Process Capability

Process Capability Ratio (PCR) CP

Cp= USL-LSL / 6σ

Cp is the quality characteristic with

both Upper & lower Specification
limit, respectively. Note that the 6σ
spread of the process is the basic
definition of process capability since
σ is always unknown
1
 More about process centering
‹ The process capability ratio was
initially developed because Cp does
not adequately deal with the case of
a process with mean centering

2
 Practical interpretation
‹ To illustrate the calculation of Cp, let
USL= 74.05mm and LSL= 73.95mm, and
from R chart we estimated σ=R/d2=
0.0099. Thus our estimate of the PCR Cp
is USL-LSL / 6σ
‹ Cp=74.05-73.95 / 6(0.0099)= 1.68

‹ This implies that the “natural” tolerance

limit in the process (three sigma above
and below the mean) are well inside lower
and upper specification limits. Thus the
process produces approximately 20ppm
(parts per million)
3
 Interpretation in quantity
‹ P=(1/Cp) 100% is simply the
percentage of the specification band
the process uses up.
‹ P= (1/1.68)100% = 59.5 i.e.. the
process uses up approx. 60%. Of the
specification band.

4
 Case –I Cp>1 PCR Cp is greater than unity this
means that the process uses up less
than 100% of the tolerance band.
Consequently, relatively few non-
conforming units will be produced by
the process.

LSL LNTL µ UNTL

USL
3σ 3σ

5
 Case –II Cp=1 PCR Cp equals unity this means that
the process uses up all the
tolerance band. For normal
distribution this would imply
about 0.27% (or 2700 ppm)
non-confirming units.

LNTL µ UNTL
LSL USL
3σ 3σ

6
 Case –III Cp<1 PCR Cp less than unity, this means
that the process uses up more than
100% tolerance band. In this case the
process is very yield-sensitive, and a
large number of non-confirming units
will be produced.

LNTL LSL µ USL UNTL

3σ 3σ

Note that all the three cases in the above figure

assume that the process is centered at the mid
point of the specification band. In many
situations this will not be the case. Hence
extensive treatment of process capability analysis
is necessary to describe this situation adequately.

7
Values of Process capability Ratio (Cp) & associated process fallout for a normally

.
Distributed Process (in defective ppm) That is in statistical control

Process fallout (in defective ppm)

PCR One-sided Specn Two-sided Specn

0.25 226,628 453,255

0.50 66,807 133,614
0.60 35,931 71,861
0.70 17,865 35,729
0.80 8,198 6,395
0.90 3,467 6,934
1.00 1,350 2,700
1.10 484 967
1.20 159 318
1.30 48 96
1.40 14 27
1.50 4 7
1.60 1 2
1.70 0.17 0.34
1.80 0.03 0.06
2.00 0.0009 0.0018
8
 The process capability ratio is a measure of the
ability of the process to manufacture that meets
the specifications
‹ The ppm quantities were calculated
using the following important
assumptions:
‹ 1. The quality characteristic has a
normal distribution.
‹ 2. The process is in statistical
control.
‹ 3. In case of two-sided
specifications, the process mean is
centered between the lower &
upper specification limit. 9
 Recommended guide lines for
minimum values of PCR
Two-sided One-sided
Specifications Specifications

Existing processes 1.33 1.25

New Processes 1.50 1.45

Safety,strength,or 1.50 1.45

Critical parameter
Existing process.

Safety,strength,or 1.67 1.60

Critical parameter
new process.
10
 LSL USL

±3σ
99.73%

-6σ -5σ -4σ -3σ -2σ -1σ µ=T +1σ +2σ +3σ +4σ +5σ +6σ

Spec Limit Percent inside Specs ppm Defective

±1Sigma 68.27 317300
±2Sigma 95.45 45500
±3Sigma 99.73 2700
±4Sigma 99.9937 63
±5Sigma 99.999943 0.57
±6Sigma 99.9999998 0.002

Normal distribution centered at the target (T)

11
 1.5 σ 1.5 σ

LSL USL

±3σ
99.73%

-6σ -5σ -4σ -3σ -2σ -1σ T +1σ +2σ +3σ +4σ +5σ +6σ

Spec Limit Percent inside Specs ppm Defective

±1Sigma 30.23 697700
±2Sigma 69.13 308700
±3Sigma 93.32 66810
±4Sigma 99.3790 6210
±5Sigma 99.97670 233
±6Sigma 99.999660 3.4

Normal distribution with mean shifted by 1.5 σ from the target

The Motorola six-sigma concept 12
 LSL USL
σ=2 Cp=2.0
a Cpk=2.0
The process capability ratio Cp
does not take into account Where the
process mean is located relative to
38 44 50 56 62 the specification. Cp simply measures
Cp=2.0 the spread of the specifications
σ=2 Cpk=1.5
relative to six-sigma spread in the
b
process. In the adjacent figure a & b
38 44 50 56 62
both have Cp=2, but the process in
both the figure clearly has lower
σ=2 Cp=2.0 capability than the process in figure
Cpk=1.0
“a” because it is not operating at the
c
mid point of the interval between the
38 44 50 56 62 specifications.
This situation may be more
σ=2 Cp=2.0
accurately reflected by defining new
d Cpk=0
process capability ratio that takes
38 44 50 56 62 process centering into account. This
quantity is Cpk= min (Cpu, Cpl).
e σ=2 Cp=2.0
Note that Cpk is simply the one-
Cpk=0.5
sided PCR for the specification
38 44 50 56 62 65 limit nearest to the process
13
average .
 Cpk= min (Cpu,Cpl) if µ= 53 & σ=2

=min (Cpu=USL-µ/3σ),(Cpl=µ-LSL/3σ)

=min (Cpu=62-53/3x2=1.5),(Cpl=53-38/3x2=2.5)
53

=1.5

Generally, if Cp=Cpk, the process is

centered at the mid point of the
specification & when Cpk<Cp the
process is off-center.
14
 Process Performance
The magnitude of Cpk relative to Cp is a direct
measure of how off-center the process is operating.
Several commonly encountered cases are in slide 12
to get a quick estimate of potential improvement that
would be possible by centering the process. If we take
Cp=1.0 (ref. slide 7) and read the fallout from one-
sided specification column, we can estimate the actual
fallout as 1350 ppm. However ,if we can center the
process, then Cp=2.0 can be achieved. Thus the table
suggests that the potential fallout is 0.0018 ppm, an
improvement of several orders of magnitude in
process performance. Thus we usually say that Cp
measures potential capabilities in the process,
where as Cpk measures actual capabilities

15
 More about Process centering
The process capability ratio Cpk was
initially developed because CP does not
adequately deal with the case of a
process with mean µ that is not centered
between the specification limits. However
Cpk alone is still an inadequate measure
of process centering.
Cpm is called the Taguchi capability
index after the Japanese quality guru.

16
 More about Process centering
Genichi Taguchi whose work on the Taguchi
Loss Function stressed the economic loss
incurred as process departed from the target
value. This index was developed in late 1980’s
and takes into account the proximity of the
process mean to a designated target T. where

T= ½ (USL+LSL)

Cpm= USL-LSL/6√σ+(µ-T)

17
 T= ½ (USL+LSL),
Cpm = USL-LSL/6√σ² +(µ-T) ²

σ=0.8

LSL=18 USL=24

µ=22

T=1/2(24+18)= 21 Cpm= 24 -18/6 √0.8²+(22-21)²

Cpm= 1/√1.64 =1/1.281=0.78
When the process mean is centered
between specification limits or the process mean is on
the target T then Cp=Cpk=Cpm. That is to say µ=T
Also, Cpm decreases as µ moves away from T.
18
 Books For Reading
‹ Introduction to
Statistical Quality Control
By Douglas C.Montgomery.

‹ Statistical Quality Control

By E.L. Grantt

‹ Facts From Figures

By M.J.Moroney.

19