Validation metrics and turbulence frequency limits for LES-based wind load evaluation for low-rise buildings

Abstract

The accuracy of Large-Eddy-Simulation (LES) in modeling wind loads, especially components and cladding loads, does not currently fulfill its potential. The lack of extensive consideration for wind loading requirements and their direct influence on LES duration and mesh size are identified as research gaps. This study presents a validation metric for LES modeling of low-rise buildings, including the wind field and aerodynamics. Most of the proposed validation requirements are adapted from Boundary Layer Wind Tunnel (BLWT) testing requirements to highlight the challenges and opportunities for LES. The turbulence spectral frequency limits significantly contribute to the cost-accuracy trade-off in LES. A stationarity test of recorded time histories is proposed to determine the optimal duration, i.e., low-frequency limit, for statistical parameters other than the peak. An analytical expression is proposed for the LES mesh size estimation from the target maximum frequency limit and corresponding spectral energies. The proposed validation metric is demonstrated with LES models of a low-rise building targeting respective BLWT tests in three wind directions and in three exposures. Despite some limitations, the peak Cp error was within ±10% of the BLWT value range. The optimal normalized durations from LES were one order of magnitude shorter than those from the BLWT cases. The peak values did not show a significant difference in trend from BLWT; hence the duration dependence of peak uncertainties remains relevant for LES. The BLWT data supports the previously proposed high-frequency limit of two. LES models can produce reasonably acceptable wind loads with proper attention to the existing and new requirements, as shown in the present study.