+- tyFlow Forum (https://forum.tyflow.com)
+-- Forum: tyFlow Discussion (https://forum.tyflow.com/forum-1.html)
+--- Forum: General Discussion (https://forum.tyflow.com/forum-2.html)
+--- Thread: cloth simulation: infinite movement (/thread-2791.html)
cloth simulation: infinite movement - redflash777 - 12-06-2021
Hi everyone!
I have met an issue I can't solve on my own. I try to simulate curtain, using CUDA solver, and it appears that during simulation the cloth desn't want to calm down. It looks like completely weightless, no matter of the gravity. It waves, it streches, it wrinkles, but never stay still, even on 2000 frame and more. How do I calm it down so it should just hang for a while without any movement? Help me please guys! Here are some of my settings. I have highlighed which settings I adjusted.
RE: cloth simulation: infinite movement - tyFlow - 12-06-2021
Can you post the file?
RE: cloth simulation: infinite movement - Espen - 12-06-2021
can you upload scene?
RE: cloth simulation: infinite movement - redflash777 - 12-06-2021
Yes, I have added zip file i the head post! Thank you!
RE: cloth simulation: infinite movement - Espen - 12-06-2021
Not sure if i understand, but if i set timing on the Force (turb1+turb2) from 0 - 50 frames it calms down after that and settle with some movement. You want the curtain to get wrinkled/curled up by the turbulence force and then come to a rest?
RE: cloth simulation: infinite movement - redflash777 - 12-06-2021
(12-06-2021, 04:54 PM)spen Wrote: Not sure if i understand, but if i set timing on the Force (turb1+turb2) from 0 - 50 frames it calms down after that and settle with some movement. You want the curtain to get wrinkled/curled up by the turbulence force and then come to a rest?
Yes friend, seems that this has solved my problem! Never could imagine that this infinite movement was caused by working perlin noise forse...
Thank you very much.
One more question before closing this thread, how to get rid of some unwanted particle sticking between each other?
As you can witness from example, a vertex from left curtain has stuck to the vertex on the opposite half.
RE: cloth simulation: infinite movement - tyFlow - 12-06-2021
Make sure the cloth has no self-intersections at frame 0...problems like that usually occur if the starting state of the mesh has intersections, which can therefore not be resolved.
RE: cloth simulation: infinite movement - redflash777 - 12-07-2021
(12-06-2021, 05:40 PM)tyFlow Wrote: Make sure the cloth has no self-intersections at frame 0...problems like that usually occur if the starting state of the mesh has intersections, which can therefore not be resolved.
I checked - none intersections at 0. Each simulation I get random particle sticking - this time two edges of the same cloth
RE: cloth simulation: infinite movement - tyFlow - 12-07-2021
Can you upload the scene with any modifications you made since your original post? Just want to make sure I'm examining the most recent version if I take a look at it...
RE: cloth simulation: infinite movement - redflash777 - 12-08-2021
(12-07-2021, 05:04 PM)tyFlow Wrote: Can you upload the scene with any modifications you made since your original post? Just want to make sure I'm examining the most recent version if I take a look at it...
No, it is the same scene I posted already. I had this problems in my every cloth simulation before too. Is it some particle binding issue? I already removed any friction, but still sometimes get this sim error.
P.S. vertex sticking sometimes occure even inside the same mesh, like in example attached.
RE: cloth simulation: infinite movement - tyFlow - 12-09-2021
Hmm that's odd...I simmed 1000 frames into your scene and didn't see any self-intersections occur. You can check the MAXScript listener while simming...it should report if it wasn't able to get rid of all intersections at the end of a time step.
Dropping the sim timestep is usually a good way to improve collision accuracy. Yours is at 1/2 frame. Try 1/4. When you lower sim timestep you can also lower bind solver iterations. Yours is at 150. At 1/4 timestep, you could try 75 to speed things up.
RE: cloth simulation: infinite movement - redflash777 - 12-09-2021
(12-09-2021, 06:53 AM)tyFlow Wrote: Hmm that's odd...I simmed 1000 frames into your scene and didn't see any self-intersections occur. You can check the MAXScript listener while simming...it should report if it wasn't able to get rid of all intersections at the end of a time step.
Dropping the sim timestep is usually a good way to improve collision accuracy. Yours is at 1/2 frame. Try 1/4. When you lower sim timestep you can also lower bind solver iterations. Yours is at 150. At 1/4 timestep, you could try 75 to speed things up.
Yes, dropping the sim step along with iterations seems to solve those issues.
Thank you very much for your help! Awesome plugin!