A coupled finite volume-lattice Boltzmann method for incompressible internal flows

Zhou, Yang; Santasmasas, Marta Camps; De Rosis, Alessandro; Hinder, Ian; Moulinec, Charles; Revell, Alistair

Abstract

We present, test and validate a two-way framework that couples macroscopic and mesoscopic methods to simulate incompressible internal flows spanning a range of spatial and temporal scales. Specifically, the unstructured finite volume method (FVM) is coupled to the structured lattice Boltzmann method (LBM). The multi-resolution domain is resolved through two strategies, i.e., non-Cartesian FVM meshes and multi-level refinement LBM using octree-like Cartesian grid points. The coupled approach divides the entire computational domain into sub-regions, each solved independently. Information exchange between these sub-regions is facilitated by a coupling library that introduces spatial interpolation and temporal iteration schemes for different scales. The effectiveness of the proposed coupled strategy is assessed against well-documented benchmark tests and further examined in scenarios involving flow over artificial porous media. The results obtained by the new coupled framework show excellent agreement with reference data and exhibit strong parallel performance for tests on up to 32768 CPU cores, demonstrating the potential of the approach for large-scale investigations.

• Journal Article Type: Article
• Acceptance Date: May 15, 2025
• Online Publication Date: May 23, 2025
• Publication Date: 2025-09
• Deposit Date: Jun 12, 2025
• Publicly Available Date: Jun 12, 2025
• Journal: Computer Physics Communications
• Print ISSN: 0010-4655
• Publisher: Elsevier
• Peer Reviewed: Peer Reviewed
• Volume: 314
• Article Number: 109686
• DOI: https://doi.org/10.1016/j.cpc.2025.109686