Detailing the effects of geometry approximation and grid simplification on the capability of a CFD model to address the benchmark test case for flow around a computer simulated person.

This paper details the use of a simplified CFD model to predict the flow patterns around a computer simulated person in a displacement ventilated room. The use of CFD is a valuable tool for indoor airflow analysis and the level of complexity of the model being investigated is often critical to the accuracy of predictions. The closer the computational geometry is to the real geometry of interest, the more accurate the corresponding results are expected to be. High complexity meshes enable elaborated geometries to be resolved. The drawback is, however, their increased computational cost. The Fire Dynamics Simulator (FDS) model (Version 5) enabled to investigate the effects of geometry and computational grid simplification on the accuracy of numerical predictions. The FDS model is based on a three-dimensional Cartesian coordinate system and all solid obstructions are forced to conform to the underlying numerical grid which is a potential limitation when dealing with complex geometries such as those of a human body. Nevertheless, the developed computational model was based exclusively on a three-dimensional rectangular geometry. At the same time, in order to limit the total number of grid cells, a relatively coarser grid than those used for similar simulations was adopted in the investigation. The developed model was then assessed in terms of its capability of reproducing benchmark temperature and air velocity distributions. The extent to which numerical results depend on different simulation settings was detailed and different boundary conditions are discussed in order to provide some guidance on the parameters that resulted to affect the accuracy of the predicted results. The comparison between numerical results and measurements showed that a simplified CFD model can be used to capture the airflow characteristics of the investigated scenario with predictions showing a favourable agreement with experimental data at least in the qualitative features of the flow (the detailed investigation of the local airflow field near the occupant can not be probably conducted apart from considering the real human geometry). Significant influence of simulator geometry and of boundary conditions was found. [ABSTRACT FROM AUTHOR]