If it is reliability you are interested you should also look at chapter 8 of the (excellent) NIST book
It was a smart move to look at the Cpk values – not many people would do that.
The coupons do not show the same variation as the actual parts. You need to find out why – talk to the people who know the process.
When you find out why, you may be able to improve the Cpk of the actual parts.
I think you have proved that the ‘coupons’ are not a valid substitute for testing actual parts, because they give different results.
You think the process for producing them is the same but some difference exists, which you need to identify.
You should do some research first. Use the search function in the top right corner of this screen.
Start with the item
Process Capability (Cp, Cpk) and Process Performance (Pp, Ppk) – What’s the Difference?
Does the concept of normality apply to this at all?
There are a finite number of possible outcomes, some two dice totals are more frequently represented in the possible outcomes.
Over time the observed results will match the relative…[Read more]