Abstract
In this paper, a study is conducted to assess the ability and the accuracy of a CFD approach concerning the behavior of a linear
heat source positioned in a test room. Firstly, the main characteristics of our numerical model (mesh generation, turbulence
model, near wall treatment, radiation model and boundary conditions) are presented after a briefly description of the test room
taken into account. In addition, we show also the significance of these issues within our CFD approach. Furthermore, comparing
the results obtained with those from experiments tests, the numerical model precision is evaluated. As a result, we notice a
correctly plume description which allow us to achieve a proper general airflow prediction in the heated room (including thermal
stratification and pollutant diffusion). On the contrary, we detect relatively important experimental – numerical differences
regarding the heat transfer at walls. Therefore, the boundary layer description as well as the radiation model has to be improved.
This must be correlated with the measurements precision in order to be able to achieve more complete and accurate experimental
data.