Remember my last post.?

Well if you did not read it, here is a summary:

Apply risk analysis before deciding Cpk requirements, 1.33 is not a miracle cure or a universal constant.

If you have an 8-cavity tool with one cavity showing Cpk of only 0.8 against specifications of 50±0.05 mm, then the customer will find 1 part in 1000 to be up to 8 microns too large – which is the size of a red blood cell.

No todays post: What happens if these parts need to be assembled with another part of just a poor quality?

First of all my guess is; nothing happens with a misfit of 0.008 mm but of course it could be a problem (leakage for instance).

Part A, 1 in 1000 parts are too large and this one part will not fit part B unless part B is in the large end of specifications as well…. Give it some thought, and you will agree with me. A large axle will still fit a large bushel and vice versa.

So, if part B also have a Cpk of merely 0.8 in one of the eight cavities we now have 1000 poor B parts and 1000 poor A parts in the assembly line’s 1.000.000 parts feeding box.

Odds that a large and a small part will meet and cause a failure is quickly calculated to be
1 in a million.

One in a million final (potential) defects with a Cpk of 0.8. Think of that :^)

Go back to the risk analysis and consider how many serious injuries or other kinds of harm this will bring to the patient using your device. What is the likelihood that the hazardous situation will even arise?

So once again, do you really need a Cpk of 1.33 in each part or is the real concern about the final assembled device?

You think about it, and in the meantime, have a look at my other interesting posts about process validation, sample size determination and risk analysis here.

For more information please contact EpsilonPlus here.