An improved model for compressible multiphase flows based on Smoothed Particle Hydrodynamics with enhanced particle regeneration technique

In the current study, an improved numerical model is proposed in the compressible fields based on Smoothed Particle Hydrodynamics (SPH), which is comprised of MUSCL interpolation in multiphase flow, enhanced particle regeneration technique (PRT) and the particle shifting technique (PST) in compressible flows. The PRT is specially proposed to deal with compressible problems, in which the volume of particles have large variation during the whole simulation process. Different to the conventional PRT [1], an interface control (IC) method is proposed to deal with the mass conservation problem which may result in the unphysical movement of interface when modeling multifluids. The multiphase MUSCL interpolation aims at dealing with over-dissipation problem that exists in the Godunov-type SPH which may result in the wrong detection of wave front. To avoid large discontinuity between different fluids, two kinds of extrapolation schemes (constant extrapolation and isentropic extrapolation) are discussed and compared at the inspiration of the ghost fluid method (GFM). The proposed MUSCL-based compressible SPH model is validated and discussed in several challenging test cases, such as multiphase shock tube like problem, the shock wave impacting on multifluids interface problem and RichtmyerMeshkov instability etc., in which good agreements are obtained. (C) 2022 Elsevier Inc. All rights reserved.

Automated summary

In this study, an improved numerical model based on Smoothed Particle Hydrodynamics (SPH) is proposed for compressible fields. The model consists of MUSCL interpolation in multiphase flow, enhanced particle regeneration technique (PRT), and the particle shifting technique (PST) in compressible flows. The proposed model addresses compressible problems and aims to overcome issues related to mass conservation, over-dissipation, and wave front detection in multiphase flows. Validation and discussion of the proposed model are performed using challenging test cases, demonstrating its effectiveness and accuracy.