Introduction

1) Statistical quality control is the collection analysis and interpretation of inspection

data to solve the problems related to quality control.
2) Statistical quality control involves analysis of the characteristics of a production
output by sampling output.
3) In SQC statistical techniques are employed to control the quality or to solve quality
problems.
4) It is always desired to use SQC methods in mass production because it will try to
eliminate variations of process, but it is impossible to achieve exact repeatability
within a process due to random or chance causes.
5) In SQC techniques, the actual physical measurement are first done by means of
inspection. The degree of variation among the measurement is analysed by
statistical methods to take action as quality control.
Objectives
1) Improved producer-customer relations.
2) Better specification achievement.
3) Better inspection methods.
4) Most efficient utilization of equipments.
5) Better utilization of raw material.
6) Less scrap and rework.
7) Better quality level.
8) Better uniformity of quality.
Advantages
1) Reduction in scrap and rework.
2) Proper use of equipments.
3) Uniformity of quality.
4) Better quality level.
5) Improved productivity.
6) Improvement in inspection standard.
7) More effective pressure on quality control.
8) Improvement in customer satisfaction.
ushna
9) Quality awareness creation in employees and improvement in employees Kr moral.
Ac
ademy

Variation Or Variability
1) Concept of variation states that no two items will be perfectly identical even if
extreme care is taken to make them identical.
2) Variation is a fact of nature variation is bound to occur at the time of manufacturing.
Ex.: 100 jobs (parts) of 20 mm diameter are to be manufactured then variations
among the parts produced must be present.

34
3) The two things are said to be alike when difference among them are so small that
we don’t care about them, but generally variations whether small or large exist in
the parts manufactured in all production processes.
Types Of Variations
1) Variations within the part itself.
2) Variations among the parts produced during the same period of time.
3) Variations among parts produced at different period of time.
There are two causes of variations:
1) Inherent or random or chance causes of variation.
2) Assignable causes of variation.
[Q: Explain inherent and assignable source of variation.]
A] Inherent Sources Of Variations
This is caused by a large number of chance factors and depends upon the rigidity
of machine tool and its process capability.
There are inherent in process or path.
They are difficult to trace and control even under best conditions of production.
B] Assignable Causes
Variations of these types are caused by:
1) Human factors 2) Tool wear
3) Tool getting loose 4) Machine wear
5) Change in initial setting 6) Difference among workers
7) Difference in environmental conditions 8) Difference in raw materials
Assignable causes if present, easily detected and corrected before defective pieces
are produced.
They possess greater magnitude as compared to chance / inherent causes.

ushna
Kr
Ac
ademy

35
 Types Of Data
(A) Attributed Data, (B) Variable Data
A] Attributed Data
1) When a record shows only the number of workpieces (items) conforming and
non-conforming to any specified requirement, it is said to be attributed.
2) Measurement methods in which item examined considered satisfactory or
unsatisfactory.
3) Go and No Go gauges provides measurement by attributes.
4) This method simply notes presence or absence of a quality charactgeristics in the
parts examined.
5) Attribute data is also known as ‘Discrete Data’.
Ex.:
(a) Number of defectives found in a sample.
(b) A sample of 25 castings taken from a lot contain 0, 1, 2 or 4 defectives.
B] Variable Data
1) When a record is made of an actual measure quality characteristics such as
dimensions in mm, the quality is said to be in variables.
2) It is also known as ‘Continuous Data’.
3) It is done with the help of precision measuring devices.
Ex.:
(a) A sample of 20 spindles taken from a lot may have diameters which vary
from 25 to 25.99 mm.
(b) Measurement of temperature in 0K.
(c) Measurement of mass 100 kg.
SQC Tools
Different SQC tools used are:
(A) Frequency Histogram, (B) Bar Chart, (c) Frequency Polygon, (D) Ogive Curve
A] Frequency Histogram

24

20
16
ushna
12 Kr
Ac
ademy

0 10 20 30 40 50 60 70 80

36
1) In this graph, the sides of the column represent the upper and lower cell boundaries
and their heights are proportional to frequency of occurances within the cells.
2) The simplicity of construction and interpretation of the histogram makes it an
effective tool in the elementary analysis of the data.
3) It is used as a tool for incoming inspection.
B] Bar Chart
1) A frequency bar chart is a graphical representation of the frequency distribution
in which the bars are centered at the midpoints of the cells.
2) The heights of the bars are proportional to the frequencies in the respective cells.
Y

24
20
Frequency

16
12
8
4

0 10 20 30 40 50 60 70 80 X
Class Intervals
C] Frequency Polygon
1) In the frequency polygon, a series of straight lines joins the small circles which
are plotted at cell midpoints with a height proportional to cell frequency.

ushna
Kr
Ac
ademy

N.D. Curve (Normal Curve)

1) In the frequency distribution, if the number of observations are increased
considerly then the number of cells will increase and the width of the cell will
becomes smaller and smaller.
2) The series of steps that constitutes the top line of the histograph will then approach
a smooth curve.
37
 Y

x X
–3 –2 –1 +1 +2 +3
68.26%
95.46%
99.73%

3) The height of the curve at any point is proportional to the frequency at that point
and the area under it between any two limits is proportional to the frequency of
occurances.
4) The frequency curves may be of different shapes. The most important of these
curves is the normal curve. It is a symmetrical bell shaped curve.

1) The normal distribution curve is symmetrical about its mean value and has bell
shaped. The curve is fully defined by ‘ x ’ and ‘’.
2) Theoretically the ND curve extends from   to   however for all practical
purposes we can consider and 3 values to the right of the mean (i.e. x  3 ).
3) The most commonly quoted limits in connection with the curve are a follows:

x  0.6745 50
x  1 68.26
x  2 95.46
x  3 99.73
4) ND curve can estimate the total % of production that will fall within engineering
specification limit.
Concept Of Central Tendency ushna
Kr
Ac

One of the salient characteristic of the distribution of the sample data is that most
ademy

of the observations tends to concentrate in the centre of the distribution. This

characteristic of the distribution is known as ‘Central Tendency’.
Central tendency is usually expressed in three ways:
1) Arithmetic mean (The average value)
2) Median (The middle value)
3) Mode (The most frequently occuring value)
38
1) Arithmetic Mean
The arithmetic mean is the average of all the values of the variables in the sample.
Let, x1, x2, x3, ..........x n are the ‘n’ values of the variable observed in the sample,
then their arithmetic mean is given by,
x1  x 2  x 3  ......  x n
x
n

x
x
n
2) Median
When all the observations are arranged in ascending or descending order, then
the median is the magnitude of middle case.
It is half the observations above it and half below it.
Case I:
n 1
If ‘n’ is odd, the median is given by .
2
Case II:
th th
n n 
If ‘n’ is even, the median is taken as the average of   and   1 of value.
2 2 
3) Mode
Mode is the value that occurs most frequently in a frequency histogram or
frequency polygon.
Ex.: The recorded observations are,
2, 3, 2, 4, 5, 2, 2, 6, 2, 4, 2, 4
Here, 2 occurs most frequently, hence, mode = 2.

Control Chart
1) A control chart is a graphical representation of the collected information.
2) In other words, control chart is a device which specifies the state of statistical
control, a device attaining statistical control and a device to judge whether
statistical control has been attained.
3) With the help of control chart, it is possible to find out the natural capability of
production process.
4) rushna the
It detects variation in processing and warns if there is any departure Kfrom
Ac

specified tolerance limit.

ademy

5) There are many types of control charts designed for different control situations,
each with its own advantages and disadvantages and with its own field of
application.

39
 Control Chart

Control chart Control chart

for variables for attributes

x-R Chart x- Chart P-Chart C-Chart

(Defective) (Defects)
Control Chart For Variables
1) Control charts based upon measurements of quality characteristics are called as
control charts for variables.
2) The variable control charts that are most commonly used are x  R chart and
x   chart.
3) Control charts for variables are often found to be more economical means for
controlling quality than control charts based on attributes.
x  R Chart
The various steps in constructing x and R chart are as follows:
1) Calculate the average x and range ‘R’ for each sub-group.
For each sample mean value and the range is calculated.
Ex.: A sample contains 5 items whose dimensions are,
x 1, x 2, x3, x4 and x 5
The sample average,
x1  x 2  x 3  x 4  x 5
x
5
Range = R = Highest value – Smallest value
2) Calculate the grand average x and average range R .
x is the average of the x values for each sub-group.

x
x
N

R
Average range, R
N
where, N = Number of sub-groups
ushna
3) Calculation of control limits for x chart: Kr
Ac
ademy

Upper control limit (UCL) = x  A 2 R

Lower control limit (LCL) = x  A 2 R
4) Calculation of control limits for R chart:
( UCL) R  D4 R
(LCL) R  D3R
40
 Process Capability
1) Process capability may be defined as the, ‘minimum spread of a specific
measurement variation which will include 99.7% of the measurements from the
given process’.
2) In other words, the process capability = 6' , since 6' is taken as a measure of
the spread of the process, which is also called natural tolerance.
3) Process capability study is carried out to measure the ability of the process to
meet specified tolerances.
4) By this study, it becomes possible to know the % of the products which will be
produced within  36 limits on either side of mean x .
5) Process capability analysis consists of:
a) Measuring the process capability to find out whether the process is inherently
capable of meeting the specified tolerance limit.
b) Discovering why a process capable is falling to meet specifications.
Procedure To Calculate Process Capability
1) Calculate average x and range ‘R’ for each sample.
2) Calculate the grand average ( x ) . This measures centering of the process.
3) Now calculate control limits and plot x  R chart. This measures capability of
the process.
4) Calculate the process capability 6' ,
R
where, '
d2
This measures piece to piece variability of the process.

1) (x – x ) > 6'
max min

In this case, the spread of the process ( 6' ) is considerably less than the difference
between the upper and lower specification limit.
******FIG
The frequency curves shows various positions in which the process might be
centered.
Conclusion: With any position ABCDE practically all the products manufactured
will meet specifications.
2) (x – x ) < 6' ushna
Kr
max min
Ac

******FIG
ademy

In this type of situation, defective parts will always be there.

4) (x – x ) = 6'
max min

******FIG
In this situation, the spread of the process is approximately equal to the difference
between upper and lower specifications limit as shown in figure.

41
 Control Chart For Attributes
1) P – Chart : ---------- Defectives
2) C – Chart :---------- Defects

An item is said to be defective if it fails to conform the specifications in any of

the characteristics.
Each characteristics that does not meet the specifications is a defect.
P – Chart:
P – chart is an attribute control chart.
Procedure For Fraction Defective Chart:
1) Record the observation table.
*****TABLE
2) Compute fraction defective (P) for each sub-group.
Number of defectives in subgroup
P
Number of jobs inspected in subgroup
3) Compute P , the average fraction defective.
Total number of defectives
P
Total number of jobs inspected
4) Find control limits.
P (1  P )
UCL  P  3 
n
P (1  P )
LCL  P  3 
n
5) Now, polt the graph.

UCL
–
P
LCL

ushna
Procedure For Percentage Defective Chart: Kr
Ac
ademy

1) Record the observation table.

2) Find the percentage defective (P) for each subgroup.
Number of defectives in subgroup
P  100
Number of jobs inspected in subgroup
3) Compute the average fraction defective.

42