Assessing Probabilistic Extreme and Fatigue Responses of Wind-Excited Structures through Integration of Both Uncertainty and Directionality with a System Perspective

Texas Tech University, Lubbock TX

Investigators

Abstract

Better modeling and quantification of wind load effects on buildings and structures are essential for structural design and wind hazard mitigation. Wind tunnel studies and structural dynamic response analysis provide information of wind-induced extreme and fatigue responses as functions of wind speed and direction. The overall structural performance to strong wind is then evaluated by a further consideration of variations of wind speed and direction. As the most unfavorable wind direction that causes the largest wind load and structural response under given wind speed does not necessarily align with the direction of strongest wind, consideration of directionality effects can result in more economical design of structures as compared to the analysis with the "worst case" approach of disregarding wind directionality. The discrepancies of predictions from current existing approaches on directionality effect are found to be significant. Furthermore, these approaches cannot account for uncertainty and directionality in a unified framework. Due to a lack of scrutiny on the accuracy and effectiveness of these approaches, and a lack of a unified approach for both uncertainty and directionality, the detailed and comprehensive wind loading information derived from wind tunnel studies may not necessarily lead to expected improvement in the design of wind-excited structures. This research will consider the uncertainty and directionality of wind, aerodynamics, and structural characteristics in a unified framework with a system perspective. Simplified procedures will be developed for possible use in future design codes and standards. The new tools and knowledge can also be used for performance-based design and assessment of structural systems against multiple hazards. The goal of this research is to investigate transparent and reliable approaches for better assessing probabilistic multiple limit state responses to support a reliability and performance-based design of wind-excited structures. This research has the following objectives: 1) modeling of probabilistic extreme and fatigue responses of non-Gaussian response processes, 2) modeling of directional extreme wind speeds, 3) prediction of probabilistic extreme and fatigue responses considering both uncertainty and directionality, and 4) assessing structural performance with the effects of uncertainty and directionality accounting for multiple limit-state responses. The research findings will help develop consensus on how wind directionality effects are best quantified, thus the large discrepancies of existing approaches can be reduced or eliminated. The framework will also help wind engineering laboratories to better integrate wind tunnel study results with wind modeling for cost-effective design of structures to wind loads.