On multiple occasions I've had boolean operations fail in Rhino5 that I was able to complete successfully when I move the file into Rhino4. Has anyone else had this issue? Is this a known problem? Is there a solution on the user end? I've uploaded a file with this issue, if anyone is interested in having a look.
Cheers,
Ryan
Tags:
Permalink Reply by Barry Mair on October 17, 2012 at 4:51pm Hi Ryan.
I am experiencing this issue at the moment.
When objects are to be joined, it fails when they are touching or flush.
If I revise so the objects actually intersect, all works as expected.
I am still going through scenarios, but it is definitely an issue.
Barry.
Permalink Reply by Barry Mair on October 17, 2012 at 5:03pm I find the same with your file.
By amending the pieces which I did purely by moving faces so as to create an intersection between pieces, all unioned as expected.
Permalink Reply by Pascal Golay on October 18, 2012 at 3:51pm Hi Ryan- I see it, thanks, I'll add this to the bug heap. I'm surprised this works in V4 actually, usually Booleans involving non-planar coincident faces will fail in those areas.
-Pascal
pascal@mcneel.com
Permalink Reply by Pascal Golay on October 18, 2012 at 4:01pm One other thing to keep in mind - since Boolean operations are essentially shortcuts for Split/Join/Delete, it is usually possible to do this and get the clean result you want rather than moving objects just so that they intersect better for a Boolean and a result you don't want. In this case, extractSrf the back (roundy) surface of the smaller part and the pair of coincident flat faces, use the small part to Trim the corresponding location on the larger part and Join.
-Pascal
pascal@mcneel.com
Permalink Reply by Barry Mair on October 18, 2012 at 4:09pm Thankyou Pascal.
I found a similar problem with 'difference' last night.
Putting a mitred end on a box, by subtracting a box aligned to an edge and rotating along that edge, failed everytime.
Barry.
Permalink Reply by Pascal Golay on October 18, 2012 at 4:28pm Hi Barry- if you find cases you think should work, please send them to me or to tech@mcneel.com for a look. Ryan's example shows the intersector in V5 is not correctly handling the coincident cylindrical faces, and there is now a good chance we can fix that.
thanks,
-Pascal
pascal@mcneel.com
© 2013 Created by McNeel Admin.
