Professional Documents
Culture Documents
=
=
6
LSL USL
Tolerance Natural
Tolerance g Engineerin
C
p
LSL USL
Copyright 2009 HKRITA. All rights reserved.
Process Potential
Historically, a C
p
of 1.0 has indicated that a process is
judged to be capable, i.e. if the process is centered
within its engineering tolerance, 0.27% of parts produced
will be beyond specification limits.
C
p
Reject Rate
1.00 0.270 %
1.33 0.007 %
1.50 6.8 ppm
2.00 2.0 ppb
Copyright 2009 HKRITA. All rights reserved.
Process Potential
a)
b)
c)
a) Process is highly capable (Cp>2)
b) Process is capable (Cp=1 to 2)
c) Process is not capable (Cp<1)
Copyright 2009 HKRITA. All rights reserved.
Process Potential
he C
p
index compares the allowable spread (USL-LSL)
against the process spread (6 ). It fails to take into
account if the process is centered between the
specification limits.
Process is centered Process is not centered
Copyright 2009 HKRITA. All rights reserved.
Process Performance
The C
pk
index relates the scaled distance
between the process mean and the nearest
specification limit.
=
3
USL
C
pu
=
3
LSL
C
pl
{ }
pl pu pk
C C Minimum C , =
Copyright 2009 HKRITA. All rights reserved.
Process Performance
C
pk
Reject Rate
1.0 0.13 0.27 %
1.1 0.05 0.10 %
1.2 0.02 0.03 %
1.3 48.1 96.2 ppm
1.4 13.4 26.7 ppm
1.5 3.4 6.8 ppm
1.6 794 1589 ppb
1.7 170 340 ppb
1.8 33 67 ppb
1.9 6 12 ppb
2.0 1 2 ppb
Copyright 2009 HKRITA. All rights reserved.
Process Performance
a) Process is highly capable (Cpk>1.5)
b) Process is capable (Cpk=1 to 1.5)
c) Process is not capable (Cpk<1)
a)
Cp = 2
Cpk = 2
b)
Cp = 2
Cpk = 1
c)
Cp = 2
Cpk < 1
Copyright 2009 HKRITA. All rights reserved.
Example 1
Specification Limits : 4 to 16 g
Machine Mean Std Dev
(a) 10 4
(b) 10 2
(c) 7 2
(d) 13 1
Determine the corresponding C
p
and C
pk
for each
machine.
Copyright 2009 HKRITA. All rights reserved.
Example 1A
( )
5 . 0
4 6
4 16
6
LSL USL
C
p
=
=
( ) ( )
5 . 0
4 3
4 10
;
4 3
10 16
Min
3
LSL
;
3
USL
Min C
pk
=
=
Copyright 2009 HKRITA. All rights reserved.
Example 1B
( )
0 . 1
2 6
4 16
6
LSL USL
C
p
=
=
( ) ( )
0 . 1
2 3
4 10
;
2 3
10 16
Min
3
LSL
;
3
USL
Min C
pk
=
=
Copyright 2009 HKRITA. All rights reserved.
Example 1C
( )
0 . 1
2 6
4 16
6
LSL USL
C
p
=
=
( ) ( )
5 . 0
2 3
4 7
;
2 3
7 16
Min
3
LSL
;
3
USL
Min C
pk
=
=
Copyright 2009 HKRITA. All rights reserved.
Example 1D
( )
0 . 2
1 6
4 16
6
LSL USL
C
p
=
=
( ) ( )
0 . 1
1 3
4 13
;
1 3
13 16
Min
3
LSL
;
3
USL
Min C
pk
=
=
Copyright 2009 HKRITA. All rights reserved.
Process Potential vs Process Performance
(a) Poor Process Potential (b) Poor Process
Performance
LSL USL LSL USL
Experimental Design
to reduce variation
Experimental Design
to center mean
to reduce variation
Copyright 2009 HKRITA. All rights reserved.
Process Potential vs Process Performance
a)
Cp = 2
Cpk = 2
b)
Cp = 2
Cpk = 1
c)
Cp = 2
Cpk < 1
Cp Cpk Missed Opportunity
Copyright 2009 HKRITA. All rights reserved.
Process Stability
A process is stable if the distribution of
measurements made on the given feature is
consistent over time.
Time
Stable Process
Time
Unstable
Process
Copyright 2009 HKRITA. All rights reserved.
Graphical Representation of
Causes of Variations (Jurans Trilogy)
F
o
c
u
s
o
f
S
i
x
S
i
g
m
a
Common Cause ?
.
UCL
LCL
CL
Special Cause ?
.
UCL
LCL
CL
Copyright 2009 HKRITA. All rights reserved.
Steps to Study Process Capability
Select critical parameters for study
Parameters from specifications, contract etc.
Collect Data
Collect 60 data or more as far as possible
Define clearly the precision of each data (no. of significant figures, eg
up to 2 decimal places)
Establish control
Control the input to the process
Analyze the data of the process collected
Assumption : The process performance is a normal distribution
Focus on mean and standard deviation of sample data
Analyze the source of variation
Find the factors that affect the process mean and process spread
(standard deviation)
Establish process monitoring system
Tool Statistical Process Control
Copyright 2009 HKRITA. All rights reserved.
Summary on Indexes
Capability
index
Formula Short or
long term
Includes
shift
and drift
Considers the
process
centering
C
p
USL LSL
6
ST
Short term No No
C
pk
Short term
No Yes
P
p
USL LSL
6
LT
Long term Yes No
P
pk
Long term Yes Yes
Copyright 2009 HKRITA. All rights reserved.
Example 2 Customer request a metal bar
from 2 suppliers
Customer require 2mm +/- 0.1mm
Supplier B
Supplier A
2 +/- 0.1 mm
Lower Specification Limit
(LSL = 1.9mm)
Upper Specification Limit
(USL = 2.1mm)
Cp = 0.62
Cp = 1.11
Mean
(= 2 mm)
Supplier A
Supplier B
Copyright 2009 HKRITA. All rights reserved.
Cp can not reflect the condition of the center shift !!
If the center of distribution was shifted, customer may not happy even receive a high Cp value.
Location Change
Lower SpecificationLimit
(LSL)
Upper SpecificationLimit
(USL)
Mean
()
Out of
specification
Same Cp value
Distribution A
Distribution B
Copyright 2009 HKRITA. All rights reserved.
CpK Process Capability Index
A measure of conformance (capability) to
specification
Compares sample mean to nearest specification
against distribution width
CpK can more precisely reflect the capability of distribution.
Copyright 2009 HKRITA. All rights reserved.
Example 3
Process Variation on Two Suppliers
LSL = 1.9 mm
USL=2.1mm
Mean
(= 2.0 mm)
= 1.92 mm
CpK = 0.22 CpK = 1.11
= 2.00 mm
Supplier A Supplier B
Remark:
Same SD but different Central
Tendency affects Cpk seriously
but remains same for Cp.
Supplier A
Cp = 1.11
Supplier B
Cp = 1.11
Copyright 2009 HKRITA. All rights reserved.
- THE END -