# Process limits calculation

Six Sigma – iSixSigma › Forums › Old Forums › General › Process limits calculation

- This topic has 19 replies, 9 voices, and was last updated 11 years, 3 months ago by Brian M.

- AuthorPosts
- October 24, 2008 at 6:43 am #51187
Our customer is giving us the demand:

Cpk = 2

Lower tolerance limit = 105

Upper tolerance limit = 155How can we define the process limits which we should respect to achieve Cpk 2? We have no parts that can be measured yet. But we need to know the process limits before we will start to produce.

Thank you very much for your answer.0October 24, 2008 at 3:53 pm #177057Using a basic Cpk calculation and assuming that you have a process that’s “in control” and have the ability to adjust the process to run at the center of the spec limits, the aim should be at 130 and the standard deviation must be 4.15 or less.

0October 25, 2008 at 9:16 am #177075

Michael MeadParticipant@Michael-Mead**Include @Michael-Mead in your post and this person will**

be notified via email.Yeah, you need to know your process capability. I

think what you should worry about are control

limits. Is that what you mean by “process limits”? I mean what are you going to do–run all the time

and sort the products that are above 145 and

through them out? KC is correct. aim for the

middle and maintain control. That is all there is

to it.0October 26, 2008 at 9:31 am #177088Thank you both KC and Michael.

The process will be as you mentioned. There is a 100% control station measuring the value. I want to hold all the parts which would be out of control limits in order to respect Cpk 2. If I consider to adjust the process on the center line 130 and sigma 4,15, than I have to define the control limits at 117,55 and 142,45? Do you maybe know some simple software of excel file than can be used for such purposes?

I would expect that the control limits would be much to narrow than the above calculated to achieve Cpk 2. Why is it so?

Thank you all very much again.0October 26, 2008 at 12:37 pm #177090

Michael MeadParticipant@Michael-Mead**Include @Michael-Mead in your post and this person will**

be notified via email.Hello Coko,I am not sure you understand the concept of control

limits. They are for averages. They are not

directly related to individual parts. a CpK of 2

does not mean that all parts will be within the

control limits. I wish I could help you more, but

this is a basic concept of statistical process

control.In which country do you reside?0October 27, 2008 at 7:08 pm #177107

yesitismeMember@yesitisme**Include @yesitisme in your post and this person will**

be notified via email.KC is right, your avegarge will be 130 and the control limits should be:

UCL= 130+3*4.15

LCL= 130-3*4.150October 28, 2008 at 10:58 am #177120A Cp of 2 implies that your process variation is twice as small as your specification limits.

You should be able to calculate your current process Cp and Cpk to see how much variation exists within your process that you intent to run the subject product on.

If you have very small variaiton you should be okay. Since you did not specify the type of process you will be running it is difficult to give you a more specific answer.

0October 28, 2008 at 6:18 pm #177137

Brian MParticipant@Brian-M**Include @Brian-M in your post and this person will**

be notified via email.Your control limits should be narrower.Figuring this backwards from the end (Cpk) to the beginning (X Bar R chart) you can use your tolerance and desired Cpk to set up a chart to get you going, with the understanding that you will need to recalculate everything once the real data begins to come in.All of these numbers are based on a sample size of 3Cpk = Tol/(6*sigma^) Sigma^ is the estimated sigma based on the average range from your control chart (sigma^ = Rbar/D2). So by doing a little juggling you can figure sigma^ (4.16) and average range (7.05).Once you know those values you can calculate your control limits for the Xbar chart. Xbar UCL = Xdouble bar + Rbar*A2 (137.2). Xbar LCL = Xdouble bar – Rbar*A2 (122.8)Your control limits for the range chart will be: UCL = Rbar*D4 (18.1) and LCL = Rbar*D3 (0).Once again, this s all just to have some place to start to see if you are in the ballpark. Once you’ve collected enough data, recalculate everything.I think thats all correct.Have fun. Brian M

0October 28, 2008 at 6:56 pm #177140Wy do you assume a sample size of 3?

0October 28, 2008 at 6:59 pm #177141

Brian MParticipant@Brian-M**Include @Brian-M in your post and this person will**

be notified via email.Just for the sake of having something to calculate… which is why I said I was using that sample size. If he is using a different sample size, he’ll need to look up the factors himself ;)

0October 29, 2008 at 8:48 pm #177195

Brian MParticipant@Brian-M**Include @Brian-M in your post and this person will**

be notified via email.Correction… Cp = Tol/(6*sigma^) NOT Cpk = Tol/(6*sigma^)

0October 29, 2008 at 9:27 pm #177198

Mike CarnellParticipant@Mike-Carnell**Include @Mike-Carnell in your post and this person will**

be notified via email.Brian M,Is this Brian from London?

0October 30, 2008 at 9:39 am #177220

Chris SeiderParticipant@cseider**Include @cseider in your post and this person will**

be notified via email.Michael,

Why say control charts are for averages only? What about I-MR charts?……they are for individuals! Maybe Coko has no rational subgroups.

0October 30, 2008 at 9:50 am #177221

Michael MeadParticipant@Michael-Mead**Include @Michael-Mead in your post and this person will**

be notified via email.Hello Ron, Here the is way I see it. (Not always right but

seldom completely wrong.)Well, he has process capability to base his control

chart on. Rational subgrouping is not a

requirement. We are looking at variation within the

subgroup and the variation between subgroups.Individuals charts are ok….but, you can’t tell

the difference between the “between” and “within”

variation. They are not sensitive to process

changes. I don’t recommend them for this reason.

There is nothing wrong with using them as an

inspection record, but not for process control.0October 30, 2008 at 10:40 am #177223

Chris SeiderParticipant@cseider**Include @cseider in your post and this person will**

be notified via email.Michael,

Agreed…..however, there are more than a few cases where individual are the most appropriate thing to chart with SPC….

I would hesitate to tell someone to “create subgroups” if no rational basis existed for the subgroup. If no rational subgroups exist, I would hesitate to look for within variation. If no rational subgroup existed, then what would be the response plan if within variation went out of control?0October 30, 2008 at 10:44 am #177225

Michael MeadParticipant@Michael-Mead**Include @Michael-Mead in your post and this person will**

be notified via email.My response would be to immediately look for loose

fixturing or dull tools, anything that would vary

between one part and the next.0October 30, 2008 at 12:32 pm #177233

Brian MParticipant@Brian-M**Include @Brian-M in your post and this person will**

be notified via email.Nope, New Jersey

0October 30, 2008 at 3:41 pm #177243

Brian MParticipant@Brian-M**Include @Brian-M in your post and this person will**

be notified via email.Wrong

0October 30, 2008 at 5:17 pm #177248Hey, I think this is infringement.

0October 30, 2008 at 5:33 pm #177250

Brian MParticipant@Brian-M**Include @Brian-M in your post and this person will**

be notified via email.I never noticed a TM in your posts ;) lol

0 - AuthorPosts

The forum ‘General’ is closed to new topics and replies.