Recall points from contour scan

[[Ri...]]

I have some trouble getting the minimum and maximum radius on some teeth of a gear segment.
I'm trying to scan the contour and recall the minimum radius and maximum radius of the tooth segments.
Recalling these single points or radius points from the contour doesn't seem to be correct.

radius point.JPG

[[ch...]]

If you want to check radius 56.4 , I think you have to recall all the points of that radius to a circle and change the evaluation to maximum inscribed circle.

[[Ri...]]

Recalling feature points into a circle doesn't seem to work properly. This picture is taken from our offline station.



After some fiddling around with stylus radius correction and re-calculate nominal geometry the measuring points seem to be correct.
Is this normal?
I doubt this measurement will be accurate/stable when we have our first products.

[[Ma...]]

It seems like you forget to switch outer/inner circle. At least your vectors points to inner circle, but you have outer circle.
Do not recalculate - if you have measured feature, then it will recalculate from measured data.
I recalculate only when i choose something from curves, but without measured data ( so recalculate from copied feature, or remove measured data from feature ).

[[Ri...]]

I use recall feature points from Curves into Lines/Circles all the time with great success.

[[Ow...]]

You should probably make sure you are using filters in the recalled feature points feature and then also make sure the filter isn't removing too much. Gear roots (especially hobbed) can get sketchy/rough and one or two rough points can throw the whole radius off.
Also, when you select the area to be recalled, you can right click on the box selection and choose to add range limits to evaluate if necessary.

[[Ri...]]

Please sign in to view this quote.

Thank for the reply. I am running into some performance issues of the software when using more than 6 to 8k of measuring points.
I need allot of points to be able to filter small parts of the contour properly. Is it advisable to first recall a part of the contour and filter this in a second step?

[[Ow...]]

I believe any time you recall feature points, you get the raw unfiltered points even if the feature it's recalled from have been filtered, so yes.
I don't deal with a lot small stuff but, one thing to keep in mind is the probe size and speed as the shaft rigidness becomes very weak on small probes. I've measured some very small radius's in the past with a 0.5mm probe and had to slow it down to 0.3 (mm) or 0.004 (inch) to get it to work accurately. With the slower speeds I was also able to reduce the points.