Process capability, Cp, tells how well a process meets customer’s specifications. It is defined as the ratio of engineering tolerance (USL – LSL) to the natural tolerance (-3σ to 3σ) . However, a process can be capable of meeting specifications, but not be meeting specifications if the process is not centered relative to the specifications. Therefore, process capability index, Cpk, is introduced to take into account of the centeredness.
Cp and Cpk can be calculated by the above equations. The definition of the symbols are quite straightforward. If the process is not centered, Cpk will be less than Cp. It needs to be noted that σ is within subgroup variation. This estimate of the process standard deviation comes from a range control chart. as shown in the following figure.
R is a measure of the variation within the subgroup. To calculate σ, you use the following equation:
R bar is the average range and d2 is a control chart constant that depends on subgroup size. d2 can be got from the table in the link.
CPK VS PPK
The difference is the variation, overall variation, s or within subgroup, σ. σ can be calculated by above equations. s is calculated by the following equation:
N is the total number of data points.
If the process is in statistic control, Cpk and Ppk should be similar since σ will be close to s. Any significant difference between Cpk ad Ppk should be put alert if the process is in statistic control or not.
The above information is from the linked website. More details and other useful information can be got from this and other posts in the website. This post is also a good one.