ec timer box product design for six...

EC Timer BoxProduct Design for Six Sigma

2

Customers: Whirlpool, Bosch, AEG, and others

Application of Product

3

Goals

To Decrease the reject rate by a factor of 10 from 10% to less than

1%

4

Test machine

Test cam

Contactsprings

LayerStacker

5

Gage R&R on Automatic test Rig Attempt 1 Gage name:Date of study:Reported by:Tolerance:Misc:

Automatic Test rig22/09/98Switchbox team0.5mmTest on switch 1a level 1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

23.34

23.39

23.44

23.49 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 1010 1111 1212 1313 1414 1515 1616 1717 1818 1919 2020

Part

Xbar Chart

Sam

ple

Mea

n

X=23.40

3.0SL=23.46

-3.0SL=23.33

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

0.00

0.05

0.10

0.15

1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 1010 1111 1212 1313 1414 1515 1616 1717 1818 1919 2020

Part

R Chart

Sam

ple

Ran

ge

R=0.03475

3.0SL=0.1135

-3.0SL=0.000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

23.35

23.45

23.55

Part

By Part

%Total Var%Study Var%Toler

Gage R&R Part-to-Part

0

50

100

Components of VariationP

erce

nt

Gage R&R (ANOVA) for Value

6

To Make a Long Story Short….

• Team determined that the test was failing good units.

• Determined failures were related to rotation speed of the cam

7

Chi-Square Test

Expected counts are printed below observed counts

Test1 Test2 Total50000 96 4 100

83.89 16.11

70000 97 3 10083.89 16.11

90000 93 7 10083.89 16.11

100000 33 37 7058.72 11.28

120000 75 25 10083.89 16.11

150000 59 11 7058.72 11.28

Total 453 87 540

Chi-Sq = 1.748 + 9.104 +2.049 + 10.670 +0.990 + 5.152 +11.267 + 58.667 +0.942 + 4.904 +0.001 + 0.007 = 105.502

DF = 5, P-Value = 0.000

Effect of test rotation speed on radius rejects

15000010000050000

0.5

0.4

0.3

0.2

0.1

0.0

Speed

radi

us re

ject

spr

opor

tion

of p

arts

rete

sted

&

Current operating speed

CONCLUSION: WITH THE CAM ROTATION SPEED OF 70000, FALSE REJECTS CAN BE REDUCED BY A

FACTOR OF 10

8

However……….We were failing our suppliers for perceived

Variation on Critical Dimensions.

29,35 29,40 29,45 29,50 29,55 29,60

LSL USL

Process Capability Analysis for A1=29,4

USLTargetLSLMeanSample NStDev (ST)StDev (LT)

CpCPUCPLCpkCpm

PpPPUPPLPpk

PPM < LSLPPM > USLPPM Total



29,4500 *

29,350029,5305

300,03188430,0286466

0,52-0,84 1,89-0,84

*

0,58-0,94 2,10-0,94

0,001000000,001000000,00

0,01 994211,00 994211,00

0,00 997523,74 997523,74

Process Data

Potential (ST) Capability

Overall (LT) Capability Observed Performance Expected ST Performance Expected LT Performance

STLT

29,35 29,40 29,45 29,50 29,55 29,60

LSL USL

Process Capability Analysis for A1=29,4


CpCPUCPLCpkCpm

PpPPUPPLPpk




29,4500 *

29,350029,5305

300,03188430,0286466

0,52-0,84 1,89-0,84

*

0,58-0,94 2,10-0,94

0,001000000,001000000,00

0,01 994211,00 994211,00

0,00 997523,74 997523,74

Process Data



STLT

Mean =24.4285 Stdev.= 0.02399Mean =29.53 Stdev.= 0.028368

Mean =15.15 Stdev.= 0.0161965Mean =23.4538 Stdev.= 0.0239959

24,5524,5024,4524,4024,35

USLLSL

Process Capability Analysis for B1=24,5

PPM TotalPPM > USLPPM < LSL



PpkPPLPPUPp

CpmCpkCPLCPUCp

StDev (LT)StDev (ST)Sample NMeanLSLTargetUSL

881859,66 0,01

881859,65

861051,50 0,18

861051,33

833333,33 0,00

833333,33

-0,39-0,39 1,85 0,73

*-0,36-0,36 1,70 0,67

0,02285390,0249450

3024,422924,4500

*24,5500

Expected LT PerformanceExpected ST PerformanceObserved PerformanceOverall (LT) Capability


Process DataSTLT

23,35 23,37 23,39 23,41 23,43 23,45 23,47 23,49

LSL USL

Process Capability Analysis for C1=23,4


CpCPUCPLCpkCpm

PpPPUPPLPpk




23,4500 *

23,350023,4547

300,01021030,0113188

1,63-0,15 3,42-0,15

*

1,47-0,14 3,08-0,14

0,00600000,00600000,00

0,00678526,82678526,82

0,00662093,02662093,02

Process Data



STLT

14,05 14,07 14,09 14,11 14,13 14,15 14,17 14,19 14,21

LSL USL

Process Capability Analysis for D1=14,1


CpCPUCPLCpkCpm

PpPPUPPLPpk




14,1500 *

14,050014,1502

300,01773050,0158137

0,94-0,00 1,88-0,00

*

1,05-0,00 2,11-0,00

0,00366666,67366666,67

0,01503750,00503750,01

0,00504204,53504204,53

Process Data



STLT

9

Tolerance Analysis Was Needed on the CAM

Cam21.15 R.

1

0.00

22.8

0.0023.28.5

28.6

26.94 (spring pad)

24.56

27.55

45.717

45.9

23.649

(18.791)

(45.109)

26.262

10

Statistical Tolerancing

Frequency Chart

mm

.000

.006

.012

.018

.024

0

30.25

60.5

90.75

121

22.150 22.173 22.195 22.218 22.240

5,000 Trials 26 Outliers

Forecast: Open Switch (1b) Postulated

• Statistical Tolerancing was performed on the switch, considering all sources of variation in the supply chain

• The following distribution was calculated

11

Compared to the Specification for an Open

• We were WAY inside the performance parameters required

• We were scrapping supplier material against unrealistic specs

LSL21.90 mm

USL22.50 mm

22.20 mm

μ = 22.197 mm

σ = .016

Z =( 22.197 - 21.90)/0.016 =0.297/.016 =

18.56

12

Which allowed….

• The Company was able to qualify a new supplier for $700k a year savings.

• The team had previously disqualified this supplier for not meeting specs.

– Until the Six Sigma team determined that current suppliers failed to meet the same specs and actually had more variation

– Why? Because the specs did not match customer expectations.

ec timer box product design for six...

Documents