Computational modeling of urban wind flow and natural ventilation potential of buildings

Natural ventilation is a significant resource towards a healthy and sustainable outdoor and indoor environment. In the past, indoor ventilation has attracted a lot of attention because of its importance for indoor air quality and thermal comfort. More recently the relevance of outdoor ventilation has been stressed, not only in terms of outdoor air quality and thermal comfort, but also for passive cooling and wind energy production. Modeling natural ventilation in the built environment is a challenging task, due to the many physical and computational parameters involved. Several methods are available to predict natural ventilation: simplified and analytical models, Airflow Network (AFN) models, and numerical methods. None of those methods is universally superior but the most suitable one for each situation is that one providing the required accuracy at a reasonable cost for the whole simulation process. In particular, Computational Fluid Dynamics (CFD) combined with experiments can provide accurate predictions of the outdoor wind flow, the indoor airflow, and their interaction. Nevertheless, AFN models are usually integrated into thermal flow analysis to assess the energy performance of buildings. In fact, due to the high computational cost and simulation time required for CFD, AFN models are more often coupled with building energy simulation (BES) tools for natural ventilation analysis. This thesis focuses on the accuracy and suitability of CFD and BES/AFN models to predict urban wind flow and natural ventilation of buildings. Detailed parametric analyses are conducted using CFD to predict outdoor and indoor airflow distribution and BES/AFN models to estimate the energy performance of natural ventilation. The study is supported by wind-tunnel data that were used to validate CFD simulations and to provide wind pressure coefficient data for BES/AFN analysis. Overall, this thesis provides some physical insights on the topic, points out relevant parameters for airflow simulations and analyzes the impact of physical and computational parameters on simulation results. First, the detailed analysis of wind-induced cross-ventilation demonstrates the accuracy of 3D steady RANS simulations to predict the interaction between outdoor and indoor airflow and can support the definition of specific guidelines for cross-ventilation through large openings. Next, the urban wind flow CFD study assesses the influence of a main street on the ventilation performances of simplified urban configurations and shows the suitability of CFD as a support for urban planning and management. Last, the sensitivity of night-ventilation rates and cooling energy savings to the source of pressure coefficients and to the characteristics of the urban environment reveals the importance of accurate input data to obtained reliable evaluations of the cooling effect of natural night-ventilation.

da compilare