The paper is devoted to different approaches of a pressure-velocity coupling method to account for density variations in cavitating two-phase flow simulations. Results obtained from two strategies are investigated in detail. A simpler engineering approach associated the variations of the local density solely with the changes of the vapor-volume fraction computed by a cavitation model and assumes incompressible vapor and water phases. A more elaborate method additionally accounts for the compressibility of the two individual fluid phases. Numerical issues of significance for engineering applications are discussed in the paper, such as the occurrence of ill-conditioned matrices or cavitation-model dependencies. The single-phase verification and validation study refers to prominent aerodynamic benchmarks, i.e. a convergent-divergent nozzle flow and the flow over a bump in a channel. Cavitating flow validations are concerned with a stationary flow over a hydrofoil. An unsteady cavitating flow over a NACA0015 hydrofoil is computed to demonstrate merits of the implemented compressible fluid method to simulate sheet and vapor cavitation including the collapse of a vapor cloud followed by a shock wave formation and propagation. Results demonstrate that the predicted cavitation pattern of the two approaches are very similar and the compressible flow model is only required when attention is directed to the evolution of pressure waves of the collapsing cavities, e.g. in erosion studies.
