Damon Lavenski

Here's the full response with that expanded into the opening paragraph:

Thank you. I tried using Sleep, but the stop/sleep attribute halts the animation quite abruptly. On top of that, waking the points back up isn't straightforward: the moment a point reactivates, gravity immediately starts pulling on it again, so instead of the strips looking like they're resting and waiting to be disturbed, they appear to melt. If I woke up only the points near an interacting object, for example, tongs picking up and dropping the pasta — the rest of the strip would stay asleep, and pulling on just the nearest points would cause stretching rather than the whole strip responding naturally. Waking the entire strip at once solves the stretching, but then reintroduces the melting problem, since the whole thing starts falling under gravity again instead of just reacting to the collision.

I experimented with different wrangle-based methods to soften that transition, but what ultimately worked was animating gravity so it wasn't pulling as hard on the points. I also used POP Drag along with some of the velocity-damping attributes within the cloth constraints, and tried a pin-to-target constraint with its strength set to 0, with velocity damping enabled and set to a low value (around 0.25, that helped somewhat.

One other significant thing I initially left out: I ran a separate simulation on the strips beforehand to get them into a rest position. After that sim, I had to delete the velocity attribute before using it as the starting point for the actual simulation — otherwise that residual velocity would carry over, and the new sim wouldn't be starting from a clean, zero-velocity state.

After caching out the simulation and bringing it into CHOPs, I used blend shapes to smooth out the motion: I took the last frame that still looked good from the time-shifted sim, froze it, and used a blend shape to gradually blend the simulated pasta back toward that frozen frame. That brought the motion to a slow stop instead of an abrupt one. I then applied a filter to remove any remaining noise for the final cleanup.

Since I come from a Maya background, I can only compare this to my experience there. In Maya, there's an option for the solver to check whether points are intersecting, for both collision objects and self-collision. It's fairly calculation-heavy, but it cleans up this kind of behavior quite well.

Another factor I found significant was scale. On this project, simulating my pasta bowl, I was working in centimeters, using real-world measurements: the pasta itself is about 3 cm long. Since I was working with flat strips rather than round spaghetti, the pasta behaved more like sheets of paper. When it hit the bowl, the solver had trouble damping the velocity down. I tried increasing substeps and collision passes, but that only helped so much. In the end, adjusting gravity, using a rest-position pass with velocity cleared, blending time-shifted frames, and cleanup in CHOPs together solved the problem.

My takeaway from all this: coming from Maya, this workaround felt overly complicated for what it was solving. I'm used to Maya having its own quirks and needing workarounds, but usually not ones this involved for an issue like this. Normally you'd expect undulation like this to come from intersections, but in this case there weren't any, it was purely residual velocity from over-calculating collisions. It makes me wonder: how would you stack papers on top of one another without them dancing or bubbling like this? I know there's a big difference between quads and triangulated geometry, and that may be part of it, but still, I'm new to Vellum and still learning its limitations and quirks. I understand it wasn't really designed for a setup like this, but there should still be a simpler way to approach it. If anyone has a good approach for setting up pasta strips in a bowl, with tongs picking them up and dropping them while they keep their shape, I'd love to hear how you set it up. Excited to keep learning, thanks for all the help! :-)

Hello everyone,

I'm currently working on a simulation of a bowl of pasta, and the main issue I'm running into with Vellum is that the simulation will not settle down. It continues to bubble, simmer, and jitter with every interaction.

I've tried several things to calm the simulation down. For example, after getting the pasta noodles to settle into the desired shape, I used a Time Shift to grab a specific frame so I wouldn't have to deal with all the jittering and bubbling caused by the excess velocity.

As another test, I created a stack of paper that falls into a tree to see if I could reproduce the same jittering and bubbling behavior. I know the cause of this behavior is most likely related to the solver settings or constraint conditions.

I also tried using a Pin to Target Constraint, setting the stiffness to zero, and enabling Velocity Blend. I also increased the damping a bit, but unfortunately, this doesn't seem to have made any difference.

I've been a Maya user for many years and have used nDynamics for most of my career, and I've never encountered this type of behavior to this extent. Obviously, you can get some bubbling and jittering in nDynamics as well, but not nearly as much as I'm seeing here.

I'm fairly new to Houdini, so I'm hoping someone on the forum might be able to point me in the right direction and help me figure out what is causing this behavior.

I've also tried layering the pieces and assigning IDs to every individual piece of paper so that I could use the layering attributes, but I haven't had much success with that either.

Thank you in advance for any help!

thank you freshbaked for the help, I added a color SOP between my geometry and the attribute transfer SOP. set the color to black and that solved the issue.

I spent an hour with Gemini and check TBT if they could give me the correct answer. which you did so eloquently so thank you.