A Taichi implementation of Smooth Particle Hydrodynamics (SPH) simulator. Taichi is a productive & portable programming language for high-performance, sparse & differentiable computing. The 2D case below can be run and rendered efficiently on a laptop thanks to Taichi.

Currently, the following features have been implemented:

• Weakly Compressible SPH (WCSPH)[1]
• Predictive-Corrective Incompressible SPH (PCISPH)[2]
• Divergence free SPH (DFSPH)[3]

Taichi Version: 0.6.10

• Demo (PCISPH, 4.5k particles)

Run python scene.py

Demos for the other two methods: python scene.py --method WCSPH or python scene.py --method DFSPH

1. M. Becker and M. Teschner (2007). "Weakly compressible SPH for free surface flows". In:Proceedings of the 2007 ACM SIGGRAPH/Eurographics symposium on Computer animation. Eurographics Association, pp. 209–217.
2. B. Solenthaler and R. Pajarola (2009). “Predictive-corrective incompressible SPH”. In: ACM SIGGRAPH 2009 papers, pp. 1–6.
3. J. Bender, D. Koschier (2015) Divergence-free smoothed particle hydrodynamics[C]//Proceedings of the 14th ACM SIGGRAPH/Eurographics symposium on computer animation. ACM, 2015: 147-155.