Sophia Buckingham, Ariane Frere, Cecile Goffaux, Denis Jacquet, Laurent Geron4
Year:
2013
Bibliographic info:
Building Simulation, 2013, Chambéry, France

In order to handle a case of pollutant dispersion in an urban environment Computational Fluid Dynam-ics (CFD) emerges as a fast and flexible method. By coupling it to a parametric optimization approach, the present study aims at simulating the photocatalytic de-struction of NOx particles as they hit the coated walls of an isolated building. This approach permits to eval-uate the impact of different design parameters as the wind speed and the sun exposure on the air quality. The first and critical step of the study presented in this article is to verify the CFD model. The focus is hence on determining, based on a comparison to a bench-mark, the most appropriate physical model and spatial resolution to correctly represent fluctuating quantities, natural convection and particles diffusion. One impor-tant conclusion is that wall resolved modelization is required, as using wall functions leads to high over-estimation of the pollutant destruction. After this first step, a design of experiments is carried-out to highlight the coating’s ability to reduce pollutant levels accord-ing to its intrinsic properties, wind intensity and solar radiation. Effects of these parameters on the convec-tive heat transfer coefficients are studied as well. Two metamodels are build, based respectively on 2D and 3D simulations. It appears that 3D effects cannot be neglected and that the freestream velocity is the most contributing factor in the air depollution rate, far ahead of the coating intrinsic depollution characteristic.