Shunlong Li

Reliability assessment of cable-stayed bridges based on structural health monitoring techniques

This paper presents the reliability analysis approach of long-span cable-stayed bridges based on structural health monitoring (SHM) technology. First, the framework of structural reliability analysis is recognised based on SHM. The modelling approach of vehicle loads and environmental actions and the extreme value of responses based on SHM are proposed, and then models of vehicle and environmental actions and the extreme value of inner force are statistically obtained using the monitored data of a cable-stayed bridge. For the components without FBG strain sensors, the effects and models (extreme values) of dead load, unit temperature load, and wind load of the bridge can be calculated by the updated finite element model and monitored load models. The bearing capacity of a deteriorated structure can be obtained by the updated finite element model or durability analysis. The reliability index of the bridges critical components (stiffening girder in this study) can be estimated by using a reliability analysis method, e.g. first order reliability method (FORM) based on the models of extreme value of response and ultimate capacity of the structure. Finally, the proposed approach is validated by a practical long-span cable-stayed bridge with the SHM system. In the example, reliability indices of the bridges stiffening girder at the stage after repair and replacement after 18 years of operation, and the damaged stage are evaluated.

Probabilistic deterioration model of high-strength steel wires and its application to bridge cables

Bridge inspections reveal that severe corrosion and fatigue are the main failure mechanisms of bridge stay cables. This paper presents an empirical modelling of the long-term deterioration process of steel wires in cables with consideration of the simultaneous occurrence of uniform corrosion, pitting corrosion and fatigue induced by a combined action of environmental aggression and cyclic loading. Accelerated corrosion experiments are conducted to determine the different corrosion levels of high-strength steel wires, and time-dependent statistical models are developed to quantify uniform and pitting corrosion depth. Corrosion-fatigue process of steel wires is subsequently simulated using the corrosion models and cyclic stress obtained through cable force monitoring data. The mechanical properties of corroded steel wires, including yield stress, ultimate stress, ultimate strain and modulus of elasticity, are experimentally characterised, and the statistical models are established through regression analysis. Finally, the deterioration models of high-strength steel wires (including crack depth, ratio of broken wires, and remaining strength, among others) is extended to probabilistically assess the time-variant conditions of bridge cables (sectional area loss and remaining capacity). The presented study on the long-term deterioration of bridge cables would provide guidance to future decision-making regarding the maintenance and replacement of bridge cables.

Uniform and Pitting Corrosion Modeling for High-Strength Bridge Wires

AbstractCable inspections revealed that severe corrosion of steel wires is one of the main failure mechanisms of cables. Accelerated corrosion experiments were conducted to evaluate the variation in the uniform and pitting corrosion depths of high-strength steel wires over time. The measured uniform corrosion depth followed a lognormal distribution with time-dependent corrosion variables at both the zinc coating corrosion stage and the steel corrosion stage. The block’s maximum pitting factors from a different exposure time were proven to be drawn from the same underlying continuous population and followed Gumbel distribution. The regression models of the scale and location parameter with increased surface area were obtained based on the experimental block’s maximum pitting factors. The extreme value distribution of the pit depth of corroding steel wires could be predicted by the developed statistical uniform and pitting factor models.

Damage Detection for SMC Benchmark Problem: A Subspace-Based Approach

A full-scale bridge benchmark problem was issued by the Center of Structural Monitoring and Control at the Harbin Institute of Technology. The data used in this problem were collected by an in situ structural health monitoring system implemented into a full-scale cable-stayed bridge before and after the bridge was damaged, which is very rare in structural health monitoring field. This benchmark problem will help to verify and/or make comparison of the condition assessment and the damage detection methods, which are usually validated by numerical simulation and/or laboratory testing of small-scale structures with assumed deterioration models and artificial damage. With respect to damage detection of girder, one of the benchmark problems, using the monitored and field testing acceleration data, this paper describes a damage detection method, based on a residual generated from a subspace-based covariance-driven identification method, to detect the damage, and give relative quantitative damage indexes. This method was applied on both two parts of the given benchmark problem, and then detailed discussions and results on this problem are reported in this paper.

An On-site Evaluation Method for Corrosion Degradation of In-service Cable Wires by Image Histogram Matching

This paper proposes a novel non-destructive method by image histogram matching to evaluate the corrosion degradation of cable wires in service quantitatively. Artificial accelerated salt spray test is adopted to simulate the corrosion environment, and adequate RGB images are taken by a consumer-grade camera to cover the whole information of corroded surface for each specimen under different corrosion periods. Histogram distribution analysis is applied to study the concentrated modes of RGB images, and joint Gaussian distribution is fitted to capture location parameters. Timevariant models of the location parameters are regressed and then the mechanical properties are learnt by tensile and fatigue tests. Once the features of RGB corrosion images of steel wires are matched to the proposed time-variant models, the corresponding corrosion degree and mechanical properties can be estimated. The proposed method is meaningful for the on-site evaluation of cable wires in service with no need to carry on corrosion and mechanical degradation tests after replacing cables.

Long-Term Condition Assessment of Stay Cables: Application to Nanjing 3rd Yangtze River Bridge

This paper presents a long-term condition assessment approach of cables under in-service loads collected by SHM system. For the cables of the bridge, the stochastic axial force response can be collected by the SHM system and described by a filtered Poisson process, through which the maximum value distribution of axial forces in its design reference period can be derived using Poisson Process theory. The long-term deterioration process of steel wires in the cables considers simultaneously the uniform and pitting corrosion due to environmental attack and the fatigue propagation induced by cyclic stress. By employing first order reliability method, the reliability of the cables under the monitored responses is further estimated in terms of the safety under the extreme traffic load distribution in the design reference period and the serviceability specified in the design specification. The discussions of the life-cycle condition assessment of the cable-stayed bridge provide guidance to the future decision-making related to maintenance and replacement.

Surface characteristics and mechanical properties of high-strength steel wires in corrosive conditions

Cables are always a critical and vulnerable type of structural components in a long-span cable-stayed bridge in normal operation conditions. This paper presents the surface characteristics and mechanical performance of high-strength steel wires in simulated corrosive conditions. Four stress level (0MPa, 300MPa, 400MPa and 500MPa) steel wires were placed under nine different corrosive exposure periods based on the Salt Spray Test Standards ISO 9227:1990. The geometric feathers of the corroded steel wire surface were illustrated by using fractal dimension analysis. The mechanical performance index including yielding strength, ultimate strength and elastic modulus at different periods and stress levels were tested. The uniform and pitting corrosion depth prediction model, strength degradation prediction model as well as the relationship between strength degradation probability distribution and corrosion crack depth would be established in this study.

Reliability assessment of long span bridges based on structural health monitoring: application to Yonghe Bridge

This paper presents the reliability estimation studies based on structural health monitoring data for long span cable stayed bridges. The data collected by structural health monitoring system can be used to update the assumptions or probability models of random load effects, which would give potential for accurate reliability estimation. The reliability analysis is based on the estimated distribution for Dead, Live, Wind and Temperature Load effects. For the components with FBG strain sensors, the Dead, Live and unit Temperature Load effects can be determined by the strain measurements. For components without FBG strain sensors, the Dead and unit Temperature Load and Wind Load effects of the bridge can be evaluated by the finite element model, updated and calibrated by monitoring data. By applying measured truck loads and axle spacing data from weight in motion (WIM) system to the calibrated finite element model, the Live Load effects of components without FBG sensors can be generated. The stochastic process of Live Load effects can be described approximately by a Filtered Poisson Process and the extreme value distribution of Live Load effects can be calculated by Filtered Poisson Process theory. Then first order reliability method (FORM) is employed to estimate the reliability index of main components of the bridge (i.e. stiffening girder).

Model updating for Binzhou Yellow River Highway Bridge considering uncertainties

This paper presents an improved finite element (FE) model updating method for Binzhou Yellow River Highway Bridge and its associated uncertainties by utilizing measured dynamic response data. The dynamic characteristics of the bridge have been studied through both three dimensional finite element prediction and field vibration previously. A comprehensive sensitivity study to demonstrate the effects of structural parameters (including the connections and boundary conditions) on the modes concern is first performed, according with a set of structural parameters are then selected for adjustment. According to the eigenequation considering uncertainties, the proposed methodology transforms model updating problem for Binzhou Yellow River Highway Bridge into two deterministic constrained optimization problems regarding the predictable part and uncertainties of structural parameters. Both the predictable part and associated uncertainties of the structural parameters could be obtained in iterative fashions so as to minimize the difference between the predicted and the measured natural frequencies. The final updated model for Binzhou Yellow River Highway Bridge is able to produce natural frequencies and associated frequency uncertainties in good agreement with measured ones, and can be helpful for a more precise dynamic response prediction.