This study investigates the influence of corrosion damage on the axial load capacity of steel angle members using an integrated experimental and numerical methodology. A total of twelve specimens, characterized by varying cross-sectional dimensions, slenderness ratios, and corrosion rates, were subjected to axial compression testing, with accelerated corrosion induced through controlled neutral salt spray exposure. Three-dimensional topographic scanning techniques were utilized to assess the morphology of the corroded surfaces, which were subsequently reconstructed using Fractal Brownian motion theory. Parametric studies were conducted to analyze the effects of cross-sectional dimension, non-uniform corrosion layer thickness (UNCLT), width-to-thickness ratio (WTR), slenderness ratio, and corrosion rate on the ultimate load reduction factor of corroded steel angles. The parametric analyses indicated that the WTR is a significant determinant, leading to over a 40% variation in the ultimate load reduction factor among steel angles with identical corrosion rates but differing WTRs. Additionally, the interaction between slenderness ratio and WTR was found to affect the ultimate load of the steel angle considerably, underscoring the necessity for their joint consideration in evaluations of corrosion damage. A practical analytical expression was established to forecast the ultimate load reduction factor of corroded steel angles, incorporating corrosion rate, slenderness ratio, and WTR as critical variables. Validation against experimental data and existing literature confirmed the robustness of the proposed method, with prediction errors consistently remaining below 15%. The results offer essential insights for assessing the residual capacity of corroded steel angles in practical engineering.

Critical infrastructure with stringent seismic safety requirements, such as nuclear facilities, necessitates three-dimensional (3D) acceleration control. While conventional seismic isolation systems effectively mitigate horizontal seismic motions, they often fail to adequately reduce vertical seismic components and may even amplify them under specific excitations, thereby compromising the overall isolation performance. To address this gap, this paper develops an integrated 3D seismic isolation system, which incorporates newly developed vertical isolation devices with vertically softening nonlinear stiffness and anti-rocking linkages. The vertical devices maintain or even extend the isolation period under gravity load, thereby ensuring vertical isolation, while the linkages suppress rocking motions without compromising vertical translation, a non-hydraulic design which ensures greater durability and adaptability than previous hydraulic anti-rocking systems. The study derives the theoretical model and design formulation for the vertical isolation devices and completes their experimental validation. Furthermore, the theoretical model for the anti-rocking linkages is derived, and integrated theoretical and numerical models of the full 3D isolation system are established for a typical small modular nuclear facility (SMNF). The influence of rocking modes on the structural system and the effectiveness of rocking suppression were analyzed. Subsequent shaking table tests on the typical model demonstrate that the responses of the 3D isolation system show good agreement with the theoretical and numerical predictions. Compared to the fixed-base configuration in the case studies, both numerical simulations and shaking table tests confirm that the 3D isolation system reduces peak vertical seismic accelerations of the superstructure by over 79.63% and 74.74%, and peak horizontal accelerations of the superstructure by over 82.65% and 84.30%, respectively, while effectively controlling rocking motions. The results demonstrate the promising potential of the present system as highly efficient 3D isolation solution.

To enhance the seismic resilience of electrical equipment, this study developed a novel slope-type sliding isolation bearing (SSIB), which features frequency-independent characteristics, inherent self-centering, and stable anti-overturning performance. The configuration of the SSIB is first described, and the theoretical model is derived. Two laboratory-scale SSIB specimens were conducted and subjected to various cyclic loading conditions to evaluate the hysteretic behavior. Subsequently, a numerical model incorporating velocity-dependent friction was developed in OpenSees platform and validated against experimental results. A converter valve was selected, as a case study, to perform a comprehensive parameter investigation and evaluate the effectiveness of the SSIB system for seismic protection of electrical equipment under earthquakes. Results show that the numerical model accurately predicts the hysteretic behavior of the SSIB. The optimum sloping angle of the SSIB should be maintained at 2.5 degrees, and dual-control of acceleration response and bearing displacement can be achieved by optimizing the friction coefficient. The SSIB system with the optimum parameters can achieve both vibration reduction (60 %-90 %), stable dynamic responses of the isolated valve, and exceptional self-centering capability, demonstrating reliable performance across all tested seismic excitations and seismic intensity levels.

This study investigates the seismic mitigation performance and energy dissipation characteristics of the tuned negative stiffness amplifying damper (TNSAD) in high-rise isolated connected structures (HICS). To this end, the simplified structural model is extended to a multi-degree-of-freedom (MDOF) system. Using stochastic response analysis under white noise excitation, the variation patterns of structural responses with respect to system parameters are elucidated. The optimal parameters were ascertained through an optimization objective focused on minimizing the deformation of the isolation floor and the acceleration response of the connecting corridor. By comparing structural responses under different types of actual ground motion excitations, the seismic mitigation performance of HICS with and without TNSAD control is revealed. Furthermore, an energy analysis is carried out to examine the full process of energy dissipation and transfer produced by TNSAD in HICS. An investigation is conducted to assess the energy transformation and dissipative capability of TNSAD when exposed to far-field, near-field non-pulse, and near-field pulse seismic events. The energy dissipation contributions of the inherent damping of the connected structure, the isolation bearings, and TNSAD are compared. The results indicate that the integration of TNSAD into HICS achieves superior effectiveness in mitigating both the displacement of the base-isolated floor and the acceleration behavior of the connecting corridor across a range of seismic excitation scenarios. It dissipates the majority of the input energy, thereby reducing the energy dissipated through the deformation of isolation bearings and markedly mitigating the risk of damage to the primary structure. Furthermore, under a moderate supplementary damping ratio, TNSAD achieves superior energy transformation and dissipative performance, comprehensively substantiating its mechanism for damping amplification and facilitating effective regulation of HICS.

Conventional seismic assessments presuming uniform corrosion (UC) in RC structures may compromise evaluation accuracy. This study compares seismic behavior between UC and random story corrosion (RSC) structures, developing an innovative pre-seismic functionality loss model with an ML framework for RSC prediction. Four ML algorithms including two single learning methods (Support vector machine, Artificial neural network) and two ensemble learning methods (Random forest, Adaptive boosting) were selected to develop the predictive model. A seismic response test dataset of 324 RSC structures, obtained through the inter-quartile range (IQR) method, was utilized to conduct a series of nonlinear dynamic and static analyses in OpenSees. Results demonstrate that the seismic performance of RSC structures deteriorates progressively with elevated corrosion rates, PGA, and variation coefficients. In extreme cases, seismic fragility shows 1.02-12.07% increment compared to UC structures at PGA= 1.5 g, while seismic resilience decreases by 1.93-22.04% at PGA= 0.1 g. Additionally, the best model for seismic performance prediction is the ANN model which has a R2 of 0.96. The RSC model, grounded in the improved pre-seismic functionality loss model and integrated within the ML prediction framework, provides a more accurate basis for engineering decision-making.

This study proposes and evaluates a steel modular building (SMB) system employing a novel boltless interlocking, which balances the requirements of rapid assembly/ disassembly with enhanced structural performance. This boltless interlocking SMB system utilizes a self-adapting and self-locking interlocking mechanism, ensuring excellent connection reliability and ultra-high strength, while significantly simplifying the construction process of the SMB. It provides an effective solution for the automated assembly and tolerance control of modular buildings, which promises to considerably accelerate the automation in construction. As the joint configuration of this system possesses highly complex geometry, it is difficult to model directly and accurately in conventional finite element environments. Therefore, this paper employs Rhino software for three-dimensional geometric modeling and imports the data seamlessly into the Abaqus platform to perform finite element analysis, thus achieving high-fidelity numerical simulations that accurately reflect the complex geometry. Using numerical simulation methods, this paper analyzes the mechanical response of a three-story SMB structure under gravity load, wind load, seismic action, and a corner column failure scenario. The results demonstrate that the boltless interlocking system substantially enhances the tensile, shear, and bending capacities of the joints, with its load capacity being approximately five times that of traditional bolted connections. Under gravity load, the maximum floor slab displacement is only 7.45 mm, indicating excellent overall stability of the structure. The structural performance of SMB under lateral loads is satisfactory, with maximum inter-story drift ratios of 0.035% and 0.062% under combined wind-gravity and seismic-gravity loads, respectively. Anti-progressive collapse analysis indicates that the SMB structure can develop an efficient alternative load path after the failure of a corner column, with an ultimate load capacity reaching 1315 kN.

Applications of tuned mass dampers (TMDs) are widely recognized as an effective control strategy for safeguarding building structures against damage caused by earthquakes. Recently, the strategy of integrating TMD with an inerter-based device has emerged as a viable alternative to TMD, offering enhanced control performance under identical conditions. However, the scarce literature reports on the performance enhancement of the existing TMD via replacing its damping element with an optimally designed tuned inerter damper (TID), which is referred to as a tuned mass inerter damper (TMID). Moreover, most previous research has overlooked the inherent difficulty in adjusting the parameters of existing TMDs. Therefore, the main contribution of this study is to investigate the performance enhancement of the existing TMD by individually adjusting the design parameters of TID, which can be considered as a more practical methodology. The analytical expression for effective damping ratio (EDR) is utilized as a performance metric to assess the control performance of TMID. Subsequently, the closed-form solution for the optimal design parameter of TID is proposed to maximize the value of EDR, thereby achieving optimal control performance. The design procedure is summarized to provide a clear description of the TMID design process. A comprehensive performance comparison between TMID and TMD is conducted, utilizing both EDR and time-history analysis, to elucidate the underlying mechanism of enhanced control performance in the existing TMD system. Analysis results demonstrate that the TMID constitutes an effective control strategy to enhance the performance of existing TMDs, particularly when the existing TMDs are mistuned.

This paper presents an innovative self-centering L-shaped column (SCLC) system, which is designed to streamline the construction of post-tensioned precast reinforced concrete columns for enhanced seismic resilience. Unlike existing research that predominantly focuses on shear walls or rectangular beam-column joints with complex connection details, this study addresses a critical research gap concerning L-shaped column—commonly used in low-rise buildings—by proposing a simplified system that utilizes prestressed steel bars extending across connection interfaces, eliminating the need for complicated connections. A comprehensive design methodology for the SCLC system is developed based on the calculation of biaxially eccentrically compressed normal section capacity and shear resistance. The proposed computational framework is validated through a practical design case study of a three-story residential building. Furthermore, an experimentally validated finite element model is developed, and the simulated skeleton curves are compared with theoretical analysis results. The findings indicate that the theoretical calculations meet the requirements for bearing capacity verification and structural design of self-centering L-shaped columns, while the finite element model accurately captures the global seismic response. A parametric study is conducted to investigate key design variables, including axial compression ratio, prestressed reinforcement ratio, effective prestress and prestressed reinforcement configuration. The analytical results provide critical guidance for the design of the SCLC system.

The active structural control system is the most effective means for suppressing structural vibrations in civil engineering structures, but the traditional active control faces practical challenges due to time delay, noise, and other variable operating conditions. A data-driven approach provides a feasible solution. This paper presents three soft actor-critic active control strategies, including a parameter real-time regulator, an independent data-driven controller, and a compensator, that can modify control signals by utilizing real-time feedback from the environment. The parameter regulator is capable of adjusting the controller parameters in response to feedback from the external environment. The independent soft actor-critic reinforcement learning controller completely replaces the conventional controller. Additionally, the compensator integrates reinforcement learning strategy with traditional control methods to dynamically adjust the control signal in real time. The constructed environment was augmented with random time delay and noise to simulate the intricate control environment, and seven criteria were selected to validate the control effectiveness. The results show that the three proposed active control strategies can adaptively adjust the optimal policy control in response to environmental changes, and maintain efficient control performance, while the LQG control algorithm loses control effectiveness under arbitrary time delay and noise. Specifically, the controlled effects of strategies A, B, and C on peak inter-drift reach 64.92 %, 77.28 %, and 84.50 %, respectively, while the peak acceleration reductions are 50.21 %, 43.36 %, and 63.45 %, respectively. Among three proposed SAC-based control strategies, strategy C demonstrates superior performance in mitigating structural response under the specified conditions of time delay and noise. From the statistical data of the evaluation criteria, strategy A demonstrates a superior control effect on acceleration response but is less effective in controlling displacement response when compared to strategy B.

Chloride-induced non-uniform corrosion of reinforced concrete is a major factor in structural durability degradation. This study establishes a connection between the microscopic chloride ion permeation process and the macroscopic corrosion degradation of reinforcing steel through 2D diffusion and convective transport mechanisms. A time-dependent bi-objective failure probability assessment framework integrating seismic fragility and resilience is established. A novel pitting corrosion and uniform corrosion (PCUC) coupling model is proposed, accounting for the evolution corrosion rate after cracking. Furthermore, an enhanced functionality loss model is presented for corroded structures. Utilizing these models, differences in time-dependent seismic fragility and resilience between structures under uniform corrosion (UC) and PCUC are compared. Structural resilience degradation and safety redundancy are quantified, and time-dependent iso-risk seismic hazard curves for both corrosion scenarios are contrasted. The results indicate that, as service time increases, the bi-objective failure probability of the PCUC structure is approximately 5% to 20% higher than that of the UC structure. The seismic fragility of the PCUC structure is 5.28% higher, and its seismic resilience is 13.21% lower than those of the UC structure. The findings highlight that the uniform corrosion assumption tends to overestimate the seismic performance of the structure.

Inerter-based dampers have attracted significant attention as efficient passive vibration control devices. However, their design optimization in multi-degree-of-freedom (MDOF) structures remains insufficiently explored, particularly regarding energy dissipation and placement. This study proposes a power flow-based optimization framework for the design and placement of tuned viscous mass dampers (TVMDs). A closed-form solution for the optimal parameters was derived by minimizing the structural energy dissipation power. Comparative analyses with viscous dampers and the classical fixed-point method demonstrated that the proposed approach achieved superior control performance with smaller damping coefficients. Time-history simulations of MDOF structures further verify that installing TVMDs at stories with maximum inter-story drift can achieve enhanced control effectiveness. Compared with non-optimal placements, this strategy yields control performance gains of up to 20.0

This study focuses on investigating the effectiveness of the novel slope-type sliding pendulum isolation (SSPI) system for seismic protection of electrical equipment through shaking table tests. A 1/5-scaled converter valve and two SSPIs with different sloping angles are designed and manufactured. The hysteretic behavior and the performance parameters of the SSPI are experimentally investigated. The seismic responses of the SSPI systems obtained from the shaking table tests are investigated and compared. Results reveal that the SSPI effectively mitigates the seismic response of electrical equipment. Compared with the non-isolated model, the seismic responses of the SSPI system exhibit reduced sensitivity to both seismic intensity and long-period ground motions. The SSPI demonstrates reliable re-centering performance, with residual displacement lower than 1.0 mm. Finally, a numerical model of the SSPI system is developed and validated using experimental results.

This study focuses on evaluating the seismic performance of bridge isolated by damping enhanced-friction pendulum system (DE-FPs) considering variable friction effect. The hysteretic model of the DE-FP considering variable friction effect is derived. The experimental tests of four DE-FP specimens are conducted and four types of variable friction stiffness are achieved. The numerical model of the DE-FP is created in OpenSees platform and validated through testing values. The seismic performance of a three-span bridge with DE-FPs under near-fault ground motions is investigated and evaluated by case study. Results show that the numerical model can effectively track the hysteretic responses of the DE-FP. The variable friction design effectively enhances the damping capability of the DE-FPs for seismic resilience of bridges. The peak girder displacement is significantly mitigated due to the increased damping capability, the maximum reductions range from 36.8% to 50.2%. However, the permanent bearing displacement and the base forces of the piers increase. Note that the peak drift of the piers is smaller than the equivalent yield drift. The advantage of the DE-FPs is to effectively mitigate the peak girder displacement, which is essential to prevent girder unseating. The achievements provide reliable guidance for variable friction design of the DE-FPs for bridges.

To enhance the seismic performance of the inter-story isolated structure, a shape memory alloy variable curvature friction pendulum (SMA-VCFP) system is developed. Additionally, a parametric optimal design method for SMA-VCFP system is also proposed based on the non-dominated sorting genetic algorithm II (NSGA-II). The hysteretic model of the SMA-VCFP is formulated by incorporating the models of the SMA device and variable curvature friction pendulum bearings (VCFP). The equations of motion for the SMA-VCFP inter-story isolated structure were derived. A numerical model of the SMA-VCFP inter-story isolated system is established on the OpenSees platform, leveraging the validated numerical models of the SMA device and VCFP. The parameters of the SMA-VCFP system were refined and optimized using the proposed design methodology. Several case studies were conducted to demonstrate the effectiveness of the SMA-VCFP system and parametric optimal design method. The results indicate that the re-centering performance of the inter-story isolated structure using the SMA-VCFP system significantly improves with the optimal parameters compared to the VCFP. Specially, the SMA-VCFP system effectively reduces the residual displacement and peak displacement of bearings, and also suppresses the increments of the base shear in the sub-columns, peak acceleration at the top of the structure and inter-story displacement. Notably, the peak displacement of bearings is reduced by 14.8 %, the residual displacement is decreased by 57 %, and both the base shear of columns and peak acceleration at the roof of the structure are suppressed to be within 10 %. The technical achievements contribute to reliable SMA-VCFP seismic isolation design, resilience enhancement and seismic applications.

Existing structural control devices such as nonlinear energy sinks (NESs) and tuned mass damper (TMD) have demonstrated outstanding control efficacy. However, they are susceptible to changes in structural frequency or excitation and require a large sliding mass or space for implementation, which impedes their widespread practical application. In this study, an asymmetric inerter (ASYMI) NES is proposed to improve control efficiency and reduce the mass of dampers. Based on a 1:8-scale two-story aluminum alloy frame structure constructed in a laboratory, the ASYMI NES is optimized and experimentally investigated. The structural parameters are identified by conducting free-vibration testing on the uncontrolled structure. In the laboratory, an adjustable inerter is designed using a gear-rack mechanism. Two seismic records are applied to the experimental structure in shaking table experiments to analyze the control effect of the ASYMI NES device under ground motion. The experimental results for a structure equipped with the control device is compared to the simulated data to analyze the reasons for the discrepancies between experimental and simulation results. The results revealed that the RMS and peak errors between the simulation and experimental results were both within 10%. Overall, the ASYMI NES demonstrated impressive control capabilities, even if the stiffness does not reach the optimal design stiffness during our experiments, indicating its strong potential for engineering applications.

Small modular reactors (SMRs) are increasingly used as a convenient and clean energy source. Seismic isolation is adopted to enhance their resistance. However, the lightweight nature of SMRs poses distinct challenges for isolation. Under seismic excitation, cyclic plastic deformation in Lead-core rubber bearing (LRB) lead cores generates self-heating, causing pronounced temperature-dependent degradation of yield strength and post-yield stiffness, which strongly affects accurate seismic performance assessment. In this study, LRBs are designed for SMRs and relevant experiments are conducted. An improved thermo-mechanical model is developed, explicitly accounting for degradation of yield force and post-yield stiffness due to heat generation and dissipation from the lead core under cyclic loading. The model is validated against experiment, showing good agreement in hysteretic response. Building on the refined model, a parametric analysis investigates the evolution of equivalent linear parameters. Furthermore, a particle swarm optimization (PSO)-based equivalent linearization framework is developed to identify optimal equivalent parameters for a given seismically isolated SMR structure, accounting for cyclic thermomechanical degradation and minimizing linearization error. Compared with conventional equivalent‑linearization approaches, the proposed PSO‑based framework quantitatively captures the multi‑factor influence of cyclic thermo‑mechanical effects on the isolation system and significantly improves the prediction accuracy of global structural response. Case study indicates that under recorded earthquakes, the influence of cyclic dependence on the response of seismically isolated structures is inherently multifactorial and is difficult to be represented reliably by a single scalar indicator. Compared with nonlinear response history analysis, the framework yields mean peak responses with errors of 1.33% for isolation interface displacement and 1.47% for base shear.

To enhance the efficiency and reliability of seismically isolated bearings design, this study proposes a design method based on an enhanced physics-informed neural network (EPINN), aiming to expedite the design of seismic isolators. According to the mechanical model for shape memory alloy variable friction pendulum bearing (SMA-VFP), a systematic mechanical performance design method is developed. Subsequently, a hyperparameter optimization method for the neural network in seismically isolated design is presented grounded in Bayesian theory. A comprehensive dataset encompassing a wide range of characteristics of SMA-VFP bearings is generated using the proposed mechanical performance design method, serving as a case study to evaluate the effectiveness and reliability of the proposed model. The case study results demonstrate that the EPINN model outperforms both traditional physics-informed neural network models and standard artificial neural networks. Specifically, it achieves a maximum reduction of 35.1% in pure data loss, 78.9% in physics-informed loss, and 91.5% in physical constraint loss, with the reliability of the bearing design reaches up to 96.6%. By incorporating a design-oriented physics-constrained loss term, EPINN effectively reconciles physical principles with design objectives even when data-driven loss computation is unavailable, offering a distinct advantage over conventional ANN and PINN approaches that lack such constraints.

The mechanism governing the frictional coefficient dependency on structural seismic responses remains insufficiently understood. To bridge this knowledge gap, an innovative experimental methodology employing specialized thermocouples was developed to enable direct temperature measurements at friction interfaces, allowing submillisecond thermal monitoring with high precision. Subsequent analyses have systematically quantified the effects of slip velocity, vertical pressure, temperature, and surface roughness on frictional behavior. Empirical relationships between frictional coefficients and governing parameters—normal stress, slip velocity, interfacial roughness, and transient temperature—are rigorously derived through controlled-variable parametric studies. A refined friction model (RFM) incorporating dynamic thermo-mechanical coupling effects was developed. This model integrates a computational framework for real-time frictional heat generation, hybridizing Newton's cooling law with lateral radiative dissipation. This proposed methodology improves temperature predictions by more than 50% compared to traditional methods, successfully addressing the energy loss issues seen in conventional Coulomb models. The derived thermomechanical coupling framework establishes a theoretical foundation for optimizing friction interfaces in energy dissipation devices and presents a paradigm-shifting methodology for performance prediction in friction pendulum systems under operational dynamic excitations.