The interference index and its prediction using a neural network analysis of wind-tunnel data

The statistical characteristics of the fluctuating pressures acting on bluff bodies in the region of flow separation are investigated in this paper. Probability density functions of the pressure fluctuations have shown a significant non-Gaussian behaviour. The lognormal density function is found to be in close agreement with the obtained experimental data. A probability distribution for the maximum magnitude of peak pressures expected during a period is proposed and the agreement between the experimental data and predicted results are good. The effects of free-stream turbulence intensity and integral scale on the statistical nature of pressure fluctuations are presented and discussed. The second part of this paper concerns the action of fluctuating wind pressure on building cladding. The effect of free-stream turbulence intensity and integral scale in model test on equivalent constant pressures for the purpose of cladding design is investigated, making use of direct integration of measured pressure data in the wind tunnels.(C) 1999 Elsevier Science Ltd. All rights reserved.