Tolerance definition based on expected Cp/cpk

[Mikey324]

Right, exactly what i mean

in that case, you are talking completely theoretical. You would need to control your process to the point it was very centered. I am by far no expert, but i would imagine you could only use 2.5% on either side. All results between -.025 and 0.025?? Thats easy to put on paper, but like Miner said, your process is the reality.

 

[drhouse77]

And i totally agree with Miner! But i Need to put on paper in this thereticall way and i do not know how to do It, can you support me?

 

[Golfman25]

Hello all,
I have a question. We are in technical review prior the offer, our customer is requesting a min Cp/Cpk of 2 on some characteristics.
Just as pure example, assuming that the dimension is 40 +/-0,3, that the Cp/Cpk in serial production has to be >2, how can I calculate in advance which should be UCL/LCL I have to keep in serial life or how should much I should close the tollerance range during setting my process in order to achieve the Cp/Cpk?
Many thanks for your help!!!

I don't know the specifics to see cpk of 2.0. But for 1.67 your can use 60% of your tolerance and 1.33 you can use 75% of your tolerance. My guess is that at 2.0 you're probably cutting your tolerance by more than half. It will depend on your standard deviation, but we have found these percentages to be very workable for us. Good luck.

 

[Bev D]

If I understand your question correctly, then your process must be centered and the 6SD width must be HALF your tolerance range to have a Cpk of 2. In addition to what Golfman said remember that a centered process with a 6SD spread that is exactly the width of the tolerance range has a Cpk of 1. It’s just a simple ratio...

 

[Niber]

Divide by two. For Cpk 2 you need 12s within the tolerances and for cpk1 only 6s. Od your mean tends to vary, take accordingly less, ie one third. This gives you +- 0.1