Statistical Process Control

Statistical Process
Control

1
 Statistical Process Control

Introduction
• What is Quality ?
Totality of characteristics of an entity
that bear on its ability to satisfy stated and
implied needs of customers.

• What is Quality Control ?

Quality control is “ the process
through which we measure actual quality
performance, compare it with a standard,
and act on the difference”
2
 Statistical Process Control

What is SPC
Controlling the process through the use of statistics

Statistical - Taking of measurements and the arrangement

of those measurements in clear pattern to allow
predictions to be made on performance

Process - Any activity involving a combination of Man,

Machine, Material, Method, Tools working together
to produce a final product

Control - Comparing actual performance with target

and what corrective action is necessary to achieve
the target
3
 Statistical Process Control

What is SPC – contd.

➢ Statistical Process Control is a standardized
method of identifying variation in a process
using numerical performance information.

➢ SPC is a decision making tool telling when

to adjust the process and when to leave it alone.

➢ SPC is a time series based methodology,

giving information as to when changes in the
process occur.

4
 Statistical Process Control

What is Time Series Management

Normal variation in the Process

Change occurred in the process. What happened !

5
 Statistical Process Control

Why should we use SPC ?

No Industrial process or machine is able to

produce consecutive Items which are identical in
Appearance, Length, Thickness, Diameter etc.

Differences will occur due to sources of

variation which are inherent in the machine such
as Vibration, Speed, Power etc.

The extent of the variation can be measured

and is called Machine Capability.

6
 Statistical Process Control

Results of SPC
➢ Improvement in Quality & Reliability

➢ Reduction in Scrap and Rework leading to

economy in raw material used

➢ Decreased Inspection cost

➢ Efficient use of men and machines leading

to reduction in cost per unit of the product

7
 Statistical Process Control

Results of SPC

➢ Quality Consciousness at all levels

➢ Scientific evaluation of standards for quality

and production

➢ Improved level of Customer Satisfaction

8
 Statistical Process Control

SPC Benefits
Why do we want to do SPC?

➢ Determine if the process is capable or not

➢ Minimize Variation
➢ Determine if process is suitable for its application
➢ Continuous Improvement initiatives
➢ Proactively address operation performance
➢ Standardize process adjustments

9
 Statistical Process Control

Detection Systems /
Traditional Quality systems
Adjust Process 5

1 The Process

Tools &
Scrap
People Equipment Materials
or
3 Rework
Mass 4
Output Inspection
Product OK
2
Methods Environment

Maintenance Schedule

10
 Statistical Process Control

Process Control Systems

SPC CHARTS
3 Action 2 Information 4 Action
on the about on the
Process Performance Output

1 The Process
Tools &
People Equipment Materials

OK
Output Product
Methods Environment

Maintenance Schedule

11
 Statistical Process Control

7 QC Tools

➢ Check sheet ➢ Scatter Diagram

➢ Histogram ➢ Control Charts

➢ Pareto Diagram ➢ Stratification

➢ Cause and Effect Diagram

12
 Statistical Process Control

Data Collection
Data is a numerical expression of
an activity or process
Types
➢ Variables / Continuous Data
– Length, Thickness, Diameter,
Bore finish, etc

➢ Attributes / Discrete Data

– Good, Bad & O.K / Not O.K
Defective / Non Defective
13
 Statistical Process Control

Data Collection
Reasons / Purposes
➢ To understand an actual situation
➢ Data for process control and acceptance or rejection
➢ For analysis
➢ For Acceptance / Rejection

Sources
➢ Past Records or already Available Data
➢ Live – Data obtained from a freshly identified problems
14
 Statistical Process Control

Check Sheet
Purpose

➢ To record factual data over a period of time

➢ To ensure all the required data are collected
➢ To make data-gathering easy and less time
consuming
➢ To facilitate quick analysis
➢ To arrange data automatically so that it can be
easily used
15
 Statistical Process Control

Check Sheet – Example 1

Week Week Week Week Cumulative

Cause 1 2 3 4 Total Total

Weights too Heavy 53 43 42 61 199 199
Reach too far 24 29 27 27 107 306
Incorrect Posture 5 28 13 30 76 382
Previous Injury 24 20 2 29 75 457
Frequency of Lifts 8 31 15 11 65 522
Cold Workplace 21 9 7 16 53 575
Wrong Footwear 22 4 10 6 42 617
Accidental Twists 1 12 26 3 42 659
Total 158 176 142 183 659

16
 Statistical Process Control

Check Sheet – Example 2

Checks
Deviation Frequency
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

-8

-7

-6

Specification -5 1

-4 2

-3 4

-2 6

-1 9

0 11

1 8

2 7

3 3

4 2

Specification 5 1

6 1

Total 55

17
 Statistical Process Control

Pareto Diagram
➢ A method of showing a table of data in graphical
format to aid understanding

➢ This gives a visual impact

➢ The technique of arranging data according to its

importance and then typing to a problem solving
frame work

➢ Focus on the issue that is to be attended first by

all concerned
18
 Statistical Process Control

Pareto Diagram
➢ Pareto Diagram helps in identifying and isolating
“Vital few” factors from “Trivial many”

➢ 70% to 80% of the defects will arise from

10% - 20% of the causes. These are called “Vital few”
which are highlighted for special attention. Elimination
of these “Vital few” cause will considerably reduce
the total no. of defects.

➢ The remaining 20% - 30% of the defects are from

various causes which are called “Trivial many”

19
 Statistical Process Control

Pareto Diagram - Example

Week Week Week Week Cumulative Cumulative

Reject 1 2 3 4 Total Total % %

Weights too Heavy 53 43 42 61 199 199 30.19727 30.1972686

Reach Too far 24 29 27 27 107 306 16.23672 46.4339909

Incorrect Posture 5 28 13 30 76 382 11.53263 57.9666161

Previous Injuries 24 20 2 29 75 457 11.38088 69.3474962

Frequency of Lifts 8 31 15 11 65 522 9.863429 79.2109256

Cold Workplace 21 9 7 16 53 575 8.042489 87.2534143

Wrong Footwear 22 4 10 6 42 617 6.373293 93.6267071

Accidental Twists 1 12 26 3 42 659 6.373293 100

Total 158 176 142 183 659

100
90
80
70
60
50
40
30
20
10
0
ce
r

ar
s
es
vy

s
re
fa

ft

st
la

we
ri
ea

tu

Li

wi
kp
o

ju

ot
os
To
H

lT
of
In

or

Fo
tP
o

h
to

ta
W
y
us
ac

ec

ng
nc

en
ts

io
Re

ld
rr

ro
ue

id
gh

ev

Co
co

eq

cc
W
ei

Pr
In

Fr

A
W

20
 Statistical Process Control

Cause and Effect Diagram

➢ Cause and Effect Diagram is a method of Brainstorming Causes
of a problem or situation

➢ Cause and Effect Diagram is a systematic way of listing down

all the possible contributing factors ( Causes ) of a problem
( the Effect )

➢ Cause and Effect Diagram helps the Team can focus specific causes

➢ This identifies problem areas where data can be collected and

analysis

➢ Cause and Effect diagram is also known as FISH-BONE

Diagram or ISHIKAWA Diagram
21
 Statistical Process Control

Cause and Effect Diagram

Man Machine Measurement

Effect

Material Methods Other factor

Step 1 - Write the Effect

Step 2 - Write all the possible causes from the table in turn
until dried up

Step 3 - Priorities and select the appropriate causes

22
 Statistical Process Control

Scatter Diagram
➢ Scatter Diagram are used to examine the relations between two
variables to find whether they are related, by controlling the
independent variable, the dependent variable is also controlled.
➢ Scatter Diagram is a Graphical plot of characteristic on Y- axis
and the influencing variable on X- axis
➢ Scatter Diagram is used in explaining the behavior of a process
and the means of controlling it.

Thickness Example
Of plating
Thickness of plating is related
Time of plating

Time of plating
23
 Statistical Process Control

Stratification
➢ Stratification is the process of separation of data into
categories, it is normally done for identifying the
categories contributing to the problem

Types

➢ Material Based

➢ Quality Based

➢ Worker Based

➢ Machine Based

➢ Process Based
24
 Statistical Process Control

Histograms
➢ It provides a simple and common language
➢ Show the variation in products coming from a process.
➢ Give a picture to the nature of the variation

Purpose
➢ To find fact and Visualization of Data
➢ To know the central value of the Group
➢ To know the extent of variation in the Group
➢ To estimate the % Non-Conformance
25
 Statistical Process Control
Histograms
➢ Cell –
The interval along the scale of measurement
of each of the ordered classes
➢ Cell Boundaries –
The values of the measurement at the ends of cells
➢ Cell Interval –
It is the difference between the upper & lower
boundaries of the cell
➢ Cell Midpoint –
It is the value equidistant from the cell boundaries
➢ Frequency –
The no. of observations in each of these Cells 26
 Statistical Process Control

Histograms – Construction Procedure

➢ Generate about 100 observations on the Quality Characteristics

➢ Observe the min. and max. value in the data

➢ Compute the Range ( R = max. – min. )

➢ Divide the range into several smaller intervals of equal width

( Width of class interval = R / total no. of intervals )

➢ Decide the class boundaries ( includes the smallest and the

largest of values )

➢ Go through the observation and make the tallies

27
 Statistical Process Control

Histograms – Construction Procedure

➢ Count the frequencies in each class

➢ Make a bar graph with class intervals in the X-axis and

frequencies in the Y-axis

➢ Draw a free hand curve connecting the mid-points of the bars

➢ Super-impose the specification limits

28
 Statistical Process Control

Histograms – Construction Procedure

Draw a Histogram from the data given below
SPECIFICATION : 130 - 150

141 140 141 141 138 140

140 142 139 139 143 141
140 139 139 140 139 139
140 137 145 138 141 141
140 140 141 143 138 139

29
 Statistical Process Control

Histograms
Cell
25
LSL USL

20

15

Frequency

10

Cell width
5

0
5-10 10-15 15-20 20-25 25-30 30-35 35-40 40-45 45-50 50-55 55-60

Class Interval
30
 Statistical Process Control

Histograms - Examples

Centered on nominal - good Skewed

31
 Statistical Process Control

Interpretation of Histograms

LSL USL LSL USL

Normal, centered, Normal, centered,

good dispersion, spread too wide,
well within limits can not meet spec. -
process must be improved
or spec. changed

32
 Statistical Process Control

Basic Statistics
Types of Data
➢Continuous Data – Variables
Data that can be divided infinitely
- Temperature, Time, Pressure, Voltage
➢Nominal / Discrete Data – Attribute Data
Categorical data
– Is it scratched? Is it chipped? Does if fit the gauge?
On time? Pass Fail?

33
 Statistical Process Control

Basic Statistics
Measures of Central Tendency
1. Average / Mean or X and  :
a) Sample Mean b) Population Mean
X = X  = X
n n

2. Median : Center of values / mid values ( to be arranged

in Ascending or Descending order )
E.g. (113 ,10, 70, 30,90) = 70

3. Mode : Most frequently occurring value

E.g. (1,2,3,3,4,5,6) Mode = 3
34
 Statistical Process Control

Basic Statistics
Measures Of Dispersion ( Spread )
1. Range : High value – low value ( Xmax – Xmin )

2. Standard Deviation: Sigma -  -

Average distance of each measurement within a
sample from the mean of the sample

Sample Population
2
 (x-x) 2  (x-x)
S= √ = √ N
n-1

35
 Statistical Process Control

Normal Distribution
AREA UNDER NORMAL CURVE
The area under the normal curve has a special relationship to the
population mean and population standard deviation.

-4s -3s -2s -1s X +1s +2s +3s +4s

68.26 %
95.46 %
99.73 %
99.994 %

36
 Statistical Process Control

Normal Distribution

➢ Most of measured characteristics and natural

phenomena exhibit this pattern of variation.

➢ Higher frequency around the central value and lesser

frequency at values away from central value.

➢ This pattern of variation is known as Normal pattern

37
 Statistical Process Control

Normal Distribution
Properties
➢ The frequency curve is Bell shaped

➢ The distribution is Symmetrical around average

➢ The distribution is completely known when its Average (X)

& Standard deviation ( S ) are known

➢ The area between any two vertical lines represents the

chance of obtaining an observation in this interval

38
 Statistical Process Control

Machine Capability Analysis

Machine capability study enables us to separate
variation due to machine and assess its capability to
produce to specifications
Purpose
➢ To identify the pattern of variation of output

➢ To estimate the capability of machine and compare

to product specification

➢ To estimate % of out of specification parts

which might occur if the machine proves incapable
39
 Statistical Process Control

Machine Capability Studies

Pre-study Requirements
➢ Specify the product characteristic to be studied and its
specifications
➢ Schedule 50 consecutive items
➢ Specify an uninterrupted run under normal operating
conditions with the machine Pre-set at nominal
➢ Ensure that all parts and materials used in the study are
themselves to the required standards and that operator
is fully trained

➢ Provide calibrated measuring equipment

40
 Statistical Process Control

Machine Capability Studies

Conditions for Study

➢ New (or) Rebuilt Equipment

➢ On-going capability Assessment for

existing equipment

41
 Statistical Process Control

Machine Capability Studies

Conducting The Study
➢ Record measurements in sequence from an uninterrupted production
run. If any non-planned interruption occurs ( Tool breakage, machine
fault ) discard the partially collected sample and start again

➢ Identify each part by its sequence. Retain all parts until analysis of
study data is over

➢ Observe process and note any unusual occurrence along with the
sequence number of the parts produced immediately after the event

➢ If a single measurement is found to be “way out of line” with others,

the part should be examined for any special cause. Record the results of
such an examination. Only if it is verified that the cause is not due to the
machine, may the part be excluded for determining the capability
42
 Statistical Process Control

Machine Capability Studies

Actions to be taken If equipment is not capable

➢ Work on the machine to improve capability and

repeat capability evaluation

➢ Check to see whether the design limits / tolerances /

engineering specifications can be extended

➢ Alternative machine or process should be sought to

provide capability

43
 Statistical Process Control

Variation and Process Control

➢ If there is no variation, there would be very few operational
problems. All parts would be identical to each other and
the entire output would be within specification ( or all
would be rejected )

➢ There is a variation among the products from any

manufacturing process

➢ We would like to completely get rid of variation but we

know that can’t be done

➢ The goal is to be able to measure / identify variation and

then work to minimize it
44
 Statistical Process Control

Sources of Variation
Equipment People
Machines Training
Fixtures Communication
Gauges Skill
Spindles Motivation
Collets Attitude

Indirect Material Methods

Tools Specification
Coolant Variation Speeds
Feeds
Bushings
Lubricants Process documentation
Chemicals

Orders Facilities
Direct Materials
Clarity Temperature
Hardness
Timeliness Cleanliness
Machinability
Product Mix Coolant system
Pick-up points
Adequacy Humidity
Quantity Noise
45
 Statistical Process Control

Variations
- are due to Causes

Causes

Chance / Assignable /
Inherent Special
Many in number, Few in number , economical to eliminate
not economical to eliminate Large in magnitude & Sporadic in nature

46
 Statistical Process Control

Chance Causes ( Inherent )

Sources of variation which are built into the process and
will be caused by such problems as Vibration, Speed, Power etc.,
These are the sources of random variation, the extent of which can
be measured and monitored

➢ A large number are in effect at any time

➢ Each has an individual effect that is too small to mention

➢ Only a change in the system will reduce that part of the variability

➢ Only management has the ability to make changes

➢ Remain constant over time

47
 Statistical Process Control

Assignable Causes
Sources of variation which will be due to specific
identifiable causes such as variation in Raw material, Operator,
Tool wear, Wrong setting etc., these removal are usually the
responsibility of someone who is connected directly with the operation

➢ Very few in effect at any time

➢ The effect is measurable

➢ They can be found and eliminated

➢ The machine operator is best able to discover and make changes

➢ They occur infrequently in an unpredictable manner

48
 Statistical Process Control

Chance & Assignable Causes

Chance Causes Assignable Causes

➢ Normal errors in ➢ Workmen Carelessness

handling material and ➢ Untrained Workmen
machine elements ➢ Poor maintenance
➢ Normal play in machine ➢ Defective material
slides
➢ Blow holes
➢ Hardness within
Tolerances ➢ Mix-up of material
➢ Small variation in job ➢ Wrong speed & feed
clamping ➢ Bush oversize
➢ Temperature change ➢ Wrong drawings
➢ Clamps broken
➢ Checking of hot jobs
with cold instruments
49
 Statistical Process Control

Process Control And Process Capability

PROCESS CONTROL
IN CONTROL
(SPECIAL CAUSES ELIMINATED)

OUT OF CONTROL
(SPECIAL CAUSES PRESENT)

SIZE

LOWER SPECIFICATION LIMIT UPPER SPECIFICATION LIMIT

IN CONTROL AND CAPABLE OF MEETING

PROCESS CAPABILITY SPECIFICATIONS ( VARIATION FROM
COMMON CAUSES HAVE BEEN REDUCED )

IN CONTROL BUT NOT CAPABLE OF

SIZE MEETING SPECIFICATIONS ( VARIATION
FROM COMMON CAUSES IS EXCESSIVE )

50
 Statistical Process Control

Control Charts
➢ Control charts are Visual representation of variations in the
selected parameters over a time
➢ A Control chart is simply a Run Chart with statistically
determined upper and lower limits drawn on either side of the
process
➢ Control charts are used to assess and maintain the stability of
the process
➢ Control chart indicates whether the process variation is natural
and to be expected ( Chance cause ) or due to special cause
(Assignable cause)
➢ Control chart provides a common language for
communications about the performance of a process

51
 Statistical Process Control

Control Charts
➢ SPC involves control followed by Improvement

➢ Process are initially brought under control by identifying and

eliminating the Assignable Cause variation. A controlled
situation is one when the process is operating only under the
influence of Chance Cause ( Inherent ) variation

➢ Control charts is a simple Graphical device which will help to

identify the presence of any Assignable Cause

➢ In short, Control chart will indicate How well the job is

running

➢ It gives when the job is running satisfactorily Or when something

has gone wrong needing corrective action
52
 Statistical Process Control

Purpose of Control Charts

The function of a Control

chart is to minimize the net
economic loss from Overadjustment
and Underadjustment

53
 Statistical Process Control

A Typical Control Charts

AVERAGE
UCL ➢ A Control chart has three
lines.

X ➢ One at the centre is central

line ( CL ) and the lines on
LCL
either side of the central line
are known as Control limits

RANGE
UCL ➢ The one above is called the
Upper Control Limit ( UCL )

R
➢ The one below is called the
LCL Lower Control Limit ( LCL )

54
 Statistical Process Control

Why Establish Controls

Stable Process : Unstable process:

Predictable operation Poor prediction or performance

SIZE

55
 Statistical Process Control

Control Charts
Control Limits Vs Specification Limits

➢ Specification limits / Drawing limits

Describe what the customer requires the process to produce
or what the product measurements must be.

➢ Control limits
Describe what the process is “actually giving” you.

If the UCL is above the USL, the LCL is below the LSL, or
both, it means that your process will predictably give you a certain
(almost certainly too high) percentage of out-of-spec output

56
 Statistical Process Control

Types of Control Charts

Variable Data Attribute Data
X - R Chart ➢ Np Chart – used for monitoring
– one unit / group no. of defective items in a
sample of given size
X – R Chart
– 2-5 units / group ➢ P Chart – used for monitoring
fraction defective
X – S Chart
- 6 or more units / group ➢ C Chart – used for monitoring
X – chart used for monitoring the defects in a single part or
assembled unit
process average
R – chart used for monitoring
➢ U chart – used for monitoring
process variability the defects, unequal sample
sizes
57
 Statistical Process Control

Control Charts - Variable

Outline for Preparing to Use Control Charts

➢ Collection - Gather data and plot the chart

➢ Control - Calculate Control Limits from

process data, using simple formulas

➢ Interpretation – Identify special causes of

variation: take local action to correct

➢ Capability – Quantify common cause variation,

take action on the system
58
 Statistical Process Control
Control Charts - Variable
Draw a Control chart from the data given below
Sl.no. X1 X2 X3 X4 X5 X BAR R

1 42 60 65 70 75 62.4 33

2 45 55 66 72 78 63.2 33

3 19 24 75 76 80 54.8 61

4 36 48 54 63 72 54.6 36

5 40 45 65 70 75 59 35

6 20 25 33 40 50 33.6 30

UCL : X – Chart = X + A2 * R = 54.6 + 0.58 * 38 = 76.64

LCL : X – Chart = X – A2 * R = 54.6 – 0.58 * 38 = 32.56

UCL : R – Chart = D4 * R = 2.11 * 38 = 80.18

LCL : R – Chart = D3 * R = 0 * 38 = 0
59
 Statistical Process Control

Construction of Control Charts

➢ Generate sample of size 4 or 5 at periodical intervals

➢ Compute Average and Range for each sample

➢ Repeat step 1 & 2 for about 25 sub-groups

➢ Compute average Range R

➢ The UCL for R - Chart = D4 * R

➢ The LCL for R - Chart = D3 * R

60
 Statistical Process Control

Construction of Control Charts

➢ Homogenize R – Chart

➢ Compute overall Average X

➢ The central line for X – Chart is at X

➢ The UCL for X – Chart = X + A2 * R

➢ The LCL for X – Chart = X – A2 * R

➢ Homogenize X - Chart

61
 Statistical Process Control

Construction of Control Charts

• Correction factors for sub-group size

n A2 D3 D4
2 1.880 0 3.268
3 1.023 0 2.574
4 0.729 0 2.282
5 0.577 0 2.114
6 0.483 0 2.004
7 0.419 .076 1.924

62
 Statistical Process Control

Chart Interpretation
2. Runs - 7 consecutive points above
1. Points outside of control limits
or below average
UCL UCL
CL CL

LCL LCL

3. Trends - 7 consecutive points 4. Patterns - 1/3 - 2/3 rule or any

moving up or down non-random pattern
UCL UCL

CL CL

LCL LCL

63
 Statistical Process Control

Chart Interpretation
➢ A Single points outside the Control limits

➢ Seven consecutive points or 10 out of 11 or 12 out of 14

consecutive readings on the same side of the Central Line

➢ Seven consecutive readings showing a continuously or

decreasing trend

➢ Readings showing a Cycle pattern or Saw-Tooth pattern

➢ Readings reaming too close to central line

➢ Readings suddenly going from one extreme to the other

➢ Readings showing large fluctuations & erratic Ups & Downs

64
 Statistical Process Control

Process Capability Analysis

➢ Process Capability concerns with the Variation

caused by all sources of variations : The Machine,
The Material used, The Methods employed, The
people involved and The environment as it affects
the product

➢ To make a process capability study, Data must be

collected over a fairly long period of time the
machine

65
 Statistical Process Control

Process Capability Analysis

• What is Cp.
– Cp is an index comparing the variation within the
process to the spec limits. The higher the number ,
the less variation in the process.

66
 Statistical Process Control

Process Capability Analysis

What is CPK
Cpk is an index (a simple number) which
measures how close a process is running to its
specification limits, relative to the target of
the process. The larger the index, the less
likely it is that any item will be outside the
specifications
Cpk Confirming Output (%) PPM

0.5 93.3 67,000 ppm

1.0 99.86 1,400 ppm
1.33 99.997 30 ppm
67
 Statistical Process Control

Process Capability Analysis

Cp & Cpk - Calculation:

Cpk = Min( USL - Xm) or (Xm-LSL)

3s
Cp = ( USL - LSL ) = Tolerance
6s 6s
s = Standard Deviation
= √(X-X)2 or R when n = 5, d2 = 2.33
n-1 d2
68
 Statistical Process Control

Process Capability Analysis

Comparison Of Cp & Cpk :

Cpk Cpk will always be

less than or at
the most equal
Low High to Cp.

Reduce
If the process is
Low Process Impossible Preferably
Cp Variability Centered, Cpk
and Cp shall be
Move Seek a equal
High process target
mean dimension
69
 Statistical Process Control

Process Capability Index

➢ 1.33 < Cpk :
The process capability is ample, review
the specification or process if excessive
➢ Cpk = 1.33 :
Ideal status. “ The goal” for all processes
➢ 1 < Cpk < 1.33 :
This status is desirable, but careful
control is necessary because defective units may be
generated when Cp approaches 1
➢ Cpk < 1 :
Since defective units exist, it is
necessary to take action, such as change operation
methods or total screening 70
 Statistical Process Control

Examples of Cp, Cpk

Cp=2, Cpk =1.5

LSL Cp =1 USL LSL USL
Cpk =1

Cp=.8, Cpk =.75

LSL Cp=Cpk =2 USL
LSL USL

71
 Statistical Process Control

6 Sigma And Cp & Cpk

3.4 ppm 3.4 ppm

+/- 1.5

Cp=2 Cp=2
Cpk = 1.5 Cpk = 1.5
Cp=Cpk =2

-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6

72
 Statistical Process Control

Control Charts – Attributes ( Fraction Rejected )

p - Chart
➢ Generate sample of size 25 or 30 at periodical intervals

➢ Observe proportion of Defective items p

➢ Repeat step 1 & 2 for about 25 times

➢ The central is at average proportion Defectives

per sample = p

➢ The UCL for p - Chart = p + 3 √p ( 1-p ) / n

The LCL for p - Chart = p - 3 √p ( 1-p ) / n 73
 Statistical Process Control

Control Charts – Attributes ( Fraction Rejected )

np - Chart
➢ Generate sample of size 25 or 30 at periodical intervals

➢ Observe the number of Defective items np

➢ Repeat step 1 & 2 for about 25 times

➢ The central is at average proportion Defectives

per sample = np

➢ The UCL for np - Chart = np + 3 √np ( 1-p )

The LCL for np - Chart = np - 3 √np ( 1-p ) 74
 Statistical Process Control

Control Charts – Attributes ( Fraction Rejected )

p - Chart
➢ Fraction Defective no. of rejected article
or p = -----------------------------
Fraction Rejected Total no. of inspected

np - Chart
Total no. of defectives
np = -----------------------------
No. of samples

Total no. of defectives np

p = ----------------------------- = -----
Total no. inspected n

Where, n = sample size 75

 Statistical Process Control

Control Charts – Attributes ( Fraction Rejected )

No.of units No.of Fraction % defective,
Days
inspected, n Defectives defective, p 100 * p For p-chart
1 50 4 0.08 8
2 50 0 0 0 UCL = 0.052 + 3√0.052(1-0.052) / 50
3 50 3 0.06 6 = 0.142 = 14.2 %
4 50 2 0.04 4
5 50 3 0.06 6 UCL = 0.052 - 3√0.052(1-0.052) / 50
6 50 5 0.1 10 = -0.442 = 0
7 50 1 0.02 2
8 50 2 0.04 4
9 50 2 0.04 4
10 50 0 0 0
11 50 3 0.06 6
For np-chart = 52/20 = 2.6
12 50 4 0.08 8
13 50 2 0.04 4
14 50 5 0.1 10
UCL = 2.6 + 3√2.6(1-0.052)
15 50 1 0.02 2
= 7.3
16 50 0 0 0
17 50 4 0.08 8 UCL = 2.6 - 3√2.6(1-0.052)
18 50 4 0.08 8 = -2.1 = 0
19 50 5 0.1 10
20 50 2 0.04 4
52 0.052 52% 76
 Statistical Process Control

Control Charts – Attributes ( non-conformities )

c – Chart – a) no. of surface defects observed in a roll of coated paper
b) no. of air bubbles in a glass bottle

➢ Generate sample periodically

➢ Observe the no. of defects ‘ C ’ in the sample piece / part

➢ Repeat sampling for 20 or 25 times

➢ The central line is plotted at average no. of Defects / part c

➢ The UCL for c - Chart = c + 3√c

➢ The LCL for c - Chart = c - 3√c

77