Despite significant advances in the modelling of interfacial heat and mass transfer, many proposed approaches remain restricted to structured or block-structured meshes and hexahedral AMR (Adaptive Mesh Refinement) grids, limiting their applicability to more complex engineering problems such as combustion, cryogenics, droplet evaporation in aero-engine transmission systems, and evaporative processes in the chemical industry. This work presents a numerical model for predicting interfacial phase change on arbitrary meshes. The model employs state-of-the-art geometric VOF methods and interface curvature computation via a Reconstructed Distance Function (RDF). We first extend an implicit method to compute mass fraction gradients normal to the interface and then employ the associated neighbour-cell mapping to build the non-linear system for the interfacial unknowns. Phase change is then obtained by solving this system, derived from interfacial temperature and mass conditions under local thermodynamic equilibrium. Simplified phase change models based solely on temperature or mass fraction gradients are also assessed. Analytical solutions and experimental data are used to validate the numerical results. For both static and moving droplets, the droplet temperatures show close agreement across the different gradient methods and phase change models, whereas the evaporation rates differ more noticeably, with the temperature-based model tending to provide more accurate results when the same mesh resolution is employed. These findings provide a model validated under sheared flow conditions, demonstrating its applicability to industrially relevant problems and clarifying when model simplifications can be made without compromising accuracy in capturing the key interfacial phenomena governing phase change.
Zanutto et al. (Mon,) studied this question.