boolean not working

Next
 From:  Mr. Yuri (MR_JURAJ)
9107.1 
Hi Guys,
can anyone try to operate these two objects with boolean union, diff, isect, merge?
It just calculating for minutes and minutes and nothing happens.

When I try to move one and do a boolean while not 'in sync', it seems working.

CPU is at 14%.
I have i7 7700
16G RAM and MoI is on SSD

This should be quite simple operation looking at the shapes on forum that others did.

Thanks,
Juraj
Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
9107.2 In reply to 9107.1 
Many surfaces are overlaping themselves...no chance to work in a reasonable time! ;)

I reduce size of one of it : that is working in 2 seconds!



format OBJ inside MagicaVoxel :)

EDITED: 22 Oct 2018 by PILOU

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Mr. Yuri (MR_JURAJ)
9107.3 In reply to 9107.2 
I tried to wait and watch the time.
Came back 14 minutes later leaving PC alone and operation seem finished but it was not joined.
(like it canceled itself or something)

I tried to scale one model by factor 0,95 and it took about few seconds. (good tip)

I don't mind waiting for complex calculations, I understand.
But why it won't even perform it when both models are 100%.

Also, it is acting really weird.
Can you please try yourself following:
Scale down one model by factor 0,99 and do a boolean union or isect.
I think it should make one object merging everything in it.
But this do something like diff with the models (taking about 10 seconds).

Also tried merge in this way and it seems like it does absolutely nothing (calculating for few seconds).
Same goes when I try diff.

Can you please try yourself and let me know?

Thanks a lot

EDITED: 22 Oct 2018 by MR_JURAJ

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
 From:  Michael Gibson
9107.4 In reply to 9107.1 
Hi Mr. Yuri - objects that have surfaces or edges that barely skim over each other adds a significant amount of complexity to the boolean calculations.

In order to calculate a boolean, objects need to be intersected with each other and well defined closed intersection loops are identified and used to cut each object up.

When you have surfaces that skim over each other it makes for a complex intersection case. There are kind of large "zones" of intersection instead of crisp intersection curves.

So because of that it isn't expected that a case with such a significant overlapping surface area like you have here will be able to be booleaned well. It's not too likely to work in any CAD program.

One thing that might help is to keep the overlapping areas to have completely identical underlying surfaces. For your case here it looks like you had a cylindrical like base shape that has some cuts taken out of it. Then after that the object was duplicated, mirrored and rotated , is that correct? The boolean could possibly have a better shot if you generated the second object by mirroring and rotating the cutting objects only and applying them to the exact same base object. When you have coincident surfaces that have the exact same underlying surface that can help.

Also it might be possible to finish your model using lower level surface trimming operations though. To do that you would use Edit > Separate on your objects to break them into individual surfaces and then work with Edit > Trim on smaller areas of the model.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged
 

Reply to All Reply to All