In the attached file I am getting corruption of the solid object when I difference the surface. This will also happen if I extend the surface edge into a closed curve, extrude that curve and then do a difference with two solids. This kind of thing will happen on rare occasion and my solution is always to rebuilt the solid. I switched to V3 beta after experiencing viewport errors with a different object in V2 (this has fixed things in the past) but as the title implies I am getting this error (although different geometry failures) in both versions.

Thanks,
FDP

Hi FDP - the bug seems to be something that's happened some time before this step, there is a messed up spot on the base object in at least one spot here:



If you zoom in to that area, it looks like a hole is there, if you then use Edit > Separate on that object and isolate just that one surface and turn on it's surface control points it looks like this:




So whatever generated that surface has done something wrong and has the trimming boundary all the way off the actual surface. Some previous bug kept the wrong part of the surface or something like that.

So at least that particular surface will need to be deleted and reconstructed, maybe others, I'll take a look.

The actual bug is most likely in some previous step that built that piece though.

- Michael

Thanks Michael!

I'll do a rebuild.

I had noticed the occasional surface void in this project before this problem, but assumed it was a viewport error (those seem to occur in V2 on rare occasion). Since the object still listed as "solid" and not "joined surface" I assumed that it was intact. Is it possible that this is just an error in the way that MoI reports whether an object is solid or not?

Best Regards,
FDP

Hi FDP yeah the problem seemed to be that one particular bad surface shown above.

In the attached file I have deleted that one surface, and built a new one by extruding one of the remaining surface edges over and then joining the pieces together to make a solid again.

With that face fixed the boolean difference now seems to be ok.

If you happen to have a previous version of this object right before that rounded spot was constructed, if you can possibly send that to me at moi@moi3d.com that's the step that did something wrong that would be good to fix up.

- Michael

Hi FDP,

> Is it possible that this is just an error in the way that MoI reports whether an object is solid or not?

Yeah sort of in a way - basically MoI will label an object as a solid if it has all edges joined between 2 surfaces (or joined across the same surface as a seam edge on a closed surface).

There are various kinds of ways that objects can be malformed that will still fit that basic criteria like with trim boundaries that have little self-intersecting loops, surfaces that intersect themselves, or pieces of the model sticking through each other, or other kinds of malformed boundaries. This particular one with the boundary actually located completely off the surface is not very common.

Unfortunately it would be extremely time consuming for MoI to try and do an intensive in depth analysis of the object to determine these kinds of malformations.

- Michael

Hi FDP,

> I had noticed the occasional surface void in this project before this problem, but assumed it was
> a viewport error (those seem to occur in V2 on rare occasion).

It's possible for something like that to be a bug in generating the display mesh, but usually it means there's some problem with the trimming boundary.

If you go to export it to a mesh format like OBJ for example, and the export mesher also made a weird looking result in that same spot, that makes it even less likely to be just a display problem.

- Michael