The aim of this paper is the study of the air quality in narrow urban canyons (ratio of the distance between two buildings to the building height less than or equal to one). This has been done by investigating two key points. The first one is the investigation of the modifications that the roof shape induces in the flow and turbulence, through the comparison of velocity fields, turbulence characteristics and air exchanges between the urban canyon and the outer flow in arrays of buildings with flat and pitched roofs. The second one is the assessment of the capability of a RANS model to correctly simulate the flow, through the comparison with laboratory water-channel simulations. Among the analysed quantities, the vertical air exchange rate ACH measures the rate of air removal from a street canyon. Results show that the pitched roof strongly modifies the flow and increases the turbulence and the air exchange between the canyon and the external flow, highlighting how the choice of the roof shape can be meaningful for building design, planning strategies and regulatory purposes.

The present work aims to assess whether a standard numerical simulation (RANS-VOF model with $k-ϵ$ closure) can adequately model experimental measurements obtained in a dam physical model. The investigation is carried out on the Sa Stria Dam, a roller compacted concrete gravity dam currently under construction in Southern Sardinia (Italy). The original project, for which a physical model was simulated, included a downstream secondary dam. However, due to both economic and technical reasons, the secondary dam may not be built. Hence, it is important to assess the flood discharge routing and energy dissipation in the modified plan. Numerical validation is performed adopting the same laboratory configuration, in presence of the downstream dam, and results show a good agreement with mean experimental variables (i.e., pressure, water level). An alternative configuration without the downstream dam is here numerically tested to understand the conditions of flood discharge and assess whether its results can give relevant information for the design of mitigation measures. The topic is of interest also from a more general perspective. Indeed, the feasibility to integrate numerical models with existing laboratory measurements can be very useful not only for new constructions but also for existing dams, which may need either maintenance or upgrading works, such as in case of flood discharge increment.