Total Runout Measurement using CMM

[romulus102]

Hello,

The CMM section in my workplace are measuring total runout on a bearing seating surface. The way it was explained to me is that they take scans axially along the surface,construct a cylinder and calculate the total runout. I am struggling to understand how this calculation exactly works.The confusion is arisen from the fact that I was always shown total runout being measured at 3 locations i.e. a DTI placed on the part, part rotated, next axial position checked and the maximum of the 3 readings was the total runout. On reviewing, Fundamentals of Geometric Dimensioning and Tolerancing by Alex Krulikowski, this book states that total runout needs to measured in a manner such that the DTI follows a helical path.

Based on the above description,I would appreciate any advise on the two questions below.
1:what is correct method to measure total runout as per ASME 14.5.
2:How does a CMM(LK Machine with a Renishaw retrofit, using Modus software) actually calculate total runout.

 

[3DDave]

ASME 14.5 makes no specific recommendations about how to measure total runout. Any discrete measurement of a continuous tolerance zone will depend on the observations of the operator as to how much deviation is likely for the uninspected portions of surfaces relative to the locations of the measured locations. Krulikowski's recommendation is a good one, but I would not take it as the only acceptable method. Like any measurements, one needs to understand all the error contributions for the equipment at hand.

Examples shown on the web suggest using V-blocks to establish a datum reference, but V-block can be unreliable for that purpose as they do not discover the true axis of a feature. For example:

https://www.youtube.com/watch?v=s-UG2CZ8VMM

where concerns are expressed at the end about how the setup is made. A blind 'helical' system would likely damage the dial indicator if it dropped into the holes.

You'll need to contact the Modus software makers for questions about how they do their calculation. The website claims "Certified mathematical algorithms"

http://www.renishaw.com/cmmsupport/knowledgebase/en/modus-measurement-software--22123

 

[pmarc]

romulus102,
From theoretical point of view the CMM, after probing the considered surface, should look for two virtual cylinders, coaxial with each other and with the datum axis, that would have minimum possible difference in radii while encompassing all probed points. The actual difference between radii of the two cylinders would be the actual amount total runout error.

The probing strategy may differ (helix, collection of circles at different axial locations, collection of lines at different circumferential locations, etc.) depending on different error contributions, but the assessment method should be the same.