An analysis of finite-difference and finite-volume formulations of conservation laws

Abstract

ABSTRACT. Finite-difference and finite-volume formulations are analyzed in order to clear up the confusion concerning their application to the numerical solution of conservation laws. A new coordinate-free formulation of systems of conservation laws is developed, which clearly distinguishes the role of physical vectors from that of algebraic vectors which characterize the system. The analysis considers general types of equations-potential, Euler, and Navier-Stokes. Three-dimensional unsteady flows with time-varying grids are described using a single, consistent nomenclature for both formulations. Grid motion due to a non-inertial reference frame as well as flow adaptation is covered. In comparing the two formulations, it is found useful to distinguish between differences in numerical methods and differences in grid definition. The former plays a role for non-Cartesian grids, and results in only cosmetic differences in the manner in which geometric terms are handled. The differences in grid definition for the two formulations is found to be more important, since it affects the manner in which boundary conditions, zonal procedures, and grid singularities are handled at computational boundaries. The proper interpretation of strong and weak conservation-law forms for quasi-one-dimensional and axisymmetric flows is brought out.

Three-Dimensional Aeroelastic Solutions via the Nonlinear Frequency-Domain Method

“…Although equation (4.22) does not require this methodology to be employed, the approach is more intuitive and has been shown to provide good results [87]. Nevertheless, this method is easily applicable to linear or quadratic temporal discretizations, commonly employed in time-marching unsteady flow solvers, but its extension to frequencydomain solution techniques is not straightforward.…”

“…However, many obstacles prevent this condition from being fulfilled in a straightforward fashion. First, it seems particularly attractive to split the GCL in separate parts for each face of the control volume as was shown in previous work for time-marching methods [87,93]. In fact, the exact volume of the cell is known at all time instances since it is only a function of the position of the cell vertices, which is known from the mesh deformation algorithm.…”

High-order finite-volume methods for hyperbolic conservation laws on mapped multiblock grids

“…The advantage of this approach is that it is strongly conservative in the sense of [41,40], high-order accurate, and freestream-preserving. It also has the advantage of using a smoothly-varying structured grid for its underlying discretization of space.…”

Horizontal patterns of water temperature and salinity in an estuarine tidal channel: Ria de Aveiro

“…The grid is therefore defined explicitly and the equations are solved using the same procedures, irrespectively of the cell geometry. Since, the equations are solved in the form of flux divergences, this method guarantees the conservation of transported properties (Ferziger and Peric 1995;Vinokur 1989). It uses an ADI-Alternate Direction Implicit scheme for the resolution of the equations.…”