6s Project

Measure & Analyze

Define

Improve

Control

Six Sigma Project

Project Number

SEL/QA/2006-001

Plant

SHOWA ENGG.LTD

Name of the Black Belt

Mr.S.DORAISWAMY

Team Members

Mr.C.SUKUMAR
Mr.D.PONNUVEL
Mr.V.VADIVELU

Mr.A.NANDHAKUMAR

Date of Start

03.07.2006

Caliper Housing
Mounting hole less wall thickness

Approved
Component

Rejected
Component
3

Define Phase

Phase 1- Problem Definition

Number of lines/presses/machines used for processing
One Machine

Objective of the Project

To Reduce this defect to Zero

Annual Savings in Rs. Lakhs if the defect is made zero and horizontally
deployed to other part numbers
Rs.1,21,584

Response
Variable / Attribute

Specification (if the response is variable)

PCD Dimension : 159.0 0.075 , Wall thickness: 4.8 mm to 5.8 mm

Is R&R study required

Yes / No

If R&R study is required, % R&R to tolerance

10%
5

Phase 1- Problem Definition

Problem Statement

Pin hole less wall thickness (Up to 4.2 mm as against 4.8mm to 5.8mm)

Part number selected for study

29320155/56

Other similar part numbers having the problem

Nil

Process stages where the Problem is detected

Opn No : 30
Pin hole drilling.

Current average rejection for last 6 months

27 Nos/month 0.11%

Maximum and Minimum rejection in last 6 months

Maximum rejection in a month - 34 Nos - June06
Minimum rejection in a month - 23 Nos - Mar06

Phase 1- Problem Definition

Process Mapping
Input Material
Casting

Opn. 40 - Feed & Bleed

M/C : Stama

Opn. 10 - Disc Milling

M/C : SPM Milling

Opn. 20 - Lug Milling

M/C : Cincinnati
Milling

Opn. 50 - Banjo Milling

M/C : Cincinnati Milling

Opn. 60 - Rough Bore

M/C : Column Drill

Rejection occurring
stage

Opn. 30 - Pin Hole

Drilling
M/C : Systec

Opn. 70 - Finish Bore

M/C :1SC & LMW

Phase 1- Problem Definition

Project Planning
Month

Phase

W1

W2

W3

Month
W4

W1

W2

W3

Month
W4

W1

W2

W3

W4

Define
Measure &
Analyze
Improve
Control

Planned

Completed
Planned
Start date

Planned
Completion
date

Actual start
date

Actual
completion date

Status

Define

03.07.2006

08.07.2006

03.07.2006

07.07.2006

Completed

Measure & Analyze

10.07.2006

05.08.2006

15.07.2006

08.08.2006

Completed

Improve

07.08.2006

26.08.2006

10.08.2006

05.09.2006

Completed

Control

28.08.2006

09.09.2006

08.09.2006

06.10.2006

Completed

Phases

Photograph of defect part

Casting

Pin hole drilling

(Problem operation)

Finished component

Min 4.8mm

Process flow
10

Process flow diagram

Sq.
Process step
no
10

Incoming source of
variation

Symbol

Desired outcomes

Type

Process characteristics

DISC MILLING

1.HARDNESS

1.DISC MILLING RADIUS

BP

1.SPEED

LOADING

2.CASTING DIMENSION

2.RADIUS DEPTH

BP

2.FEED

( TOOLING LOCATION W.R.TO

3.CENTER SHIFT OF RADIUS

BP

3.COOLANT FLOW

CENTER OF CASTING )

4.CLAMPING PRESSURE
5.TABLE MOVEMENT
6.LOCATING PIN IN FIXTURE
7.BUTTING THE CASTING BY
OPERATOR
8.SETTING DIMENSION FOR RADIUS
MILLING,RADIUS DEPTH AND
CENTER SHIFT OF RADIUS

20

LUG MILLING

CENTER SHIFT OF RADIUS

LOADING

1.FLANGE THICK NESS

BP

1.SPEED

2.DIMN. 11.5 MIN. FROM HOLE CENTER

BP

2.FEED

3.DIMN 24.20 / 23.80 MM FROM "V" FACE

BP

3.SPINDLE AXIAL AND RADIAL

PLAY

30

PIN HOLE

1.CASTING DIMENSION

1.PIN HOLE DIAMETER

BP

1.SPEED

DRILLING

( SYMMETRICITY,CAVITY TO

2.PIN HOLE POSITION

BP

2.FEED

LOADING

CAVITY,CENTER SHIFT OF

3.WALL THICKNESS

BP

3.COMPONENT BUTTING LOCATION

RADIUS )

4.CLAMPING PRESSURE

2.RADIUD DEPTH
3.CENTER SHIFT OF RADIUS

11

Process
FMEA
12

Process FMEA
PROCESS
STEP

FAILURE MODE

DISC MILLING 1.DISC MILLING

POTENTIAL CAUSES

1.HARDNESS

OCC

CURRENT PROCESS
CONTROL (TYPE - B)

NIL

CURRENT PROCESS
CONTROL (TYPE - C)
RECEIVING INSPECTION

RADIUS

AS PER SHOWA SINGLE

OVERSIZE

SAMPLING PLAN

UNDERSIZE

( ACCEPTANCE LEVEL 0 )

CURRENT PROCESS
CONTROL (TYPE - A)

DET

RPN

NIL

RESP : INSPECTOR
2.RADIUS DEPTH
OVERSIZE

2.CASTING DIMENSION

NIL

RECEIVING INSPECTION

UNDERSIZE

( TOOLING LOCATION W.R

AS PER SHOWA SINGLE

TO CENTER OF CASTING )

SAMPLING PLAN

3.CENTER SHIFT

( ACCEPTANCE LEVEL 0 )

OF RADIUS

RESP : INSPECTOR

NIL

OVERSIZE
UNDERSIZE

3.BUTTING THE CASTING

NIL

BY OPERATOR

POKA - YOKE

ONCE IN 2 HRS BY QA

PROVIDED

PRE-DESPATCH
INSPECTION
AS PER SHOWA SINGLE
SAMPLING PLAN
( ACCEPTANCE LEVEL 0 )
RESP : INSPECTOR

4.TABLE MOVEMENT

NIL

IT IS IN PM CHECK

NIL

LIST

13

Process FMEA
PROCESS
STEP

FAILURE MODE

POTENTIAL CAUSES

1.CENTER SHIFT OF RADIUS

OCC

CURRENT PROCESS
CONTROL (TYPE - B)

PROCESS CONTROL

CURRENT PROCESS
CONTROL (TYPE - C)

NIL

CURRENT PROCESS
CONTROL (TYPE - A)

LUG

FLANGE

ONCE IN 2 HRS BY QA

MILLING

THICKNESS

CHART

PRE-DESPATCH

VARIATION

CHECK 1 NO/HOUR

INSPECTION

RESP: OPERATOR

AS PER SHOWA SINGLE

DET

RPN

SAMPLING PLAN
( ACCEPTANCE LEVEL 0 )
INSPECTOR

2.SPINDLE AXIAL AND

RADIAL PLAY

NIL

ADDEED IN THE
PM CHECK LIST

14

Process FMEA
PROCESS
STEP

FAILURE MODE

POTENTIAL CAUSES

PIN HOLE

WALL

1.CASTING DIMENSION

DRILLING

THICKNESS

( SYMMETRICITY,

UNDERSIZE

CAVITY TO CAVITY &

OCC

CURRENT PROCESS
CONTROL (TYPE - B)

NIL

CURRENT PROCESS
CONTROL (TYPE - C)
RECEIVING INSPECTION

CURRENT PROCESS
CONTROL (TYPE - A)

DET

NIL

RPN

BOSS DIAMETER )

2.RADIUS DEPTH

PROCESS CONTROL

NIL

100% IN FINAL AREA

CHART

USING GAUGE

CHECK 1 NO/HOUR

ONCE IN 2 HRS BY QA

RESP : INSPECTOR

PRE-DESPATCH

20

INSPECTION
3.CENTER SHIFT OF

PROCESS CONTROL

NIL

AS PER SHOWA SINGLE

RADIUS

CHART

SAMPLING PLAN

CHECK 1 NO/HOUR

( ACCEPTANCE LEVEL 0 )

RESP : INSPECTOR

RESP : INSPECTOR

15

Calibration

16

Variation due to
Instrument
17

Data Collection
1. Instrument Selected : Vernier Caliper
2. Instrument Number

: SM 0200

3. Least Count

: 0.01 mm

4. No. of Samples

: 5 Nos

5. No. of Appraiser

:3

18

Bias Estimation
Following are the readings obtained during calibration of Vernier Caliper
Master Value = 20 mm
Error = 0.00001 mm
S.No

Reading

20.01

20.00

20.00

20.01

20.02

Master Error = 0.01 microns = 0.00001 mm

Average = 20.008
Error = Actual Value Master Value
Error = 20.008 20.000
Error = 0.008
19

Bias Estimation
Standard Deviation ( s ) = 0.008
Standard Deviation of Averages ( s (X bar) ) = s / Sqrt n
= 0.008 / Sqrt 5
= 0.008 / 2.236
= 0.0036

t value for 5 data

= 2.7764

Random Uncertainty

= t * (s (X bar))
= 2.7764 * 0.0036
= 0.009995

20

Bias Estimation
System Uncertainty
Total Uncertainty = Sqrt( (RU)2 + (SU)2 )
= Sqrt( (0.009995)2 + (0.00001)2 )
= 0.0099
Calculating Upper and Lower Boundary
UL = Error + TU

UL = 0.008 + 0.0099
UL = 0.0179
LL = Error - TU
LL = 0.008 - 0.0099
LL = - 0.0019

0 lies between the 2 limits. Hence the Error is 0.

21

Conclusion for Uncertainty

Calculate TU/Tolerance % = 0.0099 / 1 * 100
= 0.99 %
0.99 % is < 25%

So the Uncertainty is Acceptable

Bias is 0
Uncertainty percentage is < 25%
Hence this Instrument can be used for our
project.

22

Linearity Estimation
Method-2 : Graphical Method
Following is the data collected after calibration
Master Slip Gauge

Instrument Vernier Caliper

Master value

20

60

90

120

140

Error

0.00001

0.00000

0.00001

0.00004

0.00003

20.01

60.02

89.99

120.02

140.00

20.00

59.99

90.01

120.00

140.00

20.00

60.02

90.02

120.01

139.98

20.01

60.01

90.01

120.00

139.99

20.02

60.02

90.00

120.00

139.99

Average

20.008

60.012

90.006

120.006

139.992

Error

0.008

0.012

0.006

0.006

- 0.008

All dimensions in mm

23

Graphical Method
0.015

0.010

0.005

Error

0.000
20

40

60

80

100

120

140

160

- 0.005

- 0.010

Master Value
- 0.015

0 lies between the highest and lowest point. So it is Linear

24

Gauge R&R
Study
25

Appraiser

Trial

Sample 1

Sample 2

Sample 3

Sample 4

Sample 5

T1

12.99

13.08

12.96

12.99

12.98

T2

13.07

13.06

12.96

12.96

12.99

T1

12.98

12.94

12.92

12.94

12.94

T2

12.90

13.03

12.96

12.99

12.92

T1

12.97

13.00

12.96

12.99

12.97

T2

12.99

13.02

13.02

12.98

12.99

S.Doraisamy

0.08

0.02

0.00

0.03

0.01

S.Ganesh Kumar

0.08

0.09

0.04

0.05

0.02

V.Vadivelu

0.02

0.02

0.06

0.01

0.02

S.Doraisamy
S.Ganesh
Kumar
V.Vadivelu

Maximum Range = 0.09

0.55
R =
15
R= 0.037
26

To check the Consistency of the range

UCL = D4*R
UCL = 3.267*0.037
UCL = 0.121
LCL = 0
With in person Variation = 0.09
UCL is > With in person Variation
So It is Consistent

27

Estimating Repeatability ( Variation with in person )

Estimating Standard Deviation

= R
d2

0.037
2.48

m=5
g=1
d2 = 2.48

s = 0.015
Estimating Variation
5.15s = 5.15*0.015
= 0.077

At 99% Confidence level

With in person

Actual

Estimation

0.09

0.077

Person to Person Variation + With in person variation

Person to Person Variation only
Total R&R
28

Estimating Person to Person and With in person Variation

Finding out the Range
Mean
S.Doraisamy

13.030

13.070

12.960

12.975

12.985

13.004

S.Ganesh Kumar

12.940

12.985

12.940

12.965

12.930

12.952

V.Vadivelu

12.980

13.010

12.990

12.985

12.980

12.989

Range = 0.052

R= 0.052

Estimating Standard deviation

0.052
=
2.48

s = 0.021

m=5
g=1
d2 = 2.48

29

Estimate Variation
5.15s = 5.15*0.021
= 0.108

At 99% Confidence level

Actual

Estimation

With in person

0.090

0.077

Person to Person Variation + With in person variation

0.052

0.108

Person to Person Variation only

Total R&R

To find Reproducibility (Person to Person Variation) we have to

remove Variation with in person from this

30

Estimating Reproducibility
Person to Person Variation
= Sqrt((AV)2 ((EV)2/(n*r)))
= Sqrt((0.052)2 ((0.090)2/(2*5)))
= 0.044
Estimating Person to Person Variation
= Sqrt((AV)2 ((EV)2/(n*r)))

= Sqrt((0.108)2 ((0.077)2/(2*5)))
= 0.105
Actual

Estimation

With in person

0.090

0.077

Person to Person Variation + With in person variation

0.052

0.108

Person to Person Variation only

0.044

0.105

Total R&R
31

Actual R&R
= Sqrt((Repeatability)2 + (Reproducibility)2)
= Sqrt((0.090)2 + (0.044)2)
= 0.100
Estimating R&R
= Sqrt((Repeatability)2 + (Reproducibility)2)
= Sqrt((0.077)2 + (0.105)2)
= 0.130
Actual

Estimation

With in person

0.090

0.077

Person to Person Variation + With in person variation

0.052

0.108

Person to Person Variation only

0.044

0.105

Total R&R

0.100

0.130

At 99% Confidence level

32

Calculating R&R / Tolerance%

Calculated R&R/Tolerance * 100 = 0.130 / 1*100
= 13 %

Calculating R&R / Part to Part Variation%

Taking s from my project

s = 0.082
5.15s = 5.15*0.082
= 0.4223

R&R/Part to Part Variation = 0.130 / 0.4223

= 30.8 %

Since R&R/Part to Part Variation is > 30%. We cannot use this

Instrument for Data Collection in our Project.
We allowed a deviation to our self to use this for data collection
since the deviation is too small.

33

Phase 1- Problem Definition

Pareto (Based on last 6 months data)
HOUSING PARETO CHART FOR
THE MONTH OF JAN'06 TO JUNE'06 ( SIX MONTHS)
163

120%

160
140

REJECT QTY

120
100

77

80
31%

60
40

21%

67 47%
40%
58

53%

47

59%

65%

70%

75%

80%

85%

90%

94%

98% 100%
100%
80%
60%

45

44

42

41

40

37

36

40%
33

29

28

20%

20

0%

PIN

PIN

HO
LE
LES

SW

ALL
THI
CK
HO
NE
LE
SS
CH
AM
FE R
DIA
O/S
FEE
BO
DP
R
OR
ED
TS
IA O
PO
/S
T
FEE
F
AC
DP
E
OR
DA
TS
MA
.FA
GE
BLE
CE
ED
CH
AT.
PO
MA
RT
RK
SE
ATI
NG
DA
MA
GE
F
EE
BLE
DP
ED
O
RT
PO
SH
RT
IFT
OU
TO
FP
OS
ITIO
LUG
N
MIL
L
I
NG
SE
STE
AL
GR
P
OO
VE
FE E
DIA
DP
O/S
OR
TC
OR
ED
IA O
/S
BO
RE
M1
STE
0T
P
HR
EA
DD
AM
AG
E
BO
RE
DA
M1
MA
0T
GE
HR
EA
DW
AS
HO
UT

CUMULATIVE PERCENTAGE

180

DEFECTS

21% of the Total rejection is due to pin hole less wall

thickness.
As per Pareto Pin Hole Less Wall thickness is the
highest rejection. Hence this defect was taken for the
project.

Qty produced
Total scrap Qty
Scrap Percentage
Pin hole out of pos. scrap
Pin hole less wall thickness %

140782 nos.
787 nos.
0.55 %
163nos.
0.11 %
34

Phase 1- Problem Definition

COPQ (Cost of Poor Quality) Calculation
Number of pieces rejected last month
(for the part number identified for study)

34 ( June06)

Number of pieces scrapped last month

34

Number of pieces reworked last month

Scrap cost/piece
Rework cost/piece

298
_

Total scrap cost (Rs. Lakhs) for last

month

Rs 10,132

Total rework cost(Rs. Lakhs) for last

month

Total Rejection cost (Rs. Lakhs) for last

month
Extrapolated Total rejection cost (Rs.
Lakhs) for one year
Horizontal Deployment

Rs 10,132
Rs : 1,21,584
N.A

35

Phase 1- Problem Definition

Listing Down Suspected Sources of Variations
Cavity to cavity variation
Dia. of the bosses
Symmetricity of the bosses
Radius depth in SPM

Man

Input Material

Loading variation in
pin hole operation

Milling
Center Shift of Radius

Pin hole less wall

thickness

Process (Pin hole drilling)

Clamping variation in Pin hoe operation

Machine

Radial play in Milling &

Systec machines

36

Listing Down all SSVs

SSVs
1.Incoming Material

1.Cavity to Cavity Variation

2.Boss diameter
3.Symmetricity of Boss
4.Radius depth in SPM Milling
5.Center shift of radius

2.Operator

6.Loading Variation in pin hole operation

3.Process

7.Clamping in Pin hole operation

4.Machine

8.Spindle Runout in Milling and Systec Machine

37

The following SSVs are Constant and Hence are not considered for
further study for the Y at this Stage

1.Loading

- Mistake Proof Arrangement

(Both in previous operation & in problem generated operation)

2.Clamping

- Constant

3.Spindle Runout in Systec machine - Constant

38

Phase 1- Problem Definition

Suspected Sources of Variation (SSVs) for the Problem statement
1st level SSVs
1. Input Material
Casting Related
1.Cavity to Cavity Variation
2.Boss diameter
3.Symmetricity of Boss
Previous (Earlier) operation Related
1.Radius Milling depth
2.Center shift of radius

2. SSVs are Nil in the Process (Drilling Operation) at which Y is created

ALL THE ABOVE SSVs ARE VARIABLES

39

Phase 1- Problem Definition

Input Material

Previous Operation
related

SSVs

Current
Specification

Boss diameter

20.0 + 0.60

Cavity number

7,8,9 & 10

Symmetricity

26.50 + 0.60

Radius depth

133.70 + 0.30

Center shift of radius

48.50 / 47.50

Response (Hole ) Y = Wall thickness 4.80 mm to 5.80 mm

40

Problem Statement
20+0.6 (Boss dia)

Cavity Number

4.8 mm to 5.8 mm.

(Wall thickness)

47.5 / 48.5

133.7 +0.3

Boss - A

Boss - B

B with respect to A

Y = Wall thickness
X = Symmetricity,Boss diameter,Cavity to Cavity Variation,Radius depth and center
shift of radius

41

Six Sigma Tools used

Phase

Tools

Define

1. Process Mapping
2. Process FMEA
3. Pareto
4. Calibration
5. Gauge R&R Studies

42

Measure and
Analyze Phase

43

Machines
SPM Milling

Systec

Root Cause = X

Response = Y

Suspected Source Machine

Generating Machine
44

DOE Tool Selection

The identified SSVs are Measurable on both Good and Bad
components and are Input related.
Hence DOE Tool PAIRED COMPARISON was selected.
Y= f(x)
X = SSVs

Y = Response
= Wall Thickness

Input Material Parameters

1.Cavity to Cavity
2.Boss

3.Symmetricity of Boss
4.Radius Depth
5.Center Shift in Radius

45

Measurement

46

Data Collection
8 Good parts (Best of Best) and 8 Bad parts (Worst of Worst)
based on Wall thickness (Response) were selected for study.
WOW parts were collected from already rejected lot as the
current rejection rate for this defect is 0.11% and need to
wait 8 to 9 thousand numbers (Which is 10 days
production) to get 8 Bad components. BOB parts were
collected from a shifts production Quantity.
The parts were marked as B1,B2,B3,B4,B5,B6B7 & B8
W1,W2,W3,W4,W5,W6,W7& W8

47

Data Collection
Earlier Process Related
(Machining Related)

Input Material SSVs (Casting Related)

S.No

Specification

Response

Cavity
No

Boss
Dia - A
200.6mm

Boss
Dia - B
200.6 mm

Symmetricity

Radius Depth
133.70.3mm

Center shift of
Radius

Response

B1

GOOD

21.12

21.38

0.70

133.67

47.82

5.57

B2

GOOD

21.00

20.88

0.55

133.75

48.32

5.58

B3

GOOD

20.98

21.17

0.60

133.62

48.10

6.00

B4

GOOD

20.80

20.80

0.80

133.80

48.00

5.57

B5

GOOD

10

21.90

21.00

0.60

133.72

47.80

5.90

B6

GOOD

10

20.97

20.90

0.50

133.55

48.20

5.80

B7

GOOD

21.80

21.00

0.80

133.66

48.37

5.70

B8

GOOD

20.80

20.80

0.50

133.70

48.20

6.00

W1

BAD

21.37

22.00

1.50

133.90

47.88

4.78

W2

BAD

21.37

21.33

1.35

134.13

47.70

4.70

W3

BAD

20.84

20.90

0.78

133.70

48.17

4.70

W4

BAD

21.30

21.33

1.10

134.21

47.95

4.75

W5

BAD

20.85

20.80

0.70

134.15

47.45

4.79

W6

BAD

10

20.80

20.80

1.12

134.02

47.80

4.78

W7

BAD

20.80

21.00

0.80

133.95

47.50

4.57

W8

BAD

20.88

21.40

0.70

134.07

47.75

4.76
48

Analysis

49

Analysis Step1

To find out the root causes among the chosen SSVs

Parameter - 1

Parameter - 2

S.No

Response

Cavity no

S.No

Response

Boss A

Boss B

GOOD

GOOD

20.88

20.80

GOOD

GOOD

20.80

20.80

GOOD

GOOD

20.80

20.88

GOOD

GOOD

20.90

20.90

GOOD

10

GOOD

20.97

21.00

GOOD

10

GOOD

20.98

21.10

GOOD

GOOD

21.10

21.17

GOOD

GOOD

21.12

21.38

BAD

BAD

20.80

20.80

BAD

BAD

20.80

20..90

BAD

BAD

20.84

21..90

BAD

BAD

20.85

21.00

BAD

BAD

20.88

21.33

BAD

10

BAD

21.30

21.33

BAD

BAD

21.37

21.40

BAD

BAD

21.37

22.40

All Cavity numbers are both in good

and bad components. Hence Cavity
Variation is not the cause for
Variation in Y

In both Boss s both Max. and Min. sizes are

in Bad Components. Hence these dias are
not the reason for Variation in Y
50

Parameter - 3

Analysis Step1
Parameter - 4

Parameter - 5

S.No

Response

Radius
Depth

S.No

Response

Center
Shift

GOOD

133.55

GOOD

47.80

GOOD

133.62

GOOD

47.82

GOOD

133.66

GOOD

48.00

GOOD

133.67

GOOD

48.10

GOOD

133.70

GOOD

48.20

GOOD

133.72

GOOD

48.20

GOOD

133.75

GOOD

48.32

GOOD

133.80

GOOD

48.37

BAD

133.70

BAD

47.45

BAD

133.90

BAD

47.50

BAD

133.95

BAD

47.70

BAD

134.02

BAD

47.75

1.10

BAD

134.07

BAD

47.80

BAD

1.12

BAD

134.13

BAD

47.88

BAD

1.35

BAD

134.15

BAD

48.17

BAD

1.50

BAD

134.21

BAD

47.95

S.No

Response

Symmetricity

GOOD

0.50

GOOD

0.50

GOOD

0.55

GOOD

0.60

GOOD

0.60

GOOD

0.70

GOOD

0.80

GOOD

0.80

BAD

0.70

BAD

0.70

BAD

0.78

BAD

0.80

BAD

Symmetricity Min & Max do not

belong to the same category. Nor
Min or Max are in both the
categories. Hence this is one of
the causes.

Radius Depth Min & Max do not

belong to the same category. Nor
Min or Max are in both the
categories. Hence this is one of
the causes.

Center Shift Min & Max are not in

the same category. Nor Min or
Max are in both the categories.
Hence this is one of the cause.
51

1ST level funneling

CONCLUSION:
From the above, Cavity to Cavity and
Boss Parameter are not the causes for the
response.
That leaves with Symmetricity,Radius
Depth and Centre shift of Radius are potential
causes for the rejection.

1st level funneling Tool

DOE-Paired Comparison

1.Cavity to
Cavity
2.Boss diameter
3.Symmetricity
4.Radius Depth
5.Center Shift

2nd level SSVs

SSVs remains after 1st level funneling
1.Symmetricity
2.Radius Depth
3.Center Shift

1.Symmetricity
2.Radius Depth
3.Center Shift

Proceeding further to find whether they are the root causes

52

Analysis Step2
SSV1 : Symmetricity
Arranging values in Ascending order
Sl. No

Symmetricity

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

0.50
0.50
0.55
0.60
0.60
0.70
0.70
0.70
0.78
0.80
0.80
0.80
1.10
1.12
1.35
1.50

Response

G
G
G
G
G
G
B
B
B
B
G
G
B
B
B
B

Top Line

Top count

= 51/2

Bottom Count = 4
Bottom Line

Total Count

= 91/2

Total Count is > 6.Hence this SSV is one

of the root causes for the response at 95%
confidence level
53

SSV2 : Radius Depth

Sl. No

Radius depth

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

133.55
133.62
133.66
133.67
133.70
133.70
133.72
133.75
133.80
133.90
133.95
134.02
134.07
134.13
134.15
134.21

Analysis Step2

Response

G
G
G
G
G
B
G
G
G
B
B
B
B
B
B
B

Top Line

Bottom Line

Top count

= 41/2

Bottom Count = 7

Total Count

= 111/2

Total Count is > 6.Hence this SSV is one

of the root causes for the response at 95%
Confidence level

54

Analysis Step2
SSV3 : Center shift of radius
Sl. No

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

Radius centre Response

47.45
47.45
47.50
47.70
47.75
47.80
47.80
47.82
47.88
47.90
48.00
48.10
48.17
48.20
48.32
48.37

B
B
B
B
B
G
G
B
G
B
G
B
G
G
G
G

Top Line

Top count

=5

Bottom Count = 4
Bottom Line

Total Count

=9

Total Count is > 6.Hence this SSV is one

of the root causes for the response at 95%
Confidence level
55

Conclusion
SSV1 = Symmetricity

SSV2 = Radius Depth

SSV3 = Center shift of Radius
Are the root causes for the Wall thickness Variation (Response Y)

56

Fixing Specification
Since all the three parameter have count more than 6,New specifications are
arrived at

Parameter

Total Count

New Specification

Symmetricity

91/2

< 0.60

Radius Depth

111/2

133.55 / 133.67

Center Shift of Radius

48.17 / 48.37

New specifications are based on the sizes from good components.

57

Proceeding Further
Can we control the new specification to with in the limits arrived at?
Radius depth and Radius shift are set dimension in the Machine. Hence can be
controlled.
Symmetricity is created at Foundry and they have agreed study to restrict
Symmetricity in 0.60 Max.

58

Analysis
Checking the Contribution of Each Root Cause
for the Response
Since there are 3 Parameters which affect the
Response, I have selected DOE tool

Full Factorial
for checking the contribution.

59

Stage 0:

Factorial Analysis

Identification of (+) Setting and (-) Setting

Parameter
Identified As

Actual Parameter

Unit of
Measurement

(+)
Setting

(-)
Setting

Symmetricity

In mm

< 0.60

> 1.00

Radius Depth

In mm

133.50

134.05

Center Shift

In mm

48.30

47.60

These are the parameters identified as root causes. (+) Settings are those which
produced Good components and (-) Settings are those which produced Bad
components.

60

Stage 1:

Making of Factorial Table

We have 3 Parameters
Hence No. of Rows

= 2n = 2x2x2 = 8

No. of Columns = n-1 = 8-1 = 7

A

Symmetricity

Radius Depth

Center Shift
of Radius

Response

A*B

B*C

A*C

A*B*C

+
61

Stage 1: Data Collection

Specification for Factorial Analysis
+

Symmetricity

< 0.60

> 1.00

Radius Shift

133.50

134.05

Center Shift

48.30

47.60

Second Run
Good

Response

Bad

Response

G1

5.54

B1

4.84

G2

5.43

B2

4.84

G3

5.32

B3

4.61

G4

5.32

B4

4.75

G5

5.28

B5

4.37

G6

5.28

B6

4.40

Mean

5.36

Mean

4.64

Range

0.26

Range

0.47

Third Run

First Run
Good

Response

Bad

Response

Good

Response

Bad

Response

G1

5.88

B1

4.24

G1

5.24

B1

4.82

G2

5.35

B2

4.46

G2

5.17

B2

4.72

G3

5.48

B3

5.20

G3

5.18

B3

4.63

G4

5.58

B4

4.38

G4

5.48

B4

4.79

G5

5.48

B5

4.78

G5

5.38

B5

4.93

G6

5.63

B6

4.56

G6

5.50

B6

4.63

Mean

5.56

Mean

4.60

Mean

5.33

Mean

4.75

Range

0.53

Range

0.96

Range

0.33

Range

0.30
62

S.No

Response

S.No

Response

S.No

Response

S.No

Response

4.24

4.92

4.79

4.70

4.46

4.89

4.94

4.71

5.20

5.26

4.76

4.70

4.38

4.75

4.78

4.69

4.78

5.11

5.06

4.14

4.56

4.71

4.68

4.71

Mean

4.60

Mean

4.94

Mean

4.84

Mean

4.61

Range

0.96

Range

0.55

Range

0.38

Range

0.57

S.No

Response

S.No

Response

S.No

Response

S.No

Response

5.21

5.17

4.97

5.88

5.38

5.50

5.23

5.35

5.50

5.11

4.99

5.48

5.17

5.30

5.00

5.58

5.28

5.27

5.04

5.48

5.24

5.45

4.99

5.63

Mean

5.30

Mean

5.30

Mean

5.03

Mean

5.56

Range

0.33

Range

0.39

Range

0.26

Range

0.53
63

Factorial Analysis
A

Symmetricity

Radius Depth

Center Shift of Radius

Response
Average wall
thickness

4.60

4.94

4.84

4.61

5.30

5.30

5.03

5.56

64

To Check Contribution of each Parameter

Response from Collected Data
A

Symmetricity

Radius
Depth

Center Shift
of Radius

Response

AxB

BxC

AxC

AxBxC

4.60

4.94

4.84

4.61

5.30

5.30

5.03

5.56

Total

+
0.64

0.10

+
2.20

0.04

+
0.08

+
0.420

+
1.10

Contribution
of parameters

+
0.16

0.025

+
0.55

0.01

+
0.02

+
0.105

+
0.275

65

Conclusion
Based on the Factorial Analysis done for the data the Conclusions are given in
the table below

Parameter

Contribution

When the Parameter is moved from- to

+,Response Increases or Decreases

0.160

Increases

0.025

Decreases

0.550

Increases

66

Using Minitab

67

Using Minitab

68

Using Minitab

69

Six Sigma Tools used

Phase

Tools

Measure and
Analyze

1.Paired Comparison
2.Full Factorial

70

Improvement
Phase

71

Validating the
Root Causes

72

Selection of DOE Tool

Since Good and Bad parts Can be created alternately,We selected B Vs C Tool
to Validate the root cause(s)

Deciding B and C Condition for the Project

The Tool used for finding out the root cause is Paired Comparison.
Hence the B condition are those corresponding to Good category and
The C condition are those corresponding to Bad condition.

Symmentricity

Radius Depth

Centre shift of Radius

B Condition

< 0.60

133.55 / 133.67

48.17 / 48.37

C Condition

> 1.00

133.90 / 134.10

47.50 / 47.70

73

Deciding Sample Size

Symmetricity

Center shift of Radius

Radius Depth

S. No

Symmetricity

Response

S. No

Radius Depth

Response

S. No

Center Shift

Response

0.50

133.55

47.45

0.50

133.62

47.45

0.55

133.66

47.50

0.60

133.67

47.70

0.60

133.70

47.75

0.70

133.70

47.80

0.70

133.72

47.80

0.70

133.75

47.82

0.78

133.80

47.88

10

0.80

10

133.90

10

47.90

11

0.80

11

133.95

11

48.00

12

0.80

12

134.02

12

48.10

13

1.10

13

134.07

13

48.17

14

1.12

14

134.13

14

48.20

15

1.35

15

134.15

15

48.32

16

1.50

16

134.21

16

48.37

Top
line

Bottom
line

Top
line

Bottom
line

Top count=51/2

Top count=41/2

Top count=5

Bottom count=4

Bottom count=7

Bottom count=4

Total count=91/2

Total count=111/2

Total count=9

# Since Total Count is less than 16,We need to Machine 6B & 6C

Top
line

Bottom
line

74

Data Collection
Deciding Sample Size

Assumption: There could be some overlap in the responses between

causes
SSVs.

B & C condition as while establishing the root

the Total counts were less then 12 for all the 3

Hence 6B & 6C were selected

Deciding whether 6B & 6C are to be in Nos. or in Batches

Since the response (Wall thickness) is Variable 6B & 6C will be in Nos.

Deciding what big Y will be Monitored
The big Y is Wall thickness (Response) and is Variable.
This will be Monitored in Actual Values
Deciding Data collection Methodology
Decided to machine 6B and 6C by creating B and C condition alternatively
75

Specifications for Current & Better

condition

Characteristic

Better condition
(New specification)

Current condition

Symmentricity

< 0.60

> 1.00

Radius Depth

133.55/ 133.67

133.90 / 134.10

Centre shift of Radius

48.17 / 48.37

47.50 / 47.70

76

Data Collection
Example No

Response Y by
B Condition

Response Y by
C Condition

5.88

4.24

5.35

4.46

5.48

5.20

5.58

4.38

5.48

4.78

5.63

4.56

Y = 4.80mm to 5.80mm

Note :
1.

During drilling operation the B Condition & C Condition were alternated for each piece

2.

Since there is no overlap, the root causes are correct

77

Analysis for 6B and 6C

Finding Whether the cause is correct (Another Way)
S.No

Condition

Response Y

4.24

4.38

4.46

4.56

4.78

5.20

5.35

5.48

5.48

10

5.58

11

5.63

Bottom Count = 6

12

5.88

Total Count

Top Count

=6
=12

Since Total Count is >6,the root causes are correct.

78

Analysis
Quantifying the amount of Improvement
The Average of B = Xb = 5.57 mm
The Average of C = Xc = 4.60 mm
The Difference between Xb & Xc = Xb-Xc = 0.97 mm
Sigma value of B Values = Sigma B = 0.1813
K Value at 95% CL for 6,6 = 2.96
K * Sigma = 2.96 * 0.1813 = 0.537
( Xb Xc ) >= K * Sigma
0.97 > 0.537
Since Xb Xc is greater K * Sigma.
Improvement has taken place at the assumed C.L of 95%
The Amount of Improvement = Xb - Xc
= 0.97 mm
79

Of the 3 root causes

Symmetricity is Foundry related dimension
Radius Depth and Center shift of radius are machining related
dimension. They are set and locked.
During the middle of the Project stage we discussed with Foundry for
maintaining to new specification of Symmetricity to 0.60 max. They said they
need a tolerance of 1.2 mm max. for their process and they will keep trying to
bring it to 0.60 max.
Hence we proceeded with the project with Symmetricity up to 1.20 mm.
6 Nos were produced with Symmetricity 1.00 to 1.20 mm and the other two
dimensions to New specification.The response is as follows.

80

BOB and WOW Parts specification

Characteristic

Better condition

Current condition

Symmentricity A

< 0.60

1.00 to 1.20

Radius Depth B

133.55

133.55

Centre shift of Radius - C

48.30

48.30

81

Data Collection
Example No

Response Y by
B Condition

Response Y by C Condition with B

Condition data except Symmetricity

5.88

4.97

5.35

5.23

5.48

4.99

5.58

5.00

5.48

5.04

5.63

4.99

Y = 4.80mm to 5.80mm

Note :
1.

Since there is no overlap, the root causes are correct

82

Quantifying the amount of Improvement

The Average of B = Xb = 5.57 mm
The Average of C = Xc = 5.04 mm
The Difference between Xb & Xc = Xb-Xc = 0.53 mm
Sigma value of B Values = Sigma B = 0.1813
K Value at 95% CL for 6,6 = 2.96
K * Sigma = 2.96 * 0.1813 = 0.54

( Xb Xc ) >= K * Sigma
0.53 < 0.54
Since Xb Xc is Less than K * Sigma.

Improvement has not taken place at the assumed C.L of 95%

83

Quantifying the amount of Improvement

The Average of B = Xb = 5.57 mm
The Average of C = Xc = 5.04 mm
The Difference between Xb & Xc = Xb-Xc = 0.53 mm
Sigma value of B Values = Sigma B = 0.1813
K Value at 90% CL for 6,6 = 2.61
K * Sigma = 2.61 * 0.1813 = 0.473

( Xb Xc ) >= K * Sigma
0.53 > 0.473
Since Xb Xc is More than K * Sigma.

Improvement has taken place at the assumed C.L of 90%

The Amount of Improvement = Xb - Xc
= 0.53 mm
84

Animation
85

86

Six Sigma Tools used

Phase

Tools

Improve

1. B Vs C

87

Control Phase

88

We have Introduced the following Controls for

the SSVs
SSV 1

Symmetricity : Foundry has agreed to study the process and reduce

the variation in Symmetricity to < 0.60. Till then they
will supply castings up to 1.20mm Symmetricity.
However,to ensure the coming castings whether they
are received with in 1.20mm Symmetricity the
following check has been introduced.
For every batch 30nos are taken at random and
Symmetricity is measured and the lot is accepted if
6s level is < = 75% of Tolerance.

89

Symmetricity Variation
Incoming material
checked on : 27.09.06

S.No

Symmetricity
in mm

S.No

Symmetricity
in mm

0.25

16

0.10

0.30

17

0.30

0.20

18

0.15

0.25

19

0.18

0.10

20

0.43

0.05

21

0.45

0.15

22

0.40

0.20

23

0.45

0.20

24

0.40

10

0.24

25

0.10

11

0.25

26

0.20

12

0.25

27

0.50

s = 0.1364

13

0.05

28

0.52

14

0.45

6s = 0.8181

29

0.30

15

0.20

30

0.10

Casting Tolerance = 1.20 mm

6s level is lesser than 75% of the Tolerance

Therefore the lot is Accepted

90

Symmetricity Variation
Incoming material
checked on : 29.09.06

S.No

Symmetricity
in mm

S.No

Symmetricity
in mm

0.70

16

0.07

0.95

17

0.08

0.45

18

0.25

0.80

19

0.70

0.30

20

0.80

0.60

21

0.91

1.42

22

0.76

0.40

23

0.53

0.15

24

0.57

10

0.83

25

0.53

11

0.65

26

0.95

12

0.60

27

0.05

s = 0.3221

13

0.65

28

0.37

14

0.40

6s = 1.9327

29

0.07

15

0.25

30

0.80

Casting Tolerance = 1.20 mm

6s level is greater than 75% of the Tolerance

Therefore the lot is Rejected

91

SSV 2 & 3

Radius Depth and Center Shift of Radius :

These are Setting dimensions in Milling Machine.Hence every
time when setting takes place or tool is changed, the setting will
be controlled based on response . The parameters are locked in
the system and operator cannot change. This can vary only due to
Wear in the Machine parts and Fixture parts. Periodical
Inspection and change of Wear parts have been introduced.
Further a Checking gauge which can give measured value for
SSVs has been provided to check the parameter at periodical
intervals.Currently the parameters are checked once in a day and
whenever fresh setting is done and cutter changed.

92

Radius checking Gauge with Master

Gauge No : S GV 7018

This Gauge is added in the

Control plan no 29320155/6
Page no: 3 of 52

93

Q Parts in the Fixture

94

Check List for Q parts in fixtures

Cell Name : Housing

Checked on : 17.08.06

Machine : SPM
Operation : Milling

Part Name

Characteristics

Specification
In mm

Actual
Dimensions

Frequency
in Months

Results

Action
Taken

Q1

R.H Side Sliding block

Working
Condition of
retainer spring

Spring load
1kg on
44/46mm
length

1kg on
44mm
length

OK

---

Q2

L.H Side butting pin 1

Height

26.50 0.01

26.505

OK

---

Q3

L.H Side butting pin 1

Height

25.50 0.01

25.504

OK

---

Q4

Bottom butting pin 1

Height

17.00 0.01

17.005

OK

---

Q5

Bottom butting pin 2

Height

23.96 0.01

23.962

OK

---

Q6

Bottom butting pin 3

Height

23.96 0.01

23.959

OK

---

Q7

Bottom butting pin 3

Height

10.50 0.01

10.505

OK

---

S.No

95

Check List for Q parts in Machine

Cell Name : Housing

Checked on : 04.06.06

Machine : SPM
Operation : Milling
S.No

Part Name

Specification
In mm

Actual
Dimensions

Frequency
in Months

Results

Action
Taken

Bearing

Worn-out

---

12

OK

---

Spindle axial play

0.02 0.03 mm

0.022 mm

12

OK

---

Spindle Radial play

0.02 0.03 mm

0.027mm

12

OK

---

X Axis
Repeatability

0 - 0.02 mm

0.017 mm

12

OK

---

Z Axis
Repeatability

0 - 0.02 mm

0.010 mm

12

OK

---

96

Variation Analysis for Root

Cause and Response

97

Step 1: Identifying Parameters

Radius Depth and Center Shift of Radius are the two root cause
parameters identified for Variation Analysis as they are
1.

Measurable and Variable

2.

These root causes are generated from the milling process and
not functional parameters.

3.

These are generated from discrete production process and not

from a batch process.

4.

Can be measured with Non-destructive method.

5.

We can detect the parameter immediately if something goes

wrong immediately after generation.

6.

Apart from the root causes,Variation Analysis for Response (Y)

I.e., Wall thickness was also decided to be done to find out
whether the estimated Variation is < = 50% i.e.,6s level

Step 2: Identifying Product / Process Stream

In this case only one Stream is there.
98

Step 3: Deciding Sample Size

The following were decided and followed
1.

Sample size

- 3 Nos

2.

Time block chosen - 25

3.

Period of Coverage - 4 Hrs

4.

Planned events between the Time blocks were captured.

5.

Randomized Time intervals were followed.

6.

Time blocks were closer because Cycle time is < 3 mins

(Actual is 56 Secs)

Step 4: Collection of Data

Instrument and Dial Gauge with Accuracy which is less than 10%
of Tolerance is used.

Step 5: Multi Vari Analysis

Not done as products are not produced from different Streams
99

Variation Analysis for

Root Causes (X)

100

Variation Analysis for Root Cause (X)

Part Details
Characteristic

Radius
Depth

Part Number

29320155/6

Gauge Number

S GV 7018

Unit of
Measurement

MM

Part
Description

Housing
Machining

Gauge
Description

Radius
Gauge

Date

08.02.06

Gauge Least
Count

0.01 mm

Gauge R&R
Value

7.98

Target Value
Tolerance

0.20 mm

USL

133.600

Study Dates
& Shift

LSL

133.400

Any other
details

06.07.06
1st Shift

----

Data Grouping and Sample details

Number of Groups (Number of Time blocks
x number of Streams)

26

Number of Samples in each group (It is

preferable to collect 5 samples continuously
from the process so that the inherent
variations are captured)

3
101

Study taken to check the variation in the Process at randomized time blocks in SPM
Milling for Radius Depth and Center Shift
Actual time=1:30
T1
S.No Response
1
0.00
2
-0.01
3
0.00

Actual time=1:35
T2
S.No Response
4
-0.01
5
0.03
6
-0.01

Actual time=1:44
T3
S.No Response
7
0.00
8
0.01
9
0.01

Actual time=2:00
T4
S.No Response
10
-0.01
11
0.01
12
-0.01

Actual time=2:10
T5
S.No
Response
13
-0.02
14
0.01
15
-0.01

Actual time=2:20
T6
S.No
Response
16
-0.01
17
-0.03
18
0.00

Actual time=2:28
T7
S.No Response
19
0.01
20
0.00
21
0.01

Actual time=2:36
T8
S.No
Response
22
-0.03
23
0.00
24
-0.02

Actual time=2:45
T9
S.No Response
25
-0.02
26
-0.02
27
0.02

Actual time=3:05
T10
S.No Response
28
-0.02
29
0.00
30
0.00

Actual time=4:02
T11
S.No Response
31
0.00
32
0.00
33
0.02

Actual time=4:11
T12
S.No Response
34
0.03
35
-0.04
36
0.02

Actual time=4:25
T13
S.No
Response
37
-0.03
38
0.00
39
-0.01

Actual time=4:38
T14
S.No
Response
40
-0.03
41
0.00
42
-0.01

Actual time=4:50
T15
S.No Response
43
-0.01
44
-0.03
45
-0.02

Actual time=4:55
T16
S.No
Response
46
-0.04
47
-0.05
48
-0.02

Actual time=5:13
T17
S.No Response
49
0.00
50
0.01
51
-0.01

Actual time=5:20
T18
S.No Response
52
-0.01
53
0.00
54
-0.02

Actual time=5:28
T19
S.No Response
55
-0.01
56
0.00
57
-0.06

Actual time=5:48
T20
S.No Response
58
-0.08
59
-0.03
60
-0.03

Actual time=5:54
T21
S.No
Response
61
-0.03
62
-0.03
63
-0.03

Actual time=6:00
T22
S.No
Response
64
-0.04
65
-0.07
66
-0.04

Actual time=6:07
T23
S.No Response
67
0
68
-0.04
69
-0.03

Actual time=6:17
T24
S.No
Response
70
-0.04
71
-0.06
72
-0.04

Actual time=6:24
T25
S.No Response
73
-0.01
74
0.00
75
0.00

Actual time=6:30
T26
S.No Response
76
0.00
77
-0.02
78
-0.02

Time between Time blocks in Minutes

1)

9)

20

17)

2)

10)

57

18)

22

3)

16

11)

19)

4)

10

12)

14

20)

5)

10

13)

13

21)

10

6)

14)

12

22)

7)

15)

23)

10

8)

16)

19

24)

25)

102

Study taken to check the variation in the Process at randomized time blocks in SPM
Milling for Radius Depth and Center Shift
T1
S.NoResponse
0.00
1
-0.01
2
0.00
3
Ave -0.003
R
0.01

T2
S.No Response
-0.01
4
0.03
5
-0.01
6
Ave
0.003
R
0.04

T3
S.No Response
0.00
7
0.01
8
0.01
9
Ave
0.007
R
0.01

T4
S.No Response
-0.01
10
0.01
11
-0.01
12
Ave -0.003
R
0.02

T5
S.No Response
-0.02
13
0.01
14
-0.01
15
Ave -0.007
R
0.03

T6
S.No Response
-0.01
16
-0.03
17
0.00
18
Ave -0.013
R
0.03

T7
S.No Response
0.01
19
0.00
20
0.01
21
Ave 0.007
R
0.01

T8
S.No Response
-0.03
22
0.00
23
-0.02
24
Ave -0.017
R
0.03

T9
S.NoResponse
-0.02
25
-0.02
26
0.02
27
Ave -0.007
R
0.04

T10
S.No Response
-0.02
28
0.00
29
0.00
30
Ave -0.007
R
0.02

T11
S.No Response
0.00
31
0.00
32
0.02
33
Ave
0.007
R
0.02

T12
S.No Response
0.03
34
-0.04
35
0.02
36
Ave
0.003
R
0.07

T13
S.No Response
-0.03
37
0.00
38
-0.01
39
Ave -0.013
R
0.03

T14
S.No Response
-0.03
40
0.00
41
-0.01
42
Ave -0.013
R
0.03

T15
S.No Response
-0.01
43
-0.03
44
-0.02
45
Ave -0.020
R
0.02

T16
S.No Response
-0.04
46
-0.05
47
-0.02
48
Ave -0.037
R
0.03

T17
S.NoResponse
0.00
49
0.01
50
-0.01
51
Ave 0.000
R
0.02

T18
S.No Response
-0.01
52
0.00
53
-0.02
54
Ave -0.010
R
0.02

T19
S.No Response
-0.01
55
0.00
56
-0.06
57
Ave -0.023
R
0.06

T20
S.No Response
-0.08
58
-0.03
59
-0.03
60
Ave -0.047
R
0.05

T21
S.No Response
-0.03
61
-0.03
62
-0.03
63
Ave -0.030
R
0.00

T22
S.No Response
-0.04
64
-0.07
65
-0.04
66
Ave -0.050
R
0.03

T23
S.No Response
0.00
67
-0.04
68
-0.03
69
Ave -0.023
R
0.04

T24
S.No Response
-0.04
70
-0.06
71
-0.04
72
Ave -0.047
R
0.02

T25
S.NoResponse

T26
S.No Response

73
74
75
Ave
R

-0.01
0.00
0.00

-0.003
0.01

76
77

-0.02

78
Ave
R

-0.013
0.02

0.00

# Part to Part Variation = Max. Value of R = 0.07 mm

# Time to Time Variation = Max. Value of Ave. Min. Value of Ave. = 0.057 mm

-0.02

The tolerance is 0.20 mm. Variation 0.07 mm is with in 75% of the Tolerance
103

Calculating Upper Control Limit

R (Average of all Ranges) = 0.027 mm
(Rounded off to one decimal more than the data)

UCL = D4 x R

= 2.575 x 0.027 (D4 Value taken from table for 3 Sample Size)

= 0.0695 mm
Rounding off to the same decimal of data = 0.07 mm

104

Step 7: Checking the Consistency of Part to Part Variation

R Chart for Response

0.08
UCL=0.07030

Sample Range

0.07
0.06
0.05
0.04
0.03

R=0.02731

0.02
0.01
0.00

LCL=0
0

10

20

Sample Number
105

Checking the Consistency of Part to Part

Variation
1.

All points lie with in the Limits

2.

The Stratification level is > 3

3.

There are no 7 points continuously increasing

4.

There are no 7 points continuously decreasing

5.

There are no 7 points in any of the 2 zones

Therefore Part to Part Variation is Consistent

106

Consistency of Part to Part Variation

Checking the Consistency of Part to Part Variation (Step 7)
Average range (R-Bar) (Round off to one
decimal more than the data)

0.027

Upper control limit (UCL) = D4*R-Bar

(Round off to the same decimals as data)

0.07

Lower control limit (LCL) = D3*R-Bar

Is the Part to Part Variation Consistent

Samples

D3

D4

3.267

2.575

2.282

2.115

2.004

YES / NO

If the Part to Part variation is consistent, STOP, do not proceed further. Plan for
DOE
Is the range chart plotted

YES / NO

Are the stratification level more than 3

YES / NO

If the stratification level<=3, Then the Part to Part Variation is very less and
parameter does not require any monitoring. STOP. Do not proceed further
Are there 7 consecutive points
Increasing, Decreasing and one side
of mean range

YES / NO

If yes write causes

(If possible)

--107

Step 9: Estimating Part to Part Variation

s = Estimated Part to Part Variation

s = R/d
2

(d2 is a Constant and depends on Sample size.d2 for 3

samples is 1.693)

= 0.027 / 1.693
= 0.0159

(To be rounded off to one decimal more than the data)

= 0.016

6s = 0.096
Estimated Part to Part Variation Confidential level = 99.73%
108

Step 10 : Normality Check

Frequency

20

10

0
-0.10

-0.05

0.00

Response
109

Step 11 : Estimating Rejection Percentage

Average

0.027

Z (usl)

4.563

0.016

Z (lsl)

7.938

6s

0.096

Estimated Rejection
above USL

USL

0.100

Estimated Rejection
above LSL

LSL

-0.100

Cpk (usl)

1.521

Cpk (lsl)

2.646

110

Variation Analysis for

Response to arrive at Control
method for Y

111

Variation Analysis for

Response (Y)

112

Variation Analysis for Response (Y)

Part Details
Characteristic

Wall
thickness

Part Number

29320155/6

Gauge Number

SM 200

Unit of
Measurement

MM

Part
Description

Housing
Machining

Gauge
Description

Vernier
Caliper

Date

07.07.06

Gauge Least
Count

0.01 mm

Gauge R&R
Value

13%

Target Value
Tolerance

1.00 mm

USL

5.80 mm

Study Dates
& Shift

LSL

4.80 mm

Any other
details

07.07.06
1st Shift

----

Data Grouping and Sample details

Number of Groups (Number of Time blocks
x number of Streams)

26

Number of Samples in each group (It is

preferable to collect 5 samples continuously
from the process so that the inherent
variations are captured)

3
113

Study taken on Response at randomized time blocks

Actual time=3:05

Actual time=3:17

Actual time=3:26

Actual time=3:38

Actual time=3:48

Actual time=4:04

Actual time=4:11

Actual time=4:21

T1
S.No Response
1
5.10
2
5.10
3
5.16

T2
S.No Response
4
5.03
5
5.24
6
5.15

T3
S.No Response
7
5.14
8
5.22
9
5.22

S.No
10
11
12

T4
Response
5.24
5.31
5.24

T5
S.No Response
13
5.21
14
5.19
15
5.18

S.No
16
17
18

T6
Response
5.14
5.14
5.19

T7
Response
5.11
5.19
5.20

T8
Response
5.14
5.17
5.16

Actual time=4:28

Actual time=4:35

Actual time=4:45

Actual time=4:51

Actual time=5:00

Actual time=5:08

Actual time=5:20

Actual time=5:33

T9
Response
5.30
5.23
5.02

T10
Response
5.25
5.13
5.05

T11
Response
5.06
5.21
5.18

T12
Response
5.10
5.23
5.14

T13
Response
5.23
5.14
5.08

T14
Response
5.14
5.19
5.10

T15
Response
5.18
5.12
5.28

T16
Response
5.28
5.40
5.06

S.No
25
26
27

S.No
28
29
30

S.No
31
32
33

S.No
34
35
36

S.No
37
38
39

S.No
40
41
42

S.No
19
20
21

S.No
43
44
45

S.No
22
23
24

S.No
46
47
48

Actual time=5:43

Actual time=5:56

Actual time=6:10

Actual time=6:20

Actual time=6:31

Actual time=6:38

Actual time=6:48

Actual time=6:56

T17
Response
5.21
4.92
5.18

T18
Response
5.17
5.15
4.93

T19
Response
5.03
5.00
5.09

T20
Response
4.97
5.03
5.09

T21
Response
5.03
5.10
5.35

T22
Response
5.08
5.17
5.10

T23
Response
5.24
5.11
5.21

T24
Response
5.04
5.09
5.02

S.No
49
50
51

S.No
52
53
54

S.No
55
56
57

S.No
58
59
60

S.No
61
62
63

S.No
64
65
66

S.No
67
68
69

S.No
70
71
72

Time between Time blocks in Minutes

Actual time=7:02

Actual time=7:12

T25
Response
5.07
5.04
5.08

T26
Response
5.10
5.11
5.01

S.No
73
74
75

S.No
76
77
78

1)

12

8)

15)

13

22)

10

2)

9)

16)

10

23)

3)

12

10)

10

17)

13

24)

4)

10

11)

18)

14

25)

10

5)

16

12)

19)

10

6)

13)

20)

11

7)

10

14)

12

21)

Total = 247 mins = 4Hrs 7Mins

114

Randomised time block

5.5

5.4

Response "Y"

5.3

5.2

5.1

5.0
T1

T2

4.9
Just before
event

4.8
5

10 15 20 25

T3
Event
Time

T4

T5

T6

T7

T8

T9

Just after
event

30 35 40 45 50

55 60 65

70 75 80

85 90

95 100

Time interval in mins

115

Step 6: Checking whether the Part to Part is highest

T1
S.No Response
1
5.10
2
5.10
3
5.16
Ave
5.12
R
0.06

T2
S.No Response
4
5.03
5
5.24
6
5.15
Ave
5.14
R
0.21

T3
S.No Response
7
5.14
8
5.22
9
5.22
Ave
5.19
R
0.08

T4
S.No Response
10
5.24
11
5.31
12
5.24
Ave
5.26
R
0.07

T5
S.No Response
13
5.21
14
5.19
15
5.18
Ave
5.19
R
0.03

T6
S.No Response
16
5.14
17
5.14
18
5.19
Ave
5.16
R
0.05

T7
S.No Response
19
5.11
20
5.19
21
5.20
Ave
5.17
R
0.09

T8
S.No Response
22
5.14
23
5.17
24
5.16
Ave
5.16
R
0.03

T9
S.No Response
25
5.30
26
5.23
27
5.02
Ave
5.18
R
0.28

T10
S.No Response
28
5.25
29
5.13
30
5.05
Ave
5.14
R
0.20

T11
S.No Response
31
5.06
32
5.21
33
5.18
Ave
5.15
R
0.15

T12
S.No Response
34
5.10
35
5.23
36
5.14
Ave
5.16
R
0.13

T13
S.No Response
37
5.23
38
5.14
39
5.08
Ave
5.15
R
0.15

T14
S.No Response
40
5.14
41
5.19
42
5.10
Ave
5.14
R
0.09

T15
S.No Response
43
5.18
44
5.12
45
5.28
Ave
5.19
R
0.16

T16
S.No Response
46
5.28
47
5.40
48
5.06
Ave
5.25
R
0.34

T17
S.No Response
49
5.21
50
4.92
51
5.18
Ave
5.10
R
0.29

T18
S.No Response
52
5.17
53
5.15
54
4.93
Ave
5.08
R
0.24

T19
S.No Response
55
5.03
56
5.00
57
5.09
Ave
5.04
R
0.09

T20
S.No Response
58
4.97
59
5.03
60
5.09
Ave
5.03
R
0.12

T21
S.No Response
61
5.03
62
5.10
63
5.35
Ave
5.16
R
0.32

T22
S.No Response
64
5.08
65
5.17
66
5.10
Ave
5.12
R
0.09

T23
S.No Response
67
5.24
68
5.11
69
5.21
Ave
5.19
R
0.13

T24
S.No Response
70
5.04
71
5.09
72
5.02
Ave
5.05
R
0.07

T25

T26

S.No Response
73
5.07
74
5.04
75
5.08
Ave
5.06
R
0.04

S.No Response
76
5.1
77
5.11
78
5.01
Ave
5.07
R
0.10

# Part to Part Variation = Max. Value of R = 0.34

# Time to Time Variation = Max. Value of Ave. Min. Value of Ave. = 0.23

Part to Part variation is highest and hence we proceeded further

116

Calculating Upper Control Limit

R (Average of all Ranges) = 0.139 mm
(Rounded off to one decimal more than the data)

UCL = D4 x R

= 2.575 x 0.139 (D4 Value taken from table for 3 Sample Size)

= 0.3579
Rounding off to the same decimal of data = 0.36 mm

117

Step 7: Checking the Consistency of Part to Part

Variation
R Chart for response
0.4

Sample Range

UCL=0.3574
0.3

0.2
R=0.1388
0.1

0.0

LCL=0
0

10

20

Sample Number
118

Checking the Consistency of Part to Part Variation

1.

All points lie with in the Limits

2.

The Stratification level is > 3

3.

There are no 7 points continuously increasing

4.

There are no 7 points continuously decreasing

5.

There are no 7 points in any of the 2 zones

Therefore Part to Part Variation is Consistent

119

Consistency of Part to Part Variation

Checking the Consistency of Part to Part Variation (Step 7)
Average range (R-Bar) (Round off to one
decimal more than the data)

0.139

Upper control limit (UCL) = D4*R-Bar

(Round off to the same decimals as data)

0.36

Lower control limit (LCL) = D3*R-Bar

Is the Part to Part Variation Consistent

Samples

D3

D4

3.267

2.575

2.282

2.115

2.004

YES / NO

If the Part to Part variation is consistent, STOP, do not proceed further. Plan for
DOE
Is the range chart plotted

YES / NO

Are the stratification level more than 3

YES / NO

If the stratification level<=3, Then the Part to Part Variation is very less and
parameter does not require any monitoring. STOP. Do not proceed further
Are there 7 consecutive points
Increasing, Decreasing and one side
of mean range

YES / NO

If yes write causes

(If possible)

--120

Step 9: Estimating Part to Part Variation

s = Estimated Part to Part Variation

s = R/d
2

(d2 is a Constant and depends on Sample size.d2 for 3

samples is 1.693)

= 0.139 / 1.693
= 0.0821

(To be rounded off to one decimal more than the data)

= 0.082

6s = 0.492
Estimated Part to Part Variation Confidential level = 99.73%
121

Step 10 : Normality Check

Frequency

15

10

0
4.90 4.95 5.00 5.05 5.10 5.15 5.20 5.25 5.30 5.35 5.40

response
122

Step 11 : Estimating Rejection Percentage

Average

5.141

Z (usl)

8.037

0.082

Z (lsl)

4.159

6s

0.492

Estimated Rejection
above USL

USL

5.800

Estimated Rejection
above LSL

LSL

4.800

Cpk (usl)

2.679

Cpk (lsl)

1.390

123

Step 12 : Decision on the type of Monitoring

6s Analysis
In our case Tolerance = 1.00 mm

6s = 0.492
6s < Tolerance
Actual Part to Part Variation = 0.34 mm
This is less than 50% of the Tolerance
Hence we decided to use Pre control Limit chart

124

Step 14 : On line Monitoring

Pre control limit for On job
Hour

Wall thickness (Response)

1st Job

2nd Job

5.33

5.11

5.41

5.61

5.28

5.39

5.14

5.26

5.09

5.06

5.45

5.51

5.12

5.25

5.47

5.21

125

Step 14 : On line Monitoring

Pre control limit for On job
USL

5.80
5.75
5.70
5.65
5.60

UCL

Wall Thickness (Response) in mm

5.55
5.50
5.45
5.40
5.35

5.30
5.25
5.20

Time interval

5.15
5.10
5.05

LCL

5.000
4.95
4.90
4.85
4.80

LSL

126

Controls
1. Drawing updated for new specification
2. Control plan updated
3. FMEA updated
4. The new parameters have been programmed and
locked in the system
5. A Setting Gauge has been introduced in milling
operation to check the new parameter settings at
periodical intervals.This has been incorporated in
Control plan and FMEA

127

Control plan
128

Control Plan
Characteristics

Product

DISC

Process

Special
Character Product/Process Evaluation/
istic class
Specification
M easurement
Tolerance
Technique

M ethods
Sample
Size

Frequency

Control
method

Reaction Plan
Resp.

Recording (Y/N)

1.RADIUS DEPTH

133.7 0.30 MM

CMM

PER DAY

IN SPECTIONIN SPECTOR

YES

HOLD,REIN SPECT

2.CEN TER SHIFT OF

47.50 / 48.50 MM

CMM

PER AY

IN SPECTIONIN SPECTOR

YES

HOLD,REIN SPECT

MILLIN G

RADIUS

3.CASTING

LEVER DIAL

DIMENSION

AND FIXTURE

100%

INSPECTION QA

NO

SEGREGATE AND
KEEP

( TOOLING
LOCATION W.R.TO
CENTER OF

CASTING )

4.HARDNESS

165 - 250 BHN

HARDNESS

AS PER

MEASURING

SHOWA

INSTRUMENT

SINGLE

INSPECTION QA

NO

SEGREGATE AND
KEEP

SAMPLING
PLAN

4.BUTTING OF

VISUAL

100%

INSPECTION OPERATOR NO

COMPONENT

SEGREGATE AND
KEEP

5.CLAMPING

PRESSURE

PRESSURE

GAUGE

ONCE

PER DAY

CHECKING

OPERATOR YES

INFORM
MAINTENANCE

129

Control Plan
Characteristics

Product

Process

LUG

1.FLANGE

MILLING

THICKNESS

2.POSITION FROM

Special
Character Product/Process Evaluation/
istic class
Specification M easurement
Tolerance
Technique
13.15 / 12.85 MM

VERNIER

M ethods
Sample
Size

Frequency

Control
method

Reaction Plan
Resp.

Recording (Y/N)

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REINSPECT

11 GG 5092

11.5 MM

VERNIER

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REINSPECT

24.20 / 23.80 MM

HEIGHT GAUGE

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REINSPECT

HOLE CENTER

3.LUG FACE TO
"V" FACE

4.CENTER SHIFT OF

S GD 7091

47.50 / 48.50 MM

S GV 7018 GAUGE

KEEP

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REINSPECT

RADIUS

130

Control Plan
Characteristics

Product

Process

WALL

Special
Character Product/Process Evaluation/
istic class
Specification
M easurement
Tolerance
Technique
4.80 / 5.80 MM

M ethods
Sample
Size

Frequency

Control
method

Reaction Plan
Resp.

Recording (Y/N)

VERNIER

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REINSPECT

1.RADIUS DEPTH

CMM

PER DA

INSPECTIONOPERATOR

YES

HOLD,REINSPECT

2.CENTER SHIFT OF

CMM

PER DAY

INSPECTIONOPERATOR

YES

HOLD,REINSPECT

INSPECTION QA

NO

SEGREGATE AND

THICKNESS

RADIUS

3.CASTING

LEVER DIAL

DIMENSION

AND FIXTURE

100%

KEEP

( SYMMENTRICITY )

5.CLAMPING

PRESSURE

PRESSURE

GAUGE

ONCE

PER DAY

CHECKING

OPERATOR

YES

INFORM
MAINTENANCE

131

Control plan &

Process FMEA
after 6 Sigma
132

Process FMEA after 6 Sigma

PROCESS
STEP

FAILURE MODE

DISC MILLING 1.DISC MILLING

POTENTIAL CAUSES

1.HARDNESS

OCC

CURRENT PROCESS
CONTROL (TYPE - B)

NIL

CURRENT PROCESS
CONTROL (TYPE - C)
RECEIVING INSPECTION

RADIUS

AS PER SHOWA SINGLE

OVERSIZE

SAMPLING PLAN

UNDERSIZE

( ACCEPTANCE LEVEL 0 )

CURRENT PROCESS
CONTROL (TYPE - A)

DET

RPN

NIL

RESP : INSPECTOR
2.RADIUS DEPTH
OVERSIZE

2.CASTING DIMENSION

NIL

RECEIVING INSPECTION

UNDERSIZE

( TOOLING LOCATION W.R

AS PER SHOWA SINGLE

TO CENTER OF CASTING )

SAMPLING PLAN

3.CENTER SHIFT

( ACCEPTANCE LEVEL 0 )

OF RADIUS

RESP : INSPECTOR

NIL

OVERSIZE
UNDERSIZE

3.BUTTING THE CASTING

NIL

BY OPERATOR

POKA - YOKE

ONCE IN 2 HRS BY QA

PROVIDED

PRE-DESPATCH
INSPECTION
AS PER SHOWA SINGLE
SAMPLING PLAN
( ACCEPTANCE LEVEL 0 )
RESP : INSPECTOR

4.TABLE MOVEMENT

NIL

IT IS IN PM CHECK

NIL

LIST

133

Process FMEA after 6 Sigma

PROCESS
STEP

FAILURE MODE

POTENTIAL CAUSES

1.CENTER SHIFT OF RADIUS

OCC

CURRENT PROCESS
CONTROL (TYPE - B)

PROCESS CONTROL

CURRENT PROCESS
CONTROL (TYPE - C)

NIL

CURRENT PROCESS
CONTROL (TYPE - A)

LUG

FLANGE

ONCE IN 2 HRS BY QA

MILLING

THICKNESS

CHART

PRE-DESPATCH

VARIATION

CHECK 1 NO/HOUR

INSPECTION

RESP: OPERATOR

AS PER SHOWA SINGLE

DET

RPN

SAMPLING PLAN
( ACCEPTANCE LEVEL 0 )
INSPECTOR

2.SPINDLE AXIAL AND

RADIAL PLAY

NIL

ADDEED IN THE
PM CHECK LIST

134

Process FMEA after 6 Sigma

PROCESS
STEP

FAILURE MODE

POTENTIAL CAUSES

OCC

CURRENT PROCESS
CONTROL (TYPE - B)

NIL

CURRENT PROCESS
CONTROL (TYPE - C)

PIN HOLE

WALL

1.CASTING DIMENSION

RECEIVING INSPECTION

DRILLING

THICKNESS

( SYMMETRICITY,

AS PER SHOWA SINGLE

UNDERSIZE

CAVITY TO CAVITY &

SAMPLING PLAN

BOSS DIAMETER )

( ACCEPTANCE LEVEL 0 )

CURRENT PROCESS
CONTROL (TYPE - A)

DET

NIL

RPN

RESP : INSPECTOR

2.RADIUS DEPTH

PROCESS CONTROL

NIL

ONCE IN 2 HRS BY QA

CHART

PRE-DESPATCH

CHECK 1 NO/HOUR

INSPECTION

RESP : INSPECTOR

AS PER SHOWA SINGLE

SAMPLING PLAN
3.CENTER SHIFT OF

PROCESS CONTROL

RADIUS

CHART

NIL

( ACCEPTANCE LEVEL 0 )
RESP : INSPECTOR

CHECK 1 NO/HOUR
RESP : INSPECTOR

135

Control Plan after 6 Sigma

Characteristics

Product

DISC

Process

1.RADIUS DEPTH

MILLIN G

M ethods

Special
Character Product/Process Evaluation/
istic class
Specification
M easurement
Tolerance
Technique

Sample
Size

133.55 / 133.67 MM S GV 7018 GAUGE

Frequency

Control
method

Reaction Plan
Resp.

Recording (Y/N)

PER HOUR IN SPECTIONOPERATOR

YES

HOLD,REIN SPECT

PER HOUR IN SPECTIONOPERATOR

YES

HOLD,REIN SPECT

(PROCESS)
2.CEN TER SHIFT OF

48.17 / 48.37 MM

RADIUS

(PROCESS)

S GV 7018 GAUGE

3.CASTING

LEVER DIAL

DIMENSION

AND FIXTURE

100%

INSPECTION QA

NO

SEGREGATE AND
KEEP

( TOOLING
LOCATION W.R.TO
CENTER OF

CASTING )

4.HARDNESS

165 - 250 BHN

HARDNESS

AS PER

MEASURING

SHOWA

INSTRUMENT

SINGLE

INSPECTION QA

NO

SEGREGATE AND
KEEP

SAMPLING
PLAN

136

Control Plan after 6 Sigma

Characteristics

Product

Process

LUG

1.FLANGE

MILLING

THICKNESS

2.POSITION FROM

Special
Character Product/Process Evaluation/
istic class
Specification M easurement
Tolerance
Technique
13.15 / 12.85 MM

VERNIER

M ethods
Sample
Size

Frequency

Control
method

Reaction Plan
Resp.

Recording (Y/N)

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REINSPECT

11 GG 5092

11.5 MM

VERNIER

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REINSPECT

24.20 / 23.80 MM

HEIGHT GAUGE

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REINSPECT

HOLE CENTER

3.LUG FACE TO
"V" FACE

S GD 7091

4.CENTER SHIFT OF

48.17 / 48.37 MM

RADIUS

(PROCESS)

S GV 7018 GAUGE

KEEP

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REINSPECT

137

Control Plan after 6 Sigma

Characteristics

Product

Process

WALL

Special
Character Product/Process
istic class Specification
Tolerance
4.80 / 5.80 MM

THICKNESS

M ethods
Evaluation/
M easurement
Technique
VERNIER

Sample
Size

Frequency

Control
method

Reaction Plan
Resp.

Recording (Y/N)

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REIN SPECT

S GG 7167

1.RADIUS DEPTH

S GV 7018 GAUGE

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REIN SPECT

2.CENTER SHIFT OF

S GV 7018 GAUGE

PER HOUR INSPECTIONOPERATOR

YES

HOLD,REIN SPECT

NO

SEGREGATE AND

RADIUS

3.CASTING

LEVER DIAL

AS PER

DIMENSION

AND FIXTURE

SAMPLING

INSPECTION QA

KEEP

( SYMMENTRICITY )

138

Six Sigma Tools used

Phase

Tools

Control

1. Variation Analysis
2. Pre-Control chart
3. Control plan

139

Benefits
Rejection reduced to Zero
Estimated Cost Saving by Rejection reduction is Rs. 1,21,589 / Annum

Rejection due to M/H Less Wall thickness

5
Qty Produced from week 39 week 45
= 39785 Nos

45

ee
k

44

ee
k

43

ee
k

42
W

ee
k

41
W

ee
k

40
ee
k

39
W

ee
k

38
W

ee
k

37
W

ee
k
W

ee
k
W

Improvement
through 6s

36

Qty in Nos

8
7
6
5
4
3
2
1
0

140

141