An efficient tree-topological local mesh refinement on Cartesian grids for multiple moving objects in incompressible flow

This paper develops a tree-topological local mesh refinement (TLMR) method on Cartesian grids for the simulation of bio-inspired flow with multiple moving objects. The TLMR nests refinement mesh blocks of structured grids to the target regions and arrange the blocks in a tree topology. The method solves the time-dependent incompressible flow using a fractional-step method and discretizes the Navier-Stokes equation using a finite-difference formulation with an immersed boundary method to resolve the complex boundaries. When iteratively solving the discretized equations across the coarse and fine TLMR blocks, for better accuracy and faster convergence, the momentum equation is solved on all blocks simultaneously, while the Poisson equation is solved recursively from the coarsest block to the finest ones. When the refined blocks of the same block are connected, the parallel Schwarz method is used to iteratively solve both the momentum and Poisson equations. Convergence studies show that the algorithm is second-order accurate in space for both velocity and pressure, and the developed mesh refinement technique is benchmarked and demonstrated by several canonical flow problems. The TLMR enables a fast solution to an incompressible flow problem with complex boundaries or multiple moving objects. Various bio-inspired flows of multiple moving objects show that the solver can save over 80% computational time, proportional to the grid reduction when refinement is applied. (c) 2023 Elsevier Inc. All rights reserved.

Automated summary

This paper presents a tree-topological local mesh refinement (TLMR) method on Cartesian grids for simulating bio-inspired flow with multiple moving objects. The TLMR method nests refinement mesh blocks of structured grids to target regions and arranges the blocks in a tree topology. The algorithm is shown to be second-order accurate in space and capable of saving over 80% computational time when refining the grid.