industry standards for Cp and Cpk…
Six Sigma – iSixSigma › Forums › Old Forums › General › industry standards for Cp and Cpk…
- This topic has 18 replies, 10 voices, and was last updated 13 years, 7 months ago by
Anonymous.
-
AuthorPosts
-
November 13, 2008 at 7:37 pm #51322
PRPatelParticipant@PRPatelInclude @PRPatel in your post and this person will
be notified via email.Hello,
I’m looking to see if there are industry stanrards for Cp. I know that 1.33 is used for Cpk, but I have not found anything for Cp.
Any thoughts?
-Pritesh.0November 13, 2008 at 9:14 pm #177646
Brian MParticipant@Brian-MInclude @Brian-M in your post and this person will
be notified via email.Cp tells you if your process is capable, Cpk tell you how centered you are. Most requirements are for Cpk as it is an indication of how much good product you are producing. A great Cp (capable process) is worthless if you are producing outside of the spec limits. A perfectly centered process will have Cp=Cpk.
0November 14, 2008 at 12:57 am #177650
marco duarteParticipant@marco-duarteInclude @marco-duarte in your post and this person will
be notified via email.I think you could use 1 for CP. This number is acceptable.
Regards!!0November 14, 2008 at 1:56 am #177651Wrong, 1 is not acceptable.
0November 14, 2008 at 3:36 pm #177663
Mike CarnellParticipant@Mike-CarnellInclude @Mike-Carnell in your post and this person will
be notified via email.Stan,
Are you sure? The original post was with a Cpk of 1.33 and this guy is ok with a Cp of 1.0. I would really like to see the data for that process.
Where does this stuff come from?
Regards0November 14, 2008 at 4:30 pm #177667In the first place its hard to get Cp<Cpk.
Jan0November 14, 2008 at 4:39 pm #177668
Mike CarnellParticipant@Mike-CarnellInclude @Mike-Carnell in your post and this person will
be notified via email.Jan,
Really.
Thank you.0November 14, 2008 at 7:23 pm #177683Many customers are requiring a Cp and Cpk of 1.67 short term and extended requirement of 2.0
The idea is to have both measurements equal.0November 14, 2008 at 7:59 pm #177686Are you saying customers expect higher long term capability than short term?
0November 14, 2008 at 9:07 pm #177691That would be a pretty dumb thing to say – don’t you think?
0November 15, 2008 at 6:11 pm #177710
Mario Perez-WilsonParticipant@Mario-Perez-WilsonInclude @Mario-Perez-Wilson in your post and this person will
be notified via email.Ron, your statement The idea is to have both measurements [Cp=Cpk] equal is well taken, however, it is a good general statement, particularly, when you are focusing only on a single domain, such as quality by reducing defects.
It may NOT be a good generalization when we are characterizing (concurrently) in more that just one domain. Quality is one domain, but there are other domains, such as speed, cost, etc.
Lets state some underlying assumptions, so the readers dont derail.A1 – The Cp can never be smaller than the Cpk.
A2 The Cp and Cpk cannot be negative.
A3 Knowledge exists on how to control the average and the variance.
A4 Dealing with variable data under the Ratio scale of measurement.
A5 Maximizing a domain does not imply maximizing its value.
A6 A domain may have multiple responses.
In the quality-domain, having the Cp equal to the Cpk, implies that the average is centered in the middle of the specification limits, and this in turn minimizes the probability to produce outside the specification limits (below the lower spec. limit and above the upper spec. limit). However, when now, we introduce a second domain, the general idea of having the Cp=Cpk, may not be the way of maximizing for both domains. When dealing with two domains, quality and cost, a better choice may be to maximize the Cp and not maximize the Cpk. Maximizing Cp may work best for quality and cost. However, maximizing the Cpk may be best for quality but not cost. As a matter of fact, maximizing the Cpk may be detrimental in the cost-domain. A good example is Gold Plating, particularly at a spot price of 750 USD per oz. In a gold plating process, I may want to maximize the Cp, but I may want to have a Cpk that meets the standard but with an average as close to the lower specification limit as I can get. This in turn guarantees that I am maximizing under the quality-domain (less probability of producing beyond the specification limits) and maximizing under the cost-domain (by providing the minimum gold or less cost, thus more earnings), and still meeting customer satisfaction.
When we characterize processes under the MPCpS Methodology we are working on maximizing five domains.
So, the statement of Cp=Cpk is good, but it may depend on the process and under how many domains I may be characterizing that process.0November 15, 2008 at 8:27 pm #177711A2 – Wrong – Cpk can very well be negative.
Jan0November 16, 2008 at 2:46 pm #177712
Gary ConeParticipant@garyaconeInclude @garyacone in your post and this person will
be notified via email.Jan,Take the time to think about what Mario is saying. This is an
ASSUMPTION. He is saying we will not choose to run a process outside
the spec.Mario is sharing something well beyond what is normally seen or for
that matter what is normally taught. Take the time to understand.0November 16, 2008 at 9:14 pm #177718Gary,Did a couple of your posts disappear?Cheers,
Andy0November 16, 2008 at 11:51 pm #177719
Gary ConeParticipant@garyaconeInclude @garyacone in your post and this person will
be notified via email.Andy,Yes.I am too political these days and criticized a big advertiser.No problem, saw your message on Lasik and am doing some
digging as I am considering.If you have time go to http://blog.gpsqtc.com and comment again.
About 100 hits per day there, so a possibility for an unedited
discussion.I value you opinion.Weigh in on Mario’s comments as well. Also an ex-Motorolan and a
good guy.0November 17, 2008 at 7:35 am #177722Gary,If someone took your post down because you criticized an advertiser that is a great shame because I think rational argument and the ability to reason is a very important part of six sigma implementation.As there is an eye in isixsigma, I hope the powers that be won’t mind me providing you with this link as your blog has a password:http://www.usaeyes.org/lasik/library/fda-lasik-quality-study.htmYes, I’ll read Mario’s post over the next few days.Do you know Dr. Pyle? I’m meeting him and his family in London next week.Cheers,
Andy0November 17, 2008 at 7:54 am #177723Mario,I’m not sure about your use of the term ‘domain’ and your claim Cpk can’t ever be negative, but your other comment is an important contribution:Another example of your multivariate approach not only applies to different variates, but also the same variate, such as a flow controller, where parts identically machined can cause defects in the centre of the specification due to stiction.Cheers,
Andy0November 17, 2008 at 1:11 pm #177729
Gary ConeParticipant@garyaconeInclude @garyacone in your post and this person will
be notified via email.I don’t know Dr. Pyle – who is he?
0November 17, 2008 at 1:23 pm #177730Hi Gary,Dr. Pyle was formerly director of IP (Motorola SPS) in Austin. He is now an IP attorney and expert witness.Cheers,
Andy0 -
AuthorPosts
The forum ‘General’ is closed to new topics and replies.