[SOLVED] Is there a way to animate Vellum constraint properties?

[haveahapyday]

Greetings!

I'm working on a shot where I'm using a sphere to push a cloth through a hole in another object. Up to a point, the sphere acting as a collision object works great. However, the cloth slips off the sphere halfway through the sim. I was hoping I could animate an "Attach to Geometry" Vellum constraint to ensure the cloth sticks to the sphere after a certain period of time. I can't seem to figure out how to animate any constraint properties though. I'm doing it at the sop level using Houdini Apprentice 17.0.352, if that makes a difference.

I tried making a simple scene with a grid and sphere through vellum constraints. The idea was to let the grid fall to the ground and animate the stiffness value on an "Attach to Geometry" to pick it back off the ground, but it doesn't work. I also tried the "Attach Frame" setting on the "Attach to Geometry" constraint, but that doesn't seem to do the trick either.

Are Vellum properties set at the beginning of the sim and unable to be changed? Or is there some secret sauce I need to add? Do I need to do this in a DOP network or change values upstream in a wrangle?

I can't seem to find anything but beginner Vellum materials online at the moment. Thanks!

haveahapyday_vellum_animatedPropertiesTest.hipnc

haveahapyday_vellum_clothHole.hipnc

EDIT: Solved! Check Noobini's responses in this thread.

Edited November 15, 2018 by haveahapyday
Solved

[Noobini]

here's one way, in DOP. Don't be fooled into thinking RBD is working fine with Vellum...no...Vellum can't push RBD back, it's not a 2 way interaction, not that I know of.

Of course you can also do it in SOP, inside the Vellum solver...same idea, animate the Activation to switch on at whatever frame.

vu_vellumsticky.hiplc

[Noobini]

here's the SOP version, the odd thing about this is if you connect the ball to vellum cloth to act as a collider...it sticks as soon as it comes into contact...rather than 'wait' until a later time.

vu_vellumsticky_SOP.hiplc

[Noobini]

I'm stumped as to why I can't get the 1st version, DOP, to work now....the SOP one still works.

Hmm...

Still NFI, so I redid the same setup from scratch...annnnnnnd...it works.....here's the new DOP scene...maybe there's something dodgy in the DOP workflow.

test1.hiplc

test1.mp4

I would stay away from the DOP method for now, after I saved the working file, closed, reopened it, it no longer works....urrrggg....

Edited November 15, 2018 by Noobini

[haveahapyday]

17 hours ago, Noobini said:

here's the SOP version, the odd thing about this is if you connect the ball to vellum cloth to act as a collider...it sticks as soon as it comes into contact...rather than 'wait' until a later time.

vu_vellumsticky_SOP.hiplc

Yes! This is exactly what I needed! I saw you could dive into the SOP Vellum solver, but wasn't certain what you were supposed to do in there. It seems to me at the moment that the SOP level Vellum nodes are kind of like an all or nothing deal, but I could be wrong. In the future I think I'll go the DOP route, but right now I just want this shot done, haha.

I found a work around to prevent it from sticking immediately using a delete and switch SOP. If the collision geo input is empty until you need it, the attach to geo constraint won't trigger. Hooking up the switch to the geo attach activation parameter to make things a little less cumbersome.

vu_vellumsticky_SOP-haveahapydayEdit.hipnc

Edited November 15, 2018 by haveahapyday
better words

[Noobini]

ah ok good thinking.....but you could even simply stuck a null in where the 'delete' is, a null by itself that is with nothing connected to it...

The only minor point with this approach is someone might insist on the ball 'pushing' around the cloth at the start, they WANT to see this...then later on it sticks....so this null or delete approach would miss out on the interaction at the start.

[Noobini]

yes...done it....

new approach, based on this thread I just posted here:

 

So now with this approach, you can see clearly the rolling ball pushing around the cloth at the start...then at a later time, it sticks.

 

vu_vellumsticky_SOP2.hiplc

[Noobini]

Grasshopper is a Shaolin student, he needs to pass the test of walking on ricepaper so lightly that it doesn't stick to his feet....at frame 72 he fails miserably....

 

 

vu_shaolintest.hiplc

[haveahapyday]

17 hours ago, Noobini said:

ah ok good thinking.....but you could even simply stuck a null in where the 'delete' is, a null by itself that is with nothing connected to it...

The only minor point with this approach is someone might insist on the ball 'pushing' around the cloth at the start, they WANT to see this...then later on it sticks....so this null or delete approach would miss out on the interaction at the start.

I did use a null initially which is probably faster, but I thought it might make a little more sense if I had to go back to it later. Six one way, half dozen the other? Or is there more to it?

15 hours ago, Noobini said:

So now with this approach, you can see clearly the rolling ball pushing around the cloth at the start...then at a later time, it sticks.

vu_vellumsticky_SOP2.hiplc

oh wow, i love the little detail created when the ball is rolling but not sticking. if i understand this correctly, this version works because both the floor and sphere are Vellum cloth opposed to the sphere being an RBD/whatever Vellum uses for default collisions?

haha that shaolin test is hilarious. great implementation. thanks for sharing.

Edited November 16, 2018 by haveahapyday
added question.

[Noobini]

yeah if you look at the docs/launch video/sneak peeks...etc....etc...they never mention Vellum with RBD in the same sentence.

Basically, Vellum uses a Position Based Dynamics approach to cloth, hair, and grains

Yet ppl keep complaining, I can't get RBD to work with Vellum...well read the above statement. So for now, simple RBD effect has to be faked with (another cloth object) very stiff cloth settings. I'm sure SESI will find ways to get Vellum and RBD to play nice together.....soon.

[haveahapyday]

Thanks for pointing that out, I don't think I would have picked that up. I'm just glad they put this sick solver in, I can wait for further improvements.

[grasshopper]

Quote

 

Grasshopper is a Shaolin student, he needs to pass the test of walking on ricepaper so lightly that it doesn't stick to his feet....at frame 72 he fails miserably....

vu_shaolintest.hiplc

 

 

Well seeing as Noobini is challenging me one answer is to set "Detach Point Chance" to "$F>70" on the Vellum Constraints node inside the Vellum Solver.

Edited January 29, 2019 by grasshopper