I have noted that most texts dealing with the control volume formulation for CFD do so for a structured grid environment. Are there any books specifically directed at unstructured grids?

(1). There are many different un-structured mesh approaches. (2). The mesh requires different ways to generate and the bookkeeping will be different from that of the structured mesh generation. (3). I don't think, there is a simple rule to follow to apply to conservation equations. (4). This is not the case in the multi-block structured mesh approach, where the bookkeeping is simple, and the cell shapes are similar. (5). If you can keep your mesh to one simple shape, then it should be easier to develop the solution algorithm. If you have changing types of cells, then I guess, you will have to deal with each one separately. (6). You can check the forum/resources/mesh generation section for more information.

All of the computational infrastructure for handling unstructured meshes has been around for a long time in the finite element community.

That being the case, look among finite element books for examples demonstrating the use of connectivity arrays to handle unstructured meshes. The basic principles of handling meshes will be the same as for your finite volume code, although your local statements of conservation will be different.

Incidentally, there are many interesting parallels in matrix theory with the concept of connectivity, too.

Hi Chris,

You will find the Thesis of Hrvoje Jasak, (1996, Error analysis ans estimation for the finite volume method with applications to fluid flows. Phd Thesis University of London) very useful and interesting.

JH Ferziger & M Peric (1997, Computational methods for fluid dynamics, Springer-Verlag) spend some pages on this topic.

Hope this help