A Higher Order Mixed Discontinuous FiniteElement Method For Incompressible Flows
Description
A higher order mixed discontinous Galerkin finite element method is employed for the un-
steady Navier Stokes equations. In order to iteratively solve the Navier Stokes equations
several problems are tackled. The coupling of the physical quantities velocity and pressure is
resolved with help of the standard pressure correction scheme. The temporal discretisation
relies on a finite difference scheme (implicit midpoint rule) and the nonlinearity due to the
advection term is resolved with a Picard iteration. On top of that, for spatial discretisation
finite elements are used, which are fully discontinous for the pressure and partially disconti-
nous (in the tangential component) for the velocity approximation. The partial discontinuity
introduces local fluxes in the weak formulation of the finite element methods, here defined
as upwind fluxes. Furthermore, the jumps in the tangential component are penalised in
the diffusion term according to the symmetric interior penalty, which also introduces fur-
ther terms for consistency and symmetry reasons. Generally, in mixed discontinous Galerkin
methods the continuity equation, which ensures incompressibility of the flow, is solved exact
by retaining the continuity in the normal component and incorporation of the continuity
equation in the residual formulation of the momentum equation, where the according trial
function acts as a Lagrange multiplier. On top of that, the finite elements are chosen to be
quadrilateral, since their tensor product structure opens more possibilities of assembly op-
timisation algorithms.
Files
Files
(63.2 MB)
Name | Size | Download all |
---|---|---|
md5:1f7b5c6c78e35a506df31ee5851c7102
|
63.2 MB | Download |