[Bug] FluidEffect2DViscosityDFSPH is broken

[kBlankii]

Describe the bug

No matter how you set the fluid_viscosity_coefficient for it, the simulation would explode, with following error being outputed:

E 0:00:03:256 godot_core::private::set_gdext_hook::{{closure}}: [panic C:\Users\runneradmin.cargo\git\checkouts\salva-c88c9745d16e1d12\b30c622\build\salva2d....\src\solver\pressure\dfsph_solver.rs:92] assertion failed: !denominator.is_zero()
<C++ 源文件> C:\Users\runneradmin.cargo\registry\src\index.crates.io-1949cf8c6b5b557f\godot-core-0.4.5\src\private.rs:343 @ godot_core::private::set_gdext_hook::{{closure}}()

To Reproduce

Just follow the official tutorial, but add FluidEffect2DViscosityDFSPH as the fluid effect

Expected behavior

The simulation should runs fine

Environment:
Godot v4.6.2.stable - Windows 10 (build 19045) - Multi-window, 1 monitor - Direct3D 12 (Forward+) - dedicated NVIDIA GeForce RTX 5070 (NVIDIA; 32.0.15.9186) - 13th Gen Intel(R) Core(TM) i5-13600KF (14 threads) - 15.78 GiB memory

• Type: Godot Rapier v0.8.32, 2d-single-simd-parallel

Example project(zip)

fluid_effects_panic_.zip

[Ughuuu]

My initial guess is it's either bug in salva that needs fixed there, or the numbers we use are too high and the solver there explodes. Either way it should be fixed in salva. We should probably open a new issue there if there isn't one already:
https://github.com/dimforge/salva

[bencmckenna]

Just wanted to chime in and say I was also getting this error while using Fluid3D (except the error path said salva3d instead of salva2d). I couldn't get it to happen consistently, but I haven't encountered it since removing the FluidEffect.