Using Multi-scale CFD Simulations to Improve Pedestrian Comfort within Built Areas in Complex Terrains

Abstract Accurately assessing wind characteristics in complex terrains, e.g., mountainous regions and cliffs, is inherently complex due to strongly non-linear interactions between the atmosphere, the terrain, and the built environment. Understanding this interaction is crucial, nonetheless, not only for structural safety but also for pedestrian comfort whenever the problem comes to populated areas. In absence of extensive on-site measurements, which are usually hard to get, high-fidelity Computational Fluid Dynamics (CFD) simulations are the only viable option. Yet, in complex terrains, these simulations are often computationally prohibitive due to: i) the need to include a large portion of the terrain around the area of interest, often of the order of kilometres, to achieve sufficient accuracy; ii) the large scale disparity between atmospheric flow and near-surface phenomena, which prevents direct scaling of the mesh resolution with the domain size. Moving from this background, this study proposes a multi-scale CFD simulation approach to assess and improve pedestrian comfort in a newly-built resort on the island of Folegandros, Greece. First, medium-scale simulations of the island are carrie d out for 12 wind directions using the Ansys WindModeller tool, with boundary conditions derived from offshore ERA5 re-analysis wind data. This provides the annual wind rose at the resort location, which serves as input for a second set of building-scale simulations with a smaller domain and the detailed geometry of the hosting facility. After validation against on-site measurements, these simulations are used to evaluate pedestrian comfort at the facility and test different mitigation measures in critical wind conditions.