I was planning on smoothing pretty much all hard edges on this model with the bevel tool. But on quite a few of the edges the tool does not work, giving chaotic results as in the picture. I have tried resetting the scale and rotation of the object and I have removed all Ngons, it still gives this result. If anyone knows what's up or has a better alternative I'd be pleased to know. Thanks!
-
$\begingroup$ Is the mesh manifold? No internal faces, no double vertices? Is the scale applied? $\endgroup$– LeanderCommented Dec 24, 2019 at 0:38
-
$\begingroup$ The scale is applied, no internal faces afik and I've used merge by distance so there shouldn't be any double vertices (I checked by moving them, there arent't along that edge). I don't know what manifold means though. Thanks for your reply. $\endgroup$– AlabastineCommented Dec 24, 2019 at 1:51
-
$\begingroup$ Nonmanifold surfaces are essentially impossible geometry in the real world. Things such as faces/edges passing through one another. Could you post your blend file if you don't mind? $\endgroup$– stphnl329Commented Dec 24, 2019 at 6:16
-
$\begingroup$ Maybe you could diagnose it better if you add those edges to a vertex group, add a Bevel modifier limited to the group, and adjusted the width while in wireframe. $\endgroup$– JohnCommented Dec 24, 2019 at 6:22
2 Answers
Looking at that screenshot, I think that the bevelled edges are crossing over each other. This is often caused when trying to bevel sharp edges that are in close proximity. You may have to manually delete the local geometry, use the knife tool ("K" in edit mode) to cut in new geometry and fill in the new faces.
I Fixed this edge in the end, I think what caused it was I didn't take proper care while extruding, accidentally extruding vertices over itself. I deleted the local geometry several times before but it didn't help. Paying close attention and with your tips I figured it was probably the extruding, so I paid a lot of attention this time and it worked. Now I just hope I can fix the entire model like this. Thanks!
-
$\begingroup$ Ok. I still had some problems, I now think it was the surface normals that were not properly ordered. $\endgroup$ Commented Dec 25, 2019 at 14:30