Runout of small groove

[[Ma...]]

How would you go about checking the run out of this torus (groove) in the attached drawing. The way I'm currently evaluating the run out is by creating 8 circles along the torus recalled into a circle. The arc of the radius that I'm able to measure with a .5mm L probe is only 60 degrees for my 8 circles on the torus. The size of the radius of the groove is only .045 (Standard) I'm getting some extremely high results that I can't replicate with table inspection (indicator/pin on -D-). Does anyone have any suggestions on how they would approach measuring this feature?

Calypso 2016, Vast XXT, Curve/FF

RUNOUT-HOWSDG.png

[[Cl...]]

One thing you could try first would be to constrain the radius of the circle you are recalling the circles into. Have you tried recalling the circle points into a torus?

[[Ma...]]

Thank you for the suggestions. I did try to recall the circles into a torus but when doing so it will not allow me to use it in a runout. When constraining the radius on my recalled circle the run out stayed about the same.

[[Ma...]]

Theoreticaly you can measure torus to get real center and then measure just circle in exact center. This should be enough - no total runout used.

[[Er...]]

I agree with making an alignment with a torus, measuring R2. Then measuring R1 with a circle.

Also, there have been instances where recalling points does not yield give as accurate a reading.

[[Ri...]]

Ideally you'd want to be able to self-center scan around the equator of the donut. That would really only be possible with a rotary table.

Otherwise, I would suggest to measure 7+ arcs of the R2, and construct a circle using recall from the 7+ arcs. You can then use that recall circle inside of the radial runout characteristic.

Tough measurement without a lot of great solutions.

[[Ma...]]

Thank you for the suggestions. Measuring the feature as a torus and using that data to measure a circle at the center is the most reliable and stable method of everything I've tried for both size and location. Thank you Martin for the suggestion.