Thanks for your reply. I am using collocated grid and primitive variable formulation. The issue is when I am trying to solve this problem for very High Reynolds no, I need to have very fine mesh around the boundaries.
I solved the problem using Finite difference but if I refine my mesh around the boundaries, I will have to perform grid transformation. And if I refine the mesh throughout the domain, it increases the computational time .
Also my ultimate objective is to solve flow field around a cylinder and an airfoil , for which I will be using FVM.
I would like to start from learning how to discretize transient 2D momentum equation for non-uniform rectangular grid
Why deny the use of calculus and vector field theory in formulating an arbitrary discretization? A Galerkin weak form in the continuum sets the stage, thereafter selecting a FE trial space basis embeds all the coordinate data, with the final step being conversion to a recursive "stencil". Optimal MODIFIED CONTINUOUS Gelerkin CFD, Wiley 2014, provides complete details.
FINITE ELEMENTS - Computational Engineering Sciences, Wiley 2012, fully details FEA methodology at an introductory level including CFD. Visit the supporting website to download a Matlab toolbox specifically designed to implement FEA algorithms.
I can recommand you differents books. If you don't have a lot of experience with Finite Volume Method. th first book you must have from my point of view (simple and easy to understand) is :
Computational Fluid Dynamics: Principles and Applications, 3rd Edition
Jiri Blazek
2015, Butterworth-Heinemann
the companion website contains the sources of 1-D and 2-D Euler and Navier-Stokes flow solvers (structured and unstructured) and other tools ...
Dear Shashank, I have used FEM in Mathematica for lid driven cavity with good results (see images in .doc file). For a use of Scilab (= Matlab), see the attached paper by Margonari. Gianluca