bigger collision splashes with Splash tank OceanFX tool

[Waasha]

Hello everzone!

I'd like to ask if there's a way to increase the splashes from collision (made by static RBD object) objects..

 

I have a setup which works (more or less :-) ) and I need the water to be traveling at this speed. But I'd like it to be splashing on the rocks and be more wild on the collisions.

 

Is there a way to increase the collision effect/add more splash particles on collision? I tried to play with the whitewater emitter settings (min/max curvature, speed, etc..), but it doesn't work..

 

Thanks for any advice!

splash_tank_test.hip

[JuriBryan]

well you can always go in and cheat some stuff...
what I try sometimes is to run a geometry vop in dops over my sim.
Then check against my collision volume and see how far away each point is.
that is done with a simple volume sample that reads your collision objects.
now that you have that you can run a fit node over the result and us that new gradient that describes the distance to your collision objects to multiply the velocity up.
that means that nothing in you sim changes, but that the particles closer to your collision objects move faster and there by splash faster. 
running a nother look up against velocity in +y to see when to kill the extra velocity is a good idea for this set up.
so that you don't have points flying into nothing but fall back down and join with the main sim.
for some extra detail in the splash you can always use a sop solver that deletes all points but your roplets.
then you point replicate your droplets and merge those new points back in with your main sim.
and boom more details in your splashes. which then means more violent collision splashes.
 

[Waasha]

Thanks JuriBryan, thanks to your advice I found a little different way to accomplish it. I am sourcing the collision with the fluid source node and I'm adding some velocity with curl noise (in a VopSop) to the collisionvel field. this way I can scale the collisionvel in the DOP and it works like a charm..

 

Thanks a lot! Your other ideas are also good-will give it a try..

 

Bye!

Edited March 31, 2014 by Waasha

[old school]

You may also want to look at the collision options in the FLIP Solver's Volume Motion > Collisions tab. You can boost the Velocity Scale here along with a few other options.

[Waasha]

Thanks for your sugestion old school, but the velocity scale in the collisions tab works only when the collision objects moves. with stacionary object (like a rock in the sea) it's velocity is zero, so scaling zero doesn't work..

 

I tried adding parameter v to the points of the stationary collision object and then scaling the collision velocity with this parameter, but it didn't work with static RBD collision. It started to work with collision sourced via the fluid source though. But this way I can scale the collisionvel directly in the source node and don't need to scale it in the collisions tab of the FLIP solver..

 

One more question - is there a way to increase the speed of the waves? I mean other than scaling the Time parameter in the wavetank initial? When I speed them up this way, the ocean spectrum goes faster, but the FLIP particles produce much less whitewater and collisions and generally this solution of speeding up the waves isn't ideal.. Any thoughts?

 

Thanks, bye!

[JuriBryan]

here is a quick file showing how to read the collision field in a geometry vop inside dops.
Harima asked for it.
volumeSampling_inDops.hip

[harima]

Thanks Juri!

[pupps09]

I am sourcing the collision with the fluid source node and I'm adding some velocity with curl noise (in a VopSop) to the collisionvel field. this way I can scale the collisionvel in the DOP and it works like a charm..

 

Hi Waasha. I am actually trying to do the exact same thing as you and I am stumped on how to go about the collision velocities. I just recently began using Houdini and I am unable to follow how JuriBryan is extracting the collision info.

Would greatly appreciate if you could go in to more details on how you are going about with your method. Uploading a sample .hip file would be even more appreciative.

Thx in advance. 

Edited October 6, 2015 by pupps09