I've been having an issue measuring concentricity. Could someone explain best practices for checking concentricity between two holes/pins at opposite ends of a part? The parts very from 4" to 12". I can spin them on the datum on my V-Blocks and show them as basically perfect with an indicator (indicator/v-block has been certified), but when I run them on the CMM I show them as being out of tolerance by like.. 1mm~/0.040". I have no idea if I'm just setting up the feature wrong or what, any thoughts?
What is the length of the datum? Are you measuring the entire length of the datum as a cylinder? How many points? How many points for the measured feature?
A print snippet or code snippet would be good for assessment as well.
I've at least partially figured it out after playing around with it for a little while. It seems that the vector of the cylinder (DAT-A) heavily influences the concentricity? I don't know why. If I construct a cylinder around DAT-A and reverse the vector, the measurement becomes closer to what I would expect to see on this part.
Also to answer your questions:
-Yes, I'm measure the full length of the datum as a cylinder (approx. 25mm)
-3 levels with 6 points on each level for the datum
- 9 hits for the feature that's being held concentric to the datum
Seems like you might have found the problem. But just a side note, PC-DMIS evaluates concentricity as TP. If you actually want to measure true concentricity on the CMM, it can be quite cumbersome. You would have to measure using diametrically opposed points at an infinite number of cross sections across the entire surface of the measured features. Then construct mid points at each cross section. Then report the deviation from center. This would give you the error at the center at each cross section.
If you are using the Geometric Tolerance command and referencing ASME Y14.5, you are able to choose whether concentricity is evaluated as median points or the axis method.
You would have to measure using diametrically opposed points at an infinite number of cross sections across the entire surface of the measured features. Then construct mid points at each cross section. Then report the deviation from center. This would give you the error at the center at each cross section.
This is the median points method.
The axis method is the same as position and is the method that the older XactMeasure command and legacy dimensions use.
