Prediction of pressure coefficient on setback building by artificial neural network

Abstract

The present study predicted the pressure (Cp), drag (Cfy), and lift (Cfx) coefficients on square shape and setback building models. The study considered a conventional (1:1:2) square model, a single side single-setback, and single side double-setback models. It is very much challenging to measure the different aerodynamic coefficients on the setback building models for the intermediate wind incidence angles (WIAs). At first, the study calculated the different aerodynamic coefficients by Computational Fluid Dynamics (CFD) method and then predicted the Cp of intermediate wind angles by the artificial neural network (ANN) method. The Cp for different WIAs directly exports from the Cp versus WIAs graph. The study found the double setback model is 4.26% and 0.6% more efficient to resist the drag and lift force compared to the single setback building. Finally, suggested the setback number takes an important role to control the frequency due to pressure and velocity.