Professional Documents
Culture Documents
Lecture 2
Rajiv Gupta
BITS Pilani
Agenda
• Module 1
– Recap of Lecture 1
• Module 2
– Xbar and S Charts
– X and Moving Range Charts
• Module 3
– Attribute Control Charts
• Module 4
– Using Control Charts
• Module 5
– Process Capability Cp and Cpk
2
• Begin Module 1
– Recap of Lecture 1
3
The Normal Distribution
4
Process Output-- Run Chart & Control
Charts
• Run chart – shows process output (e.g. X-bar plotted
as a function of subgroup number).
• Control Chart– Run chart with control limits (i.e. LCL,
UCL)
• Note: Control limits are not the same as specification
limits (i.e. LSL, USL)
• Two types of control charts
- Variables Control Charts
- Attribute Control Charts
5
Variables Data
• Points falling outside the control limits can be
investigated, assignable causes can be
identified, process can be fixed, and natural
variability of the process can be computed.
• The following pair of charts are used for SPC
X-bar and R
X-bar and S
X and RM
6
Example of Control Charts: X-bar
and R Charts
7
• End of Module 1
8
• Begin Module 2
– Xbar and S Charts
– X and Moving Range Charts
9
Example Xbar and S Chart
Pc 1 Pc 2 Pc 3 Pc 4 Pc 5 Xbar S
11
Calculation of UCL and LCL
13
X and Moving Range (MR) Charts
• X and Moving Range (MR) Charts are used
when each subgroup consists of only one
piece
• This typically happens when the testing is
destructive testing or the part has a long cycle
time/lead time
• MR = difference between two successive parts
sampled
14
X and Moving Range Chart
Subgroup Value (X) Moving Range
1 8
2 8.5 0.5
3 7.4 1.1
4 10.5 3.1
5 9.3 1.2
6 11.1 1.8
7 10.4 0.7
8 10.4 0
9 9 1.4
10 10 1
11 11.7 1.7
12 10.3 1.4
13 16.2 5.9
14 11.6 4.6
15 11.5 0.1
16 11 0.5
17 12 1
18 11 1
19 10.2 0.8
20 10.1 0.1
21 10.5 0.4
22 10.3 0.2
23 11.5 1.2
24 11.1 0.4
16
For the MR Chart
17
X and Moving Range Charts
18
• End of Module 2
19
• Begin Module 3
– Attribute Control Charts
20
Attribute Measures
• Number of defective units in each subgroup are
identified during inspection.
• Proportion of defectives in each subgroup is computed
and plotted against subgroup number. (For P chart).
• The number of defective units are plotted against
subgroup number. (For NP chart.)
21
Attribute Charts-Sample Sizes
• Attribute charts require inspection of large number of
items – which is especially true as occurrences of
defectives and defects are rare (or lower).
• To develop a P (or NP) Chart, number of items (n) in each
sample should be at least 1/p. So that we get at least 1
or 2 non-conforming items on a frequent basis in each
sample.
• P can be reported as DPM (Defects per million) parts.
22
P Chart
Subgroup n Defectives (np) p
1 100 6 0.06
2 100 7 0.07
3 100 10 0.1
4 100 7 0.07
5 100 5 0.05
6 100 13 0.13
7 100 9 0.09
8 100 7 0.07
9 100 8 0.08
10 100 9 0.09
11 100 8 0.08
12 100 12 0.12
13 100 6 0.06
14 100 15 0.15
15 100 8 0.08
16 100 5 0.05
17 100 8 0.08
18 100 8 0.08
19 100 10 0.1
20 100 8 0.08
21 100 4 0.04
22 100 6 0.06
23 100 6 0.06
24 100 9 0.09
25 100 8 0.08
23
P Chart for
constant
sample size
24
For the P Chart
25
P Chart for constant sample
size
26
NP Chart For Defectives
Subgroup n Defectives np p
1 30 1 0.03
2 30 3 0.10
3 30 2 0.07
4 30 1 0.03
5 30 2 0.07
6 30 3 0.10
7 30 2 0.07
8 30 0 0.00
9 30 4 0.13
10 30 2 0.07
11 30 2 0.07
12 30 1 0.03
13 30 1 0.03
14 30 2 0.07
15 30 3 0.10
16 30 2 0.07
17 30 5 0.17
18 30 2 0.07
19 30 3 0.10
20 30 2 0.07
27
NP Chart for
defectives
28
Centerline = Average defectives = 2.15
29
NP Chart for defectives
30
• End of Module 3
31
• Begin Module 4
– Using Control Charts
32
Purpose of Control Charts
• The purpose of statistical control charts is to
detect assignable (special) causes of variation
in the process so that we have the chance to
eliminate those causes and get the process
into control.
• The natural variability of the process (σ) can
be estimated only after the assignable causes
are removed and the process is in statistical
control.
33
Processes and Control
• A process is said to be in control when special
causes of variation have been eliminated to
the extent that the points plotted on a control
chart remain within the control limits and
exhibit random (normal) variation between
control limits.
• Note: A state of control does not indicate a
capable process.
34
Initializing Control Charts
• When we are first setting up a control chart, we need
to chart some initial data, at least 25 points
• From these 25 data points we need to ascertain
whether the process is in control. If there are
assignable causes for variation, we should remove
those points and recalculate the control limits
• Generally we first plot the Range values and try to
get stability, before we plot the Xbar values
• Once the process I shown to be stable it can be used
to monitor the process
35
Monitoring the Process
• It is customary to recalculate the control limits every
month to reflect the current state of the process
• The control chart must be displayed in a prominent
place on the shop floor so that everyone may be
aware of the status of the process
• While monitoring the process, any points outside the
control limits must be investigated and any
assignable causes identified and removed to bring
the process back in control
36
Control Chart Zones
(From: Kolarik)
37
Runs
Rules
(From: Kolarik)
38
Reducing Fraction Defectives
39
• End of Module 4
40
• Begin Module 5
– Process Capability Cp and Cpk
41
X-bar Chart, Process Capability and
Specification Limits
42
Capability Analyses
• Process Capability = 6 σ
• Tolerance = USL-LSL
44
Potential Capability Index = Cp
45
Actual Capability Index = Cpk
46
Actual Capability Index = Cpk
47
Relationship between Cp and Cpk
48
• End of Module 5
49
Control Limits for variables charts
50
Constants for
computing
control limits of
variables charts
(From:
Besterfield’s
Textbook)
51
Constants for computing control limits of variables charts
52
Control Limits for Attribute Charts
53