From: Stan Hilliard
Date: 02 Apr 2000
Time: 00:16:37
Greetings Todd,
Neither of those criteria (AQL and C=0) are adequate for designing acceptance sampling plans. A customer should not be concerned about AQL. this is explained at:
www.samplingplans.com/aqlprimer.htm
The problem of using C=0 as a design criterion for acceptance sampling plans is discussed in many messages on this forum. Click on "search for words" at the upper right and search for the word zero.
To answer your questions specifically, the consumers risk for C=0 plans is large and usually unacceptable - once it is known. The program TP105 is very good for demonstrating this by calculating, for any n and C, the producers and consumers risks, and it will show the whole oc curve.
You can specify the risks that you want and the software will calculate the right sample size and acceptance number. See:
www.samplingplans.com/software_oc.htm
If you do not know the oc curves of your sampling plans, you cannot match the plan to the quality requirements of the producer and consumer. There is no other way.
Sincerely, Stan Hilliard