Soil Dynamics and Earthquake Engineering最新文献

Tunnels embedded in liquefiable soil are frequently subjected to uplift and sustain serious damage during major earthquakes. Mitigation methods to prevent the flotation of these tunnels must be developed and implemented in the natural environment. For this purpose, this study proposes a new method that uses stone columns to enhance soil drainage and mitigate soil liquefaction around tunnels. A circular tunnel in liquefied soil is simulated using a 2D finite element model, PLAXIS 2D, and subjected to a sinusoidal input motion. The tunnel's pre-construction and post-construction scenarios are examined. Parametric studies are carried out to investigate the effect of changing several parameters, such as the distance between stone columns and tunnel springing, the diameter of stone columns, the spacing between stone columns, the number of stone column rows, and the stone column arrangement patterns on the effectiveness of liquefaction mitigation. The study reveals that liquefaction mitigation is enhanced by using stone columns closer to tunnel springing, with larger diameters, less spacing, and more rows of stone columns that are arranged in a square pattern. It also emphasizes the importance of timely implementation of stone columns for maximum benefit.

The noise horizontal-to-vertical spectral ratio (NHV) has been applied for inverting subsurface velocity structures, but the non-uniqueness issue in the inversion remains prominent. Parameter sensitivity analysis is crucial for understanding the extent to which parameters influence inversion results, providing reasonable value ranges for each parameter, and offering rational constraints to mitigate the non-uniqueness of solutions. This study takes 636 sites from the KiK-net network as benchmark sites and employs the Monte Carlo method based on Toro's statistical model to generate 200 random samples for each site's shear wave velocity (V_S), compressional wave velocity (V_P), layer thickness (h), and density (ρ) parameters. These random samples are then combined into six scenarios, serving as the stochastic site models for uncertainty analysis. Finally, based on the diffuse field assumption (DFA) for NHV forward modeling, the NHVs and their standard deviations are calculated for each scenario. The standard deviations of NHV are further utilized to conduct a sensitivity analysis of soil layer parameter uncertainties' impact on NHV. The results indicate: 1) The NHV peak frequency (f_peak) is most sensitive to the V_S and h combination, followed by the V_S and ρ combination. Among single-parameter scenarios, f_peak is most sensitive to V_S, followed by h, and relatively insensitive to V_P and ρ. 2) The NHV peak amplitude (A_peak) is most sensitive to the V_S and ρ combination, followed by the V_S and h combination. Among single-parameter scenarios, A_peak is most sensitive to V_S, approaching the sensitivity level of the V_S and h combination, followed by ρ, and relatively insensitive to h and V_P. 3) Within the 0.1–50 Hz frequency band, the NHV curve is most sensitive to the V_S and h combination, followed by the V_S and ρ combination, and the single-parameter V_S scenario. Among single-parameter scenarios, the NHV curve is secondly sensitive to h, while its sensitivity to V_P and ρ is relatively low. The sensitivity patterns of NHV to soil parameters revealed in this study are widely applicable, providing effective constraints for NHV inversion, assisting in optimizing parameter combinations, assessing the feasibility of inversion schemes and the credibility of results, and offering beneficial insights for improving the efficiency of the inversion process.

This paper presents a novel analytical model based on a fictitious saturated soil pile (FSSP) model, Biot's theory, and the Novak plane strain model for analyzing the horizontal vibration of a floating pile in saturated soil. First, the horizontal displacement of the saturated soil and horizontal resistance to the pile were determined using the separation variable method and potential functions. Next, the horizontal vibration response of the pile was obtained by deriving differential equations specific to the FSSP, considering its complete contact with the solid pile. The analytical model was then validated by reducing the proposed solution and comparing it with solutions presented in the literature. Finally, the effects of various soil parameters, such as the FSSP length, damping ratio, elastic modulus, and porosity, on the horizontal vibration response of the floating pile were analyzed.

The environmental vibration problems caused by underground trains have recently received widespread attention. An accurate prediction method is essential for vibration assessment around metro lines and for implementing necessary vibration isolation measures. Previous studies have indicated that a hybrid method that combines numerical modelling and experimental measurements can effectively reduce prediction uncertainty with wide adaptability. However, few studies have reported hybrid methods for predicting environmental vibrations caused by underground trains. However, these methods are limited owing to inconvenient excitation experiments in tunnels. Therefore, this study proposes a convenient hybrid prediction method. Subsequently, an experimental study was performed to validate the applicability of the Betti–Rayleigh dynamic reciprocal theorem to the proposed method. A case study was conducted using numerical simulations to verify the feasibility and accuracy of the proposed hybrid method. Finally, a numerical study was conducted to investigate the influence of adjacent hammer spacing and line-source length on the prediction results. The study results demonstrated that the Betti–Rayleigh dynamic reciprocal theorem is applicable to the proposed hybrid prediction method. Hybrid prediction method has been proven to exhibit high accuracy. The adjacent hammer spacing and line-source length can affect the prediction accuracy. Accordingly, the adjacent hammer spacing should be smaller than 19.2 m, and the line-source length should be larger than 80 m in underground train-induced vibration prediction projects under similar conditions.

Data driven models are useful in various classification and regression problems in bridge engineering. Although machine learning applications in identifying the performance and characteristics of bridges has been increasing, repair cost modeling using real earthquake damage data is only sparsely reported in the literature. This study assesses various machine learning models for repair cost estimation of reinforced concrete (RC) bridges damaged by the 2015 Gorkha earthquake in Nepal. Ensemble learning interpretation is used to hierarchize bridge attributes based on their relative importance in repair cost prediction. The impact of individual features is also assessed using various combinations of bridge features. The results indicate that two categorical features, discrete damage state and foundation type, are the most important in predicting repair cost. Of the various models tested, extra trees (ET) ensemble outperformed any other ensemble as well as base learner methods. The findings indicate that, for the case-study data used here, repair cost is better estimated by a classification model for damage state in series with a regression model than just a regression model.

As structural safety emerges as a paramount concern and with the advancement in vibration control technology, a notable gap persists in accurately simulating structural state nonlinearity, especially within the field of inerter-based seismic control technology. Dealing with the need to upgrade the seismic performance of practical nonlinear structures, this study proposes a structural state nonlinearity-based design method for inerter-incorporated civil structures by incorporating the target seismic performance, structural nonlinearity level, and ground motion effects. By establishing mechanical models of inerter-based systems with straightforward parameters and a prototypical bilinear model for the primary structure, governing equations for inerter-incorporated structures were derived. Later, structural state nonlinearity-based design method for inerter-based systems, including optimization criteria and parameter distribution pattern, is elaborated. Simultaneously, this study also provides structural state nonlinearity-based design curves and modification formulae related to structural state nonlinearity and peak ground acceleration of earthquake. Validated using a 10-story multiple-degree-of-freedom (MDOF) structure subjected to earthquakes, the results highlight the effectiveness of the method, emphasizing its potential in controlling seismic responses of structures within the predetermined performance target. In addition, the absolute acceleration mitigation effect of inerter-based systems is investigated. The efficiency and robustness of the proposed method are also verified under near-fault and far-fault earthquakes. In essence, this research pioneers an approach in inerter-based design, bridging structural state nonlinearity, and potentially reshaping seismic control strategies to align with diverse structural states.

Many grain silos in earthquake intensity areas are at significant risk of post-seismic damage, which compromise their functionality and pose an enormous challenge to post-disaster rescue and social stability. However, the current specifications based on fixed-base foundations are not fit for the seismic design of silos in soft soil areas. Therefore, it is of great practical significance to study the seismic disaster prevention of siloes considering soil-structure interaction (SSI) for food security and post-disaster supply. In this research, a column-bearing silo in a soft soil area is taken as the research object. The relative displacement response, elastic-plastic development and storage lateral pressure of the structure under different ground motions are studied in detail when the state of filling storage material is empty, half-filled and fully filled. Compared with the fixed-base model, the mechanism of the ground motion response and the influence of the SSI effects on the dynamic characteristics of the column-bearing silo structure under different storage conditions are revealed. In addition, structural vulnerability analysis is carried out with the incremental dynamic analysis (IDA) method. Finally, the structural damage probability with and without SSI effects under different storage conditions is further discussed. The results demonstrate that the SSI effects have a certain damping effect on the column-bearing silo. The amount of storage material changes the failure probability of the structure. Moreover, full-silo is the most dangerous condition, indicating that the filling state of storage material affects the stiffness degradation. This study provides theoretical insight to the influence of the SSI effect on the seismic resilience of structures.

In recent years, there has been a growing recognition of the importance of vertical ground motions in the seismic design of engineering structures. A comprehensive understanding of the small-strain constrained modulus M₀, which is a key input soil parameter, is essential for conducting a reliable analysis of vertical site response. Natural soils in engineering scenarios are often subjected to various anisotropic stress states, and the role of such loading on M₀ is a critical concern that remains incompletely understood. This paper presents a systematic experimental program aimed at addressing this issue. Using a triaxial apparatus, sand specimens initially isotropically consolidated were subjected to various anisotropic stress states, including triaxial compression and triaxial extension. The evolutions of M₀ at different stress states were captured by exciting elastic compression waves through embedded bender-extender elements. The specimens were tested under a wide range of states in terms of void ratio, axial stress, and radial stress. The study demonstrates that the impact of stress anisotropy is complex, depending on the magnitude of the stress ratio, the loading mode, and the initial state of the specimen. A practical model is suggested for the improved characterization of M₀ under the anisotropic stress states. This model considers two primary mechanisms that are associated with the effects of stress anisotropy.

This paper proposes a re-centring and self-balanced inerter (RSBI). The rhombic linkage achieves self-balanced of the torque on the screw, which releases the constraints demanded at the end of the screw and reduces the working wastage and cost of the screw. The flywheel's re-centring feature is guaranteed by introducing a self-resetting spring and enhances the stability of the system. More specifically, firstly, the resonance analysis method provides the optimized mounting angle of the RSBI's rhombic linkage. Then, the user-friendly optimal design strategy of the four-parameter inerter system, are derived by applying the fixed-point theory, and the validity of the optimized parameters is verified by parameter and time history analysis. Finally, nonlinear model of the RSBI is performed to account for the nonlinearity due to the variation of the rhombic linkage angle at large working strokes. The nonlinear amplitude frequency response function of the system is obtained using the harmonic balance method and the working stroke of the inerter is classified by comparing it with the linear frequency response function. The device proposed in this paper provides a good control effect on the displacement and acceleration control of the main structure under multiple seismic waves. The nonlinearity of the device is appropriately exploited to almost double the working stroke of the inerter, which can effectively reduce the size of the device.

Due to the limited features and poor accuracy of current methods for predicting the dynamic response of subgrades, this paper proposes an innovative approach that combines subgrade dynamic response field tests and machine learning (ML) technology. This method uses Bayesian optimization XGBoost (BO-XGBoost), Bayesian optimization LightGBM (BO-LightGBM), and Bayesian optimization CatBoost (BO-CatBoost) models to analyze the effects of physical properties and stress conditions on the dynamic stress, dynamic acceleration, and dynamic displacement of the subgrade. The optimal ML model was selected on the basis of the residuals, coefficient of determination (R²), mean squared error (MSE), and mean absolute error of the prediction results. Using SHapley additive exPlanations (SHAP), the global importance, feature importance, and feature interaction behaviours of the optimal ML model input features were explained, and the main controlling features affecting the dynamic stress, dynamic acceleration, and dynamic displacement of the subgrade were obtained. The research results indicate that the prediction results of the BO-XGBoost, BO-LightGBM, and BO-CatBoost models for dynamic stress, dynamic acceleration, and dynamic displacement are mostly within the 10 % error range, and the R² values of these three models are greater than 0.98. On the basis of the comparison results of the hyperparameter combinations, the objective of MSE (MSE_obj), and the error evaluation metrics, the BO-CatBoost model yields the highest prediction accuracy, making it the optimal ML prediction model. This prediction method can quickly and intelligently obtain the main controlling features of dynamic stress, dynamic acceleration, and dynamic displacement, including depth (H), axle load (P), frequency (f), and moisture content (w). The boundary conditions for these four features are as follows: H > −1.3 m, P > 10 ton, f > 3.7 Hz, and w >18.1 %. The research results contribute to enhancing the service performance and lifespan of expressways.