I am having a strange problem with the Boolean Difference Modifier. Below, is a screenshot of a tower leg section. On the left is the original section, and on the right, part of the bottom has been sliced away, with the Boolean modifier, to match an identical piece, that sits at an angle:
Look closely at the top images (Top Ortho). Notice how the last ribs at the top, on the right, have been removed. I cannot understand why this is happening. I have tried redoing it, countless times. I have used edit mode, to remove any duplicate vertices, and I have recalculated the normals. I even used ctrl-alt-shift-M, to look for abnormalities. The cube, used for the slicing, is much larger than the piece. So, that should not be the problem. Further, I performed the same action, on the identical, angled piece, with a different cube, without this problem.
Can anyone tell me why this is happening, or what I might be doing wrong?
What I am trying to do, is cut the bottom at an angle, so that I can join it with a similar piece, to form a bend. The cut that I am trying to accomplish, can be seen in the smaller before-and-after ortho views, at the bottom (the top ortho is a top view, and the bottom ortho is a side view). I honestly don’t know why that smaller protrusion keeps disappearing, no matter what I do.
Get rid of the n-gons. Check presence by Select -> Faces by Sides where n=5 or more in Edit mode.
Instead of booleans use Shift-d selected mesh (top or bottom faces), move and rotate. Bridge selected face loops on sections.
Didn’t work. I checked for n-gons. There are none. Just in case, I tried subdividing any suspect faces, in edit mode. The results were even more unwanted and unpredictable. I don’t really understand the shift-D method described.
This is booleaned bottom, duplicated top faces and Bridged under weird angle. Shift-d can duplicate selection. Nothing’s got missing.
Maybe extruding in some predefined direction is easier. Could be Spin would work for you, or Shear.
Bring in some real thing; as you can see talking much doesn’t quite help.
Success! Thanks Eppo. Actually, your very first suggestion (eliminating n-gons) is what eventually led me to a solution. Though the "select’ function could find no n-gons, I began to experiment with knife cuts, across the problem end faces, to try to eliminate the possibility of hidden n-gons. Initially, that didn’t work, and I got even stranger results. But, eventually, I found that holding down “control”, while making the knife cuts, fixed the problem. I was now able to do the Boolean subtraction that I needed, without losing any of the ribs. Why that worked, I still have no idea.
Well, I’m having the same problem again! I had to redo the mesh shown above, and this time. the trick that I used before, with the Boolean Modifier, will not work. In fact, I can’t get any of the modifiers to work properly on this mesh.
At least, this time, the non-manifold (ctrl-alt-shift M) is flagging the problem edge. Here is a screenshot of the non-manifold result:
But, how can I fix this? That edge HAS to be there! The two faces meet at an angle. There are no duplicate vertices, and the normals have been recalculated. I tried removing the hidden face. But, that only made matters worse. Nothing I have tried, will fix this—not even any of the Mesh-Cleanup commands.
Thanks Eppo, I got it to work, by making three knife cuts, these two shown, and one on the opposite face:
I then deleted the hidden, interior face, and it all worked fine.
Funny though, I swear that I tried this same approach yesterday, and it didn’t work. Yet, today, just like before, after shutting the computer off for a while, and coming back to this problem later, suddenly, the simplest solution suddenly seems to work. Could the computer be getting tired? Does Blender somehow get old stuff stuck in its memory?
