Unregistered (imported) wrote: Fri Mar 20, 2009 10:24 am You randomly choose several parts (or stories in the case of the archive) then you inspect them for quality. The size of the batch and the tolerance for error determines the size of the sample. For example a batch of 10,000 would require a sample size of 100 or 1% if 99%+ is the desired quality tolerance (it's actually a bit more complicated than that but for this discussion I think it will suffice). If all the samples pass the quality inspection then the batch is passed and certified for shipment. If any of the samples proves defective the procedure is to take a larger sample of the batch (usually 10%) to get a more precise picture of how many are bad parts. From the second inspection we determine whether it is worthwhile to glean out any good parts or if we should just simply scrap the whole batch and start over. If 50% or more of the original sampling fails inspection we don't bother with a follow up inspection we just scrap the whole batch and remake the parts. We use a similar method to determine if a vender sent us good parts or not.
Hi Unregistered,
Your account of QC sampling was fascinating. Purely as a matter of interest how come some parts fail and some don't? A batch of 10,000 would presumably be made on automatic machines which would do exactly the same thing every time.
For machined components would it be due to a slack tool-tip checking regime (Which is all I can think of) or what?
Regarding rewriting someone's story for them. Yes, rewriting may be a bit much but I once edited a college essay for my daughter to show her how it could be improved and once she saw the improvement she kind of got the idea and her subsequent writing was much better. Writing skills can definitely be taught (or perhaps learnt?).
Kortpeel