Are mutual affectors between flip fluids working?

[pclaes]

Hi,

I've been trying to get mutual affectors between two flip fluids, but I guess I am missing something as I can't get it to work .

I checked the affector matrix in the details spreadsheet for the sim ( http://www.sidefx.com/docs/houdini11.0/dynamics/relationships#affectormatrix ) , and it shows that they are mutual affecting but with different solvers... But when you look at it, one flip sim is dominant and it is not mutual at all.

I am expecting the following behaviour but with particle fluids:

http://www.afterworks.com/FumeFX/text/vdSolver.wmv

I've set up another dopnet where both flip fluids are emitted from the same source and I define their relationships myself, one of the fluids has the density set to 100, which should make it lighter so it would float on top of the other (this is explained in the buoyancy example with the floating ball... the ball is an rbd though and I think it might be different between flip fluids).

I increase the feedback on the autodopnet to 10 to help the mutual to come to a solution, but that does not seem to help.

Has anybody got this working? Have a look at the attached scene file.

fluids_interacting_02.hip

[Macha]

Peter, just to clarify, you want two different flip solvers, right? One solver with two fluids is not what you're after?

[Macha]

If you make them 1 fluid object (but with different solvers and sources) you can make them interact correctly.

fluids_interacting_me.hip

[bunker]

I tried to separate the 2 flipfluid objects data by renaming one of them to "Geometry2".

Unfortunately, everything in the FlipFluid Solver is referring to "Geometry" only.

So I would say that everything has to be done in one solver.

[pclaes]

Thanks for the advice, and thanks for the file. It is good to know you can use different solvers with flipfluids. Now all I need is to get it to work with two flipfluidobjects.

Yesterday I also reverted back to the one solver and made some bubbles:

I separate them out with a sopsolver so it is a hack really (it works well enough though). The reason why I would like to use two flipfluidobjects is to be able to modify that density information and really play with simulating different materials with different densities. I would assume I can modify that data somewhere in the solver too...

So it would be fine to have one flipfluid object as long as I can modify the different pressures and density settings per solver. With your file, Marc, I think I should be able to open it up and override it before it comes in.

And Julien, I think you are right as well in the sense that I will need two sets of data. Geometry1 and Geometry2... Or two sets of densities attached to the single geometry..?

I'll have to break down those assets completely to understand them better.

[Hazoc]

I wonder what happens if you use 1 flip object, but create it's points using the pop solver and assign and modify those different density atibutes there based on separate groups or something.

EDIT: Rest Density parameter adds to particle mass attribute. However if I use pop solver to give some particle streams different mass than others, it has no effect on the behavior despite mass in the details view is clearly changed for some parts of the fluid geo.

Edited March 22, 2011 by Hazoc

[pclaes]

I wonder what happens if you use 1 flip object, but create it's points using the pop solver and assign and modify those different density atibutes there based on separate groups or something.

EDIT: Rest Density parameter adds to particle mass attribute. However if I use pop solver to give some particle streams different mass than others, it has no effect on the behavior despite mass in the details view is clearly changed for some parts of the fluid geo.

I had a go as well, but it seems velocity is the only thing that gets picked up. I think mass is there so that if you use any of the pop operators it is taken into account for the pops, which really eventually feed back down into the velocity. (It also states on the input of the flipsolver "particle velocity"... so kinda makes sense). I am testing if I can override that density in the single fluidobject on a per point basis. Not even emitting new fluid.