operating stresses in aircraft-engine crankshafts and connecting

UNCLASSIFIED

AD NUMBER

LIMITATION CHANGESTO:

FROM:

AUTHORITY

THIS PAGE IS UNCLASSIFIED

ADB803953

Approved for public release; distribution isunlimited.

Distribution authorized to DoD only;Administrative/Operational Use; AUG 1945. Otherrequests shall be referred to NationalAeronautics and Space Administration,Washington, DC. Pre-dates formal DoDdistribution statements. Treat as DoD only.

NASA TR Server website

MR No. E5H18

NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS

WARTIME REPORT ORIGINALLY ISSUED

August 1945 as Memorandum Report E5H18

OPERATING STRESSES IN AIRCRAFT-ENGINE CRANKSHAFTS

AND CONNECTING RODS

II - INSTRUMENTATION AND TEST RESULTS

By Francis J. Dutee, Franklyn W. Phillips and Howard F. Calvert

Aircraft Engine Research Laboratory Cleveland, Ohio

WASHINGTON

NACA WARTIME REPORTS are reprints of papers originally issued to provide rapid distribution of advance research results to an authorized group requiring them for the war effort. They were pre- viously held under a security status but are now unclassified. Some of these reports were not tech- nically edited. All have been reproduced without change in order to expedite general distribution.

E-191

3 1176 01364 8291

HACA MR Ho. E5KLS

HATI0HA1 ADVISORY COMMITTEE FOR AEHOHA.UTICS

. . -MEMORAHDOM REPORT

for the

Air Technical Service Command, Army Air Forces

OPE&SLTIHO STRESSES IiT AIRCRAFT-EH&INE CEAHESHATTS

AND OCNIiECTING BODS

II - IKSTREKEITTATIOH ASS TEST RESULTS

By Francis J. Dutee, Franklyn W. Phillipe and Howard F. Calvert

SUMHARY

An investigation was conducted to develop methods and equipment for the measurement of operating stresses in aircraft-engine crankshafts and connecting rods. Ranges of alternating stress were measured under motoring and po\jer conditions at several locations on the master connecting rod and at one location on the crankshaft of a nine-cylinder radial air-cooled engine modified for test purposes. The results indicate that the measurement of stresses in connecting rods is feasible. Some resonant bending vibration occurred in the rod under all engine operating conditions investigated, but no ap-oreciable resonant axial vibrr.tion was found.

INTRODUCTION

At the request of the Air Technical Service Command, Army Air Forces, a program is being carried out at the HA.CA Cleveland laboratory to develop methods of measuring strains in critical sections of rotating-shaft systems. Severe!, crankshaft failures have been caused by excessive strains due to bending vibration of the crankshaft at its natural frequency. Crankshaft stresses are extremely difficult to determine bv analytical methods because many factors, such as natural frequencies, forcing functions, damping capacity, and bearing clearances, are involved and some of these factors cannot be accurately determined. Torsional vibration is commonly measured for experimental analysis of crankshaft stresses; bending

KACA m So. E5H1S

vibration of the nhaft, however, cannot "be determined from torsional- vibration data. Wire strain gages have therefore "been used for meaourement of 'bending vibration (reference l).

No literature on the measurement of operating stresses in connecting rods of high-speed engines has come to the attention of the authors. The importance of investigating vibration conditiors in connecting rods has been recognized hut, because suitable instrumentation ."Mid methods were lacking, experimental strain-measurement investigations have not been made heretofore.

The instrumentation and methods that were used to measure stresses in the crankshaft and the master connecting rod of a modified test engine are described herein and the test results are pre- sented. Wire strain gages were employed for the measurement of dynamic strains in the crankshaft and the inaster connecting rod. Strain gage8 mounted and used as described yielded an over-all strain signal of relatively large amplitude, which minimized the effect of distortion caused by slip rings and high-gain amplifiers.

DESC?.IPTIOi:T Off APPARATUS

TeBt Engine

The strain-meaBuremer.t investigation was conducted on an aircraft engine that had be^n converted to a three-cylinder test engine. The iflpster-rod cylinder (FO. l) was operated under power, whereas cylinders h and 7 were motored with exhaust and intake valves installed but with valve push rods removed. The power was absorbed by an electric dynamometer. In all tests, standard laboratory equipment was used to measure engine speed, power output, and cylinder temperatures. The engine was oporpted with "traospheric inlet pressure inasmuch as supercharging facilities were not availablo.

Strain Gages and Circuit Diagram

The strain gages used, vere attached by means of Balrclite 3C- 6035 cement. All gages were of 120-ohm nominal roslstanco with a strain-sonsitive area of l/S by i/o inch Bquare for each gage. The strain wan measured by four strain gages, which were mountod on the strained member closely adjacent to each other at the desired locations. (See fig. l(aX) The strain-sensitive filaments of all four strain gages were orientated in the direction in which it was desired

KAOA MB Ho. E5EL5

to measure the strain. She filaments of two of the gages were made of Advance wire and those of the other two gages were made of Kovar wire. Advance wire and Kovar wire have strain-sensitivity factors that pro opposite in sign and approximately equal In magnitude. She four strain gages were interconnected to form an electrical resistance bridge of the "&eatstone type (fig. 1(h)) with Advance wire gages in two opposito arms of tho bridge and Kovar wire gages in the other two opposite arms, ^he effect of the resistance change of all gages was to change the bridge balance by an amount proportional to the average strain applied to the four strain gages.

The term "strain picZaip" is usod herein to refer to the combi- nation of four strain gages described in the foregoing paragraph. She effective strain-Bensitive area of each strain pickup was 5/l6 inch long by 3/8 inch wide.

Crankshaft Strain-Measuring Installation

One strain pickup was located on the propeller side of the front crack arm in tho position shown in figure 2. This strain pickup measured tho total strain, which consisted of the sum of tho alternating strains duo to bending and average tension and compression loads in the crank arm. She circuit diagram for this strain pickup is shown in figure 3« Xho slip rings (fig. h) were mounted at the outer end of the propeller shaft on the dynamometer coupling flange.

Installation for Measuring Strains in {faster Connecting Sod

Six strain pickups were located on tho master connecting rod in tho positions shown in figure 5» All strain pickups were orientated with strain-sensitive filaments located longitudinally with respect to the connecting rod. Tho circuit diagram for a typical strain pickup is shown in figure 6. Six slip rings wero installod on the master connocting rod (fig. 7)* T^° strain pickups were divided into threo groups of two with the battery terminals for each pair connected in parallel. Six slip rings were therefore sufficient to operate one group of strain pickups at a time. The lead wires from the slip rings could be transferred from one group of strain pickups to another by shifting short load wires on the master connecting rod.

-rl— ü '-•£££&.±£:£L. L'. _1^>Üä£LJ^J • ».»•'•"-iij;-..' — -J _> 'i • m ' *..-_.-

k NACA HE Fo. E5H1S

Slip Sings and Brushes

A description of tho slip rings and the "brushes used in this investigation and their performance characteristics is ^ivon in reference 2. Hie slip-ring and "brush contact surfaces were porpon- dicular to the axis of rotation. She slip rings were made of shim brasB 0.0l6 inch thick with a Tickers hardness number of 150. Silver-graphite brushes l/3 inch in diametor and 9/l6 inch long were used throughout those tests. Four "brushes were used for oach slip ring with a contact pressure of 250 pounds per square inch for the slip rings mounted on tho master connecting rod and for the slip rings fastenod to the outer end of the propeller shaft on tho dyna- mometor coupling hub. $ho slip rings mounted on the connecting rod wore oporatod in oil and the slip rings mounted on the dynamometer coupling hub woro operated dry. All slip rings and tho "brush holders mounted on the crankshaft (fig. 7) were attached with B?Jcelite BC-6035 cement.

Becording Instruments

Tho dynamic strains were recordod "by a moving-coil typo of oscillograph with a galvanometer element having an undamped natural froquency of 2C00 cycles per second. Hie froquoncy response of the moving-coil type of recording-oscillograph system was found to ho constant within +3 percent from 15 to 500 cycles per second. A moving-coil olemont with an undamped natural frequency of 500 cycles per socond recorded the timing signal, which was obtained from a pickup coil consisting of a few turns of \riro wrapped around one of the spark-plug wires.

A cathode—ray oscilloscope in conjunction with a rotp.ting-drum camera was used aB a single-channel rocording oscillograph to obtain a few strain records for comparison with the records obta_ned with tho moving-coil oscillograph. The frequency response of the cathode- ray oscillograph was found to he constant within +5 percent from lM- to 2000 cycles per second. Figure S shows- comparable records obtained with "both oscillographs from the same strain pickup under the same engine operating conditions.

EIST PB0CEDUB3B

Strain records were obtained from each of the strain pickups for ongine spoeds of 1250, 1500, 1750, and 2000 rpm for both motoring and power conditions. All power runs were made at full throttle

:; •?••« • ":—rr^rrr? >^-:'^;^:. • • SS??^--_ ••.!£:;•

H4CA MR No. E5ELS

without "boost. Power was determined on an indlcatod "basis "by adding the powor required to motor the engine to the "brake horsepower.

All strain measurements were made with a strain-gage ourrent of 30 milllampcres and with the same amplifier-gain setting. The ordinatcs of all tho oscillograms havo tho same scalo and thereforo aro diroctly comparable, übe film spood was 10 inches per second for all of tho osoillograms except for tho record from the cathode- ray oscillograph shown In figure 8, which was 5*V inchos per second.

Tho amplifiers and the oscillographs wero calibrated "by recording tho strain signal from a strain pickup mounted at a point of known alternating strain on a vibrating csmtilovor "beam. Bio same strain-gago current and amplifier-gain setting wero used during the calibration as wero used during the ongino strain-moaBuring tosts. She accuracy of calibrations doponds upon the equality of the strain sensitivity of tho various strain pickups* Bio strain sensitivity of a large number of strain gages of tho type usod in this investigation was measured, and the variation was dotormined to be not groator than +1 porcont.

E3SULTS AI2D LISCUSSIOjJ

Strosses in Crankshaft

She strain in tho crank arm at tho location shown in figure 2 was moasurod by strain pickup 1« This pickup was locatod at a convenient position on tho crankshaft to provide a strain signal that would domonstrato tho suitability of the instrumentation for moasuring dynamic strains in tho crankshaft, ilo attompt was mado to locate tho strain pickup at tho point of maximum stross, inasmuch as dovolopmont of instrumentation tochniquos was tho prime objective in this preliminary phase of tho work. The oscillograms from strain pickup 1 aro shown In figure 9» Tho highest stress range measured at this location for power operation was 12,700 pounds per square inch at speeds of I5OO and 1750 ?pm. Xho maximum stross rango when tho engino was motored was 5150 pounds per Bquaro inch at a speed of 1750 rpm. Thoso stressos wero computed from tho oscillograms of figure 9 with the assumptions that the strain pickup was orion- tatod in the direction of tho major principal axis of strain and that tho stress was uniaxial.

j'Tg-3.'>tif.-av ••?->^^.^^^ffi^x^i •*it- c _-,

EA.CA MR Ho. E5ELS

Stresses in Piaster Connecting Bod •

Bosonant vibrations were found in all tho oscillograma (fig. 10) obtained from strain piclcop 5» which was locatod on the flange of tho I-beam soction of the master connocting rod ono-third of tho distance from the center lino of the mastor-rod piston-pin hoaxing to the center lino of tho mastor-rod crankshaft hearing. The resonant vibrations wero not sustained hut varied from a maximum to a mini- mum amplitude sovoral times throughout the engino cyclo. Some of the OBcillograms of figure 10 are roproducod to a largor scale in figure 11 to show the wave form more clearly.

Strr.in pickup 2 was locatod on tho center lino of the wob of tho mast or connecting rod at the same cross soction as pickup 5- This pickup was so close to tho noutral axis of the I-beam soction that it could not have been greatly affectod by bending) end the oscillograma aro thorofore fairly representative of the average longitudinal strain in the connocting rod at this cross section. Little rosonant vibration was obtained from this pickup. (See fig. 12.) It is thoroforo concluded that tho rosonant vibration obtained in tho records of figuro 10 was largol-' due to bending and that littlo a Trial vibration of tho mastor connecting rod occurrod.

Tho rango of alternating stress was determined for oach strain pickup and test condition by comparing tho signals from the strain pickups with the signal from a strain pickup mounted on a calibrator bar subjectod to a known uniaxial beading stross. Shis procoduro is basod on the assumption that the stross in the connecting rod is uniaxial with tho major principal axis along tho longth of tho rod. She results aro listed in table I. Tho maximum rango of stross obtoinod in tho mastor connocting rod for any of tho locations and conditions tosted wns 32i200 pounds por square inch occurring in pickup 6 at an ongine spocd of 2000 rpra.

Tho measured range of stress in tho master connecting rod as determined from tho six strain pickups varied directly with tho square of tho ongine spood (fig. 13) owing to inertia effects whon the ongine was motored by means of tho dynamomotor. In tho analysis of tho oscillograma for inertia stress rango at the lowor engino speeds, tho poak oomprossion-stress amplitude at the timo of peak compression pressure and the marl mum tensile-stress amplitude that immediately followod wero disregarded.

Tho data prosontod herein demonstrate that it ii practicable to measure dynamic strains in connocting rods. Tho engine power output could not be oxtendod to high moan effective pressures bocause boost

HACA MR Ho. E5ELS

pressure was not available. Ho difficulty should be involved with the strain-measuring system if the program is extended to operating conditions at higher engine speeds and mean effective pressures.

SUMMARY Q? RESULTS

Tho results of tests of the dynamic strain-measuring system dovelopod for the E-1820-25 toot ongino ind do scribed horoin showed that I

1. Clear and reproducible oscillograms were obtained from strain pickups locatod on the crankshaft and the master connecting rod.

2. Resonant bending vibrations of tho master connecting rod in a plane perpendicular to tho crankpin were found to occur in appreciable magnitude for all conditions of engine operation tested.

3- No apTjrociablo resonant axial vibration was found in the master connecting rod.

h. The experimental valuos of tho inortia stress range measured with pickups on the master connecting rod varied directly as the squire of the engino speed.

Aircraft Engino RoBenrch Laboratory, iTp.tional Advisory Committee for Aeronautics,

Clevoland, Ohio, August 13, 19^5«

REFEKEiTCES

1. Walstroin, Douglas P.: Mensuremont of Operating Stresses in an Aircraft Engine Crankshaft under Power. iiACA ARR No. E5BCI, 19^5-

2. Dutoo, Francis J., Phillips, Pranklyn V., and Kemp, Richard H.: Operating Strosses in Aircraft-Engine Crankshafts p-nd Connecting Rods. I - Slip-Ring and Brush Combinations for Dynamic-Strain Measurements. HACA MR No. E5C3O, March 30, 1945.

3 EACA MR Ho. E5H18

TABLE I - MASanSRr-COFjECTING-EOD STHESSES FOE ALL STEAM-PICKUP

LOCATIONS AND ENGINE OPERATE«} CONDITIONS INVESTIGATED

Strain Engine Indientod Range of pickup spoed mean alternating

(rpm) effectives stress pressure (lD/aq. in.) (ID/ST. in.)

1 1250 07.2 10,900 1500 97-6 12,700 1750 100.7 13,600 2000 94-5 11,500

2 1250 100.?. 23,300 1500 100.7 24,600 1750 103.0 24,1100 2000 101.2 25,300

3 1250 96.8 24,000 1500 93.6 21,400 1750 99.9 23.S00 2000 91-9 20.5U0

1* 1250 102.5 21,100 1500 101.1 20,900 17S0 102.5 20,700 2000 GO.5 19,700

5 1250 106.5 29,300 1500 109.1 30,000 1750 107.0 29,200 2000 92.g 27,900

6 1250 103.0 23,900 1500 IO7.3

107.4 31,100

1750 3I.5OC 2000 96.S 32,200 j

'National Advisory Committee for Aoronauticn

NACA MR No. E5HI8

Kovar Advance

i

Advance Kovar

(a) Relative location of four strain gages

Kovar dvance

To galvanometer

(b) Basic wiring diagram.


Figure I. - Strain pickup.

NACA MR No. E5HI8

Ax i s oj or ientat ion

Strain pickup I


Figure 2. - Location of strain pickup on propeller side of the front crank arm.

NACA MR No. E5HI8

Advance-w ire stra in gage , I20Q res istance

R^ Kovar-wire strain gage, I20Q res istance

Port i on of circuit attached to crankshaft

Slip r i ngs on crankshaft (operated dry)

Record ing osc iI lograph 4 IH


Figure 3. - Circuit diagram used to measure alternating strain in crankshaft.

NACA MR NO. E5H18

Figure 4. - slip ring used on crankshaft

NACA C-9548 4-7-45

IF

NACA MR No. E5HI8

\


Figure 5. - Location of strain pickups on antiprope 1ier side of master connecting rod. All pickups have strain-sensitive filaments orientated longitudinally with respect to the rod.

NACA MR No. E5HI8

9 v

Amp 1 if ier Record i ng osc i1lograph

Advance-wire . strain gage, I20Q resistance

Kovar-wire strain gage , 1202 res istance

port ion of c ircu it attached to master connect i ng r od

-Slip rings on master connecting rod (operated in oil)

-Slip r i ngs on era nk- shaft (operated dry)

4 1 •—J


Figure 6. - Circuit used to measure alternating strain in master connecting rod.

NACA MR No. E5H18

NACA 04549 2-10-46

Figure 7. - View of master connecting rod and crankshaft assembly showing strain-measuring instal I at ion.

Crankshaft rotation, 720° s

ft

i\v

z > >

X)

z o

CD

s

Record from cathode-ray oscillograph

• äT* «MEt,-,?..*?" aMw^-itsBk »•<>•<, r *>f

*-*äS-

1$»

«*•» pfMp^i>MMgtla*«^*aiMM*MM>a^i*IM ^m? i^r*> » r»i"

Record from moving-coil oscillograph

Figure 8. - Comparison of records obtained by the moving-coil-type oscillograph and the cathode-ray type of oscillograph

NACA C- 11842 7.24-45

NACA MR No. E5H18

i» .i •.

4ÖV*/*>*•»?!

•"_»•»<•- •»• - t, ] -.^-^ ^vf, w,»i — .„^4

A ""* <r * v. *M«lMM*i

* *JT * V» .- '

J>' - v«

1250 rpm

*VvvsVWVvv%V^

1500 rpm

i mep , 103. 2 lb/«q in.

i mep, 105.1 Ib/sq in

v^/v^ ^^fif^^f^ 12.7-00 (b/sq in.

1750 rpm i mep, 10 2.0 Ib/sq in,

Möto ring

2000 rpm

• «•! I Jl I ••-

i mep, 94.9 (b/sq in

Power

Figure 9. - Osci I log rams from strain pickup I on crankshaft for all engine conditions tested. (Reduced to one-half original size.) Ignition timing, 20.° B. T. C.

NACA C- t 1643 7.24.45

ll •• i mil ••• i

NACA MR No. E5H18

••<-**. -\» i- > j

mep, 106.5 Ib/sq in 1250 rpm

-^ Ig rtit ion

1500 rpm

W^^/Vr«^^

i mep , 109.1 Ib/sq in

Jv4A*UA/> 1750 rpm

imep, 107.0 Ib/sq in

imep, 92.8 Ib/sq in.

20 0 0 rpm

Motoring Po we r

Figure 10. - Osci I log rams from strain pickup 5 on master connecting rod for all engine conditions tested". (Reduced to one-half original size.) Ignition timing, 20° B.T.C.'

NACA C 11844 7-24-45

-^

!» '. /•< > t , ,f

•c**v7jvt iff- r

,yyt « •

^ff^Wf/fr^ v Ignition

Moto ri ng

1500 rpm

Tv-'f *\ ' '•&•» * * <\ • C •

>• 1• ^ ,-* *'

1 •"*

>M;3^.„B

7 50 rpm Powe r

NACA C.11845 7.24.45

z > o >

33

Z o

m x CD

Figure II. - Oscillograms from strain pickup 5 on master connecting rod for engine speeds of 1500 and 1750 rpm. (Original size.) Ignition timing, 20° B.T.C.

NACA MR No. E5H18

r i—UTtf tr"ii~l—r"v " nir1 • L i r .. . . j 1

1mep, 100.2 Ib/sq in. I 250 rpm

-"-T

S^/SW^vv^vv^v^

&- •*„ ^1 ~ ,*£/>.>*

I 5)00 rpm

Ignition

i mep, 100.7 I b/sq in

w^i mep, 103 I b/sq in

1750 rpm

^vwwwww^

2000 rpm i mep, 10 1.2 I b/ sq in

Mo to r i n g Po we r

Figure 12.- Oscillograms from strain pickup 2 on master connecting rod for all engine conditions tested. (Reduced to o n e-h a I f original size.) A, compression pressure peak; B, explosion pressure peak; ignition timing, 20° B.T.C.

NACA C 11846 7-24.45

NACA MR No. E5HI8

12,00C

KT 8.00C V)

8 » 10 V u *» CO

4,000

Engine speed, rpm Figure 13» - Effect of engine motoring speed on the range of

alternating Inertia stress measured In the master connecting rod. Abscissa scale proportional to square of engine speed.

NACA MR No. E5HI8

12.00C

er 8.00C u

g 4,000

a 10 «

CO


1000 1550 I5ÖCT

Engine speed, xpm Figure 13. - Effect of engine motoring speed on the range of

alternating Inertia stress measured In the master connecting rod. Abscissa scale proportional to square of engine speed.

AYO- 8302

Instrumentation and Test Results UVMON

none AUTHORS): Dutee, F. J.; Phillips, F. W. ORIGINATING AGENCY: National Advisory Committee for Aeronautics, Washington, D. C.

ouo. Aamci m.

PUBLISHED BY: (Same) rUBUIHNO AOCMCV HO.

Mia Aiw '45

DOC a a. TInrlafn.

COUNTIT

U.S. tAHOUAOi

' Eng.. FAOO

21 nunriAiiom Bhotos, tables, diajrs, graphs

ABSTRACT:

Dynamic strain measuring system was developed as a means of determining operating stresses in aircraft engines. System was tested in special R-1820-25 radial engines. Reproducible oscillograms were obtained from strain pickups on crankshaft and master connecting rod. Resonant bending vibrations of master connecting rod occur In appreciable magnitude at aii operating conditions, but little resonant axial vibrations occur. Experimental values of Inertia stress range measured with pickups on master connecting rod varied directly as the square of the engine speed.

DISTRIBUTION: Request copies of this report only from Originating Agency DIVISION: Power Plants, Reciprocating (6) SECTION: Components (11)

ATI SHEET NO.: R-8-11-17

SUBJECT HEADINGS: Engine components - Vibration testing (32830.4); Stress analysis - Test Installations (90805)

Air Documenta DlvUion, Intel I ig« nc« Dopartmint Air Material Command

Ala iCCHNICAL INDEX Wright > atrsruii Air Forco Bato Dayton, Ohio

operating stresses in aircraft-engine crankshafts and connecting

Documents