Purpose: The problem for the traditional control chart is that it is unable to monitor the very small fraction of nonconforming and the underlying distribution is the normal distribution.
control chart is useful where it controls the vert small fraction on nonconforming. In this study, we will design the
control chart in order to use under non-normal process. Methods:
is calculated not by failure rate based on attribute data but using variable data. Control limit for non-normal
control chart is designed based on
-risk calculated by cumulative distribution function of Burr distribution.
-risk, which is for performance evaluation, obtains in the Burr distribution's cumulative distribution function and control limit. Results: The control limit for non-normal
control chart is designed based on Burr distribution. The sensitivity can be checked through ARL table and OC curve. Conclusion: Non-normal
control chart is able to control not only the very small fraction of nonconforming, but it is also useful when
distribution is non-normal distribution.