Smoke is "killing" me

[Activate]

While the subject may have been true in the Bay Area last month, the subject is much more about the vexation of getting smoke right. I had found a little bit of time to go through Steven Knipping's Rigids II video and decided to add some debris and smoke. To source the debris and smoke, I processed the constraint breaks. For the smoke source, I made a tiny pop sim with noise and some velocity filtering. The smoke for the explosion has pretty high velocities, e.g. around 30 units/s. Initially, the smoke was basically just smooth mushrooms. I then introduced a high value disturbance against the velocity field in the low density areas. This broke up the smoke but I never was able to get it right. Here is where I ended after I had decided I had spent too much time; see video below - ignore obnoxious sound track :-)

Would anyone here be willing to point me in a direction on how to improve this?

I ran a bunch of wedges against lower resolution simulations. The overall smoke container ended up being really large, so sim-ing and rendering low resolution wedges took a few hours of turn around.

 

Hip file attached for anyone who wants to take a look.

 

Thank you,

 Niels.

rigids2_v7.hiplc

[Activate]

Hmm. I was hoping that someone could share some sage wisdom but perhaps my questions are too naive :-)

[Atom]

I did take a look at your file but the first thing I would try is the zero velocity trick for removing mushrooming. Then your correction systems might not have to work as hard. Basically @v=set(0,0,0); before you create your velocity field. Then create the density field using a particle system or a point affected by forces or gravity. So the velocity only exists in the popnet, not the dopnet.

I have an experiment on this approach at the link for creating a single falling dust stream.

 

Edited December 20, 2018 by Atom