Numerical and physical modeling of bluff body flow and dispersion in urban street canyons

Abstract

This investigation is focused on bluff body flow, wind loads and transport of pollutants or chemical and biological (CB) agents in urban environments, which require accurate measurements of the basic flow fields for carefully controlled, well known conditions. The research consists of two components: physical urban street canyon modeling in a boundary layer wind tunnel and numerical urban street canyon modeling using finite-volume numerical methods. Fluid modeling in an industrial wind tunnel provides an opportunity to produce accurate simulations of the bluff body flow and transport of urban pollution or simulations of CB agents associated with urban terrorism incidents. A basic building shape, the Wind Engineering Research Field Laboratory building (WERFL) at Texas Tech University, is used for this study. Many studies have been performed on pressure fields, flow and dispersion patterns based on the isolated WERFL building during the past ten-year Colorado State University/Texas Tech University Cooperative Program in Wind Engineering. An urban street canyon matrix was built by using the 1:50 scale WERFL model and surrounding it by models of similar dimensions. These buildings were arranged in various symmetric configurations with different separation distances and different numbers of buildings upwind, laterally and downwind. A series of measurements were made over the generic urban street canyon arrangements using flow visualization, anemometry, pressure transducers and gas chromatography. Experimental data include visualization, video sequences, velocity and turbulence intensity profiles, surface pressure on the building and dispersion of tracer gas. Results were examined for the influence of surroundings on the pressure and concentration distributions. Results were also compared to three-dimensional numerical models of the same configuration using the commercial code, Fluent 5.4 and the Large Eddy Simulation (LES) code Fire Dynamics Simulator (FDS). The effects of grid resolution, boundary conditions, source placement and selection of turbulence model (kappa-epsilon, RNG kappa-epsilon, Reynolds stress, LES, etc.) were examined in a series of sensitivity calculations.