Jump to content

Recommended Posts

Hello everyone.

I was going to try Wire Solver for some tests similar to the ones that Ben Watts recently posted: https://www.instagram.com/p/Br1uOBqH8wm/

He said that he used Wire Solver for this and the result is amazing.
So I started with a simple test: just a line with around 100 subdivisions which should fall to the floor. I used Wire Solver for that but can't get any remotely good results. The rope is self-intersecting and behaving strange. I used the shelf tool to create the setup and changed only the Elasticity parameters in the Material settings for the Wire Object. Set the Angular Spring Constraint to 0 and Damping to 0.001 and ticked on the Self Collide setting as they recommend in the Houdini help for wires or rope. I tried increasing the Substeps in the Wire Solver and in the DOP Network itself without any luck. The rope is still self intersecting.

I also watched the M04 | Next Steps | Procedural Animation | Wire Solvers masterclass. But found nothing new ther. And of course I read help of the Houdini where they suggest increasing Substeps to get better results if there are self-intersections.

Any help is greatly appreciated. Are there any resources to lean how to achieve the results similar to what Ben has in his test? Or any good tutorials on using Wire Solver for simulating ropes. I can't seem to find anything in Google.

Edited by shishkovdima

Share this post


Link to post
Share on other sites

Hello

It would be more easily to get help if you attached your hip with setup. This is an example file and i hope it helps you.

wires_fall.hipnc

 

test1.gif.4eaa3ffc3bbed50061c6d4c0c94ac929.gif

Edited by vicvvsh
  • Like 1

Share this post


Link to post
Share on other sites
6 hours ago, vicvvsh said:

Hello

It would be more easily to get help if you attached your hip with setup. This is an example file and i hope it helps you.

wires_fall.hipnc

 

test1.gif.4eaa3ffc3bbed50061c6d4c0c94ac929.gif

Thank you very much for the example file. I tried to modify Wire Object and Solver parameters in my test scene after taking a look at your file but still getting pretty weird results. I am attaching the HIP along with the GIF preview. I would be very grateful if you could take a look. Thank you in advance.

 

2019-01-12_18-16-42.gif

rope_test_01.hipnc

Share this post


Link to post
Share on other sites

As in your case one wire object I've only changed Collision Handling parameter in solver to Local Geometric and decreased Linear Damping Constant, set Angular Spring Constant to 0.01 in wireobject and that's it. You should increase substeps in DOP Network node to get rid of vibration.

rope_test_01_fix.hipnc

  • Like 1

Share this post


Link to post
Share on other sites

If this is a purposeful exercise for Wire solver, then ignore this post

but if you want to get nice, fast and robust rope sim, you should have a look into Vellum Solver

Wire solver is still useful for very stiff wires, but for rest of the softer "wire" sims like ropes, threads, strings, hair, ... Vellum may be much better choice

  • Like 1

Share this post


Link to post
Share on other sites
3 hours ago, vicvvsh said:

As in your case one wire object I've only changed Collision Handling parameter in solver to Local Geometric and decreased Linear Damping Constant, set Angular Spring Constant to 0.01 in wireobject and that's it. You should increase substeps in DOP Network node to get rid of vibration.

rope_test_01_fix.hipnc

Thank you very much for the fix. will be doing some more tests. Could you please explain 2 things to me:

1. Collision types. From Houdini help it is not clear to me what are the differences between 3 types. The only thing I understood is that Global Geometric should be more precise. Do you know what the differences are and what type is best for what case?

2. When should I increase Substeps in Wire Solver vs Substeps in DOP Network?

Again thank you for your time and help.

Share this post


Link to post
Share on other sites
2 hours ago, anim said:

If this is a purposeful exercise for Wire solver, then ignore this post

but if you want to get nice, fast and robust rope sim, you should have a look into Vellum Solver

Wire solver is still useful for very stiff wires, but for rest of the softer "wire" sims like ropes, threads, strings, hair, ... Vellum may be much better choice

I am going to try Vellum for this thing definitely. I started with Wire, because Ben Watts said that he tried Vellum and it did not work for him and he went with the Wire Solver for his test. As I want to try and achieve something similar, I decided to try create it with Wire Solver too. But for now even simple test with falling rope does not work as I would expect. Maybe Wire Solver is not the best choice for soft wire/rope but in Ben Watts' and Ben Koppel's tests the wires look pretty soft and Wire Solver produces beautiful result. Though their test are centered more on wire to wire collisions. This is something I want to try next. But just to get my hands dirty with Wire Solver I decided to create a simpler as I thought example with a rope falling to the ground and it is not as simple as I expected.

Thank you for the input. Appreciate it a lot.

Share this post


Link to post
Share on other sites
11 hours ago, shishkovdima said:

Thank you very much for the fix. will be doing some more tests. Could you please explain 2 things to me:

1. Collision types. From Houdini help it is not clear to me what are the differences between 3 types. The only thing I understood is that Global Geometric should be more precise. Do you know what the differences are and what type is best for what case?

2. When should I increase Substeps in Wire Solver vs Substeps in DOP Network?

Again thank you for your time and help.

I can't add more then help says unfortunately. I usually do plenty of tests before i find look i like. Substeps in solver influences on calculations that solver do in each simulation frame whereas substeps in DOP Network node influences on all simulation. i.e. if number of substeps in DOP Network is 3 we get 3 simulation frames in DOP per 1 frame. I usually increase substeps in solver and if still some artifacts i increase substeps in DOP Network node.

  • Like 1

Share this post


Link to post
Share on other sites
1 hour ago, vicvvsh said:

I can't add more then help says unfortunately. I usually do plenty of tests before i find look i like. Substeps in solver influences on calculations that solver do in each simulation frame whereas substeps in DOP Network node influences on all simulation. i.e. if number of substeps in DOP Network is 3 we get 3 simulation frames in DOP per 1 frame. I usually increase substeps in solver and if still some artifacts i increase substeps in DOP Network node.

Will be playing with Wire Solver more. Will make some test with "wire to wire" collisions later today. Thank you very much for your tips and help with scene setups. Helped me a lot. Greatly appreciated!

Share this post


Link to post
Share on other sites

Hard to say witch one is best, both needs some fiddeling. Wire solver isnt that good to interact with other solvers, same goes with vellum.

Heres a basic setup comparing them.

wire_vs_vellum_01.hip

Vellum is blue and wire is red

wire_vs_vellum.gif

Edited by ThomasPara
added gif
  • Like 3

Share this post


Link to post
Share on other sites
4 hours ago, ThomasPara said:

Hard to say witch one is best, both needs some fiddeling. Wire solver isnt that good to interact with other solvers, same goes with vellum.

Heres a basic setup comparing them.

wire_vs_vellum_01.hip

Vellum is blue and wire is red

wire_vs_vellum.gif

This is gold. Thank you very much for this comparison file. I will study it tomorrow.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×