ABSTRACT Seismic assessment is essential to maintain pile-supported wharves (PSW) functionality and limit earthquake losses. However, existing studies have not fully addressed the stochastic dynamic response of PSW systems and the evolution of their time-dependent seismic reliability. To address this, a Probability Density Evolution Method (PDEM)-based framework is developed to quantify reliability degradation under stochastic excitation. The results indicate pronounced non-Gaussian evolution of the response probability density over time, including broadening, peak reduction, skewness, and multimodality. Consistently, the system reliability exhibits a non-smooth, stepwise decline, implying that the underlying response process deviates from conventional Poisson or Markovian assumptions.

ABSTRACT The 2020 earthquakes in Croatia caused significant damage, particularly to churches, museums, cultural assets, and older buildings located in areas of pronounced topography in Northern Croatia (intensity VI, EMS). This study aims to derive insights from available macroseismic data on historical and recent seismic events, with the objective of characterising localised topographic damage patterns. Our specific inquiry is whether local ground effects, especially irregular topography, contribute significantly to damage beyond the influence of a building’s inherent structural condition. This research also highlights gaps in empirical evidence, historical data, and methodologies for analysing topographic effects.

ABSTRACT This paper examines the combined effects of an elastomeric rubber bearing and a submerged internal block on the nonlinear seismic performance of a rigid rectangular liquid container under earthquakes with various frequency contents. A nonlinear finite element model is developed incorporating a Coupled Eulerian and Lagrangian (CEL) formulation to address the nonlinear sloshing effects of liquid. The developed model has been validated using the benchmark results available in the literature. Parametric analyses have been carried out for different internal block configurations and isolation parameters to study their influence on essential seismic responses, namely slosh amplitude, hydrodynamic pressure components, and base shear. Results show that isolation is effective in shifting the fundamental sloshing frequency along with significant reductions in hydrodynamic pressure and total base shear. On the other hand, an internal block further improves the slosh suppression effectiveness. The nonlinear analyses also yield higher sloshing amplitudes compared to their linear counterparts, demonstrating the necessity of nonlinear modeling for accurate freeboard estimation. Overall, the present study establishes that the hybrid implementation of a submerged block and rubber bearing isolation is an effective vibration control strategy for liquid storage tanks subjected to earthquakes with varying frequency characteristics.

ABSTRACT The current reconstruction methods tend to use global normalization strategies, which fundamentally conflict with the local characteristics of missing seismic data traces. This often leads to mean and variance shifts, making it difficult to handle the heterogeneity of the missing regions. This paper suggests the Dynamic Region-Aware Normalization Network (DRA-Net), which employs a dynamic region-aware normalization mechanism to address the issue at hand. Through a staged, region-specific normalization strategy, it enables differentiated representation of missing and complete regions. Two modules are designed in this approach: the DRA-B module and the DRA-L module, which target the downsampling and upsampling stages of seismic data, respectively. The first module utilizes the original mask to guide region segmentation and applies regional normalization to mitigate the interference of global statistical bias on local features. The second module dynamically generates potential region masks by modeling spatiotemporal correlations and performs global affine transformations using learnable parameters, thus enhancing feature fusion between regions. Experimental results demonstrate that the DRA-Net exhibits excellent reconstruction performance on both synthetic and real-world seismic datasets. This method provides a new approach for high-precision seismic data reconstruction under complex missing patterns, with practical applications in reducing exploration costs and enhancing the reliability of oil and gas reservoir characterization. Our code is available at https://github.com/dingding-king/DRA-Net.

ABSTRACT This study proposes a unified framework for assessing bending-buckling failure of slender rock-socketed piles in liquefiable soils. Using a validated beam on nonlinear Winkler foundation (BNWF) model, it quantifies how axial loads amplify p-Δ effects, causing significant stiffness degradation. Study employs Receiver Operating Characteristic (ROC) analysis to establish a displacement-based buckling threshold and derives a novel governing equation incorporating embedment and slenderness ratios. Unified with the plastic hinge equation, it creates a failure mode partition diagram. Validated through case histories of Showa Bridge and Gaiko Warehouse, this approach offers a practical design tool for seismic pile foundations in liquefiable sites.

ABSTRACT This paper presents a set of fragility curves and machine learning-based damage classification models developed using strong motion data from the 2023 Turkey-Syria earthquakes. The models rely on post-event performance evaluations of 322 buildings and strong ground motion records measured during the events and use as input ground motion intensity measures/parameters computed directly from the ground motion records, such as Peak Ground Acceleration, Peak Ground Velocity, Cumulative Absolute Velocity, Root-Mean-Square Acceleration, Arias Intensity, Araya-Saragoni Intensity, Housner Spectral Intensity, and Pseudo-Spectral Acceleration. The study explores the damage-predicting capability and the correlations between the aforementioned intensity measures and the observed structural damage. The damage predicting models consist of a set of fragility curves generated using the maximum likelihood method, and a random forest classifier with a feature importance analysis that includes a SHapley Additive exPlanations (SHAP) model. The results show that, for the dataset studied, the Peak Ground Acceleration and Araya-Saragoni Intensity are the most effective damage predictors. The models demonstrate how post-earthquake data can inform seismic performance assessment and enhance risk-informed decision-making.

ABSTRACT Existing models for estimating expected losses due to future earthquakes are mainly based on empirical, site-specific data collected after real earthquakes, and generally refer to the maximum damage level observed in buildings. However, the expected direct economic losses also depend on the extension and distribution of damage along the building height. In this paper, a methodology for deriving analytical functions to estimate direct economic losses by explicitly accounting for both the severity and distribution of damage along the building height has been developed and applied to some RC building types widespread in Italy and Southern Europe.

ABSTRACT This paper provides a summary of the seismic behaviour exhibited by “rocking” partition walls, featuring dual-slot tracks, designed with different construction details tested under quasi-static reverse cyclic drift demands. The variations in construction details encompass different types of end-studs, reuse and placement of aluminium angles, horizontal gaps between adjacent plasterboards, and the L-shaped walls. It presents the discussions on seismic response, failure mechanisms, and onset of damage states of partition walls. Consequently, several key conclusions and recommendations are derived to enhance the seismic performance of the “rocking” walls.

ABSTRACT To address the limitations of traditional viscoelastic dampers in installation flexibility, architectural integration, and maintenance efficiency, a composite window-type viscoelastic damper (CWVED) is proposed. The force–displacement behavior of CWVEDs with different configurations is derived, and the influences of aspect ratio, opening ratio, shear length, and shear height on their mechanical performance are examined using a finite element model validated through experiments. Finally, a dynamic analysis is conducted on a 10-story reinforced concrete frame equipped with CWVEDs of different configurations. The results indicate that CWVEDs exhibit excellent seismic performance, and the four-point-connection CWVED achieves the greatest reduction in seismic response.

ABSTRACT The geological conditions are extremely complex in the coastal and riverfront cities in China, some shield tunnels are designed to inevitably pass through various saturated strata, while the dynamic response of tunnel structure passing through multiple saturated soil layers remains unclear. This paper investigated the dynamic responses of a shield tunnel passing through saturated sand and clay deposits under transverse and longitudinal earthquake excitations. The practical two-surface plastic model and hysteretic model were adopted to capture the saturated sand and clay, respectively. The sophisticated tunnel model was proposed to simulate the cyclic mechanical behaviour of segment joints. The numerical results reveal that tunnel structures near the soil interface present uneven upward deformation, resulting in a significant increase in lining force and joint displacements. Comprehensive parametric studies indicate that earthquake magnitude, clay type, sand density, and interface inclination have a major impact on the dynamic response of shield tunnels near the soil interface. It was determined that the grouting measure and rubber buffer layer could effectively alleviate the differential uplift of tunnel segments and the displacement of segment joints near the interface. The research results could serve as a theoretical reference for the purpose of seismic design of underground structures.