As in any CFD approach, you can calculate the drag/lift by integrating the stress tensor along the surface of the obstacle. There exists however another method, the momentum exchange method, which is typical for lattice Boltzmann. It works for bounce-back boundaries and is straightforward. Simply evaluate the balance between incoming momentum (the incoming f’s times their lattice vector) and the outgoing momentum.

The momentum exchange method is for example documented in this review paper[/url]. Furthermore, I have written an article in which I use the momentum exchange method to compute the drag force acting on a rectangular obstacle. I saw at the ICMMES 08 that Bo Liu has extended the work in this paper (see B. Liu talk in [url=http://icmmes.org/media/program/ICMMES%202008%20program%20FINAL.pdf]the ICMMES program on page 8) and achieved higher precision. You may want to contact him if you are interested in his results.

I am simulating flow over an asymmetrically placed cylinder in a channel. Inlet velocity is 0.1, rho = 1.0, Re = 100. At the inlet and outlet, Zou/He pressure boundary conditions are applied. At the channel walls, the bounce-back boundary condition are used. At the cylinder boundary, Bouzidi interpolation-based model is performed. The momentum exchange method is applied to calculate lift and drag coefficients. Finally, I got the result of coefficients but they are very large (normally, Cd_max = 3.2 and Cl_max = 1.0). Can anybody give me a help and show me what is wrong in my work? Thank you very much.