Locating theoretical features on unusual geometry

[[Th...]]

I have some slightly awkward parts to measure. I need to find the location of a point along the edge of a CAD model where a flat plane meets a line on the model that is perpendicular to that plane. The wall geometry this line is on is weird, the kind of geometry that Calypso can't create features out of. If I can't measure this directly, how do I infer its location?

I attached a zoomed-in picture. Hopefully it shows enough to give an idea of what I'm trying to do.

intersection.PNG

[[Ma...]]

From the screenshot you provided looks like you can scan a plane on one surface and radius on the other then intersect them. Line you have selected looks like beginning of the radius.... those lines are there just to show you boundaries of each feature

[[Th...]]

Problem is that plane isn't a plane, if you're referring to the short feature on the left of the 2D line. It's a swept cylinder of some type. I spent way too long trying to do exactly that before I discovered it wasn't flat.

[[Ma...]]

Please sign in to view this quote.

Ahh I see, if you click to extract a feature you can see what it defaults to.

Then you should be able to just do two circles and intersect them..... if it tells you that they don't intersect, which they should, then you can use Kink Point

Intersection.JPG

[[Cl...]]

Curve.

[[Th...]]

I can't really get circles either. Using the point set creation tool, I can ID the CAD entity as an elliptical cylindrical face, although Calypso somehow manages to extract a plane feature out of it normally. Sectioning the feature, creating either circles or ellipses off the resulting CAD data doesn't appear to match the surface at all.

Please sign in to view this quote.

I do have Curve. My problem there is I don't know how I would use Curve to find this point.

[[Ri...]]

Use Curve, and then the Kink Point.

[[Th...]]

On closer inspection, both those features are "elliptical cylindrical faces". I don't seem to be able to get suitable features out of either of them. Kink point requires nominal features to compare the curve to, doesn't it?

Also I'd like to clarify, since it's not very visible in that picture, that the intersection of the two features on the sidewall is not itself what I'm trying to measure, but that line's intersection with the purple-highlighted wall. It's that point that is dimensioned on the print, as a linear distance to another feature that's not shown.

[[Ma...]]

Please sign in to view this quote.

If you try to extract the feature from CAD if it does recognize feature type then you should be able to capture it, intersect each one to the sidewall using Shell of the features. The created intersections should internally be seen as circles which themselves should be able to be intersected for the location you are looking for.

Other method would be to create 4 circles total with "Define Circle on Cylinder" tool. Two sets near one edge and two near the other, intersecting each set will yield two points from which you can create a line and intersect with the sidewall

[[Th...]]

Calypso will extract features from these faces, but Calypso is wrong. The short face extracts as a plane, which it absolutely is not, and the longer face extracts as a regular cylinder, but the feature type is elliptical cylinder, and with a 30 inch diameter I can't visually see if it properly matches. I'm inclined to believe it doesn't. At any rate, the intersection construction does not provide a result. As for circle paths, I can't extract circles off the short face. Calypso will only generate a planar feature off of it.

I'm currently of the opinion that I can't directly measure any of this, so what about indirect? Can I use a fitting alignment with a bunch of points across these features to define the line location?

[[Cl...]]

I have in the past intersected two 2d curves. Have you tried that?

[[Th...]]

That's gotten me somewhere. I've sectioned the part along that feature, created a curve across, and intersected the curve with the theoretical line. I can do this on either side of the wall and recall the intersections into a line and then intersect *that* with the plane on the underside. It looks like this, I also have a 3D line recalling the intersections and an intersection of the 3D line and the flat face, but I left those off the selection because it's all overlapping features at this point.

curvetest.PNG

[[Ma...]]

Please sign in to view this quote.

Intersecting with that theoretical line will always give you perfect result left to right, only possible reading would be top to bottom. The line represents where one feature ends and the other one begins, the only way to get a result is by capturing the two features in any way and intersecting them.

You could do profile or points along those features to alignment to see if those two surfaces fall into tolerance.

[[No...]]

Please sign in to view this quote.

That's because Calypso uses feature approximation. It tries to extract the best matching regular geometry when it is unable to identify it clearly.