Jump to content
ParticleSkull

Create multiple FEM over time?

Recommended Posts

But even here, you get to a point where it breaks. You can up the steps, and that will of course help, but still, there will always be some point where it breaks again.

45Cwgzs.gif

Share this post


Link to post
Share on other sites

Didn't have the chance to open the scene yet, but I'd like to check whether there's a way to control the maximum forces of the tetrahedal mesh inside of the spheres.

 

Share this post


Link to post
Share on other sites
4 minutes ago, Pancho said:

Didn't have the chance to open the scene yet, but I'd like to check whether there's a way to control the maximum forces of the tetrahedal mesh inside of the spheres.

 

No, doesn't seem like you can, or at least not easily - as in, messing about inside the DOP solid object- and finite element solver subnets. 

So grains is the more controllable choose here. 

Share this post


Link to post
Share on other sites

We should bring an employee of SFX into the discussion to get some decent explanations. Would be a pitty if this doesn't end with an "Ahhhhh!" moment. Not to be confused with "Aaargh!".

 

Share this post


Link to post
Share on other sites

Well, personally I don't work like this, or at least not in production, it's like, well, this doesn't work, and you move on, hehe, find what does... So stuff that doesn't work is a bit of a non issue, if you get why I mean. I don't allow myself to get stuck fighting windmills, so to speak. ;)

FEM to me is for FEM only stuff and if it starts to cause issues, I just move on, because as I said, it's a solver that doesn't play nice (or at all) with other solvers. And I've used grains a lot, it's awesome because they are POPs, and FLIP and Bullet are point based as well, so all the POP stuff us available as well as you can do anything with it in a SOP solver. :D

Oh, and though there are SESI staff running here, it's more likely to get them into s thread in the SESI forum than here. 

Share this post


Link to post
Share on other sites
12 hours ago, Pancho said:

Didn't have the chance to open the scene yet, but I'd like to check whether there's a way to control the maximum forces of the tetrahedal mesh inside of the spheres.

 

Yes! relatively easily. 

FEmAtt.hipnc

TetAtt.gif

  • Like 1

Share this post


Link to post
Share on other sites

Looked at your scene, but I'm not sure whether that is what I've meant. You are defining SHAPESTIFFNESS, but that's more or less just an attribute of the object, e.g. softer or harder (marshmallow vs rubber). As far as I can predict, this won't change the intercolission behaviour in the sim (the interpenetration). I expect the penetration to happen because of forces piling up in the tetrahedralized mesh to such an account, that the colission detection can't avoid interpenetration anymore. I thought there is a way to check the forces inside the mesh and if the force (which will be a vector, the force amount its length) is abouve a threshold then clamp it at this point. I expect the penetration to stop because of this. If you can't push harder than with a certain force, things shouldn't break.

P.S.: In the attributecreate node's group settings, where do the numbers come from (1-2 5-6 ....)? Did Houdini create them or you?

Share this post


Link to post
Share on other sites

@Pancho sounds like you want to plastic the object which the solver doesn't natively support. BTW a 1 dimension vector is a scalar :)

Share this post


Link to post
Share on other sites

No, I don't want to plastic the object. THat would totally change its behaviour look. I want to crop the forces above a level which might be the reason for the interpenetration. At one point there wouldn't be a difference if there are 4 spheres above the lowest sphere or twenty, since the force doesn't increase.

@pskull: Saw the settings the first time and enabled them in the FemAtt.hip. Seems to give a good indication of what's going on. I just guess that setting them manually won't change a thing. Looks more like they are telling you what forces are happening at the moment, or do I miss something?

Share this post


Link to post
Share on other sites

Maybe you can use this energy attribute to give more stiffness or drag to those "energized" areas. I think just a multiplication could work.

Can't make tests on ir right now but I'll give it a try asap

Share this post


Link to post
Share on other sites

Guess I don't know what 'crop the forces' mean in a Fem solver tbh. I can't think it through but if you can draw a free body diagram of it that would help. 

Share this post


Link to post
Share on other sites

O.k., changed the scene slightly to make it quicker in terms of intercolission.

Please set frame distance in the solid object back to 10 so that there are more eyes.

What I noticed is that potentialdensity is rising as soon as the spheres pile up. I guess that's the "stored" energy. The spring, which gets more and more under pressure and want's to release it energy into kinetic energy, movement.

While this might be right from a technical viewpoint, it seems to collide with what we expect to happen. Actually I would expect this material not to be a spring. While it might be able to store some energy, at a certain point there's no way to increase the potential energy. There would be a force in the sphere preventing the spring to get loaded anymore, but keep it at a certain threshold.

Sounds a bit funky, but hey, I hate interpenetrations and strange sim behaviour. : )

flask.v4.Odforce.FF.hiplc

Share this post


Link to post
Share on other sites
26 minutes ago, Pancho said:

While this might be right from a technical viewpoint, it seems to collide with what we expect to happen. Actually I would expect this material not to be a spring. While it might be able to store some energy, at a certain point there's no way to increase the potential energy. There would be a force in the sphere preventing the spring to get loaded anymore, but keep it at a certain threshold.

Yes - that is why the solver needs to support plasticity. The three main parts of mechanical response to a load is elasticity, plasticity and fracturing...

Share this post


Link to post
Share on other sites

Ahh!

So, what do we do about it? Use the grain solver? Is FEM the wrong approach for this kind of softbody sim?

Share this post


Link to post
Share on other sites

I just mailed SideFx. Hopefully they take a look at this thread. Would be very nice.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×