The Elsmar Cove Wiki More Free Files The Elsmar Cove Forums Discussion Thread Index Post Attachments Listing Failure Modes Services and Solutions to Problems Elsmar cove Forums Main Page Elsmar Cove Home Page
Google
  Web Elsmar.com
*Please be aware that SOME RECENT forum threads may not yet be indexed by Google.

View Full Version : How to measure Process Capability ( Cp and Cpk)


astudent
25th October 2004, 08:05 PM
How do you measure process capability? By estimating (somehow) percentage non conforming or by data from control charts?

If you have more products with different capability what is the "overall company cabability"?

thank you

Jennifer Kirley
25th October 2004, 10:24 PM
How do you measure process capability? By estimating (somehow) percentage non conforming or by data from control charts?

If you have more products with different capability what is the "overall company cabability"?

thank you

Here are a few progressively detailed sites I found that should help explain this better than I could.

Measuring process capability

http://www.itl.nist.gov/div898/handbook/ppc/section4/ppc46.htm

http://www.npd-solutions.com/proccap.html

http://www.statsoft.com/textbook/stprocan.html

Here is what I found for "Overall Process Capability." It deals with process capability within a larger quality management structure.

http://www.statsoft.com/textbook/stprocan.html

I hope this helps!

cyberjyothi
26th October 2004, 06:01 AM
hi all,

how to calculate(Conduct) the process capability study for the attribute control charts (especially for p-charts).

thanks,
Jyothiswar

Darius
26th October 2004, 10:48 AM
How do you measure process capability? By estimating (somehow) percentage non conforming or by data from control charts?

IMHO data from control charts, Donald Wheeler said that you can't traduce (without some assumptions that almost never are accomplished) percentage of non conforming to capability indicators (So he recomended to take in account both indicators apart and calculate the % of non conforming from the raw data).

I am agree that there are everywhere on the net tables for translation between them even here on the board found a wonderfull link to the calculus, but the normality assumption strikes my head. One tink that you should consider is that if you calculate an indicator is just AN ESTIMATE, wich depends on the amount of data, the representability of the data for such process (rational sampling) and the specs vality.
:bonk: