|
*Please be aware that SOME RECENT forum threads may not yet be indexed by Google. |
|
View Full Version : Gage R&R constants - Appraisers vs K2 for 1 to 5 Appriasers - Seeking table in a file vargadomonkos 12th September 2005, 05:04 AM I need a little help in GRR studies, I couldn't find the constants of GRR tests even in the MSA Manual. Here is what I need: Appraisers vs K2 for 1 to 5 Appriasers Parts vs K3 for 1:50 Parts Could any of You attach a table of these in a file? THX Jim Wynne 12th September 2005, 11:07 AM I need a little help in GRR studies, I couldn't find the constants of GRR tests even in the MSA Manual. Here is what I need: Appraisers vs K2 for 1 to 5 Appriasers Parts vs K3 for 1:50 Parts Could any of You attach a table of these in a file? THX Check pages 115-116 of the 3rd Edition AIAG MSA manual; I think the information you're looking for is there. vargadomonkos 13th September 2005, 04:59 AM Right and THX, but I still cannot calculate for higher part number, over 20. Jim Wynne 13th September 2005, 09:33 AM Right and THX, but I still cannot calculate for higher part number, over 20. Why on earth would you want to? Measuring 21 parts isn't likely to tell you anything useful that measuring 20 won't. vargadomonkos 13th September 2005, 10:25 AM In some cases the variation between parts can be high for small part number size, which can be reduced by increasing the part number. Please prescind the mathematical unexactness of reducing the variation. The greater the partnumber the more realistic the view of the process. In most cases small part number is used, but sometimes more exact results are needed. THX Jim Wynne 13th September 2005, 10:31 AM In some cases the variation between parts can be high for small part number size, which can be reduced by increasing the part number. Please prescind the mathematical unexactness of reducing the variation. The greater the partnumber the more realistic the view of the process. In most cases small part number is used, but sometimes more exact results are needed. THX Please prescind the mathematical unexactness of reducing the variation.I'm not sure what you mean, but what you say might make sense for say, 10 parts vs. 15 (although with part-to-part variation that high it seems that you might have a different problem), but again, how will 50 parts help you more than 20, or 15 insofar as detecting sources of error is concerned? vargadomonkos 13th September 2005, 10:42 AM What happens if the Measurement system is very close to the "discision line", where intervention is needed. I just want to check what would happen if the part number was increased. Otherwise I do not understand, why I couldn't find the answear anywhere. Sorry for beeng a chap that likes to believe, what he sees with his own eye. Jim Wynne 13th September 2005, 11:06 AM What happens if the Measurement system is very close to the "discision line", where intervention is needed. I just want to check what would happen if the part number was increased. Otherwise I do not understand, why I couldn't find the answear anywhere. Sorry for beeng a chap that likes to believe, what he sees with his own eye. No need to apologize. It's good to make sure of your data. I was merely suggesting that you might be overdoing it, but it was just a suggestion. I just had a glance at the AIAG SPC manual (1st Edition; I don't have the new one yet) and there's a table on page 143 that shows d2 values for subgroup sizes up to 25, if that's any help. Juran's Quality Control Handbook (4th Edition) also has values for up to 25 (page AII.3). Since you apparently already have the values for up to 20, here are the numbers for 21-25: 21 3.778 22 3.819 23 3.858 24 3.895 25 3.931 vargadomonkos 14th September 2005, 03:02 AM Thank You |
|