Dynamical computational fluid dynamics modeling of the stochastic wind for application of urban studies

For decades, the borders of building studies were restricted up to the exterior walls. With better understanding the side effect of urbanization on human health and with excessive progress in new investigation tools, the area of building studies were enlarged to neighborhood environment where mass and air transport vividly interact with the buildings. Unlike the indoor studies, one can emphasize the significance of the stochastic wind in outdoor studies. Many experimental and simulation research have been conducted to verify the contribution of the wind velocity over/within street canyons/buildings. These works were mainly designed to observe the relations between wind flow and natural ventilation, pollution dispersion, pedestrian thermal/wind comfort, and drag resistance of the buildings. All these efforts continuously proved the significance of vortices caused by upstream wind as they interact on physical phenomena within the street canyons. However, the main drawback of these studies is not attributed to the extracted results, but the way in which wind, an inherently transient and stochastic phenomenon, is presented by a steady state consideration as applied through the boundary conditions. For example, upstream wind in terms of direction and magnitude is widely considered as a constant and steady state profile induced through the street canyons. However, due to the transient and stochastic nature of encountered wind in the urban environment, these assumptions fail to capture the physics of the problems. Therefore, the main point of concern in street canyon modeling should be described as constant and steady state reflection of stochastic and transient phenomena through the boundary conditions. Although there are existing studies related to the application of transient boundary conditions, a practical procedure to model the dynamic and stochastic behavior of wind direction is barely addressed in literature. In light of the lack of stochastic wind modeling, this study intends to introduce for the first time an approach in order to generate dynamic wind. This is followed by a brief discussion of existing approaches in urban-scale wind modeling and their major shortcomings. For this purpose, a computational Fluid dynamics (CFD) model is developed considering a novel cylindrical computational domain. An analytical and a parametric study have been also conducted to obtain suitable sizes for the domain geometries. The advantages of the proposed model comparing to the traditional approaches are depicted using a case study of array of cuboids.