Meshing or discretization is the process of dividing the continuous fluid domain into a discrete computational domain, allowing the generally nonlinear partial differential equations of fluid mechanics to be numerically solved (further details will be discussed in the solver theory chapter).
Generally, smaller mesh sizes will yield more detailed and accurate computational results but will increase the number of elements, thereby requiring higher computational efforts.
Before discuss deeper into meshing, it’s important to first understand several terminologies used in meshing, as illustrated in Figure 2.1 below:
Figure 3.1. Basic Mesh Nomenclature
With the following definitions:
- cell – the control volume where the domain is divided
- node or vertex – the endpoint of a grid,
- cell center – the central point of a cell,
- edge – a side boundary of a face,
- face – a side boundary of a cell,
- zone – a group of nodes, faces, and cells,
- domain – a group of node, face, and cell zones.
It is important to understand the difference between mesh and geometry terminology; some software has its own terms, but generally, the “point”, “line”, and “surface” for mesh are used in the same way as “node”, “edge”, and “face” for geometry.
Some important considerations when creating a mesh include:
- Resolution or mesh detail
- Type of mesh, and
- Hardware used for the simulation (generally, higher resolution mesh needs higher RAM to store the mesh data during meshing)