A Galerkin implementation of the generalized Helmholtz decomposition for vorticity formulations (Q5935512)

From the conclusions: We present two approaches for determining boundary conditions appropriate for the vorticity form of Navier-Stokes equations. Both approaches are based on a Galerkin implementation for generalized Helmholtz decomposition (GHD). There are several advantages both numerically and conceptually in using a Galerkin formulation as opposed to the more traditional point-collocation formulations. A Galerkin finite element method is implemented to solve the vorticity equation using the GHD to provide appropriate boundary conditions as discussed above. The vorticity equation is linearized again using the GHD to determine the interior velocities. The driven cavity problem at a Reynolds number of 400 is considered as a benchmark. Both vorticity formulations (Neumann and Dirichlet) are shown to provide more accurate results than a primitive variable formulation for the same level of discretization. In fact, the vorticity formulations using 1681 grid points compared very favorably to a multigrid finite difference method using 16,641 grid points.