Optimize Scene

[Aeonx21]

Hey odforce,

I am currently working on a scene that involves a bullet shooting through a glass of water. I figured out how to do the effect but when I run the entire simulation it just stops at the frame the glass breaks. I was wondering why it might do that. I am not rendering yet but I have built the materials. Any ideas as too why that would happen? I am on a decently powerful machine.

- Jake

[Alexander Bong]

it's probably cooking your simulation. the solver is very efficient when there are no possible collision nearby, which could give the illusion that it should go faster.

[ikarus]

Could you elaborate on how the simulations 'stops'? Is it freezing houdini? Is it cooking a specific microsolver? etc

[Hudson]

Alexander is right.

I think your bullet is probably going really fast. and in the frame after the glas breaks all pieces will have been spread out quite a lot and also the flip fluid will be disturbed by that = long simulation times. Have you tried only to break the glass without the liquid inside?

I would do that first of all just to see if the bullet's speed is correct and that the glass pieces does not fly all over the place. It is also worth mentioning that you will either need to crank up your substeps or simulate in a higher FPS or even the both. So start of with just the glass. then add the fluid afterwards when you got your simulation settings right

[Aeonx21]

So when I say the simulation stops, I mean that it seems too freeze. But if I hit escape a few times it will go back too normal at frame one. The bullet is going pretty fast but I have done the rigid bodies on there own and runs at a decent speed. Most of the pieces are still in the shot after the end of 100 frames. My sup steps are 6. The rigid bodies look decent but when the flip fluids are added it starts too slow down which is normal. It runs until frame 26. At 27 it fractures and breaks. That is the frame where it seems too stop. I have left it alone for over an hour and still nothing.

[ikarus]

Does it happen even when you used a cached out version of the fracture simulation?

[Hudson]

I would do as Raymond says but it is worth to keep in mind that only the glass pieces will affect the flip particles and not vice-versa. But it is a way of analysing the issue. I would also try different methods for the collision volumes of the pieces. Which methods have you tried so far?

It is worth trying to do the same simulation and not breaking the glass(keep it and the bullet static). Just to see how long it takes to simulate the fluid alone.

[Aeonx21]

Yes it does it every time and is cached out. I am using the static object dop node as the collision. Right now I'm redoing the flip because everything else worked until then. Previously i used a circle to fill up my class and had it transform up in Y. Now I'm just shrinking my object and adding points and letting gravity take over until I get a rest state.

[ikarus]

I've had a couple occasions where the gas build collision mask caused weird cooking issues with specific collision geometry configurations. I would try building the collision + collisionvel fields in sops and using a source volume dop to add them directly to the sim fields, instead of using a static object.

[Aeonx21]

I tried using volumes but same problem. So not i decided to redo the flip cause I thought that might be the problem, however now the water collapses on its self and I dont know why