Electronic Thesis and Dissertation Repository

Degree

Doctor of Philosophy

Program

Civil and Environmental Engineering

Supervisor

Gregory A. Kopp

Abstract

Severe wind storms such as tropical cyclones, tornados, downbursts etc. can cause significant damage to infrastructure. Damage surveys following these events have shown that the roofs of residential, wood-frame construction are particularly vulnerable to failures. While damage surveys provide detail information of what components fail, they cannot provide the loads at which these failures occurred or how they initiated. Wind tunnel pressure models provide detailed information of the wind loads on buildings, however, they are not able to predict failures or how these loads are transferred through the structure. In order to better understand the response of wood framed houses when subjected to high wind loads, realistic fluctuating wind loads were applied to a full scale two-story wood frame house. In addition, individual component tests were conducted on toe-nailed roof-to-wall connections to examine their behaviour to fluctuating wind loads. The testing of individual toe-nail connections under realistic fluctuating wind loading has found that the nails are incrementally withdrawn at peak loads. However, the maximum load applied during the fluctuating load tests matches well with the failure capacity determined from ramp loading experiments, even though damage to the connections initiates at much lower loads. Tests performed on the roof of a house have shown that the uplift capacity of the roof is significantly higher than that predicted using the individual connection results. The higher uplift capacity of the entire roof is attributed to significant load sharing between adjacent connections so that failures likely initiate at multiple connections up to the entire roof and the effective tributary area of the roof-to-wall connections is substantially larger than that of a single truss. Since toe-nailed roof-to-wall connections are partially withdrawn during peak wind gusts, how the loads are transferred through the structural system, to the connections, changes as the connections become increasingly damaged. This implies that while static testing is suitable to determine the capacity of individual toe-nail connections, testing of the full structure must be conducted using realistic fluctuating wind loads. Despite the significant damage accumulated at the toe-nail roof-to-wall connections, there was little evidence of damage to the interior of the house, indicating that there may be significant undocumented damage to homes following tropical cyclones.


Share

COinS