📝 SummarXiv

Abstract

Submarines are vital for maritime defense, requiring optimized hydrodynamic performance to minimize resistance. Advancements in Computational Fluid Dynamics (CFD) enable accurate predictions of submarine hydrodynamics for optimal design. This study compared the meshing capabilities of OpenFOAM and commercial software as well as the performance of High-Performance Computing (HPC) and standard PC resources upon hydrodynamic characteristics. The RANS turbulence model with was employed to analyze the resistances of the MARIN's BB2-class submarine. CFD simulations were conducted at a model scale (1:35.1) at a speed of 1.8235 m/s ( of 21 knots) upon various mesh densities from 1 to 97 million cells. Empirical equations were initialized for turbulence parameters. Mesh sensitivity and iteration convergence ensured validated results. The findings showed that the results were validated with errors ranging from 0.3% to 10% across different mesh densities. The lowest error (0.3%) was achieved with 97 million cells generated by the commercial meshing tool with HPC, while 13 million cells by OpenFOAM with a standard PC resulted in a 3.4% error. Accuracy improved with precise initialization of turbulence parameters, mesh strategy, numerical schemes, and computing resources. The application of a standard PC with the OpenFOAM meshing tool was able to produce an acceptable accuracy, with less than 5% error for lower mesh densities. Thus, it can be suggested that using a standard PC was beneficial for preliminary hydrodynamic simulations. However, HPC with commercial software was essential for detailed industrial analyses, such as full-scale resistance and propulsion simulations.