Process Capability Index (Cpk) measures the process potential and performance of processes. Process capability evaluates the productivity of an in-control process to the requirement limits by using capability indices. It can be applied only when the output response displays a normal distribution.
The effort of a process typically has one or more than one measurable characteristics that are used to specify outputs. These can be evaluated statistically. The input of a process is likely to meet up customer requirements, specifications, or product tolerances. Engineering should conduct a process capability study to find out the degree to which the process can meet these potentials by improving constantly every process for planning, production, and service. This can be carried out through process capability index.
Cpk calculation emerges from the estimate of sigma based on the average range. Various ratios and indices have been developed to explain this relation between product specifications and the measured process performance. Cpk comprises of a centering factor as well as the variant factor. It has been extensively used in the manufacturing industry as an instrument for measuring process performance. Cpk computes the differentiation between the actual process average and the contiguous requirement limit over standard deviation. The two things which Cpk measures:
- The interpretations of the mean as how close they are to the centre of the upper and lower spec limits?
- How extensively the readings are spread?
The higher the range of Cpk, the improved is the ability of the process to complete its necessities. Many quality control researchers and practitioners have mentioned that Cpk is yield-based and is autonomous of the target. This may lead to failure in accounting for process centering with balanced tolerances, and resulted in presenting an even greater problem with imbalanced tolerances. To come out from the dilemma, several generalizations of Cpk have planned to tackle the processes with asymmetric tolerances
Cpk is used for measuring the potential capability of a system to meet customer needs. It should be used to examine a system’s ability to perform. And if Cpk is below than one, then the process is considered as incapable. And if it is equal to one or greater than one, it is considered to be capable of generating a product in the limits of specification. In the process of Six Sigma, Cpk equals 2.0. Moreover it is related to the variability or standard deviation of a process. As the Cpk is higher, the narrower is distribution of the process when contrasted with the limits of specification, and homogeneous is more to the product. With the increase in standard deviation, the index of Cpk gradually decreases. On the other hand, the capability to produce product increases outer surface of the limits of the specification at the same time.
The values of Cpk should be positive. It will be equivalent to zero in the stage when the average actual process is related or exempts one of limits of the requirement. The index of Cpk does not exceed the Cp at any cost, it can rather only be equivalent. It comes at the time of the actual average process declines in the mid of the limits of the specification.
The formula (Mean-LSL)/ (3 x sigma) or as you can consider the Cpk = either (USL-Mean)/ (3xSigma) by consider the smaller one. With capability of 3 sigma (Cpk=1.0), a process will produce approximately 99.73% good product or .27% bad product. For some people, this represents an extreme high level of poor products. For the higher standard of quality, 4 standard deviations between the mean and the adjoining specifications. The variation was curtailed so that, instead of 3 sigmas between the closest specifications, 4 sigmas were fitted. This may be equal to a Cpk = 1.33. At this phase, the process will manufacture approximately 99.9937% good product or .0063% bad product.
It should be noted that the least importance should be given to Cpk, when it appears that the process is out of control level. Customers demand high-quality Cpk values and also some assurance for the future. From that point of view, Cpk will be reliable or made in a better position over preceding capability studies. Cpk should not be stressed when the process is out of control because of the fundamental statistics that are generally applied in calculating Cpk. Since, Cpk uses a sort, a process seem to appear "superior” because the available range is not a reasonable representation of the process changeability when the process is out of control. However, if the process seems to be in control, then it can be assumed that the range is adequately available for calculating Cpk.