Subject: | |
From: | |
Reply To: | TechNet E-Mail Forum. |
Date: | Thu, 25 Feb 1999 10:39:02 -0600 |
Content-Type: | multipart/mixed |
Parts/Attachments: |
|
|
Ron,
Usually the formula for calculating Cpk is written something like as follows:
Cpk = min of ((USL - AVG)/3 Std. Dev, (AVG - LSL)/3 Std. Dev.)
Where: USL = upper spec limit
LSL = lower spec limit
AVG = process average
Std. Dev. = standard deviation of the process
When making a calculation, use the smaller number of either USL - AVG, or AVG - LSL and divide by 3 Std. Dev.
Technically, if the process average exceeded the spec limits you would get a negative number since you would be
subtracting a larger number from a smaller number. However, this would represent a situation where most
product built (> 50%) does not meet specficiation, so this should be corrected immediately (if it ever even got
that far).
Jay Soderberg
LEGAL NOTICE: All opinions stated above are mine and mine alone. They do not necessarily reflect the opinion
of ownership or managment of this radio station.
> Subject: Additional Cpk Questions
> Date: Thu, 25 Feb 1999 09:12:02 -0500
> From: "Ronald J. Rhodes" <[log in to unmask]>
>
> Hi All,
>
> I have two questions for those of us that use Capability Potential and
> Capability Performance Indices.
>
> 1) When the mean of the data falls outside the specification limits, is
> Cpk negative or zero?
>
> 2) Does anyone have specific references that delve into the details
> of Cp and Cpk?
>
> Thanks for your help.
>
> Ron Rhodes
> Conductor Analysis Technologies Inc.
> Phone: 732-424-1919
> Fax: 732-424-1886
> Email: [log in to unmask]
> Web: http://biz.swcp.com/cat
|
|
|