ec timer box product design for six...
TRANSCRIPT
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
PPM < LSLPPM > USLPPM Total
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
USLTargetLSLMeanSample NStDev (ST)StDev (LT)
CpCPUCPLCpkCpm
PpPPUPPLPpk
PPM < LSLPPM > USLPPM Total
PPM < LSLPPM > USLPPM Total
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
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
PPM TotalPPM > USLPPM < LSL
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
Potential (ST) 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
USLTargetLSLMeanSample NStDev (ST)StDev (LT)
CpCPUCPLCpkCpm
PpPPUPPLPpk
PPM < LSLPPM > USLPPM Total
PPM < LSLPPM > USLPPM Total
PPM < LSLPPM > USLPPM Total
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
Potential (ST) Capability
Overall (LT) Capability Observed Performance Expected ST Performance Expected LT Performance
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
USLTargetLSLMeanSample NStDev (ST)StDev (LT)
CpCPUCPLCpkCpm
PpPPUPPLPpk
PPM < LSLPPM > USLPPM Total
PPM < LSLPPM > USLPPM Total
PPM < LSLPPM > USLPPM Total
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
Potential (ST) Capability
Overall (LT) Capability Observed Performance Expected ST Performance Expected LT Performance
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.