Equal-order finite elements of hydrostatic flow problems (original) (raw)
Related papers
A mimetic finite difference discretization for the incompressible Navier–Stokes equations
International Journal for Numerical Methods in Fluids, 2008
A mimetic finite difference discretization of the three-dimensional, incompressible Navier-Stokes equations is introduced, based on ideas that have been applied successfully to geophysical flows over the last four decades. The proposed method is mass conservative and vorticity preserving, in the sense that a discrete form of the vorticity equation is derived naturally from the discrete momentum equation by application of a mimetic rotation operator. A vorticity preserving discretization of the viscous terms and an appropriate treatment for rigid wall boundary conditions are also proposed. The relationship of this approach to other similar techniques is discussed. Several test cases are considered, in which the proposed method is compared to a widely used finite difference discretization. The results obtained in these tests demonstrate the advantages of the proposed method, especially when strong vorticity production takes place at the boundaries.
Three-dimensional hydrodynamics on finite elements. Part I: Linearized harmonic model
International Journal for Numerical Methods in Fluids, 1987
The linearized three-dimensional hydrodynamic equations are solved numerically for periodic motions, subject to a linear slip condition at the bottom. The structure of the linearized equations allows an exact uncoupling of the horizontal and vertical computations, so that they may be achieved sequentially rather than simultaneously, and without iteration. The solution strategy involves simple horizontal Co finite elements for the description of free surface elevation. Vertical variations in velocity may be treated analytically for some special variations of viscosity with depth; more generally the finite element method is employed with onedimensional linear elements. Because of the uncoupling, the entire three-dimensional solution scales as a twodimensional vertically averaged problem. The limiting two-dimensional problem may be solved as a Helmholtz-type problem for elevation alone, using established techniques. Solutions for test problems are compared with known analytic solutions. Some simple gridding rules are established for the vertical discretization. Finally, a field application is shown involving the tidal response of the Lake Maracaibo (Venezuela) system.
Enforcing the continuity equation in numerical models of geophysical fluid flows
Applied Mathematics Letters, 2001
Abstraet--A method is described for modifying the velocity field in geophysical fluid models so as to enforce the continuity equation. A corrective mass flux is introduced, which derives from a scalar potential, The latter is the solution of a Poisson problem which is formulated in such a way that a suitable lmrm of the corrective velocity be minimum. It is seen that a generalised vertical coordinate may be used. Finally, an elementary, one-dimensional illustration of the functioning of the method suggested is provided. (~)
Three-dimensional hydrodynamics on finite elements. Part II: Non-linear time-stepping model
International Journal for Numerical Methods in Fluids, 1991
The linearized three-dimensional hydrodynamic equations are solved numerically for periodic motions, subject to a linear slip condition at the bottom. The structure of the linearized equations allows an exact uncoupling of the horizontal and vertical computations, so that they may be achieved sequentially rather than simultaneously, and without iteration. The solution strategy involves simple horizontal Co finite elements for the description of free surface elevation. Vertical variations in velocity may be treated analytically for some special variations of viscosity with depth; more generally the finite element method is employed with onedimensional linear elements. Because of the uncoupling, the entire three-dimensional solution scales as a twodimensional vertically averaged problem. The limiting two-dimensional problem may be solved as a Helmholtz-type problem for elevation alone, using established techniques. Solutions for test problems are compared with known analytic solutions. Some simple gridding rules are established for the vertical discretization. Finally, a field application is shown involving the tidal response of the Lake Maracaibo (Venezuela) system.
A finite element method for 3D hydrostatic water flows
Computing and Visualization in Science, 1999
In this paper we propose a numerical method for the solution of the “Quasi-3D” hydrodynamic equations. This approach uses a combination of standard linear finite elements along the vertical direction and non-conforming Raviart-Thomas elements in the horizontal planes. We describe, also, a suitable approximation for the convective terms.
On stable equal-order finite element formulations for incompressible flow problems
International Journal for Numerical Methods in Engineering, 1992
In this work we discuss stable equal-order finite element formulations for incompressible flow problems based on Petrov-Galerkin methods, constructed by adding to the classical Galerkin formulation leastsquares of the governing equations. Continuous and discontinuous pressure interpolations are considered. Numerical results are presented reinforcing the numerical analysis.
Journal of Scientific Computing, 2006
Any solution of the Navier-Stokes equations in a three-dimensional axisymmetric domain admits a Fourier expansion with respect to the angular variable, and it can be noted that each Fourier coefficient satisfies a variational problem on the meridian domain, all problems being coupled due to the nonlinear convection term. We propose a discretization of these equations which combines Fourier truncation and finite element methods applied to each two-dimensional system. We perform the a priori and a posteriori analysis of this discretization.
2012
We have developed a second-order numerical method, based on the matched interface and boundary (MIB) approach, to solve the Navier-Stokes equations with discontinuous viscosity and density on non-staggered Cartesian grids. We have derived for the first time the interface conditions for the intermediate velocity field and the pressure potential function that are introduced in the projection method. Differentiation of the velocity components on stencils across the interface is aided by the coupled fictitious velocity values, whose representations are solved by using the coupled velocity interface conditions. These fictitious values and the non-staggered grid allow a convenient and accurate approximation of the pressure and potential jump conditions. A compact finite difference method was adopted to explicitly compute the pressure derivatives at regular nodes to avoid the pressurevelocity decoupling. Numerical experiments verified the desired accuracy of the numerical method. Applications to geophysical problems demonstrated that the sharp pressure jumps on the clast-Newtonian matrix are accurately captured for various shear conditions, moderate viscosity contrasts and a wide range of density contrasts. We showed that large transfer errors will be introduced to the jumps of the pressure and the potential function in case of a large absolute difference of the viscosity across the interface; these errors will cause simulations to become unstable.