Obi Official Forum
2D Simulation? - Printable Version

+- Obi Official Forum (https://obi.virtualmethodstudio.com/forum)
+-- Forum: Obi Users Category (https://obi.virtualmethodstudio.com/forum/forum-1.html)
+--- Forum: Obi Fluid (https://obi.virtualmethodstudio.com/forum/forum-3.html)
+--- Thread: 2D Simulation? (/thread-4069.html)



2D Simulation? - Blitz54 - 08-12-2023

Hello again! Been a while, had a quick question that I can't find a direct answer for in the forums.

Quite simply, I'm curious if Obi fluid can do a proper 2D simulation. Basically, I'm using Zibra Liquids and kinda tricking it to work in 2D. It runs 3D, but by tweaking the values enough (gravity, viscosity, etc) and keeping the box just barely bigger than a particle, it will kinda act like it's 2D. The problem now is I updated to the latest version, which changed the effectiveness of some values, and it's a pain in the ass getting it to work properly again. It will sometimes slide down the front and back invisible walls of the box.

So ya, curious if Obi can pull this off and just totally ignore Z depth. Also, how sticky can the fluid be? Zibra struggles with super sticky fluids, and now the fluid is bouncing off walls since the update. Even with friction maxed out.

Final question, I see you have GPU simulation arriving with obi 7. It looks like a similar structure to Zibra with a 3D bounding box. Will the GPU version be able to do 2D sims?

Thanks lots! I see you're very active in the forums and I appreciated the quick response last time I was here.


RE: 2D Simulation? - josemendez - 08-12-2023

(08-12-2023, 09:13 AM)Blitz54 Wrote: Hello again! Been a while, had a quick question that I can't find a direct answer for in the forums.

Quite simply, I'm curious if Obi fluid can do a proper 2D simulation. Basically, I'm using Zibra Liquids and kinda tricking it to work in 2D. It runs 3D, but by tweaking the values enough (gravity, viscosity, etc) and keeping the box just barely bigger than a particle, it will kinda act like it's 2D. The problem now is I updated to the latest version, which changed the effectiveness of some values, and it's a pain in the ass getting it to work properly again. It will sometimes slide down the front and back invisible walls of the box.

Yes, the ObiSolver component has a 2D-specific simulation mode. All data just loses one dimension and the simulation is done in 2D. See: http://obi.virtualmethodstudio.com/manual/6.3/obisolver.html

Quote:Also, how sticky can the fluid be? Zibra struggles with super sticky fluids, and now the fluid is bouncing off walls since the update. Even with friction maxed out.

You can have arbitrarily sticky fluids in Obi. Friction can also be set to 1 and fluids won’t slide off surfaces.

For viscosity (which is completely different from stickiness), you’re only limited by how small a timestep you can afford. Obi 7 has a much improved anisotropic viscosity model that’s also uncoditionally stable, so this limitation only applies to Obi 6 and previous versions.

Quote:Final question, I see you have GPU simulation arriving with obi 7. It looks like a similar structure to Zibra with a 3D bounding box. Will the GPU version be able to do 2D sims?

Yes, Obi 7 has a GPU mode. See:

http://obi.virtualmethodstudio.com/manual/7.0/whatsnew.html

However it isn’t related to Zibra in any way. Zibra is a hybrid particle/grid method (MLS-MPM), the fact that simulation happens on a grid means there must be some spatial bounds to it. Obi is purely particle-based (XPBD) which means there’s no need for a bounding box, fluids are free to move anywhere.

2D mode is also supported by the GPU backend. It does everything the regular CPU backend does, but allows for much larger simulations, all that’s needed to use GPU mode is flick a switch.

Obi 7 is currently in beta, you can send us a request to try it.

kind regards,