Predicting peak pressures from computed CFD data and artificial neural networks algorithm

Abstract

The goal of this paper is to predict the peak pressure coefficients by combining two simulation models, steady‐state Reynolds averaged CFD model and Artificial Neural Networks (ANN). Many previous studies have shown that CFD can predict mean pressure coefficients, Cp well if inlet profiles, grid adaptation and the turbulent model are well chosen. However, the design codes for wind loads are based on peak pressure coefficients in wind tunnel experiments. The combination of two simulation methods, CFD and ANN, allows us to predict the peak pressure coefficients. The peak surface pressure values on master WERFL models inside urban street canyons are determined by the prognostic model FLUENT using the k‐epsilon turbulence model and Artificial Neural Networks algorithm. The results are compared against fluid modeling from wind tunnel tests.

Key words:

• CFD
• artificial neural networks
• wind loads
• wind tunnel

Notes

Corresponding author. (Tel: [phone] ext. 2670; Fax: [phone]; Email: [email protected])