This paper presents a scalable parallelization of an Eulerian-Lagrangian method, namely the three-dimensional front tracking method, for simulating multiphase flows. Operating on Eulerian-Lagrangian grids makes the front tracking method challenging to parallelize and optimize because different types of communication (Lagrangian-Eulerian, Eulerian-Eulerian, and Lagrangian-Lagrangian) should be managed. In this work, we optimize the data movement in both the Eulerian and Lagrangian grids and propose two different strategies for handling the Lagrangian grid shared by multiple subdomains. Moreover, we model three different types of communication emerged as a result of parallelization and implement various latency-hiding optimizations to reduce the communication overhead. Good scalability of the parallelization strategies is demonstrated on two supercomputers. A strong scaling study using 256 cores simulating 1728 interfaces or bubbles achieves 32.5x speedup. We also conduct weak scaling study on 4096 cores simulating 27,648 bubbles on a 1024 3 1024 3 2048 Eulerian grid resolution.
This paper presents a scalable dynamic load balancing scheme for a parallel front-tracking method based multiphase flow simulation. In this simulation employing both Lagrangian and Eulerian grids, processes operating on Lagrangian grid are susceptible to load imbalance due to moving Lagrangian grid points (bubbles) and load distribution based on spatial location of bubbles. To load balance these processes, we distribute load keeping in view both current processor load distribution and bubble spatial locality and remap interprocess communication. The result is a uniform processor load distribution and predictable and less expensive communication scheme. Scalability studies on the Hazel Hen supercomputer demonstrate excellent scaling with exponential savings in execution time as the problem size becomes increasingly large. While moderate speedup is observed for strong scaling, speedup of up to 30% is achieved over nonload-balanced version when simulating 13824 bubbles on 4096 cores for weak scaling studies.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.