Efficient Inflow Generation for Wind Loading Predictions for Low-Rise Buildings

Wind loading predictions for low-rise buildings require an accurate representation of the roughness sublayer within the incoming atmospheric boundary layer. However, to correctly predict the roughness sublayer, a substantial computational cost is required, as the terrain roughness needs to be resolved. In this study, we explore a framework that combines synthetic turbulence generation at the inflow and volume forcing within the roughness sublayer to represent the effect of the roughness elements in a computationally efficient manner and reproduce a wide range of target boundary layers.

To establish the framework we first build a database of LES simulations for a range of roughness configurations representative of setups that can be obtained with the Terraformer at the University of Florida Boundary Layer Wind Tunnel. The simulations resolve the terrain roughness through an immersed boundary method and serve as the foundation for developing a model that correlates terrain roughness with boundary layer characteristics, thus allowing us to identify a suitable roughness configuration for generating specific target boundary layer conditions. Next, we integrate synthetic turbulence generation and computationally efficient volume forcing techniques to reduce the number of roughness elements that have to be represented with the immersed boundary method while still reproducing the target boundary layer flow.