Structural health monitoring and fatigue life reliability assessment of a flexible structure in extreme wind

Li-Wei Tsai,Alice Alipour

DOI: https://doi.org/10.1007/s13349-022-00658-1

2023-01-18

Journal of Civil Structural Health Monitoring

Abstract:Many failures of cantilever traffic signal structures have been reported in the USA in the past two decades, which have revealed the vulnerability of such structures to wind force and created a need to study their wind-induced behavior and fatigue performance. Previous studies have provided a thorough understanding on the wind-induced behavior of traffic signal structures in regular wind conditions. However, the wind-induced behavior of such structures in the extreme wind conditions has never been studied before. In this study, a cantilever traffic signal structure was selected for long-term monitoring. It was the first time that a traffic signal structure was monitored during a derecho, which has a wind speeds of more than 240 miles and wind gust at least 58 mph. In the first part of the study, the monitoring data during the August 2020 Iowa Derecho were analyzed to understand its wind-induced behavior in extreme wind conditions and the monitoring stress data were used to evaluate the wind speeds could possibly create fatigue damage on the structure. In the second part of the study, to more accurately quantify the fatigue damage on the structure, a data-driven algorithm for estimating fatigue life and reliability of wind-excited structures was proposed. The monitoring stress data in both regular and extreme wind conditions were used to develop fatigue damage model, which addressed the issue of insufficient monitoring data at high wind speeds from previous studies. Fatigue life was then estimated by combining the fatigue damage model and the local wind speed probability. In reliability analysis, uncertainties were considered as the wind speed probability, the fatigue resistance of the mast arm base, and Miner's sum, and Monte Carlo simulations were conducted to generate a probability-of-failure curve. Finally, to demonstrate the algorithm, wind speed probability data from three cities in Iowa were used to estimate the fatigue life and reliability. The proposed data-driven algorithm could be widely used on other wind-excited structures, and the results from reliability analysis can serve as a reference in determining the period of regular maintenance for such structures.

engineering, civil