Just wanted to provide an update on the state of things.

I'm currently working with an updated mesh that's behaving much better with Vellum.

I'm currently reworking how the buttons work, and improving them over the initial test.

In the initial test, it didn't really feel like a volume was holding the shirt together, but just points on the shirt's mesh, and I wanted to solve that.

UPDATE

I've solved the issue, the new button setup pops very similar to the old one, not quite exactly but with a bit of tuning, this will be much better.

What I'm currently doing is now I'm simulating a collision mesh for the cloth simulation to solve some cloth sim clipping issues.

This is basically just a basic simulation on a modified version of Carol's mesh, and simulating problematic parts and prevent clipping and keep parts of the mesh from intersecting, like the armpits, breasts on the belly, and the part where the thighs meet the belly. This will just make sure the mesh doesn't have those problematic intersections where the cloth sim will start to clip at.

Initial results look good, and even the unpolished collider mesh has solved some major issues, most of the tuning now is thickening the model slightly so that it sits over the original better, and when my next set of iterations are done, I'll record a video and upload it here!

So after a bit more time, I found a superior method that works slightly different to the method I used before, and I wanted to talk about it a bit.

I was using a similar method to how I stitch the panels of the dress together, a weld constraint. This weld constraint allows me to "weld" the points together in post, and they basically act like one vertex during simulation until they break.This worked okay for the buttons when it really was just one vertex holding them together, but attempting to add additional vertexes to add "mass" to the pinning, it welded all of the vertexes in close proximity into a single point, and obviously that's not what I wanted.

I found an alternative method to stitch the points together, and one I was originally avoiding because of my issues with it in the past: the stitch constraint.At first, it sounds perfect, but I always had issues getting it to work properly, and after a fair bit of experimentation, I found a solution to get it working.

However, I've run into another small hiccup with how the stitch break constraints work, and they work a bit differently to how the weld break constraints work.

To update the break threshold in the welds was simple, I just had to tell the solver to update the threshold to match an animated attribute that I created before it reaches the solver.Normally, Houdini and Vellum just take the initial starting frame attribute, and run with that, and doesn't allow you to update those attributes with animated attributes, unless you tell it to within the solver with something called a SOP Solver, and a bit of code.

I solved this for the weld constraint as it was pretty straight forward.

The problem now is that the stitch constraint works in a slightly different way.A vellum node usually has 3 outputs: Geometry, Geometry Constraints, and Collision.The weld's break threshold is contained within the regular Geometry output, and can be referred to easily.

Accessing the constraint geometry requires a slightly different setup, and conversion from point attributes into primitives, which is what the constraint geo is made up of, as opposed to the regular geometry which uses point attributes.

This is where I'm currently at.I need to convert the point attributes into primitive attributes that I can then use to break the constraints that hold the buttons together.

I haven't solved it yet but I'm really close, and I've also updated my simulation geo, and made some improvements to fix some clipping issues I had with previous simulations.

When I solve the stitch break constraints, this simulation should be perfect.I really cannot wait to get this into the project. I already know I can basically just take it straight from Houdini into Blender after some touch ups after the simulation, and even the initial tests made me feel like I'm really headed in the right direction with this stuff.

I'm working on this now because in the future, with different types of outfits and such, this information and experience will be vital to getting them all up and running in a process I'm familiar with, and can refer back to with this project.And I want to share my work and experience with this in hopes that maybe someone will benefit from this information.

I just wanted to give a quick dive into what typically goes into this experimentation, so you can get an understanding, or somewhat of an understanding, of the problems I'm currently working on solving.

I just know that it'll be so worth it, because I've seen cloth tearing done in lots of ways in other various animations, whether it be small meshes to simulate the cloth tearing like flesh bulging, or even some simulations that aren't too far off of how this breaking works, but doesn't quite have the stretch, stress, and direction that this process can provide (and the collider distance because I really don't like my cloth simulations being a large, measurable distance away from the base mesh).

Thanks for coming to my ramble talk :)