Ying Zhou

Accurate and rapid prediction of nonlinear structural dynamics is central to structural design, disaster mitigation and resilience assessment, yet conventional finite-element analysis remains too computationally demanding for real-time or portfolio-scale applications. To address this computational bottleneck, we develop SeisGPT, a physics-informed foundation model for high-fidelity response prediction across multistorey buildings with diverse structural types and topologies. Its architecture integrates structural-mechanics priors, storey-resolved structural representations and a spectral decoder that captures modal response propagation while accommodating learnable nonlinear corrections. The model is trained on more than 2 million nonlinear elastoplastic simulations derived from 270,000 automatically generated, code-compliant structural designs and 694 real-world buildings, together comprising over 10 billion response time steps. For previously unseen buildings, SeisGPT predicts displacement and acceleration histories with normalized errors below 5% and achieves an approximately 40,000-fold computational speedup relative to conventional finite-element analysis. By assimilating sparse sensor measurements, SeisGPT reconstructs full-building responses from limited observations and, in shake-table tests, shows closer agreement with measured responses than finite-element simulations, supporting structural-health monitoring and post-earthquake damage-state assessment. These results indicate that physics-informed foundation models can support high-fidelity structural-response prediction, sparse-sensor reconstruction and response-informed downstream structural assessment at inference speeds compatible with real-time and portfolio-scale use.

Promoting the construction of new urban infrastructure and building resilient cities represent a key trend in urban development. Achieving resilient power plants is crucial for improving their sustained operation capability, adaptability, and recovery capacity in the face of uncertain hazards, while also helping to reduce economic losses and negative social impacts. Functional quantification forms the basis for analyzing functional loss and recovery processes, as well as for evaluating the resilience of power plants after disasters. However, because power plants consist of various types of equipment and piping, understanding the functional relationships between components and systems presents a challenge. Although quantitative studies have been conducted on the probability of failure or safety margin under basic events or natural disasters, research on functional quantification methods that can effectively evaluate post‐disaster functions is still limited. In this paper, a calculation method for different types of components in thermal power plants is presented, along with examples that include components with single‐input single‐output, single‐input multiple‐outputs, multiple‐inputs single‐output, and multiple‐inputs multiple‐outputs, categorized based on their input–output functional relationships. In order to demonstrate the application of the proposed method in quantifying the functions of thermal power plants after disasters, this paper conducts a case study on the functional loss analysis of the coal handling and coal pulverizing systems of a 2 × 660 MW ultra‐supercritical unit under earthquakes. This method can provide a theoretical basis for the quantification of functional loss and recovery, as well as for the resilience assessment of power plants after disasters.

Self‐centering (SC) precast concrete wall structure is one of the effective systems to enhance seismic resilience within the low damage structures community, which fulfills the demand for low damage and high cost‐effectiveness. Most past studies have focused solely on the structural level, for example, on general seismic performance, the influence of different energy dissipation devices, and design methods. Little attention is paid to the coordination between SC wall and non‐structural components, such as infill walls, the latter of which plays an essential role in terms of achieving the integral seismic resilience of buildings. To this end, this study reports on the experimental results of SC wall structures incorporating various innovative infill walls. The primary objective of the proposed new infill walls is to coordinate the deformation between the structural skeleton and infill walls, thereby fostering an integrated seismic resilient structural system that is promising in earthquake‐prone areas. Moreover, the study explores the effectiveness of energy dissipators equipped between infill wall panels. One of the purpose of using addition damper in infills is to reduce energy dissipation demand for SC walls while preserving ease of replacement. The outcome of test results covers various aspects, including force–displacement relationships, degradation of structural performance indexes, residual displacement, response of slotted beams, energy dissipation, and some detailed results for the SC wall. Results indicate that the tested specimens incorporating infill walls exhibit great seismic performance in terms of achieving low damage and self‐centering behavior as expected. The inclusion of dampers between infill wall panels enhances energy dissipation capacity for structure. Notably, specimens equipped with U‐shaped steel dampers and viscoelastic dampers demonstrate significant improvements to energy dissipation capacity, as increased by 32.1% and 22.4% under design target displacement, respectively, compared with those without dampers.

A confined precast prestressed hollow‐core wall structure was proposed to improve the construction efficiency of the building. The system comprises three standardized components that are prestressed hollow‐core wall panels, precast confined frames, and prestressed hollow‐core floor slabs, with in‐situ concrete for topping and beam–column joints. To verify the seismic performance of the system, a shaking table test was conducted on a full‐scale three‐story building that was designed using the proposed concept. The building was subjected to several shaking table tests with a range of intensity ground motions, incorporating both unidirectional and bidirectional loading. The test results showed that the structural system performed effectively during the intense series of tests. The building exhibited good seismic performance and achieved the performance objectives. The damages were mainly concentrated at the hollow‐core wall toe and the horizontal and vertical connection joints between wall panels. The damage state was low and did not result in strength reduction of the building during DBE earthquakes. The test results validate the seismic safety of the system, demonstrating the proposed system is a reliable solution for precast construction.

In recent earthquakes, poor performance and collapse of buildings attributed to inelastic torsion responses have been reported. The torsional responses of buildings have been extensively studied with analytical models, but large‐scale experiments considering seismic‐induced torsion demands on buildings are rare. A shake‐table test of a two‐storey, low‐damage concrete wall building with strength asymmetry in the plan view was recently conducted, and the response of the building to both unidirectional and bi‐directional loadings at the design‐level earthquake intensity is presented. Analysis of the results includes the global responses of the structural components and the centre of mass for the test building, the strength response of unbonded post‐tensioned walls, the interaction between the translational strengths and torque, the total absorbed energy related to the torsional rotation, and variations in the eccentricities of the centres of rigidity and strength. Furthermore, the seismic response of the test building is explained with a comparison to previously published torsional mechanisms. Key outcomes from the analysis of the test results included: (1) an increase in the strength asymmetry and bi‐directional loading both amplifies the torsional rotations and effects the displacement and strength responses of the structural components along the perimeter of the diaphragm; (2) the comparison to previous torsional mechanisms shows a desirable level of agreement and supports the utilization of these mechanisms to assess the torsional performance of low‐damage wall buildings with irregularities; and (3) despite exhibiting a strong torsional response, the test building with the asymmetric strength configurations still achieved the low‐damage performance targets.

Thick rubber bearings (TRBs) have been proven to be effective in mitigating horizontal shaking induced by earthquakes as well as railway-induced vertical vibration. During earthquake excita...

Accurate crack detection is essential for structural damage assessment after earthquake disasters. However, due to the gap between the target domain of the detected structure and the source domain, it is challenging to achieve high-precision crack segmentation when performing crack detection based on deep learning (DL) in actual engineering. This article proposes a crack segmentation transfer learning method based on a self-supervised learning mechanism and a high-quality pseudo-label generation method, which can significantly improve the detection accuracy in the target domain without pre-made annotations. Besides, to improve the crack segmentation model’s ability to extract local and global features, this article proposes a SegCrackFormer model, which embeds convolutional layers and multi-head self-attention modules. An experiment of the crack segmentation transfer learning method is performed on two open-source crack datasets, METU and Crack500, and a newly proposed LD dataset. The experimental results show that the crack segmentation transfer learning method proposed in this article can improve the mean intersection over union (mIoU) by 38.41% and 15.66% on the Crack500 and LD datasets, respectively. The proposed SegCrackFormer is evaluated through comparative experiments, which demonstrate its superiority over existing crack segmentation models on the METU dataset. Additionally, the proposed method is shown to require significantly less computational resources than other existing models, which highlights the potential of SegCrackFormer as a powerful and efficient model for crack segmentation in practical applications.