I believe you mean tolerances that are "not bilateral", not binomial.
For unilateral tolerances, Cp and Cpk are the same value. They measure the 3 sigma distance from the mean to the applicable limit (LSL or USL). If you only have an upper limit then:
Cp = Cpk = (USL - Xbar)/3 sigma
For something with only a lower limit, substitute (Xbar - LSL) in the numerator.
It's open to arguement whether or not it's appropriate on these types of features, since they are generally not normally distributed. (Runout should tend more to the exponential due to the nature boundary at 0). You could consider some sort of data transformation to make it more normal.
Community metadata Could not be loaded. No pre-configured community forum parent found for the incoming fid: . You need to add this forum into custom_code/community.sql
Send to friend
Comments
qdigest 8/26/2003
I believe you mean tolerances that are "not bilateral", not binomial.
For unilateral tolerances, Cp and Cpk are the same value. They measure the 3 sigma distance from the mean to the applicable limit (LSL or USL). If you only have an upper limit then:
Cp = Cpk = (USL - Xbar)/3 sigma
For something with only a lower limit, substitute (Xbar - LSL) in the numerator.
It's open to arguement whether or not it's appropriate on these types of features, since they are generally not normally distributed. (Runout should tend more to the exponential due to the nature boundary at 0). You could consider some sort of data transformation to make it more normal.
Good Luck.