Process Capability

Process Capability
Process Capability is an important concept
in SPC. Process capability examines
-- the variability in process characteristics
-- whether the process is capable of
producing products which conforms to
specifications

Process Capability
Process capability studies distinguish between
conformance to control limits and conformance
to specification limits (also called tolerance
limits)
-- if the process is in control, then virtually all
points will remain within control limits
-- staying within control limits does not necessarily
mean that specification limits are satisfied
-- specification limits are usually dictated by
customers
Process Capability: Concepts
The following distributions show different
process scenarios. Note the relative positions
of the control limits and specification limits.
Process Capability: Concepts
In control and product
meets specifications.

Control limits are within
specification limits
UCL: Upper Control Limits
LCL: Lower Control Limits
USL: Upper Specification Limits
LSL: Lower Specification Limits
Process Capability: Concepts
In control but some
products do not meet
specifications.

Specification limits are
within control limits
Data from process with low capability
Data from process with
medium capability
Data from process with high capability
relative capability
Process capability: capability index
The capability index is defined as: Cp = (allowable
range)/6 =T (Tolerance)/ 6= (USL - LSL)/6s


The distribution of process quality
is often assumed to be
approximated
with a normal distribution.

Px3=99.73%,
Normal Distribution
x
f (x)
0
=0.5
=1
=2

0
f (x)
x

1

2

The normal distribution N(,
2
) has several
distinct properties:
--The normal distribution is bell-shaped and is
symmetric
--The mean, , is located at the centre
-- is the standard deviation of the data

The smaller the steeper the curve

For same changing the value of
is to move the curve without any
change in its shape
3 Principle






XN(,
2
)
P{-<X<+}=(1)-(-1)=2(1)-1=0.6826
P{-2<X<+2}=2(2)-1=0.9544
P{-3<X<+3}=2(3)-1=0.9974

-3 +3
0.9974

The probability for X to fall within (-
3,+3)is 99.74%, and the probability
for falling outside this interval is only
0.26% which is considered almost
impossible. That characteristic of normal
distribution is called 3 principle.
Applying this principle in QM can judge
whether there is abnormity appearing in
the process, since three standard
deviations above and below the process
mean represent almost all the fluctuation
range of the process.
Process capability: capability index
The capability index (T/6) show how well a process
is able to meet specifications. The higher the value
of the index, the more capable is the process:
-- Cp < 1 (process is unsatisfactory)
-- 1 < Cp < 1.6 ( process is of medium relative
capability)
-- Cp > 1.6 (process shows high relative capability):
better to analyze the actual specifications
(Tolerance) and process technics to save resources
in enhancing the process capability, such as the
increased accuracy of equipment.
Process capability:
process performance index
The capability index
-- considers only the spread of the
characteristic in relation to specification limits
-- assumes two-sided specification limits
The product can be bad if the mean is not set
appropriately. The process performance
index takes account of the mean () and is
defined as:
Cpk = min[ (USL - )/3, ( - LSL)/3]
Process capability:
process performance index
The process performance index can also
accommodate one sided specification limits
-- for upper specification limit:
Cpk = (USL - )/3
-- for lower specification limit:
Cpk = ( - LSL)/ 3
Process capability: the message
The message from process capability studies
is:
-- first reduce the variation in the process
-- then shift the mean of the process towards
the target
This procedure is illustrated in the diagram
below:

Basic Forms of Statistical Sampling for
Quality Control
Sampling to determine if the process is within
acceptable limits (Statistical Process
Control).

Sampling to accept or reject the immediate
lot of product at hand (Acceptance
Sampling).

Process Control
Process Control is concerned with
monitoring quality while the production or
service is being conducted.
UCL
LCL
Samples
over time
1 2 3 4 5 6
UCL
LCL
Samples
over time
1 2 3 4 5 6
UCL
LCL
Samples
over time
1 2 3 4 5 6
Normal Behavior
Possible problem, investigate
Possible problem, investigate
Statistical Process Control
-- Control Charts
Control charts
Processes that are not in a state of statistical control
-- show excessive variations
-- exhibit variations that change with time
A process in a state of statistical control is said to be
statistically stable. Control charts are used to detect
whether a process is statistically stable. Control
charts differentiates between variations
-- that is normally expected of the process due chance
or common causes
-- that change over time due to assignable or special
causes


Control charts: common cause variations
Variations due to common causes
-- have small effect on the process
-- are inherent in the process because of:
the nature of the system
the way the system is managed
the way the process is organized and operated
-- can only be removed by
making modifications to the process
changing the process
-- are the responsibility of higher management

Control charts: special cause variations
Variations due to special causes are
-- localized in nature
-- exceptions to the system
-- considered abnormalities
-- often specific to a
certain operator
certain machine
certain batch of material, etc.
Investigation and removal of variations due to special
causes are key to process improvement
Note: Sometimes the delineation between common and
special causes may not be very clear.
Control charts: how they work
The principles behind the application of control charts
are very simple and are based on the combined use of
-- run charts
-- hypothesis testing
The control limits most commonly used in
organizations are plus and minus three standard
deviations. We know from statistics that the chance
that a sample mean will exceed three standad
deviations, in either direction, due simply to chance
variation, is less than 0.3 percent (i.e., 3 times per
1000 samples). Thus, the chance that a sample will
fall above the UCL, or below the LCL because of
natural random causes is so small that this occurrence
is strong evidence of assignable variation.
Control charts: how they work
The procedure is to
-- sample the process at regular intervals
-- plot the statistic (or some measure of
performance), e.g.
mean
range
variable
number of defects, etc.
-- check (graphically) if the process is under
statistical control
-- if the process is not under statistical control, do
something about it

Control charts: types of charts
Different charts are used depending on the nature of the charted
data. Commonly used charts are:
-- for continuous (variables) data
Shewhart sample mean ( -chart)
Shewhart sample range (R-chart)
Shewhart sample (X-chart)
Cumulative sum (CUSUM)
Exponentially Weighted Moving Average (EWMA) chart
Moving-average and range charts
-- for discrete (attributes and countable) data
sample proportion defective (p-chart)
sample number of defectives (np-chart)
sample number of defects (c-chart)
sample number of defects per unit (u-chart)
x
Control charts: assumptions
Control charts make assumptions about the
plotted statistic, namely
-- it is independent, i.e. a value is not influenced
by its past value and will not affect future
values
-- it is normally distributed, i.e. the data has a
normal probability density function

Normal Probability Density Function
The assumptions of normality and independence enable
predictions to be made about the data.
Control charts: interpretation
Control charts are normal distributions with an
added time dimension.
Control charts: interpretation
Control charts are run charts with superimposed
normal distributions.
Control charts: run rules
Run rules are rules that are used to indicate out-of-
statistical control situations. Typical run rules for
Shewhart X-charts with control and warning limits are:
-- a point lying beyond the control limits
-- 2 consecutive points lying beyond the warning limits,
namely within the area of 2 ~3 or even beyond the
control limits
-- 7 or more consecutive points lying on one side of the
mean
-- 5 or 6 or more consecutive points going in the same
direction (indicates a trend)
-- Other run rules can be formulated using similar principles
UCL
CL
LCL
X
LCL
CL
UCL
X
attention investigate Prompt action
Trend
Control charts: run rules
If only chance variation is present in the
process, the points plotted on a control
chart will not typically exhibit any pattern.
If the points exhibit some systematic
pattern, this is an indication that
assignable variation may be present and
corrective action should be taken.
Attribute Measurements (p-Chart)
p =
T o t al N u m b er o f D efe ct i v es
T o t al N u m b er o f O b s e rv at i o n s
n
s
) p - (1 p
=
p
p
p
3 - p = LCL
3 + p = UCL
s
s
Given:
Compute control limits:
Example of Constructing a p-chart:
Sample n Defectives p
1 100 4 0.04
2 100 2 0.02
3 100 5 0.05
4 100 3 0.03
5 100 6 0.06
6 100 4 0.04
7 100 3 0.03
8 100 7 0.07
9 100 1 0.01
10 100 2 0.02
11 100 3 0.03
12 100 2 0.02
13 100 2 0.02
14 100 8 0.08
15 100 3 0.03
1. Calculate the
sample proportions, p
(these are what can
be plotted on the p-
chart) for each
sample.

Example of Constructing a p-chart:
2. Calculate the average of the sample proportions.
0.036 =
1500
55
= p
3. Calculate the standard deviation of the sample
proportion
.0188 =
100
.036) - .036(1
=
) p - (1 p
=
p
n
s
Example of Constructing a p-chart:
4. Calculate the control limits.
3(.0188) .036
UCL = 0.0924
LCL = -0.0204 (or 0)
p
p
3 - p = LCL
3 + p = UCL
s
s
Example of Constructing a p-chart:
5. Plot the individual sample proportions and the control
limits
UCL
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Sample number
p
UCL
CL