No abstract

This paper addresses the temperature-induced variations of measured modal frequencies of steel box girder for a suspension bridge using long-tem monitoring data. The output-only modal frequency identification of the bridge is effectively carried out using the Iterative Windowed Curve-fitting Method (IWCM) in the frequency-domain. The daily and seasonal correlations of frequency-temperature are investigated in detail and the analysis results reveal that: (i) the identified modal frequencies using IWCM provide an effective indication for changes of the bridge due to the ambient temperature variations; (ii) temperature is the critical source causing modal variability, and there is an overall decrease in modal frequency with temperature for all the identified modes; (iii) the random variations in measured modal frequencies mainly arise from the identification algorithm due to the nonstationary loadings, which can be effectively eliminated using multi-sample averaging technique; (iv) the daily averaged modal frequencies of vibration modes have remarkable seasonal correlations with the daily averaged temperature and the seasonal correlation models of frequency-temperature are suitable for structural damage warning if future seasonal correlation models deviate from these normal models.

No abstract

No abstract

This paper focuses on developing an online structural condition assessment technique using long-term monitoring data measured by a structural health monitoring system. The seasonal correlations of frequency-temperature and beam-end displacement-temperature for the Runyang Suspension Bridge are performed, fi rst. Then, a statistical modeling technique using a six-order polynomial is further applied to formulate the correlations of frequency-temperature and displacement-temperature, from which abnormal changes of measured frequencies and displacements are detected using the mean value control chart. Analysis results show that modal frequencies of higher vibration modes and displacements have remarkable seasonal correlations with the environmental temperature and the proposed method exhibits a good capability for detecting the micro damage-induced changes of modal frequencies and displacements. The results demonstrate that the proposed method can effectively eliminate temperature complications from frequency and displacement time series and is well suited for online condition monitoring of long-span suspension bridges.

A reliability assessment method of fatigue life based on the long-term monitoring data is developed for welded details in steel box girder, and the application research is presented with examples of welded rib-to-deck details in Runyang Bridges. Firstly the fatigue damage limit-state function is established based on S-N curves and Miner's rule, and the probability distribution characteristics of the coefficients in the function are discussed in detail. The uncertainties in fatigue loading effects are mainly studied based on long-term monitoring data. In the traditional studies, only the uncertainty of equivalent stress range is considered in fatigue reliability assessment. However, stress cycle number is also treated as a random variable in this paper because we know traffic flow every day differs in a thousand ways. Then the optimization method is employed to calculate the fatigue reliability. After studying the changing law of the reliability indices with time and the effect of the randomness of stress cycle number on reliability, the effect of the traffic growth on the reliability is studied. This study shows that the uncertainty in the fatigue life of the welded details can be well studied based on structural health monitoring, so it is necessary to carry out long-term strain monitoring of the welded details for accurate fatigue reliability assessment during the whole service period. steel box girder, fatigue loading effects, equivalent stress range, stress cycle number, fatigue reliability, traffic growth Citation:Deng Y, Ding Y L, Li A Q, et al. Fatigue reliability assessment for bridge welded details using long-term monitoring data.

A workable realization procedure for damage alarming of frame structures is put forward based on energy variations of structural dynamic responses decomposed using wavelet packet transform in the paper. The WPT-based method consists of two steps: 1) calculation of the wavelet packet energy spectrum from the measured structural dynamic responses; 2) extraction of the damage alarming index ERVD from the wavelet packet energy spectrum, which is sensitive to structural local damage and insensitive to measurement noise. The ASCE benchmark experiments demonstrate the practicability of the damage alarming method for fame structures, which reveal that the WPT-based damage alarming index ERVD is a good candidate index that is sensitive to structural local damage affected by the actual measurement noise. Also, the experimental results reveal that the lower decomposition level and dominant frequency bands are sufficient for the detection of the damage occurrence using the index ERVD, which makes the proposed damage alarming procedure practical.

The monitoring data makes it feasible to quickly evaluate the cracking of the prestressed concrete box-girder bridge. The live-load strain can accurately quantify the load effect and cracking of bridges due to its explicit datum point of signal. Based on the live-load strain data from bridge monitoring system, this study develops a comprehensive data-driven method of state evaluation and cracking early warning for the prestressed concrete box-girder bridge. The feature of vehicle-induced strain is extracted using the deep learning and classification of long short-term memory network. The vehicle-induced strain features are clustered via Gaussian mixture model, and the cracking early warning of the bridge is conducted according to the reliability of heavy vehicle clustering data. This method can be used as an indicator for the bridge inspection, truck-weight-limit and reinforcement work. The results demonstrate that (1) using the long short-term memory network, a deep learning model can be trained to intelligently classify the non-stationary and stationary sections of vehicle-induced strains, of which the test accuracy of classification surpasses 99%, and (2) according to the Gaussian mixture model probability distribution of data, the vehicle-induced strain features can be clustered by the corresponding Gaussian mixture model crest, which is the premise for reflecting relational mapping between vehicle loading and strain response.