ABSTRACT
A methodology using a reduced-scale air model to predict the temperature and airflow patterns in naturally ventilated buildings is presented. Three cases for naturally ventilated buildings are investigated using the reduced-scale air model: buoyancy-driven (no wind), wind assisted (balance of buoyant and wind forces), and wind-driven. Through dimensional analysis and similitude, the air model looks to meet the criteria of geometric, thermal, and dynamic similarity. Computational fluid dynamics (CFD) is used to validate the temperature variation throughout the model and the airflow patterns and velocities, given the boundary conditions obtained through experiment. Results obtained through experiments with the reduced-scale air model are then compared with a prototype full-scale building.
