Khaled, may I suggest that you be very careful about using CFD in bluff-body architectural aerodynamics to obtain design peak cladding pressures. The turbulence models (RANS, LES, ect.) have varying degrees of success; particularly peak negative pressures prove very difficult to get right. This is not the place to describe all the problems associated with a complex bluff-body cityscape in attempts to use Computational Wind Engineering (CWE), but I have a paper in JWE&IA (my co-author Russ Derickson) outlining many of the issues. More recent publications by Bert Blocken and Robert Meroney, both in B&E, would also be a wise read before doing too much on the topic. Many are working on CFD/CWE for building studies, but success is rare due to closure issues and coarse grids. I saw a presentation last year at the APCWE in Aukland by some Taiwanese researchers and they had some very impressive comparisons with wind-tunnel data. However, the huge number of cells (hundreds of millions from memory) meant each wind direction run took a couple of weeks to complete. I don't want to dwell, but read carefully the limitations - way of the future, but many users of CFD in architectural areodynamics are missing the mark when compared to full-scale or wind tunnel. LC