Computational and experimental air pollution simulation - an integrated approach

摘要

The paper describes an integrated approach to air pollution modelling in which a computational simulation of a specific site using the techniques of Computational Fluid Dynamics (CFD) is combined with an experimental verification using tests in an atmospheric wind tunnel. The approach is specifically tailored to sites where the topography plays an important role in the air pollution dispersion. The topography is surveyed using the Global Positioning System (GPS) which results in random, unstructured surface elevation data. For the computational approach, this data is used to generate a structured grid. The computational simulation solves the Reynolds-Averaged Navier-Stokes equations using the k-ε turbulence model. The computational grid is also used to construct the wind tunnel model of the terrain. The model is instrumented for testing in a specially designed atmospheric wind tunnel test section which allows testing of any wind direction. An atmospheric boundary layer (ABL) is generated by an adjustable roughness section of the wind tunnel. The experimental results include surface pressures and flow visualisation data and are compared with computational results which include all the flow properties at each of the volume grid points. The advantage of wind tunnel testing as opposed to full-scale experimental testing is the controlled experimental conditions that can be obtained. This allows the researcher to isolate the influence of parameters. The integrated approach presented shows that computational and experimental simulation can be used to enhance each other.