Parallel Eulerian-Lagrangian Method with Adaptive Mesh Refinement for Moving Boundary Computation

Abstract

In this study, we present a parallelized adaptive moving boundary computation technique on distributed memory multi-processor systems for multi-scale multiphase flow simulations. The solver utilizes the Eulerian-Lagrangian method to track moving (Lagrangian) interfaces explicitly on the stationary (Eulerian) Cartesian grid where the flow fields are computed. Since there exists strong data and task dependency between two distinct Eulerian and Lagrangian domain, we address the decomposition strategies for each domain separately. We then propose a trade-off approach aiming for parallel scalability. Spatial domain decomposition is adopted for both Eulerian and Lagrangian domains for load balancing and data locality to minimize inter-processor communication. In addition, a cellbased unstructured parallel adaptive mesh refinement (AMR) technique is implemented for the flexible local refinement with efficient grid usage and even-distributed computational workload among processors. The parallel performance is evaluated independently for the Cartesian grid solver and sub-procedures in cell-based unstructured AMR. The capability and the overall performance of the parallel adaptive Eulerian-Lagrangian method including moving boundary and topological change is demonstrated by modeling binary droplet collisions. With the aid of the present techniques, large scale moving boundary problems can be effectively computed.

Simulation of droplet dispersion in COVID-19 type pandemics on Fugaku

“…This insertion and removal of Lagrangian particles is efficiently carried out using the particle's global index based hash table. With regards to the domain decomposition for parallel computing, a spatial decomposition strategy is employed in this work [14]. The computational domain is decomposed such that on any given partition the Eulerian BCM mesh and the Lagrangian marker particles spatially overlap.…”

CUBE: A scalable framework for large-scale industrial simulations

“…As there are two separate domains, namely, the Lagragian domain and Eulerian domain, combined or separate domain decomposition strategies may be employed. Three of the most commonly used approaches for decomposing a Lagrangian–Eulerian system are task parallelism, atomic decomposition, and spatial decomposition, respectively (Kuan et al, 2013). In task parallelism, the Lagrangian and Eulerian workload are assigned to separate set of processors.…”

Communication analysis and optimization of 3D front tracking method for multiphase flow simulations

“…For example, Darmana et al (2006) and Nkonga and Charrier (2002) parallelized the independent particles using a domain decomposition method for both Eulerian and Lagrangian grids. Kuan et al (2013) parallelized the connected particle Eulerian-Lagrangian method by applying domain decomposition to both grids. They spatially decompose the Lagrangian grid similar to the Eulerian grid.…”

“…Kuan et al (2013) parallelized the connected particle Eulerian–Lagrangian method by applying domain decomposition to both grids. They spatially decompose the Lagrangian grid similar to the Eulerian grid.…”

“…The main focus of the prior work about the parallelization of the Eulerian–Lagrangian methods for multiphase flows (Nkonga and Charrier, 2002; Darmana et al, 2006; Kuan et al, 2013) relied on domain decomposition and focused less on the resulting communication overhead. In this work, we model the resulting communication in depth and implement two strategies for the Eulerian–Lagrangian method.…”