Process Capability Index (CPK)
Process Capability Index (CPK)
Process Capability Index (CPK)
When we speak of a manufacturing process, the first and foremost thing that comes bundled with
it is the process capability. Process Capability can be defined as the repeatability and uniformity
of a process with respect to the customer’s demands in terms of product parameter specification
limit. The specification limit measures the degree to which the process meets the customer
requirement. Many organization have their own product standard which are based on their
process capabilities. For example, a stainless steel making unit has a product standard for its end-
product like coils and slabs as per its various production entities capabilities.
Now that we have our process capability defined, it is required to measure the process
consistency which can be carried out graphically. These graphical representations which portrays
the measure of the customer requirement to the process consistency are called capability indices.
Capability indices helps us to portray the process distribution with respect to the product
specification limits. Capability indices can be used to determine if the process, given its natural
variation, can achieve the established specifications.
Where, Cp is the potential capability indicating how well a process could perform if it were
centered on target and considers only the spread within-subgroup variation and doesn’t take into
account the closeness of the estimated mean to the specification limits, Cpk, measures the
probable process capability by measuring the distance between the process average and the
specification limits, when compared to the process spread.
CpL measures the relative closeness of the process mean to the lower specification limit
CpU measures the relative closeness of the process mean to the upper specification limit
Cpk equals the lesser between CpL or CpU
The difference between the two indices is the way the process standard deviation (σ) is
calculated. Cpk uses σ which is estimated using (R-Bar/d2) or (X-Bar/C2).
A higher Cpk value paves way to a better process. For instance, Equipment A has a Cpk value
of 1.6 and Equipment B has a Cpk value of 1.2. By comparing the Cpk values, one can infer that
Equipment A is better than Equipment B. As Cpk counts on specification limits and sigma, we
can also decide on the process efficiency, yield processed and machine losses.
Cpk Process
0.5 86.8%
0.8 98.4%
1.0 99.7%
1.2 99.97%
1.33 99.99%
Cpk ≥ 1.33 indicates that the process is capable and meets its specification limits. If the Cpk
value is less than 1.33, it suggests that process variation is too wide when compared to the
specification limits or the process average is shifted from the requisite value.
Natural
Design Process
specificati Variation
on (USL (X-Bar)
& LSL)
Process
Variation
(σ) spread
Process Capability
Four possible outcomes arise when the natural process variability is compared with the design
specifications or customer requirement or expectations.
monitoring.
Case 3 Cpk < 1 This process will produce
& (The higher number of non-
Case 4 Process is conforming product.
not The process average is off-
Capable) center.
In such cases, modification
A Non-Capable Process: Voice of the Process >
can shift the process mean Customer Expectations.
back to target.
If no action is taken, a
substantial portion of the
production will fall outside the
specification limit even though
the process might be in
A Non-Capable Process: Voice of
statistical control. Process>Customer Expectations.
Capability indices described here attempt to represent with a single number of the process
capability considering certain assumptions, precautions and conditions like process distribution
is approximately normal or bell shaded, data chosen includes all natural variations and the
number of samples used for calculation has a significant impact on the accuracy and correctness
of the Cpk estimate.
Reference:
2. Wikipedia.org
3. Chapter-6, wiley.com.