marks are a t right angles to the stress. It is for thls reason that airplane engine parts are, whenever possible, finished so that the grinding marks run parallel to the major stress direction.

Corroded or etched surfaces are particularly detrimentai and where the fatigue limit is important, surfaces produced by casting or forging are smoothed by machining or grinding.

In addition to the smoothness of finish, any metallurgical change between the surface and the underiying metal is particularly important. A decarbur~zed layer a thousandths or two deep on the surface of strong steel very greatly reduces the fatigue strength, for a small fatigue crack forms in the weak decarbonized layer which then proceeds through into the underiy~ng metal. This condition is true even thougli the surface is protected by a hot-dip coating or an electro plate. Fatigue cracks may start easily in the electro plate layer and proceed from there Into the strong steel underneath.

Usually all electro plating is detrimental to fatigue life as is likewlse hot-dipped galvanizing. In the latter case, fatigue cracks occur in the brittle iron-zinc alloy whlch can propagate into the steel even though the thickness of this layer is less than a thousandth of an inch. These protective coatings, however, may serve a very usefui purpose in preventing direct corrosion.

PRINCIPLES O F MAGNETIC PARTICLE TESTTNC

There are many methods of improving the strength of the surface of steel, for instance, by carburizing, nitriding, cyaniding and flame or induction hardening. In all of those instances the fatigue strength of the surface of the part may be very materially increased, some. times to such an extent that the fatigue crack forms a t the junction between the hardened surface and the normal underlying steel. In this case, whiie the fatigue crack actually begins below the surface, i t progresses rapldly outward through the strong layer of the external surface, and in any event the crack lies close enough to the surface to be readily found by Magnaflux inspection.

Another method of strengthening surfaces has become extremely widely used-the improvement by means of shot peening. Surfaces treated in t h ~ s manner are somewhat roughened by the shot blast but the surface is hardened by cold work and, therefore, strength- ened; and a t the same time the surface is put into initial compression so that under repeated loads the maximum tension is iess than would have been the case in the absence of such treatment. I t is as yet impossible to say what proportion of improvement is due to the changed stress condition and what to the increase in strength and hardness due to cold work, but in any event shot peening can markedly improve the fatigue strength of parts whether of steel or of other metals. I t is current practice in many cases to shot blast instead of grind as a method of improving forgings, for it has been found that a rough forg~ng, shot blasted, may be as good as or better than the same forging after a careful surface grind. Quite obviously, in parts of complicated shape, shot blasting is f a r more economical. I t must be remembered that this shot blasting must be done with proper care and control: for there i s a very distinct maximum in the results obtained by variations in the size of shot, the velocity of the shot and the length of time of biasting.

A11 metais show a great reduction in fatigue strength if the part is subjected to corrosion during the application of load. It would appear that surfaces under repeated stress are much more readily attacked than surfaces under static load. Furthermore, the attack is continuous and the fatigue limit continues to decrease way below the normal fatigue limits. One can, therefore, say that there is no real fatigue limit if a part is tested under water. The strength will. be reduced continuously d u r ~ n g the time of operation. If the part

DEFECTS AND THE STRENGTH OF DiATERIALS

is steel, ope~ating in salt water, the reduction 1s very rapid and good engineering demands absolute protection against such conditions. Many englneering parts have to operate in an unprotected manner, for instance, hollow drill steel in ~ h i c h water is circulated to remove stone dust from the bottom of the hole. After prolonged operation, fatigue cracks start from the inside diameter of the drill steel. Sucker rods for pumping oil wells likewise have to withstand corrosion fatigue and have a rather limited life. Tail shafts of ships are protected from direct contact with the water inasfar as possible by bronze liners and rubber gaskets of many different types. In spite of the fact that such shafting operates a t a very low fiber stress, failure is inevitable if tile water tightness of the packing is not perfect.

In the same way turbine blades operating in steam have a tower fatigue limit than they would have in air and the fatigue resistance of these blades must be studied in relation to corrosion rather than as to stress cycles. There is strong evidence that fatigue testing in air results in lower fatigue limits than though such tests were conducted in a vacuum, brass, for instance, showing a 20% improvement in some cases when tested in the absence of air. While this is interesting, fatigue testing is usually carried on in an air atmosphere without humidity control, although perhaps humidity control would reduce some of the variations which are normally found.

8. ~TOTCHES AND STRESS RAISERS

Another great source of variation between static testing and fatigue testing is in the very important effect of notches in relation to fatigue. A notch produced by a change in section on the surface serves to concentrate the stress and in this manner produces a very definite lowering of the fatigue strength. For ~nstance, a small hole in a comparatively large plate under tension produces a three to one stress concentration. This means that the stress immediately adjacent to the hole is three times the average stress in the plate and in calculating the fatigue strength of a pal+ with an oil hole, such factors must be taken into consideration. A stress concentration a t a fillet will, of course, depend on the radius of the fillet and the relative diameters of the two parts connected by the fillet. A square notch is, therefore, more dangerous than a large radius a t such a transition.


In most bolts there is a square notch under the head, unprotected by any fillet. Such a condition is extremely weakening to the fatigue strength if the bolt is in repeated tension. A fillet a t this point will markedly increase the strength of the bolt and such fillets are provided in proper designing of connecting rod bolts. The thread end of a bolt is likewise a series of notches and has a very detrimental influence on fatigue strength. A "V" bottom thread, especially if produced by grinding, is far more detrimental than a rounded root as in the Whitworth system. The stress raising effect of such mechanical notches may easily be measured by photo-elastic means which accurately measures the stress distribution a t notches of all types where the stress can be considered in a plane.

Inasmuch as stresses computed from the theory of elasticity or measured by means of photo-elasticity assume the material to be perfectly elastic! it is found on fatigue testing that the notch severity is not quite as pronounced as theory indicates, because the metals themselves are not perfectly elastic. The less elastic the material, the greater is the discrepancy between the theory and practice. This means that the harder, stronger and more elastic metals are more injured by notches than the softer and less perfectly elastic metals. Fatigue tests are, therefore, conducted to determine the notch sensitivity of different materials and a knowledge of this factor is important for proper selection. It frequently happens that a less notch-sensitive carbon steel is superior in the notched condition to a stronger and more elastic alloy steel on account of this notch effect. It is important to realize that the shallow notches, extremely sharp a t their roots, are more detrimental than deeper and more rounded notches and the defects found by Magnaflux are primarily detrimental to fatigue because of their extreme sharpness rather than on account of their depth.

The notches caused by grinding cracks and heat-treating cracks are, by their nature, extremely sharp regardless of their depth and, therefore, constitute the worst form of stress raiser. Notches caused by coid shuts in forgings or laps in rolled bars may also be sharp and, if they run perpendicular to the direction of applied load, constitute a very serious weakness. The lap resulting from an over fill in rolling bar stock lies parallel to the direction of rolling. If this bar stock is made into a bolt which is primarily loaded only in tension or shear, the presence of the notch may be quite unimportant, merely on account of the direction of stress. If: for any

A P P ~ D I X DEFECTS AND THE STRENGTH OF MATERIALS -

reason, this same stock were used in repeated torsion, as is the case in a helical spring, the notch 1s then a serious source of weakness. In order to decide whether or not any particular discontinuity in the metal is important and constitutes a defect, i t must be considered in relation to the direction of loading.

Notches are likewise caused by the presence of non-metallies in steel. These non-metallies are distributed in the direction of rolling or forging and, if they are parallel to the stress, may be relatively unimportant, becoming more important as the stress is applied a t right angles. For instance, seamless tubing used in tension or compression may contain longitudinal non-metallies which are quite harmless but if this tubing is loaded by hoop tension, as is the case when fluctuating hydraulic pressure is applied, the exact same non- metallic~ may constitute serious weakness. The same argument holds in regard to slight imperfections in welded tubing. These may he extremely serious under internal-pressure conditions in the tubes, as for instance, in Diesel engine oil injection lines, but if the tubing is used for airplane construction, where loads are dis- tinctly longitudinal, the same discontinuities may be harmless. I t is to be noted that bars or tubing in torsion are ser~ously weakened by longitudinal discontinuities, for in torsion the maximum tension stress lies a t an angle of 45' to the axis. Helical springs must, therefore, be particuiarly carefully inspected and all serious longitudinal discontinuities removed.

Non-metallies in steel, particularly the manganese suphide type, present rounded surfaces rather than sharp discontinuities and are, therefore, realtively harmless as compared with cracics, cold shuts, or deep seams. In almost all cases the stress raising effect of the non-metallics is far less than the stress ra~sing effect of notches put in by improper design. If a part, such as a bolt or crankshaft, already contains mechanical notches of considerable severity, there will be no weakening produced by non-metallics whose effect is considerabiy less than that of the mechanical notch unless the non- metallic occurs a t a position of maximum load. I t is, therefore, customary to ignore non-metallics which occur elsewhere than a t known high stressed positions. A severe non-metallic stringer a t an oil hole may be dangerous, whereas in the metal a short distance from the oil hole it may be harmiess.

There is a relationship between grain direction and stress under fatigue loading, cross-grain metal being weaker than with a longi-

PRINCIPLES OF bIAGNETIC PARTICLE TESTING

tudinal grain, primarily because the distribution of the non- metallic~ follows the flow lines of the steel in complicated forgings, The direction of the flow lines is frequently important. These flow lines may be shown by Magnaflux inspection where a heavy current and the wet continuous method is used. The same information can also be found by deep etching. Inasmuch as the flow lines in the steel are always present in forgings, Magnaflux indications of the flow lines must not be considered a source of weakness but rather as an indication of the structure of the metal. Investigation of flow lines sl~ould be of interest to the metallurgist rather than the inspector.

While non-metallics themselves may not he of importance in many cases, they may lead to serious difficulty. For instance, steels which are quenched outright may be cracked during heat treatment by the stress raising effects of non-metallics and in this case the effect of the non-metallic may be serious as an indication of steel quality. The same is true of ball-bearing steels in which the presence of non-metallics a t the worklng surface of the ball may lead to early failure.

Magnaflux inspection is much used in determining the quality of steel in regard to its non-metallic content. It is quite customary to inspect aircraft quality steel and tool quality steel by this means, not because the non-metallics are a source of weakness in the bar which is inspected, but purely because such an inspection is quicker and cheaper than examination by means of microscope. Such tests are again of a metallurgical rather than of an engineering importance.

9. INFLUENCE OF A SERIES OF NOTCHES

It is well known that a series of notches is less detrimentai to fatigue strength than a single notch of the same shape and depth. A part which is threaded throughout its length is, in fatigue, stronger than a part containing a single notch or which is threaded only for half its length. A single non-metallic of considerable sjze in an otherwise clean steel is, therefore, a more severe stress raiser than a well distributed system of non-metallics throughout the material. In the same way, a slngie blow-hole in otherwise sound metal is more detrimental than a large area of porosity caused by small blow-holes. In general, blow-holes have rounded contours and

APPENDIX

DEFECTS AND THE STRENGTH O F MATERIALS

do not constitute a severe source of weakness. Porosity and blow holes in welds are, therefore, relatively harmless although they may appear to be very pronounced under x-ray examination. Sand pockets act in the same way in castings. Shrinkage cracks in castings, however, are usually intergranular and constitute sharp notches, injuring the material f a r more than porosity and sand. Such shnnkage cracks lying near the surface are usually removed and repairs are made by welding.

It has long been recognized that gray iron is remarkedly free from notch sensitivity. This behavior is usually explained by the fact that the material already contains many interior notches in its grain structure a t the ends of the graphite flakes. As these notches are more severe than the external notch caused by suitable fillets, the effect of the mechanical notch 1s obscured by the original weakness of the metal. Here again we have a system of a very large number of small notches, the behavior of which is included in the characteristics of the metal.

The same may be said of free cutting steels in which non-metallies are introduced for the purpose of breaking the chip during machining. Such steels, although extremely dirty in the usual sense, are in fact quite trustworthy, for their weakness is generally known and considered in the design. If, however, a part is designed for clean steel and a substitute IS then made to free machining steel; the results may be unsatisfactory, particularly if the load is not directly parallel to the rolling direction. Free machining steels are, therefore, substantially as good as clean steels in tension and compression

but are weaker under transverse loads, or torsion.

It is now known that there is a relationship between rate of loading, temperature of steei, and severity of a notch, all steels becoming more brittle as the temperature is lowered. The transition from the ductile to the brittle behavior is very clearly marked and is called the drop-off point. The temperature a t which this occurs i s characteristic of the type of steel used, its alloy content and its method of deoxidation in the furnace. In general, Bessemer steel has a higher drop-off point than open hearth steel and rimmed steel is also more sensitive to lowered temperatures than fully killed steels. The addition of alloys, particularly nickel, lowers the drop-off point.


Whatever the inherent temperature sensitivity of the steel, if it breaks in a brittle manner a t a higher temperature, the more severe is the notch condition, and exactly as in the case of fatigue, the sharpness of the notch rather than its depth, is important. Parts which are to operate under impact, particularly under low temperatures, must, therefore, be free of stress raisers, both those produced by design and those produced by accident. Non-metallics may, therefore, be important as in the case of fatigue, if the stresses are not parallel to the rolling direction, particularly a t low temperatures. Very much less information is available here than in the case of fatigue but direct experiments on full-sized parts may always be used to substantiate any questionable conditions.

Increased brittleness due to low temperature, leading to failure under impact, is characteristic of ferritic steel only. Non-magnetie steels do not show this effect nor do any other engineering alloys, and containers for liquid oxygen and other low-temperature chemical equipment are not made of magnetic steel.

All metals exhibit two forms of strength, one the shearing strength in which different elements of the structure slide over each other without immediate failure thereby allowing a change of shape under load. When the shearing strength is relatively low, a metal is said to be ductile. Metals also have a definite cleavage strength, meaning the force required to pull the structure apart under direct tension. All brittle materials, both metals and such substances as glass and rock, fail by cleavage. Whether a metal fails in a ductile or brittle manner depends on the relation between the two types of strength. In all magnetic steels the shearing strength increases as the temperature is lowered, more rapidly than the cieavage strength, and when the two strengths are equal, OY the shear strength is the higher, a brittle failure results.

The effect of rate of loading is likewise to increase the shear strength so that the rate of loading required influences the drop-off point in the same way as a lowering of temperature. This characteristic of steel has long been known. Steel rails, for instance, are more prone to break the lower the temperature a t which they are operated. Low carbon steel, such as used in chain, becomes un- dependable at low temperatures, and chain is frequently heated before belng used on wrecking cranes or other equipment operated in severe winter weather. A careful workman likewise warms his

APPWDK DEFECTS AND THE STRENGTH OF MATERIALS

cold chisel before chipping when the weather is severe. It is generally recognized that welded steel ships are more apt to sustain severe damage from cracks in cold weather.

From all that has been said it is evident that before i t can be determined whether or not a discontinuity constitutes a defect, it is necessary to know the conditions of service, including both the amount and the direction of those stresses for which the part is intended. As fa r as the stress distribution in service parts is concerned, this subject is in the hands of a relatively new profession, that of stress analysts. The new engineering tools of Stresscoat and the wire strain gage have made it possible to determine accurately what stresses are encountered in service. The more accurately these factors are known, the more fully the distinction may be made between harmless and harmful conditions. This a r t is rapidly ad- vancing and should form a part of the education of all those dealing with the strength of materials.

I t is customary for engineers to design parts with a certain factor of safety. When a failure may be extremely costly it is customary to either use a large factor or inspect the part with additional care. For instance, the failure of an airplane engine part may result in very great damage as compared with a corresponding failure in an automobile. It IS, therefore, good engineering to ask for a more rigorous Inspection of airplane parts than of automotive parts, particularly if the failure of the automotive part does not endanger the vehicle. Springs may, therefore, be used a t a lower factor of safety and with less rigorous inspection than parts of the steering gear or of the brake system. Much skill and judgement is requlred to reach the best compromise between weight, cost and safety.

A. V, de Forest Feb. 1945

A P P E N D I X

BIBLIOGRAPHY

1. SAXBY, S. M., Magnetic Testing of Iron. Engineering, Vol. 5, P. 297, 1868.

2. BURROWS, C. W., On the Best Method of Demagnetizing Iron in Magnetic Testing. Bulletin 78, United States Bureau o f Standards, Pp. 205-274, 1908.

3. STEINHAUS, W . & GUMLICH, E., E x p e ~ i m e n t d Investigations in the Theory. of Ferro-Magnetism. Verhandlungen der Deutschen ~ h ~ s i k a l i s c h e i Gesellsehaft, No. 21, 1915.

4. STEINMEW,, CHARLES P., Transient Electnc Phenomena and Oscillation. 1920.

5. MITRA, S. K., On the Demagnetization o f Iron by Electromag- netic Oscillation. J . de Physique, P. 259, 1923.

6. SPOONER, THOMAS, Properties and Testing of Magnetic Mate- rials, 1927.

7. TSCHETVERIKOVA, M., Demagnetization of Iron Compounds by Electric Oscillation. Ztschr, f . Physik, 44, No. 139, 1927.

8. CULVEE, CHAS. A., Electricity and Magnetism. October, 1930.

9. WATTS, T . R., Magnetic Testing of Butt Welds. Journai o f the American Welding Society, Voi 9, No. 9, P. 49 f., September, 1930.

10. WILLIAMS, SAMUEL R., Magnetic Phenomena. 1931.

11. GEROLD, E., Kritische Betrachtung der Magnetischen Verfah- ren fuer Werkstofpncefung. Stahl und Eisen, Vol. 51, No. 14, P. 428, April 2; 1931.

12. MEHL, R. F., Nondestructive Testing Methods in the Iron and Steel Indus iq . Yearbook American Iron and Steel Institute, Pp. 126-184, 1932.

13. DE FOREST, A. V., Magnetic Testing of Iron and Steel. Iron and Steel Engineer, Vol. 9, No. 4 , P. 170, April, 1932.

491


14. DE FOREST, A. V., Magnetic Test Locates Flaws in Valve Springs. Iron Age, Vol. 130, No. 3, P. 107, July 21, 1932.

15. DOANE, F. B., Magnetographic and X-ray Tests o f Pipe Welds Compared. Iron Age, Feb. 2, 1933.

16. JACOBS, F. C., A Crack S l e u t L M a g n e t i c Powder Inspects Turbine Blading. Power, May, 1933.

17. DOANE, F. B., Magnaflux Inspection for Cracks and Seams. Iron Age, September 26,1933.

18. MCCUNE, C. A., Magnafluz in the Transit Field. Transit Jour- nal, September 20, 1933.

19. VON SCHWARZ, M. & KRAUSE, J., Magnetische Untersu- chung z u m Fehlernuchweis in Ferromugnetischen Werkstoffen. Masehinenschaden, Vol. 11, No. 7, P. 107, 193.4.

! i 20. HATFIELD, H. S., T h e Action of Alternating and Moving Mag-

i netic Fields Upon Particles of Magnetic Substances. Physics Society (London) , 46 (5), P. 604, 1934.

21. DE FOREST, A. V., Magnafluz Testing for Discontinuities in Magnetic Materials. U.S. W a r Dept., Air Corps Technical Report, 1935.

22. DEFOREST, A. V., Dynamic Strength of Machine Parts Affected by Quality of Surface. Iron Age, Feb. 21, 1935.

23. DAVIS, C . W., Some Observations on the Movement and Demug- netization o f Ferromagnetic Particles zn Alternating Current Magnetic Fields. Physics, Vol. 6, June, 1935.

24. WARREN, A. G., Flash Magnetization for Examining Castings and Forgings. Engineering, Vol. 140, No. 3638, Pp. 353-354, Oct. 4, 1935.

25. DE FOREST, A. V., The Rate o f Growth o f Fatigue Cracks. Journal o f Applied Mechanics. March, 1936.

26. METROPOLITAN VICKERS ELECTRIC COMPANY, Fault Detection in Metals by Magnetic Methods. Engineering, V o f . 141, No. 3669, P. 504, May 8, 1936.

27. NOBLE, H. J., Magnajlux Inspection Methods in Airplane En- .gines. Metals and Alloys, Vol. 7, No. 7, P. 167, July, 1936.

492

BIBLIOGRAPHY

28. YANT, J . W.; The Magnaflux Inspection of Pressure Vessel Welds. Paper presented Oct. 23, 1936, at A S M E Symposium, Cleve-

I land, Ohio.

29. BERTHOW, R., Fault Detection by the Magnetic Powder Proc- !

ess. Engineering Progress, Berlin, Nov., 1936.

30. W I L H E L M , H. & ELSASSER, H., A Method of Detecting Cir- cumferential Cracks in Hollozu Cylinders. Mitteilungen Verein Grosskessel-Bezitzers, December 30, 1936.

31. HODGE, JAMES CAMPBELL, Welding o f Pressure Vessels. Jour- nal o f American Society for Naval Engineers, Pp. 519, 521, Nov., 1936.

32. BITTER, FRANCIS, lntrod~cction to Ferromagnetism. 1937.

33. SKILLING, HUGH H., Transient Electric Currents. 1937.

34. BERTHOLD, R. & GOTTFEW, F., A New Aid in Testing Welds. Stahlbau, Vol. 10, No. 4, P. 31, Feb. 12, 1937.

35. CAVANAGH, R . F., Magnaflux Inspecticm o f Boiler Drums and Unfired Pressure Vessels. Mechanical Engineering, Vol . 59, P. 153, March, 1937.

36. RATHBONE; T. C.: Detection of Fatigue Cracks by M a g n a f l m Methods. Mechanical Engineering, Vol. 59, P. 147, March, 1937.

i i

37. DIXON, E. O., Metallurgical Qualities in Forgings. Aero Digest, July, 1937. !

38. ADAMS, C. A., The Nondestmctiwe Tests o f Welded Joints. i Paper presented at the Eleventh General Meeting o f the National Board o f Boiler and Pressure Vessel Inspectors, a t New York . May 25,1937.

39. YAMIN, I. V., New Method of Detecting Cracks in Welded Seams. Zavodskaya Laboratoriya, Vol . 6, P. 1284, 1937.

40. DE FOREST, DOANE, MCCUNE, Nondestructive Tests o f Welds. American Welding Society Handbook, 1938.

41. WEVER, F. & HAENSEL, H., Beitrag zur Magnetkchen W w k - stoffpruefung. Mitteilungen aus dem Kaiser-Wilhefm Institut, Voi. 20, No. 8, 1938.

42. DE FOREST, TABER, Detection of Fatigue Cracks. California Oil World, March 26, 1938.

PRINCIPLES OF DLAGNETfC PARTICLE TESTING

I 43. HAENSEL, H., Die Fehlererkennbarkezt bei der hfagnetzschen 1 Zerstoerungsfrezen Pmefung. Archiv flier das Eisenhuettenu~ese~,

Vol. 11, No. 10, April, 1938.

44. GRIGOROV, K. V., Fixation of the Preczpitate of Magnetic 1% spection. Zavodskaya Laboratoriya, Vol. ?, Pp. 737-738, 1938. Ab- stract in Chem~cal Abstracts, Vol. 33, Pp. 101-109, 1939.

I 45. MCBRIAN, RAY, Magnetic Testing of Car and Locomotive Parts. Railway Electrical Engineer, July, 1938.

46. MCCUNE, C. A., Magnaflux Inspection of Gas Cylinders. Iron Age, July 14, 1938.

47. MCBRIAN, RAY, D. & R. G. W . Installs Magnaflux Inspection. Railway Electrical Engineer, September, 1938.

48. BERTHOLD, R. & SCHIRP, W., The Physical Bases of the Mag- netic Powder Process. Atlas der Zerstoerungsfreien Prufverfahren,

i published by Johann Ambrosius Barth, Leipzig, 1938.

49. BERTHOLD, R., Technical Aids of the Magnetic Powder Process. Ibid, Leipzig, 1938.

1 50. LEE, J O H N G., Airplane Demagnetization and Neutralization.

I Journal of the Aeronautical Sciences, Vol. 2, No. 12, Oct., 1939.

I 51. PORTEVIN, ALBERT M., Interpretation. of Magnetic Patterns. Metal Progress, April, 1939.

52. LESTER, H. K., SANFORD, R. L. & MOCHEL, N. L., Nondest?uctive Testing tn the United States of America. I.E.E. Journal, Vol. 84, No. 509, May, 1939.

I 53. BERTHOLD, R., Nondestructive Testing Based on Magnetic and Electrical Prznczples. I.E.E. Journal, Vol. 84, No. 509, May, 1939.

54. W H I T E , A. E., Changes in High-Pressure Drum to Eliminate Recurrence of Cracks Due to Corrosion Fatigue. Paper presented at the Annual Meeting of the ASME at New York , Dec. 5, 1938. Published in the Transactions, ASME, Aug., 1939.

55. MCCUNE, C. A., Detection of Flaws in Steel and Steel Parts. National Safety Council News, October, 1939.

56. GRAF, S. H., Industrially Significant Nondestructive Testing Methods for Engineering Materials and Machine Parts. Paper pre-

h

i b APPENDIX

BIBLIOGRAPEY :

/ sented to SAE, Seattle, Wash~ngton, March 10, 1939 and ASM, ! Portland, May 12, 1939. 1

57. T I M M O N S , J. J., Magnaflux inspection of the Power Plant. i Power Plant Engineering, April, 1940.

: 58. DE FOREST, A. V., Nondestructive Testing of Metals. Mining i and Metallurgy, July, 1940.

59. MCBRIAN, RAY, New Denver and Rio Grande Laborator?) in Step with Modem Progress. Railway Purchases and Stores, P. 457,

: Oct., 1940.

60. JAEKEL, W.; Magnetic Powder Method. Die Giesserei, Vol. 27, Pp. 262-265, 1940.

61. SCHRADER, H., Practical Experiences with Magnetic Powder Method of Detecting Cracks. Stahl und Eisen, Vol. 60, Pp. 634-645 and 655-660, 1940.

62. MCCUNE, C. A., Magnaflux Inspection of Railroad Steel Pwts . Brotherhood of Locomotive Fireman & Enginemen's Magazine, Vol. 109, No. 6, P. 397, Dee., 1940, Southern and Southwestern Railway Club Proceedings, July, 1940.

63. Principles of Magnaflux inspection. First Edition, DOANE, F. B., Magnaflux Corporation, January, 1940.

64. DE FOREST, A. V., Magnaflux on the Railroads. New England Railroad Club, April 8, 1941.

65. Recommended Practice for Magnaflux Rating. S A E Journal, Vol. 48, Pp. 18-19, January, 1941. S A E Handbook, P. 335.

66. Prevention of the Failure of Metals Under Repeated Stress. Battelle Memorial Institute, 1941.

67. Airplane Welding and Materials. Second Edition, JOHNSON, J . B., 1941.

68. Inspection of Metals. PULSIFER, HARRY B., 1941.

69. Magnafluz Aircraft Inspection Manuat. DOANE, F. B. & THOMAS, W . E., Magnaflux Corporation, August, 1941.

70. DAVIS, C. W., A Rapid Practical Method of Demagnetization, Involving High Frequency. Nature.


71. HASTINGS, CARLTON, The Magnetic Powder Method of Inspect- ing Weldments and Castings for Sub-surface Defects. Welding Journal, January, 1943.

72. THOMAS, W . E. & DE FOREST: T., Bolt and Nut Inspection by the Magnetic Particle Method. Published by Magnaflux Corporation, July, 1943.

73. BETZ, C. E., Use of Magnuflux by the Railroads. Proceedings, Southern and Southwestern Railway Club, September, 1943.

74. BETZ, C. E., Magnetic Particle Inspection. Canadian Metals and Metal Industries, May, 1944.

75. MAGES, M. L., Demagnetization During Magnetic Particle In- spection. Aeronautical Engineering Review, Vot. 3, No. 10, Oct., 1944.

76. OWENS, ZAMES E., Routine Inspection and Salvage of M 5 chinenj Weldments. The Welding Journal, October, 1944.

77. Tentative Method for Magnetic Particle Testing and Inspec- tion of Heavy Steel Forgings. American Society for Testing

1 Materials, Standards, 1944, Part I , A 275-44 T .

78. Tentative Method of Magnetic Particle Testing and Inspection of Commercial Steel Castings. American Society for Testing Ma- terials, Standards, 1944, Part I , A 272-44 T .

79. FFIELD, PAUL, Conditionzng of Steel Castings to Standards o f Quality. Transactions, American Foundrymen's Association, Vol. 52, Pp. 173-204, 1944.

80. Tentative Recon~mended Practice for Determining the In- clusion Content of Steel. American Society for Testing Materials, Standards, 1944, Part I , E 45.

i 81. COTTON, JOHN F., Magnetic Powder Inspection of Large Cast- zngs. Transactions, American Foundrymen's Association, 1944.

82. Magnetic Particle Inspection. Inspection Handbook for Manual Arc Welding; American Welding Society, 1945.

83. BROWN, A. L. & S M I T H , J. B., Failure of Spherical Hydrogen Storage Tank. The Welding Journal, March, 1945.

, APPENDIX

BIBLIOGRAPHY

84. EVANS, SIDLEY O., Quality Control Durzng Production of Elec- t n c Resistance Welded Tubing. The Welding Journal, September, 1945. I i 85. Symposiz~?n on Magnetic Particle Testing. Published by American Society for Testing Materials, 1945.

86. Transcript of First Railroad Magnaflux Conference. Published by Magnaflux Corporation, February, 1946.

87. MCMASTER: R. C. & BANTA: H. M., Progress Report on Drill String Research Nondestructive Testing of Drill Pipe. Battelle Memorial Institute, Drilling Contractor, August 15, 1946.

83. FREAR & LYONS, Nondestructive Testing of Castings. In- dustrial Radiography, Winter, 1946-1947.

89. WALSH, D. P., Practical Tool Inspection and Quality Control Methods. The Tool Engtneer, December, 1946.

90. MCERIAN, RAY, Testing with Magnafluz on the D. & R. G. W . R Y . Railway Eng~neering and Maintenance, February, 1947.

91. MUELLER, H . M. & YEAST, W . E., Magnetic Particle Inspection of Chromiu?a Plated Tools. Metal Progress, March: 1947.

92. Transcript of Conference on Weld lnspection with Magnafluz. Published by Magnaflux Corporation, May, 1947.

93. FICK, N. C., Present Methods and Trends with Notes on Literature. Metals Review, May, 1947.

94. THOMAS, W . E., Inspection Methods Using Magnafl~cz and Zyglo in Production Industrtes. Nondestructive Testing, Fall, 1947.

95. MARADUDIN, A. P.; Inspection o f Pressure Vessels and Tanks. Western Metals, April, 1948.

96. MCCUTCHEON, D. M., Application o f Nondestructive: T e s t i ~ ~ g to Auto7notivc Parts. Paper presented at S.A.E. Summer Meeting, June. 1948.

97. RODA, DONALD E., iltagnetic Particle Inspection m Enginrw- ing. Iron Age, Vol. 162, No. 6, August 5, 1948.

98. RODA, DONALD E., Standwds of Magnetic Pwticle 1n.spcct.ion. Iron Age, Vol. 162, No. 7, August 12, 19.18.

497


99. HITT, WM. , The Valzie of Scienti,iic Ins?~ection to Indzcstq. Western Machinery and Steel World: July, 1948.

100. Introdz~ction to Complex '17ariables. CHINCHILL , R. V . ; McGraw-Hill, 1948.

101. Pr~ncivles of Electricity. PAGE, L. & ADAMS, Y . I.; V a n Nostrand, 1949.

102. Electromagnetic Felds, Tfol. I: Mapprng of Fields. WEBER, ERNST; John Wiley & Sons, 1950.

103. Static and Dynamic Electqicity. SMYTHE, W . R.; McGraw. Hill: 1950.

104. Handbook of Expe?imental Stress Analysis. HETENYI, M.; John Wiley & Sons, 1950.

105. CAINE, JOHN B., What's Wrong with Castings? Found~y, June, 1951.

106. SWEET, JOHN W., Weldment Inspection in Aircraft Constme- tion. Nondestructive Testing, Fall, 1951.

107. GILL, STANLEY A., Various Inspection Methods Used in Heat Treating Shops. Metal Treating, May-June, 1952.

108. Dictionary of Conformal Representations. KOBER, H.; Dover, 1952.

109. Conformal Mapping, NEHARI; Z.; McGraw-Hill, 1952.

110. Electronbagnetics. KRAUS, JOHN D. ; McGraw-Hill, 1953.

111. THOMAS, W . E., Econonlic Factors of Nondestructive Testing. Nondestructive Testing, March, 1953.

112. ALLEN, ARTHUR H., Improved Tools Expedite Magnetic Par- ticle and Penetrant Inspection. Metal Progress, January, 1954.

113. OYE, LLOYD J., Nondestructive Testing of Structures. The Welding Journal: March, 1954.

114. MCCLURG, G. O., Theory and Application of Coil Magnetiza- tion. Journal Society for Nondestructive Testing, Jan.-Feb., 1955, 11 23-25.

115. MIGEL; HAMILTON, Magnetic Particle, Penetrant and Related inspection Methods as Production Tools for Process Control. Steel Processing, February, 1955.

116. W I L S O N , T . C., Nond~~strzccfzve Trstiny o f Refinery E ~ ~ u i p - a~ent . The Petroleum Engineer, Joky, 1955.

117. S M I T H , 0. G., Magnetic Partzcle Techncqzfe Makes Billet In- spectzon Positme and Eficcent. Iron Age, RZay 5, 1955.

118. DEVRIES, A. J.: Productizre Inspection ASSZL~L'S qua lit^ and Reduccs Scrap. Ch~cago Mid\vest Metal Worker, January, 1956.

119. CATLIN, FRANKLIN S., A'ondestmctive Sai?&plc Testing for Cracks .kids Her~f Treafzng. Metal Treating, AIarcll-April, 1956.

120. MCCABE, J . L. & HIRST, B., Use of Tap Wate, in Magnetic Inspection. Metal Progress, July, 1956.

121. ~ I A R R E R , J O H N R., Greater Acceptance of Welding Through the Use of Inspection Methods. The Welding Journal, March, 1957.

122. Standardization Azds Weld Inspection of Tractors. Editorial, The Welding Engineer, April: 1957.

123. Steel Hzents Harder for Flnzus. Business Week, May 4, 1957.

124. CATI,IN, F. S., Test Methods Valzcabie in improving Casting Destgn. Foundry, March, 1957.

125. The Making, Shaping and Treating of Steel. United States Steel Corporation, Seventh Edition, 1957.

126. Brittle Behavzor of Engineering Materzals. PARKER, EARL R.; John Wiley & Sons, 1957.

127. MARSCHALL, H. G.; Magnetic Particle Testing and Its Latest Means of Indication. Werkstatt und Bei-trieb, Vol. 91, Pp. 640-653, Nov., 1958.

128. Magnetic Particle Inspection. The Welding Handbook, Fourth Edition, 1958; American Welding Society.

129. Nondestructive Testing. HINSLEY, J . F.; McDonald & Evans Ltd., London, England, 1959.

130. Progress i n Nondestrz~ctive Testing, Vol. I. STANFORD, E. G. & FEARON: J . H. ; The McMillan Co., New York (Heywood & Co., London), 1959.

131. ARROTT, A. & GOLDMAN, J . E., Fundamentais of Ferro- Magnetism. Electrical Manufacturing, Pp. 109-140, March, 1959.


132. Handbook, Society for Nondestructive Testing, Vol. ZI, Sec- tions 30 t o 34, 1959.

133. Ferro-Magnetism. RICHARD M. BOZORTH; D. V a n Nostrand Co., Inc., 1959.

134. Physics of Magnetism. % S H I N CHIKAZUMI; John Wiley & Sons, 1959.

135. Black Light Pin Poznts Flaws in Tube-Conditioning Line. Iron Age, Editormi, Nov. 24, 1960.

136. CHRISTENSEN, A. E., Method Control i n Testing-Key to Reliability. Tool and Manufacturing Engineer, Dee., 1960.

137. STARR ROY C., Shop Quality Control Safeguards Automotive Sp.ring Manufacture. Canadian Machinery, May, 1960.

138. Allowable Stresses in Flash-Welded Joints. Product Engi- neering, Editorla], Oct. 31, 1960.

139. WALSH, D. P., Reducing Maintenance Cost Through Znspec- tion. Mining Congress Journal, August, 1960.

140. BETZ, CARL E., The Nondestructive Testing Engineer-To- day's Career Opportunity. Nondestructive Testing, Jan.-Feb., 1960.

141. Progress in Nondestructive Testing, Vol. ZI. STANFORD, E. G. & FEARON, J. H.; The McMillan Co.; New York (Heywood & Co., London), 1960.

142. Symposium on Fatigue of Aircraft Structures. ASTM, 1960.

143. Magnetic Transitions. BELOV, K. P.; Consultants Bureau, N. Y., 1961.

144. VANCE, SAM, Magnetic Particle Inspection at Regular Zn- tervals Can Prevent Major Troubles. Pipe Line Industry, February, 1961.

145. N I X O N , FRANK, Simple Fatigue Testing as an Essential Corollary to Nondestructive Testing. Nonde~tructive Testing, March-April, 1961.

146. BOGART, HENRY G., The Place for Nondestructive Tests zn the Field of Plant and Eqnipment Overhazll. Presented, Winter Annual sleeting, American Society of Mechanical Engineers, Dec. 1, 1961.

APPMDIX

BIBLIOGRAPHY

147. Magnaglo Takes a Close Look. Wireco Life , July-August, 1961, American Steel &Wire Div., U.S. Steel Corp.

148. Missile Parts Get Fast Check. Iron Age, Editorial, July 13, 1961.

149. EOBANKS, PAUL E., Making Efficient Use of Magnetic Par- ticle Inspection. Foundry, July, 1961.

150. Engineering and Technology Report-Nondestructive Test- ing. Factory Magazine, Jan., 1961.

151. E M I S H , C. F., Tubmg Quality is Built In-Tests Cannot Pro- duce It. Steel, June 5, 1961.

152. MULLER, E. A. W., Materzal-Pruefung Nach dem Magnet- Pulver Verfahren. Akademische Verlagsgesellschaft, Leipzig, 1961.

153. Metals Handbook. 9th Edition: Vol. I . American Society for Metals, 1961.

154. Analysis of Service Faih~res. Republic Alloy Steels Hand- book, Republic Steel Corporation, 1961.

155. SITARAIN, R. V., A Useful Engineernzg Theonj of High Stress Level Fatigue. Aerospace Engineering, Oct., 1961, P. 18.

156. Glowing Lines Outline Seams in Steel Billets and Blooms. Iron Age, Editorial, Oct. 4, 1962.

157. Magnetic Particle Testing of Malleable Castings. Report, AFS Malleable Division, Finishing and Inspection Committee (6 G ) , American Foundry Society Transactions, Vol. 70, 1962.

158. THOMAS, W . E., Conditioning Practices and Their Effect on Product Quality and Cost. Iron and Steel Engineer, Nov., 1962.

159. DAi;E, DON. The Crack of Doom. Sports Car Graphic, May, 1962.

160. MCPARLAN, Jos. L., How Nondestructive Testing Aids Power Station Mazntenance. Power Engineering, March, 1957, P. 57.

161. HARZ, J O H N J., HOZU to Find Trouble Before It Starts. National Engineer, August, 1962.

162. Billets Handled Automatically in Nondestructive Test Unit. Automation, November, 1962.

501

PRINCIPLES OF RIAGNETIC PARTICLE TESTING

163. Progress in Nondestructive Testing, Vol. III. STANFORD: E. G. & FEIARON, J. H . The McMitlan Co., New York (Heywood & Co.: London), 1962.

164. Principles and Practice of Nondestructive Testing. LAMBLE, J. H.; John Wiley & Sons, 1963.

165. WALORDY, ALEX, Magnaflux-The Key to Engine Reliability. Cars, April, 1963.

166. HENDRON, J . A., Automated Magnetic Particle Testing of Pressure Vessels. Nondestructive Testing, July-August, 1963.

167. BRIGGS, CHAS. W., Sigwi.fcance of Discontinziities in Steel Castings on the Basis of Destructive Testing. Materials Research and Standards, ASTM, June, 1963, Pp. 472-479.

168. G O H I N , G. R., Tlte Mechanism o f Fatigue. Materials Research and Standards, AST & M, Vol, 3, No. 2, Feb., 1963, P. 106. Fatigue o f Metals.

169. HARDRATH, H. F., Crack Propagation and Final Failure. Ihid., P. 116. Fatigue of Metals.

170. PETERSON, R. E., Engineering and Design Aspects. Ibid., P. 122. Fatigue o f Metals.

171. PARK, F. R., Fractz~re. International Science and Technology, March, 1963, P. 24.

172. Micronzagnetics. W M . FULLER BROWN; Intel-science Publica- tion, 1963.

173. BEZER, H. J., Magnetic Metlrods of Nondestructive Testing. The British Journal o f Nondestructive Testing, Vol. 6, No. 4, Dec., 1964, P. 109.

174. Metals Handbook, Vol. II, 8th Edition, 1964. American Society for Metals.

175. Sonics Unite with Magnetic Glow for Automatic Billet Test- ing. Iron Age, Editor~al, Oct. 15, 1964.

176. Merry-go-Rozind Test System. Tooling and Production, Edi- torial, September, 1964.

177. MIGEL, HAMILTON, Acceptance Standards for Nondestructive Tests. Mechanical Engineering, April, 1964.

602

A P P ~ D I X BIBLIOGRAPHY

178. Synzposiun~ on Nondestructive Testing i n the hlissile In- dustr?~. AST&X, 1964.

179. VLAHOS, C. J., Nondestructive Testing Pays Its Way in Maintenance. Mill and Factory, August, 1964.

180. HAYES, D. A,, The Use of Meclzan7.zed Nondestructive Testing in Quality Control. Given at Chicago Technical Meeting of American Iron & Steel Institute, October 20, 1965.

181. WULPI , D. J., HOW Components Fail-Modes of Fracture. Metal Progress, Vol. 88, No. 3, Sept., 1965, P. 72.

182. WULPI, D. J., HOZU Compo?tents Fail-Types of Loading. Ibid., No. 5, Nov., 1965, P. 83.

183. WULPI, D. J., HOW Components Faid-Effects of Va?-iables. Ibid., No. 6, Dec., 1965, P. 65.

184. Defects and Failures in Pressure Vessels and Piping. HELMUT THIELSCH ; Reinhold Publishing Corp., 1965.

185. Table of Specifications and Standards for Nondestructive Testing. Materials Evaluation, Vol. XXIV , No. 3, March, 1966, P. 158.

186. HOMER, RICHARD G., Assuring Quality Forgings by Non- destructive Testing. Metal Progress, March, 1966.

187. THOMAS, W . E., DeveIopment and Application of Nondestruc- tive Testzng zn the Aerospace Industry. Third Aerospace Confer- ence on Nondestructive Testing, Georg~a Institute o f Technology, Atlanta. May 2, 1966.

188. Testing Combination Sets High Production Standards. Tool- Ing & Production, June, 1966.

189. BETZ, C. E., Nondestrz~ctive Testing-Recent Achievements. industrial Quality Controi, Vol. 23, No. 1, July, 1966, P. 19.

190. Manual of Experz?nental Stress Analysis, Second Editzon. Edited by W . H. Tuppeny, Jr. and A. S. Kobayashi, SESA, 1966.

SUBJECT INDEX

Abrasive Wheel, f o r sectioning parts, 403

Acceptance limits, 459 Acceptance standards, 460 Accept-Reject Decision, 64, 417 Accept-Reject Standards, 338,

460, 461 Aircraft overhaul, 428, 429, 461

F.A.A. Requ~rements, 461 Air Force, U.S., 57, 438

Equipment Specs, 461 Air gap, defined, 125 All-over patterns, 383 Alternating current (A.C.)

Defined. 120 Demagnetizing with, 314 Firs t use of, for magnetizing, 54 Loops, for demagnetizing, 319 Magnetizing with, 54, 157, 193 Permanent Magnetization with;

167

Sensitivity compansons, with D.C. & Half Wave. 234. 236

Skin effect of, 152; 159, 195, 196

Sources of, 157 Vs. D.C., factor in Equipment

Design, 347 Vs. D.C., for masnetizlng, 151,

23 1 Vs. D.C., for surface and

subsurface cracks, 231, 232, 2 3 L , 234

Yokes. 145 , American Soc~ety of Mechanical

Engineers, 457 Anierrc;~n Soc~ety for filetals, 457 Americaii Soc~etv for Testing

I I I I ~ Rl;ltenilis, 457 Recommended Prod Spticings

Z I I I ~ eurrt*tit riliuas of, 205, 201;

S~~er i f ic ;~ t lons of, 463 Amerlc:lrb Soclety ol' Tool E11gl.s..

457 ~ & e r l c : ~ i i Welding Soc~cty, 457,

463

Ammonlum Persulphate Etch, 404 Ampere, defined, 120

Magnetizing, half-wave, 15 5 Ampere turns, defined, 124: 145

Rule for determining, f o r longitudinal magnetizing, 181

A m ~ e r p ~.~ -- .. Rule for determining, for

Circular magnetizing, 198 Prod magnetizing, 206 Per inch of prod spacing, 206

Analog methods, for field distribution. 175

Analysis, Experimental stress, 414 Appearance of Indications

Affected by spread of emergent field, 376

Factor in Interpretation, 343, RAA

~ppi icy t lons of Mag. Part. Testing Classification of, 419 Industrial, 419

Applying Magnetic Particles Dry Powders, 250 Factor in design, 345 Wet Bath, 266

Arc. Eye damage from, 153 Par ts damaged by, 151 Quenching of, on A.C. hreak,

157. 158 Austeiitic steels, 102 Automatic equipment, 333, 342

Advantages of. 343 Aiialysls of testing problems

for, 344 Bolts, testing with, 335 Compromises in design of, 341,

348

SUBJECT INDEX

Method, factors in desjgn of: 345 Pipe couplings, oil field,

testing with, 337 Reasons for uslng, 342, 343 Resldual vs. continuous method,

factor in design of, 348 Specification for design of, 348 Steps leading to design of, 343 Uses of, 243

Automatic accepting and rejecting, 417

.4utomotive applications, 430

Balls, bearing, Testing of; 349 Bars, rotor, aluminum, 438, 439 Bath, for wet method. 469

Appiying, 267 Cleanliness, Need for, 272 Costs of. 259 Deterioration of, 266 Dirt in, 2F3, 265, 266 Drag-out ioss

of bath liquid, 346 of magnetic particles, 265

Dumping of. 266 Errors in making up, 467 Evaporation of, 265 Foam~ng of, 346 Incorrect strength of, 468 Maintenance of. 265 Mak~ng up, directions for, 263,

7G5

~ i i i i r . 257, 258 Characteristics of, 258 Specifications for, 220. 469

Pre-mixed, 224, 272 Skln, protection from, 271 Strength of, 261

Chart for, 263, 277, 471 Incorrect, 468 Tests for, 262, 278, 469, 472

Rust preventives in, 258, 271 Uniformity of, 261 Water used for. 58. 220, 256.

346 Wrong particles used in, 467

Bearing Balls, testing of, 349 Betlr~ng Races, testing of? 350 Berthold field gauge, 170 Bibliography, 176, 475 Billet testing. 345. 353. 437

Classifications In, 354 Equrpment for, deslgn

considerations for. 35fi First application of. 58

Seam depth in, 354 Binocular microscope, 398 Black Light, defined, 125, 275,290

Eyeball Fluorescence from, 304 Eye Fatigue from, 288 Filter, transmission curve for,

293 Fluorescence and, 274 Indications, brightness of, 299 Intensity required, 283, 284,

299 Mercury arc spectrum. 292 Meters for measnrmg, 301, 474 Photography with, 410 Sources of, 292 Sunlight a s a source of; 291 Yellow filters, to protect eyes,

289. 305 Black light intensity

Achiev~ng adequate, 304 Distribution of, 300 Measurement of, 301, 473, 474 Meters for measuring, 301,

302; 474 Variations in, causes of, 302

Black Light lamps, 295 Causes of output variations of,

302, 303 Commercially available, 297 Distribution of black light from,

xnn ~nc losed mercury arc: 283,

284, 296 Eye fatigue from: 288 Filters for, 285, 292, 296 Flood type, 283, 297 Health hazards of, 288 Incandescent, 296 Intensity of light given by,

284. 300-302 Life of, 285, 302 O ~ e r a t i n c characteristics of.

284, 303 Spectrum of mercury arc of, Spot type, 283, 297 Tubular type, 295

Blemish, defined, 71 Block, test, f o t bath strength,

472, 473 Blooms, testing of, 345 Rlow holes 77 - ~ . B I " ~ I & of i;itigue fractures, 402 Boothi, h1:ick light inspection,

?Rn 2R1 - - -, - - - Boundary zones in welds, 390

SUBJECT INDEX PRINCIPLES OF MAGNETIC PARTICLE TESTING

Brazed joints, non-relevant indications of, 390

Brazing, indications of, 74 Brittle coating stress analysis,

103, 104, 110 Brittle fractures, 98, 101 Buildings, welded, testing of, 446 Burning, of steel: 87 Bursts, forging, 85

Cadmium coating, for contrast, 222 Caiculation of field distribution,

7 79 -." By analog methods, 175 By conformal mapping, 175 By transformation methods, 172

Califorma Div. of Highways. 464 California State Dept. of

Education, 447 Camera setup fo r black light

photograpny, 410 Car parts, racing, 435, 436 Castings, magnetic particle testing

a t By expendable method, 451 Fluorescent narticle bath for.

451 Gray iron; 66, 451 Handling cracks in gray iron, 88 Large, 357-359, 340 Overall method for. 358. 359,

440, 450 Prods, used for, 357, 440, 451 Malleable, 66, 452 Shrinkage In, 88

Castings, Nondestructive Testing as a design tool for, 451

Centrai Conductor, defined, 125 Effect of placement of. 147 Field of a cylinder magnetized

by, with D.C., 182 Chart. Bath standardization guide,

A 7 1

char;, Bath strength, 263 Chart. Fluorescent bath strength,

277 Chipping

To identify defects, 400 For repaxr and salvage, 400

Circuits. Resonant, for demagnetizing, 319

Circular fields. 186 Described. 135 Distortion of. 138 Removal of. 308. 313

Circular magnetization with central conductor, 138, 147 Crack onentation, effect of, 136 Current required for, 183, 198,

208, 241 Defined. 147 Described, 135 Distortion of

due to external Iron, 138 due to shape, 138

Dry method with, 247 With electric current, 147 Field strength for, proper, 163 Fi rs t application of. 49 Of an I-shaped bar, 201, 202 Of irregular shaped parts,

148. 198 Of non-uniform cross sections.

148 Of missile motor cases, 361 Permeability, minimum required,

183 Of a rectangular bar, 199; 200 Of ring-shaped parts, 147 Of a square bar, 199 Of tubes. 147 Wet method with, 267

Cleaning after testing with Dry method, 246 Fluorescent wet method, 289 Wet method, 270

Cleavage fractures, 98, 100 Coating of parts to improve

contrast, 222 Coercive force, defined, 115, 117 Coil magnetization

Minimum permeability for. 183 Ruie fo r ampere turns for, 182,

208 Coils, for demagnetizing, 314,

321, 332 Coils, for magnetization, 145

Diameter, effect of, 145 Fixed, 250 Length, effect of. 146 Split, 249

Coil Shot, defined, 125 Cold Work, non-relevant indication

of, 387. 389 Color Response of human eye, 287 Colors

Choice of, magnetic particles, 346 Of dry powders, 222, 225, 227,

236. 248 Of wet materials, 225, 227, 236

Company specifications, 456. 457 Compass, magnetic, 119, 133, 134 Concentration of \vet bath. 262

Chart for. 263 Measurement of, 262

Conditioners, f o r water-base baths. . ~

228 Conditioning, of steel billets. 353 Conductivity, defined, 120

Skin effect of AC, effect of; On, ~

195 Conductor, field around. 143

Central, field generated by, 125, 147, 182

Field in and around, a h e n carrylng alternating current, General case for. 196 Magnetic; hollow, 194, 195 hiapnetic, solid, 193, 194

Field. in and around, when carrying direct current, 143, 186. 197 Magnetic. general case. 190 hfapnetic, hol!ow, 189 Magnetic, solid. 189 Non-magnetic. hollow, 187. 188 Non-magnetic. solid. 187

Conformal mapptng of fields. 175 Consequent poles. defined, 133 Consolidated Gas Co. Building,

Detroit, Michigan. 446 Constricted rnetai path, non-

reievant indication of. 384 Construction weldments, 446 Contact bu rn~ng , 346 Contact heating, 151, 346 Contacts, head, 128, 149 Contacts, Prod, 128. 203, 207 Continuous method, defined. 266

Advantapes of. 239, 240 Care requlred in use of. 240 Requirements of. 267 Residual method, comparison

with, 297. 240 Vs. Residual, factor in

equipment design; 348 Seiisitivity of, 240 Soft steels, requlred on, 240 Time of contnct for, 267 Wet method with, 2G6

Contrast, of indications with surface, 221. 222 Improved, by coating surfaces

bv cadmium pintinc, 222

Control, current, self-regulating, 328

Cooling cracks. 83, 84 Corrosion f :~ t~pue , 94. 100, 112 Corrosion, stress, 94 cost s anngs , by overall method,

450, 451 Cost of testing, factors affecting,

259. 346 ~ r a c i s

In bar magnet. 134 Cooling. 83 Corrosion fatigue, 94, 112 Corrosion, stress, 94 Deviating from stralght line,

135, 137 Etching. 91. 92 Fatigue, 69, 94, 102, 107, 370 Flash line, in forgings, 87 Grrnding. 90, 91 Handling. In gray iron, 452 Heat-treating, 89 Ingot, 79 Orientation of

Effect in circular magnetization. 1 3 6

Effect in longitudinal magnetization. 134

From over-stressing, 94. 95 Pickling, 91, 92 Plating. 93 propapation, rate of, 112 Quenching, 89 Service. 93, 95 Stra~ghtening, 90 Surface, limits of detectability

of, 369, 370, 372 Threads. in roots of, 385 In weldments, 89

Crxnxshafts. Diesel, 335, 427 Creep, 100. 113 Crystalline fractures, 98, 100, 102 Cupping, in steel bars, 83, 84 Curie Point of steels. 310 Current, alternating (A.C.)

Defined. 120 Cycle, 123 Vs. Direct, for magnetizing. 151 Full wave rectified. 121. 155 Half wave rectified, 122. 154,

207 Meter readings for, 154. 155

hfaanetizing with. 193 Permanent magnetization with.

157


Rectified, 122, 152, 154, 155 Single phase, 122, 154 Skin-effect of, 152, 195, 196 Sources of, 157 Three phase, 123 Three phase rectified, 121, 156

Current, ASTM recommended for Prod magnetization, 205

Current, f o r ClrC~lar magnetization, 183,

208 Choice of, f o r given test, 231,

235 Control, self-regulating, 328 Direction, factor in equipment

deslgn, 348 Current. Direct

vs. A.C. for magnetizing, 151 Defined, 121 Sources of, 152

Plating generators, 153 Rectifiers, 153, 154, 155, 156 Storage batteries, 153, 243 Welding generators, 153

Current. Hard 'steels, effect on

required, 238 Induced, 124. 160 Oscillating, 317 Prod magnetization,

requirements for, 203, 208

D a m adaptation, of the eyes, 285, 287

Dark adaptation, perception chart, 286

Daylight fluorescence, 291 Decay, defined,lZl

of Currents, 121 of Fields, 121 Demagnetization with, 315

Motor generators, 315 Saturable-core reactors, 316 Step-down switches, 315 Variable transformers, 316

Defects . . .. Characteristics of, influencing

choice of techniques, 230 Classified, 75, 367 Deep-lying, 374

Concept of depth of, 375 Definition of terms for, 374 Emergent field at , 376 Least favorable shape, 379 Most favorable shape, 379

Defined, 71, 126, 416 Detectable, 366 Evaluation of, 413 Identification of, 392, 406 Interpretation of; 392 Removal of; for salvage, 400,

An1 so&Kes of; 70 Sub-surface, 368

Characteristics of, 368 Detection of; 373 Importance of, 373 Groups of, 373

Surface, 367 Characteristics of, 367 Detection of, 370 Importance of, 369, 370 Threshold, 370

de Forest,,A. V.; 48, 49; 50, 225 Demagnetization, 115

With A.C. coils, 314, 321, 322 With A.C. loops, 319 With A.C.; through-current, 321 With A.C. yokes, 145 Apparent, 311, 313 Before Magnetic particle testing,

9nc """ Between iongitudinai and

c~reular shots, 308 Choice of methods for, 321, 323 Circular fields, removal of; 313 Degree of, checks for, 320

Field meter for, 320, 321 With direct current. '316, 322;

323 With D.C. single shot, 317 With D.C. yokes, 318 Earth's field, influence of; on,

310, 324 Equipment for, 315-319. 322,

332

SUBJECT INDEX

Limits of; 309, 320, 324 Longitudinal fields, removal of,

?1? --- With magnetizing equipment,

322 Not required, 307, 308 With oscillating current, 317 Oscillations, damped, in. 319 Reasons for, 306, 313 Requirements -.- for accomplishing, 313

Resonant circuits for. 319 Self-demagnetizing effect. 181 Of small parts, 314, 318 Specifications for, 311, 321 Vibration, effect of, on, 324

Dendritic segregation. 390 Density, of magnetic particles, 213 Depth, concept of, for deep-iying

discontinuities, 375 Depth of defect, effect on

detectability, 371, 375 Depth, of seams, in billets, 354 Desien

Of automatic equlpment. 343 Of billet testing equipment, 353 357 . . . , . . .

Factor of safety in, 96: 97 Fo r fatigue, 106, 107 Knowledge of, for evaluation of

defects, 414 Nondestructive testing, tool for.

451 Of special equipment, 341 Stress ralsers, 103, 107

Detachment of Fluorescent Pigment, 468

Detectability, of defects Depth, effect of, on, 371, 375 Factors, miscellaneous, affecting,

380 Height, effect of, on, 377 Least favorable defects for, 379 Length, effect of, on, 377 Limits of, 372 Magnetization, effect of

method of, on, 380 Most favorable defects for, 379 Orientation, effect of, on. 37s.

379 Permeability, effect of, on, 380 Requirements for, 370, 371 Of scratches, shallow, 371 Shape, effect of; on, 378, 379 Of surface cracks, 369

Width, effect of, on, 377 Detectable defects, 366

Deep-lying, 374 Depth, effect of, 371 Laps, detection of, 372 Least favorable, 379 Nost favorable, 379 Non-metallic inclusions. 373, 374 Requirements, for surface. 370 Scratches, shallow, 371 Sub-surface, 373 Surface, 369 Threshold. 370. 373

Dial-Amp current controi, 328 Diamagnetic materials, defined,

, I 0 .a"

Diameter, of coil, effect of, 145 Direct current (D.C.), defined, 121

vs. A.C.. factor in eouinment . . deslgn, 347

vs. A.C.. for magnetizing, 151, 2"' 0"" "no

D ~ I .,I, L U G , GOO

nagntization with, 316, 317, ~~ -~

322, 323 Penetration of field of; 232 From rectified A.C., 156 SOL----- -' "" Frc

field d i r ~ Of field Of field

Effect 1 RR

~ f f e c t of shape on, 184, 185, IaG

Directional properties of a field. 134 Discontinuities

Chnraeter~stics of, 367 Classes of, 75 Classification system for, 75 Defined, 71, 126 Importance of detecting, 373 Inherent. 75-80 Magnetic. 72. 127 Metallic. 72. 127


Surface, 62, 65, 231, 367, 369, '15n ., ,"

Discrtmtnation, of seam depth in steel billets. 354, 355, 356

Disstmil;lr metals, Junction of, 73 Distortion. of fields. 61. 126. 138. . .

185 Distortion, of parts, 351 Distribution of field

Calculation of, 172 In~por tance of knoleinf, 165 In , r rcpu l ;~r shaped oblccts, 198,

--- 111 symn~etrtcttl objects, 178

Doane, F B., 19, 50. 51 Drag-out, loss O F wet bath by,

265, 346 Dry concentrates, fo r fiuorescenl

method. 278 Dry eoncmtr;ttes, for wet method,

263, 265 Dry method

Advnntapcs of, 245 Chotcc of, ss . met method, 235 Circul i~r mngoetization for, 247 Cleaninp ;li ter testing, 246 Comp:lred to \vet method, 235. 977 -., .

Dis;ldvnntnpcs of, 245 E;~rly uses of, 241, 215 I-listoty of, 244 Lonp~lud in ;~ l m:~(rnetiz:~tion for.

2.19 h 1 n ~ e t i z ; t t i o n for. 247 hlntert;*ls for, 244. 245 P t~ t te rns . wi!h d rv notsders. 60. . .

2533 I'roti conl:icts, use tn. 207, 443 Size, of d ry mt~gnet ic particles,

210 Steps III itppiyti~g, 246 Surface, prep;ll.;itiort for. 246 Weld inspection rrith. 250, 254,

4 4 3 Dry powders

A i r pun for, 251 Anplication of, 250 A\,;lilublc, list of, 225-227 Ch:~rac te r~s i i es of. 245 Coerci\.e force of, 216 Colors of , 221. 246 Contl';lst of; 221. 223

Density, effect of, 212 Flo\vtng of, 212 Nysterests curves of, 216-219 nlobility of? 220

Produced by A.C., 220 Permeability of, 215 With resldual method, 239 Size, of particles, effect of, 210 Shape, of p:~rticles, effect of, 213 Snueeze hottie for, 2 5 i Visibility of, 221, 223

Duor.ec", 1 2 i Dyes, Huorcscent, 292

L:ziti em,s.;ion s:,ecfrllm of . 294

Z;,r111..s #:<!l<i. IGl l t~ f l~ tence 0 8 ) demagnetization

limits, 310, 321 h1;ip~retiztltion by, 310

E;rrthrs mttgnetic poles. 141 Edrly current methods, 54

Conipikrtson with magnetic pttrtrcle testing, 69

F o r \reid testinp. 443 E f f c ~ t i \ ~ c ~~ern?exbi i i ty . 180

E;!u;~tion fn r detrrmlninp, 181 Electric rests t i~nce \\.elded

steel ntne, 447, 418, 449 Elcct rom:i(rtlet, defitted, 124 Emcrpent lenl;;~ge Held, spread of,

:z71i ~. . . Emrss~on spectrum, of fluorescent

cl\'c. 294 Endurttnce limil, of steel, 106 Et ru~t~rnen t , fo r m; l~ne t ic t);~rticIe

testing .4lter11;1tiitp current. 328-331 Automsltic, 56. 243. 333, 342 Blacl; light, 28:3 Chn,?e of. 242 Coil.;, fixed, 550 Coils. s i~l i t . 249 F o r dem;~met iz ; t t io~~ , 315, 332 Direct current, 331 F o r the dry melhod. 247 E;~rl ies t , 54. 325 For Huorescent mrthods. 280-282.

330 He;iyy rlutv, D.C.. 331. 332 Elislor?. o f devcIopn>ent of, 325 I<its; fo r mtirnetiztnrr. 327 I.ecclit~s. 248 Pu'<.riI for. 325 Portt~blc. :327

I.ar(re. 328

SUBJECT INDEX

Prod contsicts, 248. 249 Purposes served by, 326 With sclf-regulating current

control. 328 Simole. 326 Specl;il, 338 Snecific;~trons for, 461 St;ttion;~ry, 329 Storage bat tery, 54, 55 \ ' ;~ r~a t ions of, 332 F o r the \vet method, 268, 330 Yokes. 326

R,li~rnmei,t ~. LJ,.s,gtl "I slll!ct~ll, :314 1:uturt. trrltds in, 365 General Purpose, defined. 334.

:1R!l Rl t~ i func t ion ln~ of, 466 Opei~ltirtg tnstructioils for. 461 Sinxie-purpose, defined. :334 Specl;il, defitlcd. :3:<:1

Fixture*. 339 S t rma st~:~li .sts, 110

Errors. m mairtt~g wet bslih, 467

Cr:~cks. 91. 405 Dcel,, zlcid. 404 F o r How lines, 404 For pr;nn s t ructure , 404

Ev;llu:~tion of defects, 413 Defined. 392, 413 Gel~eral rules for. 416 Problem of. 413

E\.;iport~tion, loss of wet bath by, 965 ---

Expend;lble bath technioues. 276. 3.16. 3358. 451

Ext,rr~mcnt;tl s t ress ant~lysts. 103. 104. 110. 414

Estern:~l fields. ?>on-relevant trtdic;ttions of, 38:3, 387

E s t ~ r n i ~ l ~ O I P S , non-relrr:~nt tntlicnttons of. 382

Eyes. hunl;ln, Arc H;~shes, d;~m:tpe of, by, 153 !'oior response char t of. 287 D. .tik .. .~rlep!:~tiot~ . of, 285 F ;~ t ipue of. 288, 305 Fluorescence of. 288, 304 I'erccptio~t ch:~rt of. 287 Ultr;~rlolet lipht; dam;~ee of, by.

291

F;ictor of safetr . 96. 97, 415 - ~~

Failure of metals, 96 Conditions leading to, 100 By corrosion fatigue, 94. 100,

t l ? --- By creep. 100. 113 By fattgue. 47, 100; 102 By tmimct. 100 Modes of, 98-102 By over-stresstt~g. 94, 100

False ttldic:~tions, defined, 126, 382 F:ist bretlk, defined, 128, 159, 160

Vxlue of, i6O F;itisue, defined. 102

Co~.ros~on. 94, 112 Cr;~cks. 69, 94. 102, 107, 370 Endurtlncc Limit. 106 Failures. 47. 100, 102 Of the huntill1 eye. 288. 304 Life. 106 Ltmit, 106 Of met;tis, 105 Prop;tgtltion, of cracks, 113 Rxte of cr;tcL prop;tg;ltion, 112 Ratio, 106 St!I\~tt~e of cracked parts. 112 Stretlgth, 106 Test inr . 105. 106 -. In torslan. 109

Federal Aeronautics Admlntstra- tl0ll Atrcraft repair s ta t ton

requtrements. 461 Ferromagnetic rn;~terrals, defined,

115 Ferromagnetic p:~rticles, 63. 250, 3K7 (See also miignetic particles)

Field, m;>gnetic, defined, 118 Around ;I bar mitarlet, 131 Calcult~tion of, 172. I 7 3

Bv analog methods, I 7 5 By conformal mapplng. 175 B r transforntaiion methods,

173 Circular, 186 Conductor, ru :tttd around. 143,

186 Conductor, tnside. 186 Corlductor, outside, 186 Direct iot~ of. effect of shape on,

184 Direction111 properties of, 143 Distorted, 126, 138, 185

PRIKCIPLES OF MAGNETIC PARTICLE TESTING

Distribution of, 165-175, 178, 198 I n a bar, I-shaped, 201, 202 I n a bar, rectangular, 199, 200 In a bar, square, 199

Ear th , around the. 141 E c c e n t r ~ c tubes, inside, 174 I n i r regular shaped parts,

198-202 I n non-uniform cross sections,

Pnn -"" In symmetrical oblects, 178 Of the earth. 141, 310, 324 Electromagnetic, 178 Emergent, spread of, 376 External, non-relevant

indication of, 383. 387 Gauges, 167 I n a loop, 144 llieasurement of. 165 Meters, 166, 167 Penetration, of D.C.. 234 In plate, magnetized with prods,

203 Plotting, 175 In a solenoid, 149 Strength of. 165-178

Measurement of, 166, 167 In prod magnetization. 203,

207 Rule of thumb for

circular magnetization, 163 longitudinal magnetization,

163 Suitable, for m;ignetic

particle testing, 163 Fields

Concept of flow of; 134 Distortion of, 61, 138 Leakage, 62, 64, 126 Miirnetic, defined, 118 Parallei, 139 Residual. 120 Resultant or Vector, defined, 128

Filing, Suppiemental Test, 398 Fill-factor, in coil magnetization.

~ i i ~ e ; welds, 445 Fillets, non-relevant indications of,

!<US "-" Filters, for blacli l irht. 285

F o r fluorescent l i rht , 305 Triinsmission curve of. 293

Final ~nspection, 419 F i n i s h ~ n f defects, 89-93

F i r e Hazard. in wet method, 224. 255. 345. 467

Fissuies, internai, in ingots, 78 Fits; forced, non-relevant indication

of. 391 F ix ing of an indication, 407 Flakes in steel, 85 Flame gouging, to remove defects,

nnl ="A

Flash-line tears, in forgings, 87 Flash magnetization, 163 Flaw, defined, 7 1 Flow lines in forgings, 390, 404 Fluorescence, defined, 126, 290

Daplight. 291 Principle of, 2-74

Fluorescent dyes, 292 L u h t emission curves of, 293,

294 Fluorescent magnetic

particles, 56, 64, 212, 274 F o r billet testing, 354 Character of, 278 Contrast of, 222, 223 Density of, 212 Dry powder concentrate of, 227;

2'ifi pas te form of, 227 Permeability of, 216, 219 Separation of pigment, 277, 468 Shape of. 213 Size of, 212 Visibility of. 222, 223, 237 VS. Visible magnetic particles,

? A 6 "." Fluorescent particle method. 274

Advantages of, 275 Bath s t renr th for, 277 Booths, curtained, inspection,

?Rfi "u"

Darli adaptation, of the eyes, 285

Disadvantages of, 276 Esebal! fluorescence. 304 Indications, examination for,

279. 283 Insn~c t ion area requirements,

279. 281 Inspection in the open. 281 Materials for, 2iG Pre-nilxed bath for, 289 Post-insi~ection cic~lning; 289 Steps in the testinn process, 279

Fluorescent n;trticle wet bath Bath r t r m p t h chart for. 277

SUBJECT INDEX

Deterioration of. 278, 279 Fluorescence of oil for. 276 Maintenance of, 278 hlixing of, 278 Prepared o r premixed, 289 Skin protection from, 289 Strength of, 277

Fluorescent penetrants, 56, 59, 68 Flux Density, defined, 115 Flux, magnetic, defined, 118 Flux meters, 166 Flux paths, 62

I n a bar magnet, 222 Direction, importance of, 134 F o r a iong bar in a short coil, 146

Flux-shunting devices f o r field strength and direction, 170

Foaming of water base baths, 346 Farced fits, non-relevant

indications of, 391 Forgings

Bursts in. 85 Fat igue cracks in. 453 Flash line tears in, 87 Flow lines in. 390 Grinding of defects, 399 Laps in, 86 Repair and salvage of, 400 Testing of large, 440. 453

Fracture- Brittle, 101 Cleavage mode, 98, 100 Crystalline, 100; 102 Shear mode, 98. 100 Silky. 100

Fracturing, f o r identification of defects, 401

Frequency. of A.C.. 123 Effect on demagnetization.

313, 314, 317 And skin effect, 152, 195

Full-wave rectified A.C.. 121, 155 Function, of magnetic particle

method, 366 Functions, transformation. 173

G a p , air, defined, 125 Gauges. field. 169 Gauss, defined, 115. 119 General Electric Co.. 255 G.E. light meter. 302, 474 General purpose equipment, 334,

339, 340 F o r testing automotive steering

parts. 342

F o r testing castings, iarge, 358. 859 . . . , . . .

Compromises in design of; 341 Reasons f o r using, 341

Government requirements, 56, 57, 460, 461

Government specifications, list of. 4fi? ---

Gram boundaries, non-relevant ~ndicat ions of, 390

Gram structure, etching for, C04 Gray iron castings; 66. 4 5 1

Handling cracks in; 452 G r ~ n d i n g , to ~nves t iga te

indications, 399 F o r repair and saivage, 400

G r ~ n d i n g cracks, 90, 9 1 Growler, demagnetizing with, 318

H a l f wave rectified A.C.. 122, 154 Applications of; 155 Meter readings for , 155 Prods, used with, 207

Hall effect, defined, 115, 168 Hall probe, f o r measuring field

s t rength, 203 Hand creams, use of, 271, 289 Hand grinder? 400 Hardness, in steel, effect on

detection of defects, 238, 239 Head contacts, 128, 149 Head shot, defined, 126 Heads, .I26 Heat-treating cracks, 89 Heating, demagnetizing by, 309,

310 Height of defect, effect on

detectability. 377 History of

Dry method, 244 Equipment f o r magnetic

particle testing, 325 Fluorescent magnetic particles,

274

and around, Magnetic. carrying A.C.. 194,

196 - * -

Magnetic, carrying D.C., 189 h'on-magnetic, carrying D.C.,

187, 188

l a f PRINCIPLES OF 8fAGNETIC PARTICLE TESTING

SUBJECT INDEX 9 proper field strength for, 163 In the wet method, 268

Loop, defined, 125, 144 A. C., demagnetization with, 319 Field in and around, 144

L over D ratio, defined, 181; 208 Low temperature, effect on

s t rength of steel, 101 Luders lines, non-relevant

indications of, 389 Luminescence, defined, 290

M a c h i n i n g tears, 89 Macro-etching, 404 Magnagloo, 274 Magnet, defined, 114, 118

bar, 131 cracked bar , 134 electro-, 124 horseshoe, 115 permanent, 119 temporary, 120

Magnetic attraction, 133 discontinuities, 72; 127 field, defined, 118 field meters, 167 Hux, defined, 118 induction, defined, 117

Magnetic Particles application of, dry, 251 choice of; for a given test,

231, 235; 346 coerclve force of, 215, 216 color of, 222, 223, 227, 236, 248 density, effect of, 212 descrlntion of. 209 dispe<sion of, wet, 214 dry, 210 Huorescent, 56, 274 fluorescent vs. visible, 346 f o r the fluorescent wet method,

276 hys te res~s curves of, 216.219 improvement of, 58 liquids, suspending, for, 220,

221, 237, 260 mobility of, 220 Permeability of, 214, 215 shape, effect of, 213 slze, effect of, 210 wet bath materials, 211. 212, 260 for the wet method, 260, 263

Magnetic particle method advantages of. 66

alrcraf t overhaul, use for, 428 461

automotive applications, 430 comparison, with other NDT

methods, 68 early development of. 48 design, a tool for, 415 factors affecting patterns in,

60, 61, 65 function of, 366 fundamental concepts of, 60 fu ture of, 59, 365 eor'ernment requirements, 56,

57, 462, 463 hlstory of, 47 industrial applications of, 419 limitations of, 67 performance of, requirements

for successful. 66 railroads, use by, 335 specifications for, 455 s tandards for, 455 steps In applyrng, 61 structural welds, testing with,

446, 447 t rends of; future, 365 w a r materials, use on, 55

Magnetic poles, defined, 119, 132 Magnetic reluctance, defined, 126 Magnetism, defined, 114

remanent, 120 residual, 120

Magnetization. 61 A.C.; used for. 54, 151, 193 circuiar. defined. 125

current r e q u ~ r e d for , 198, 241 for dry method, 247 first use of, 49, 186, 240 for wet method, 267, 268

wiih coils, 145, 181, 182 by continous methods, 267 current, effect of types of, on, 151 f o r the d r y method. 247 by the earth's field. 141, 310 with electric currents, 143, 178,

207. 208. 267 Hash. ' 163 indicator, Japanese, 171 by induced current method. 160,

352. 363 longitudin&, 126, 184, 240

for dry method. 247 for wet method, 268

of mis.;ile motor cases, 361 multl-direct~onal. 127, 207

with prods, disadvantages of; 151 rules f o r ampere turns, for, 182 sequence of, 308, 311, 313, 356 solenoids for , 145. 181. 182 of steel billets, 353 by s t ray fields. 311 by the surge method, 155 with translent currents, 159 f o r the wet method, 267, 268 with yolces, 145

hfagnet iz~ng ampere, half uzave, 155

~ G n e t i z l n g force, defined, 118 a t peak of A.C. cycle. 232

Magnetizing techniques, f o r large object, 442

hfagnetographs, defined, 119 of field around a bar magnet,

132 of field around c~rcu la r ly

magnetized tube. 148 of field between prod contacts,

383 a s means for indicating field

distribution, 169. 170 Maintenance inspection, 419, 432,

49% --- Malfunction~ng of equipment, 466 Malleable iron castings. 66. 452 Manuals f o r t ra inlng operators,

A5X .-- Manufacturers' standards, 464 Mapping of fields. 175 Nartna City Towers, Chrcago,

. . -. - - -. . , . Marlne shaft , fa t igue in, 453 Material permeability, 179 Mathematicai methods f o r field

strength and distribution, 172 Maximum materral permeability,

179 Measurement. of fields. I f i5 -.. Mercury vapor lamp, 283, 296

construction of, 296, 297 spectrum of output of, 292, 297

Metallic discontinuities, 72. 127 Metallography, 97, 406 Metallurgical laboratory, 395. 406 Metallureist, 395, 396

Metallurgy, 97 Metals. failure of. 96, 98

s t rength of, 98 Mcters,

alternating current, 159 Black Liaht , 302 fiux, 166 hsllf Itrave current, 155

Method factors in des~gn . 345 M ~ t h n d c

, . .. c o i h r i s o n of wet and dry. 235 continuous vs. residual, 237 drv. 244 eakiy testing, 47 expendable bath, 276, 346 fluorescent, 274 induced current; 160, 352 lacquer, 273 matbematicai, f o r field s t rength

and distribution, 172 oil and whiting. 48 overall, A qn 58. 149, 207, 243, 358,

="" penetrant, 56. 59 residual vs. continuous, 237 transformation f o r field

s t rength calculation; 172 wet, 255, 345

Microscope, binocular, 398 Microscop~c examination, 406 M i n ~ m u m permeability required

for circular magnetization, 183 for 1 R? longitudinal magnetization,

A--

for magnetic particle testing, 1 R7

Miss?Kmotor cases, 359, 433 domes, testing of, 434 magnetization of, 361

circular. 361 longitudinal, 361-362

port opentngs, testing of: 363, 364

t e s t i o n s i d e r a t i o for , 360 Mobility

of dry particles, 220 produced by A.C., 220 of wet particles. 220

Modes of failure of metals, 98, 100, 102

hfoore, R. R., fatigue testing machine. 105. 106


n~otor-generators f o r demagnetization, 316

i\fulti-directional magnetization defined, 127; 207 use of; 243, 450

Multiple test systems, 364

Near-ultfiivio~et l ight , 274 Nickel, magnetic p;Irtlcle testing

of. 183 Nondestructive testing

::7 ; ~leslgn tool, 451 .,. -.... !-, 457 ll<.t',; . . . .....! .. . 1 ; Society far. ::,a

Nonmetallic lnciuitons. 77. 373 s tr ingers of. 2:15. 374

Nonrelevant trtdications. 382 of boundary zones in welds, 390 of brazed joints. 390 of cold work. 387-389

ho\v to recognrze, 385 :it constrtctions in metal path. .. .

384 holy to recopnlze. 385

defined. 127, 382 of dendritic segl-egation. 390 of external fields. 383, 387 of external poles, 382 tit lillets, sharp, 385 of flow lines in forgings, 390 of forced fits, 391 of pratn boundaries, 390 of joint behveen dissimilar

magnetic materials, 391 of luders lines. 389 of magnetic writing, 387

hoiv to recogntze, 388 of root openings, design,

In fillet welds. 445 of scale natches, edges of. 383 a t thread roots. 386

how to recoanize. 386 'orth Pole, mapnetic.

of a c ~ i c h e d bar magnet, 134 of the earth. 141 of a magnet, defined. 119

Notch sensitivity, 101

O e r s t e d , defined. 118. 119 P e r inch of nrod spactng.

205, 206 Ohm, defined, 122 Oil a s suspending liquid for

wet bath, 220, 257-259

Adrlition of r u s t preventive to, 758

F i F i h a z a r d in use of, 224, 7 5 5 346 ---, - ~.

Odor of; 259 Specifications for; 258, 469 vs. water for wet bath, 224, 345 Viscosity, importance of, 258

Oil. :iccurnui;itioti in bath, 266 Oil arid whiting methoij. 48 O p e ~ i t i n p instructions, f o r

esutpment. 461 Oli..~!i"l.

T ,,, .. ,.,,. ! ,V.."P n+ ';<I ict.~;i:irrz?len!s lei., X!1.5. :<!I#; Sources o i knowledge for. 394,

39ii als for. 458

~f cracks ametiznt ion.

. . . . . . . . . . . . Circular m ,, ~~ ~ ~ ,

effect in, 136 Detect;ibilit?, effect on. 378 1.nneitodinal mzmetiz:ition.

effert in. 134 - ~ ~ - . ~ 0rtent;ition bf sub-surf;~ce

discotltinuitics. Effect on cietectabilit?, 378

Oscillatinp current, for dem;ipnetizlnp, 317

Oscillations, dnniped, f o r demagnetiztng, 319

Ouelxll m s m e t i z n t ~ o n . 58. 149. 207, 245, 358, 450 Cost s a v ~ n g s of, 450

Oser-fills. tn rolls L;ips from. 81 Seams from. 81

Overhaul inspections ;\ircr;ift, 428, 429 Aulornotitre. 430 Plant equipment. 431. 432 R:riIroad. 427 Sti.an~ turbtnes. 431. 432

Over tns~ec t ion . 416 Over-Stressitip, cr;icIis from, 94.

100 F:zilure due to. 414

Piiltit, renlo\-:iI requtred. 247. 266 Pn~.:illrl fields. 139 Par;imaancttc m;iter~als. defined.

119 Par t , defined. 127 Parts . ch;tr;icterrstics of.

~nf luenc>na choice of techn~que, 230

SUBJECT INDEX

P;~ste , form of \vet nlaterlals, 224, 255

Pe;ik. mspnetiztnp effect. of A.C. cycle, 232

Penetrant methods, 56, .59 Companson with magnetic

p;irtlcles, 68 I>enctration of D.C. fields. 234 Percention chart of the human eye. - 286' Permanent mapnet, defined, 119

B1apnetiz;ition with, 141. 142 Permanent mapnet yokes, 142 prrninr!c!it mngnr.ilz:~liu!l with

.LC., 157 Permeability, defined. 119. I79

Effectire, 180 Eou;ition for. 181 1niti;il. 180 Of m;ipnetlc materials. I79 Of mzignetic t>z~rticles, 214, 215 R1;iterlal. 179 ~ I a s i m u r n m;ttertal. 179 BIinimum requtred, 182

for circular magnetization, 183 for ionpitudinal magnetization,

I83 Sliin effect of A.C.. effect on. 195 V:tri:ttions of. 138

Phosphorescence, rlefinerl. 190 Photo-el;istic s t ress nn+lysis. 110 Photograph?, to record

tndications. 409 Bl;iclr lipht. 409

C a m r r ; ~ set-up for, 410 Color, 412 Pol;lroid canier;l, use of. 409

Piclilinp, of billets. 353 Picklinc cr:icks, 91 Pipe, In inzots. 7G Pine. !rclderl. 436 Pitlsburgn Testinp Laboratory, 50 PI:ittt ~quinn len t m:llntenance, . .

431. 4:G! Platinp. C;tdmlum. for contrast.

222 Platina cracks, 9:< Pli~t inp. r e m o ~ a i required. 247. 266 Plottina. of field distribution, 175 Pol:irls nltssile motor case.

testina of, RG1-364. Por t areas of. testing, 363

Pole.-, RI;ipnetic, drfined. 119 of s h;tr rnzipnet. 132 Consetiuent. 13.3

Of the ear th, 141 North and South. defined. 119

1'ort;lble equipment, 327. 329 for \veld inspectton, 447

Pre-mixed wet bath. 224, 272 Pressure vessels, testlnp of, 418

449 Pressurized cans, for wet bath,

272, 289 Problem of interpretation. 393

Of evaluation; 413 Procedul.al gutdes for magnetic

Particle trstinp. 457 I3ro;id. 457 Company. 457

Process specifications. 417. 418 Suppested co\'erage In, 458

Proce.sslng defects Fints l l1na,~75. 89. 93 Prtmary, 15, 80. 89 Second:ir\.. 89

Prod Conl;iets, defined. 128 D r ? Por\.der used with, 207 Field distribution xvlth, 203 Half-\v;iue current used with,

207 L;irpe c;istinps. f o r testing. 357 Mapnetizatiotl with, 149. 203.

m7 - Spac t i~p of, 203. 207

Amneres per tnch of, 205. 206 ASTRI Recommended. 205

Oersteds, per Inch of. 205. 206 Structur.al welds, testing with,

447 W e l d m ~ n t s . testtrig \rlth. 443

Product specificattons, 458 Production of indications. 392 Propagation of fatigue cracks.

112. 113

Quenc~linr Qutci; Brr;lk. cr:iclis. defined, 89 128, 159, 160

R;ices. Be;irtnp, testing of. 350. 353 D:im;ige of. durlng testing, 351

Raclnc car narts. 435, 436 R:~rliopriipn.v

Coml,:irison lvitll rntlanetic p;irticle testitig, 68

~ ~ i d ~ ~ s t r l i ~ l use. earls. 48 ~ & ( e r . H. I I . , 48 Weld Testinp with, 443. 447

~ ; ~ i l ~ ~ s d s . Usr ot magnetic particle testing bv. 335, 427


Reactance, Inductive, defined, 122 Reactors, Saturable core. 316 Records, 407

Fixlng Indications with lacquer, A f 1 7

~ i i i i n g indications wtth scotch tape. 407 rpith strlpable lacquer, 408.

Color. 412 Directions for , 410, 411 Polaroid camera for. 409 VBlue of, 407

Recianrular bar, Field distribution in. 199, 200

Rectified A.C., defined, 122 Full wave, 155 Half Wave, 122, 154 Three Phase, 156

Rejection Standards o r limits, 460. 461

Reluctance, Magnetic. defined. 120 Remanent magnetism, defined. 120 Repair station requirements,

arrcraft. 461 Residual field, defined, 117, 120 Resldual magnetism, defined, 120 Residual method,

Advantages of. 237, 239 Application procedures for, 270

Curtain flow of bath, 270 Immersion, 270

Choice of, vs. Continuous, 237 Vs. continuous, Factor in

equlpment design, 348 Defined, 269 Hardness of part, effect of. 238 Wet method, used in, 269

Residual s t resses , a s cause of, Corrosion, 94 Etching cracks. 91 Failures of par ts , 414 Grinding cracks. 91 Plat ing cmcks, 9 3

Resistance, electrical, defined, 122 I Jn i t of , ohm defined, 122

Resistance-welded steel pzpe, 446, 447, 448

Resonant circuits, f o r demalmetization. 319

Resultant fields, defined, 128

Results, recording of (See Records), 407

Retentivity, defined, 120 Effect of, in dry method. 254

Reversal of field, frequency of. fo r demagnetizing. 313, 314. 317

Reversing D.C.. for demagnetizing, 316, 322

Right hand rule, 143 Ringing methods, 48 Ripple, in rectified AC, 155 Rolling, overfills and underfills, 82 Root openlng. i n fillet welds, 445 Roots of threads, non-relevant

indications of, 385 Cracks in, 385

Rotor bars, aiuminum, 438, 439 Rule

fo r determlnlng amperes f o r clrcular magnebzing, 198, 241

for determining ampere tu rns fo r longitudinal magnetizing

182, 198, 241 Rules, general, for evaluation of

defects, 416 Rules, general guides to

magnetizlng, 416 Rust prevention, w ~ t h wet method,

271

S a f e t y factor, 97, 415 Saturable core reactors, f o r

demagnetrzlng. 316 Saturation polnt, defined. 11;. 120 Sauveur, Dr. Albert, 98 Scabs, on ingots. 79 Scale patches, non-relevant

~ndicat ions of, 383 Scarfing, of steel billets. 353 Scotch tape, for l i f t ~ n g indications,

407 Seams

in bars, rolled. 80, 81 in billets. 353 Depth of, in billets, 354, 355 F r o m ingot cracks, 79 From over-fills in rolls. 82 In sprrng wire, 109 From under-fills in rolls 82 ~~~ -

Seam depth discrimination' -- in billet testing, 354; 355, 356

Seamless steel tubing. 426 Secondary processing defects, 89

93 S e c t i o n l n ~ , through a defect. 403

SITRJECT INDEX

Segregation, indications of, 74. 390 Dendritic. 390 Related tb cupping, 83 In steel, 77

Self-demagnetizing effect, 181 Self-regulating current control,

328. 447 Sensitivity

Control, in billet testing, 356 Comparison of, f o r A.C.. D.C.,

and half wave on subsurface defects, 234

Comparison of, for wet and dry methods, 239

Of fluorescent wet method. 276 Highest, f o r fine defects, 240 Of the prod method. 150

Service cmcks, 93, 95 Service failures, 98 Settling of wet mapnetic part., 211,

212. 221 Settling test f o r fluorescent bath,

278, 469. 470 Settling test f o r wet bath,

262, 469, 470 Shaft, marine; fat igue cracks in.

453 Shape of defects, effect on

detectability. 378 Shape of magnetic particles, 213 Shape of parts, effect on field

distribution, 184. 185 Shear fr;icture, 98 Shot, coil. defined. 125, 267 Shot, duration of. 267 Shot. head. defined. 126, 267

S h u n t i n r devices. flux. 170 Sinple phase A.C.: 122,154

Full wave rectified. 121, 155 Half urave rectified. 122, 154 Rectified. 122. 154

Sinrle-purpose equlpment. 333, 120 --"

F o r testing bearing balls, 319. :251 ..., &

F o r testirlp be;lrlng rollers, 314 For testing couplings,

oil field prpe, 337 F o r testinp cmnksh;~fts , 340 Compromises, in design of, 341 F o r t rs t inp electrlc apelded steel

plpr, 446, 447, 448 F o r testink mlssile motor cases.

3.59. 364

F o r testing propeller blades. steel. 335. 336

Reasons for using, 3.11 Single shot, D.C. demagnetizing,

217 .,. >

Size, of magnetlc particles. 210 Of dry powders, 210 Of fluorescent particles. 212 Of wet method parllcles, 211

Skill, requlred of inspector, 253 Skln effect, of A.C., 152, 159, 195,

*""

Skin effect, o i half wave, 155 Skin protection in wet method, 271

In fluorescent method, 289 Small parts, demagnetization of,

91 A

Society o Automotive E n ~ i n e e r u , A57

Specifications of, 463 Soclety f o r Nondestructive Testing,

A57

Specifications of, 463 Solenold, defined, 125

Diameter, effect of. 145 Field in and around, 145 BI;ignetization with, 145

Solid mapnetlc conductors, Field I n and around D.C., 187

Solid non-magnetic conductors, Field. in and around

A.C.. 193, 194. 19G D.C., 189, 190

Sour-ces of defects. 70. 95 Sources of knowledge, for

inspectors, 394, 396 South Pole, m;lgnetlc

Of the B i r t h , 141 Of ;l cracked m;lxoet. 134 Of a mapnet, 119

Space-ape rrqulrements. 59 Spacing o i prod contacts, 203. 207

ASTM, recommended, 255 Snec1;11 ~ouipment . :$:I:$

An;~lysls of testrng nroblem, 344 f'omi~romisrs III desikn of , 348 Ikfiried, :$:$:< F;ictors in dcelgn of

Ai,ply~ns current, method of, : < A X ..

Aplllyinp m;tpnctic i l~~rt ic les . method of, :145

Current. type of. 947 Redisu;ll vs. cantinous method,

348

SLIRJECT INDEX PRINCIPLES O F ZlAGh'ETIC F:iRTICLE TESTING

t c s t ~ n g , 358, I44 Standard enulnmr.nt - ~ . ~ ~ ~ . ~ -~~~

~ e e d for , 338 Defined, 333 Single purpose speclal, 33.1, 339 Speclai fixture, 339 Special purpose, defined, 334 Standards fo r magnetic particle Specification fo r design of, 348 testing, 455 S teps le;iding to desrgn of, 344 fo r accentance and rejection. 460

Specla1 purpose euuinment, 333 Statlonary m:ignetizrng equipment, F o r testing benrlng races, 350. 329

353 Steam turbrrtes, 4:31, 432 For testing car axles, railroad, Stccl billets, tesrlng of. 3-15. 353.

:3:4li -* <> 8

F"&. !esl inr <,?,?r?k:.c!t,?f~.?, 3211;. $37, Y4il

F o r testing propellei- blades, 335. 336

For t e s t ~ n g steei billets, 353-357 Steps lez~dinp to deslpri of, 3-14

Specificailon, design, 348 Specific a t ' tons

Demagrtetiz~ng, 311. 321 Equlunlent, 461 Government, list of, 462 Industrv. 456

Sicp-down sivltches. Z15 Stol;es in\%,, 201 Storage b;ktteries. u s e of, 153, 243 Straightening cracks, 90 Str;tirb gauges. 110 Strength of fields

And distributiorr $11 Irregular shiiued ohjects, 198, 202

And d i s t r ibu t~on in symmetrical OkJJeClS, I78

Measurement of, 16.5 With flux-shuntina devrces.

l n t e r s t i t e commerce comrnrssron. IT0 fo r testrng g a s tank trucks, 464 In plate, magnetized with prod.;.

Magnetic p;irtlcle testing, 455 203. 207 Adv;inlages derived from, 456 S l rcns th of m;tter~:lis, 415 Less desirable results from. Stress ana lvs~s . 98. 103. 110. 414.

4.56 T V W S of, 456, 457

fill;i~nten;lnce, 459 Rl;~nuf;icturer's S1andardiz;ltion

Socret?, Tor steel valve castings, 4F.l

Ovrrhnul, 459 Process, 417; 458 For petroleum-base liqurds fo r

\\,el bath, 469 Qil;tIific;~tinn o f operators, 4li0 Techn~c:il societies, list of, 463 Welded s t ructu~.es , for testina.

46.1 Spectrum

Of emrssion of fluorescent dye. oq'l -. .

O f encloseti mercury arc, 292. ?!I::

~ i i ' h i t v light. 291 Speed of ivsting, 313 Split coils, 249 Spray cans f o r \vet bath, 224, 272.

289

Stress corrceitil.:it~oi~, 103 Stress-corrosion cracking, 94, 112.

370 R;i!c o f i)rop>,:lrr;ltion of, 113

S t 1 . c ~ ~ rii.;lrihuiion, 97. 413 SLress-i.:i~sers, lil:3s 107, 414. .lit> Stress-reiievrng, for cold worli, 336

Refore e t c i i ~ n r , 'In5 S t ~ . i n r e r s of ~ ro~~-n~c t ; i i l i c

inclusions, 77, 235 Strrpable lacuuer, 408 Structorztl <velds. tcsirng of. 446 Slutsm;~ri 0 . E.. 255 Suh-surface discontinuities, G l . 65

Loc;~trct \viln proti magnetlz:~tion. 203

Suni isht , :is rool~cc o f bi;lck light. 29.1

S t i ~ ~ ~ ~ l ~ ~ n ~ ~ ~ ~ ~ l ~ t l iwts . :b~<l in inlrr l~ret ;~t ioi i , 397 Rillor-iii:ti. nricroscor,e, use of'. 398

Blueing of fatigue m c k , 402 Penetrant , 56, 59. 68 Cil~!>prng, 400 Supplemental, for defect Cut surf;lce, ex:inirrlatlon of, 403 ldentifieation. 3 9 i

by deer> etching, 404 Teslinrr of, by light eichrng, 404 Ail.cr-nit parls, 429 by hlncro-etcl~~iig. 2104 ~\ iuni lnum roicr b:irs, 136, 459

\Vith aninionlunt Auiomotrse s tce r ins knucl;les. nrrsulnhnte, 40.1 420

~l icroscoplc examination, 406 C:lstlnrs. Small. 423 Filing, 39P Converai \i.lieels, '125 Flame gouging, 401 Diesel er;inksl~aft. ,127 F ~ i c t u r i n p : 401 Forgings, small, 423 Gri: ,dii~g, ,399 .GP:L~s, 422. 424 S t ~ e ! s o ! ~ ~ n ~ J e t ensine nnrts. .l!?il: 129

Rr : ihmsire heel, 40?, blissile coniponents. 433-435 \tJitR core drill (11-eD;innlng). Plant equiument, 431. 432

4 0:i Racing c a r parts. 435 Bv s:~\oing, 403 Raiv materials, 421

Stress-relievrl~p before etchrnf, Seainless tubing. 426

Surface discoirti~iuitres, 62, 65 Surface i,rcn:~i':ttton

Pol. ti,? ow method, 246, 2.17 For the wet method. 266

S u r s e method of magnci iz~ng. 156 Suspending li<iulds fo r \vet baths.

220 224 . Suecific;lliol~ lor b:lth oil, 469

Sxvilclies, sten-do\r~ti. 315 S t . 1 I 273

S!cani turbine ~comoonai ts . 431. 432

Steel billets S: blooms, 4 3 i Tools, 420 Truck comnonerits. 130 Underseas telephunr cable, ,137.

A?il

welded plpe, 436 LVtre ends, 421. ,113

Tests, fo r corltroi of eqtirpn~ent, .ti;& . . ..,.,

~ ~ s i i i i l . : R. c., 274 Tests, fo r control of nrocesses. 466 Systems, multiple test, 364 Tests, fo r petroleum base b;ith

T ~ I I I < S , steei, testing \r-e~ds of , 450 T e e r s tl:ish line, 8 i

I-lot, I I I castings, 88 hl~tvlbrrti~rg. 69

Tect1111c;tl uz.cle1.s. U.S. Air Force. 458

Trchriirjtivs. v;rr!;ltions in, 229 'l'elrnhoitr c;thle, unrlerse:ts. ,1:!7.

'$:is Tempcl.:lturi. effrct on ?tee1 fai1ul.e.

ill1 . \ iempal~:li.o mngllci, rlefinrd, 120 Tust bioel;. for bath s t rength.

d i ? . 473 Test. rettliils. fo r \vet b;ith. 262,

d C 9

! Test ins , j Creep. 100. l1;l

~ s l i 105. 106

l l o u ~ d s , 469 Tests, supplrment:ii, aid rn

Interpretation, 397 (See "Supplement;ll tests")

Threads, non-relevant iiidie;ttions

Three ubnsr AC., defined, 123 Rectified, 121, l5C

Threshold currents fo r A . C.. D.C.. ;tnrl h;ilf \rs;ive for soli,-surfact,

rlr~iects. 2?,4 Tliresltul~l of detectability of

drfecls, 372 Tit:t~ii I I I m~ss i l e motor ciise. 360 Tomron, fatigue in, 101 Tr:tnsfo~rn:ition methotls, fo r

r;tlcul:~tirw field distribution, li.3 Tr:titsi'ormers, vnrl;lbie, 316 T~.;ritsirrtt currents , defined, 12;

M : ~ ~ r ~ e r r ~ : i t i o n \\,rth. 159 Trr.ix of sro\r.th of N.D.T., 5: Tre/~:ittiliiii', fo r cross-sectinn~ng,

411:i

PRINCIPLES O F lWAGNETlC PARTICLE TESTING

Truck components, testing of, 430 Testing of, 433, 435 ~ u b i n g , eccentric, field in. 174 Welded steel pipe, testing of, 436 T u b ~ n g , seamless steel, 426 Weldments, defects in, defined Turbines, steam. 431, 432 Cracks. 89 Turns, ampere, defined. 124 Lack of fusion. 89

Lack of penetration, 89 Undercutting, in ~velds, 89 Uudercutting, 89 Undei,filled rolls, scanis from, 82 weidments; testing of U. 5 . A l r Force T. 0. '~. 57. 458 Current, types of, used in, 441

Certification of operators. 57, .- Dry powder, used for , 252, 253, 4b8

Ultrasonic methods. 56 ComDarison with magnetic

particie testing. 68 F o r mapping of laminations. For weld testing. 443. 447

Ultra-violet light. 274, 290 Black light. 125. 274. 290 Filters for, 291 Near-ultraviolet light, 274 Used for mineral prospecting

275

Variable transformers, for demagnetization, 316

Variations in techniques, 229 Vector fields, defined, 128

Creation of, 308 Vessels, presure, tes t ing of, 448 Vibration, effect of

As a cause of failure, 415 I n demagnetization, 324 I n magnetization, 324

Virginia S ta te Road Commission, S~ecif icat ion for welded

structures, 464 Viscosity, effect of, in oil baths, 258 Visibility of maanetic uarticles.

221, 223 Visible vs. fluorescent magnetic

particles. 346 Volt, defined, 124

W a t e r , 220, 256, 260 Suspensoid for magnetic

narticles. 58 ,~ .~ ~ . Cor~dir~oi~cr . i , z\.:iilnolr. 228 Corrosion of etluipmrnl u!.. 260 F o n m l n ~ of b:lth. Rlii - ~~ ~~~~~~~ .-- Freezing of bath; 259 Hazards from use of; 260

vs. oil f o r wet baths, 224, 345 Weld defects defined. 440, 441 Welded ]unction, indication of, 73 Welded steel missile motor cases,

359

443 Half wave vs. A.C. for , 445 Large, 410 1,eerh contacts used in, 248 Magnetizing techniques for , 442 Other N.D.T. methods for; 443 Pipe, eleetrrcal res~s tanee welded,

447, 448; 449 Prod contacts used in. 248, 443 Rough beads, effect of, 253 Structural, of buildings, 446 Yokes, used f o r testing, 442,

444, 445 Vessels, pressure, 448, 449

Welds, boundary zone indications, 2sn ---

Weston light meter, for black light, 302, 474

Wet method Advantages of, 256 Agglomeration of fine particles

in. 212 ~ a t l ; constituents, 257 Choice of, vs. dry. 235 Circuiar magnetization, 267, 268 Cleaning a f te r testing, 270 Continous method in, 239, 267 Cost of, 259 Disadvantages of. 257 And dry, compared, 235, 237 Errors , in mi r ing of bath, 467 A s a factor in equipment design,

345 ire-hazards of. 224. 255 Fluorescent, advantages of. 275

Disadvantages of, 276 Fluorescent particles for, 212,

274 276 - . . , - . . Lacquer method in, 273 Longitudinal magnetization for , 26R

~ i i i - t e n a n c e of bath for, 265 Making up bath for, 263, 265 Materials for, 223, 257, 260 il1e;isunng bath. s t rength, 262.

SUBJECT INDEX

Paste form of, 221, 26.1 Permeability of, 215 F o r residual method, 239, 269 Shape of particles, 213, 214 Size of particles. 211 Suspending liquids. 220, 255, 257 Visibility of, 221, 223 Wrong particles f o r bath, 467

White light; exclusion of, f o r Fluorescent testing, 279

White light spectrum, 291 Wright Aeronautical Company, 255 Wright Field, 255 Wri t ing , magnetic; 386

257 X-rays, first use of, 47 Uniformity nf "1 Of bath* Importance Lester, H. H., industrial use. 48

- - < --- Viscosity of bath for , 258

Wet method materials Agglomeration of, 212 Available, list of. 225, 227 Coercive force of, 216 Colors of, 221, 268, 346 F o r continuous method, 239, 267 Contrast of, 221. 223 Density of. 212 Dispersion of, 214, 221, 223, 224 Dry concentrates, 224. 263 Fluorescent, 212, 274, 276 Hysteresis curves of, 216, 219 Mobility, 220 Nature of, 223

Yellow glasses. for black light inspection, 289, 305

Yokes, 326 Alternating current. 145 Demagnetizing with, 145, 317,

318 Direct current. 145 Electromagnetic, 145, 243 Permanent magnet, 142, 243 F o r vee butt welds, 442 F o r weld testing, 444

Zones, boundary, in welds Non-relevant indications of. 390

With MAGNAFLUX Corporation, he was appointed Manager of the Eastern Region in 1940, with offices In New York City, where he remained until the fall of 1944. He then returned to Chicago a s F i rs t Vice President and was made Executive Vice President in 1945.

Carl was a member of the American Soclety for Metals. The American Soc~ety for Test- ing Materials, the Soelety f o r Nondestr~rctive Testing, and the Amer~can Association for the Advancement of Science. He frequently ad- dressed chapters of those and other societies in many cities from coast to coast.

He developed many of the mater~ais and techniques of the magnetic particle testing method and this book is drawn largely from his own personal knowledge and experience. He is the author of many published papers in the field of nondestructive testing and is co- author with the late F; B. Doane of the book "Principles of Magnaflux." He is the author of "Princlpies of Penetrants" published by S'AGNAFLUX Corporation in 1963. and a contributing author to the "ASNT Hand- book."

I n 1959 Carl was honored by the Society for Nondestructive Testing and gave the Lester Honor Lecture a t that Society's annual con- vention, under the title "The Nondestructive Testing Engineer-Today's Career Opportu- nity." At the National meeting of the Society for Nondestructive Testing in 1965 he was again honored by belng made a n Honorary Life Member of the Society.

Carl retired from MAGNAFLUX Corpora- tion in 1963.

On Januarv 26.1981. Cari Betz passed away In Los ~ l t o ' Hills. CA. On his frlends, h ~ s company, and h ~ s world he left a l a s t ~ n g mark. All will mlss hlm.

principles of magnetic testing ce betz

Documents