Hello all!
Happy to announce another update to version 0.5.0. In this version:
- Log files can now be produced for each simulation, with time statistics and used hardware
- Simulation time has been improved once again, by utilizing faster data writes, thereby pausing less in simulation steps. Thanks to @jipolanco for this improvement by providing the draft function to achieve this.
- An FAQ section has been made: Frequently Asked Questions (FAQ) · AhmedSalih3d/SPHExample Wiki · GitHub
Now I have started to perform dedicated benchmarks up against DualSPHysics (state of the art SPH solver) and am getting these performance readings on the same hardware:
And attached here is a video showing the result at resolution = 0.02 on the same hardware (Intel i5). Res=0.01 has more particles than Res=0.02. Top row uses DBC (Dynamic Boundary Condition) with no density diffusion, whilte bottom row shows it with DDT enabled. Left column is SPHExample, right column is DualSPHysics.
Notice how SPHExample has a cleaner result in the density field (and thereby pressure too, since it is a weakly compressible implementaton) and that the noise around boundaries is non-existent compared to DualSPHysics + it is faster at these resolutions.
So it is turning out to be an awesome package for developing SPH methods, hope I can get some of the Julia community to give it a go!