Theory

Nassu solver is developed under Lattice-Boltzmann Method (LBM) framework Oliveira et al.[1]. The LBM is a Computational Fluid Dynamics (CFD) method based on the discretization of the Boltzmann equation from gas kinetic theory. Rather than discretizing the Navier-Stokes Equations (NSE) directly, the LBM evolves discrete particle populations on a lattice through a kinetic/mesoscopic description, from which the macroscopic NSE behaviour is recovered. The scheme is deterministic at the population level; what differs from conventional CFD is the mesoscopic viewpoint, not a probabilistic one.

The result is a highly localized computational method which is extremely suited for high performance parallel GPU processing. The LBM holds second-order accuracy and is run with the recursive-regularized BGK (RR-BGK) collision operator as the production default, with a hybrid variant (HRRBGK) also available. To perform Large-Eddy Simulations (LES) the solver applies the Smagorinsky Sub-Grid Scale (SGS) model.

Spatial resolution is varied through a static 2:1 multiblock grid refinement, with the rate-of-strain rescaled at block interfaces so that the LES shear stress remains continuous across refinement levels; the substantial memory saving comes separately from storing only macroscopic quantities per node rather than the full population set. Turbulent inflow can be prescribed through the Synthetic Eddy Method (SEM) or through the precursor-based PODFS approach. The solver aims to reproduce wind tunnel experiments, hence Dirichlet and Neumann boundary conditions are available, together with equilibrium wall-model boundary conditions (WallModelEqLog, WallModelEqTBL). In order to delineate complex boundary walls, the Immersed Boundary Method (IBM) can be adopted. Passive scalar transport is also supported. The theoretical aspects regarding the mentioned features are disclosed in this section.

---

[1]

Waine Oliveira, Rodrigo S. Romanus, and Alan Lugarini. Nassu: a high performance LES solver for computational wind engineering. Journal of Wind Engineering and Industrial Aerodynamics, 274:106465, 2026. doi:10.1016/j.jweia.2026.106465.