Chapter Two

Methods of Statistical Process Control (SPC)

and Capability Analysis

2.1 Methods and Philosophy of Statistical Process Control

2.2 Statistical Basis of the Control Chart
2.3 Control Chart for quantitative variable
2.4 Control Charts for Attributes
2.5 Process and Measurement System Capability Analysis

1
 2.1. Methods and Philosophy of SPC
 If a product is to meet or exceed customer
expectations, generally it should be produced by
a process that is stable or repeatable.
 More precisely, the process must be capable of
operating with little variability around the target
dimensions of the product's quality
characteristics.
2
 Statistical process control (SPC): is a
powerful collection of problem-solving tools
useful in achieving process stability and
improving capability through the reduction of
variability
 Statistical process control (SPC): is an
effective method of monitoring a process
through the use of control charts. 3
 Control charts enable to distinguishing variation
based on statistical techniques.
 Much of its power lies in the ability to monitor both
process center and its variation about that center.
 It has an emphasis on early detection and prevention
of problems(variations).
 In addition to reducing waste, SPC can lead to a
reduction in time required to produce the product.

4
 SPC indicates when an action should be taken in a
process, but it also indicates when no action should be
taken.
 The main objective of SPC in any production process
is to control and maintain a satisfactory quality level
for its product. It is achieved through the technique of
control charts invented by W.A.Shewhart in 1924.
 Product control limits aims at guarantying a certain
quality level is being maintained by the producer.
5
2.2 Statistical Basis of the Control Chart
 A major objective of SPC is quickly detect the
occurrence of assignable causes of process shifts
so that investigation of the process and corrective
action may be undertaken before many
nonconforming units are manufactured.
 One of the most important tools of production
management and control of quality in the
manufactured product is the ‘control charts’ 6
 The ‘control chart’ is a graphical display of
quality characteristics that has been measured or
computed from a sample various sample number
of time.
Some of the control charts to be discussed:
 Quantitative (X-bar, range, and standard
deviation charts),
 Attributes (p, np, and c charts) and

 Cumulative Sum (CUSUM) chart. 7

 The control chart contains a center line that
represents the mean value (target value) of the
quality characteristics corresponding to the in-
control state or process.
 Two other horizontal lines, called the upper
control limit (UCL) and the lower control limit
(LCL) are also shown on the chart.

8
 Control Limits are used to determine if the process is
in a state of statistical control (i.e., is producing
consistent output).
 As long as the point plots are within the control
limits, the process is assumed to be in control, and
no action is necessary;

 otherwise the process is out-of-control, and

correction actions are required to find and eliminate
the assignable causes. 9
 Some analysts are using two tests of limits on
control charts.
 These limits are found between the control
limits that is used to decide the process is in
control with warning and called warning limits.

10
11
2.3. Control Charts for quantitative variables
“control charts” is the technique of process
control.
Variable control charts are used when quality is
measured as variables (length, weight, tensile
strength, etc.).
The main purpose of the variable control charts
is to monitor the process mean and the standard
deviation. 12
a) X-bar control chart
 A typical quality control plan requires sampling one
or more items from a production process
periodically.
 CL, UCL, LCL, UWL and LWL
How do we choice these lines?

 We want to choose the control limits so that

P( is outside the control limits given that the process

13
 Similarly, if we want to have warning limits we
want to choose the warning limits so that

P(X-bar is outside the warning limits given that

the process is in control) to be small.

 We might use the 95% and 99% rule, i.e., solve

• P(LCL < X-bar < UCL/ process is in control) = 0.99,

• P(LWL < X-bar < UWL/ process is in control) =0.95

14
 If the distribution of X is normal with mean μ
and variance σ2, then the distribution of X-bar
is normal with mean μ and variance σ 2/n.
• UCL = μ+2.58*,
• UWL = μ+1.96*,
• CL = μ,
• LWL = μ-1.96* and
• LCL = μ - 2.58*

15
 Another way of choosing the limits is to use 2-sigma

and 3-sigma methods. Therefore, if we use the 2 and

3-sigma methods, the above formula becomes:
• UCL = μ+3*,
• UWL = μ+2*,
• CL = μ,
• LWL = μ-2* and
• LCL = μ - 3*.
16
 During the 1920's, Dr. Walter A. Shewhart
proposed a general model for control charts.
 Let w be quality characteristic of interest (e.g;
thickness), and the mean of w is μw, with a

standard deviation of σw. Then the center line,

the UCL and the LCL are:

17
 UCL = μw + k σ w

 Center Line = μw

• LCL = μw - k σ w , Where k is the distance of the

control limits from the center line, k is set to 3
(accepted standard ), we speak of 3-sigma control
chart.

• In general, σ is unknown. Here we replace σw with a

given standard value, or we estimate it by a function
of the average standard deviation. 18
19
 3-sigma X-bar control limits:
• UCL = μ+3*
• CL = μ
• LCL = μ - 3*
 If m samples each of size n are taken, the
grand mean is:

20
 Here we may estimate from either the ranges
or standard deviations of the m samples.

 The average range is:

 An unbiased estimator of  based on range is:

where is tabulated for various sample sizes.

21
 Hence, the central line and 3-sigma control
limits for X-bar control chart are:
• UCL =
• CL=
• LCL =

or
UCL = and LCL=
and tabulated for various sample sizes 22
 Suppose we have m preliminary samples at our
disposition, each of size n, and let si be the
standard deviation of the i-th sample. Then the
average of the m standard deviations is:

 An unbiased estimator of  is:

Where c4 is tabulated for various sample sizes

23
 Hence, the central line and 3-sigma control
limits for X-bar control chart are:

24
Possible Decision Rules

The most common decision rules are:

 The process is “out-of-control” if and only if X-bar is

outside the control limits; it is “in control” if and only if it is

within the warning lines; otherwise it is an issue of warning.

 If we do not use warning limits we can use the rule: the

process is “out-of-control” if and only if it is outside the

control limits; and it is “in control” if and only if it is within

the control limits. 25

26
How do we Measure the Performance of a Control
chart?

i) The Power Function

• The power function is the probability of the

process is out-of-control at sample i (regarded as
a function of parameters mean and variance at
sample i).

27
 If the control chart is good the power function will be
small (close to 0) when the process is in control and
large (close to 1) when it is out-of-control.

ii) Average Run Length

• The time until a process is found to be out-of-

control is called a run.
• This time is a random variable. Its expectation is
called the Average Run Length (ARL).
28
 Ideally we want the ARL to be large when the
process is in control and small when it is out-of-
control.
Exercise: i) Suppose that when the process is in
control the measurements made for a variable Xi
are identically and independently normally
distributed with mean 0 and variance = 1 If a
sample of size 4 is taken at each time point and we
say “the process is out of control” 29
if the ith observation is outside the control limits,
calculate
a) the power function (Ans: 0.0026)
b) average run length (Ans: 285)

ii) Process is out of control with mean =0.2 and

variance =1.
c) the power function (Ans: 0.0047)
d) average run length (Ans: 213)
30
b) Range (R) chart:
 This chart controls the process variability.
 An unbiased estimator of  based on range is

31
32
33
34
Where C4

35
36
37
 If we assign A3 = , then
UCL = + A3 , CL = and LCL = - A3.

Factors for Calculating Limits for and R Charts

X

38
Example:

39
40
41
42
Example:2

43
44
45
Range chart

46
S chart

47
48
Changing Sample Size on the X-bar and R-charts
 and R charts have been presented by assuming that the
sample size n is constant from sample to sample.
 There are situations in which the sample size n is not
constant.
 Another situation is that of making a permanent (or
semi-permanent) change in the sample size

49
Because of cost or because the process has exhibited good
stability and fewer resources are being allocated for
process monitoring.
 In this case it is easy to re-compute the new control
limits directly from the old ones without collecting
additional samples based on the new sample size.
 Let
• = average range for the old sample size
• = average range for the new sample size
50
• = old sample size
• = new sample size
• d2(old) = factor d2 for the old sample size

• d2(new) = factor d2 for the new sample size

 For the chart the new control limits are
•UCL = +
•LCL = - where the center line is unchanged and the factor
A2 is selected for the new sample size.

51
For the R chart, the new parameters are:
•UCL =
•CL = =
•LCL = max{0, } ,

where D3 and D4 are selected for the new sample

size.

52
The X-bar and S-charts with Variable Sample Size

 If ni is the number of observations in the i th sample,

then use
• = and =

For x-bar chart

• UCL = + A3 ,

• CL = and

• LCL = - A3A3, is based on an average sample size of

ni’s. 53
For s chart

s
LCL  s  3 1  c 24
c4
or
UCL = B4 , CL = and LCL = B3

Where B3 and B4 are based on an average sample size of ni’s.

54
 Control Chart for S2

• UCL =
• CL =
• LCL =
Note: .

55
Individual measurement chart (X chart)
 This occurs when we take one item at each time point (n=1)
•Moving range (MR): MRi = |xi - xi-1|, then the UCL, CL and
LCL for a control chart for individuals are

LCL= -3
56
• And for a control MR chart for moving ranges
CL  mr
LCL  D3 mr  0
 X Charts can be interpreted similar to X-bar charts.

 MR charts cannot be interpreted in the same way as X-bar or R

charts.

 MR chart cannot really supply useful information about process

variability.

 More emphasis should be placed on interpretation of the X

chart. 57
58
59