Untrimming should also work across seams


(Gustavo Fontana) #1

I have to redo some holes patterns on a bunch of parts, and the first step is untrimming the existing ones. Rhino can only remove the easy stuff: holes that fall completely inside the boundaries of a surface. and it skips all these:

So I have to bring toe parts to Fusion to get it done, like this:

BTW, if I tried to manually do this with detach trim, etc, all of the other seams of the model start to fall apart and become naked edges, so I lose my solid. So that’s not an option.

Rhino should deal with this. I know it’s hard, maybe even not possible to do them all at once, but one at at time would be an acceptable compromise, that’s how also F360 understand those faces to remove are part of the same hole. If I tried to select more than one it fails too.

G


#2

This needs to be addressed there is a whole mess with seams and singularities a hold over from old rhino. I bet tech will say use, I think it’s merge seam can’t remember off the top of my head then try to remove hole? This is why Rhino should not use a nurbs paradigm as it’s base level but that’s another thing.
RM


#3

I suspect you could do this in Rhino faster than what your video
shows. The process would be to extract the 3 (or 4) surfaces that have the holes in them and then duplicate the outer boundary of those joined surfaces and then UntrimAll and then retrim the surfaces and join. After that you can window select all the little cylinders (the holes) and delete them.


(Gustavo Fontana) #4

Nope, I can do that, but I end up with a mess of naked edges.


#5

That probably means you aren’t doing what I suggested.
If what I suggested were impossible then the Rhino developers wouldn’t be able to do anything to help either.


(Gustavo Fontana) #6

I think you might be thinking that the joining/common edges of each patch are the same as their untrimmed edges? that’s not the case. they are more like this:

Look the selected (yellow) trim boundary after untrimming.

G


#7

No that’s not what I was thinking. I’ve done this thousands of times, I know what’s involved in removing or moving features that cross over boundaries

When you redo the trims that would include trimming the surfaces against each other (if it was necessary) to recreate the seam without the holes.
If you are asking what will happen if Rhino can’t determine the intersection between those surfaces in their untrimmed state? Then its likely that the command you are asking them to create will fail. One nice thing about Rhino is the users have at their disposal the same functions that the developers do.

In the couple of examples you have shown, it looks like the surfaces are tangent and thus the Rhino Surface-surface intersector is likely to not be able to find complete intersections. So that means to get a clean trouble free boundaries that aren’t just a bunch of broken pieces cobbled together (waiting to cause future problems) will require some level of understanding of how to interpolate clean complete boundaries from partial boundary information. The developers have shown little aptitude or even interest in solving that problem up to now.


#8

Ouch. Even if it does seem to be true.


#9

“Untrimming should also work across seams”

Fillet edge too, but often that is where the breakdown in fillet happens.

Just spent days rebuilding a trimmed patchwork class A polysurface to consist mainly of merged surfaces - no seams where it matters. Was a bitch! Had another reason to do so, so as long as I was at it decided I’d do it ‘right’…which was a major time suck.

But low and behold, Rhino now multi rounds (variable fillets) the thing as good as Creo!

(Creo ate the ‘bad’ seamed solid for lunch and asked for seconds. Fusion crapped out too btw.)


#10

You’re actually making it hard work in Fusion, you could just window select all the holes and remove all in one go.

Mark


(Gustavo Fontana) #11

Are you sure? go ahead at try it. I will blow up your model or will do nothing but giving you an error message. This only works one at a time, and not it all holes. So I was actually better off doing it all by hand in Rhino and then bringing the model with the unstitched/naked edges that Rhino messed up to re-stitch into a solid in Fusion360.

All CAD sucks… just at different things :rofl:

G


#12

Yes probably depends on the model. You could still window select the hole rather than select each face to speed things up.

Mark


(Gustavo Fontana) #13

good point Mark, I’m a total newbie at F360 and I struggle still with navigation and selection. I realized I was not window-selecting in shaded mode because I though it would also select the larger trimmed surfaces. Not the case.


(Gustavo Fontana) #14

Do you mean they could not import them correctly?

One thing I’ve found frustrating about Fusion and Onshape when I was using that it that sometimes they just say ‘failed import’ and leave you hanging, with no model, nothing. Like if that’s an acceptable solution. I much rather prefer Rhino’s approach: if you import a hairy model, it will import a hairy model and you can work with it.

I think the non-committal approach of bringing-in only easy/clean geometry is unacceptable.


#15

both imported out of rhino.

creo rounds successful, fusion (and rhino native) failed to round (fillet).

this is common with ‘badly’ (using that term loosely) constructed seamed polysurfaces. creo still deals with most, fusion sometimes, rhino almost never.

once rebult as merged, rhino rounded just as easily as creo. didn’t try same in fusion. assuming that it probably would work too. no time to try…

got to treat rhino ‘right.’ that takes more time. my experience at least