18 replies to this topic

[moreffects]

I'm trying chocolate caramel simulation where I'm trying  chocolate getting break into pieces and while breaking caramel inside started getting stretch and wrinkled like cloth shape.I have tried using SPH particle but only thing i lacking is the wrinkling please help me ,if anybody know how to approach that.

I'm attaching image so you can get the idea

Attached Files

•  cho.jpg   55.04K   35 downloads

[hopbin9]

As far as I'm aware Houdini can not do thick viscosity effects with fluids. I spent the better part of a few months trying to simulate chocolate and caramel without any success. You're going to have to come up with some kind of hack or other solution to fake it. I've not been able to reproduce anything close to the viscosity effects you can do in Realflow.

There is a stack error problem in the way Houdini calculates viscosity that the fluid will collapse onto itself, it'll cause bouncing, errors and separation of the particle streams. I've had better results with static volumns, but once you start using fluid emitters the solver will have problems with any high viscosity values.

That is why the only example SideFX has posted with high viscosity uses a static volume, and it looks more like jello then fluid viscosity.

[moondeer]

i wonder if the cloth engine would do the trick here. it has stretch and break that would approximate the motion.

then, coat the cloth with a bit of FLIP fluid set to viscous.

the just surface the whole lot and shade to with carmel.

[Erik_JE]

hopbin9, on 23 July 2012 - 02:24 PM, said:

As far as I'm aware Houdini can not do thick viscosity effects with fluids. I spent the better part of a few months trying to simulate chocolate and caramel without any success. You're going to have to come up with some kind of hack or other solution to fake it. I've not been able to reproduce anything close to the viscosity effects you can do in Realflow.

There is a stack error problem in the way Houdini calculates viscosity that the fluid will collapse onto itself, it'll cause bouncing, errors and separation of the particle streams. I've had better results with static volumns, but once you start using fluid emitters the solver will have problems with any high viscosity values.

That is why the only example SideFX has posted with high viscosity uses a static volume, and it looks more like jello then fluid viscosity.

https://vimeo.com/42285647

There are some people on vimeo that have had luck with H12 Flip and High Viscosity. Have not had time to play with it too much myself tho.

[moreffects]

moondeer, on 23 July 2012 - 03:57 PM, said:

i wonder if the cloth engine would do the trick here. it has stretch and break that would approximate the motion.

then, coat the cloth with a bit of FLIP fluid set to viscous.

the just surface the whole lot and shade to with carmel.

Thx replying Eric , I have tried that and giving me decent result but second part i didnt get it "coat the cloth with a bit of FLIP fluidsetto vicous" . can u explain me that a bit more .

[Pazuzu]

hopbin9, on 23 July 2012 - 02:24 PM, said:

As far as I'm aware Houdini can not do thick viscosity effects with fluids. I spent the better part of a few months trying to simulate chocolate and caramel without any success. You're going to have to come up with some kind of hack or other solution to fake it. I've not been able to reproduce anything close to the viscosity effects you can do in Realflow.

There is a stack error problem in the way Houdini calculates viscosity that the fluid will collapse onto itself, it'll cause bouncing, errors and separation of the particle streams. I've had better results with static volumns, but once you start using fluid emitters the solver will have problems with any high viscosity values.

That is why the only example SideFX has posted with high viscosity uses a static volume, and it looks more like jello then fluid viscosity.

I think the H12 solver is very stable and fast with very high viscosity values in comparison with RealFlow; I have try it even with an ice cream simulation (multiple stack fluid) with very nice results.
I'll post my results very soon in my vimeo channel.

https://vimeo.com/lordpazuzu/videos

Edited by Pazuzu, 24 July 2012 - 09:51 AM.

[moreffects]

Pazuzu, on 24 July 2012 - 09:48 AM, said:

I think the H12 solver is very stable and fast with very high viscosity values in comparison with RealFlow; I have try it even with an ice cream simulation (multiple stack fluid) with very nice results.
I'll post my results very soon in my vimeo channel.

https://vimeo.com/lordpazuzu/videos

I dont have question whthr houdini is better for particle fluid simulation or not, my problem is nothing but to get look what posted as image

[hopbin9]

Pazuzu, on 24 July 2012 - 09:48 AM, said:

I think the H12 solver is very stable and fast with very high viscosity values in comparison with RealFlow; I have try it even with an ice cream simulation (multiple stack fluid) with very nice results.
I'll post my results very soon in my vimeo channel.

https://vimeo.com/lordpazuzu/videos

Already "liked" your vimeo post

Your results so far, have far exceeded anything I could come close to in H11. I still have a few of my old "failed" H11 scene files, but I'm not sure what happens if you open them in H12.

For the caramel stream, is the emitter swinging back and forth to generate the folding of the fluid, or is the folding the result of viscosity feedback in the stream? That was something I was never able to achieve.

[hopbin9]

moreffects, on 24 July 2012 - 09:59 AM, said:

I dont have question whthr houdini is better for particle fluid simulation or not, my problem is nothing but to get look what posted as image

Yes, but getting the "look" is technically very difficult. Especially if the caramel is to be animated.

[Pazuzu]

This is a fast approach for this type effect, play with the divergence attribute to initialize the tear zones, and with the viscosity.

Attached Files

•  tear_fluid.hip   1.72MB   102 downloads

[Pazuzu]

hopbin9, on 24 July 2012 - 11:23 AM, said:

Already "liked" your vimeo post

Your results so far, have far exceeded anything I could come close to in H11. I still have a few of my old "failed" H11 scene files, but I'm not sure what happens if you open them in H12.

For the caramel stream, is the emitter swinging back and forth to generate the folding of the fluid, or is the folding the result of viscosity feedback in the stream? That was something I was never able to achieve.

Thank you!!

The emitter is moving back and forth; But right now I'm working with a sim of ice cream that needs feedback of viscosity to have those yummy folds . It's possible to do that remapping the velocity distribution in the shape of the emitter, for example if is a star shape emitter you can have high values towards the spikes and low at the center with a ramp for example, that way is like faking pressure force by collision.

[hopbin9]

Pazuzu, on 24 July 2012 - 11:46 AM, said:

The emitter is moving back and forth; But right now I'm working with a sim of ice cream that needs feedback of viscosity to have those yummy folds . It's possible to do that remapping the velocity distribution in the shape of the emitter, for example if is a star shape emitter you can have high values towards the spikes and low at the center with a ramp for example, that way is like faking pressure force by collision.

Wow, that sounds awesome. I haven't looked at fluids in H12 but I wish I had this a year ago.