A sharp-interface mesoscopic model for dendritic growth

Abstract The grain envelope model (GEM) describes the growth of envelopes dendritic crystal grains during solidification. Numerically growing are usually tracked using an interface capturing method employing a phase field equation on fixed grid. Such approach as diffuse interface, which can lead to numerical artefacts that possibly detrimental. In this work, we present sharp-interface formulation GEM eliminates such and thus track with high accuracy. new uses adaptive meshless discretization solve diffusion in liquid around grains. We use ability operate scattered nodes accurately describe i.e., envelope. proposed algorithm combines parametric surface reconstruction, surfaces, global solution construction procedure partial differential operator approximation monomials basis functions. is demonstrated two-dimensional h -adaptive diffusive dendrites assessed by comparison conventional diffuse-interface fixed-grid GEM.