SPC

STATISTICAL PROCESS CONTROL

DISCUSSION BY

GP CAPT Y RAJA RAO(RETD)
-BE-MECH OU-1979;
-MS in QUALITY MGT-2001(iiqm-JAIPUR/BITS PILANI);
-SEC-SYSTEMS ENGG COURSE;
-TSOC-TECH.STAFF OFFICERS COURSE-IAT PUNE;
-SMOC-SENIOR MGT ORIENTATION COURSE.
-WITH IAF 26+ YRS (FROM 1979-2005)
-CO-OPTED MEMBER AIRCRAFT ACCIDENT INVESTIGATIONS.
-CO-OPTED MEMBER DTE OF AIR-STAFF INSPECTION.
-CQA-CHIEF OF QUALITY ASSURANCE-11 BRD,OZHAR NASIK.

-I/C QA/QC NICKEL GROUP-HBL SINCE Nov-2005

SPC
STATISTICS AND INDUSTRY- The long range contribution of
Statistics depends not so much upon getting a lot of highly
trained statisticians into industry, as it does on creating a
statistically minded generation of engineers, chemists,
physicists and others who have a hand in developing and
directing production processes.
SHEWART, DEMING-1939
IT IS TRUE EVEN TODAY-
Statistics is the core of SIX SIGMA; LEAN MFG initiatives

-are we statistically minded at HBL?
The question will be repeated after this
discussion!!!


SPC
Child Games?

Riding bike/ cycle without holding the
handle.(before and after servicing)
Man on scooter with weight on one-side.
Car-driving without holding the steering wheel.
(wheel balancing, alignment).
Newtons first law.
SPC
QUALITY -A GOAL-POST APPROACH-ACCEPTANCE OF WHAT IS WITHIN SPECS.
Have we ever measured how close to the NOMINAL VALUE- A PROCESS OUTPUT IS?-
CASE OF -SONY TV.
CASE OF DOMESTIC POWER SUPPLY-230 +/- 10V





LSL
NOMINAL VALUE-X BAR
USL
SAN DIEGO-US
JAPAN
SONY TV
COLOUR/ CONTRAST-POTMETER
ADJUSTMENT
FAIL
URE
ZO
NE`
FAIL
URE
ZON
E
SPC
DAILY FIGHT-
HAVE WE DONE THE JOB CORRECTLY
-UNDESIRABLE POST-MORTEM
PRODUCED
ITEMS
GOOD!!
-Prodn
Mgr
BAD ?
CUSTOMER,
QA/QC.
SPC

Post-Production INSPECTION (POST MORTEM)/ detection-
Is unproductive.

Serves short sighted approach of limiting the number defectives
reaching customer/ progressing to next stage of mfg.

Checking, rechecking, 100%-checking, chasing/
expediting.apologising for short supply, delayed delivery etc
etc is WASTE, WASTE and MORE WASTE.

ONE CANNOT BUILD QUALITY BY INSPECTION, INTO WHAT IS ALREADY PRODUCED,
SPC
THE QUESTION HAVE WE DONE THE JOB CORRECTLY

ONE SHOULD ACTUALLY START WITH CAN WE DO THE JOB
CORRECTLY

THE SOLE AIM SHOULD BE TO PRODUCE A DEFECT FREE PRODUCT/ SERVICE
THE VERY FIRST TIME, EVERY TIME-CONSISTANTLY


SPC


WHAT IS A PROCESS- IT IS TRANSFORMATION OF A SET OF INPUTS
INTO DESIRED OUTPUT.

THE OUTPUT IS TRANFERRED TO NEXT STAGE OR FINALLY TO CUSTOMER.

AT EVERY SUPPLIER-CUSTOMER INTERFACE (NEXT STAGE OR END-USER)
IS A PROCESS.

EVERY TASK OR AN ACTIVITY IN AN ORGANISATION IS A PROCESS.


SPC
PROCESS
TRANSFORMING A SET OF INPUTS (MATERIAL, ACTIONS, METHODS/ OPERATIONS)
INTO DESIRED OUTPUT(PRODUCT, INFORMATION, SERVICE)
Applicable to all areas of organisations-m/c shops, stores, data entry, sales-
tendering, finance, purchase, training, invoicing-billing, banking, absenteeism,
injuries, even giving a lecture.



PROCESS
MATERIALS INCLUDING PAPER
METHODS,PROCEDURES/INSTRUCTIONS
SPECS, INFORMATION
MAN-TRG, SKILLS, KNOWLEDGE
ENVIRONMENT
RECORDS
EQUIPMENT, PLANT, TOOLS
PRODUCT
SERVICES
INFORMATION
PAPER
WORK
INPUTS
OUTPUTS
SPC

What is CONTROL-data gathering measurement of performancefeed-
back corrective actions.

A. HAVE WE DONE THE JOB CORRECTLY-BY INSPECTION

1.CAN WE DO THE JOB CORRECTLY IS CAPABILITY
2.ARE WE DOING CORRECTLY-operator, eqpt, tools, procedure
ISCONTROL

IF THE ANSWER TO 1, 2 QUESTIONS IS YES THEN THE NEED TO CARRY
OUT INSPECTION TO ANSWER QUESTION-A DOES NOT ARISE

Ensuring INPUTS-CAPABLE OF MEETING PROCESS REQUIREMENTS-is a
managerial responsibility-A QUALITY ASSURANCE FUNCTION




SPC



CONTROL OF QUALITY-CAN TAKE PLACE ONLY AT THE POINT OF
TRANSFORMATION OF INPUTS TO OUTPUT, AT THE POINT OF
OPERATION OR PRODUCTION WHERE A LETTER IS TYPED OR
AN ITEM IS MADE. The act of product-inspection is NOT
Quality control.


SPC
SPC: STAGE -WISE CONTINUAL INPROVEMENT.

CAN WE DO THE JOB CORRECTLY- Knowing whether the
process is capable of meeting the requirements-
CAPABILITY.

ARE WE DOING THE JOB CORRECTLY-Knowing whether
the process IS meeting the requirements at point in
time-adjust if required-CONTROL.

Have we done the job correctly-QUALITY ASSURANCE.

COULD WE DO THE JOB BETTER-IMPROVEMENT

SPC



SOME OF THE COMMON PROBLEM SOLVING TOOLS
FOR SPC ARE-

DATA GATHERING.-STRATIFICATION.
HISTOGRAMS/ PIE CHARTS
PARETO ANALYSIS.
CAUSE AND EFFECT DIAGRAMS.
SCATTER DIAGRAM.
CONTROL CHARTS.


SPC


STATISTICAL PROCESS CONTROL- IS A PHILOSOSPHY, A TOP MGT STRATEGY
FOR CONTINUAL IMPROVEMENT
THE AIM IS
TO REDUCE VARIABILITY-THE ROOT CAUSE OF MOST QUALITY PROBLEMS
VARIATION OCCURS- IN PRODUCTS-
IN TIMES OF DELIVERY-
IN WAYS OF DOING THINGS-
IN MATERIALS-
IN EQPT IN USE-
IN MAINT. PRACTICES-
IN PEOPLES ATTITUDE..




SPC
Sources of Variation

Man everybody does things differently.
Machines using different machines even different
individual machines of the same model
might produce different results.
Method different methods emphasize different things.
Materials different materials effect results.
Environment differences in temperature, humidity, etc.,
can affect results.
Measurement system the more accurate and precise
the measurement tools- smaller or lesser the range of
variation.
SPC
Causes of Variation
Special Causes :- Anything that causes variations that
is not part of the stable process is called a special
cause, assignable cause will be able to eliminate
using Cause-Effect /WHY-WHY/8D analysis, without
much effort.
Random/ Common Causes :- The cause of variations
in a stable process is called a Common Cause. A
common cause is an inherent/ natural variation in
the system.-will be difficult to eliminate one can
only minimise-thus the effort towards SIX SIGMA
Random variation-
Common Causes
Special Causes
Total Variation
Product
Measurement System
Repeatability Reproducibility
Man Methods Materials Machines Environment
Variable error Attribute error
Visual errors
Repeatability & Reproducibility
Repeatability : Variation that results when one person
measures the same sample over and over and over using
same measuring instrument
Same operator
Same instrument
Same set-up procedure
Same part
Same environmental conditions
Reproducibility :The variation that results when others
measuring the same sample using the same measuring instrument
Different operator
Same instrument
Same set-up procedure
Same part
Same environmental conditions
SPC
SPC
Mean
SELECTION OF CONTROL CHARTS
SPC
ATTRIBUTES: Cannot be measured as Variable. Therefore less sensitive to Variation.
To be counted as present or absent.
Conforming/ non conforming.

Defectives-individual flaw, error, or fault.-Binomial Distn
np chart-Number of non-conforming units/ constant sample size. e.g
defective plates for a constant batch size-say 1000 plates.
p chart-Proportion of non-conforming units per varying sample/ size. E.g
% of plating discolouration for tabs of different batch size

Defects- contains one or more flaws, errors, faults or defects. Poisson or
exponential distn
c chart-Number of defects for constant sample size. E.g. Number welding
splatter/defects on a PE panel of regular size 2.5mx3m. Number of
errors in tender, invoice, PO, Prodn plan, customer queries/order etc
U chart Number of defects per unit-for varying sample size. Cases of dry
particles/ sintered area 10cmx10cm
22

Control Chart methodology
Establish state of statistical control
Monitor process and signal when it
goes out of control
Determine process capability
Developing Control Charts-variables
Collection of data
Calculate mean (X-bar) for each sample
Calculate grand mean (X-double-bar)
Calculate range (R) for each sample
Calculate mean of all sample ranges (R-bar)
Calculate UCL and LCL for means
Plot grand mean and control limits on X-bar
chart
HBL POWER SYSTEMS LTD.
NCPP DIVISION
Statistical Process Control
Brequitting section Type : KBH 590(- Ve Briquette)
Brequitte weight Spec: 3.00 3.0%
Brequitting machine : 04 Interval time : 15 minites
2.30
2.4
5
3.00 3.15 3.30 3.45 4.00 4.15 4.30 4.45 5.00 5.15
5.3
0
5.4
5
6.0
0
6.15 6.30 6.45 7.00 7.15 7.30
2.97 3.06 2.99 3.03 3.09 3.08 3.10 3.05 2.98 3.02 3.04 2.99 3.02 2.98 3.06 3.02 3.04 2.99 3.02 2.98 3.06
2.97 3.09 3.03 3.00 3.10 3.08 3.06 3.04 2.99 3.06 3.01 3.02 3.00 3.01 3.04 3.06 3.01 3.02 3.00 3.01 3.04
3.01 3.06 3.02 3.02 3.08 3.06 3.08 3.04 2.99 2.99 2.98 2.98 3.05 3.06 3.08 2.99 2.98 2.98 3.05 3.06 3.08
3.05 3.09 3.02 3.05 3.08 3.10 3.04 2.99 3.03 3.01 3.04 2.97 3.03 3.07 3.01 3.01 3.04 2.97 3.03 3.07 3.01
3.01 3.04 2.98 3.03 3.04 3.14 3.02 3.03 3.02 3.04 2.99 3.02 2.99 3.06 3.09 3.04 2.99 3.02 2.99 3.06 3.09
2.97 3.02 3.01 3.02 3.06 3.09 3.05 3.02 3.00 3.00 3.04 2.99 3.01 3.05 3.09 3.00 3.04 2.99 3.01 3.05 3.09
2.96 3.00 3.04 3.00 3.05 3.12 3.09 3.06 2.98 2.98 3.07 3.04 3.02 3.03 3.06 2.98 3.07 3.04 3.02 3.03 3.06
3.01 2.98 3.02 3.02 3.09 3.07 3.06 3.05 3.03 3.02 3.03 3.03 3.06 3.03 3.00 3.02 3.03 3.03 3.06 3.03 3.00
X bar 2.99 3.04 3.01 3.02 3.07 3.09 3.06 3.04 3.00 3.02 3.03 3.01 3.02 3.04 3.05 3.02 3.03 3.01 3.02 3.04 3.05
R 0.09 0.11 0.06 0.05 0.06 0.08 0.08 0.07 0.05 0.08 0.09 0.07 0.07 0.09 0.09 0.08 0.09 0.07 0.07 0.09 0.09
X Double Bar
3.031
R Bar 0.078
Control Limit Factor
Table
( n ) ( A
2
) ( D
3
) ( D
4
)
Std Deviation
= R bar /d2;
d2= 2.059
UCL = X dble bar + A2*R 2
1.8
8
.. 3.27
0.023
3.060 3
1.0
2
.. 2.57
Process Capability
Cp = USL-LSL/ 6 LCL = X dle bar - A2 * R 4
0.7
3
.. 2.28
1.528 3.002 5
0.5
8
.. 2.11
CpL = X - LSL/3 R UCL = D4 * R 6
0.4
8
.. 2.00
-12.36 0.144 7
0.4
2
0.08
1.
92
CpU = USL- X /3 R LCL = D3 * R 8
0.3
7
0.14 1.86
15.41 0.011 9
0.3
4
0.18 1.82
Process Capability Cpk =min (CpU: CpL)
10
0.3
1
0.22
1.
78
Index
-
12.36
3.060
3.031
3.003
0.144
0.078
0.011
3.25
3.30
3.35
3.40
3.45
3.50
3.55
3.60
3.65
15.30 15.45 16.00 16.15 16.30 16.45 17.00 17.15 17.30 17.45 18.00 18.15 18.30 18.45 19.00
X - bar Chart for Briquette weights of B.M/c-4 X-Bar
X-Bar Avg
USL
LSL
X UCL
X LCL
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
15.30 15.45 16.00 16.15 16.30 16.45 17.00 17.15 17.30 17.45 18.00 18.15 18.30 18.45 19.00
R Bar Chart for Briquette weights of B.M/c-4
R Bar
R Bar Avg
R UCL
R LCL
Normal Distribution
=0 1o 2o 3o -1o -2o -3o
95%
99.74%
How does one decide if a point is part of
an uncontrolled process?

One point beyond the 3control limit
2 out of 3 consecutive points plot outside of the 2 control limit
4 out of 5 consecutive points plot outside of the 1 control limit
9 consecutive points on the same side of the centerline
6 consecutive points increase or decrease
14 consecutive points oscillate up and down
15 consecutive points inside of the 1 control limit.
These are the Western Electric Rules
35
Typical Out-of-Control Patterns
Point outside control limits
Sudden shift in process average
Cycles
Trends
Hugging the center line
Hugging the control limits
Instability
Shift in Process Average
Identifying Potential Shifts
Cycles
Trend
Process Capability
Process capability compares the output of an in-control process to the
specification limits by using capability indices. The comparison is made by
forming the ratio of the spread between the process specifications (the
specification "width") to the spread of the process values, as measured by
6 process standard deviation units (the process "width").
mm mm mm
C
LTL UTL
C
p
p
0868 . 0 25 . 75 . 10 is ion specificat Part : Example
minimum) often the more (1.5 capable as defined is 1 if
6
=
>

=
o
o
96 . 0
0868 . 0 6
50 . 10 00 . 11
=
-

=
p
C
( )
Target the is T
1
2
2
o
T
C
C
p
pm

+
=
( )
( )
10.7171mm at centered is process assume but above, as same : Example
2
where 1
, min
3
3
Tolerance
T
K K C C
C C C
LTL
C
UTL
C
p pk
pu pl pk
pl
pu

= =
=

Process Capability CONTINUED..

086 . 1
0868 . 0 3
7171 . 10 0 . 11
=
-

=
pu
C
834 . 0
0868 . 0 3
5 . 10 7171 . 10
=
-

=
pl
C
( )
8977 . 0
868 . 0
75 . 10 7171 . 10
1
960 . 0
2
2
=

+
=
pm
C
46
Capability Versus Control
Control
Capability
Capable


Not Capable
In Control Out of Control
IDEAL
47
Control Chart Selection
Quality Characteristic
variable attribute
n>1?
n>=10 or
computer?
x and MR
no
yes
x and s
x and R
no
yes
defective defect
constant
sample
size?
p-chart with
variable sample
size
no
p or
np
yes
constant
sampling
unit?
c u
yes no
48
SPC Implementation Requirements
Top management commitment
Project area selection
Initial workable project
Employee education and training
Accurate measurement system
Some Recollections on session.....

1. What is mean by variation ?
2. What are the sources of variation ?
3. What is mean by SPC ?
4. Where to apply SPC ?
5. Which are set of statistical tools ?
6. What is Repeatability and Reproducibility ?
7. How to develop control charts ?
8. What is the formula of Cp and Cpk ?
9. What are the western electric rules ?
SPC
OLD OR CURRENT DIALOGUE:
CUSTOMER: YOUR PRODUCT QUALITY IS POOR Vs YOUR COMPETITOR.
SALES: I WILL LOWER THE PRICE.
CUSTOMER: GOOD THAT WAS WHAT I WANTED.

LATEST OR FUTURE DIALOGUE:
CUSTOMER: YOUR PRODUCT QUALITY IS POOR Vs YOUR COMPETITOR.
SALES: I WILL LOWER THE PRICE.
CUSTOMER: PRICE IS OK. QUALITY IS NO GOOD.
SALES: WAS ANY OF OUR PRODUCT OFF-SPEC?
CUSTOMER: NO.
SALES: I DO NOT UNDERSTAND.
CUSTOMER: YOUR DATA IS VERY CLOSE TO TOLERANCE. THERE IS LARGE
VARIABILITY. YOUR COMPETITOR IS MUCH LESS VARIABLE.

Are we ready to answer?
SPC
LESSER VARIATION- A MEASURE OF GOOD QUALITY
USL
LSL
HBL COMPETITOR
Are we ready to answer?

mean
ARE WE STATISTICALLY MINDED????
THANKS

Constants for Xbar and R chart
n A2 D3 D4 d2
2 1.880 0 3.268 1.128
3 1.023 0 2.574 1.693
4 0.729 0 2.282 2.059
5 0.577 0 2.114 2.326
6 0.483 0 2.004 2.534
7 0.419 0.076 1.924 2.704
8 0.373 0.136 1.864 2.847
9 0.337 0.184 1.816 2.970
10 0.308 0.223 1.777 3.078


SPC
IMPLEMENTATION
BENEFITS
REDUCED COST OF EXTERNAL FAILURE.
REDUCED COST OF INTERNAL FAILURE AND WASTE.
HIGHER PRODUCTIVITY.
BETTER CONSISTENCY, MARKET SHARE AND IMAGE.


SPC
ENABLERS
KNOWLEDGEABLE, ENTHUSIASTIC AND
COMMITTED TOP MGT.
ALLOCATION OF ADEQUATE RESOURCES.
TRAINING( THIRD PARTY),