How to fix the thicken mesh that twisted in the process of thickening

Hi everyone,

I have a minimal surface (output of kangaroo). When I use Weaverbird mesh thicken, or Weaverbird mesh offset, or even mesh offset components to make the initial minimal surface thicken a part of the thickened mesh has a problem as you can see in the image attached. please let me know how can I fix this. (the second image is the initial mush (Kangaroo output) to thicken)


Can you post your definition or at least a part of the result before the thickening?

Well let use my :crystal_ball:. You’ll have to fix the normal.

1 Like

New (2.7 MB)

Thank you. But the question is How :wink: ! Can you use your :crystal_ball: again?
Actually, I want to understand why the output mesh of Kangaroo has various normals and in this case, how can I simply fix the normals?
To avoid such a problem again, I need to understand what is wrong in my def.
Thanks again @laurent_delrieu

Those plugins… Good luck

1 Like

you can’t really fix the normal calculations. A normal is perpendicular to the surface. Your surface is distorted due to the way the ellipse are oriented.

One solution could be to replace all normal that are on naked edge with what you think is good. And offset the surface your self using deconstruct mesh and construct mesh.
For the side it is possible to make them with a loft mesh of the 2 polylines of Naked Edges.

Here is the way
Extract the naked edges

Orient the planes toward the exterior

I choose to offset (loose to keep the same number of points) then move up so it makes a 45° chamfer. But you could or not move the curve.

Then the idea is to deconstruct the mesh and replace normals by the vectors on naked edges

Then making the sides, very classical, don’t forget that closed polylines have an extra point (the end is the beginning)

mesj naked (908.3 KB)

And if you still have problem with normals, attractor could be used. For example
New “Normals” = Normals * (1-factor) + Vector * factor
Factor is 1 near the maked edges and decreasing after. It will no be as simple but seeing your definition you must understand the idea.

1 Like

I appreciate you @laurent_delrieu
I see that naked edges could work properly; however, as it is obvious in resulted mesh, the problem still remains. I have attached an image here to explain better.

this ellipse section has this weird problem while the other ones are working properly!
Indeed, the naked edges of the main mesh are at the same plane (x, y, z = 8) and as soon as the offset mesh is made, we see the naked edges of the offset mesh are at different planes (x, y, z = various). it prevents me from having a clean mesh. This messy mesh is not good for fabrication and even presentation!

The problem is there depending of the offset you choose, as you have very thin faces problem is there at quite low offset.
I think attractor could work. So you will have a better transition between vectors.
You could also smooth value with my tool to smooth values on vertex (Nautilus tool), with fixed values on boundary!!
No smoothing

5 steps

10 steps

mesj naked edges (910.6 KB)

It is the implementation of this


I appreciate you so much @laurent_delrieu
I was arranging the components to find out what’s going on. I need to check the code more and read more about what you told and suggested. If possible, I reach out to you in upcoming days to ask possible questions and to be sure that I got the points properly.

Thank you again for you time and help!

1 Like

You have the file in the answer.