A grid‐free upwind relaxation scheme for inviscid compressible flows

A new grid‐free upwind relaxation scheme for simulating inviscid compressible flows is presented in this paper. The non‐linear conservation equations are converted to linear convection equations with non‐linear source terms by using a relaxation system and its interpretation as a discrete Boltzmann equation. A splitting method is used to separate the convection and relaxation parts. Least squares upwinding is used for discretizing the convection equations, thus developing a grid‐free scheme which can operate on any arbitrary distribution of points. The scheme is grid free in the sense that it works on any arbitrary distribution of points and it does not require any topological information like elements, faces, edges, etc. This method is tested on some standard test cases. To explore the power of the grid‐free scheme, solution‐based adaptation of points is done and the results are presented, which demonstrate the efficiency of the new grid‐free scheme. Copyright © 2005 John Wiley & Sons, Ltd.     

Inviscid flow modelling using asymmetric implicit finite difference schemes

Asymmetric spatial implicit high‐order schemes are introduced and, based on Fourier analysis, the dispersion and damping are calculated depending on the asymmetry parameter. The derived schemes are then applied to a number of inviscid problems. For incompressible convection problems the proposed asymmetric schemes (applied as upwind schemes) lead to stable and accurate results. To extend the applicability of the proposed schemes to compressible problems acoustic upwinding is used. In a two‐dimensional compressible flow example acoustic and conventional upwinding are combined. Evaluation of all presented results leads to the conclusion that, of the studied schemes, the implicit fifth order upwinding scheme with an asymmetry parameter of about 0.5 leads to the optimal results. Copyright © 2005 John Wiley & Sons, Ltd.     

Multidimensional upwind schemes for the shallow water equations

A multidimensional discretisation of the shallow water equations governing unsteady free-surface flow is proposed. The method, based on a residual distribution discretisation, relies on a characteristic eigenvector decomposition of each cell residual, and the use of appropriate distribution schemes. For uncoupled equations, multidimensional convection schemes on compact stencils are used, while for coupled equations, either system distribution schemes such as the Lax–Wendroff scheme or scalar schemes may be used. For steady subcritical flows, the equations can be partially diagonalised into a purely convective equation of hyperbolic nature, and a set of coupled equations of elliptic nature. The multidimensional discretisation, which is second-order-accurate at steady state, is shown to be superior to the standard Lax–Wendroff discretisation. For steady supercritical flows, the equations can be fully diagonalised into a set of convective equations corresponding to the steady state characteristics. Discontinuities such as hydraulic jumps, are captured in a sharp and non-oscillatory way. For unsteady flows, the characteristic equations remain coupled. An appropriate treatment of the coupling terms allows the discretisation of these equations at the scalar level. Although presently only first-order-accurate in space and time, the classical dam-break problem demonstrates the validity of the approach. © 1998 John Wiley & Sons, Ltd.     

Positivity‐preserving,flux‐limited finite‐difference and finite‐element methods for reactive transport

A new class of positivity‐preserving, flux‐limited finite‐difference and Petrov–Galerkin (PG) finite‐element methods are devised for reactive transport problems.The methods are similar to classical TVD flux‐limited schemes with the main difference being that the flux‐limiter constraint is designed to preserve positivity for problems involving diffusion and reaction. In the finite‐element formulation, we also consider the effect of numerical quadrature in the lumped and consistent mass matrix forms on the positivity‐preserving property. Analysis of the latter scheme shows that positivity‐preserving solutions of the resulting difference equations can only be guaranteed if the flux‐limited scheme is both implicit and satisfies an additional lower‐bound condition on time‐step size. We show that this condition also applies to standard Galerkin linear finite‐element approximations to the linear diffusion equation. Numerical experiments are provided to demonstrate the behavior of the methods and confirm the theoretical conditions on time‐step size, mesh spacing, and flux limiting for transport problems with and without nonlinear reaction. Copyright © 2003 John Wiley & Sons, Ltd.     

A flux‐splitting solver for shallow water equations with source terms

This paper introduces a stable flux‐splitting solver for one‐dimensional (1D) shallow water equations. This solver is specifically designed to satisfy a strengthened consistency condition for stationary solutions that ensures the stability and accuracy of the scheme. It applies to channels with variable depth and width, including terms modelling friction at bottom and vertical walls. Some numerical tests by comparison to both analytical solutions and experimental measurements show the good performances of the scheme. Copyright © 2003 John Wiley & Sons, Ltd.     

Developing a unified FVE‐ALE approach to solve unsteady fluid flow with moving boundaries

In this study, an arbitrary Lagrangian–Eulerian (ALE) approach is incorporated with a mixed finite‐volume–element (FVE) method to establish a novel moving boundary method for simulating unsteady incompressible flow on non‐stationary meshes. The method collects the advantages of both finite‐volume and finite‐element (FE) methods as well as the ALE approach in a unified algorithm. In this regard, the convection terms are treated at the cell faces using a physical‐influence upwinding scheme, while the diffusion terms are treated using bilinear FE shape functions. On the other hand, the performance of ALE approach is improved by using the Laplace method to improve the hybrid grids, involving triangular and quadrilateral elements, either partially or entirely. The use of hybrid FE grids facilitates this achievement. To show the robustness of the unified algorithm, we examine both the first‐ and the second‐order temporal stencils. The accuracy and performance of the extended method are evaluated via simulating the unsteady flow fields around a fixed cylinder, a transversely oscillating cylinder, and in a channel with an indented wall. The numerical results presented demonstrate significant accuracy benefits for the new hybrid method on coarse meshes and where large time steps are taken. Of importance, the current method yields the second‐order temporal accuracy when the second‐order stencil is used to discretize the unsteady terms. Copyright © 2009 John Wiley & Sons, Ltd.     

Three-point scheme and two-point boundary formulations for unsteady diffusion with flow

^** Email: ron.smith{at}lboro.ac.uk A three-point scheme of Crank–Nicolson type and two-pointboundary formulations are derived that accurately model bothlong- and short-scale branches of gradual time-variation solutionsfor unsteady diffusion on low-resolution grids. Inadequaciesin the resolution are compensated, to second order in the rateof change, with interior and boundary weights that have elementaryfunction dependence on the grid spacing. Numerical testing suggeststhat the computations are robust to rapid variation, improvewith interior grid points at tuned uniform spacing and becomeaccurate if the design assumption of gradual time variationis satisfied.     

COLLATION AND UPWINDING FOR THERMAL FLOW IN PIPELINES: THE LINEARIZED CASE

Simulating thermal effects in pipeline flow involves solving a coupled non-linear system of first-order hyperbolic equations. The advection term has two large eigenvalues of opposite signs, corresponding to the propagation of high-speed sound waves, and one eigenvalue close to or even equal to zero, representing the much slower fluid flow velocity, which transports temperature. Standard collocation methods work well for isothermal flow in pipelines, but the stagnating eigenvalue causes difficulties when thermal effects are included. In a companion paper we formulate and analyse a new numerical method for the non-linear system which arises in thermal modelling. The new method applies to general coupled systems of non-linear first-order hyperbolic partial differential equations with one degenerate eigenvalue. In the present paper we focus on a linearized constant coefficient form of the thermal flow equations. This substantially simplifies presentation of the error analysis for the numerical scheme. We also include numerical results for the method applied to the fully non-linear system. Both the error analysis and the numerical experiments show that the difficulties that come from the application of standard collocation can be overcome by using upwinded piecewise constant functions for the degenerate component of the solution.     

MULTIDIMENSIONAL UPWIND RESIDUAL DISTRIBUTION SCHEMES FOR THE CONVECTION–DIFFUSION EQUATION

Multidimensional residual distribution schemes for the convection–diffusion equation are described. Compact upwind cell vertex schemes are used for the discretization of the convective term. For the diffusive term, two approaches are compared: the classical finite element Galerkin formulation, which preserves the compactness of the stencil used for the convective part, and various residual-based approaches in which the diffusive term, evaluated after a reconstruction step, is upwinded along with the convective term.