iSixSigma

MSA %P/T

Viewing 7 posts - 1 through 7 (of 7 total)
  • Author
    Posts
  • #37527

    TP
    Member

    When we do MSA’s at my company, we look at the %P/T –  % Process to Tolerance.  I am having a hard time understanding what that is.  To the best of my knowledge, it is just comparing the process tolerance to the spec limit, and whether or not you can really tell whether or not you have a good part or a bad part.  We work in  a low volume, high mix environment and it is difficult for us to get a sample of parts that truly represent the whole process.  So I think that this is just a sanity check to take into account that we do not get the whole deviation in parts. 
    I was told that many companies do not use this term.  Just looking for some feedback as to what it means exactly.

    0
    #110689

    Mikel
    Member

    You have two different questions which should be answered with MSA. The first is whether a measurment is adequate for acceptance judgements. P/T does that. The second is whether the measurement is adequate for studying the variation in a process. This is P/TV.
    Which is most important? If you are making acceptance decisions with the measurement, it’s P/T.

    0
    #110690

    Dog Sxxt
    Participant

    You are not doing MSA, you are doing garbage because you do not understand what you are doing. Get a good MSA book and learn its basic first.

    0
    #110786

    TP
    Member

    Oh ok Dog Sxxt
    yeah… thanks for the advise, but if you find an MSA book with %p/t let me know.  I haven’t been able to find that in any book. 

    0
    #110788

    Dog Sxxt
    Participant

    1. AIAG’s MSA Guidebook.
    2. Mario’s Gauge R & R. Can be available from the author or ASQ.
    3. There are bunch of good articles from Quality Technology Journal on MSA.

    0
    #110799

    Chris Seider
    Participant

    TP,
    Doug’s second post was invaluable on the sources on MSA.  I only use the AIAG book on MSA because it pretty much tells you all you need to know.
    Using the %P/T is ideal for your situation since you don’t have many samples which can give the total process variation.  The % tolerance measurement is what I refer to as the final determination of a valid gage.  Many BlackBelts in training misapply the Gage R&R tool and don’t get the proper variation for a product line so looking at the % tolerance works well. 

    0
    #110960

    Ron Manubay
    Member

    What it actually gives you is the info whether your gauge is still capable of detecting the normal variations in your process. This is important for continuous improvement. As your process improves (variation decreases), there may be a time when your current gauge may no longer detect the shift or improvement in your process since it is already too small for your gauge to detect. This will signal you to look for a better gauge (higher resolution?. I agree with you that it is hard to get samples that would represent the total variation in your process. Just get the best representation you can.When you compare your GRR with your tolerance, this is to see whether your gauge can still distinguish the bad from the good, taking into consideration the risk of having a misclassification. The higher your %GRR, the more likely it is that you will get some misclassifications specially in borderline situations.Rgds.

    0
Viewing 7 posts - 1 through 7 (of 7 total)

The forum ‘General’ is closed to new topics and replies.