You are on page 1of 45

6

Six Sigma Project


Project Number

TACL 110

Plant

TACL FARIDABAD

Name of the Green Belt

Narender Singh

Team Members

Date of Start

02.01.2010

Slide number: 1

Phase 1- Problem Definition

Problem Statement

Part number selected for study

ID Closing & Final Inspection

Current average rejection for last 3 lots

ID Closing

Process stages where the problem is detected

NA

Last manufacturing process stage where technically the problem can get
generated

182191KSP9300C

Other similar part numbers having the problem

Splitting Problem

3.3 %

Maximum and Minimum rejection in last 3 lots


Maximum rejection in 3 inspection lot - 6.2 %
Minimum rejection in 3 inspected lot - 1.5 %
Slide number: 2

Phase 1- Problem Definition

Number of lines/presses/machines used for processing

1 Presses

Objective of the Project


To reduce the current Scrap percentage from 3.32 % to 0.66.%

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

1.08 Lakhs

Response
Attribute

Specification (if the response is variable)


NA

Is R&R study required


- No

Slide number: 3

Phase 1- Problem Definition


Process Mapping
GUILL GRAPHITE

MAKE SHELL

ANNEALING

MAKE SHELL

ANNEALING

PACK GRAPHITE IN
COPPER SHELL

CLOSE OD

CLOSE ID

BLANK

INSPECTION

Slide number: 4

Phase 1- Problem Definition


Project Planning
Phase

DEC2009
W1

W2

W3

JAN10
W4

W1

W2

W3

FEB10
W4

W1

W2

W3

W4

Define
Measure
& Analyze
Improve
Control
Phases

Planned Start
date

Planned
Completion
date

Actual start
date

Actual
completion
date

Status

Define

24.12.09

1.01.2010

2.01.10

11.01.2010

COMPLETE

Measure &
Analyze

4.01.2010

18.01.2010

12.01.2010

5.02.2010

Complete

Improve
Control

Slide number: 5

Phase 1- Problem Definition

SHELL

SPLITING PROBLEM

GRAPHITE
LAYERS

ID CLOSING

ASSEMBLY

OD CLOSING
Slide number: 6

Phase 1- Problem Definition


Rejection Trend Based On 6 month data
Rejection Trend for 18291KSP9300C July'09~Dec'09
7.0

Average
Rejection=3.3%

Percentage Rejection

6.0
5.0
4.0
3.0
2.0
1.0
0.0
July'09

August'09

Sep'09

Oct'09

Nov'09

Dec'09

Months
22.6 % is the average rejection as calculated
from 3 lot inspection

Slide number: 7

Phase 1- Problem Definition


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

824 Nos.

Number of pieces scrapped last month

824 Nos.

Number of pieces reworked last month

NIL

Scrap cost/piece

7.75

Rework cost/piece

NA

Total scrap cost (Rs. Lakhs) for last


month

0.064 Lakhs

Total rework cost(Rs. Lakhs) for last


month

NIL

Total Rejection cost (Rs. Lakhs) for last


month

0.064 Lakhs

Extrapolated Total rejection cost (Rs.


Lakhs) for one year

1.08 Lakhs

Slide number: 8

Phase 1- Problem Definition (Data collection and


analysis)
Monthly rejection trend for last 6 months
Rejection Trend for 18291KSP9300C July'09~Dec'09
7.0

Average
Rejection=3.3%

6.0
Percentage Rejection

5.0
4.0
3.0
2.0
1.0
0.0
July'09

August'09

Sep'09

Oct'09

Nov'09

Dec'09

Months

Slide number: 9

Phase 1- Problem Definition


SUSPECTED SOURCES OF VARIATION
Input Materials
1)
2)
3)
4)

Forming Height
Burr
No of graphite layers
Closing Process

Slide number: 10

FF

VS

OBSER

MVA

CC

CS

PPS

PC

SSVs

MCS

Phase 1- Problem Definition

HARDNESS
BURR
NO OF GRAPHITE LAYERS
CLOSING PROCESS

Legend: PC Paired Comparison, PPS Product/Process search, CS Component Search, MCS Modified Component Search, CC Concentration chart
MVA Multi-Vari analysis, OBSER Observation, VS Variable Search, FF Full Factorial
Slide number: 11

Phase 2- Measure and Analyze


Analysis # 1

Objective: Is Burr the cause for split problem

Technique/Tool used: Product Process Search

Slide number: 12

Phase 2- Measure and Analyze


CALCULATION OF COUNT FOR BURR
SNO.

SSV 'BURR'
scale

RESPONSE

NOT OK

NOT OK

NOT OK

NOT OK

NOT OK

NOT OK

15
12

NOT OK

NOT OK

10

OK

11

OK

13

OK

14

OK

16

OK

OK

OK

OK

Using a 3 level scale for burr


Scale 1= No Burr
Scale 2= Low Burr
Scale 3= High Burr

As the SSV corresponding to scale 3


Is both OK and Not OK So count is
zero.
Hence it is not a cause

Slide number: 13

Phase 2- Measure and Analyze


Objective: Is Forming Height the cause for split
problem

Technique/Tool used: Product Process Search

Slide number: 14

Phase 2- Measure and Analyze


CALCULATION OF COUNT FORMING HEIGHT
COMP. No.

FORMING
HT.

RESPONSE

184

6.01

OK

291

6.03

OK

174

6.03

OK

176

6.03

OK

264

6.04

OK

272

6.04

OK

281

6.04

OK

163

6.04

OK

167
171

6.04
6.04

OK

116

6.19

NOT OK

223

6.2

NOT OK

25

6.21

NOT OK

293

6.22

NOT OK

249

6.22

NOT OK

315

6.24

NOT OK

137

6.24

NOT OK

209

6.25

NOT OK

33

6.25

NOT OK

333

6.26

NOT OK

222

6.26

NOT OK

237

6.27

NOT OK

OK

TOP COUNT=8
BOTTOM COUNT=8
TOTAL COUNT =16

New Spec = 5.60.4 mm


Variation in height = 0.26
Derived Tolerance = 0.8
Ratio
=(0.26/0.8)*100
= 32.5 %
So less than 75% of tolerance
so It can be controlled.
Hence It is a root cause

Slide number: 15

Phase 2- Measure and Analyze


Objective: Is No. of graphite layers the cause for
split problem

Technique/Tool used: Product Process Search

Slide number: 16

Phase 2- Measure and Analyze


CALCULATION OF COUNT No. OF GRAPHITE LAYERS
SNO.

No. Of Graphite
layers

RESPONSE

OK

15

OK

12

OK

13

OK

OK

OK

OK

OK

OK

11

OK

14

OK

OK

OK

10

Not Ok

16

OK

OK

Since 4 layers response is both ok and


not ok . So count is zero
Hence it is not a cause

Slide number: 17

Phase 2- Measure and Analyze


Total Analysis summary

Cause(s) identified for the problem: Forming Height

Root cause(s) identified for the problem: Forming


Height

Slide number: 18

Phase 3 - Improve
Validation of the Root cause(s)- B Vs C
1

Part number selected


for validation

Better Condition
Current Condition

Sample size

6B,6C / 3B,3C

Sample type

Pieces / Lots

Response decided for


monitoring

Historic rejection

Lot quantity

Is alternating between Yes / No


B and C condition
possible

Slide number: 19

Phase 3 - Improve
Validation of the Root cause(s)- B Vs C
Piece / Lot

Better ( B )

Current ( C )

1
2
3
4
5
6

Slide number: 20

Phase 3 - Improve
Data

Slide number: 21

Phase 3- Improvement
Analysis - B Vs C
1

Part number selected


for validation

Average of B
Average of C

Total count

Xb Xc (Amount of
Improvement)

Sigma (B)

Is Xb-Xc greater than


Sigma (b)

Confidence level of
improvement

Number of iterations
done
Slide number: 22

Phase 3 Improve Actions Summary


No

Root cause(s)

Actions Planned

Planned
Completion date

Resp

Actual
completion
date

Status

Slide number: 23

Phase 3 Improve
Photograph of actions

Slide number: 24

Phase 3 Improve
Photograph of actions

Slide number: 25

Phase 3 Improve
Photograph of actions

Slide number: 26

Phase 4 Control Phase


No

Root cause

Variation
analysis done

Type of control
method decided

Status of implementation

Slide number: 27

Phase 4- Control
Analysis # 1
Objective:
Tool/technique used:
-

Slide number: 28

Phase 4- Control
Steps in Variation analysis
Step 1 : Identify the Parameter that requires Variation analysis
Parameter :

Step 2 : Identify are there multiple Product and Process streams


Number of streams :
Type of streams
:

Slide number: 29

Phase 4- Control

Slide number: 30

Phase 4- Control
Data

Slide number: 31

Phase 4- Control
Data

Slide number: 32

Phase 4- Control
Multi Vari analysis - Conclusion
Part to Part variation
Time to Time variation
Stream to Stream variation -

Slide number: 33

PHASE - 4

CONTROL PHASE

Step: 7 Check the consistency of the part to Part variation

Paste your range chart here

Slide number: 34

PHASE - 4

CONTROL PHASE

Slide number: 35

PHASE - 4

CONTROL PHASE

Estimated part to part variation at 99.73%


confidence level is :

Slide number: 36

PHASE - 4

CONTROL PHASE

Slide number: 37

PHASE - 4

CONTROL PHASE

Slide number: 38

PHASE - 4

CONTROL PHASE

Slide number: 39

Phase 4- Control
Analysis # 1
Objective:
Conclusion:
-

Slide number: 40

Phase 4 - Control
Results

Before

After

Slide number: 41

Phase 4- Control
Project Summary
Number of drill downs (funneling) done to reach the root
cause(s):
Number of predominant root cause(s) identified:
Time in months for completing the project:
Tools/Techniques used:

Slide number: 42

Phase 4 - Control
Benefits derived through the project
1.

Overall scrap reduced from 1.95% to 1.02%.

2.

Crack scrap reduced from 0.62% to 0.02%.

3.

Cycle time reduced from 460 seconds to 422 seconds.

4.

Estimated cost benefits is Rs. 2,18,996 per year.

Slide number: 43

Phase 4 - Control
Current control plan

Slide number: 44

Phase 4 - Control
Proposed Control plan

Slide number: 45