Numerical investigation of wind load effects due to downburst and atmospheric boundary layer flow over building

• Main Author: Sim, Tze Siang
• Other Authors: -
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Aeronautical engineering
• Online Access: https://hdl.handle.net/10356/164972
• Institution: Nanyang Technological University

The wind load effects and flow field of the downburst and Atmospheric Boundary Layer (ABL) in urban terrain are numerically investigated in 2D and 3D simulations using Computational Fluid Dynamics. A downburst is a non-stationary wind hazard to structures on the ground. Upon impinging against the surface of Earth, the wind flows radially near the surface of the Earth at high speeds, culminating in structural failure to buildings. Previous experimental and numerical studies focus on the investigations of the flow field characteristics and the wind load acting on isolated buildings. The present work addresses the six different aspects of the wind load effect of a downburst on buildings, that is i.e. (i) the effects of the Reynolds numbers and the numerical modelling of microburst-like wind and the wind drag acting on an staggered arrays of buildings; (ii) the effects of microburst-like wind direction on the wind drag acting on a city consisting of low-rise building arrays and high-rise building; (iii) The effects of roof-corner radius of buildings with flat roof and hip roof on drag; (iv) a numerical study of the influence of street canyon characteristics on wind drag reduction on an idealised urban morphology acted upon by thunderstorm microburst; (v) a numerical study of microburst-like wind load acting on different block array configurations modelled using impinging jet as a model; (vi) the temporal behaviour of the mean pressure field of an impinging jet for modelling dry microburst-like wind using URANS. Secondly, the environment in the urban terrain around buildings is also significantly influenced by the lower region of the Atmospheric Boundary Layer (ABL). To the author’s knowledge, the wind load effect of concave roof building is not often been numerically investigated precisely. In addition, the use of hybrid RANS/LES technique for simulating ABL flow around building is suitable for wind engineering but is not well studied currently. The present work also investigates (A) The wind load effect of a concave roof on high-rise building, (B) Comparative study on the hybrid LES/RANS approach for modelling turbulent boundary layer flow around a 1:1:2 building. Therefore, the wind load effects of the concave roof building and the practicality of some common hybrid RANS/LES models, such as the Delayed detached-eddy simulation (DDES), Improved-Delayed-Detached-Eddy Simulation (IDDES) and scale-adaptive simulation (SAS), are investigated in the current work. An expression is proposed to estimate the microburst’s spatially-averaged wall shear stress along the centreline of the staggered array that model the wall shear stress of the staggered array (idealised urban array). The estimation is about 14.6% - 27.6% deviating from the computed results for the investigated array. The viscous drag force relative to the total drag force for various packing densities is insignificant to the total drag force for microburst-like wind, and therefore it can be neglected. This finding will provide a preliminary assessment of the error uncertainty present in the ‘pressure-tap’ experimental method for determining the total drag exerted on staggered arrays by microburst-like wind in laboratory via CFD approach. For the concave roof study, when the sag depth is constant, the increase in building height from 0.12m to 0.16m will produce a more significantly higher CP at the windward façade than an increase of U∞ from 5m/s to 15m/s (i.e. 200% increment), whereas the magnitude of CP at the leeward façade changes slightly. From a design wind load perspectives, more emphasis should be placed on the building height. From the investigation of the hybrid turbulence models, the SAS model produces the best agreement with the experiment as compared to the other hybrid turbulence models investigated.