dbnsCavitatingFoam: A density-based solver with equilibrium cavitation models in the OpenFOAM framework

TL;DR

This paper introduces a density-based cavitation solver in OpenFOAM that models phase transitions and shock waves, enabling better prediction of cavitation erosion through validated simulations and a specialized post-processing tool.

Contribution

The paper develops a novel density-based cavitation solver in OpenFOAM that considers phase compressibility and shock wave effects, with an integrated erosion risk detection tool.

Findings

Solver accurately matches analytical, numerical, and experimental results.

It captures shock waves from cavitation collapse effectively.

The post-processing tool identifies high erosion risk areas.

Abstract

This paper presents the development of a density-based solver suitable for cavitating flows in the OpenFOAM framework. In this solver, the thermodynamic equilibrium mixture approach is adopted to model the presence of and the phase transition between liquid and vapor phases. Using this approach, two cavitation models are implemented in a separate library, although more cavitation models can be easily added. The two are a temperature-dependent cavitation model and a barotropic cavitation model developed by Egerer et al. (2014). One of the main advantages of the solver is that it considers the compressibility of all phases. This feature combined with using the density-based approach enables capturing shock-waves created upon the collapse of cavitating structures which are known to be one of the main mechanisms of cavitation erosion. The implementation also includes a post-processing tool which detects aggressive collapse-induced shock-waves based on the method proposed by Mihatsch et al. (2015) and identifies the areas with a high risk of cavitation erosion. In order to validate the implemented solver and post-processing tool, four cases with progressively increasing complexity are simulated. The results from simulation of these cases are compared with analytical solution, similar numerical simulations as well as available experimental results. All of these comparisons show that the numerical results obtained by the implemented solver agree well with the reference analytical solution and numerical and experimental results.