| Windstorms create some of the most devastating natural disasters in the United States. Much of this damage occurs to low-rise buildings. Consequently, wind damage to low-rise buildings continues to receive growing attention from researchers and engineers. The design and construction of cost-effective stronger buildings in hurricane-prone regions is a top priority for engineers and researchers in the wind engineering community. The large negative pressure fluctuations, generated in the flow separation regions close to the leading edge on the roof surfaces of low-rise buildings, are a major cause of damage in windstorms. While considerable research has been performed to investigate wind loads on flat roof buildings, relatively little work has been conducted on gable roof buildings. This dissertation investigates various characteristics of roof pressure fluctuations on a gable roof house in the hope that the findings will help further the understanding of the mechanism of wind damage to low-rise buildings.; Experiments were conducted in the boundary layer wind tunnel of the Wind Load Test Facility at Clemson University. Mean, minimum, maximum and root-mean-square (RMS) values of pressure coefficients were obtained for every 10° from wind azimuths between 0° to 90°. Time histories of roof pressures were taken only at selected taps for the wind azimuths of 0°, 30° and 90°. The spatial distributions over the entire roof surfaces of the mean, minimum pressure coefficients and the RMS values of the pressure coefficients were studied in great detail. High suction pressure zones were identified for the ten wind directions from 0° to 90°. Considerable efforts were devoted to fitting probability distributions to the overall and large negative pressure fluctuations. It was determined that a shifted Weibull distribution could be used to describe the entire record. Weibull distributions of the largest negative pressure fluctuations closely matched field results. Power spectra of pressure fluctuations were analyzed and their typical patterns were identified. The regions where different patterns occur were located. Features of peak positive and negative pressure fluctuations were studied. Three different techniques, i.e., direct cross correlation analysis, peak pressure fluctuation detection and wavelet analysis, were used to study pressure fluctuation propagation across the roof surface. The results produced consistent and somewhat complementary descriptions of the propagation of pressure events across the roof surface. |
