Dyverso - Caronte Explosions

[Alex]

Hi again,
I think we just managed to fix your permissions. Maybe you can try to download the scene attached on the previous post.

Just tested the Google Drive link again and it works ok for me on a cookies-free (incognito) Chrome window.

Yes, you can modify the mass with the Density value. Use something much smaller than 1000 to see your objects float.

[hmercado]

Thanks again.

I downloaded the example file but there are still some things not quite right.
- When I simulate the example file the sim is a lot different than your playblast. Why would that be?
- Still if I stop the simulation and try to resume it the toruses reset their position and stop responding to the sim. This only happens if I try to resume from a frame other than the frame where I stopped the simulation.
Here's a screen capture: https://youtu.be/bo7RXEIIGwU

I was able to create an initial state for the toruses without realflow crashing like it does on 2015 but still they stop interacting with dyverso when I resume the simulation.

Any ideas?

Thanks

[Alex]

I think I enabled OpenCL for the simulation I posted. There can be some slight differences because of that, especially when multithreading is involved.

I didn't leave OpenCL enabled because I don't know which graphics card does your computer have, and there is a bug (already fixed) that may make crash RealFlow when changing the emitter Volume parameter. But, if your graphics card is powerful, you should experience a very good performance boost enabling CUDA or OpenCL.

Regarding objects moving or not, I think in this case it is totally my fault. If you check your export settings, the DY_DomainRigid node does not export anything.
Actually, you may want to set the objects in wireframe mode and display the DY_DomainRigid domain particles to check the actual state of the sim.

[hmercado]

Yeah I thought about the GPU making a difference in the simulation. I enabled my GPU and it looks like your playblast now.

I did notice that the rigid domain was not enabled for export, however I did a new test with the rigid domain enabled for export and I am getting the same issue. It seems the objects are getting detached from the particles when resuming the simulation at a frame other than the last simulated frame, the domain itself its resuming without any issues.

Take a look here: https://youtu.be/aOav6jZiEtk

[Alex]

This definitively looks like a bug.

I've checked it out and if you create initial states (for the objects, and both domains) it works. But it should work the way you are using it.

Thank you very much for letting us know.