Failed boolean operation in NX 12

[NX10user]

Dear Experts,

I'm trying to perform a subtraction boolean operation on my geometry, but the boolean operation fails with an error message- "Unable to perform boolean". I have tried to see visually, that my geometry has body intersections but I don't know how can I solve these? The body is faceted and I have tried to move the faces individually but couldn't. Can anybody advise me how to solve these body intersections? You can find the .stp file here:

https://we.tl/t-dlIQgWebSK

Regards
Chaman

 

[jerry1423]

Extend the faces of the weave beyond the face of the block, just a little passed the block faces will work.
You may be able to use Offset Face to do it.
It also appears that you are working in an assembly, which changes quite a bit.
So, here is what I would do.
In your assembly Promote all those interior solids the solid bodies.
Offset the side walls of the cube -0.5 mm (so they will move inward).
Your subtraction should work, but I did not try it.

Jerry J.
UGV5-NX11

 

[NX10user]

Hey Jerry.

Thank you for the suggestions. Yes I tried both of your recommendations but unfortunately, the body intersections still remain. I don't know how can I remove them. Is there a way to locally move the edges of the faceted weave (the edges which surround the places of body intersection) and the surfaces could follow it?

Regards
Chaman

 

[jerry1423]

Can the end faces of the weave move ? It depends on how it was modeled. Unfortunately these weaves were modeled where the side face of the weave end at exactly the end faces, so beyond those end faces they are no defined so therefore the end faces cannot move outward, only inward.

Correction : I went into one of the weaves and I was able to offset outward the end face, so what you will need to do is to go into each weave and offset the end face (0.2 mm will be fine) so that the end face extend beyond the cube face.
Then what you can do with the assembly is export the entire thing as a parasolid, bring import it into a new part, and then do the subtraction in there so the analysis can be done.

Jerry J.
UGV5-NX11

 

[NX10user]

Dear Jerry.

Yes, the weaves end faces can move inward as it just a representation and is completely parametric, but would it lead the interpenetrations to resolve? I'm trying to understand the physics behind your recommended method.

Thank you very much.
Chaman

 

[jerry1423]

The current problem is that the end face of the weave is exactly on the face of the cube, and when doing a subtraction it cannot determine if there should be a hole there or not (it theoretically is a zero thickness face, which is not possible) , so therefore it comes up as an error. When extending the faces of the weaves outward before the subtraction you are simply extending the depth of the cut - in the same when a machinist drills a thru hole he does it more than just the metal thickness.

Jerry J.
UGV5-NX11

 

[Toost]

There is the move edge feature that could be used to modify the shape, but i rather think you should re-think the entire process.
Since I do not know why or what this model is, it's hard to propose a method.
I do know a few details based on the model you supplied.
This is a pretty small model, the block is 8.7 mm, each strand is ~1,3 mm across
( based on this , maybe you should tighten the Modeling distance tolerance. )
the strands might be the output from a FEM System ?
one of the strands has ~4400 faces, the other 9 probably similar number.
each face ( seems consistent) is a single patch b-surface 1x1 degree. I guess this is an approximation of a smooth shape.( it is not smooth as is in NX.)
If i guess, the vertices of these faces are not very accurate, a deviation from the ideal position is probably allowed in that other system , a close approximation in NX should work as well.
depending on what you want to do with this model, a remodel of the strands might be the thing to do.
What system do the strands come from ?
what is the purpose of this model ?

Regards,
Tomas