Author(s)

MATHEUS ALMEIDA, MOHAMMAD JADIDI, MAHDI AZARPEYVAND

Abstract

This study investigates the role of upstream roughness representation in atmospheric boundary layer (ABL) flow around a surface-mounted cubical building using a combined experimental and numerical approach. Wind tunnel measurements were performed in the University of Bristol boundary layer wind tunnel using planar particle image velocimetry (PIV) for a cube fully immersed in a neutrally stratified ABL. Two high-resolution lattice Boltzmann method large eddy simulations immersed boundary method simulations, each using approximately 200 million cells, were carried out. In Case A, the upstream roughness elements were explicitly resolved and combined with synthetic eddy method inflow, whereas in Case B the inflow developed over a short upstream fetch without geometrically resolved roughness. The resolved-roughness case reproduced the incoming ABL profiles and turbulence intensity with substantially better agreement against the PIV data and provided more accurate predictions of the windward stagnation location, wake reattachment behaviour and canonical wake structure. By contrast, the short fetch case showed weaker agreement with the measured inflow, larger errors in key flow features and a visibly distorted recirculation region. Although both simulations reproduced the inertial subrange behaviour in the wake, the dominant shedding characteristics differed between the two cases, indicating a strong sensitivity of bluff body aerodynamics to upstream boundary layer development. Overall, the results demonstrate that explicit resolution of upstream roughness is essential for realistic prediction of ABL flow around bluff bodies and provides a more reliable high-fidelity framework for urban flow applications such as ventilation, pollutant dispersion and pedestrian wind comfort.

Keywords

atmospheric boundary layer (ABL), wind tunnel experiments, large-eddy simulation (LES), lattice Boltzmann method (LBM), upstream surface roughness, cubical building