Numerical simulation of particle-laden turbulent flows-Environmental applications

In first part of the thesis a detailed study of the particulate pollutant distribution by wind flow over a building in an urban area was performed. The accuracy of RANS-RSTM and LES turbulence models for predicted airflow over a square cylinder was first evaluated. These models are then applied for simulating wind flows over the scale-model of the Center of Excellence (CoE) Building. Comparing the simulation results with the experimental data of Kehs et al. (2009) showed that the RSTM predicted the pressure distribution on the building consistent with the measurements, but it could not capture the details of the airflow velocity field around the building. The LES simulation, however, showed good agreement with the PIV data. The LES model was then used for analyzing the particulate pollutants transport and deposition analysis.;Particle motion was modeled using a one-way coupling, Lagrangian approach. Particular attentions were given to the effect of the turbulent velocity fluctuations on particles dispersion and deposition. Instantaneous turbulent velocity fluctuations were simulated using the Langevin stochastic differential equation. The particle transport model in turbulent flows was validated by comparing the predicted deposition velocity for vertical and horizontal channel flows with the existing experimental data and numerical simulation results. Finally the particulate pollutant dispersion and deposition around the scaled CoE Building were investigated using the LES and unsteady particle tracking approach.;In addition, the size-concentration distribution of secondary organic aerosols (SOAs), as an indoor air aldehyde pollutant, was numerically modeled. The population balance equation of the SOAs was solved using the method of moments (MOM). To close the model, particle size distribution was assumed to follow a lognormal distribution, which was based on the experimental data of Chen and Hopke (2009). The nucleation of SOAs from the chemical reaction of &agr;-pinene (a common emission from indoor furniture), and ozone in the air, as well as, their Brownian coagulation and the surface growth were considered in the numerical model. The computational model was evaluated by comparison with the experimental data of Chen and Hopke (2009).;The MOM was used for modeling the distribution of the SOAs in an office space. The concentrations of SOAs in the breathing zone of an occupant in the room were evaluated for two mixed-mode ventilation systems. The simulation results showed that the pollution concentration in the ventilation system with the air outlet placed in the ceiling was smaller than the one in which the air outlet was in the floor behind the manikin model.