Flows and pollutant dispersion over flat rural terrain have been investigated for decades. However, our understanding of their behaviours over urban areas is rather limited. Most cases have either focused on street level or in the roughness sub-layer (RSL) of urban boundary layer (UBL). Whereas, only a handful of studies have looked into the coupling between street-level and UBL-core dynamics, and their effects on pollutant dispersion. In this thesis, computational fluid dynamics (CFD) is employed to examine the flows and pollutant transport in and over urban roughness. Idealised two-dimensional (2D) street canyons are used as the basic units fabricating hypothetical urban surfaces. A ground-level passive and chemically inert pollutant source is applied to simulate the flows and pollutant dispersion over rough surfaces in isothermal condition. Large-eddy simulation (LES) with the one-equation subgrid-scale model is used to solve explicitly the broad range of scales in turbulent flows. Arrays of idealized street canyons of both uniform and non-uniform building height are used to formulate a unified theory for the flows and pollutant dispersion over urban areas of different morphology. The geometry of roughness elements is controlled by the building-height-to-street-width (aspect) ratio (0.083 ≤ AR ≤ 2) and/or the building height variability (BHV = 0.2, 0.4 and 0.6), in which the characteristic regimes of skimming flow, wake-interference and isolated roughness are covered. A detailed analysis on the roof-level turbulence structure reveals parcels of low-speed air masses in the streamwise flows and narrow high-speed down-drafts in the urban canopy layer, signifying the momentum entrainment into the street canyons. The decelerating streamwise flows in turn initiate up-drafts carrying pollutants away from the street canyons, illustrating the basic pollutant removal mechanism in 2D street canyons. Turbulent transport processes, in the form of ejection and sweep, are the key events governing the exchanges of air and pollutant of street canyon. Air exchange rate (ACH) along the roof level is dominated by turbulent transport, in particular over narrow street canyons. The LES results show that both the turbulence level and ACH increase with increasing aerodynamic resistance defined in term of the Fanning friction factor. At the same AR, BHV greatly increases the friction factor and the ACH in dense built areas (AR ≤ 0.25). The turbulence intensity is peaked on the windward side of street canyons that does not overlap with the maximum velocity gradient near the leeward building corners, suggesting the importance of background turbulence in street-level ventilation. Over the building roughness, pollutant plume dispersion after the ground-level area source in cross flows resumes the self-similar Gaussian shape in the vertical direction in which the vertical plume coverage is proportional to the square root of downwind distance in the streamwise direction. Moreover, the vertical dispersion coefficient is proportional to the one-fourth power of friction factor over idealised street canyons. Conclusively, friction factor can be used to parametrise ventilation and pollutant dispersion over urban areas. / published_or_final_version / Mechanical Engineering / Doctoral / Doctor of Philosophy

No description available.

No description available.

No description available.

A large-eddy simulation methodology (LES) has been developed for predicting the turbulent physics of an incompressible flow in simple and complex geometries. The Cartesian form of the governing equations was first verified, and then later used to investigate a three-dimensional shear-driven cavity flow. The investigation involved Reynolds numbers of 2000, 3200, 5000 and 10000 and focused on the unsteadiness and turbulent characteristics of the flow. At the low Reynolds numbers (Re ~ 5000) where the cavity flow is fully laminar, direct numerical simulations (DNS) were conducted whereas the LES methodology was adopted to predict the cavity flow at the higher Reynolds number (Re = 10000). Determining the parameters in the damped subgrid scale (SGS) turbulence model for this complex flow was guided by the DNS results at Re = 5000. The SGS model was also verified against DNS results at Re = 7500 where the cavity flow was known through laboratory experimentation to be locally transitional. The LES results using the damped SGS model verified the published experimental evidence as well as uncovered new flow features within the cavity. LES computations were also carried-out of the three-dimensional shear driven cavity flow at a high Reynolds number where the SGS turbulent field was represented by a dynamic model. Lilly's least-squares expression was tested for determining Smagorinsky's coefficient in the model without ad hoc measures such as ensemble-averaging or filtering. However, zero cutoff of negative total viscosity (kinematic plus turbulent eddy viscosity) was necessary to maintain stable solutions. A discretized filter function was derived for the test filter. Both qualitative and quantitative comparisons to experimental data show that the dynamic model performed quite well. The dynamic model gave better comparisons to the experimental evidence than the damped model did. Vortex formation in the wake of a circular cylinder and their subsequent downstream transport was also numerically investigated by LES. Here however, the curvilinear form of the governing equations was necessary to perform the computations. A new generalized dynamic model was derived to represent the SGS stress field in the curvilinear space. This new model introduced the contravariant velocity components as part of the field variables. New downstream boundary conditions were also formulated to permit the shed vortices to exit with minimum disturbance. The focus of the investigation was at Re = 5600 with some verification of the computations at Re = 200 and Re = 3000. At all of these Reynolds numbers. the upstream boundary layer was laminar but the wake was fully turbulent at Re = 3000 and Re = 5600. The LES results of the many interesting characteristics of the wake showed good comparisons to the experimental data. / Ph. D.

In view of the worsening air quality in the world, more concerns are focused on the environment. This thesis uses the technique of CFD and develops the computer model to investigate the wind and pollutant transport, as well as the chemistry of reactive pollutants in idealized two-dimensional (2D) street canyons. Three scientific questions are raised in this thesis. The first task is to find out the po- sition with the most favorable pollutant removal along the ground level over 2D idealized street canyon of different building-height-to-street-width (aspect) ratios (ARs). The di- mensionless parameter, C, represents the pollutant removal performance. In the isolated roughness regime, the two local maximum C locate at the reattachment point and the windward corner. In the wake interference regime, C is peaked on the windward side. The number of vertically aligned recirculations depends on the street depth in the skimming flow regime. The sizes of the secondary recirculation upstream and downstream deter- mine how the maximum C shifts from the street centre. After identifying the position of peaked pollutant removal rate at the ground level, the emission source should be placed with the highest constant C in order to remove the pollutants upward more quickly to safeguard the street-level air quality. After understanding the best pollutant removal in the street canyon of different ARs, the second task is to find out what AR is the most favorable for the ventilation and pollutant removal across the roof level. The three parameters, namely friction factor, air exchange rate (ACH) and pollutant exchange rate (PCH), are introduced to quantify the pressure difference to sustain the mean flow, the ventilation and pollutant removal, respectively. The turbulence contributes more than 70% to the total ACH and PCH in all the three flow regimes. By increasing the atmospheric turbulence in building geometry as well as the surface roughness, the ventilation and pollutant removal performance can be improved. The linear relation between the friction factor and ACH demonstrates the larger resistance that in turn promotes the air exchange over the roof level. The physical dispersion is studied; however atmospheric pollutants are seldom in- ert but chemically reactive instead. The last task is to include the three common air pollutants, NO, NO2 and O3, in the simple NOx ?O3 mechanism in terms of the photo- stationary state and reaction rates. The Damkohler numbers of NO and O3, DaNO and DaO3, are parameterized by the concentrations of the sources NO and O3. The normalized mean and fluctuation NO, NO2 and O3 are separately considered. The integrated pho- tostationary state (PSS) in the first canyon increases with DaO3 under the same DaNO. The integrated PSS of the second to the twelveth street canyons are compared with each case, the monotonic increase in the PSS from the second to twelveth canyon is perceived in DaNO/DaO3 1, 0.03, 0.02, 0.001 and 0.000333. Further decreases the DaNO/DaO3 to 0.000143, 0.000125, 0.000118, 0.000111 and 0.0001, the PSS is found to be non-linear and the trough appears in the fourth and fifth canyons. / published_or_final_version / Mechanical Engineering / Master / Master of Philosophy

No description available.

Yuen, Man Chun. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2013. / Includes bibliographical references (leaves 58-[60]). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstracts also in Chinese.

In the present study syngas-air diffusion flames are simulated using LES with artificial neural network (ANN) based chemical kinetics modeling and the results are compared with previous direct numerical simulation (DNS) study, which exhibits significant extinction-reignition and forms a challenging problem for ANN. The objective is to obtain speed-up in chemistry computation while still having the accuracy of stiff ODE solver. The ANN methodology is used in two ways: 1) to compute the instantaneous source term in the linear eddy mixing (LEM) subgrid combustion model used within LES framework, i.e., laminar-ANN used within LEMLES framework (LANN-LEMLES), and 2) to compute the filtered source terms directly within the LES framework, i.e., turbulent-ANN used within LES (TANN-LES), which further dicreases the computational speed. A thermo-chemical database is generated from a standalone one-dimensional LEM simulation and used to train the LANN for species source terms on grid-size of Kolmogorov scale. To train the TANN coefficients the thermo-chemical database from the standalone LEM simulation is filtered over the LES grid-size and then used for training. To evaluate the performance of the TANN methodology, the low Re test case is simulated with direct integration for chemical kinetics modeling in LEM subgrid combustion model within the LES framework (DI-LEMLES), LANN-LEMLES andTANN-LES. The TANN is generated for a low range of Ret in order to simulate the specific test case. The conditional statistics and pdfs of key scalars and the temporal evolution of the temperature and scalar dissipation rates are compared with the data extracted from DNS. Results show that the TANN-LES methodology can capture the extinction-reignition physics with reasonable accuracy compared to the DNS. Another TANN is generated for a high range of Ret expected to simulate test cases with different Re and a range of grid resolutions. The flame structure and the scalar dissipation rate statistics are analyzed to investigate success of the same TANN in simulating a range of test cases. Results show that the TANN-LES using TANN generated fora large range of Ret is capable of capturing the extinction-reignition physics with a very little loss of accuracy compared to the TANN-LES using TANN generated for the specific test case. The speed-up obtained by TANN-LES is significant compared to DI-LEMLES and LANN-LEMLES.

published_or_final_version / abstract / toc / Mechanical Engineering / Master / Master of Philosophy