NPLIC: A machine learning approach to piecewise linear interface construction

Volume of fluid (VOF) methods are extensively used to track interfaces in numerical simulations, and many VOF algorithms require that the interface be reconstructed geometrically. For this purpose, Piecewise Linear Interface Construction (PLIC) technique is most frequently used, which for reasons geometric complexity can slow difficult implement. Here, we propose an alternative neural network based method called NPLIC perform PLIC calculations. The model trained on a large synthetic dataset solutions square, cubic, triangular, tetrahedral meshes. We show data-driven approach results accurate calculations at fraction usual computational cost, single system reconstruction different mesh types.