PROCESS CAPABILITY

Pradeep Kumar
Professor
Mechanical & Industrial Engineering
Indian Institute of Technology Roorkee
 Common Questions for Investigating an
Out-of-Control Process

• Are there differences in the measurement accuracy of instruments /

methods used?
• Are there differences in the methods used by different personnel?
• Is the process affected by the environment, e.g. temperature/humidity?
• Has there been a significant change in the environment?

3
 Common Questions for Investigating an
Out-of-Control Process

• Is the process affected by predictable conditions such as tool wear?

• Were any untrained personnel involved in the process at the time?
• Has there been a change in the source for input to the process such as
a new supplier or information?
• Is the process affected by employee fatigue?

4
 Common Questions for Investigating an
Out-of-Control Process

• Has there been a change in policies or procedures such as maintenance

procedures?
• Is the process frequently adjusted?
• Did the samples come from different parts of the process? Shifts?
Individuals?
• Are employees afraid to report “bad news”?

5
 Estimating Sigma
The Control Chart Method for Variables Data

1. Construct the control chart and remove all

special causes.

NOTE: special causes are “special” only in that

they come and go, not because their impact is
either “good” or “bad”.

6
 Estimating Sigma
The Control Chart Method for Variables Data

2. Estimate the standard deviation. The approach used

depends on whether a R or S chart is used to monitor
process variability.

^ _ ^ _
 = R / d2  = S / c4

7
 Process Capability

What it is

• Capability is a process which gives a set of

indexes which tell us how capable we are in
meeting customer specifications, in the
short term and in the long term.

8
 Process Capability

What it is

• A simple view of Process Control and Process Capability

• Process control - refers only to the “voice of the process” –

looking at the process using an agreed performance
measure to see whether the process forms a stable
distribution over time.

9
 Process Capability

Process Control

Process Control
Note - no reference to
specs !

In Control
(Special Causes Eliminated)

Out of Control
(Special Causes Present)

10
 Process Capability

What it is

A simple view of Process Control and Process Capability

Process capability - measures the “goodness of a process” -

comparing the voice of the process with the “voice of the
customer”. The voice of the customer here is the specification
range (tolerance) and/or the nearest customer specification
limit.

11
 Process Capability
Process Capability

Process Capability

In Control and Capable

(Variation from Common
Lower Spec Limit Causes Reduced)
Upper Spec Limit

In Control but not Capable

(Variation from Common Causes
Excessive)

12
 Process Capability

Why Processes Fail?

Process variation (spread) is too large

 Process average is not properly centered
 Process average is not properly centered and
Process Variation is too large

13
 Process Capability

Objective ?

• The objective is to get as close to the theoretical best

that your process can achieve by eliminating special
causes of variation,

• so that only common (natural) causes are acting on the

process, and then to reduce these to a minimum,
whenever possible.

14
 Process Capability

Measurement

– We are often required to compare the output of a

stable process with the process specifications and
make a statement about how well the process meets
specification.

–To do this we compare the natural variability of a

stable process with the process specification limits.

15
 Process Capability

– Assume that a process involves a quality characteristic

that follows a normal distribution with mean , and
standard deviation, . The upper and lower natural
tolerance limits of the process are

– UNTL =  + 3
– LNTL =  - 3

Actual Process Spread = UNTL – LNTL

Allowable Process Spread = USL - LSL

16
 Process Capability
Measurement
– A capable process is one where almost all the
measurements fall inside the specification limits. This can
be represented pictorially by the plot below:

Actual Process Spread < Allowable Process Spread 17

 Process Capability
Lower Upper
Specification Specification Actual Process Spread =
Allowable Process Spread

A. Process variability
matches specifications
Lower Upper
Specification Specification

Actual Process Spread <

Allowable Process Spread
B. Process variability
well within specifications Lower Upper
Specification Specification

Actual Process Spread >

Allowable Process Spread

C. Process variability
exceeds specifications 18
 Process Capability
Process Capability for Bilateral Specification

A process producing a characteristic with a bilateral

specification meets the minimum requirement of
capability when it is stable, and has no more than
0.135 percent of its output for this characteristic
outside either specification limit.

LS L U S L

0 .1 3 5 % 0 .1 3 5 %

0 5 10 15 20

19
 Process Capability
Process Capability for Unilateral Specification

A process producing a characteristic with a unilateral

specification meets the minimum requirement of
capability when it is stable, and has no more than
0.135 percent of its output for this characteristic
outside the single specification limit.

L S L

0 .1 3 5 % 9 9 .8 6 5 %

0 5 10 15 20

20
 Process Capability Indices
1 2
CAPABILITY PERFORMANCE
Short Term Long Term
In Control Not in Control
Pooled std dev Total / overall std dev
compares
How spec range
a much (tolerance) Cp Pp
variability to process
width
how close
process
How
b
centred
centre is to Cpk Ppk
nearest
spec limit

Cp or ENTITLEMENT is the theoretical best

that a process can achieve.
21
 Explaining the Terms
1 2
CAPABILITY PERFORMANCE
Short Term Long Term
In Control Not in Control
Pooled std dev Total / overall std dev
compares
How spec range
a much (tolerance) Cp Pp
variability to process
width long term
how close
process
How
b
centred
centre is to Cpk Ppk
nearest
spec limit

600

550

500

450

400

350

300
Time

short term
22
 Explaining the Terms
1 2
CAPABILITY PERFORMANCE
Short Term Long Term
In Control Not in Control
Pooled std dev Total / overall std dev
compares
How spec range
a much (tolerance) Cp Pp
variability to process
width
how close
process
How
b
centred
centre is to Cpk Ppk
nearest
spec limit

LSL USL

Cp, Pp compares specification range (tolerance) to process

width regardless of where the process is centred.
23
 Process Capability
Cp, Pp compares specification range (tolerance) to process
width regardless of where the process is centred.

It is the ratio of the specification width to

6 x process standard deviation.

Cp (Process Capability Index) = USL-LSL

6sP

Pp (Process Performance Index) = USL-LSL

6sT

sP = pooled standard deviation sT = total standard deviation24

 Explaining the Terms
1 2
CAPABILITY PERFORMANCE
Short Term Long Term
In Control Not in Control
Pooled std dev Total / overall std dev
compares
How
X
spec range
a much (tolerance) Cp Pp
variability to process
width
how close
process
How
b
centred
centre is to Cpk Ppk
nearest
spec limit

LSL USL

Cpk, Ppk measure how close the process centre is to the nearest
customer spec.
It is (difference between the process average and the nearest specification)
(3 x the process standard deviation).
25
 Process Capability

Cpkl = X-LSL Ppkl = X-LSL

3sP 3sT

Cpku = USL-X Ppku = USL-X

3sP 3sT

Cpk = min {Cpkl, Cpku} Ppk = min {Ppkl, Ppku}

26
 Example

Humidity level in the foundry shop is expected to be

between 60% and 80%. Random samples taken yield
the following results:

60, 75, 70, 85, 60, 65,76,80, 70, 75

Average of the above ten values: 71.6,

SD = 8.24

27
 Example
USL  LSL 80  60
Cp    0.40
6 6(8.24) It is required to have at

least 1.0, so this is not a
x  LSL 71.6  60 good process. If possible,
C pkL    0.47
3 3(8.24) reduce the variability
 or/and center the process.
USL  x 80  71.6
C pkU    0.34
3 3(8.24)

C pk  min (C pkU , C pkL )  min (0.34,0.47)  0.34

28
 Process Capability

The Cpk can also be expressed as Cpk = Cp(1-k), where k

is a scaled distance between the midpoint of the
specification range, m, and the process mean,

29
 Example

For a certain process the USL = 18 and the LSL = 6.

The observed process average, = 14, and the standard deviation, s = 2.

USL  LSL 18  6
Cp   1.0
6 6(2)

This means that the process is capable as long as it is located

at the midpoint, m = (USL + LSL)/2 = 12.

30
 Example

But it doesn't, since = 14. The k factor is found by


mx
2
k   0.333
(USL  LSL) / 2 6

CPK  CP (1  k )  0.667

It is required to have Cpk at least 1.0, so this is not a good

process. If possible, reduce the variability or/and center the
process.
31
Process Capability Index

32
Translating capability into "rejects"

USL - LSL 6 8 10  12 
Cp 1.00 1.33 1.66 2.00
Rejects 2700 ppm 64 ppm 0.6 ppm 2 ppb
% of spec used 100 75 60 50

33
 Process Capability

use as a rule of thumb the following chart

Cp Cpk Pp Ppk
Red (Bad) < 1.00 < 1.00 < 1.33 < 1.33
Yellow (OK) 1.00 - 1.33 1.00 - 1.33 1.33 - 1.67 1.33 - 1.67

Green (Good) > 1.33 > 1.33 > 1.67 > 1.67

34
 Process Capability

How it helps

• A Process Capability study is the end result of the

analysis done on process output data.

• A Process Capability study reflects what our customer

sees.

• It indicates what our short term and long term situation

looks like.

35
 Process Capability

How it helps

• Recognise however that our processes need to be stable

(in statistical control) as well as capable.

• Before anything, we must have faith in our

measurements, hence the need for a measurement
system analysis.

36
 Process Capability
Major uses of data from a process capability analysis

1. Predicting how well the process will hold the

tolerances.
2. Assisting product developers/designers in selecting
or modifying a process.
3. Assisting in Establishing an interval between
sampling for process monitoring.

37
 Process Capability
Major uses of data from a process capability analysis

4. Specifying performance requirements for new

equipment.
5. Selecting between competing vendors.
6. Planning the sequence of production processes when
there is an interactive effect of processes on tolerances
7. Reducing the variability in a manufacturing process.

38
 Process Capability

–Setting Tolerances on Assemblies and Components

Assumption
–The process making assemblies and components are barely capable
–i.e. 6σ = USL – LSL
–or Cp = 1

39
Process Capability

Y = X1 + X2 + X3 +X4

Y = X1 – X2

40
 Process Capability
In general (LC)
Y = a1X1+a2X2+…..+akXk
k

= a x
i1
i i

–Xi (component dimension) & Y (assembly dimension) are random variables

–Xi → with mean µi and SD σi
k
If component dimension are independent Vari (y) = y2 =  ai2 var  xi  =
k

k a  i
2
i
2

a 
i1 i 1
y =
i i
i 1

i = mean of Xi (variable)
ai = constant.

41
Process Capability
Y = X1 + X2 + X3 +X4

 y  1  2  3  4
 y   12   2 2   3 2   4 2

NT onY  y  3 y

Y = X1 – X2

 y  1   2
 y   12   2 2

NT onY  y  3 y42
 Process Capability

Components Mean Length Tolerances

A 2 2±0.3
B 5 5±0.2
C 6 6±0.2
D 7 7±0.1

–Design specifications of assembly length are 20±0.3

–Find the NTL for the assembly length.

–Comp. A σ1 = (2.3-1.7)/ 6 = 0.1 µ1 = 2

– B σ2 = (5.2-4.8)/ 6 = 0.067 µ2 = 5
– C σ3 = 0.067 µ3 = 6
– D σ4 = 0.033 µ4 = 7 43
 Process Capability
–Comp. A σ1 = (2.3-1.7)/ 6 = 0.1 µ1 = 2
– B σ2 = (5.2-4.8)/ 6 = 0.067 µ2 = 5
– C σ3 = 0.067 µ3 = 6
– D σ4 = 0.033 µ4 = 7

 y  1   2  3   4
 2  5  6  7  20 NT onY  y  3 y
 y   12   2 2   3 2   4 2 20  3(0.142)  (19.574,20.426)
 y  0.12  0.67 2  0.67 2  0.332
 y  0.142

44
 Process Capability

ZUSL = (20.3 – 20)/ 0.142 = 2.11

USL = .0174 
ZLSL = (19.7 – 20)/ 0.142 = -2.11   Total Area = .0348 = 3.48%
LSL = .0174 
45
 Process Capability
Components Mean Length Tolerance on Individual Components.
A 2
–Design specifications of assembly length are
B 5 20±0.3
C 6 –Find the NTL for all the components
D 7

 y  1   2  3   4 A  2  3 x0.05
20.3  19.7 B  5  3 x0.05
y   0.1
6 C  6  3 x0.05
 y2   1   2   3   4  0.01
2 2 2 2
D  7  3 x0.05
assuming equal distributi on of
variance on all components A = (1.85, 2.15)
 y 2  4 2 B = (4.85, 5.15)
C = (5.85, 6.15)
 2
.01
2  y   0.0025 D = (6.85, 7.15) 46
4 4