Statistical investigation of wind duration using a refined hurricane track model

Abstract

A complete characterization of the atmospheric boundary-layer winds essentially involves speed, direction and duration. While extensive statistical analyses have been carried out to determine wind speed and directionality for appropriate wind design of buildings, there is a lack of similar research effort for wind duration. Although wind duration may have limited effects on linear, elastic structural response, it can present significant importance in the consideration of nonlinear, inelastic response due to damage accumulation and cyclic deterioration of strength and stiffness. Recent advances of performance-based wind design methodology allow the controlled nonlinear, inelastic deformation of buildings under strong winds, and hence place a demand on statistical investigation of wind duration. In this study, the wind data used for probabilistic duration analysis result from a refined hurricane track model. Specifically, both environmental flow and Coriolis force (the beta effect) are integrated into the translation module for improved modeling of hurricane movement. Based on the numerically generated wind data, the wind duration measured with the over-threshold method (i.e., uniform duration) is obtained. Furthermore, a statistical analysis framework is developed to jointly consider wind duration, speed and direction. A case study is conducted to demonstrate the efficacy of the proposed statistical analysis framework for characterization and quantification of duration in wind climate consideration. Finally, structural wind performance of a simplified single-degree-of freedom system is examined to highlight the importance of incorporating duration consideration into the performance-based wind design.