Earthquake Engineering & Structural Dynamics最新文献

Viscoelastic (VE) dampers are capable of dissipating energy over a variety of input vibration frequencies. They supplement both displacement- and velocity-dependent restoring forces, thus, provide both stiffness and damping to the structure. For this, they are able to effectively mitigate both frequently occurring wind loads and seismic forces. They are sensitive to loading frequency, temperature, and strain level. Their combined sensitivity to both loading frequency and temperature is extensively researched through three-dimensional finite element (3D-FE) methods considering heat generation and transfer. However, they can experience a significant nonlinear reduction in dynamic mechanical properties under large strain levels. Pursuant to these, the previously developed 3D-FE method is extended in this study by combining with a nonlinear strain level–sensitive constitutive rule to investigate the behavior of a full-scale multilayer VE damper. This nonlinear 3D analysis method agrees accurately well with experimental results. Additionally, the 3D-FE analysis results suggest an approach to one-dimensional (1D) time-history analysis. Despite of large strain level, the energy dissipation obtained from 3D-FE analysis is uniform, further suggesting a framework for a simplified 1D approach, that is, considering uniform strain distribution. These 1D methods accurately predict VE damper global responses.

Machine learning (ML) has recently been used as an efficient surrogate to estimate different steps of performance-based earthquake engineering (PBEE), from dynamic structural analysis to fragility and loss assessments. However, due to the varied data, models, and features in existing literature, the relative efficiency of ML models across different PBEE steps remains unclear. Additionally, the black-box nature of advanced ML algorithms limits their ability to provide design-oriented insights, hindering the broader application of ML in PBEE-based design. This study provides a comprehensive comparison of the accuracy and explainability of design-oriented ML models across different steps of PBEE using a consistent database of 621 steel moment frames with varying designs and geometry. Eight ML algorithms were used in a careful training workflow comprising feature selection, hyperparameter tuning, cross-validation, and model inference. The sensitivity of model accuracy to representative PBEE outputs—maximum responses, median fragility, and expected annual loss—was assessed using statistical measures. In addition, the explainability of the best models for each step was examined to explore the relationship between design parameters and the corresponding PBEE output. The results show that while ML models can reasonably map design parameters to all different PBEE outputs, models accuracy was higher for drift responses, median fragilities, and component-based loss metrics. In addition, the optimal algorithm remained the same across different PBEE steps, where support vector machines and random forests provided the highest accuracy with an average R² of 0.93 and 0.91 over different outputs on the test set. Although the selected feature sets varied across outputs and algorithms, height, number of stories, fundamental period, and the minimum of the beams’ moment of inertia were influential for both models and notably affected different PBEE outputs.

This article presents the authors' experience with large-scale shaking table tests conducted in Japan using the E-Defense shaking table. The discussion focuses on four criticisms often addressed regarding the utilities of large-scale shaking table tests. Potential solutions to mitigate such criticisms are discussed based on shaking table tests conducted for a pair of three-story wooden houses. The first criticism is that the test specimen anchored rigidly to a rigid shaking table is not a reproduction of actual structures supported by soils and foundations. A model ground was developed in a large sandbox, which occupied about 85% of the total specimen weight, supported the house, and the entire soil-structure system was shaken. Considerable sliding occurred, having lessened the earthquake forces exerted and resultant damage to the superstructure. The second criticism is that a single specimen test, regardless of its size, cannot provide sufficient information for generalizing the behavior and performance. Empirical equations between the maximum story drift and the change in the natural frequency were developed from a series of shaking table tests. Using such empirical equations might promote quick damage assessment of individual houses when suffering from actual earthquakes. The third criticism is the importance of public appeal and eventual support from the general public to secure the budget to operate large-scale testing facilities. The example test featured two nearly identical specimens placed on the table with different support conditions. The apparent difference in response revealed the effect of support conditions on seismic performance. The fourth criticism is the importance of increasing the number of experimental projects to balance the operation budget. Most of the preparation in the example test was accomplished in an open yard adjacent to the shaking table, and the test specimens were quickly assembled on the table using indoor cranes. The table occupation was four out of 35 weeks of the entire test duration.

A three-dimensional (3D) base isolation system can enhance the seismic performance of steel frame structures. However, the notable coupling effects and rocking behavior in the superstructure will complicate the design of base isolators and weaken the seismic resilience of the 3D base isolation system. To realize the 3D isolation function and constrain the coupling effects and rocking behavior of steel frame structures simultaneously, a layered 3D isolation system is proposed through the combination of a horizontal base isolation with friction pendulum bearings and a vertical floor isolation with steel coil springs installed beneath the floor slabs. A series of full-scale shaking table tests of a two-storey steel frame structure were conducted to verify the effectiveness of the proposed system. The test results indicate that the layered 3D isolation system with flexible horizontal displacement constraint devices decoupled the horizontal and vertical motions of the superstructure and improved the seismic behavior of steel frame structures. The horizontal–rocking coupling effects of the superstructure typically associated with traditional 3D base isolation were significantly suppressed. Due to the vertical acceleration isolation effects, the layered 3D isolation system led to stabler normal pressures on the friction pendulum and lower friction forces compared to the horizontal base isolation system, which mitigated the horizontal–vertical coupling effects of the structure and the stick-slip motions of friction pendulums to result in higher horizontal acceleration isolation effects and self-centering capacities. The unfavorable influence of elevating the floor slab on the seismic behavior of the superstructure was compensated by the horizontal isolation at the base.

This study investigates the seismic response of lead rubber bearings (LRBs) under constant and time-varying axial forces using real-time hybrid simulation (RTHS). Although shake table testing can provide realistic seismic responses, it is often expensive and quite challenging for large-scale structures. RTHS, however, offers a cost-effective alternative by experimentally testing only the structural component of interest while analytically modeling the remaining structure. With the use of an advanced real-time force control method, this study implemented RTHSs for a bridge isolated with two LRBs, where the LRBs are subjected to various constant axial forces due to self-weight as well as time-varying axial forces induced by vertical ground motions. The lateral response of LRBs was found to be significantly influenced by the magnitude of constant axial force, highlighting the importance of incorporating the effect of axial force due to self-weight into numerical simulations. Additionally, it is crucial to satisfy the axial force boundary condition when conducting RTHS or cyclic loading tests to obtain more reliable test results. On the other hand, the time-varying axial force generated by the vertical vibration of the bridge due to vertical ground motions has a limited impact on the lateral response of LRBs.

The implementation of seismic isolation in lightweight structures, such as houses, is limited, except for Japan, where nearly 5000 houses are seismically isolated with highly engineered systems. The use of such systems in countries of growing economies is prohibitive due to their cost, difficulties in locally fabricating components, and incompatibility with local construction practices. Previous studies on low-cost isolation systems showed that isolation devices with rolling rather than sliding elements are more suitable for lightweight structures. A low-cost isolation system based on a deformable rolling bearing developed at the University at Buffalo has simple components that can be easily manufactured, has a large displacement capacity, and features a fail-safe mechanism. The bearing is composed of one flat and one concave concrete plate and a rubber rolling ball. This paper presents the results of shake table testing of half-length scale versions of prototype houses and small bridges equipped with this deformable rolling isolation system in three-directional seismic excitation. The results showed that deformable rolling bearings provide significant seismic protection of lightweight structures against strong ground motions, which in the testing reached a maximum peak ground velocity of 1.2 m/s in the test scale and a peak ground acceleration of 4 g in the horizontal direction and up to 1 g in the vertical direction. The isolation system was particularly effective in the horizontal direction, whereas it did not provide any vertical isolation. The tests included cases of extreme response, which included vertical impact within the bearing's components and uplift of the isolated structure.