Ellipse Length and Width

[[Da...]]

Anyone know if there is a way to report out the length and width of an ellipse?

[[Ri...]]

Can you not report the D1 and D2?

[[Da...]]

Yes, but ideally what I need to do is to report out physical length and width of the ellipse. Attached is a screenshot of a scan of an elliptical feature. There are some irregularities due to welding that I would like to take into account.

I think I found a way to do it in Calypso by creating an XZ and XY plane and using those to intersect the feature. From there I can combine the two intersection results.

Capture.PNG

[[Ri...]]

Ah OK.

If that's the case, I would just use Curve, and create the intersections that you are looking for.

[[Da...]]

Use curve, then ask for curve distance. You can ask for any coordinate. It's in Size, Curve dimension.

[[Da...]]

Thanks for the tips. Much appreciated!

[[Re...]]

Not sure if this is a correct way to do it, but it seems to work for me if the minor and major diameter is not too extreme. I use the cmm with an active head and program a median circle size and report the maximum circumscribed and minimum circumscribed size or the min/max 2 point diameter. I'm not sure about the welding irregularities - perhaps filtering?

[[Ri...]]

The Max/Min would not be good for an ellipse. Those algorithms are based on mathematical equations of calculating the 3 most extreme points.

The 2-point diameter method could work.

[[Da...]]

Hello, make it a closed curve then ask for CURVE WIDTH. It's in Size, Curve dimensions. Ask twice, once in each direction. That's what that feature is for.

[[Re...]]

Richard - I often thought that the mathematical algorithm on the circle evaluation would not be the best method and that the 2 point diameter method would be better. However I would always get the same result by either method. Maybe its due to our ellipse being highly precision ground. actually our ellipse are not a true ellipse, they are more of a spline. But I have to report the min and max. I would hate to change something that seems not to be broken, however I'm open to learn of a better method. Thanks for your input.

[[Re...]]

Dave - For my so-called ellipse I tried to use curve, but I had a difficult time with clocking. Our ellipse has nothing to time to but itself. any good methods for establishing rotation? If I could establish a good method to clock to I would give curve another try.

[[Da...]]

If you can't clock the ellipse then none of the other methods will work either. Do you have a model? If so, You can use best fit to the model allowing rotation in the Z axis. The amount of rotation is reported ( you may have to use a results element to find it ) and you can create a new alignment with the rotation applied. Now you can use curve expansion in both directions. Does that make sense ? I get a bit excited about this type of thing. 🤣

[[Ch...]]

I would try and get Dave's recommendation to work. I measure weird slots/ellipses and I do exactly as Dave does. I measure it with curve then use Curve Width to measure it.

[[Re...]]

Unfortunately I do not have accurate cad models for these parts, however the cad model would have the same min and max value. But the results seem to be correct with the circle scan using an active head and the different evaluations. I originally tried to use curve but was unsuccessful. I'ts been awhile but I did try the best fit thing, but if it wasn't near the cad rotation it would not work, and when it was close to clocking the result were not repeatable.

[[An...]]

Please sign in to view this quote.

Use the two focal points of the ellipse to establish a planar alignment.