Surface Profile Gage R&R Study

[weisusu]

Hi there. Have anyone done a Gage R&R study for surface profile for instance 0.005"|A|B|C? Would you use the min deviation, maximum deviation, or 2X worst case deviation as the response for the Gage R&R study? Any thoughts is greatly appreciated.

 

[PaulJackson]

Hi there. Have anyone done a Gage R&R study for surface profile for instance 0.005"|A|B|C? Would you use the min deviation, maximum deviation, or 2X worst case deviation as the response for the Gage R&R study? Any thoughts is greatly appreciated.

I recommend… none of the above!

Use +0.0025 and -0.0025 for your specification limits and use the actual deviation from the basic profile + or - as the measured value.

How ever many points you choose to monitor is up to you and your reasonable expectation of random variation.

You could choose to pick one, two, or all of the points per part that you chose to gather from the surface… but so that your GRR has is most likely to succeed… you should control the measurement process… how, where, and how many point measurements are gathered and what criteria is used to select a subset of those points for the measurement results i.e.
1) largest deviation from basic… either most positive (least negative) or most negative (least positive).
2) both extreme deviations from basic… most positive (least negative) and most negative (least positive)
3) all measured points

To see how to assign the + and – to the measured deviations see the other thread you started.

http://www.eng-tips.com/viewthread.cfm?qid=249113&page=1

Paul