CMM inspection for Cylindricity

[BigKoz]

We are producing a pinion, approx. 30mm long that calls for the cylindricity of the I.D. to be within 7 microns. Using a CMM we initially programmed the probe to scan in a spiral motion throughout the length of the pinion I.D. (except the first millimeter on each end). The results on six pieces were 8-12 microns. Our customer informed us that the check the parts with eight hits at three points in the I.D. Using their methodology the results were from 2.5-6 microns. Which is the proper methodology, and why the varying results? Is it due to the scanning throwing out the upper and lower readings?

Regards,

Tom Klosowski
Delta Research Corp.

http://www.delrecorp.com

 

[R1100S]

I would agree with your method BigKoz, but if the customer is happy, what is the problem. They're the ones calling out the spec.

 

[BigKoz]

I agree R1100S, the customer is "ALWAYS" right!
I guess my main question was two-fold; which methodology would be the preferred, and why the great variance between the two checks. We are concerned that variance may be due to a software issue that could raise its head in other areas, but if the variance we saw is normal we can just go on with life and chalk it up to experience.

Regards,

Tom Klosowski
Delta Research Corp.

http://www.delrecorp.com

 

[R1100S]

If I am understanding you correctly, when you performed the test by their method, you measured a circle with 8 hits at three locations along the 30mm length(5,15&25mm for example).

By scanning in a spiral motion, you were checking to see how closely the actual part matched to the theoretical profile of a cylinder. With their method you were checking the roundness of three circle and the concentricity of those circles. The results were much better, which is to be expected, with the second method.

 

[BigKoz]

Thank you for your response. We are actually 2nd tier on this item, and it was our customer who found out how their customer was checking the part. We felt, as you illustrate above, that the CMM is actually calculating the centerlines between the three diameters and calling out the distance between the high and low from the calculated centerline. We will check their parts in the manner they prescribe and continue to inspect other parts as we have in the past.
I appreciate your help.


Regards,

Tom Klosowski
Delta Research Corp.

http://www.delrecorp.com

 

[Tmoose]

Were any of the 3 point measurements within 2 mm of the ends? If there was much bell mouth, barrel, taper, etc, then your first method was catching it, and the 2nd generation method was not.