Hello,

I am trying to simulate a geometry in vellum with cloth constraints.

When I use multiple simple grids with fairly high subdivisions, I manage to get good results. However, when I try to simulate the same with gigh polycount geometry (an array of cylinders) I end up having very long cache times. So it has been very hard to update the vellum simulation (forces like wind etc.)

Is there a way to simulate with a grid and change these after the simulation to more complex geometry?

I tried to use point deform, however the results are quite bad.
I looked in to vellum rest blend but I believe it is not that helpful, since I need to input same amount of polycount at the begining.

Am I doing something wrong? How can I be more efficient?

Thanks in advance
Omer

1. Instead of cylinders use curves, sim as hair and sweep afterwards, that should be fast enough

2. For point deform way
- your inputs 2 and 3 on point deform are swapped

- also make sure you create unique integer or string attribute on scattered points so that both grid and cylinders inherit it and you can use as Piece attribute in Pointdeform to avoid cross-capturing among separate grids

- And you can still use curves to pointdeform to grid and sweep afterwards to avoid distortions on tube crossection shape

Hello,

Thank you very much for your help.

This was very helpful in general and yes, I didn't realize I was mixing the inputs at pathdeform...


I tried to simulate both with hairs and curves and with grid and then used pathdeform. When I simulated with hairs, of course I suddenly had a lot of flying single 'pipes' and it didn't give me the result I aimed.

So I used surfaces to simulate, I used pointdeform to map bunch of splines on these surfaces and I sweeped them.

The path deform didn't work but I followed Entagma tutorial recreating the pathdeform with pointwranlges.

https://www.youtube.com/watch?v=Qj4qX7TBKZc&ab_channel=Entagma [www.youtube.com]


Again, thank you very much for your reply.

You'd have to constrain individual curves together either using glue or stitch constraints if you wanted them to stay together
The advantage is that you have freedom in how you constrain them so still can get nice splay ends or mid parts like in your image where the tubes are likely held together by intersections which may also be part of the slowness

Point Deform (not Path Deform) should work fine to Deform hi res geo with low res