From: chris
Date: 22 Jul 2003
Time: 13:02:31
Thank you for your current help.
I will try to explain the problem further.
There are millions of train wheels in use and I am to design a type of sampling scheme or plan to test the wheels. The tests are done ultrasonically with EMAT or PET and require time and are expensive to have done.
These millions of wheels have been split up into smaller lots based on similar attributes. The goal is to come up with a reliable testing scheme. The problem that I have run into is that this is slightly different from most sampling plans in that a lot will not be rejected and there is no producer's risk or AQL because the wheels are already in service. It has not yet been determined how the wheels are distributed whether hypergeometrically or binomial or possibly another. I also have to figure out the error associated with the tests and don't know what equations would be applicable in this situation. Sorry for the tardiness of my response and I look forward to hearing back from you. Thanks
Chris