Development of a simulator for indoor airflow distribution in a cross-ventilated building using the local dynamic similarity model

In this study, the evaluation of cross ventilation is presented based on the simultaneous analysis of inside and outside conditions and for wind directions other than 0 degrees (i.e. for wind flow that is not aimed directly or normally to the inflow opening). The first part of the paper considers a conventional CFD analysis and compares the performance of the widely used k-ε turbulence model as well as the modified Durbin k-ε turbulence model. While a CFD approach can give good results it is very labour and computationally demanding. The second part of this paper consists of a description of a simplified approach based on the Local Dynamic Similarity Model. In the simplified approach, outdoor boundary conditions are established directly at the inflow opening by means of either wind tunnel experiment or CFD analysis of external flow conditions. Results of the simplified model are compared with experimental analysis and shown to give good agreement. The method is also shown to be valid for single and multi room applications.