Validation of the particle finite element method (PFEM) for simulation of free surface flows (original) (raw)

Purpose -The purpose of this paper is to evaluate the possibilities of the particle finite element method for simulation of free surface flows. Design/methodology/approach -A numerical simulation of a number of examples for which experimental data are available is performed. The simulations are run using the same scale as the experiment in order to minimize errors due to scale effects. Some examples are chosen from the civil engineering field: a study of the flow over a flip bucket is analyzed for both 2D and 3D models, and the flow under a planar sluice gate is studied in 2D. Other examples, such as a 2D and 3D "dam break" with an obstacle are taken from the smooth particle hydrodynamics literature. Findings -Different scenarios are simulated by changing the boundary conditions for reproducing flows with the desired characteristics. Different mesh sizes are considered for evaluating their influence on the final solution. Originality/value -Details of the input data for all the examples studied are given. The aim is to identify benchmark problems for future comparisons between different numerical approaches for free surface flows. This is the basis of the volume of fluid (VOF) technique. This scalar function is convected according to the flow velocity field once a suitable discretization of the space is provided. This allows using existing Eulerian codes and this justifies the success of the VOF method in the CFD community. This formulation permits to deal naturally with separation (or reattachment) of parts of the fluid domain; nevertheless some concerns remain particularly on the imposition of the Dirichlet boundary conditions on the free surface. Even if all the advantages of Eulerian methods on fixed meshes can be retained, the VOF approach tends to introduce some diffusion in the position of sharp interfaces (see for examples Zalesak's circle benchmark ).