Validation of numerical simulation tools for wind-driven natural ventilation design

This paper presents a validation of airflow network (AFN) and computational fluid dynamics (CFD) simulations for a naturally ventilated office building using wind tunnel measurements as the reference for external pressure coefficients and effective airflow rate prediction. The CFD simulation model is also used to study the effect of partially open windows on the effective flow rate. This study also includes a design exercise for a naturally ventilated office building that analyses the differences in predicted average window open area for a typical weather year. The results obtained show that, for simple isolated buildings, CFD can predict pressure coefficients with less than 20% average error. For cases with interfering surrounding buildings or more complex building geometries the average error is less than 40%. Average errors in bulk flow rates are lower: typically less than 25%. The largest errors occur in effective flow rate predictions for cases where the openings are exposed to recirculations and shear driven flows.