|
*Please be aware that SOME RECENT forum threads may not yet be indexed by Google. |
|
View Full Version : How do I draw OC curve? How do I plot this curve in Excel? vinay_bull 6th August 2006, 03:54 PM Lot size = 1500 Sample size = 150 Acceptance no = 3 How Do i plot this curve in Excel? Also my AQL is 0.05% and LTPD is 6%, how do I describe the protection offered? Wesley Richardson 6th August 2006, 04:09 PM Lot size = 1500 Sample size = 150 Acceptance no = 3 How Do i plot this curve in Excel? Also my AQL is 0.05% and LTPD is 6%, how do I describe the protection offered? Hi Vinay, Use the equation P(r) = ((e^(-np)(np)^r)/(r!) for your example, n = 150 r = 3 p = lot fraction defective P(r) = Probability of exactly r defectives in sample of n. Do this calculation for various values of p, for example over the range from p = 0.005 in 0.005 increments up to p = 0.10. Using the above equation, or built in function in Excel, this can be done quite easily. Plot the value of P(r) as a function of p. This is the operating characteristics curve. Wes R. vinay_bull 6th August 2006, 05:29 PM Hi, Thanks for the reply. Any idea about the second question Also my AQL is 0.05% and LTPD is 6%, how do I describe the protection offered? Wesley Richardson 6th August 2006, 07:24 PM Hi, Thanks for the reply. Any idea about the second question Also my AQL is 0.05% and LTPD is 6%, how do I describe the protection offered? AQL is normally associated with a sampling plan such as ANSI/ASQ Z1.4 for attributes or ANSI/ASQ Z1.9 for variables. The AQL is used to define for a given plan, the accept and reject numbers. Since you have defined your own plan, the AQL is not applicable. You can determine the AOQ and AOQL from the method I have given above. Calculate AOQ = (p)(P(r)) for each of the values you found above. AOQL is where AOQ is at the maximum. One correction for my previous equations, since r = 3, you will also need to calculated P(r=0), P(r=1) and P(r=2), as well as P(r=3) for each p. Add the P values together for a given p to get the total P_a (probability of acceptance) under the OC Curve. LTPD is lot tolerance percent defective. It is normally associated with Dodge-Romig sampling tables. Common values are LTPD = 5% or 10%. You pick a value, e.g. LTPD = 5%, and then read off the OC Curve the probability of acceptance at that value, P_a. Wes R. |
|