Statistical Analysis of Stress Signals from Bridge Monitoring by FBG System

• Published in: Sensors (Basel, Switzerland) Vol. 18; no. 2; p. 491
• Main Authors: Ye, Xiao-Wei, Su, You-Hua, Xi, Pei-Sen
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI AG 07-02-2018; MDPI
• Subjects: Arch bridges; Arches; Bayesian analysis; Bayesian information criterion; Bridge decks; Bridges; Civil engineering; FBG sensor; finite mixture distribution; genetic algorithm; Genetic algorithms; Highway construction; Interferometry; Investigations; Metal fatigue; Monitoring systems; Orthotropic bridges; orthotropic steel bridge; Parameter estimation; Probabilistic methods; Ribs (structural); Sensors; Signal monitoring; Signal processing; Statistical analysis; Stress analysis; Stress concentration; stress concentration factor; structural health monitoring; Temperature effects; Wavelet analysis; wavelet multi-resolution analysis; Welded joints; genetic algorithm; stress concentration factor; finite mixture distribution; wavelet multi-resolution analysis; structural health monitoring; FBG sensor; Bayesian information criterion; orthotropic steel bridge
• ISSN: 1424-8220; 1424-8220
• DOI: 10.3390/s18020491

In this paper, a fiber Bragg grating (FBG)-based stress monitoring system instrumented on an orthotropic steel deck arch bridge is demonstrated. The FBG sensors are installed at two types of critical fatigue-prone welded joints to measure the strain and temperature signals. A total of 64 FBG sensors are deployed around the rib-to-deck and rib-to-diagram areas at the mid-span and quarter-span of the investigated orthotropic steel bridge. The local stress behaviors caused by the highway loading and temperature effect during the construction and operation periods are presented with the aid of a wavelet multi-resolution analysis approach. In addition, the multi-modal characteristic of the rainflow counted stress spectrum is modeled by the method of finite mixture distribution together with a genetic algorithm (GA)-based parameter estimation approach. The optimal probability distribution of the stress spectrum is determined by use of Bayesian information criterion (BIC). Furthermore, the hot spot stress of the welded joint is calculated by an extrapolation method recommended in the specification of International Institute of Welding (IIW). The stochastic characteristic of stress concentration factor (SCF) of the concerned welded joint is addressed. The proposed FBG-based stress monitoring system and probabilistic stress evaluation methods can provide an effective tool for structural monitoring and condition assessment of orthotropic steel bridges.