Forecasting of pressure coefficient for wind interference due to surrounding tall building on a tall rectangular building using CFD data trained machine learning models

This study employs Computational Fluid Dynamics (CFD) to investigate the interference effects of side tall buildings (Interfering Building) on a tall rectangular building (Primary Building). Various configurations involving an interfering building (IB) and a primary building (PB) are examined, with the IB being shifted in the x, y, and x-y directions. Pressure Coefficient (C_P) and Mean Pressure Coefficients (C_PMEAN) on the PB is calculated on the faces of PB to quantify the impact of IB locations. The results of CFD are used in training and testing of 6 Machine Learning (ML) models. In which Wide Neural Network (WNN) outputs are very similar to CFD output. The validation of WNN is done by the calculating the mean pressure coefficient on front face of PB for critical location of IB. Critical locations of IB are identified, emphasizing the importance of specific displacement scenarios on building faces. Additionally, C_PMEAN analysis highlights the heightened sensitivity of certain faces to IB positioning. Notably, out of 6 models, WNN model predicted result is best fitted with the CFD result. The combination of CFD and ML models suggest the economical and efficient approach to predicted the pressure coefficient for any location of IB without wind tunnel test.