International Journal of Lifecycle Performance Engineering (12 papers in press)
Uncertainty analysis of the subway vehicle-track coupling system with fuzzy variables
by Hongping Zhu, Ling Ye, Shun Weng
Abstract: Dynamic analysis of the vehicle-track coupling system is important to the structural design, damage detection and condition assessment of the structures. The structural parameters of the vehicle-track coupling system include many uncertainties intrinsically, it is rational to treat the parameters in an uncertainty way. In this paper, the uncertain system parameters are modeled as fuzzy variables instead of deterministic values or conventional random variables with known probability distributions. Afterwards, the dynamic response functions of the coupling system are transformed into a component function based on the high dimensional representation approximation. Then the Lagrange interpolation method is used to approximate the component function. Finally, the bounds of the system dynamic responses can be predicted by performing Monte Carlo method on the interpolation polynomials of Lagrange interpolation function, avoiding the expensive Monte Carlo process on the subway vehicle-track coupling system. A numerical example is analyzed by the proposed method to predict the bounds of the subway vehicle-track coupling system responses. The results are compared with the direct Monte Carlo simulations, which show that the proposed method is effective and efficient to predict the bounds of the system responses with fuzzy system parameters.
Keywords: subway vehicle-track coupling system; fuzzy variables; high dimensional representation approach; Monte Carlo simulation.
Special Issue on: System Identification and Health Monitoring of Civil Structures in the Presence of Uncertainties
Multi-scale finite element model validation method of cable-stayed bridge based on the Support Vector Regression
by Peijuan Zheng, Zhouhong Zong, Qiqi Liu, Jie Niu, Haifei Zhou, Rumian Zhong
Abstract: In this paper, the multi-scale Finite Element model (FEM) of a composite cable-stayed bridge, Guanhe bridge, was established based on the Arlequin method firstly. Then a two-step multi-scale FE model updating method was proposed. Furthermore, based on structural health monitoring (SHM) system of Guanhe bridge, support vector regression (SVR) method was employed to analyze the uncertainty quantification and transmission. It was shown that the errors between the calculated frequencies from the updated multi-scale FEM and the measured frequencies from SHM were less than 3%. In the procedure of inverse uncertainty propagation, the coincidence indexes of the structural parameters were larger than 65%. The deviations between the optimal values of the updated parameters and the corresponding statistical mean values were very small (<5%). Finally, the analysis results indicate that the distributions of the parameters agree well with the assumed normal distribution.
Keywords: Multi-scale simulation; Finite element model (FEM) validation; Support vector regression(SVR); Uncertainty quantification and propagation; Composite cable-stayed bridge.
A parametric study of bridge load effect under stochastic vehicular load
by He-Qing Mu, Hou-Zuo Guo, Tian-Yu Zhang, Cheng Su
Abstract: Bridge vehicular load effect inference is one of the key factors in bridge reliability and life-cycle assessment. The stochastic vehicular load can be decomposed into two parts: 1) the random vehicular inflow, describing the information about the properties of the inflow vehicles entering the bridge; 2) the stochastic vehicular flow, describing the information about the vehicle-following pattern of vehicles on the bridge. In order to investigate how the bridge load effect is affected by the stochastic vehicular load, two key parameters of the stochastic vehicular load are selected: 1) the probability of the existence of vehicle (PoV), controlling the traffic volume of the vehicular inflow entering the bridge; 2) the probability of random slowing down (PoSD), controlling the vehicle-following pattern of the stochastic vehicular flow. With different values of the PoV and the PoSD, the samples of the vehicular load effects are achieved by embedding the samples of the stochastic vehicular load and the influence lines of the load effects. The Gaussian Process Regression (GPR) is utilized to obtain the relations between two parameters of the stochastic vehicular load (the PoV and the PoSD) and the statistical moments (the mean and the standard deviation) of the simulated load effect samples. it turns out that the load effect is interactively influenced by the PoV and the PoSD.
Keywords: Bridge vehicular load effect; Gaussian Process Regression; Nagel-Schreckenberg model; Uncertainty Quantification.
Operational Modal Analysis and Bayesian Model Updating of a Standing Seam Metal Roofing System
by Heung Fai Lam, Jun Hu, Kwok Fai Chung
Abstract: The metal roofs have been widely adopted in public facilities due to its simple construction procedures. However, it is well-known that the metal roof is vulnerable to repeated wind load and temperature effects, and clip detachment is found to be a common failure mode. Clip damage leads to the reduction in the stiffness at some supports of the metal roof and alter its dynamic characteristics. Therefore, it is possible to detect the damage of clips by monitoring the modal properties, such as natural frequencies and mode shapes, of the system. To assess the dynamic properties of metal roofs, part of a full-scale standing seam metal roof was constructed on a supporting steel frame in the laboratory. Ambient vibration tests were carried out under laboratory conditions. Modal parameters, such as natural frequencies, damping ratios and mode shapes, were identified from the measured acceleration responses of the metal roof using frequency domain decomposition method (FDD). The finite element model of the metal roof together with the supporting system was developed, and the clip stiffness was estimated by Markov chain Monte Carlo (MCMC)-based Bayesian model updating method utilizing the identified modal parameters. Both the modal identification and model updating results are reported in detail. The results presented in this paper are valuable and important for the development of a structural health monitoring system for metal roofing system.
Keywords: Operational modal analysis; Bayesian approach; Model updating; Metal roof.
Fractal signal processing method of acoustic emission monitoring for seismic damage of concrete columns
by Yong Huang, Changsong Shao
Abstract: In this paper, a new fractal theory-based Acoustic emission (AE) signal processing method is proposed. It is found that both the curve lengths and fractal dimensions (FD) of AE signal are related with damage evolution. The AE tests of pseudo-static experiment of a reinforced concrete column, a reinforced nano-concrete column and a concrete-filled GFRP tube are then performed for validation. For each specimen, several Piezoelectric ceramic (PZT) patches and one AE sensor are bonded at different positions of the specimen surface to monitor the AE signals. The results show that the fractal theory-based damage method can assess damage evolution effectively. In addition, the damage can be localized approximately by the diversity of damage assessing index values from various PZT detectors.
Keywords: fractal dimension; acoustic emission; Pseudo-static experiment; damage assessment; damage localization.
Special Issue on: Data-driven Structural Damage Identification and Performance Assessment
Acquisition of thermal field of spatial steel structure based on temperature measurements
by Wei Lu, Kai Huang, Jun Teng
Abstract: A spatial steel structure is a kind of high-order statically indeterminate structure, and the deformation caused by repeated changes in temperature will cause fatigue damage to its members. Therefore, for the design and safe use of spatial steel structures, it is of great significance to study the thermal effect. A method of acquiring the overall thermal field based on the limited temperature measurements is proposed in the paper. Firstly, the simulated thermal field of the structure is acquired by superposing the ambient and simulated temperatures. Secondly, the optimal mode for temperature estimations is determined by inverse distance weighted interpolation. Thirdly, the thermal field of the spatial steel structure is acquired based on the temperature measurements and the optimal mode. Finally, the feasibility of the proposed method is verified in the case of the large shell of the Zhuhai Opera House.
Keywords: structural health monitoring; thermal field; inverse distance weighted interpolation; Zhuhai Opera House.
Lifetime wave overtopping assessment of coastal defences under changing environments
by Mehrdad Bahari Mehrabani, Hua-Peng Chen
Abstract: Coastal defence structures become more vulnerable due to changing environments such as sea level rise caused by climate change, hence requiring accurate time-variant reliability analyses in order to provide optimum maintenance strategies. This paper presents a method for assessing wave overtopping risk of the coastal defences subjected to changing operation conditions. The joint probability of sea water level and wave height is used to assess future hydraulic conditions on the basis of the UK climate projections. Structural resistance degradation of coastal defences such as the crest level settlement of earth sea dykes is also considered in the process of risk analysis and failure probability evaluation. A case study of a typical earth sea dyke section affected by changing environments is used to demonstrate the effectiveness of the proposed method. The results show that the overtopping failure probability of the earth sea dykes will increase significantly due to changing environments, and appropriate maintenances are needed to reduce the risk of the wave overtopping failure.
Keywords: Sea defences; Overtopping; Reliability; Climate change; Performance deterioration.
Structural damage detection using wavelet packet transform combining with principal component analysis
by Zhenhua Nie, Enguo Guo, Hongwei Ma
Abstract: In structural health monitoring (SHM), data-driven is a model free method that mines the internal characteristics of the raw measured responses, does not need considering the structural model. Since it involves only tracking changes in signals, it is appropriate for the continuous monitoring of the performance of structures. This paper proposes a new data-driven damage detection method based on wavelet packet transform (WPT) combining with principal component analysis (PCA). In this method, the responses are firstly decomposed by WPT, and each layer of wavelet packet energy of each measurement are obtained. Since each wavelet packet component contains information about the signal in a specific time-frequency window, the magnitude of the component energy may vary significantly. Those wavelet packet component of small energy magnitudes are easily contaminated by noises and should be discarded. PCA is utilized to reduce and eliminate redundant information of wavelet packet energy, and a damage index named the rate of the wavelet packet energy score change (WPESCR) is proposed to locate the damage. A simple beam is employed to demonstrate the effectiveness of this method. The numerical and experimental results show that the method combined with WPT and PCA can successfully locate the damage. As a data-driven method, the new WPESCR indicator is a better candidate for continuous online SHM.
Keywords: Structural health monitoring; Damage detection; Principal component analysis; Wavelet packet transform.
Structural damage identification of steel-concrete composite bridge under temperature effects based on Cuckoo Search
by Minshui Huang
Abstract: Civil structures are generally exposed to varying temperature conditions. Temperature variations in structural components not only cause quasi-static responses like displacement and stress, but also lead to changes in vibration features, such as frequencies and mode shapes. Damage identification is usually based on the vibration characteristics, which is easily affected by temperature. This means that the accuracy of identification is not guaranteed without considering the temperature effects. In the paper, temperature is considered as a variable in material properties, and the finite element model of I-40 steel-concrete composite bridge is established based on MATLAB platform in order to figure out its vibration features. Then Cuckoo Search (CS) is introduced to damage identification under temperature variations. It is shown that Cuckoo Search is able to distinguish experimental structure damages from temperature variations.
Keywords: Damage identification; Steel-concrete composite bridge; Temperature effects; Cuckoo Search (CS).
Identification of Unknown Inputs Considering Structural Parametric Uncertainties
by Ying Lei
Abstract: Due to the inevitable uncertainties in structural parameters, deviations may occur between the actual structural dynamic characteristics and the theoretical status, which in turn affects the identification results via the inversed problem. In practical engineering, it is often difficult or even impossible to measure external excitations to structures. Therefore, it is necessary to consider the effects of structural parametric uncertainties in the identification of unknown inputs to structures. In this paper, the identification of unknown external inputs considering structural parametric uncertainties is investigated. Two identification approaches are proposed to account for the different scenarios of sensor deployments. The first algorithm is based on the improved Kalman Filter with unknown input (KF-UI) recently proposed by the authors, in which acceleration responses are measured at the locations where unknown inputs applied. The second method is based on modal Kalman Filter with unknown input (MKF-UI), which is Kalman Filter with unknown input in modal domain, to consider the scenario that acceleration responses at the locations of unknown inputs are unmeasured. For the uncertainties of structural parameters, probability model or interval model are studied, respectively. Some numerical examples are used to demonstrate the performances of the two proposed approaches. Monte Carlo simulation is applied in comparison to validate the effectiveness and accuracy of the identification of unknown inputs.
Keywords: excitation identification; Kalman Filter; unknown input structural parametric uncertainties; probability model; interval model.
Special Issue on: Data-driven structural damage identification and performance assessment
Special Issue on: System identification and health monitoring of civil structures with the present of uncertainties
An efficient method for Bayesian system identification based on Markov chain Monte Carlo simulation
by Jia-Hua Yang
Abstract: This paper proposes an efficient method for identifying a dynamic system using measured accelerations. A practical mathematical model of a dynamic system is developed based on modal superposition for predicting system response with affordable computational time. To explicitly address uncertainties in system identification, a probabilistic model is used to describe the prediction error between measured and predicted response. Following Bayes theorem, the posterior probability density function (PDF) of the uncertain parameters of a system model is then derived. System identification is then viewed as a Bayesian inference problem where the objective is to identify the posterior PDF conditional measured data. Unless a very simple system is considered, the posterior PDF is usually very complicated in the sense that its region of significant probabilities is concentrated in the neighborhood of an extended and extremely complex manifold in a high-dimension parameter space. An effective Markov chain Monte Carlo (MCMC) algorithm is developed to sample from the posterior PDF. Instead of pinpointing one optimal model, multiple models of a model class are considered with the posterior PDF measuring the plausibility of each model. Given the samples generated on the manifold by the MCMC algorithm, a framework is proposed to systematically considered multiple models whose relative plausibility is quantified by the weightings depending on the PDF values of the samples. A simple two-story shear building system is used to illustrate that the proposed method can globally identify the structural system. A complicated space frame is then employed to illustrate that the proposed method can also identify a posterior PDF with a complex manifold and accurately predict structural response when multiple equally important models are considered.
Keywords: system identification; Bayesian updating; Markov chain Monte Carlo; robust prediction.