Center for the Built Environment

This paper describes the development of an empirical model for the prediction of wind-induced indoor air motion in naturally ventilated buildings, as needed for the assessment of thermal comfort. The model is based on correlations developed from a large set of experimental pressure and velocity data collected from architectural models in a boundary layer wind tunnel. The goal of the study was to examine and formalize the relation between indoor air motion parameters (velocities and turbulence intensities) and the external surface pressure distribution on sealed models for which data bases amt correlations are now available. This was accomplished through two series of tests. In the first, indoor air speed and turbulence intensity distributions were measured in models with openings for various wind directions and building configurations. In the second, the external surface pressure distribution was measured on a sealed model for the same building configurations. The number of architectural configurations tested, which were selected to cover a wide range of possible building-airflow interactions, included the effects of wind direction, upwind obstructions, building shape, window size, and other architectural features such as wing walls and roof overhangs. The resulting correlations, based on nearly 300 tests, predict indoor air motion with a good level of accuracy as a function of wind direction, window size, and the external pressure distribution measured on sealed models.