What if Cpk is 0.89
Six Sigma – iSixSigma › Forums › Old Forums › General › What if Cpk is 0.89
 This topic has 7 replies, 6 voices, and was last updated 15 years, 3 months ago by Ruddy.

AuthorPosts

June 2, 2007 at 10:01 am #47145
Mohsin SiddiquiParticipant@MohsinSiddiqui Include @MohsinSiddiqui in your post and this person will
be notified via email.If I calculate CPk for a process and it is found to be 0.89, so what does it really shows?
Can i calculate the rate of rejection from this value? or amount of rejection from this value??
and how can i find the repeatibility of my process??
pls help!!!0June 2, 2007 at 11:28 am #156850
Fake Gary AlertParticipant@FakeGaryAlert Include @FakeGaryAlert in your post and this person will
be notified via email.Your process is not capable and not centered
0June 2, 2007 at 6:27 pm #156860CpK of .89 means that your process is not centered.
Calculate Cp = (USL – LSL)/6*Std.Dev
To Calculate Non Conformint Rate: NCR=2P(Z<2.67) Assuming Cp is .89 and same as your CpK value. To find P(Z<2.67) you must have index table for Z
In this case P(Z<2.67)=.00379256 Thus, for a centered process with a Cp of .89 the nonconforming rate is 7585.12 ppm
Once again this calculation assumes that your Cp and Cpk are the same.
Hope this helps
CT0June 2, 2007 at 6:39 pm #156861
Jim ShelorParticipant@JimShelor Include @JimShelor in your post and this person will
be notified via email.Mohsin,
You cannot tell the reason for the low capability of your process by looking only at Cpk. You need to examine both capability indices (Cp and Cpk) and both performance indices (Pp and Ppk) to determine the problem.
Cp = What is my process capable of doing assuming the process is centered and all process variation is common cause variation?
Cpk = What is my process capable of doing using the actual process mean and assuming all process variation is common cause variation?
Pp = What is my process capable of doing assuming the process is centered and using actual total process variation?
Ppk = What is my process actually doing using the actual process mean and the actual total process variation?
To determine the reason for a low process capability or a low process performance, the following comparisons are used;
If Cp >> Cpk a mean centering problem exists.
If Cp >> Pp a total variation problem exists.
If Cpk > Ppk a total variation problem exists.
If Cp >> Ppk both a total variation and a centering may exist.
If any of these indices are <1, your process is not capable of meeting your customers requirements. Any of the indices being <1 will cause Ppk to be <1.
I hope this helps.
Jim Shelor0June 3, 2007 at 5:43 am #156873
Fake Gary AlertParticipant@FakeGaryAlert Include @FakeGaryAlert in your post and this person will
be notified via email.Do you think that we can calculate Cp,Cpk,Pp and Ppk for the same process data,regardless the special status of each data distribution (subgroups or individual data)??
0June 3, 2007 at 3:14 pm #156877
Jim ShelorParticipant@JimShelor Include @JimShelor in your post and this person will
be notified via email.All 4 indices are calculated using the same set of data. The data can be subgrouped to calculate all four, or it can be individual to calculate all four.
You cannot use some indices calculated from subgrouped data and other indices calculated from individual data for any single analysis of all four indices.
I am not sure what you are driving at with your question.
Jim Shelor0June 3, 2007 at 4:53 pm #156880Why do you think my process is not centered?
0June 4, 2007 at 5:16 am #156884hello moshin,
if your Cpk is at 0.89, then definitely your process is not capable. what you could do is to check again your process based on the following :
1. shift of your distribution – is it shift going to higher side of the specs or lower.
2. check for chronic factors involving your distribution. factors that leads to defects.
3. repeatability is one of the major contributions for non capability of the process. your product may be actually within specs but your equipments of process maybe erratic and sometimes gives misleading results. Use GRR.
4. go back to your 5m’s1e and identify contributions leading to rejections.
hope these would help
regards.0 
AuthorPosts
The forum ‘General’ is closed to new topics and replies.