Hello Houdini community,

I'm currently working on a project where I need to simulate a pillow falling using the Vellum cloth simulation. The challenging part is that I need to achieve a very precise control over the end result.

To give you more context, I have an approved layout where the locations of all pillows are fixed. I need the simulation to adhere closely to this predetermined layout while still maintaining a realistic look and feel.

So far, I have tried using the 'pin to target' option in Vellum Constraint to achieve this, but the results have not met my expectations. I'm struggling to control the end result with the precision that I need.

I'm reaching out to you in hopes that someone might have encountered a similar problem or could share some insights or techniques on how to improve this simulation. Any advice on controlling the end result of a falling pillow simulation in this context would be immensely appreciated.

Thank you in advance for your time and assistance.

I'm struggling to control the end result with the precision that I need.

Hey,
what exactly is the issue with pin to target? Have you tried animating the constraints? So one idea could be to start with a super loose pin constraint and straighten it during the end of the sim to force it to the desired endframe.

I've had a similar problem some time ago where I needed to bring the vellum sim back to the initial state and I found this VEX code:


You need to dive into the solver, add a geometry wrangle and put the initial state SOP into the second input.

It tries to match the initial shape 100% but maybe if you animate the strength and keep the blend very low, it can bring you closer.

Good luck!

Xue_Yue

I'm struggling to control the end result with the precision that I need.

Hey,
what exactly is the issue with pin to target? Have you tried animating the constraints? So one idea could be to start with a super loose pin constraint and straighten it during the end of the sim to force it to the desired endframe.

Thank you for your suggestion.

To clarify the issue I've been having with 'pin to target': The simulation doesn't fully follow the target. Instead, it starts to jiggle around near the starting position and doesn't move towards the intended location with the precision I need.

I like your idea of starting with a looser pin constraint and gradually tightening it. I'll definitely try it out. But considering the jiggling issue I mentioned, do you think this method will still work?

And if you have any more details on how to animate the constraints effectively, I would really appreciate it.

Thanks again for your help!

Glad if it helps you. Most of the time you can get rid of the jiggle in vellum by increasing the damping ratio or via a simple pop drag in your dynamics network.

Hey, this is a small setup i hacked together. I didn't think of the issues that occur while having a high damping from the start. So I decided to animate the strength of the constraint and the damping ratio in the vellum solver. It works, but I think the best way to get a nice transition might be to use something like this blend setup in the sim, so that you get close to your target geo, and then do a smooth position blend after the sim. Hope it helps a bit.

Sim in reverse?

Xue_Yue
Hey, this is a small setup i hacked together. I didn't think of the issues that occur while having a high damping from the start. So I decided to animate the strength of the constraint and the damping ratio in the vellum solver. It works, but I think the best way to get a nice transition might be to use something like this blend setup in the sim, so that you get close to your target geo, and then do a smooth position blend after the sim. Hope it helps a bit.

Thank you so much for your efforts and for sharing the setup you've created. It's great to see your approach on this issue.

The concept of animating the strength of the constraint and the damping ratio in the vellum solver sounds promising. It could be the solution I need for achieving the precision in the final layout.

I also find the idea of using a blend setup within the simulation and then smoothly transitioning the position after the simulation quite interesting. I'll definitely give that a try.

I'll take a look at the project file and start playing around with the techniques you've suggested. I'll report back on how it works out.

Your help is greatly appreciated. Thanks again!

eaniix
I've had a similar problem some time ago where I needed to bring the vellum sim back to the initial state and I found this VEX code:

v@force = normalize(point(1,'P',@ptnum) - v@P) * ch('strength');
v@P = lerp( v@P, point(1,'P',@ptnum), ch('blend') );

You need to dive into the solver, add a geometry wrangle and put the initial state SOP into the second input.

It tries to match the initial shape 100% but maybe if you animate the strength and keep the blend very low, it can bring you closer.

Good luck!

Thank you for sharing your experience and the VEX code. The idea of using a geometry wrangle within the solver and animating the strength and blend value sounds promising.

I'm eager to test your method along with the other suggestions. I appreciate your advice and will update the thread with my progress. Thanks again!