How to retain advected velocity of particles as new particles are crea

[magneto]

Hi,

 

I am bringing some particles into a smoke object using SOP Geometry DOP and update it every frame. Then I am advecting particles onto velocity field using Gas Advect DOP for the Velocity Update input of the pyro solver.

 

It works but as I move forward in time, the particles' velocity is cut off because the new particles have no velocity.

 

How can I update the particles' position as they are advected and the new particles come through so the old particles are not reset by the SOP Geometry again?

 

 

Thanks

[cwhite]

I think the simplest approach is to do a POP sim for the particles, and use the POP Advect by Volumes node with the velocity field from your pyro sim.

[magneto]

Thanks Cameron, that works great. But if you were to do it in DOPs, do you know how one would achieve this?

 

I thought I would need to do these:

SOP Geometry reads the SOP data every frame.

I copy my calculated values (vel in this case) onto these particles using id, so the old particles get correct values

I perform my operation that calculates the velocity for all particles

then repeat

 

Just for learning practice

[rtep]

Hi Ryan,

 

Any chance you could put up an example file?  I'm not quite sure what you are doing with the particles.

[magneto]

I uploaded it now

particles.hip

[rtep]

Does this do it?

 

particlestest.hip

[magneto]

Thanks Richard. If I am not wrong, you are creating the particles in DOPs, right? Also some particles are not moving.

 

Another thing is I might have another means of generating the geometry, like scattering some points each frame and merging them. So am looking for a way to retain the velocity set for the old points whatever method is used in SOPs.

[rtep]

Hi Ryan, here's another attempt.  I'm probably an idiot but I'm still not quite sure what you are looking for.  Could you maybe give a specific example of when you would use this setup?

 

This example still uses the multisolver but it's calling the pop emitter from the particles at sop level so you could do any sort of animation or geo scattering that you want.

 

particlestest_v02.hip

 

it's interesting how you just added a sop Geometry to your smoke object and started pumping points into your smoke dop, but I wouldn't actually expect that to work since you don't have any kind of rbd or pop solver.  Is all of that being handled by the microsolvers you have going into the smoke solver?

 

So hang on try this one:

 

particlestest_v03.hip

 

This one uses the sop geo you had set up.  The pop net reads that sop geo to emit particles in the multisolver.

 

I don't fully understand how your setup is working but it has some unexpectedly awesome results.  If you feed full geo into the sop geometry instead of points it produces a heat distortion rippling effect on the geo.  I think what's going on is since you don't have a particle solver in the initial setup it is updating the points from the sop geo back to their non-simulated position, applying the advection vel, solving the geo on that frame, then on the next frame returning the geo to it's unsolved position, and repeating.

 

Does any of this rambling help?