Numerical prediction of thermal and airflow conditions of a naturally ventilated atrium and validation of CFD models

Abstract

Natural ventilation as a major passive cooling strategy is growing in popularity, and atria, as a feature of buoyancy-driven natural ventilation, have been used with greater frequency in large modern buildings to enhance the flow regime of buildings. To have a reliable and suitable design, building simulation tools have become increasingly important in assisting designers to predict the temperature and airflow behaviours throughout atria. However, proper validation is required to ensure the accuracy of the prediction. Among the studies undertaken on atria, only a limited number have relied on full-scale experiments for validation. This study applied a full-scale four-storey atrium with solar assisted natural ventilation to validate four different turbulence models—“Realizable” k-ε, Standard k-ε, RNG k-ε, and SST k-ω—coupled with the Surface to Surface radiation model. The numerical predictions of all the computational fluid dynamic models were found to have good agreement with the experimental results for the temperature distribution and the air velocity rate. However, among the four tested models, RNG k-ε had the greatest similarity to the predicted temperature results. The results also show that direct ventilation in the atrium improves the airflow conditions of the attached space at the ground floor level with the inlet opening, even more than the atrium itself.