Scene Crash

[omere]

can you please check why my scene is crushing?
thank you

here is a link to the file:
http://we.tl/5J7lG0C4OK

[Ivaylo Katev]

can't reproduce it, can you tell the build info?

[omere]

it's 2.2.5 nightly build ID: 1512086525

[Ivaylo Katev]

ok, what is happening? is it crashing during the loading, the simulation or rendering?

[omere]

it crash the moment i click the start button.
sometimes immediately some times after 1 or 2 frames.
it basically happens the moment that there is interaction between the particles and the can.

[Ivaylo Katev]

can you attach the log file? use the about button to reach it

[omere]

i managed to solve it by flipping and unifying the Normals and performing Reset Xform on the model of the can.
but now i can't get a decent simulation.

i'm trying to fill up the can, then start simulating the pour.
i can't find a way to make an initial state. i tried to simulate with static can until frame 50, then load frame 50 with load& start. but the particles are deleted very fast.
when i try to simulate from frame -50, so the simulation will actually start in frame 0, the result of the simulation is very chunky and very water like.

can you simulate it and post the file so i will see what am i doing wrong?

[Ivaylo Katev]

create an object that defines the init fill up area, right click it, select phoenix fd properties, and enable the "initial fill up" uption

[omere]

Yes, exactly what I did.
Tge simulation is looking very bad and there is no initial State.
These are two separate issues.
First tge liquid simulation looks chunky and not like liquid.
The second is that i need to letbtge liquid to relax inside tge can berore starting to simulte. But there isn't an option to simulate an initial state over one frame. Like in max cloth simulation for example

[Ivaylo Katev]

there is a button load&start, it does exactly this.
about the simulation issues,when the liquid does not behave smooth enough, you have to increase the spf, or to use bigger scale. in our beer sample scene we use bigger scale to avoid incredible high spf, in the most cases nobody notices that the scale is twice bigger. but the main parameter that gives "low scale" behavior of the liquid is the surface tension, set it to 1.

[omere]

the load&start is not exactly doing that.
in my case the can is animated. so i need to let the liquid to calm down while the can is not animated. so what i'm looking for is a button the simulate local, over and over again in frame 0. like it is in realflow or in max cloth sim. so the final result will be that the liquid calmed down in frame 0 and then i can start simulate while the can is moving around pouring the liquid.

i've tried a work around where like this:
1. i duplicated the can and deleted the animation.
2. i simulated with non animated can until the liquid calmed down (frame 50).
3. i brought back the animated can and hide the static can.
4. using load&start, i've loaded frame 50 to start and simulate from frame 0.

this method failed because when using load&start the liquid appears in frame 0 not the same as it was in frame 50. then the liquid particles slowly disappear like something is killing them.

[Ivaylo Katev]

i see now, we can add this functionality in future, but for now you have to do this manually, disabling the animation of the can
about the disappearing particles - i think your geometry is open, phoenix supports open geometry, but the liquid must intercat with the face side of the mesh, not with the back side like in your case.
in general, it's always better to use closed geometry. i think the easiest way to fix this is to use thin shell modifier.
btw, why you need to simulate the liquid inside the can? it's not a bottle, you can't see it, the usual technique is just to create an animated source at the can hole.

[omere]

the behavior of the liquid look more realistic when it is being poured from a filled object rather then from emitter with constant stream or even animated emitter with animated velocity

[omere]

mmm...but the scene is still crashing.
here is a link to the updated scene:

http://we.tl/ZKqBL2CcdT

it crash after simulating 1 frame

[Ivaylo Katev]

ok, i will try it again, but we can't always reproduce the crash, that's why we have logging and many times we fix the bugs using only the log. attach it please, to have at least info where the problem may be