Wind Tunnel Pressure Series Statistics for the Case of a Large Span Canopy Roof

This paper investigates pressure coefficient and peak factor statistics for a hyperbolic paraboloid canopy roof used as a tensile structure to cover a soccer arena. The non-Gaussian properties of the pressure coefficient processes are measured at different roof locations for different wind angles of attack. Peak factor statistics, estimated using pressure coefficient time histories experimentally measured in wind tunnel tests and measured on the bottom and on the top of the roof, are compared with corresponding peak factor statistics estimated through use of four analytical models available in the literature, namely the Davenport, classical Hermite, revised Hermite, and translated-peak-process (TPP) models. It was found that: (1) all analytical models underestimate the mean and standard deviations of the experimental peak factors; (2) the non-Gaussianity region is significantly affected by the position on the roof, i.e., up and down, whereas it is less affected by the wind angle; (3) the two Hermite models provide accurate estimates of peak factor mean and standard deviations.