Earthquake Engineering & Structural Dynamics最新文献

In the aftermath of an earthquake, damage detection and performance evaluation of structural components are imperative for assessing the residual seismic capacity of a building. In this study, an integrated image processing and deep learning approach was developed to evaluate the degradation in strength and stiffness (i.e., strength reduction and stiffness reduction) of reinforced concrete (RC) shear walls. The approach comprised two main tasks: detecting and localizing visible seismic damage from photographs and evaluating strength and stiffness degradation based on this information. The semantic segmentation network, Damage-Net, was used for damage detection and localization. A novel crack morphological processing layer and a patch feature extraction layer were developed for damage feature extraction and compression. A lightweight deep convolutional neural network named DegradeEval-Net_v2, featuring the upgraded dilated and separable convolution block and multi-layer perception, was developed to link the damage feature with strength and stiffness degradation. A database comprising test data and photographs of 14 RC shear wall specimens with a flexural-dominated behavior mode and high to intermediate ductility was constructed to train and test the DegradeEval-Net_v2 network. The results indicate that DegradeEval-Net_v2 substantially improved the performance assessment accuracy of damaged RC shear walls, with a 35% smaller root mean square error (RMSE) for stiffness degradation evaluation and 75% smaller RMSE for strength degradation evaluation, compared with the provisions specified in JBDPA and FEMA guidelines. Moreover, evaluation results on test sets demonstrate that introducing the damage feature extraction and compression layers effectively preserved local crack information and improved the accuracy with which stiffness reduction was evaluated. In addition, DegradeEval-Net_v2 outperformed ResNet18 and MobileNet V3 in terms of balanced efficiency and accuracy. Interpretability analysis demonstrates that the model learned the distinct contribution patterns of various visible damage indexes to stiffness and strength degradation across different loading levels.

Real-time hybrid testing (RTHT) is an efficient method to simulate the dynamic behavior of complex engineering systems. A novel offline RTHT method has been developed in recent years, wherein the computation of the numerical substructure and the loading of the experimental substructure are independent. Offline RTHT has obvious advantages in terms of accuracy, stability, and cost compared with conventional online RTHT. However, due to the excessive number of iterations, the application range of the existing offline RTHT methods is limited. This paper proposes an accelerated time history iteration (ATHI) method based on system identification and virtual iteration. A two-loop parameter optimization (TLPO) method is developed to obtain an accurate discrete transfer function. Virtual iterations are performed by replacing the real system with an identified transfer function, which can reduce the number of real iterations. Physical tests were performed on structures equipped with a tuned mass damper or active mass damper, where resonance, nonlinearity, closed-loop control, and measurement noise exist. The test results suggest that the real system can be accurately represented by the identified transfer function when adopting the TLPO method. The proposed ATHI successfully accelerates the convergence process while ensuring stability and accuracy.

The analysis of soil–structure interaction (SSI) problems has been established successfully in recent decades. In particular, the solution in the frequency domain provides an exact and efficient method for computing the response of the coupled system. Despite this, the state of practice as a first attempt to incentivize time domain analyses compatible with standard finite element packages introduces the so-called dimensionless flexible-base frequency. This frequency, which depends solely on the structure-to-soil-period ratio, allows transforming the frequency domain analyses into time domain analyses using frequency-independent soil impedance values. However, if such frequency exists for the combined system, it must depend on several physical variables. In this work, we propose a supervised approach to obtain the flexible-base dimensionless frequency at which the frequency-independent soil impedance should be used. The analysis is carried out using five dimensionless parameters, and the importance of each one to the estimation of the dimensionless flexible-base frequency is investigated. We use an inverse problem based on ensemble Kalman inversion (EnKI) to obtain the optimal frequency of the interaction. The data obtained are then employed in a machine-learning framework to map a set of dimensionless parameters to such a frequency. The generated mapping is finally verified, and a significant improvement in time-domain simulations is shown compared to the state of practice.

Recent destructive seismic events have underlined the need for increasing research efforts devoted to the development of innovative seismic-resilient structures able to reduce seismic-induced direct and indirect losses. Regarding steel Moment Resisting Frames (MRFs), the inclusion of Friction Devices (FDs) in Beam-to-Column Joints (BCJs) has emerged as an effective solution to dissipate the seismic input energy while ensuring a damage-free behavior. Additionally, recent studies have demonstrated the benefits of implementing similar damage-free solutions for Column Bases (CBs). In this context, the authors have recently experimentally investigated a Self-Centering CB (SC-CB) aimed at residual drift reduction. Previous experimental tests only focused on the response of isolated SC-CBs under cyclic loads. Conversely, the present paper advances the research through an experimental campaign on a large-scale steel structure equipped with the proposed SC-CBs, providing valuable insights into the global structural response and improved repairability. A set of eight Pseudo-Dynamic (PsD) tests were conducted considering different records and configurations of the structure. The experimental results highlighted the effectiveness of the SC-CBs in minimizing the residual interstory drifts and protecting the first-story columns from damage, thus enhancing the structure's resilience. Moreover, the consecutive PsD tests allowed investigating the effectiveness of the reparation process in restoring the seismic performance of the ‘undamaged’ structure. An advanced numerical model was developed in OpenSees and validated against the global and component-level experimental results. Incremental Dynamic Analyses were finally performed to investigate the influence of the SC-CBs on the structure's seismic response while accounting for the record-to-record variability.

In the absence of experimental investigations on column members of underground structures, full-scale column specimens were tested to explore the seismic behavior of shallow-buried subway stations at various depths. The axial compression ratios of internal column specimens were set as 0.16, 0.33, and 0.40. Both hybrid simulations and quasi-static tests were performed on the station columns. The hybrid simulations illustrated the drift demands of internal columns, while the load-carrying capacity and deformation capacity were obtained from the quasi-static tests. Hybrid simulations at low, moderate, and high-intensity levels were conducted to study the seismic responses of shallow-buried rectangular stations. The hybrid simulations suggest that the most severe damage occurred in the station when the axial compression ratio of the tested column reached 0.40. Central columns suffered severe stiffness deterioration under high-level earthquake excitation, especially in stations at greater depths. Meanwhile, the quasi-static test results indicate that the ultimate load of the central columns increases with increasing axial compression, but this leads to a significant decrease in the ductility of columns. Besides, the sectional analysis results show that the central columns are prone to tension-controlled failure, and the safety margin for flexural response deteriorates with an increasing axial compression ratio. The test results indicate that shallow-buried rectangular stations are susceptible to flexure-controlled structural failure when their central columns possess a relatively low axial compression ratio and a high shear span-to-depth ratio. The failure mechanism of station columns is revealed by both the hybrid simulations and quasi-static tests, and the findings from the full-scale tests are beneficial for the practical design of shallow-buried rectangular stations.

Underground diaphragm walls are commonly used as a support system for the construction of subway stations, working together with inner side walls of subway stations to withstand the pressure from surrounding soils. However, the effect of diaphragm walls on the seismic response of subway stations is still not well understood yet, or at least not well considered during design. In this paper, a series of 1 g shaking table tests is designed to investigate the seismic response of a typical two-story and three-span subway station considering the influence of underground diaphragm walls. The stratum is simulated by synthetic model soil (a mixture of sand and sawdust), and the model structure and diaphragm walls are made by granular concrete with galvanized steel wires. A test case of the structure without diaphragm walls is also involved and taken as a benchmark comparison to understand the impact of diaphragm walls on the seismic response of subway station. The seismic excitations for the test include actual seismic records with the amplitude of 0.2, 0.4, and 0.8 g, respectively. Based on the test data analysis, a comprehensive discussion is conducted on the influence of diaphragm walls on the seismic design of underground structures. Current misconceptions that ignoring the role of diaphragm walls is a conservative way in seismic design of underground structures are also reviewed. Results show that the presence of underground diaphragm walls would enhance the lateral stiffness of the structure, and thus significantly reduce the lateral deformation of subway stations during earthquakes. Notably, the structure with diaphragm walls also exhibits a significant amplification in acceleration response and experiences greater dynamic earth pressures on the sidewalls, and furthermore the strains at the connection between the sidewalls and diaphragm walls are dramatically amplified during the earthquake. It is worth noting that these adverse effects of the diaphragm walls on the amplification of dynamic earth pressures on the structure as well as the increase of internal forces at the sidewalls end-diaphragm walls connection should be carefully considered in the seismic design of underground structures.

In this paper the results of 2D dynamic finite element analyses of a zoned earth dam are presented and discussed with the aim of detecting proper intensity measures able to describe the variation of the crest permanent settlement with the characteristics of the input ground motion. A large set of horizontal components of earthquake records has been selected and used as input motion considering both upstream and downstream directions. This allowed to explore the activation of plastic mechanisms in the dam embankment and the amplification of the horizontal accelerations. Starting from the analysis results, two original intensity measures, related to the amplitude, energy and frequency content of the input motion, have been detected and original empirical predictive formulas have also been derived to estimate the permanent crest settlement of the dam. The effectiveness of the proposed intensity measures and the reliability of the novel relationships between these intensity measures and earthquake-induced permanent crest settlements have been checked against (i) numerical results of further dynamic analyses carried out for the same dam using additional sets of input motions, (ii) numerical results available in the literature for two other earth dams and (iii) field data relative to the crest settlements induced in four earth dams by two recent large earthquakes. The whole set of analysis results allowed defining an empirical equation that appears as a promising general predictive tool for the earthquake-induced crest settlement of earth dams.

Bidirectional analyses covering a range of incidence angles for several records are essential yet challenging for routine design. Recent studies have shown that the response to bidirectional shaking may be estimated per straightforward unidirectional analysis in the most preferred orientation in conjunction with the orthogonal combination rules, and this can also remove the need for various incidence angles. Further, it is well known that the mean response can be estimated with reduced dispersion using fewer spectrum-compatible records. Combining the advantages of both, a pair of companion spectra independent of period and orientation has recently been developed. While the basis of the companion spectra is established therein from a sound mathematical perspective, the applicability of the proposed spectra to real structures may be revisited, especially to structures sensitive to bidirectional interaction, before recommending the companion spectra for practical design. To achieve this end, following a brief scrutiny of the companion spectra, the performance of the same is critically examined for idealized SDoF analogues and real structures such as bridge piers representative of SDoF systems and buildings with asymmetry using single storey and multistorey models (MDoF systems). The results show that the response of the real structures to bidirectional seismic loading can be reasonably estimated with reduced dispersion using fewer records compatible with the pair of the companion spectra. As such, the critical response parameters may be computed by actual bidirectional analyses or by combining responses to unidirectional analyses under fewer spectrum compatible records. Hence, the pair of companion spectra that can be used as “target” to independently match two components of a chosen record is useful for practical design.

Damage assessment of tall buildings after an earthquake is important for efficient postearthquake management due to social and economic reasons. Structural health monitoring (SHM) system enables rapid, remote, and objective condition assessment for tall buildings by controlling dynamic properties of structures. However, tracking only the changes in dynamic properties of tall buildings may not be sufficient for damage assessment. In this paper, changes in modal frequencies and maximum interstory drift ratio are investigated as damage assessment indicators because they can be obtained by analyzing vibration data recorded by SHM system. On the other hand, limited number of sensors are used due to economic reasons. Therefore, in this paper, firstly, a unique methodology on development and optimization of nonlinear simplified model for tall buildings is presented to estimate responses of noninstrumented floors from instrumented floors. After that, new threshold values are suggested for changes in dynamic properties and interstory drift ratios to reliably decide performance level of structures after earthquakes. Finally, the proposed method was validated with vibration record of a real damaged building.

Experiments have shown that the coefficient of friction in steel-polymer interfaces is influenced by several interacting parameters, making it difficult to accurately model the instantaneously varying coefficient of friction in sliding seismic isolation bearings. A new model to characterize the variation of the coefficient of friction in sliding seismic isolation bearings in which a thermoplastic slides relative to a steel surface is presented. The model takes into account the variation in the coefficient of friction produced by changes in sliding velocity, normal pressure, and stick-slip phenomena, as well as degradation due to frictional heating. In particular, the model takes into account the differences in behavior during reversing and nonreversing stick-slip transitions. It is shown that the proposed model is able to reproduce very well experimental results of friction tests using various steel-polymer interfaces.