Lightweight lattice Boltzmann

Area of Science:

• Computational physics
• Materials science
• Fluid dynamics

Background:

• Traditional lattice Boltzmann methods (LBM) require significant memory for simulating soft materials.
• High-performance computing is crucial for complex soft matter simulations.

Purpose of the Study:

• To introduce a regularized lattice Boltzmann method (RLBM) for efficient soft material simulations.
• To reduce memory requirements and data access costs in LBM simulations.

Main Methods:

• Reconstructing distribution functions from hydrodynamic variables (density, momentum, pressure tensor).
• Avoiding storage of the full set of discrete populations.
• Validating the method through benchmark tests relevant to soft matter, including fluid droplet collisions.

Main Results:

• Demonstrated significantly lower memory requirements compared to standard LBM.
• Showcased reduced data access costs.
• Validated the method's efficacy in simulating soft matter phenomena.

Conclusions:

• The RLBM offers a memory-efficient and cost-effective alternative for soft material simulations.
• This method is well-suited for high-performance simulations on future exascale computers.
• The approach is particularly relevant for complex systems like colliding fluid droplets.