6

SD BASKET SHAFT
DIAMETER VARIATION &
MWO DIMPAL HEIGHT VAR.
 Customer VP MD
Six Sigma Theme Registration Facilitator HOD

Approval
Chetan
Deptt : Production Team Leader: Sanjay Arora Khosla

Theme SD Basket Shaft Diameter Improvement Project Schedule

Event Time Schedule
Period 10/01/2000 - 20/03/2000
Step
Ist wk 2nd wk 3rd wk 4th wk 5th wk 6th wk 7th wk 8th wk
Team Members Sandeep Sethi, Sanjay Jain, AK Jindal Items

1 M
More Dia. Results In Tight Fitting Of SD
Basket In Bellow Assembly,Displacing
Theme The Spring OF Bellow,Causing Leakage 2 A
Description Through Spin Side. This Water Falls On
The Spin Motor , Leading To Burning Of 3 I
Motor.

4 C

Definition Of Yield
From existing defect rate of Shaft diameter
Target
56,000 ppm to 3.4ppm.
Commitment / Help Required
Cost saving = Rs 75,000
RM Engg. ,the manufacturers of the
S Shaft.
Process Map
define measure analyse improve control
6
Scrap Sdbasket
assm.
Parting Pressing Packing
Check
Insert Dispatch
Turning
Rivet to Brite
on lathe r Emery
Flaring of e paper a
Milling Rivet IQC at j c
brite e c
rejected
Centre Polishing c
Inspection e
accepted
Drill (oemery ) t p
Moulding
e Segregation t
Centre Final d e
less Inspection Check for
d
runout rejected
Grinding
Sleeve Dispatch to IQC at
Knurling Insertion LGEIL
LGEIL
accepted
 6
measure
Measure/Gage R & R define analyse improve Control

sample operator diameter

1 1 13.9780
1 1 13.9780 Source %Contribution %Study Var %Tolerance
2 1 13.9775
2 1 13.9780 Total Gage R&R 8.78 29.63 10.47
3 1 13.9765 Repeatability 6.59 25.66 9.07
3 1 13.9760
Reproducibility 2.20 14.82 5.23
4 1 13.9760
4 1 13.9760
operator 0.37 6.05 2.14
5 1 13.9770 operator*sample 1.83 13.53 4.78
5 1 13.9770 Part-To-Part 91.22 95.51 33.74
6 1 13.9780 Total Variation 100.00 100.00 35.33
6 1 13.9780
7 1 13.9780
7 1 13.9780 Number of Distinct Categories = 5
8 1 13.9770
8 1 13.9775
9 1 13.9770
9 1 13.9770 Gage name:
Date of study:
10 1 13.9770 Gage R&R (ANOVA) for diameter Reported by:
Tolerance:
10 1 13.9765 Misc:
1 2 13.9780 Xbar Chart by operator operator*sample Interaction
1 2 13.9780 13.978
1 2 13.978
operator
Sample M ean

2 2 13.9775 1

Averag e
3.0SL= 13.98 2

2 2 13.9780 13.977
X=13.98
-3.0SL= 13.98 13.977

3 2 13.9760
3 2 13.9760 13.976 13.976
0
4 2 13.9760 sample 1 2 3 4 5 6 7 8 9 10

R Chart by operator By operator

4 2 13.9760 1 2 13.978
0.0005
5 2 13.9770 3.0SL= 4.90E-04
Sample R ang e

0.0004
5 2 13.9770 0.0003
13.977
6 2 13.9775 0.0002

0.0001
R=1.50E-04

6 2 13.9775 0.0000 -3.0SL= 0.00E+ 00

13.976
7 2 13.9780 0 operator 1 2

7 2 13.9780 Components of Variation By sample

8 2 13.9770 100
%Total Var
13.978

%Study Var
8 2 13.9775
Percent

%Toler

9 2 13.9765 50 13.977

9 2 13.9765
13.976
10 2 13.9770 0
Gage R&R Repeat Reprod Part-to-Part sample 1 2 3 4 5 6 7 8 9 10
10 2 13.9770
 6
measure
Measure/Process
Measure Capability define analyse improve control

Process Data
USL 13.9860
Target * Process Capability Analysis for diameter
LSL 13.9750 Calculations Based on Weibull Distribution Model
Mean 13.9760
Sample N 30 LSL USL
Shape 17692.1
Scale 14.0

Ov erall (LT) Capability

Pp 1.64
PPU 2.40
PPL 0.36
Ppk 0.36

Observ ed LT Perf ormance

PPM < LSL 0.00
PPM > USL 0.00
PPM Total 0.00

Expected LT Perf ormance 13.970 13.974 13.978 13.982 13.986

PPM < LSL 139190.55
PPM > USL 0.00
PPM Total 139190.55
ANALYSE / Fish Bone Diagram define measure
analyse
improve control
6
Man Method

Inability to concentrate

Min. cut possible Inconsistent view of the

scale

Manual feed
Inconsistency in selecting
the point of checking Dia.
var.
in
shaft
No. of cuts Position of shaft

Manufacturing
system Inspection
Use of
Depth of cut
method wrong
Not accounted for measuring
Dressing of grinder system
wheel
Material Method Gages Micrometers

Innovation with 6Achieve Super A! LGEIL

ANALYSE / No. of cuts
define measure analyse improve control 6
As p value is less than 0.05 so no. of cuts is a factor for
dia.variation in S - Shaft

Boxplots of dia. of by factor

(means are indicated by solid circles)

One-way Analysis of Variance

13.981
Analysis of Variance for dia. of
Source DF SS MS F P
factor 1 0.0000528 0.0000528 77.13 0.000 13.980
Error 18 0.0000123 0.0000007
Total 19 0.0000651
Individual 95% CIs For 13.979
Mean

dia. of shaft
Based on Pooled StDev
Level N Mean StDev --------+---------+---- 13.978
1 10 13.9765 0.0006
2 10 13.9797 0.0010
13.977
Pooled StDev = 0.0008 13.9770 13.9785
13.9800
13.976

13.975

Level 1: 5 cuts factor

2
Level 2 : 6 cuts

Innovation with 6s Achieve SuperA

ANALYSE / Measuring Sys.
define measure analyse improve control 6
As p value is less than 0.05 so measuring system is a factor for
dia.variation in S - Shaft

Boxplots of dia. by factor

One-way Analysis of Variance
(means are indicated by solid circles)

Analysis of Variance for dia.

Source DF SS MS F P 13.979
factor 1 0.0000153 0.0000153 28.05 0.000
Error 18 0.0000098 0.0000005
Total 19 0.0000251 13.978
Individual 95% CIs For Mean
Based on Pooled StDev
Level N Mean StDev -+---------+---------+--
13.977

dia.
1 10 13.9777 0.0008 (---*--) 13.976
2 10 13.9759 0.0006 (--*--)

Pooled StDev = 0.0007 13.9755 13.9770 13.9785

13.975
13.9800
factor

2
Level 1: mech.
Level 2: Digi.

Innovation with 6s Achieve SuperA

 6
improve
define measure analyse control
IMPROVE / DOE
No of cuts Mea.sys. data
6 mech 13.982
Main Effects Plot (data means) for data
6 mech 13.983
6 mech 13.979
5 mech 13.977 ch i
5 6 me di g
5 digi 13.976 13.9810

5 digi 13.975
13.9798
6 digi 13.982
5 digi 13.975

data
13.9786
5 mech 13.977
6 mech 13.982 13.9774
6 digi 13.981
5 mech 13.977 13.9762
No of cuts Measuring sy
6 digi 13.98
6 mech 13.979
5 mech 13.978
5 digi 13.975
6 digi 13.9795
6 digi 13.98
The graph suggests that no. of cuts is a
5 digi 13.976 major factor & the measuring system
5 mech 13.978
is a minor factor.
 6
improve
define measure analyse control
IMPROVE / DOE
No of cuts Mea.sys. data
6 mech 13.982
Main Effects Plot (data means) for data
6 mech 13.983
6 mech 13.979
5 mech 13.977 ch i
5 6 me di g
5 digi 13.976 13.9810

5 digi 13.975
13.9798
6 digi 13.982
5 digi 13.975

data
13.9786
5 mech 13.977
6 mech 13.982 13.9774
6 digi 13.981
5 mech 13.977 13.9762
No of cuts Measuring sy
6 digi 13.98
6 mech 13.979
5 mech 13.978
5 digi 13.975
6 digi 13.9795
6 digi 13.98
The graph suggests that no. of cuts is a
5 digi 13.976 major factor & the measuring system
5 mech 13.978
is a minor factor.
 OPTIMUM CONDITIONS OF THE FACTORS
OBTAINED AFTER DOE

FACTORS OPTIMUM CONDITIONS

1. NO. OF CUTS SIX

2. MEASURING SYSTEM DIGITAL MICROMETER

3. DRESSING ( CUTTING WHEEL ) EVERY ONE HOUR

4. DRESSING ( CONTROL WHEEL ) EVERY TWO DAYS

ANALYSE / Fish Bone Diagram define measure
analyse
improve control
6
Innovations Practical Result
 Four Vendor persons and one LG
 Improved the Process Capability person involved in 100 % inspection
To more than Six Sigma level. removed.
 Avery tight tolerance of 11microns
(10 -6m) earlier thought unachievable
 Educated Vendor Top Management & was met.
all Operators about the seriousness of  A Chronic problem of Washing
quality of this dimension and also how machine line : SD Shaft loose or tight
to maintain it. was solved.
 Leakage problem on the line due to
 Provided the right equipment & scratches generated because of excessive
educated the vendor supervisor & handling was also eliminated.
operators about how to use the  Total RTN reduced from 56000 to
equipment. Zero .
 Involved the Vendor supervisors in  Three stages of emery paper rubbing ,
developing the control mechanism & rejection from Brite IQC and
plan as well as its execution. Segregation at our end were eliminated.

Innovation with 6Achieve Super A! LGEIL

Improve/Process Capability define measure analyse improve
control
6
BEFORE 6 AFTER 6
Cp = 1.5 Cp = 2.42
Process Data
Cpk=0.36
USL 13.9860 Process Capability Analysis for diameterCpk=2.12
Target * Process Capability Analysis for diameter
LSL 13.9750 Calculations Based on Weibull Distribution Model
Mean 13.9760 LSL USL
Process Data
Sample N 30 LSL USL ST
USL 13.9860
Shape 17692.1 Target * LT
Scale 14.0 LSL 13.9750
Mean 13.9800
Ov erall (LT) Capability Sample N 30
Pp 1.64 StDev (ST) 0.0007579
StDev (LT) 0.0010117
PPU 2.40
PPL 0.36
Ppk 0.36 Potential (ST) Capability
Cp 2.42
Observ ed LT Perf ormance CPU 2.62
CPL 2.22
PPM < LSL 0.00
Cpk 2.22
PPM > USL 0.00
Cpm *
PPM Total 0.00 13.975 13.977 13.979 13.981 13.983 13.985 13.987

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

Expected LT Performance 13.970 13.974 13.978 13.982 13.986 Pp 1.81 PPM < LSL 0.00 PPM < LSL 0.00 PPM < LSL 0.30
PPM < LSL 139190.55
PPU 1.96 PPM > USL 0.00 PPM > USL 0.00 PPM > USL 0.00
PPM > USL 0.00
PPL 1.66 PPM Total 0.00 PPM Total 0.00 PPM Total 0.30
PPM Total 139190.55 Ppk 1.66
Control /Xbar R Chart 6
Xbar/R Chart for diameter

13.9815
3.0SL=13.98
Sample Mean

13.9805
X=13.98
13.9795

13.9785 -3.0SL=13.98

Subgroup 1 2 3 4 5 6

0.005 3.0SL=0.004976
Sample Range

0.004
0.003
R=0.002353
0.002
0.001
0.000 -3.0SL=0.00E+00
 CTQ No. X-R Control Chart In Charge Supervisor
IQC -02 (S Shaft Diameter)
14.000-0.014 Checked
Part Name S SHAFT( SD Basket) Spec. Model All Washing Machines By
-0.025
Spec. Upper 13.986
Measuring Unit mm Application W/M Assembly Line
Limit Lower 13.975
Cpk
Measuring Method Digital Micrometer Periodical Check Every 2 hours Time period 01/03/2000~15/03/2000

Date 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
X1
X2
Measured
X3
Value
X4
X5
Total ΣX
Average x
Range R
13.9860

13.9815

X Chart 13.980

13.9785
13.9750

0.005
0.004
0.003
R Chart 0.002
0.001
0.000

Remarks
 CTP (Check Sheet)
CTQ LINE
S Shaft (SD Basket ) 8003 IQC

CTP No. CTP Item Spec. How to check Frequency Remarks

IQC-02-01 No. of cuts 6 Update W.I Every 2 Hours
IQC-02-02 Measuring System Digital Update W.I Every 2 Hours
IQC-02-03 Grinding Wheel Dressing Update W.I Every Hour
IQC-02-04 Control Wheel Dressing Update W.I After 2 days

CTP Name
Month ( )
Date 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
9:00
11:00
No. of Cuts
2:00
5:00
9:00
11:00
Measuring
2:00
System
5:00
9:00
Cuting Wheel 10.00
11.00
12.00
3:00
4:00
5:00
Control Wheel 2 days
Signature of
 60000 56k

50000 Zero
R Rejection
e 40000
j
e REJECTION
c 30000
t 22k

I 6 SIGMA
o 20000 LEVEL
n
10000
0.1k

0
Jan Feb Mar Apr May
 60000 ANNUAL SAVING = 6.27 LACS

50000 Zero
R Rejection
e 40000
j
e REJECTION
c 30000
t 38 k
I 6 SIGMA
o 20000 LEVEL
n
10000
62 k 63 k 63 k

0
Jan Feb Mar Apr May
CTQ Registration Format. 6
Registration No. PjtLdr. Super A CTQ
R&D QP
Incharge
CTQ REGISTRATION

Sign
( SD Shaft Dia Variation) Date

CTQ Description (what is the CTQ) Measuring System (How to measure) Target Present

Digital Micrometer ( LC-0.001 mm) Zst 6.0 7.26

Shaft Diameter
Zlt 4.5 6.36

Graphical Description (show sketch or photo) Gage R&R Specifications Control Limits
%Study
Variation % Tolerance LSL USL LCL UCL
SD 13.9750 13.9860 13.9785 13.9815
Shaft 29.63% 10.47% mm mm mm
mm
CTP List Document Check List

S.No. Name Area of CTP Spec. S.No. Document

 Doc. Reference
Application  Number
1 Number if cuts Vendor end 6

2 Measuring sys Vendor end Digital

Dressing every
3 Grinding Wheel Vendor end hour
Dressing every
Control Wheel Vendor end 2 days
4
 Approval

Period
18 NOV.99
6 Theme Registration to
17 FEB .00
M.D.
V.P.
OPTIMISING SET UP COND FOR PROJ. WELDING
Theme : FOR MS-283/304 MWO CAVITY. Dept. Head Sanjay Arora
Detailed Description : Guide C. Khosla
* HIGH REJECTION DUE TO UN OPTIMISED SET UP CONDITIONS. Pjt. Leader A.K.Jindal
Name Sign

MEMBERS
* THE LOSS IS APPROX RS.400 PER CAVITY. H.G.Choi

* IT IS A CTQ AS IF THE Pcs GO UNDETECTED THESE CAN LEAD N.C.Upreti

TO FAILURE OF THE M/C & HAMPER MARKET POSITION.
V.Madaan
Q.Fn.(Y) Specification Gauge Name Gauge L.C. Current ( Y ) Target ( Y )
NO. OF
NIL NIL NIL 80,000 PPM 3.4 PPM
SET UP REJ.
Process Map With Yield of each process Logic Tree of Worst Process
Lo-Up
Inspection Welding
U-Bending Set Up Rej.
of sheets
of MS-283/304
Fr. Long Back Short Back Long
Welding Welding Welding

Welding Welding
Fr. Short TT Bkt Hng. Plt
Pressure Current
Welding Welding Welding

Shift To Balance Wv Gd/Mgt Sheet Dimple Contact

MW Line Checking Welding Chemistry Height Time

Estimated Contribution after Project Request To M.D.

Qualitative Quantitative (Rs. Lacs)
* Improvement in
Rs. 400,000
Rejection
 Measure
Measure / Gage R&R Define Analyse Improve Control
6
As Gage R&R % Study Variation & % Tolerance is less than 20% hence we can use this gage for dimple height
variation study

Gage Used For Measurement of Dimple Height : Puppy Dial

Least Count= 0.001 mm
StdDev Study Var %Study Var %Tolerance
Source (SD) (5.15*SD) (%SV) (SV/Toler) Sample Jindal Jindal2 V.Madaan V.Madaan2
1 1.295 1.293 1.295 1.292
Total Gage R&R 1.36E-03 0.006997 6.44 5.60 2 1.299 1.299 1.299 1.299
Repeatability 1.34E-03 0.006878 6.33 5.50 3 1.289 1.29 1.284 1.287
Reproducibility 2.49E-04 0.001283 1.18 1.03 4 1.322 1.322 1.322 1.322
Operator 2.49E-04 0.001283 1.18 1.03
Part-To-Part 2.11E-02 0.108451 99.79 86.76
5 1.305 1.3 1.302 1.305
Total Variation 2.11E-02 0.108677 100.00 86.94 6 1.279 1.279 1.279 1.279
7 1.298 1.3 1.298 1.298
8 1.302 1.302 1.302 1.302
Number of Distinct Categories = 22 9 1.348 1.348 1.348 1.348
10 1.333 1.33 1.33 1.331

Innovation with 6  Achieve Super A! LG Electronics

 Measure
Measure / Gage R&R Define Analyse Improve Control
6
From measurement of dimple Ht. ; which we know is an important factor for quality of welding
we know that it is a primarily a problem of variation.

Four Block Diagram

Process Capability Analysis for
Dimple Ht.
LSL USL
Process Data
USL 1.37500
Within
Target *
LSL 1.22500 Overall
Mean 1.29044
Sample N 125
StDev (Within) 0.0258164
StDev (Overall) 0.0421137
A B
Potential (Within) Capability
Cp 0.97
1.5
CPU 1.09
CPL 0.84 Zshift
Cpk 0.84 C D
1.15 1.20 1.25 1.30 1.35 1.40
Cpm *

Overall Capability Observed Performance Exp. "Within" Performance Exp. "Overall" Performance 4.5
Pp 0.59 PPM < LSL 40000.00 PPM < LSL 5624.12 PPM < LSL 60100.56
PPU 0.67 PPM > USL 40000.00 PPM > USL 527.61 PPM > USL 22329.16 Zst
PPL 0.52 PPM Total 80000.00 PPM Total 6151.73 PPM Total 82429.72
Ppk 0.52

Innovation with 6  Achieve Super A! LG Electronics

 6
Analyse
Analyse / 4-M Diagram Define Measure Improve Control

Machine Man
VOLTAGE
LEVEL
PRESSURE
CAVITY TPR
DURING WELDING

ELECTRODES
ELECTRODES GAP
DRESSING

Set Up
Rejection
DIMPLE HEIGHT
VARIATION
PARAMETERS
NOT DEFINED PROPERLY
PROFILEOF
DIMPLE SHEET CONTACT
CHEMISTRY TIME

Material Method

Innovation with 6  Achieve Super A! LG Electronics

 Measure
Analyse / Dimple Profile Define Analyse Improve Control
6
Two proportion test to examine the effect of dimple profile on no. of weak joints during
projection welding.
Test and CI for Two Proportions

Sample X N Sample p
Tomb Type 3 42 0.071429
Conical 13 42 0.309524
Sheet

DIMPLE WITH TOMB PROFILE Estimate for p(1) - p(2): -0.238095

95% upper bound for p(1) - p(2): -0.103783

Test for p(1) - p(2) = 0 (vs < 0): Z =

-2.92

Sheet P-Value = 0.002

DIMPLE WITH CONICAL PROFILE

We have established that the dimple profile is having an effect on welding quality.Tomb
profile is better.

Innovation 6 
with Achieve Super A! LG Electronics
 6
Analyse
Analyse / Dimple Profile Define Measure Improve Control

The press die dimple profiles were changed as per previous test inference. Also the punches
which were varying in height were made same.With this direct improvement the process
capability of dimple height improved from 0.97 to 2.63.

Process Capability Analysis for Process Capability Analysis for C45

Dimple Ht.
LSL USL LSL USL
Process Data Process Data
USL 1.37500 USL 1.37500
Within Within
Target * Target *
LSL 1.22500 Overall LSL 1.22500 Overall
Mean 1.29044 Mean 1.29498

Sample N 125 Sample N 125

StDev (Within) 0.0258164 StDev (Within) 0.0095059

StDev (Overall) 0.0421137 StDev (Overall) 0.0162463

Potential (Within) Capability Potential (Within) Capability

Cp 0.97 Cp 2.63

CPU 1.09 CPU 2.81

CPL 0.84 CPL 2.45

Cpk 0.84 Cpk 2.45

1.15 1.20 1.25 1.30 1.35 1.40 Cpm * 1.22 1.24 1.26 1.28 1.30 1.32 1.34 1.36 1.38
Cpm *

Overall Capability Observed Performance Exp. "Within" Performance Exp. "Overall" Performance
Overall Capability Observed Performance Exp. "Within" Performance Exp. "Overall" Performance
Pp 1.54 PPM < LSL 0.00 PPM < LSL 0.00 PPM < LSL 8.25
Pp 0.59 PPM < LSL 40000.00 PPM < LSL 5624.12 PPM < LSL 60100.56
PPU 1.64 PPM > USL 0.00 PPM > USL 0.00 PPM > USL 0.42
PPU 0.67 PPM > USL 40000.00 PPM > USL 527.61 PPM > USL 22329.16
PPL 1.44 PPM Total 0.00 PPM Total 0.00 PPM Total 8.68
PPL 0.52 PPM Total 80000.00 PPM Total 6151.73 PPM Total 82429.72
Ppk 1.44
Ppk 0.52

Innovation 6 
with Achieve Super A! LG Electronics
Improve / SS Cavity Define Measure Analyse Improve Control 6
We know from past setting experience that three parameters (PG - Pressure Gauge. PSP &
Voltage ) have a crucial impact on welding quality. We did RSM to establish the best machine
condition.
Response Response Surface Regression: r versus PG, voltage
PG PSP voltage O.K N.G.
(%o.k) The analysis was done using coded units.
3.000 1.000 270 2 10 0.167 Estimated Regression Coefficients for r
4.500 1.000 270 1 11 0.083
3.000 1.500 270 4 9 0.308 Term Coef SE Coef T P
Constant 0.4285 0.04016 10.671 0.000
4.500 1.500 270 0 13 0.000
Block -0.0383 0.04016 -0.955 0.356
3.000 1.000 320 11 1 0.917 PG -0.1796 0.04399 -4.082 0.001
4.500 1.000 320 2 10 0.167 voltage 0.1925 0.04399 4.375 0.001
3.000 1.500 320 11 2 0.846
4.500 1.500 320 6 7 0.462 S = 0.1606 R-Sq = 72.4% R-Sq(adj) = 66.5%
3.750 1.250 295 4 8 0.333
3.750 1.250 295 8 5 0.615 Analysis of Variance for r
3.750 1.250 295 3 9 0.250
3.750 1.250 295 7 6 0.538 Source DF Seq SS Adj SS Adj MS F P
Blocks 1 0.02351 0.02351 0.02351 0.91 0.356
2.525 1.250 295 11 1 0.917
Regression 2 0.92387 0.92387 0.46194 17.90 0.000
4.975 1.250 295 5 8 0.385 Linear 2 0.92387 0.92387 0.46194 17.90 0.000
3.750 0.842 295 4 8 0.333 Residual Error 14 0.36127 0.36127 0.02581
3.750 1.658 295 5 8 0.385 Lack-of-Fit 6 0.21210 0.21210 0.03535 1.90 0.198
3.750 1.250 254 2 10 0.167 Pure Error 8 0.14918 0.14918 0.01865
3.750 1.250 336 8 5 0.615 Total 17 1.30866

After Analysis of RSM model , we realise that PSP is not a vital few factor and hence is
removed from the regression equation.

Innovation with 6  Achieve Super A! LG Electronics

 6
Improve
Improve / SS Cavity (Wireframe Plot) Define Measure Analyse Control

By the Wire Frame plot we establish the best condition for SS-cavity

Surface Plot of r ( %age O.K.)

1.0

0.5
r

330340
330
0.0 320
310
300
270
290
280
270
voltage
3 260
260
250 Best Condition :
PG 4
5
Voltage :310
PG :2.8

Innovation with 6  Achieve Super A! LG Electronics

 6
Improve
Improve / GpSp Cavity Define Measure Analyse Control

We know from past setting experience that three parameters (PG - Pressure Gauge. PSP &
Voltage ) have a crucial impact on welding quality. We did 2 level factorial design to establish the
vital few.
Main Effects Plot (data means) for resp

pg psp volt ok rej resp

0 0
3.0 4 .5 1 .0 1 .5 27 32

0.24

0.18

3.00 1.00 270.00 3.00 9.00 0.25

resp
0.12

0.06

4.50 1.00 270.00 0.00 13.00 0.00 0.00

pg psp volt

3.00 1.50 270.00 4.00 8.00 0.33 Interaction Plot (data means) for resp

4.50 1.50 270.00 0.00 13.00 0.00 pg 0.30

3.00 1.00 320.00 0.00 13.00 0.00

4.5
0.15
3
0.00

4.50 1.00 320.00 0.00 12.00 0.00

psp 0.30

1.5
0.15

3.00 1.50 320.00 5.00 8.00 0.38 volt

0.00

4.50 1.50 320.00 0.00 13.00 0.00

We conclude that PG & PSP seem to have more effect than voltage in this case.There is
also mild interaction between all factors.

Innovation with 6  Achieve Super A! LG Electronics

 6
Improve
Improve / GpSp Cavity Define Measure Analyse Control

We have done RSM to optimise the welding machine settings for GpSp cavity.We have used the
Central composite design.

Response Surface Regression: resp versus pg, psp

pg psp volts ok rej resp The analysis was done using coded units.
3.00 1.00 270 3 9 0.250
4.50 1.00 270 0 13 0.000 Estimated Regression Coefficients for resp

3.00 1.50 270 4 8 0.333 Term Coef SE Coef T P

4.50 1.50 270 0 13 0.000 Constant 0.0309 0.03810 0.812 0.432
3.00 1.00 320 0 13 0.000 Block -0.0368 0.02865 -1.284 0.222
pg -0.1856 0.03113 -5.963 0.000
4.50 1.00 320 0 12 0.000 psp 0.0634 0.03113 2.035 0.063
3.00 1.50 320 5 8 0.385 pg*pg 0.1402 0.03262 4.298 0.001
4.50 1.50 320 0 13 0.000 S = 0.1137 R-Sq = 82.6% R-Sq(adj) = 77.2%
3.75 1.25 295 1 11 0.083
3.75 1.25 295 0 13 0.000 Analysis of Variance for resp
3.75 1.25 295 0 12 0.000 Source DF Seq SS Adj SS Adj MS F P
3.75 1.25 295 0 13 0.000 Blocks 1 0.043849 0.021291 0.021291 1.65 0.222
2.53 1.25 295 12 1 0.923 Regression 3 0.751804 0.751804 0.250601 19.39 0.000
Linear 2 0.513070 0.513070 0.256535 19.85 0.000
4.97 1.25 295 0 12 0.000 Square 1 0.238735 0.238735 0.238735 18.47 0.001
3.75 0.84 295 0 13 0.000 Residual Error 13 0.168013 0.168013 0.012924
Lack-of-Fit 5 0.130239 0.130239 0.026048 5.52 0.017
3.75 1.66 295 3 10 0.231
Pure Error 8 0.037773 0.037773 0.004722
3.75 1.25 254 0 12 0.000 Total 17 0.963666
3.75 1.25 336 0 13 0.000
We conclude that the main effect is of PG(Pressure gauge) & PSP has some effect.But
voltage which was very important in SS case is not important in GpSp.
Innovation with 6  Achieve Super A! LG Electronics
 6
Improve
Improve / GpSp Cavity Define Measure Analyse Control

Through Wire frame plot we establish the best condition for GpSp Cavity.

1.0

0.5

resp
0.0
1.61.7
1.5
1.31.4

3 1.01.1
1.2
psp
0.9
4 0.8 Best Condition :
pg 5
PSP :1.3
PG :2.8

Innovation with 6  Achieve Super A! LG Electronics

Control / GpSp Cavity Define Measure Analyse Improve Control 6
We have chosen the np chart for number of defective projection welding points (checked during
time check & setup) .
This plot shows the number defective for the last 20 lots.

NP Chart f or NG
1.5

UCL= 1. 305

1.0
Sample Count

0.5

NP= 0.15
0.0 LCL= 0

0 10 20
Sample Number

Remark: Upto Two weak dimples ( welding joints are acceptable)

Innovation with 6  Achieve Super A! LG Electronics

Rejection Trend Define Measure Analyse Improve Control 6
PPM

80000
60000 6  Project
40000
20000 PPM
0
March
January

May

July
November

Innovation with 6  Achieve Super A! LG Electronics