High Cpk's and Control Charts - Is it justified to stop using control charts?

[MSAFAI]

Hi everybody,

In processes, where we have high Cpk's (like 2 or more), is it justified to stop using control charts (like X-bar and R) and instead, calculate Cpk every month or so ?


Thanks on beforehand.

[This message has been edited by MSAFAI (edited 04 August 2001).]

[Rick Goodson]

MSAFAI,

It depends on why you are running control charts to begin with. If customer mandated you need to discuss it with the customer. Even if you decide to discontinue them I would not leave the process run without some checks. You might consider pre-control charts as an alternative.

Rick

[AJLenarz]

MSAFAI,

I agree with Rick. It would not be wise to totally abandon process monitoring. If you have a high CPK, you may want to consider reducing the frequency of inspection and charting (with the consent and acknowledgement of the customer of course).

Also, the comment was made to stop control charts and calculate CPKs every month or so. The SPC manual’s definition of CPK is: “The capability index for a stable process…”. If you discontinue the use of the SPC chart then you will not know if your process is stable and capable. If you don’t know if your process is capable, can you calculate and report a CPK value?

[MSAFAI]

Dear Rick

Thanks for your input.
1- In this specific case customer has not mandated SPC. Actually, in my country OEM's are just starting to define special characteristics and SPC for them.

2- I didn't mean leave the process without any control. I meant just not to use control CHARTS. Normal periodic checks, of course, remain.

Thanks again
MSAFAI

[MSAFAI]

Dear AJLenarz,

Thanks for your comment.

About calculating Cpk on stable processes, I agree with you.

But for calculating Cpk we usually use control chart (at least for 20-25 subgroup) to establish control, then we proceed to Cpk calculation.

Thanks again,
MSAFAI

[This message has been edited by MSAFAI (edited 06 August 2001).]

Too all,

Would and X-Bar MR chart be applicable in a case when stability and capability have been proven over time and special causes removed from the process?

Also, if applicable, could the variable control chart be replaced by an attribute chart under the same circimstances?

If so or not, why?

ASD...

[MSAFAI]

Guys,

by accident, I found an interesting forum discussion from 1999. Guess where?
http://Elsmar.com/ubb/Forum10/HTML/000058.htm

The same subject is discussed, with good inputs from Don Winton.

Regards
MSAFAI

[MSAFAI]

Originally posted by Al Dyer:
Too all,

Would and X-Bar MR chart be applicable in a case when stability and capability have been proven over time and special causes removed from the process?

Also, if applicable, could the variable control chart be replaced by an attribute chart under the same circimstances?

If so or not, why?

ASD...

Dear Al,

1- As far as I know, X-MR is used in special cases where taking more than one sample each time is not practical (like in a very slow process).

If you mean X-MR can reduce sampling, you can achive the same by less frequent sampling.

I'm not an expert in this, but I think X-MR has a disadvantage that it is sensitive to non-normality of the underlying distribution (central limit theorem does not hold anymore)

2- About using an attribute chart, I think you would get a flat line (zero defect) almost all the time, even if the process goes out of control (because of high Cp/Cpk)

I hope it makes sense

Regards
MSAFAI