Crimper Process Optimizationusing DOE

Koushik GaneshMonish NavapariaSanket Deshmukh

Under Guidance of ,Sara Wolfson

Subramanian Iyer

associated with:

Contents

• Project Charter• Crimping tools and process• Pareto Analysis• Root cause analysis• Capability analysis• Backstop Standardization• Measurement system ( Gage R&R)• Our Approach• Future work

Project Charter Feb 2016 – April 2016

Background:At WC103 Crimper downtime is accounted for 2.8% of total downtime. The total defect rate in the line is 4259 PPM. Leakers account for 25% of all defects in themonth of January. Currently the crimp quality is determined by line operators, mechanics and quality, but there is no standard for acceptable crimp.

Deliverables:Determine standards for good crimp and ways to measure or quantify, and the optimal crimp settings.

Scope:All products on WC103 JG crimper

Goals:Increase RFT by 0.2%, reduce defects PPM by 200 and increase OEE by 0.5%

Defining tools

Top belt Inner sleeve

Outer sleeve Collet and Backstop

Defining tools

Defining tools

Collet assembly Whole assembly

The process

Crimp Quality definition

Loose

Loose

TightTight

Good

Pareto Chart Analysis

Out of 96 products produced at WC103, the pareto analysis helps us to focus on the vital few products. Standardization of crimping process is initially focused on the vital products.

NAUT VOY M EDT NS 1

00ml EU

BEY RISE RG EDP 100ml 1

4 IV

GS NIGHT TST

EDT 50ml 13 IV

GS NIGHT SF

EDT 50ml 13 IV

BEY WILD

RG EDP 100ml 14 IV

BEY HEAT RG EDP 30ml

GS NIGHT RG EDT 30ml w

/SEC TAG 15 USA

BEY KISS RG EDP 50ml 1

5 IV

BEY WILD

SF EDP 30ml 1

4 USA

CK DTW RG EDP 50ml 13 IV

GS M RG EDT NS 7

5ml 10 IV

NAUT VOY N83 RG EDT 30ml 14 USA

NAUT VOY N83 RG EDT 100ml 13 IV

BEY WILD

SF EDP 15ml 1

4 IV

VWG EMB RG EDT NS 30ml P

ERIW&IRIS

15 IV

BEY HEAT XM10 SF EDP 30ml E

U

BEY HEAT Beyonce

Tour EDP100ml

BEY MNIGHT SF

EDP 30ml 12 IV

BEY WILD

RG EDP 30ml 14 IV

GS M TST

EDT NS 75ml 1

0 IV

KP KILL Q

RG EDP 100ml 13 USA

NAUT LIFE

RG EDT NS 50ml S

EC TAG15USA/PR

NAUT LIFE

SF EDT 50ml C

TN WIP

NAUT VOY M TST

EDT NS 100ml

0100200300400500600

0.00%20.00%40.00%60.00%80.00%100.00%120.00%

103 Crimper/product Percentage

9

Selection of vital product

Product DescriptionUnits

Produced 103 Crimper/product 103 Rework/ product

GS MARC W RG EDP 100ml 10 IV 248673 433 73

BEY HEAT RG EDP 100ml 448608 390 49

GS MARC M RG EDT NS 100ml 10 IV 149175 319 35

BEY RISE RG EDP 100ml 14 IV 185411 286 167

BEY MNIGHT RG EDP 100ml 12 IV 157627 212 208

BEY KISS RG EDP 100ml 15 IV 163716 194 47

NAUT VOY M SF EDT 30ml WHITE LOGO 10 IV 139999 185 0

GS NIGHT TST EDT 50ml 13 IV 12423 172 0

GS MARC W TST EDP 100ml 10 IV 20122 165 0

BEY HEAT RG EDP 100ml HEAT TAGGING 10USA 135694 162 164

Vital products were selected to attain more benefit with lesser effort.1. Beyoncé Heat 100ml and it’s flankers2. Guess Marciano 100 ml

Cause effect analysis

The effects were discussed with varied authorities Co-ordinator, Line In-charge, Mechanics, GPWs

Potential causes of variation

•Backstop HeightNo robust mechanism to calibrate standard collet backstop assembly

•Stroke lengthVariable stroke length leads to inconsistency in crimping

• Inner SleevePosition of inner sleeve is another potential reason for bad crimping

• Machine HeightMachine height is not standardized product-wise. Height is set “manually” leading to variation

Theoretical versus Actual measurementEssential Rules for good crimp:

1 - Correct shaped crimp collet

2 - Well adapted backstop

3 - Crimping Diameter

4 - Crimp Height

Theoretical value of crimp height for Beyoncé Heat

h= 2.6±0.2mm

Jc= 1.07mm (7% compressed)= 0.995mm

2e= 0.81mm

H= 4.405±0.2mm

Crimp Height:H=2e + Jc + h

h= height of glass carnet

e=thickness of ferrule

Jc=thickness of gasket

Measured values of crimp height

Capability studyTo evaluate the Sigma quality of crimping process at WC 103

• Capability analysis on measured crimp height for Beyoncé Heat• Analysis was done on 8 heads over 150 bottles• Observed CpK = 0.273 and Sigma quality = 2.336

One way analysis

Lower Limit – 24.85 mm Upper Limit - 25.20 mm

Backstop height standardization

• A cylinder used as a gauge to adjust the back stop height

• Two marked red lines in the cylinder indicate the limits

• The collet head is calibrated between the two lines

Advantages:• Consistent crimp quality in all heads• Simple implementation• Faster changeover times• Better the currently existing trial and error system

Measurement System Analysis ( Gage R&R)• Gage R&R study to standardize backstop height• Repeatability & Reproducibility effects were studied• Total gauge R&R is 7.08% (<10%)• Significant variation was only between the different crimp qualities

Our approach• Select a product and find a theoretical value of crimp height• Set the machine and collet assembly to perform an initial crimp• Compare the crimp height with theoretical value• Change the collet backstop assembly by rotating backstop at

intervals of 45 in either direction⁰• Measure and note the crimp height• Go up to the quality of crimp is unacceptably deteriorated (too

tight/ too loose• Mark that level as upper/lower limit• Rotate the backstop in the other direction and continue the

process• Setting Upper Specific and Lower Specific Limit based on the

95% confidence limits

Future work• Poka-Yoke design

Outer sleeve design Top belt design

Advantages:• Prevention of wrong pin insertion• No problems of loose outer sleeve

Future Work

• Hardened tool steel (typically A3) for inner surface of inner sleeve

• As the inner sleeve is under constant movement, the inner surface faces abrasive wear and effective diameter increases

• Effective diameter increase leads to inefficient crimping

• A3 carbon steel can withstand more cycles without wearing off

Hardened tool steel for inner surface

Thank You

Special thanks to -Robert Leder-Vasanth-William Morin-Ramon-Dr. Timothy Clapp-Dr. Steve Jackson