As a critical component of an offshore wind turbine foundation, grouted connections are susceptible to fatigue failure under long-term loading conditions, which makes the assessment of their fatigue behaviour essential for the overall structural integrity and maintenance. Compared to costly experimental characterization and computationally expensive numerical modelling, there is a high demand for developing a fast, efficient and accurate surrogate model to predict fatigue life. This paper develops a novel hybrid Physics-Informed Neural Network (PINN) model that integrates both simplified physical constraints and hidden physical laws to predict the fatigue life of axially loaded grouted connections, where the physical knowledge is the relationship between fatigue life and fatigue-related parameters. The results show that the developed hybrid PINN model achieves superior prediction accuracy compared to the current codes of practice, the conventional Deep Neural Network (DNN) model, the PINN model integrating simplified physical constraints (S-PINN), and the PINN model integrating hidden physical laws (H-PINN). To enhance the interpretability of the model, Shapley Additive Explanations (SHAP) analysis and physical consistency analysis are conducted to assess the contribution of each feature to the fatigue life and to investigate the distribution of predictions with respect to physical consistency. It’s found that the new hybrid PINN model produces predictions that exhibit a higher degree of physical consistency than the purely data-driven DNN model, demonstrating the reliability and robustness of the model.
Carbonation curing provides a practical route to reduce the carbon footprint of cement by converting CO2 into stable carbonates while improving early-age performance. However, its efficiency is limited by moisture-controlled diffusion barriers and inconsistent curing protocols. This study couples process optimisation with waste-derived biochar to enhance CO2 uptake and strength development in cement pastes. An L9 orthogonal design using rice husk biochar (RHB) identified water-to-cement ratio, preconditioning duration, and biochar dosage as the governing parameters. The optimal combination-W/C = 0.40, 5% RHB, and 3 h preconditioning-produced substantial improvements in CO2 absorption, pore refinement, microhardness gradients, and both early and 28-day strengths. To evaluate the generality of this approach, two types of additional agricultural waste biochar, straw biochar (SB) and coconut shell biochar (CSB), were incorporated under the optimized curing regime. SB mainly acted as a filler, limiting CO2 transport, whereas CSB introduced a favorable meso-macropore network that enhanced CO2 transport, achieving carbonation degrees above 40% and early-strength gains of approximately 20 MPa at 5% dosage. These findings highlight the value of transforming bio-waste into functional porous additives that enable more efficient CO2 curing, which supports scalable, circular-economy strategies for low-carbon cement production.
The efficacy of carbonation curing in cement-based materials is governed by the pore network, which controls moisture transport and CO2 diffusion. This study investigates how preconditioning, via natural ventilation or oven drying at 60 degrees C, regulates the moisture state of cement composites by tailoring pore structure and influences early-age carbonation. Cement pastes with water-to-cement (w/c) ratios of 0.35 and 0.18 were preconditioned to specific moisture levels prior to carbonation. The coupled mechanisms of moisture distribution, pore evolution, and chemical reaction were elucidated through multi-scale characterization, complemented by pH, mass change, and compressive strength measurements. Preconditioning increased overall porosity and coarsened the pore structure by removing free water and interconnecting micropores to form continuous transport pathways. Subsequent carbonation partially counteracted this coarsening through the precipitation of CaCO3, which refined finer pores and provided nucleation sites for secondary C-S-H formation. Concurrently, decalcification enlarged capillary pores, thereby maintaining the continuity of CO2 transport. The outcomes of preconditioning were found to be strongly dependent on the initial pore structure: in high w/c pastes, extended preconditioning generated a balanced bimodal pore network that facilitated both CO2 ingress and continued hydration, resulting in significant early-age strength gains. In contrast, low w/c pastes reached equilibrium rapidly; prolonged drying induced the formation of irreversible macropores, which enabled high carbonation efficiency but limited strength improvement due to insufficient microstructural densification. These findings underscore the critical need to optimize preconditioning protocols to synchronize the moisture state, pore connectivity, and reaction fronts, thereby maximizing CO2 uptake without compromising mechanical performance.
Quasi-brittle materials, such as concrete or rock, exhibit pronounced anisotropy due to heterogeneous spatial distribution and complex geometric topology of microcracks. This study presents a novel anisotropic damage model capable of capturing both the self-similar and kink growth of microcracks at the microscale. Using homogenisation techniques, the model quantifies the contributions of penny-shaped microcracks in various states to the macroscopic mechanical properties. Furthermore, it emphasises the influence of kinked microcracks, derived from these initial microcracks, on the overall properties of the material. The stress intensity factor at the tip of the kinked microcrack is derived, leading to an enhanced fracture criterion. The model defines a critical limit state arising from the unconstrained growth of kinked microcracks and proposes two optimisation schemes: one for identifying the onset of this limit state and the other for achieving stable numerical results. The model results match well with the uniaxial compression tests. Notably, the limit state of the pure micromechanical model and the microscopic and macroscopic model results are thoroughly examined. These findings underline the model’s ability to provide a detailed description of the microscopic mechanism underlying the anisotropic damage of quasi-brittle materials.
Recycled polyethylene terephthalate (PET) particles have been increasingly incorporated into cementitious composites to improve toughness, yet their poor thermal stability remains a major obstacle for applications under elevated temperatures. Therefore, the thermal behavior of microcapsule-incorporated fiber-reinforced high-volume ground granulated blast furnace slag (GGBFS) cementitious composites containing 0–30% recycled PET after exposure to 25–750 °C was systematically investigated, with particular emphasis on the coupled evolution of PET, the matrix, and their interface across multiple length scales. A multiscale framework combining residual mechanical testing, phase analysis, microstructural observation, pore structure characterization, and density functional theory simulations was employed to elucidate the temperature-dependent interaction between PET and the GGBFS-based matrix. The results show that within 25–300 °C, composites with low PET content retained or slightly improved their residual mechanical properties because thermally stimulated GGBFS hydration promoted the formation of additional calcium aluminosilicate hydrate (C-(A)-S-H) and matrix densification. In contrast, excessive PET introduced more interfacial defects and weakened this compensating effect, while 10% PET delivered the best overall performance. Above 450 °C, PET decomposition coincided with pronounced macropore formation, loss of interfacial continuity, and rapid deterioration of the physical and mechanical properties. Meanwhile, progressive dehydration and destabilization of the cementitious matrix further contributed to structural damage. These findings reveal a temperature-dependent transition from low-temperature matrix compensation to high-temperature polymer/matrix interface deterioration, and provide mechanistic guidance for designing thermally resilient GGBFS-based cementitious composites incorporating PET.
Accelerated carbonation curing (ACC) has emerged as a promising technology for enhancing CO2 sequestration and early-age performance of cement-based materials. However, the role of porous additives in balancing improved gas transport with the potential reduction in early-age compressive strength due to binder dilution remains insufficiently understood. This study systematically investigates the comparative effects of expanded perlite (EP) and coral sand (CS), with varying particle sizes and dosages, on the carbonation kinetics and early-age mechanical performance of cement composites. The results demonstrate that incorporating 5% CS significantly enhances carbonation efficiency, achieving a 46% increase in mass gain compared to the control, whereas EP exhibits a more moderate improvement of 27%. Thermogravimetric (TG) and microstructural analyses reveal that the interconnected macroporous network and intrinsic chemical composition of CS not only facilitate CO2 transport and provide abundant nucleation sites but also promote the formation of highly crystalline calcite (CaCO3), leading to improved early-age compressive strength. In contrast, although EP enhances carbonation reactions, its fragile honeycomb structure and weak interfacial bonding result in a pronounced reduction in early-age mechanical performance. Overall, a 5% dosage of coarse CS (2 mm particle size) achieves the optimal balance between carbonation efficiency and mechanical integrity, maximizing CO2 uptake and microstructural densification. These findings highlight the potential of mineralbased porous additives for developing low-carbon, high-performance cementitious materials.
Distributed fiber-optic sensors (DFOSs), which are based on optical frequency-domain reflectometry (OFDR), provide high-resolution strain measurements and have promising application in structural health monitoring. This study introduces a novel steel fibe-reinforced polymer composite bar (SFCB) with self-sensing, structural reinforcement, and damage control features designed to evaluate the response and damage status of concrete members. Investigating the force transfer mechanism between SFCB and concrete is essential for understanding concrete cracking behavior and establishing a reliable damage evaluation approach. Initially, tension tests were conducted on SFCB concrete members to investigate the effects of cover depth and bonding mechanism (concrete type and surface treatment of the SFCB) on the end effects, along with a test procedure designed to effectively eliminate the end effects. The results indicate that the use of members with small cover depths, surface sandblasted SFCB, and geopolymer concrete (GPC) can reduce the impact of the end effects. The tracking and quantification of particular crack progressions were subsequently assessed through the integration of a digital image correlation (DIC) system and a DFOS system. Finally, based on the results from which the influence of end effects has been eliminated, a theoretical model for the response of SFCB concrete tension members was proposed, along with a model for damage variables that is independent of geometry and material behaviors.
Accelerated carbonation treatment is recognized as an effective method for enhancing recycled aggregates (RA), but its potential in structural concrete, particularly with respect to seismic performance, remains underexplored. To address this gap, this study is the first to integrate mesoscale modeling with structural finite element analysis (FEA) to systematically investigate the seismic behavior of carbonated recycled aggregate concrete (CRAC) shear walls under dynamic loading. At the material scale, uniaxial compression tests on CRAC cylindrical specimens with varying replacement ratios were conducted to evaluate their stress-strain behavior and mechanical properties. A mesoscale model of CRAC was developed using a random aggregate placement method, and FEA was employed to extend the analysis of replacement ratios. At the structural scale, a CRAC shear wall FEA model was established, incorporating the material-level stress-strain relationships into cyclic lateral loading simulations. Parametric analysis revealed that increasing both the axial load ratio and the replacement ratio significantly reduced the seismic performance of CRAC shear walls, with a maximum reduction of 21.7%. Based on these findings, recommended ranges forRA replacement ratios and axial load ratios are proposed, providing practical guidance for the structural application of CRAC.
In response to the scarcity of freshwater resources and the high environmental impact of cement production, this research explores the use of seawater to prepare alkali-activated cementitious materials (AAMs). To avoid the adverse effects of high concentrations of Cl-in seawater on performance, the feasibility of synergistic application of hydroxysodalite and seawater in AAMs systems was proposed. By setting the hydroxysodalite content and mixing water type, the effects on the early hydration behavior, Cl-binding capacity, and microstructural characteristics of AAMs were systematically investigated. The results show that hydroxysodalite improves the initial flowability (8 %-14 %) while shortening the setting time (initial setting and final setting are shortened by 32 %-40 % and 26 %-33 %, respectively). Compared with the deionized water mixing system without hydroxysodalite, hydroxysodalite increased the compressive strength of the seawater mixing system by 17.2 %. Seawater causes OH-in hydroxysodalite to migrate out of the beta cage, and induces the formation of chlorine-bearing sodalite. At the same time, hydroxysodalite promotes the formation of more C-A-S-H gel, which not only reduces the total porosity but also increases the gel pore ratio. The synergistic effect of these two aspects significantly enhances the chloride ion binding capacity of AAMs. In addition, the binding behavior of AAMs towards Cl-follows the Freundlich model. Overall, the introduction of hydroxysodalite provides an effective strategy for significantly improving the early performance and Cl-binding capacity of AAMs mixed with seawater.
Carbon fibers have excellent properties, including high strength, light weight, corrosion resistance, and high durability; therefore, they are widely used in various fields. Carbon fibers possess excellent electrical conductivity and electrochemical stability, and they can be used as electrode materials for functionalized applications in civil engineering. This study explores the evolution mechanism of the electrochemical properties of carbon fibers and carbon fiber composites used as anodes. This study further focuses on the collaborative intervention technique of impressed current cathodic protection and structural strengthening (ICCP-SS) for reinforced concrete structures, as well as the non-destructive recycling of carbon fibers based on their electrochemical evolution mechanism. This study aims to provide new ideas for the functionalization of carbon fiber composites in civil engineering.
Early-age carbonation curing in cement pastes can simultaneously enhance early strength and sequester CO2, however, its long-term performance largely depends on whether preconditioning establishes a moisture state that promotes carbonation without limiting post-carbonation hydration excessively. This issue is particularly important for semi-dry compacted cement systems, where limited pore water and discontinuous capillary networks may alter the balance between carbonation and later hydration. This study investigated the effects of sequential preconditioning and carbonation on the post-carbonation hydration behavior of cement pastes with two representative water-to-cement (w/c) ratios, w/c = 0.35 for conventional casting and w/c = 0.18 for semi-dry compaction. The results show that moderate preconditioning in the w/c = 0.35 system (14 h ventilation) promoted more relatively homogeneous carbonation and finer CaCO3 precipitation, leading to early densification and strength enhancement while preserving sufficient transport accessibility for later hydration and pore refinement. In contrast, the semi-dry w/c = 0.18 system (2 h ventilation) showed mainly localized carbonation, retained a coarse pore network, and formed a dense carbonated structure that strongly limited post-carbonation hydration. Overall, the performance of carbonation-cured materials was governed by the balance between moisture availability, CO2 transport, and residual hydration capacity, rather than by carbonation extent alone.
Current research on cement primarily focuses on macroscopic properties, yet microscopic aspects are equally crucial for understanding its complex behavior. To investigate cement at the mesoscale, a coarse-grained force field for cement materials is developed. In this study, a bottom-up approach was adopted to develop the coarse-grained field, where the structure and mechanical properties of the cement system were simulated using molecular dynamics (MD) and then reproduced in the coarse-grained model. First, the molecular model of cement is established, and the radial distribution function and Young's modulus of cement are obtained by MD simulation. The model is then coarse-grained into mesoscopic representations, and the force field parameters (sigma, epsilon, cutoff distance) of the Lennard-Jones potential (12-6) are determined through iterative coarse-grained dynamics and uniaxial tensile simulations. These parameters are applied to mesoscopic models with varying crack widths and filling rates, and stress-strain curves are obtained through uniaxial tensile tests. Results indicate that increased crack width reduces model strength, while higher particle filling rates enhance tensile strength. Mean-square displacement and radial distribution function analyses reveal that stronger interparticle interactions contribute to enhanced mechanical performance. Additionally, creep protocols elucidate the material's mechanical evolution pathway, capturing the progression from stable elastic response and time-dependent delayed yielding to rapid plastic densification under increasing external loads. This work establishes a robust and practical coarse-grained force field for cement using a straightforward methodology, laying the groundwork for future refinements and applications in mesoscale cement research.
Improving the accurate determination of the fracture energies in cementitious composites is highly significant for engineering practice. This research employed X-ray computed tomography (XCT) and digital volume correlation (DVC) to investigate fracture behavior in fiber-cement composites. In-situ XCT loading experiments were conducted for tracking failure progression, facilitating three-dimensional volume image reconstruction. The internal volumetric strains of the samples were calculated by DVC based on the volume images before and after deformation, enabling computation of deformation energies outside the fracture domains. The fracture energies of the samples were obtained after taking the deformation energy into account. The outcomes showed that the fracture energy will be overestimated by about 10 % (the value will increase for fiber reinforced materials) when the deformation energy is not considered.
Calcium sulfoaluminate (CSA) cement is widely blended with Portland cement (PC) to enhance crack-healing capabilities; however, the mechanisms underlying CSA systems' healing capacity remain insufficiently understood. Investigating the factors influencing crack healing in CSA cement is essential to developing more durable, self-healing cement-based materials. This study investigated CSA clinker with varying sulfate-to-ye'elimite molar ratios (M value) of 0.1, 1.0, and 2.0, and explored their effects on crack-healing capabilities. Mortar prisms were pre-cracked at 1, 7, and 28 days (target crack width approximate to 250 mu m) and then exposed to deionized water for healing. Remarkably, CSA clinkers demonstrated rapid recovery, achieving full crack closure within three days under optimal conditions. The healing process was driven by the hydration of ye'elimite, which induced ettringite supersaturation and facilitated ettringite precipitation, filling cracks. Ettringite preferentially nucleated and grew in the 'open space' around cracks, and its precipitation kinetics were shown to be governed primarily by the supersaturation degree (beta), which was effectively regulated by pore-solution pH; higher pH markedly accelerated precipitation and crack healing. Minor formations of amorphous AH3 and crystalline CaCO3 from natural carbonation of ettringite were identified through XRD, TG, and 27Al NMR analysis. Furthermore, variations in sulfate content and cracking age influenced the hydration of CSA cement, modulating pore-solution chemistry and thereby determining both the crack-closure ratio and the kinetics of ettringite formation around cracks. This work establishes a framework for pore solution, supersaturation, and precipitation in autogenous crack healing in CSA systems, providing fundamental knowledge to develop durable, self-healing cementitious materials.
This study presents a novel self-sensing steel fiber-reinforced polymer composite bar (SFCB). The SFCB combines damage control, self-sensing, and structural reinforcement functions using distributed fiber optic sensing (DFOS) technology. By combining DFOS strains with theoretical and numerical models, a multilevel performance method for damage assessment is proposed from the perspectives of safety, suitability, and durability. Stiffness is a metric used to assess the complete service history of the reinforced concrete (RC) structure, which was used to define the damage variables. Initially, a basic correlation is created between the SFCB strain and several performance characteristics, such as moment, curvature, load, deflection, stiffness, and crack breadth, at characteristic points. The threshold values of damage variables for safety, serviceability, and durability were determined based on loading peak, mid-span deflection limits, and crack width limits corresponding to the damage variables. Then, a modified fiber damage model based on DFOS strain data is proposed to improve identification, quantification, and tracking for fiber damage. Finally, the reliability of the proposed theoretical and numerical models was verified by three-point flexural tests of SFCB-RC beams, and the test beams were analyzed using the proposed method. The results show that increasing the reinforcement ratio can lower the threshold at all levels and improve the ability of the flexural beams to control damage. This study contributes to advancing the intelligence of RC structures and offers valuable insights for the design of intelligent RC structures. (c) 2024 THE AUTHORS. Published by Elsevier LTD on behalf of Chinese Academy of Engineering and Higher Education Press Limited Company. Thi s is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
While existing methods offer guidance for designing concrete structures against impact loads, however, under varying impact energy levels, the damage state of structures poses significant challenges for quantification and assessment, especially when no clearly visible cracks or spalling are present on the structural surface. This study proposes and validates a novel self-sensing steel-FRP composite bar (SFCB)-RC beam that integrates impact resistance with damage monitoring. Through combined drop-weight impact tests, post-impact residual static tests, and high-fidelity finite element simulations, we demonstrate that SFCB reinforcement significantly enhances structural recoverability and damage control compared to conventional steel rebar, attributable to its stable post-yield stiffness. The self-sensing capability of SFCBs successfully correlates distributed residual strain patterns with structural resilience, enabling effective post-impact assessment. A comprehensive parametric analysis identifies a post-yield stiffness ratio of 0.07 as a critical threshold for optimizing resilience. Furthermore, a quantitative predictive model is established, explicitly linking the restitution coefficient to the equivalent reinforcement ratio and post-yield stiffness ratio, thus providing a vital tool for the performance-based design of impact-resilient structures. Finally, a performance-based design framework is established to guide the development of impact-resilient SFCB-RC beams with built-in health monitoring functions.
To address the crack healing and durability enhancement of concrete, a novel and pragmatic vascular self-healing approach has been developed, and a quantitative characterization of vascular healing efficiency has been presented. A biocompatible Pickering emulsion containing no surfactant or other organic modifiers, consisting of natural montmorillonite co-stabilized with carboxymethyl cellulose on methyl palmitate, is developed as a 3D printing ink to construct hollow pipelines in cement in situ, which is characterized by means of Fourier transform infrared spectroscopy (FT-IR), zeta potential, confocal laser scanning microscopy (CLSM), scanning electron microscopy (SEM) and rheological property measurement, showing that it can be long-term stored stably, 3D printed smoothly, self-supporting steadily in cement and self-dissolving forming hollow pipeline following cement curing. The manufacture of pipelines in various diameters is facilitated by variable control settings in 3D printing. X-ray computed tomography is used to characterize the internal pipelines, revealing excellent connectivity and coherence in their architecture. Bending and compression tests indicate that larger pipeline diameters exhibit better strength recovery at lower pumping pressures. Watertightness tests demonstrate the pipeline's remarkable crack-healing capability across cement specimens of varying widths, achieving nearly complete seals for cracks 100 mu m and 200 mu m wide. This is attributable to the pipeline's outstanding transport performance. SEM and X-CT analyses confirm excellent adhesion of the polyurethane. Vascular self-healing in cementitious materials with Pickering emulsion-based 3D printing provides outstanding crack healing capability and versatile construction freedom.