The analysis methods for mechanical behavior and safety in engineering still fail to meet real demands. Simply applying available artificial intelligence (AI) methods cannot fundamentally address the strict requirements on the stability and reliability of output in engineering. To tackle this issue, by simulating the thinking and decision-making process of human experts, the ‘mechanism’, represented by mechanical analysis methods, and the ‘data’, obtained after multi-source information assimilation, are integrated in real time. Centering around the mechanical models of engineering, three main methods for AI of Engineering are established, namely the multi-source data assimilation and data quality evaluation method, the mechanism-data coupling-driven AI method, and the cross-engineering synergistic analysis method. These methods are progressively implemented into the framework of AI of Engineering, forming a new generation engineering intelligent agent, and achieving a qualitative change from ‘one-way AI for engineering’ to ‘integrated AI of engineering’. AI systems developed therefrom are applied to landslide dams, slopes, and wind turbine generators to predict the performance. The applications indicate that AI of engineering is not constrained by the limited quantity, unstable quality and weak correlation of multi-source data in practice. It also addresses the limitations of mechanical methods under complex conditions and the difficulties in accurately obtaining computation parameters. AI of engineering integrates multiple functions, such as deformation source tracing, mechanical behavior prediction, risk early-warning and risk regulation, providing solid supports for engineering projects.

In recent years, viscoelastic damping technology has been introduced into the vibration control for seismic performance enhancement of civil structures. In this study, a high-damping viscoelastic dampers (HDVED) is developed to solve the critical issues of VED for a middle-high temperature environments serving in civil structures and for low-frequency damping control under seismic action. To reveal the influence rule of frequency, temperature and displacement amplitude on the mechanical properties and seismic energy dissipation performance of HDVED, the dynamic mechanical properties of HDVED are experimentally investigated at different frequencies, temperatures and displacement amplitudes. The results show that HDVEDs exhibit excellent vibration damping and energy dissipation capabilities and environmental adaptability under middle-high temperature range and low frequency loading. Their dynamic mechanical properties and energy dissipation capabilities show strong dependence on frequency, temperature and displacement amplitude with obvious coupling effects. On this basis, a high-order fractional derivative model with equivalent internal variables of HDVED is proposed, and the validity and accuracy of the model are verified by tests. The results show that the proposed high-order fractional derivative model with equivalent internal variables can comprehensively describe the effects of excitation frequency, ambient temperature and displacement amplitude on the dynamic mechanical properties of HDVED. The model can accurately predict the mechanical properties of HDVED at different frequencies, temperatures and displacement amplitudes. It can provide an important basis for dynamic response analysis and design of viscoelastic damping structures.

A structure is treated as a multi-degree-of-freedom system, and each bouncing person on the structure is treated as an individual spring-mass-damper system. The influence coefficient and mode decomposition methods are used to derive the equation of motion for the coupled system in canonical coordinates. Furthermore, the frequency response function linking the state vector of coupled system with the crowd bouncing load time history is established by the state space method. The spectral analysis solution for the equation of motion of coupled system is derived by the stochastic vibration theory, and then it is used to calculate the auto-power spectrum and root mean square value of structural acceleration response. A comparison between theoretical predictions and field-measured responses of large-span floors demonstrates that the proposed spectral analysis model and the computational method are rational, and they can be applied to analyzing and caleulating structural responses during crowd bouncing activities.

The prefabricated components in a monolithically precast subway station aremainly connected by the cast-in-situ concrete and splice sleeve. The design objective is to make sure that the structural performance is equivalent to that of a cast-in-place structure. But the structure is different from a cast-in-place structure, so the above mentioned objective should be investigated. Based on a real subway engineering project at Beijing, 1-g shaking table model tests of the site-monolithically precast subway structure were performed to study the seismic structural displacement and deformation responses, including the horizontal, vertical absolute and relative displacement, and angular displacement, etc. Results show that the structural horizontal responses were dominant by the responses of the surrounding soil. But the vertical responses were not sufficiently restrained. The angular displacement were observed to be within 0.03 degree. The story drifts of the basement-1 andbasement-2 were large with unrecoverable deformation smaller than 0. 2 mm in the both horizontal and vertical directions. However, almost no structural damage was observed after the experiments, indicating that the prefabricated components have relative displacement at the joint regions.

Suspension bridges with closely spaced parallel dual main cables may experience significant aerodynamic interference effects, and particularly are prone to significant wind induced vibration due to the low frequency and rapidly varying cross-section shape of the main cable in the construction phases without the constraint of slings.In order to investigate the vibration characteristics of the parallel dual main cables in construction phases, five typical segmental models of dual main cables in the parallel configuration were designed for a very long-span suspension bridge with a main span of 2180 m. Based on the wind tunnel tests, the wind-induced vibrations of the dual main cables on one side in construction phases were studied and then compared to those of the single main cable on one side with the same parameters. The results show that the wind-induced vibrations may occur for the five segmental models of dual main cables with the maximum standard deviation of the amplitude up to 8.7 H. Evident differences in the vibration characteristics of the dual main cable can be observed under different wind direction angles and wind attack angles. The possible mechanisms of wind-induced instability of the dual main cables in construction phases are summarized as follows: the coupling mechanism of unstable main cable shape and aerodynamic interference between cables, the single mechanism of aerodynamic interference between cables, and the single mechanism of unstable main cable shape. At the early stage of the erection of the dual main cables, the wind-induced vibration of the cables is the most problematic, but the maximum amplitude of the dual main cables caused by instability decreases with the erection process pf the cables.

Taking three-layer composite beams as the research object, the finite element analysis method considering the geometric nonlinearity of double interface slip is studied. Firstly, the displacement governing differential equation of the three-layer composite beam is established based on the basic equations of elasticity, namely equilibrium, geometry and physics equations. The geometric linear element stiffness matrix with “precise” characteristics in local coordinate system (co-rotation coordinate system) is derived by direct stiffness method. Secondly, the transformation matrix between the structural coordinate system and the local coordinate system is derived by the co-rotation coordinate method of geometric nonlinear analysis, and the geometric nonlinear tangential stiffness matrix of the three-layer composite beam element under the structural coordinate system is established, and the corresponding equivalent node force matrix can be obtained simultaneously. Finally, the rigid arm and force equivalence principle are used to consider the eccentricity effect caused by the inconsistency between the point of concentrated force applied to the three-layer composite beam & the boundary constraint point and the node of the finite element model. Based on the above results, a finite element analysis method for three-layer composite beams considering geometric nonlinearity of double-interface slip is established, and the corresponding program is developed. Three typical examples of three-layer composite beams are analyzed, and the correctness and reliability of the proposed method and the developed program are verified by comparing with the calculation and test results of existing literatures.

An optimization algorithm for ultrasonic tomography of foundation pile based on three-dimensional space path tracking is addressed in this paper. This algorithm can overcome the limitations of 3D tomography technology and traditional sonic logging techniques for piles testing and the shortcomings. By optimizing the distribution of spatial field source points, the Dijkstra algorithm is used to accurately track the elastic wave propagation path in three-dimensional space. And combined with the improved SIRT algorithm, the visualization of the internal structure of concrete foundation piles and the rapid location and quantitative assessment of potential defects are achieved. Subsequently, numerical simulation verification was carried out through the finite element analysis software ABAQUS. The impact of key technical parameters including transducer arrangement density, field source point encryption coefficient and defect characteristics on the three-dimensional tomography results was analyzed. Finally, the proposed optimization algorithm was applied to a bridge pile foundation on the Qiantang River in Hangzhou City, Zhejiang Province. The results demonstrate that the technology can accurately locate concrete defects and provide references for defect types and severity, offering theoretical and technical support for innovations in sonic logging methods for piles testing.

Polymer-modified bentonite (PMB) has been extensively investigated for industrial waste containment due to its low hydraulic conductivity and salt resistance. Its performance under actual field conditions, however, remains insufficiently documented. Swelling and heavy-metal adsorption behaviors of PMB and sodium bentonite (NaB) were systematically compared under varied contaminant solutions. The hydraulic performance of polymer-modified bentonite-sand (PMB-S) and sodium bentonite-sand (NaB-S) mixtures was evaluated. Furthermore, the feasibility of using in situ phosphate mine waste as a substrate material in PMB-based barriers was discussed. Results show that across the pH range of 2 to 12, PMB always maintains a free swelling index greater than 50 mL/2 g, significantly higher than that of NaB. Additionally, its adsorption performance for Pb2+ and Cu2+ is superior to that of NaB, with the removal rates increasing by 58.42% and 61.31%, respectively, particularly in acidic environments. The hydraulic conductivity of PMB-S is significantly lower than that of NaB-S, and may consistently achieve values below 1×10-8 cm/s in most contaminated environments. The mixtures of PMB with three phosphate mine wastes have comparable hydraulic performance to PMB-S, confirming its feasibility as a barrier material. Furthermore, field demonstrations in alkaline lead-zinc and acidic phosphogypsum tailings validate the robust hydraulic and contaminant barrier properties of PMB-S under actual conditions.

To investigate the effects of alpine freeze-thaw cycles and saline soil environments on the microstructure of fiber-reinforced concrete, a comprehensive durability test was conducted on polyacrylonitrile fiber-reinforced concrete (PANFRC) under combined salt corrosion and freeze-thaw (salt-freezing) conditions. Using both ultrasonic plane testing and opposite-side measurement methods, the thickness of the damaged layer and the development of internal defects in PANFRC were evaluated. Furthermore, the evolution of the pore structure was analyzed using NMR technology and pore fractal theory. Results indicate that under the combined action of salt-freezing cycles, the thickness of the damaged layer increases progressively, and its compressive strength decreases, leading to a corresponding reduction in the overall compressive strength of the concrete. A linear relationship was observed between these two parameters. The ultrasonic velocity in the central region of the PANFRC specimens decreased sharply, and the area of the defect zone expanded significantly, making this region the most vulnerable part under load and prone to fracture. When the PANF content exceeds 1.5 kg/m³, the loss in ultrasonic velocity is minimized, and the fibers effectively inhibit the growth of internal defects, thereby enhancing the toughness of the concrete, particularly during the middle to late stages of salt-freezing cycles. As the number of salt-freezing cycles increases, the proportion of micropores in PANFRC decreases, while the number of macropores and microcracks increases. The total proportion of small and medium-sized pores also rises. Overall, the development of the pore structure follows a trend of increasing complexity, with small pore throats interconnecting and transforming into super-large and fissure-type pore throats.

A test system for the serviceability of highway subgrade considering the principal stress rotation was developed to accurately simulate the evolution of the three-dimensional stress state and assess the serviceability of the subgrade under traffic loads. The subgrade serviceability test system is composed of the core loading system and the auxiliary modules such as full digital servo control system and data acquisition system. The loading system with four dynamic and three static actuators was designed to achieve the rotation of the principal stress axis in subgrade soil elements at various levels based on the coordination of static and dynamic loading. A theory for three-dimensional dynamic stress recurrence in highway subgrade was developed based on the theoretical model of the subgrade’s dynamic response under traffic load and the loading system’s mechanical and spatial characteristics, and furthermore a method of transforming the “stress spectrum” into the “load spectrum” was proposed. Using the test system, the three-dimensional stress state in the highway subgrade under vehicle load was examined. The results show that the actual output load deviation of each static or dynamic servo actuator is less than 1%, confirming that the developed subgrade serviceability test system can accurately simulate the three-dimensional stress state of the subgrade under traffic loads. The time history curves of each dynamic stress component and the attenuation pattern of vertical dynamic stress closely match the theoretical ones. The vertical dynamic stress in the subgrade gradually decreases with depth, and the stress path experienced by the soil element resembles a “heart shape.” The “seven-cylinder linkage and coordinated loading” method can replicate the dynamic response of soil elements at various depths in the subgrade working area during the “far-near-far” movement of traffic loads, despite the loading device’s position remaining unchanged. Therefore, the subgrade serviceability test system developed in this study introduces a new concept, approach, and technology for investigating the evolution of highway serviceability under actual traffic loads.