How to perform Gauge R & R on Automated Test & Measurement Equipment

[cutsdean]

I am looking for help and/or training material on how to perform MSA/Gauge R & R on automated equipment such as a CNC controlled CMM or automated test equipment.

Any help or direction would be greatly appreciated

Cutsdean

 

[Miner]

A little more information is necessary. While there may be no Operator reproducibility, are there multiple stations that could introduce station to station reproducibility? Also, can multiple measurements be taken of the same part?

 

[cutsdean]

It is single station equipment that once the part is set up in fixturing etc is measured or tested without human intervention. Therefore the main source of variation would be in the set up, but i am unsure how to capture that.

 

[Miner]

If the same part can be measured multiple times, you can use what is called a Type 1 R Study. This assesses Repeatability only.

However, if there is a possibility that the fixturing of the part could affect the measurement, you should still use a standard R&R study.

 

[Jojoj]

I understand for GRR 0-10% is acceptable criteria. What if we are only measuring Repeatability only, what would be the acceptable criteria?

 

[Miner]

The criteria remain the same.

 

[Bev D]

the other cautions I have are that different operators can effect the positioning of the part in the fixture. if using a CMM, this positional difference can lead to 'measurement error' (the tip hits the part at different locations and/or angles or the shank hits before the tip does...) If you have within piece variation, the positional change can have a substantial effect on measurement error. Remember you are not testing the CMM, you are testing the whole system.

the same holds for any ATE from electrical board testing to NDE and destruct testing of physical properties to vision systems to functional tests of any kind

I would recommend a standard multiple part, multiple readings, multiple operators design.

 

[WCHorn]

I would recommend a standard multiple part, multiple readings, multiple operators design.

I agree. We do standard 10/10/3 gage R&R on our video CMM. The operator can have a significant effect on reproducibility, especially if you're measuring a part that is made of a less-than-rigid material, like rubber or plastic. We remove the part from the machine for every measurement, so positioning, fixuring and other variables can be accounted for. On a video CMM, lighting can affect the outcome significantly as well.

 

[Jojoj]

Thanks for the inputs. We are using a handler machine to insert the units for test to the socket. So I guess operator variable is negligible, they only pour in the units into the input bowl and the handler machine will do the rest of the activities until test is completed.