On the use of deterministic and intelligent techniques to predict the air velocity distribution on external openings in single-sided natural ventilation configurations

Four single-sided natural ventilation experiments were carried out in a tes t cell, a full-scale outdoor facility in Athens, Greece. During the experim ents air velocity data at various heights at the vertical centerline of the opening were collected. Various mathematical approaches were used for the prediction of the air velocity at the door level, using experimentally deri ved parameters as inputs. Deterministic as well as intelligent techniques w ere applied in an attempt to determine the potential of each of these appro aches in predicting the air velocity with accuracy. In the frame of the fir st category, correlation techniques, pressure models and CFD models were st udied. The intelligent techniques involved an application of fuzzy theory f or the prediction of the air velocity at the opening level. The response of each of the above methodologies is discussed in this paper. (C) 1999 Elsev ier Science Ltd. All rights reserved.