[University home]

School of Mechanical, Aerospace and Civil Engineering

What our graduates say

Top performing MSc student 2009-2010

Aerofoil

Sam Gorji was awarded a prize of £100 and a certificate in recognition of his excellent performance on the MSc program in 2009-2010.


PhD student:

Sam Gorji

Project title: A Study of Pollution Dispersion in an Isolated Two-Dimensional Street Canyon
Degree: MSc in Thermal Power and Fluids Engineering
Country: Iran
Expert group: Computational Fluid Dynamics
Awards: Top performing MSc student 2009-2010
Supervisor: Professor D Laurence

I did my BSc in Mechanical Engineering (Solid Design) in Azad University of Takestan. Afterwards, I came to University of Manchester to continue my studies in the Mechanics of Fluids. University of Manchester is one the leading institutions in Turbulence Mechanics and CFD. I had the chance to learn many things, discuss issues and be judged by professionals in CFD and Turbulence Mechanics. I enjoyed the whole year. At the moment I am working on an EPSRC PhD project in the University of Aberdeen. I work on an experimental project on unsteady turbulent flows over rough surfaces.

 

Project abstract

Geometry - Elevations and key quantities for 2-D abstraction of cityscape
   Geometry                        Flow Streamlines               

Pollutant concentration

  Turbulent kinetic energy   Pollutant concentration

A Study of Pollution Dispersion in an Isolated Two-Dimensional Street Canyon


In the present work, numerical simulations have been applied to investigate the flow field and pollutant dispersion inside an isolated street canyon. Two-dimensional numerical model based on the Reynolds averaged Navier-Stokes and species transport equations were applied with different turbulence closures. The performance of two layer zonal and low-Reynolds number turbulence models were examined on different free stream velocities of atmospheric boundary layer.

The two layer zonal turbulence models employed in this study are realisable, Renormalization Group (RNG), standard k-ε and Reynolds stress transport model (RSM) schemes. The performance of low-Reynolds number k-ε Launder-Sharma model was also examined and corresponding comparisons were outlined for each of the models. The experimental database obtained from the atmospheric boundary layer wind tunnel at the Meteorological Institute of Hamburg University, Germany (Journal of wind engineering and industrial aerodynamics 62 (1996) 37) was used for validation purposes of the pollutant concentrations on the edges of the canyon. In this study, it has been shown that computational fluid dynamics (CFD) can be applied for determination of flow and pollutant concentration fields in meteorological application.

The results obtained from these simulations have shown the applicability and robustness of CFD for determination of air quality in urban areas.

Figure: Elevations and key quantities for 2-D abstraction of cityscape

 

In this study, the aspect ratio of unity was set for the geometry configuration. It is shown that in such geometry the main task of pollutants dispersion is worked out by a clockwise vortex formed inside the canyon. The centre of this vortex is close to the centre of the canyon. While other two tiny vortices generated at the corners tend to transport the pollutants from the source to the bottom of the leeward edge.
Dimensionless concentration number which was calculated for the validation purposes shows insignificant variations to the changes made in the inlet wind velocities. However it has been shown that by increasing the wind velocity the pollutants concentrations within the canyon decreases and therefore more pollutants are escaped from the canyon.

Concerning the turbulence models performances, it is shown that all of the models capture the qualitative trend of the concentration distributions. All of the models employed in this study have underestimated the concentration trends on the edges of the canyon. However the performance of these models have been different in each wind velocity. This is while the RSM scheme shows relatively better performance among the rest and predictions of the Launder-Sharma k-ε model were found to be the poorest.

Top of page