MSA/GRR on position and profile

[[Ch...]]

Fellow CMM Programmers :
.

Regarding MSA/GRR gage study (assume type 1 for now)

Regarding position and (surface)profile where the reported value is kind of doubled from actual measurements, do you apply the 10/20/30% MSA tolerance to the end reported value, or should you apply to the measured deviation/axes instead?

If my GRR limit was 30% and I have a diametral zone position, how much then would I allow each axis to vary? 15% each ?

I think applying the MSA/GR Tolerance % to the end value result would be a bit too much..., sure I would love that to repeat <10% or less than 30% but I think we're hurting ourselves that way.

❓ ❓

Thank you in advance.

Chris

[[Ri...]]

Please sign in to view this quote.

I have debated this in the past for Tur Position or Profiles.
Examples are based on True Position (TP) and XY axis.
You can have a TP of the same feature that is reading the same and or similar value but have varying coordinates opposing each other.
This can create a misleading GRR and or MSA.
If you track the X and Y coordinates you will have a better understanding of what is happening when applied to a GRR or MSA.
Create an MSA and populate with TPs values only.
Then create an MSA and populate with coordinates of those TPs only.

I've done this before and shown where an MSA showed the results were about 10% and the coordinates showed the results were more towards 17%.
Several customers I have worked with now generate MSA's for both TP and Coordinates.
The above doesn't imply results would always be worse, I would guess at times the results could be better.