Quality Control and Process

Improvement
Unit 4 & 5
Statistical Process Control
 Statistical Process Control
Statistical process control is a method of quality control which employs statistical
methods to monitor and control a process. This helps to ensure that the process
operates efficiently, producing more specification-conforming products with less waste.
 Statistical Process Control
Principles underlying Process Control

1) Variability is inherent in any process.

2) Total variability is composed of two parts internal to the process and external to the process.

✓ Variations internal to the process (random variations) - They are inevitable, difficult to identify
and we can do nothing about them unless the whole process itself is changed. These variations are
due to common causes.

✓ Variations external to the process (variations caused due to assignable causes) - They can always
be traced to the operating conditions of the process and corrected
Statistical Process Control - Variations
 Control Charts
The most common device used for this purpose is Shewhart Control Chart introduced
in 1931. A Control Chart is a visual display of statistically significant changes that may
occur in a process

1. The central line denotes the mean value of the quality characteristic
2. The upper control limit is located at 3 standard deviation (3σ) above the central line
3. The lower control limit is located at 3 standard deviation (3σ) below the central line
 Types of Control Charts
Variable – Properties or quality characteristic capable of being measured are called
variables. e.g. - diameter, length, weight, temperature, voltage, hardness etc.

Attribute – Characteristic difficult to measure, and are categorised as good or bad,

defective or non-defective, acceptable or not-acceptable etc. e.g. – colour. gloss,
texture, cracks, imperfections etc.

Control Charts for Variables

ത R chart)
1. The average and Range chart ( 𝑋-
ത S chart)
2. The Average and Standard Deviation Chart (𝑋-

3. Individual Moving Range Chart (ImR / XmR Chart)

 Types of Control Charts
Control Charts for Attributes

1. Control Chart for fraction defectives (p-chart)

2. Control Chart for number of defectives (np-chart)

3. Control Chart for number of defects (c-chart)

4. Control Chart for number of defects (u–chart)

Control Chart for Variables
 Control Limits
Control limits are calculated from the process data and therefore, represent the voice
of the process (VOP).

The control Limits (UCL. LCL ) are set at ±3σ.

UCL - Upper control limit, LCL - Lower control limit

When calculating control limits, it is prudent to collect as much data as practical.

Many authors suggest at least 25 subgroups or samples.

Subgroup or sample sizes are generally held constant with variable control charts.

However, subgroup sizes may vary with specific attribute control charts
ത Chart
𝑋-R
 ത Chart
𝑋-R

A2 = a constant (Refer table to get value of A2 for different sample size)

D3 & D4 = two different constants to calculate control limits for Range chart
Steps to Prepare -R Charts
Step 1 : Collect Data

Step 2 : Calculate the mean (X) and range (R) of each sample

Step 3 : Calculate grand average ( ) and mean range ( )

Step 4 : Calculate control limits and set up control charts for average
1. If all points fall within the control limits no modification is required
2. Else eliminate points outside control limits and recalculate the new limits.
3. Repeat the process until all points fall within the control limits

Step 5 : Calculate control limits and set up control charts for range
1. If all points fall within the control limits no modification is required
2. Else eliminate points outside control limits and recalculate the new limits.
3. Repeat the process until all points fall within the control limits
 Example
Data is collected in a face and plunge operation done on a lathe. The dimensions
being measured in the groove inside diameter (ID) which has a tolerance of 7.125 +
0.010. Four parts are measured every hour. The values are given in the table below.
Example
Control Chart
ത Chart
𝑋-S

si = Standard deviation of the ith subgroup

ത Chart
𝑋-S
Example
Example
 ത
𝑋-mR Chart
In some situations, a sample size of one is used like very slow processes or processes in
which the measurement is very expensive to obtain such as with destructive tests.

If the sample size is one, an individual and moving range (also known as ImR or XmR)
chart is appropriate.

The moving range is calculated by taking the absolute value of the difference between
each measurement and the previous one.

For this reason, the moving range chart has one less point plotted than the
corresponding individual chart.
-mR Chart
Control limits for the individual and moving range chart are given
by the following formulas.
-mR Chart
Example
-mR Chart
Example
Reading Individual Data Element mR
1 290
2 288 2
3 285 3
4 290 5
5 291 1
6 287 4
7 284 3
8 290 6
9 290 0
10 288 2
288.3 2.89

E2 = 2.660, D3 = 0, D4 = 3.267
-mR Chart
Example
E2 = 2.660, D3 = 0, D4 = 3.267
-mR Chart
Attribute Charts
 Attribute Charts
Attribute charts are used for count data where each data element is classified in one of
two categories, such as good or bad.
p charts and np charts are used to plot percentage or proportion defective.
c charts and u charts are used to plot counts of defects.
The np chart and the c chart require constant sample sizes.
However, the p chart and the u chart permit variable sample sizes.
When a lower control limit computes to a negative value, it is artificially set to zero
since it is not possible to have a negative percentage defective or negative defect
counts.
 Control Chart for fraction defectives (p-Chart)
The control limits on the chart appear ragged because they reflect each subgroup’s
individual sample size.
The formulas for the center line and the upper and lower control limits for the p
chart are
 Control Chart for fraction defectives (p-Chart)
𝑝ҧ = average fraction defectives
Control Limits =
n = sample size

Since the value of p can never be negative, hence if the LCL works out to be
negative it is taken as zero
 Control Chart for number defectives (np-Chart)
np-chart unlike p-chart records number of defectives instead of fraction or
proportion of defectives.

Control Limits =

𝑝ഥ = Total number of defectives /Total number of units inspected

n = sample size

Since the value of p can never be negative, hence if the LCL works out to be
negative it is taken as zero
 Example
10 samples, each of size 50, of a pipe were inspected in pressure testing. The results
of the inspection are given below

Sample No. : 1 2 3 4 5 6 7 8 9 10
No. Of defectives : 2 3 2 0 2 3 2 1 2 3

Draw a suitable control-chart and state your conclusion

 Example
A test was conducted to determine the presence of the Rh factor in 13
samples of donated blood. The results of the test are given in the table below
 Example
Number of Sample
We need to calculate Sl #
Defectives Size
p UCL LCL

the control limits for 1 12 115 0.104 0.207 0.027

each value of ni using 2 14 125 0.112 0.203 0.031
3 18 111 0.162 0.209 0.026
the formula 4 13 133 0.098 0.201 0.033
5 17 120 0.142 0.205 0.029
6 15 118 0.127 0.206 0.028
7 15 137 0.109 0.200 0.035
8 16 108 0.148 0.210 0.024
9 11 110 0.100 0.209 0.025
10 14 124 0.113 0.204 0.030
11 13 128 0.102 0.202 0.032
12 14 144 0.097 0.197 0.037
13 17 141 0.121 0.198 0.036

SUM 189 1614

P-bar 0.1171
Example
np Chart Example
np Chart Example
Control Chart for number of defects (c-Chart)
c chart is used when the sample size is constant. We use this chart
to plot the number of defects. Example :
Quality of glassware is determined by the number of bubbles in it.
Quality of fibre is judged based on number of weak spots in a given length.
The formulas for the center line and the upper and lower control limits for the
c chart are
 Example
Ten woolen carpets were studied critically for total number of defects in their
texture. The details of number of defects in each carpet are given below

Carpet No. : 1 2 3 4 5 6 7 8 9 10
No of defects : 2 4 3 5 1 3 2 3 4 3

Construct a control chart and comment on the result

 Example
Panes of glass are inspected for defects such as bubbles, scratches,
chips, inclusions, waves, and dips. The results of the inspection are
given in the table below
Example
 U-Chart
u chart is used when the sample size varies.
We use this chart to plot the number of defects per unit.

The control limits on the chart appear ragged because they reflect each
subgroup’s individual sample size.

The formulas for the center line and the upper and lower control limits for the u
chart are
U-Chart
u-chart - Example
Panes of glass are inspected for defects such as bubbles, scratches,
chips, inclusions, waves, and dips. The results of the inspection are
given in the table below
Sl # No. of Defects Sample Size
1 4 125
2 8 111
3 3 133
4 7 120
5 5 118
6 5 137
7 6 108
8 10 110
9 4 124
10 3 128
11 4 144
12 7 138
13 11 150
u-chart - Example
Sl # No. of Defects Sample Size UCL LCL LCL
1 4 125 0.1048 -0.0113 0
2 8 111 0.1084 -0.0148 0
3 3 133 0.1030 -0.0095 0
4 7 120 0.1060 -0.0125 0
5 5 118 0.1065 -0.0130 0
6 5 137 0.1022 -0.0087 0
7 6 108 0.1092 -0.0157 0
8 10 110 0.1086 -0.0151 0
9 4 124 0.1050 -0.0115 0
10 3 128 0.1041 -0.0106 0
11 4 144 0.1009 -0.0073 0
12 7 138 0.1020 -0.0085 0
13 11 150 0.0998 -0.0062 0
Sum 77 1646
u-bar 0.0468
u-chart - Example
Constants for Control Chart
n A2 A3 D3 D4 d2 B3 B4 E2
2 1.880 2.659 0.000 3.267 1.128 0.000 3.267 2.660
3 1.023 1.954 0.000 2.574 1.693 0.000 2.568 1.772
4 0.729 1.628 0.000 2.282 2.059 0.000 2.266 1.457
5 0.577 1.427 0.000 2.115 2.326 0.000 2.089 1.290
6 0.483 1.287 0.000 2.004 2.534 0.030 1.970 1.184
7 0.419 1.182 0.076 1.924 2.704 0.118 1.882 1.109
8 0.373 1.099 0.136 1.864 2.847 0.185 1.815 1.054
9 0.337 1.032 0.184 1.816 2.970 0.239 1.761 1.010
10 0.308 0.975 0.223 1.777 3.078 0.284 1.716 0.975
Monitoring – Statistical Process Control