Experimental and numerical prediction of indoor air quality.

This paper is based on a dual approach (experimental and numerical) in order to predict the indoor air quality for small ventilated enclosures. The experimental part employs a ventilated test room and a tracer gas technique (constant method as gas injection) to estimate the diffusion of a pollutant. The gas used is the sulphur hexafluoride (F6S). The numerical approach is a CFD simulation, adding a convection - diffusion equation (to determine the local mass fraction of the pollutant) to the equations normally used to solve a turbulent flow. As the injected quantity of the gas is extremely low, the convergence of the pollutant transport equation can be only reached after an important CPU. Therefore we propose a strategy for the initialization of the problem in terms of tracer gas concentrations using the experimental data. This allows us to obtain important savings in time simulation. The experimental and numerical concentration tracer gas fields are compared for isothermal and non-isothermal conditions. Further, in order to characterise in a more general manner the ventilation efficiency and the indoor air quality of the two cases studied, we evaluate the ventilation efficiency index based on the experimental and CFD values. Moreover, we add to these results the values obtained by an improved zonal model. The results achieved reveal a good agreement between the experiment and CFD calculation. On the contrary, there are important discrepancies between the ventilation efficiency index values predicted by the zonal model and the values based on the experimentation.