Implicit Frictional Boundary Handling for SPH.

Area of Science:

• Computational physics
• Fluid dynamics
• Numerical simulation

Background:

• Smoothed Particle Hydrodynamics (SPH) simulations often struggle with accurately representing rigid boundaries.
• Existing methods like density maps improve boundary handling but have limitations.
• Problems include inaccurate pressure forces and complex surface representations.

Purpose of the Study:

• To develop a novel, robust method for handling static and dynamic rigid boundaries in SPH simulations.
• To improve upon the density maps approach by precomputing boundary volume contributions.
• To achieve more accurate and efficient simulations with reduced computational cost.

Main Methods:

• Precomputation of boundary geometry volume contributions, stored on a spatial grid.
• Utilizing standard SPH kernel differentiation for force calculations.
• Seamless integration with existing SPH frameworks and flexibility with different kernels.

Main Results:

• Achieved smooth pressure forces comparable to density maps, even at lower resolutions.
• Reduced precomputation times and memory requirements by over two orders of magnitude.
• Demonstrated seamless integration and compatibility with various SPH methods and kernels.

Conclusions:

• The proposed volume maps method offers significant advantages over density maps for SPH boundary handling.
• This approach leads to more efficient, accurate, and flexible fluid simulations.
• The method is readily applicable to various SPH simulations, including those with surface tension and viscosity.