Análise e modelagem matemática da evolução temporal de emissões contínuas em um dossel urbano idealizado

Abstract

The present study evaluates the flow and dispersion of a scalar inside an array constituted by obstacles of rectangular dimensions. Numerical simulations are performed using a transient street network model, in whose basic idea is to divide an urban area into connected boxes, which may be streets or intersections. The purpose of the model is to provide the average concentration values within each location. It is known that in real situations, wind speed can be highly variable in time and space. Thus, modeling this information is not simple. When it comes to approaches based on computational fluid dynamics (CFD), results obtained from direct numerical simulations (DNS) are more representative of reality, since in this method all scales of a turbulent flow are resolved. Therefore, in this study, the flow and dispersion of a passive scalar through a network of streets idealized from data from DNS were evaluated, where initially, an analysis of the general characteristics of the dispersive process within the array was carried out, through the evaluation of the temporal evolution of velocities responsible for advective transport. It was verified that the velocities present a highly three-dimensional and fluctuating character among the streets of the array, with the greatest fluctuations occurring in street intersect regions. In addition, the results showed the influence of topological dispersion on the distribution of the climbing plume within the canopy, in which the intersection regions act by amplifying transport to the subsequent streets. In view of the above, the applicability of the street network model was verified based on comparisons with DNS data, demonstrating its ability to predict concentrations and the general behavior of the plume, where the main aspects associated with dispersion were evidenced. In view of this, in order to identify the influence of velocities on the scalar dispersion, reference data from DNSsimulations were adopted to obtain the results of the street network model. Being the velocities characterized in four different heights, 0.03125H, 0.25H, 0.5H and 0.75H. The concentration levels were generally more significant in places around the emission site. In addition, it was observed that as the velocity input data were modified, there were changes in the dispersive process, mainly with regard to the temporal variability of climbing between canopy streets. Finally, changes in concentration fluctuation levels within the canopy were also identified. It was evident that the concentration fluctuations were more intense in the whole area evaluated for the more stochastic velocity field, with the most significant occurring for velocities in regions close to the ground.