Computational modelling of pollution dispersion in the near wake of a vehicle

Richards, Kathryn (2003) Computational modelling of pollution dispersion in the near wake of a vehicle. PhD thesis, University of Nottingham.

[img]
Preview
PDF
10Mb

Abstract

The feasibility of using CFD to model the dispersion of pollutant in the near wake of a model vehicle was investigated through a series of experimental and computational studies. The near wake structure of the MIRA 33% scale reference vehicle (fastback) was measured using Particle Image Velocimetry (PIV), and hotwire anemometry and the dispersion of a tracer measured using Flame Ionisation Detectors (FIDs). The experimental data not only provided insight into the dispersion character of the model vehicles's near wake but more importantly produced data for the validation of the numerical simulations of the measured near wake and dispersion fields. The numerical simulations of the near wake flow field were conducted using the CFD code STAR-CD with the standard, RNG, Chen and nonlinear (quadratic) k- e models in combination with Upwind (UD), Linear Upwind (LUD) and the Monotone Advection and Reconstruction (MARS) differencing schemes. Validation showed the predicted flow field to be in good agreement with the measured flow field. Using the numerical flow field predictions as a foundation the dispersion of a passive gaseous pollutant was simulated by modelling the dispersion of scalar quantity released into the computational domain using a fluid injection technique. The numerical predictions of both mean velocity and concentration fields were validated against the experimental data using various statistical validation techniques. Several short investigations into the influence of vehicle speed and exhaust mass flow rate were also conducted to further assess the applicability and use of the technique in investigating dispersion in the near wake of a vehicle. Relative successes in both the velocity field and dispersion simulations were demonstrated in making predictions of the mean velocity and concentration fields. However there is clearly the need for more development and in particular the application of time-dependent techniques for the underlying velocity solutions in order that peaks in mean concentration resulting from build-up due to unsteadiness in the flow field are fully captured. Nevertheless the study demonstrated the potential for the use of STAR-CD to investigate and understand in more detail the dispersion of pollution close behind a road vehicle and possibly assess the concentration levels, at the road side of different pollutants released.

Item Type:Thesis (PhD)
Supervisors:Wright, N.G.
Uncontrolled Keywords:pollution dispersion computational fluid dynamics CFD car modelling
Faculties/Schools:UK Campuses > Faculty of Engineering > School of Civil Engineering
ID Code:108
Deposited By:Nigel Wright
Deposited On:13 Jul 2005
Last Modified:06 Feb 2009 14:43

Repository Staff Only: item control page