Computation of the dynamic stress of long suspension bridges under multiloadings is essential for either the strength or fatigue assessment of the bridge. This paper presents a framework for dynamic stress analysis of long suspension bridges under wind, railway, and highway loadings. The bridge, trains, and road vehicles are respectively modeled using the finite-element method (FEM). The connections between the bridge and trains and between the bridge and road vehicles are respectively considered in terms of wheel-rail and tire-road surface contact conditions. The spatial distributions of both buffeting forces and self-excited forces over the bridge deck surface are considered. The Tsing Ma suspension bridge and the field measurement data recorded by a wind and structural health monitoring system (WASHMS) installed in the bridge are utilized as a case study to examine the proposed framework. The information on the concerned loadings measured by the WASHMS is taken as inputs for the computation simulation,...

Influence line (IL) has emerged as very promising damage indices for bridge damage detection. This study proposed a method to localize and quantify damage in beam structures by estimating section flexibility change from deflection IL (DIL) change. To this end, the relationship between second derivative of DIL change and flexibility change was established. To remove noise interference in measurement, piecewise quadratic functions were used to fit and replace noisy DIL change curves, wherein the coefficients of quadratic function were determined via a sparse regularization method, considering the sparsity nature of damage that typically takes place in only a limited number of elements. The feasibility and accuracy of the proposed method are verified through numerical examples and laboratory experiments. Through four hypothetical damage scenarios of a simply supported beam with one or two damaged locations, its ability to quantify minor damage and its anti-noise robustness were well verified. Finally, a laboratory experiment on a simply supported aluminum beam illustrated that the location and extent of damage could be successfully identified in the single-damage and double-damage cases. The numerical and experimental results indicate that the proposed method is promising for future damage localization and quantification of bridge structures.

Previous damage detection of frame structures mainly focuses on the detection of beam and column element damage. It has been shown that beam-column joints in frame structures are more susceptible to damage than the other members in the structure. Joint damage may be represented by the reduction of beam-column connection rigidity. Therefore, damage detection of a frame structure with joint damage includes the identification of joint connection stiffness in additional to those of beam and column stiffness, which involves the difficulty of identifying a large number of unknown structural parameters. In this paper, an algorithm based on a two-step Kalman filter approach is presented for the damage detection of frame structures with joint damage under earthquake excitation using partial measurements of structural acceleration responses. Recursive solutions for unknown structural parameters and structural state vector are derived by a two-step Kalman filter, respectively. Therefore, the number of unknown variables to be estimated in each step is reduced compared with the conventional Extended Kalman filter (EKF) approach. Structural damage is detected from the degradation of the identified stiffness values of joints, beam and column elements of frame structures. A numerical example and a lab experiment test data are used to validate the performances of the proposed algorithm for damage identification of various joint damage scenarios in frame structures under earthquake excitation.

As image acquisition devices have outstanding potential for gathering vibration information, computer vision has received a lot of interest in structural health monitoring (SHM). In this work, a fully automated peak picking methodology based on computer vision in tandem with deep learning is proposed to realize vibration measurements and identify natural frequencies from the plot of the power spectral density transmissibility (PSDT). A deep-learning-enhanced image processing technology was used to extract the vibration signals with automatic active pixel selection, while a convolutional neural network was used to further process the vibration measurements so that the frequencies could be identified from PSDT-based functions. The proposed method was verified by three case studies, including the dynamic testing of two laboratory models and the field testing of the stay cable. The findings showed that the proposed deep-learning-enhanced approach has a high potential for use in SHM by automatically performing vibration measurement and frequency extraction.