The first step is to divide the solution domain into a finite number of control volumes or computational cells (grid generation). Two types of grid, namely, structured and unstructured grids, were briefly introduced in Section 1.2 (and are illustrated in Fig. 1.12). Methods of grid generation will not be discussed in this book. Some of the references useful for grid generation and some of the available grid generation tools are cited in Chapters 7 and 8, respectively. For the purpose of discussing the finite volume method, here we consider a simple, structured grid arrangement. Different methods can be employed to generate a structured computational grid (that is to select positions of computational nodes and boundaries of computational cells). However, usually boundaries of computational cells are decided as a first step and then a computational node is assigned at the center of each computational cell or control volume (CV) (as shown in Fig. 6.1a), rather than selecting node positions in a first step and assigning cell faces at the midpoints of each pair of nodes (Fig. 6.1b). It is important to note that CVs should not overlap and each CV face should be unique to the two CVs, which lie on either side of it. Here we illustrate the procedure using a Cartesian grid. A typical finite volume grid and notation is shown in Fig. 6.2.


II •




0 0

Post a comment