double flank rolling system vs. GearPro

[[Se...]]

Good morning,

we measure a lot of worm shafts with Gear Pro. This is not a problem. But a lot of our costumers measure their parts with an double flank rolling system. There are often big differences in results. Somtimes they measure ok parts and we measure nok and sometimes we measure ok and they nok. I tried to verify it with runout measurement self centering in Claypso but there are also differences. I know that these are two complete different methods but we dont have an double flank rolling system. Has someone an idea or experience how I can measure with GearPro similar to double flank rolling system?

[[Ch...]]

Composite roll testing is different than analytical - it will never be apples to apples.

Composite is a combination of errors. In most cases if composite is good, then analytical should be good, however there are exceptions, such as a lead that is angled and loaded on the edge of the tooth ,etc.

I've only "heard" of a Mitutoyo CMM where the table moves and the bridge is stationary- that the software can "simulate" roll testing with a master.

In theory, you could come close with software and technology now -a-days. I think you would need to map and 'actual' master- not a theoretical perfect nominal mating gear. You would also need to do at least 3 lead and profiles- probably all teeth.

Then you would also need to do 3 lead and profiles on all teeth of actual part - which ends up taking a long time. Composite is meant to be economical and easier to save time.

[[Ow...]]

What Chris said, it will never be apples to apples.
Some of the differences can be related to the fixtures/abors/centers they use in the composite inspection not being true or possibly bent.

[[Kl...]]

There are two important boundary conditions for such measurements:
- GearPro knows calculated against the CAD model, normvector surface
- CALYPSO only knows the target values ​​(point + normal vector).
In the case of a single-flank measurement, GearPro can calculate out inaccuracies in the scanning path, caused by deviations in dimensions, position and shape, using the CAD model (e.g. as with a circle on a cone, ... in CALYPSO).
Single flank by CALYPSO requires a CAD, or special transformation to get surface normal vectors (picture).
With a flank with e.g. 45 °, a path deviation of 0.010 mm leads to a form deviation of the same size not compensated!
In order to stabilize the path, CALYPSO usually scans in a self-centering manner.

Summary: There are a number of reasons for different results. Results are only approximately identical if the measurement strategy is comparable.