Dear Folks
I am a project manager in the automotive parts company in Thailand. Our company got ISO9000 last year and now going for QS9000.
I have learn a lot about SPC in the past cople year. However, most of the materials are only mention about theory of SPC. None of them explain how to start. This made me mad. Since, I have tried to ask the following question to our consultant but nothing clear for me. Two days ago, I found this web page and hope that someone out there can give me a light.
My question is quite simple and I would like to ask by uning the following model.
Model: Suppose that I have one new machine which operate only one process , let say drilling dia 10 +/- 0.1 mm and roughness is 6.3Rz . Our customer request us to use SPC for boths value where the requirements are
1) For PPAP, Ppk must greater than 1.67
2) For ongoing SPC Cpk must greater than 1.33
Question: 1)Why we use Ppk at the PPAP stage? (Ppk is for long term study but doe not require for stability???)
2)Do we need to use X bar R chart during our run for PPAP?
3)If yes in 2, How can we set a UCL and LCL? (If the answer is, we can have it after we finish sampling, then calculate, so what about during the sampling, how can we sure that our process is in control, and if the number of require subgroup is large, we don't to end up do it all over again after finish sampling????)
4) For one side spec (roughness in this case),if the requirement is Ppk greater than 1.67. It means that the average value will near 3.2Rz which also means that we have to design our process not for 6.3Rz but for 3.2Rz Otherwise, we will not earn those. For this case, it will cost a lot more, so can e use Pp for this.
5) When will we apply Cpk instead of Ppk (when it stable??)
6) UCL, and LCL of X bar and R chart of Cp, Cpk come from???
7) Since Cpk is a study of short term, so how can we use Cp for ongoing process??

That would be all of my questions. Sorry for a long question but really need answer.
Regards,
Narakorn Ratchapolsitte
PS: Sorry for my English, it does not be our first langauge here in Thailand

------------------

Narakorn,

Question 1: Why we use Ppk at the PPAP stage?

Ppk is the term used for the 'preliminary' process capability. You use Ppk when you are determining process capability prior to production.

Question 2: Do we need to use X bar R chart during our run for PPAP?

Need, no, but you must be sure your process is in a state of statistical control prior to accepting Ppk as a satisfactory measure of process capability.

Question 3: …how can we sure that our process is in control…

You monitor the process for a period of time (typically 30 days) and evaluate its state of control. If in control, perform your capability study. If not, take corrective action on the process and repeat (monitor the process for a period of time, etc).

Question 4: It means that the average value will near 3.2Rz which also means that we have to design our process not for 6.3Rz but for 3.2Rz.

This depends on variation. Since variation is not provided, I offer this:

USL = 6.3
Mean = 3.2, then standard deviation must be less than 0.619 to achieve Ppk of greater than 1.67. As standard deviation decreases, your mean can increase.

Question 5: When will we apply Cpk instead of Ppk?

When you have satisfied the requirements of #3 response above.

Question 6: UCL, and LCL of X bar and R chart of Cp, Cpk come from?

I am not sure of your intent of this question. Are you asking how to calculate Cp and Cpk from an Xbar-R chart?

Question 7: Since Cpk is a study of short term, so how can we use Cp for ongoing process?

Both Cp and Cpk are metrics of on-going processes.

For a more detailed discussion of process capability, you may want to see:

http://Elsmar.com/ubb/Forum10/HTML/000015.html
http://Elsmar.com/pdf_files/cpk.pdf

Hope the above helps.

Regards,
Don

Dear Don
Thank you for your answer. I think that I am now getting some light. BTW, can you recommend any material that I can use for implementing SPC in my company.
Regards,
Narakorn

Narakorn,

Glad it helped.

Hum, reading material. I have found that most reading material available usually falls into two categories. Theory oriented and application oriented. From your request, let’s try the latter:

SPC Essentials and Productivity Improvement; A Manufacturing Approach; William A. Levinson and Frank Tumbelty; ASQ Quality Press; 1997

Handbook of Statistical Methods in Manufacturing; Richard Barrett Clements; Prentice Hall; 1991

Also, Grant and Leavenworth’s Statistical Quality Control is very good, but I do not have the publisher info. Try a search at ASQ. I particularly like this one.

You may also want to try www.altavista.com (http://www.altavista.com) and in the search box type {“statistical process control””manufacturing”} with the text in the brackets exactly as typed above. That will give about 100,000 hits. When you are here, visit as many as you can and collect as much as you can.

Go to www.amazon.com (http://www.amazon.com) and try a search.

The Quality Council of Indiana publishes a series of primers intended mainly for preparing for the ASQ’s certification exams. While not books, they are good educational and learning tools.

I personally have many publications from private sources that are very good, but they are no longer in print and not available on the web (yet). I also have a couple of PowerPoint presentations (old) that were developed for training years ago. I do not attest to their accuracy or content, but I can forward if you like. I would need a while to remove dated information.

Well, that should keep you busy for a while. Enjoy.

Regards,
Don

narakorn,

Forgot two:

Process Quality Control by Ott and Schilling and Sun-Tzu's The Art of War.

recommend any material that I can use for implementing SPC in my company.

Definitely, Sun-Tzu.

Regards,
Don

Dear Don
I now got a Statistical Qaulity Control by Grant and .... Very interesting and this will keep me busy for a while. It really good book. Thank you, and if I have any question, you will be the one that I will contact, that's for sure.
Regards,
Narakorn R

narakorn,

Glad I could help. Have fun.

Regards,
Don

Thanks for the great help, Don!

Glad I could help. Was I accurate?

Regards,
Don

For those of you that happened onto the archives, you may want to consider newer information on SPC if you are dealing with precision machining. New charting methodology incorporates the non-normal continuous uniform distribution found from tool wear. Check out the following link for preparing the X hi/lo-R chart, preparing control limits and calculating capability for precision machining at: Statistical process control for precision machining (http://elsmar.com/Forums/blog.php?b=79) Stop the Xbar-R madness!

For those just coming across this issue in the archives, I am updating the responses to current requirements:

Since, I have tried to ask the following question to our consultant but nothing clear for me.

Go figure...:rolleyes:

My question is quite simple and I would like to ask by using the following model.
Model: Suppose that I have one new machine which operate only one process , let say drilling dia 10 +/- 0.1 mm and roughness is 6.3Rz . Our customer request us to use SPC for both value where the requirements are

1) For PPAP, Ppk must greater than 1.67
2) For ongoing SPC Cpk must greater than 1.33

Question: 1)Why we use Ppk at the PPAP stage? (Ppk is for long term study but doe not require for stability???)[/


It is probably not applicable at all. Drilling is not a normal distributions, and surface finish is not a bilateral specification. See below:

AIAG PPAP 4th Edition
2.2.11.5 Processes with One-Sided Specifications or Non-Normal Distributions

NOTE: The above mentioned acceptance criteria (2.2.11.3) assume normality and a two-sided specification (target in the center).

When this is not true, using this analysis may result in unreliable information.

NOTE (cont.): These alternate acceptance criteria could require a different type of index or some method of transformation of the data. The focus should be on understanding reasons for the non-normality (e.g. is it stable over time?) and managing variation.

2) For ongoing SPC Cpk must greater than 1.33
Question: 2)Do we need to use X bar R chart during our run for PPAP?

For drilling, use a X hi/lo-R chart. Xbar -R chart is the worst chart you can use. See Statistical process control for precision machining (http://elsmar.com/Forums/blog.php?b=79). You might be able to use an X-bar-R chart for surface finish if it is bad enough to exhibit a normal distribution.

Question 3)If yes in 2, How can we set a UCL and LCL?

You need to do a capability study to determine the true control limits for the surface finish.

Question 4) For one side spec (roughness in this case),if the requirement is Ppk greater than 1.67. It means that the average value will near 3.2Rz which also means that we have to design our process not for 6.3Rz but for 3.2Rz Otherwise, we will not earn those. For this case, it will cost a lot more, so can we use Pp for this.

See Question 1.

Question 5) When will we apply Cpk instead of Ppk (when it stable??)

See Question 1.

Question 6) UCL, and LCL of X bar and R chart of Cp, Cpk come from???



They originate from the capability study.

Question 7) Since Cpk is a study of short term, so how can we use Cp for ongoing process??


See Question 1.