Control Charts For Variables

Adapted from:
Prof. Dr. Sha’ri Mohd Yusof
UTM

DISCLAIMER
The information contained in this document is compiled from various sources and provided on an "AS IS" basis for teaching and learning
purposes only without any representations, conditions or warranties whether express or implied, including any implied warranties of satisfactory
quality, completeness, accuracy or fitness for a particular purpose.
Introduction
 Concept of variation
 No 2 things are alike
 Variation exists - even if variation small and
appears same, precision instruments show
differences
 Ability to measure variation necessary before
can control.
Introduction
Basically 3 categories of variation in piece
part production (e.g. Light bulbs, washer,
nuts, etc.)
1. Within piece - e.g. surface roughness
2. Piece to piece - eg. dimensions
3. Time to time - e.g. morning & afternoon, tool
wear, workers tired
Sources of variation
 Many factors contribute to variation
 Source of variation - combination of equipments,
materials, environment, operator, etc.
 Equipment - tool wear, electrical fluctuations
for welding
 Material - tensile strength, moisture
content (e.g. raw material)
 Environment - temperature, light, humidity etc.
 Operator - method, SOP followed,
motivation level, training
 Inspection - inspector, inspection equipment,
environment
Causes Of Variation - Chance &
Assignable
 Chance or random causes are unavoidable
 As long as fluctuate in natural/expected/stable
pattern of chance causes of variation which are
small – it is OK
 This is in ‘state of statistical control’
 When causes of variation large in magnitude; can
be identified, classified as assignable causes of
variation. If present, process variation is excessive
(beyond expected natural variation)
 ‘state of out of control’ – assignable cause
Control Chart Method
 Control chart - means
of visualizing variations
that occur in the central
tendency and
dispersion of a set of
observations
 Graphical record of a
particular quality
characteristic –
hardness, length, etc
Control Chart Method
 Control limits are not specification limits
 CL are permissible limits of a quality
characteristic
 Evaluate variations in quality subgroup to
subgroup
 Limits established at 3 standard dev. from
central line; for normal distribution – we
expect 99.73% of items would lie within the
limits
Control Chart Method
Objectives Of Variable Control
Chart

 For quality improvement

 To determine process capability
 For decisions in product specifications
 Provide information on production processes for
current decisions
 SOSC – leave alone
 SOOC – investigate, solve, rectify, improve
 Make decisions on recently produced items -
release next process, customer or other disposition
method, sorting, rework, reject
Variable Control Chart –
x (average)- R chart
Variable control chart – monitor measured data
quality characteristic
Steps
1. Select quality characteristic
2. Choose rational subgroup
3. Collect data
4. Determine trial limits and central line
5. Establish revised central line and control limits
6. Achieve the objective
Variable Control Chart –
x (average)- R chart
1. Select quality characteristic
 Measurable data (basic units, length, mass,
time, etc.)
 Affecting performance, function of product
 From Pareto analysis – highest % rejects, high
production costs
 Impossible to control all characteristics -
selective or use attributes chart
2. Choose rational subgroup
 Rational subgroup - variation within the group
due only to chance causes and can detect
between groups changes
 Two ways selecting subgroup samples
1. Select subgroup samples at one instant of time or as
close as possible
2. Select period of time products are produced
- Rational subgroup from homogeneous lot : same machine,
same operator
- Decisions on size of sample empirical judgment + relates to
costs
 choose n = 4 or 5  use R-chart
 when n  10  use s-chart
- frequency of taking subgroups often enough to detect process
changes
- Guideline of sample sizes/frequency using
 Say, 4000 parts/day
 75 samples
 if n = 4 19 subgroups
 or n = 5 15 subgroups
3. Collect data
 Use form or standard check sheet
 Collect a minimum of 25 subgroups
 Data can be vertically / horizontally
arranged

Subgroup Number

Measure 1 2 3 4 5 …… ….. …. 25

x1 35 34
x2 40 40
x3 32 38
x4 37 35
x5 34 38

x 35.6 37.0

R 8 6
Example
Example …

(6.35 + 6.40 + 6.32 + 6.37) / 4

= 6.36
Example …

(6.40 – 6.32)
= 0.08
4. Determine trial control limits
 Calculate Central line g
R = i
g Ri
 X =  xi 1
i 1
g
g UCL x  x  3 σ x  x  A 2 R
 X = avg. of subgroup avg.
 xi = avg. of ith subgroup LCL x  x  3 σ x  x  A 2 R
 g = no. of subgroups
 R = avg. of subgroup ranges
 Ri = range of ith subgroup
 Where A2, D4, D3 are factors -
vary according to different n UCL R  R  3 σ x  D4 R

LCLR  R  3 σ x  D3 R
X = 160.25 / 25 = 6.41 mm
R = 2.19 / 25 = 0.0876 mm

UCL x  x  3 σ x  x  A 2 R
= 6.41 + (0.729x0.0876) = 6.47 mm

LCL x  x  3 σ x  x  A 2 R
= 6.41 - (0.729x0.0876) = 6.35 mm

UCL R  R  3 σ x  D4 R
=(2.282 x 0.0876) = 0.20 mm

LCLR  R  3 σ x  D3 R
=(0 x 0.0876) = 0 mm
Example …
5. Revised Control Limits
 First plot preliminary data collected using control
limits & center lines established in step 4
 Use/adopt standard values, if good control i.e. no
out-of-control points

X  Xo R  Ro

 If there are points out-of-control discard from data,

look at records – if show an assignable cause – don’t
use
 x  x d
X new 
g  gd
= 160.25 – 6.65 – 6.51 = 6.4 mm
25 - 2
R  Rd
Rnew 
g  gd
= 2.19 – 0.3 = 0.079 mm
25 - 1
•Limits for both charts
become narrower after
revised
•Revised limits used to
report / plot future sub-
groups
•For effective use – chart
must be displayed and
easily seen
6. Achieving objective
 Initiate control charts results in quality
improvement
 Less variation in sub-group averages
 Reduction in variation of range
 Can reduce frequency of inspection - monitoring
purpose – even once/mth.
How Control Chart Helps In QI
 Psychological effect to do better – example -
maintenance group helps adjust process center
 Purchasing involved in changing material
supplier to ensure consistent quality
 Production – standardize work methods,
use/develop new tooling
 Improvements must be from investigation of
assignable causes (need technical back up)
 Sample Std. Deviation Chart
(x - s control chart )
 Both R and s measure dispersion of
data
 R chart - simple, only use XH
(highest) and XL(lowest)
UCLx  X  A3 s
 s chart - more calculation - use
 X  A3 s
ALL xi’s  more accurate, need LCLx
calculate sub-group sample UCLs  B4 s
standard deviation LCLs  B3 s
so s
 When n<10, R chart  s chart o  
C4 C4
 n  10 - s chart better , R not
accurate any more
State Of Control
 When assignable causes eliminated and points
plotted are within C.L.- process state of control
 Further improvement through changing basic
process, system

What are the characteristics of process in control?

(natural pattern of variation)
 34% within 1 from Center Line
 13% between 1 & 2
 2.5% of plotted points - 2  3
 Points located back & forth across center line random
way
 No points out of control
Advantages of Process In Control
 Individual parts will be more uniform – less
variation and fewer rejects
 Cost of inspection will decrease
 Process capability easily attained
 Trouble can be anticipated before it occurs
 Percentage of parts fall between two values
can be predicted with highest degree of
accuracy, e.g. filling machines
 Type I and Type II errors
 Subgroup averages forms frequency distribution which is normal
distribution and limits – established at 3 from center line.
 Choice of 3 is economic decision with respect to 2 types of error
Process Out Of Control
Process Out Of Control

2. Unnatural runs of variation

even within 3 limits
 7 or more points above or
below center line (in a row)
 10 out of 11 points on one
side
 12 out of 14 points on one
side
 6 points
increasing/decreasing
 Z out of 3 in Zone A (WL)
 4 out of 5 in Zone B
Process Out Of Control

3. For two zones 1.5 each

 2 or more points beyond
1.5
ANALYSIS FOR OUT-OF-CONTROL

Patterns
1. Change/Jump in level
 shift in mean
 Causes - process parameters
change, diff / new operator,
change in raw material

2. Trend or steady change in

level
 drifting mean – common,
upward or downward direction
 tool wear, gradual change in
temp. viscosity of chemical
used
ANALYSIS FOR OUT-OF-
CONTROL
3. Recurring cycles
 wavy, periodic high & low
points
 seasonal effects of mtl.
 Recurring effects of temp.,
humidity (morning vs
evening)

4. Two populations (mixture)

 many points near or
outside limits
 due to: large difference in
material quality, 2 or more
machines, different test
method, materials from
different supplier
Exercise
 Control charts for X-bar and R are kept on the
weight in kilograms of a color pigment for a batch
process. After 25 subgroups with subgroup size of
4, the result can be conclude as follows:
 Numbers of sub-group, g = 25
 Total X-bar = 52.08 kgs
 Total R = 11.82 kgs
 Assuming the process is in a state control, compute
the X-bar and R chart central line and control limit
for the next production period. Table
Answer
Diberi g=25, n=4
Untuk Carta x-bar ; daripada Jadual, A2 = 0.729
= Xbar / g = 52.08/25 = 2.0832

R bar = R / g = 11.82/25 = 0.4728

UCL = 2.0832 + (0.729 x 0.4728)
= 2.428
LCL = 2.0832- (0.729 x 0.4728)
= 1.739
Untuk Carta R;
UCL = 2.282 x 0.4728 = 1.079
LCL = 0
 References

1. Besterfield (2001). Quality Control, Prentice Hall.

2. Besterfield, D. H.(2013) Quality Improvement. 9th ed.

Pearson.