Вы находитесь на странице: 1из 2

Per definition, a mesh is a network that is formed of cells and points.

It can have
almost any shape in any size and is used to solve Partial Differential Equations.
Each cell of the mesh represents an individual solution of the equation which, when
combined for the whole network, results in a solution for the entire mesh.

Solving the entire object without dividing it into smaller pieces can be impossible
because of the complexity that is within the object. Holes, corners and angles can
make it extremely difficult for solvers to obtain a solution. Small cells on the
other hand, are comparably easy to solve and therefore the preferred strategy.

Brief History
The history of the mesh and meshing techniques is closely related to the history of
numerical methods. The paper by Courant, Friedrichs and Lewy1 is arguably the
fundamental starting point for the Finite Differences Method (FDM), where concepts
such as the C.F.L. (Courant, Friedrichs, Lewy) stability condition were introduced.

Historically, rectangular and Cartesian grids are associated with the Finite
Differences, due to the easy identification of point neighbors, used in the
derivatives approximations.

The Finite Element Method (FEM) allowed for mixed types of mesh cells, and
unstructured meshes became feasible. Variational formulations being used to solve
problems numerically can date back to Lord Rayleigh and Ritz works, from the end of
the 19th century to the first decade of the 20th century2.

Introduction
The first steps for numerically solving a set of partial differential equations
(PDEs) is the discretization of the equations and the discretization of the problem
domain. As mentioned earlier, solving the entire problem domain at once is
impossible whereas solving multiple small pieces of the problem domain is perfectly
fine. The equations discretization process is related to methods such as the Finite
Differences Method, Finite Volume Method (FVM) and Finite Element Method, whose
purpose is to take equations in the continuous form and generate a system of
algebraic difference equations. The former, namely the domain discretization
process, generates a set of discrete points and cells that cover the continuous
problem domain.

square domain discretization


Figure 1: Example of a square domain discretization

A mesh is, by definition, a set of points and cells, when connected to form a
network. This network can have many forms of geometry and topology, as will be
discussed later. Often, meshes are also called grids, and that is generally related
to the intrinsic organization of the mesh and/or when those meshes are related to
Finite Differences problems.

Each cell or point of the mesh will hold a local solution of the equations,
depending on whether the equations were discretized over the cells or points. The
choice whether to discretize the equations over the set of cells or points is a
project decision, and many aspects can affect the choice.

In general, point discretization is used when the equations are approximated using
the Finite Differences Method, where the PDEs are usually approximated by Taylor
series expansions at each point�s neighbors. Sometimes the point discretization can
be used together with the Finite Volume Method, however, cells are going to be
implicitly used around points.

When the equations to be discretized are considered in the weak form, integral
form, or conservative form, it is common to solve the integrals over discrete
cells. For example, when considering transport phenomena, the Finite Volume Method
can be formulated as the discrete cells representing small volumes. Then, the
fluxes can be balanced through those cells, while supposing that the solution is
constant inside them.

Вам также может понравиться