0

I used the tissue modifier to make this but the mesh needs to be thicker and bigger at the end points, like this:

enter image description here

enter image description here

4-K
  • 505
  • 2
  • 5
  • 20
  • https://blender.stackexchange.com/questions/58959/how-to-model-braided-nylon-sleeve – Duarte Farrajota Ramos Feb 10 '23 at 11:35
  • @DuarteFarrajotaRamos Did you look at my source image? Will try this and get back to you. – 4-K Feb 10 '23 at 16:24
  • Do you mean this sort of clumping/thickening? https://imgur.com/a/LWOlu6z My method for this is a bit scraggly atm, but could be simplified enough to be an answer.. – Robin Betts Feb 11 '23 at 17:34
  • @RobinBetts yes! much better – 4-K Feb 12 '23 at 18:46
  • @RobinBetts Please share the method if possible. – 4-K Feb 13 '23 at 10:30
  • @4-K coming up.... OK. But the method in my answer is flawed. Ratios are clamped by being on a vertex-group. That needs a fix, sampling 'Nearest' in GN on an attribute instead of relying on Tissue to pass the value in a vertex-group. But this is a start, and you're in a hurry. I'll fix the answer when I can, soon. – Robin Betts Feb 13 '23 at 11:19

1 Answers1

3

The hosiery on the bulb is thinner on the swollen area because it is stretched. One approach to simulating this would be start off with a base-tube, and stretch it ourselves:

enter image description here

This can be proportionally edited as a shape-key, or just on a duplicate.

Once made, the stretched version can be compared to the base version with a GN group, storing the ratio of lengths into a vertex-group:

enter image description here

(In the end, here, the attribute has to interpolated onto points, not edges, which isn't perfect. But that's why it's OK to stash onto a pre-existing vertex-group, which can be directly used by the Tissue tesselation)

A weight-map on the bulb results:

enter image description here

When you tesselate using the Tissue add-on, you can ask for the vertex-group to be mapped to the tesselation. This method tesselates a tileable mesh-line, without thickness:

enter image description here

Then uses a GN modifier on the tesselation to generate thickness, guided by the mapped vertex-group. Ideally, as a band is stretched, to preserve volume, the cross-sectional area varies as 1 / length, and its radius varies as the square root of that:

enter image description here

The result is subtle, but you could fudge the square-root to another Power to exaggerate it.

This is the sort of result:

enter image description here

I'm not that fond of the workflow.. it has destructive steps. You could do both the stretching and the tessellation inside GN to improve it, but that's another question.

Robin Betts
  • 76,260
  • 8
  • 77
  • 190