GRR on ID with set of pin gauges

sinned

Registered Visitor
#1
Hi all,
I've repeatedly see engineers conducting GR&R (ANOVA method) with a set of standard pin gauges in certain increment, like 0.01mm. With such kind of measurement method, is there potential issue that will be challenged by customer or auditor ?
 

Miner

Forum Moderator
Staff member
Super Moderator
#2
The biggest potential issue I see is that you may not have sufficient resolution. Low resolution can affect your results by making them better or worse than they truly are.
 

sinned

Registered Visitor
#3
Hi Miner,
Yes. You are correct that the resolution will be limited by the increment of those pin gauges. You generally can't go further than 0.005 increment and I doubt if the hand feeling of 0.005 is distinguishable or not. So I usually recommend engineer to explain to customer and seek for their endorsement. The chance customer will accept this is not low to my experience especially when the holes are functioned to have shaft to fit into, ie. the pin gauge measurement simulate better the real world situation as in their assembly process. If you use point touch measurement equipment you may not get the least thru hole diameter.
:)
 

Bev D

Heretical Statistician
Staff member
Super Moderator
#4
to expand on Miner's point - the resolution is a relative thing. If the tolerance range is small compared to the pin resolution, then you get 'chunky data'. chunky data is defined as when there only ~2-5 possible results in the tolerance range. This will result in an overestimate of the measurement error SD and your R&R will most likely fail.

If pins are the best choice (because you need to detect the minimum through hole diameter) then I recommend using a categorical gage R&R. This would be a more accurate and useful assessment of the measurement system.
 

bobdoering

Stop X-bar/R Madness!!
Trusted
#5
Pins are really no different than the resolution of the lines on a dial or any other gage. I have used "tenths sets" as variable data very readily - if the process variation or tolerance (whichever was applicable) was wide enough. Some people think it is not "continuous" data, but, again, it as continuous as the dashes on your dial gage! One of the best ways to measure MMC including fit.

As far as explaining to a customer, never assume they are more expert than you. Be prepared to train them, if need be.
 
Top