Wind induced peak pressures on low-rise building roofs via dynamic terrain computational methodology

Abstract

Computational Wind Engineering is expanding rapidly in the last decade and is expected to be introduced as a design tool for wind-induced loads in future code provisions and standards worldwide. The paper discusses the current CWE shortcomings of the state-of-the-art and proposes a methodology that targets to close the current research gaps, from a practical point of view. Monitoring wind tunnel experiments, significant deviations of instantaneous velocity profiles due to high turbulence intensity are expected to be critical for design pressures. This inspired the novel modeling technique - named Dynamic Terrain - that reformulates the turbulence characteristics and assumes that each instantaneous profile is produced by changing the corresponding terrain conditions. Successively derived profiles from the wind tunnel are introduced as inlet conditions in coarse LES computational domain and propagate inside the domain to interact with the building. The incident flow is successfully modelled, in terms of mean, turbulence intensity and spectral content. Mean, standard deviations and peak pressure coefficients correlate well with experimental results and the procedure is advantageous compared to similar computational techniques. The target accuracy is achieved, benefits and limitations of the method are discussed, and conclusions based on experimental and computational observations are drawn.