Even after many years of the application of numerical CFD techniques to flow through porous fences, still there is disagreement between researchers regarding the best turbulence model to be implemented in this field. Moreover, different sources claim to have achieved good agreement between numerical results and experimental data; however, it is not always possible to compare numerical and experimental results due to the lack of information or variations in test conditions. In this paper, five different turbulence models namely; K-ε models (standard, RNG and Realizable) and K-ω models (Standard and SST), have been applied through a 3D CFD model to investigate air flow behind a porous panel, under the same conditions (boundary conditions and numerical schemes). Results are compared with wind tunnel experiments. Comparison is based on the vertical velocity profile at a location 925 mm downstream of the fence along its center line. All models were capable of reproducing the velocity profile, however, some turbulence models over-predicted the reduction of velocity while it was under-predicted by other models, however, discrepancy between CFD modelling and experimental results was kept around 20%. Comprehensive description of the turbulence structure and the streamlines highlight the fact that the criterion for selecting the best turbulence model cannot rely only on the velocity comparison at one location, it must also include other variables.

This content is only available via PDF.
You do not currently have access to this content.