A novel pressure-based algorithm for the simulation of compressible interfacial flows is proposed.
•
Interface discretisation method that retains the acoustic properties of the compressible flow.
•
Accurate propagation, reflection and transmission of acoustic waves in interfacial flows.
•
Accurate capturing and prediction of shock waves and rarefaction fans, including shock-interface interaction.
Abstract
A pressure-based algorithm for the simulation of compressible interfacial flows is presented. The algorithm is based on a fully-coupled finite-volume framework for unstructured meshes with collocated variable arrangement, in which the governing conservation laws are discretised in conservative form and solved in a single linear system of equations for velocity, pressure and specific total enthalpy, with the density evaluated by an equation of state. The bulk phases are distinguished using the Volume-of-Fluid (VOF) method and the motion of the fluid interface is captured by a state-of-the-art compressive VOF method. A new interface discretisation method is proposed, derived from an analogy with a contact discontinuity, that performs local changes to the discrete values of density and total enthalpy based on the assumption of thermodynamic equilibrium, and does not require a Riemann solver. This interface discretisation method yields a consistent definition of the fluid properties in the interface region, including a unique definition of the speed of sound and the Rankine–Hugoniot relations, and conserves the acoustic features of the flow, i.e. compression and expansion waves. A variety of representative test cases of gas–gas and gas–liquid flows, ranging from acoustic waves and shock tubes to shock-interface interactions in one-, two- and three-dimensional domains, is used to demonstrate the capabilities and versatility of the presented algorithm in all Mach number regimes. The propagation, reflection and transmission of acoustic waves, shock waves and rarefaction fans in interfacial flows are predicted accurately, even for difficult cases that feature fluids with shock impedance matching, transonic shock tubes or strong shocks in gas–liquid flows, as well as on unstructured meshes.
