Process Quality
Control
Haritha Saranga
Conformance Design
Quality Quality
Process Product
Capability Quality
Supplier Process Product Customer
Quality Design Design Needs
Production Marketing
& Quality & Design
Control
Types and causes of variability
◼ Normal Variability
Occurs due to common causes
Need to redesign the system to reduce this
This is
what SPC
◼ Abnormal Variability tracks
Occurs due to special causes
Can be eliminated easily in the short run
Types of Control Charts
◼ Variable based control charts
X bar and R Charts
◼ Attribute based control charts
P chart
C chart
Measurement Methods
◼ Variable Based
Detailed observation of the characteristic (such as length, diameter,
weight)
measurement will be expensive and more time consuming but will
provide a wealth of information about the process
◼ Attribute Based
simple clustering of the characteristic into a few categories (such as
good or bad)
Two frequently used attribute measures are:
◼ Proportion of defects (denoted as p)
◼ Number of defects (denoted as c)
measurements are easy to make, quick & less expensive, but reveal
very little information about the process
Sample Data – Collected 5 times a
day for 20 days
Day
Time 1 2 3 4 5 6 7 8 9 10
9a.m 81 82 80 74 75 81 83 86 88 82
11a.m 73 87 83 81 86 86 82 83 79 84
1p.m 85 88 76 91 82 83 76 82 86 89
3p.m 90 78 84 75 84 88 77 79 84 84
5p.m 80 84 82 83 75 81 78 85 85 80
Day
Time 11 12 13 14 15 16 17 18 19 20
9a.m 86 86 88 72 84 76 74 85 82 89
11a.m 84 83 79 86 85 82 86 85 84 80
1p.m 81 78 83 80 81 83 83 82 83 90
3p.m 81 80 83 79 88 84 89 77 92 83
5p.m 87 83 82 87 81 79 83 77 84 77
Process Control Chart
Process Measure
Time
◼ Information: Monitor process variability over time
◼ Control Limits: Average + z * Normal Variability
◼ Decision Rule: Ignore variability within limits as “normal”
Investigate variation outside as “abnormal”
◼ Errors: Type I - False alarm (unnecessary investigation)
Type II - Missed signal (to identify and correct)
Criterion Notation Explanation
Sub-group size n Each sub-group consists of n
observations
Sub-group measures Variable measures Based on the n observations, each of
▪ Mean - X these measures can be computed for
▪ Range – R each sub-group
Attribute measures
▪ Proportion of defects – p
▪ No. of defects – c
Notations for Number of sub- m m measurements are used for
control charts groups establishing the sampling
distribution and control chart
Mean of the Variable measures These means are nothing but the
▪ Mean - X =
sampling distribution X process average (center line) of the
of the sub-groups m
respective control charts formed
using each of these measures.
▪ Range – R =
R
m
Attribute measures
▪ Proportion of defects –
p=
p
m
▪ No. of defects - c =
c
m
Standard deviation of Attribute measures Three standard deviations on either
the sampling ▪ Proportion of defects – side of the center line are the control
distribution of the p(1 − p) limits for the respective charts.
sub-groups p =
n
▪ No. of defects - c = c
X bar and R Chart
Setting control limits for small samples
Computations for the X chart
Process average or center line: X =
X
m
Upper Control Limit : UCL X = X + A2 R
Lower Control Limit: LCL X = X − A2 R
Computations for the R chart
Process average or center line: R=
R
m
Upper Control Limit : UCLR = D4 R
Lower Control Limit: LCL R = D3 R
* The values A 2, D 3 and D 4 can be read directly from tables (see table in the next slide)
Coefficients for computing LCL and UCL in
X-bar and R charts*
Sample 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 0.076 1.924
8 0.373 0.136 1.864
9 0.337 0.184 1.816
10 0.308 0.223 1.777
Source: Juran, J.M. and F.M. Gryna, (1995), “Quality Planning and Analysis”, Tata McGraw-Hill, 3rd Edition, New Delhi, pp 385.
X Bar ( X ) Chart
UCL
LCL
◼ Average, X = 82.5 kg
◼ Average Range R = 10.1 kg
◼ Control Limits = X + A2 R
= 82.5 + (0.577)(10.1) = [76.67, 88.33]
◼ Process is “in control” (i.e., the mean is stable)
Range (R) Chart – small samples
20
UCL
15
Range
10
5
0
LCL
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Day
◼ Average Range R = 10.1 kg
◼ Lower Control Limit : D3 *10.1 = 0*10.1=0
◼ Upper Control Limit : D4 *10.1 = 2.144*10.1=21.654
◼ Process is “in control” (i.e., variation is stable)
P chart
Setting the control limits
Computations for p chart
Process average or center line: p=
p
m
p(1 − p)
Upper Control Limit : UCL p = p + 3
n
p (1 − p )
Lower Control Limit: LCL p = p − 3
n
P Chart
An example
p Chart
p Centre Line UCL LCL
0.21
0.18
Proportion of defects
0.15
0.12
0.09
0.06
0.03
0.00
1 2 3 4 5 6 7 8 9 10 11 12
Sample No.
How to use control charts?
Stable
Unstable
Trend
Cyclical
Shift