USER FORUM
(you are viewing a thread; or go back to list of threads)
naked edges in a trivial double extrusion (by Alex B)
Guys,
Trying to make this model ran into a problem. Whether I add material from one side or the other, I'm getting red highlighting of naked edges. Any ideas ? Thanks!
Trying to make this model ran into a problem. Whether I add material from one side or the other, I'm getting red highlighting of naked edges. Any ideas ? Thanks!
(no subject) (by Andrew)
Curved surfaces are a often cause trouble in Solvespace. However I can give you a workaround. Make your first extrusion slightly longer, and create the second extrusion so that it overlaps the first extrusion, making it the right length. This offset can be a very small amount, like 0.01, as in the attached sketch.
(no subject) (by Alex B)
I see, thank you. I've ran into similar problems with more complex designs before, but never with one this simple. Really seems like a bug.
(no subject) (by Andrew)
I looked at that too early in the morning, and forgot the simple answer to that sort of problem, and that is use the the
force NURBS surfaces to triangle mesh option on the extrusion
that causes problems. The NURBS library used is known to have bugs that cause the problem you saw.
force NURBS surfaces to triangle mesh option on the extrusion
that causes problems. The NURBS library used is known to have bugs that cause the problem you saw.
Nice exampleof NURBS failure - thanks (by ruevs)
Hi,
Thank you for the nice simple example of NURBS failures. The NURBS engine does not handle many corner cases well.
However case you are creating is "nasty" for any engine. It has "singularities" see the links in my comments here:
https://github.com/solvespace/solvespace/issues/470
Apart from the solutions already proposed (having the first extrusion go into the second one is the better one) you can also change the boolean operation for the second one from "union" to "assemble" in this case.
Thank you for the nice simple example of NURBS failures. The NURBS engine does not handle many corner cases well.
However case you are creating is "nasty" for any engine. It has "singularities" see the links in my comments here:
https://github.com/solvespace/solvespace/issues/470
Apart from the solutions already proposed (having the first extrusion go into the second one is the better one) you can also change the boolean operation for the second one from "union" to "assemble" in this case.
(no subject) (by Alex B)
Interesting... what are the implications of Assemble vs Union in such example ?
(no subject) (by ruevs)
Hard to say. Depends on what you are going to be using the model for.
What happens in the code is that all edges (and faces) from both extrusions are kept in the resulting model. In other words there is a "zero thickness" face (two in fact) "splitting" the model where the extrusions meet. Strictly speaking this is not a clean "one manifold" solid but for most cases it may be OK.
What happens in the code is that all edges (and faces) from both extrusions are kept in the resulting model. In other words there is a "zero thickness" face (two in fact) "splitting" the model where the extrusions meet. Strictly speaking this is not a clean "one manifold" solid but for most cases it may be OK.
(no subject) (by Alex B)
This is to be 3D-printed. I guess it then depends on what the slicer makes out of it.
naked edges in a trivial double extrusion (by Salvatore Fazia)
Hi Alex, I think if you make the second working plane the back of the first extrusion it will be fine
Post a reply to this comment: