Provides scalable particle fluid simulation code.
This project presents geometric feature-size based adaptive sampling algorithms for Lagrangian particle fluids. This adaptive sampling strategy allows using smaller (and thus more) particles in geometrically complex regions, while less particles are used for thick flat fluid volumes. Additionally, a novel distance-based particle surface definition is implemented which hides the particle granularity and allows dynamic resampling near the fluid-air interface. The code is implemented in C++ and should compile on linux.