This study provides a comprehensive analysis of ultrahigh-performance concrete (UHPC) compressive strength, focusing on data augmentation, predictive modeling, and model interpretability. The research utilized 808 experimental data points with 16 features to form the original UHPC data set. An optimized Gaussian copula generative adversarial network (Opt-CopulaGAN), enhanced via Bayesian hyperparameter tuning, was employed to augment the data set. Postaugmentation, the inception-fully connected network (Incep-FC Net) and machine learning models were used for prediction, resulting in a significant R-2 value increase from 0.9035 to 0.9545. Partial dependence plots (PDP) and Shapley additive explanations (SHAP) analysis were applied to understand model behavior, revealing that high silica fume and fly ash concentrations improved UHPC strength, while nano-silica had a minor impact. SHAP also highlighted the negative effects of low-dosage silica powder and fine aggregates during extended curing. Elevated temperatures and humidity positively influenced UHPC compressive strength, especially during the initial curing phase.

This study explores vertical insertion of steel rebars into 3D printed concrete (3DPC) using a robotic arm, enabling reinforcement during the printing process. A robotic system comprising a mobile base, a 6-axis robotic arm, a force/torque sensor, and a gripper was employed, alongside a dedicated algorithm for self-calibrating surface detection and force-controlled insertion. To enhance bond performance, grouting with epoxy or highflowability cementitious materials was investigated. A total of 19 specimens were fabricated for the pull-out test, while 3 specimens were reserved for visual inspections. Pull-out tests identified four failure modes: concrete splitting, rebar pull-out with or without splitting, and rebar fracture. Steel cable confinement prevented splitting failure and increased bond strengths. Grouting improved bond strength up to 24-42 %, whereas 0.5 vol % PVA fibers in the printed matrix reduced bond performance by 11.55 %. Confined specimens with 100 mm bond lengths and grouting exhibited rebar fracture without concrete damage, indicating sufficient bond capacity for structural applications. Visual inspections of specimens indicated that non-grouted specimens exhibited minor voids in the upper region, while grouting effectively filled voids along the insertion path. Existing bond-slip models were evaluated for robotically inserted rebars, showing reasonable predictions for splitting failure but limited accuracy for pull-out failure. An analytical model was developed based on thick-walled cylinder and fictitious crack models. Overall, robotic rebar insertion (combined with grouting) achieved an acceptable bond performance by observing rebar fracture during the pull-out test, demonstrating the feasibility and potential of this automated reinforcement strategy in 3DPC.

The study evaluates the effect of fiberglass mesh, structural adhesive on the mechanical behavior and microstructure of 3D printing concrete (3DPC). The research investigated the interplay between fiberglass mesh strip thickness and structural adhesive application thickness in phase I. Phase II further determined ideal pairings of fiberglass grid density and polypropylene (PP) fiber dosage based on Phase I. The experiment incorporated mechanical evaluations (compressive, flexural, and oblique shear strength testing) and microscopic performance characterization through scanning electron microscopy (SEM) imaging and digital image correlation (DIC) analyses. The 0.3 mm fiberglass mesh and 3 mm structural adhesive configuration (M0.3-A3) is determined as the optimal reinforcing solution to achieve peak compressive (52.15 MPa), flexural (11.04 MPa), and shear strengths (17.36 MPa) in Phase I. Additionally, based on the optimized solution in Phase I, the IN4-P3 composite (4 mm grid interval, 3 vol% PP fibers) demonstrated better microstructure and mechanical strength with 59.86 MPa compressive strength, 14.18 MPa flexural strength and 25.9 MPa shearing strength. Specifically, PP fibers concentrations exceeding 6 vol% impede strength enhancement due to excessive agglomeration while tightest mesh grid interval attained enhanced mechanical performance via a bridging effect. Consequently, tighter 4 mm grid interval combined with moderate 3-6 vol% PP fiber dosage effectively enhances mechanical and microstructural performance. Fiberglass mesh and PP fibers-modified structural adhesive exhibit a synergistic effect that improves mechanical properties, enhancing the construction applicability of 3DPC.

Pre-excavation dewatering design for deep excavations still relies on uniform well layouts, engineering experience, and repeated 3D finite-element trial-and-error analyses, hindering rapid and systematic multi-objective optimization. This study examines whether a physics-consistent surrogate-assisted workflow can support reliable optimization of dewatering well layouts while reducing repeated numerical evaluations. A physics-constrained residual surrogate combines a multi-task U-Net with interior-domain PDE regularization and soft correction. On 1200 simulated cases, CNN surrogates outperform an operator-learning baseline for this geometry-sensitive task: U-Net, ResNet, and FNO achieve drawdown RMSE(s) of 0.835, 0.838, and 0.871 m. Physics regularization improves physical consistency, and Soft-corr PCRS provides a better balance between predictive accuracy and PDE consistency. Coupled with NSGA-II and MOPSO, 100 k surrogate evaluations finish in about 1 min, substantially reducing dependence on repeated 3D finite-element analyses and enabling rapid optimization of dewatering well layouts for engineering design under irregular excavation geometry and coupled constraints.

Electromagnetic pollution from electronic devices, cellular base stations, and high-voltage transmission lines has increased the demand for cementitious materials with electromagnetic wave (EMW) absorption. However, the regulation of pore structures, 3D printing paths, and impedance matching characteristics in cementitious composites remains unclear. In this study, pre-saturated superabsorbent polymer (PSAP) was introduced into 3D-printed cementitious composites. The effects of printing paths, including parallel, circular, cross-parallel, diagonal-parallel, and diagonal-cross paths, and uniform or gradient PSAP interlayer distribution schemes on EMW absorption, pore structure, and mechanical properties were investigated. Results revealed that the introduction of PSAP between layers generally enhanced measured EMW absorption performance. The specimen printed with a circular path and a uniform nominal PSAP volume fraction of 60 vol% achieved a reflection loss of −26.61 dB and an absorption bandwidth of 14.77 GHz, while maintaining a compressive strength of 32.3 MPa. Uniform PSAP distributions were preferable in the 2–10 GHz range, whereas gradient schemes were more effective in the 10–18 GHz range. Regardless of the PSAP gradient scheme, the circular printing path consistently provided the best absorption performance. However, compressive and flexural strengths decreased with increasing nominal PSAP volume fraction, and gradient schemes yielded lower strengths than uniform schemes. A conceptual 3D-printed PSAP-induced irregular pore-structure model was proposed to explain the measured absorption performance based on the observed irregular and locally adjacent or partially connected cavities. Scanning electron microscopy (SEM) and mercury intrusion porosimetry (MIP) analyses confirmed that PSAP incorporation increased the SEM-based porosity and the fractions of larger pores.

Due to the weakness of concrete’s tensile properties, reinforcement is required for 3D concrete printing (3DCP) structures. Among all methods, the one of steel rebar reinforced concrete confined by 3DCP formwork shows potential without affecting the design flexibility. However, investigations on the bond-slip performance of the steel rebar embedded into concrete confined by 3DCP formwork (SR_3DPF) are limited. Moreover, despite 3DCP formwork as the permanent formwork, it remains unclear whether it can work as a normal concrete cover to offer confinement for embedded steel rebar. To fill these gaps, the bond-slip performance of steel rebar in normal concrete without 3DCP formwork (SR_CANP) and SR_3DPF was studied experimentally and numerically. Pull-out test results showed that the bond strength of SR_3DPF is lower than that of SR_CANP, even though 3DCP concrete is stronger than cast concrete, suggesting that 3DCP formwork cannot work as the normal concrete cover. Through combining simulation and pull-out test results, it was determined that the low bond strength of SR_3DPF was attributed to the poor interfacial bond between cast concrete and 3DCP formwork. Additionally, simulation results revealed that besides improving interfacial mechanical properties, enhancing the mechanical properties of cast concrete and increasing the proportion of cast concrete in the concrete cover can benefit the bond strength of SR_3DPF.

Following publication, concerns were raised regarding the relevance of a few references in this publication [...]

The infrastructure construction in the South China Sea confronts crises in raw materials and construction methods. The 3D printed coral sand mortar emerges as a feasible solution due to the abundant coral sand resources on the islands and the advancements in automation technology. However, the intrinsic high porosity and irregular structure morphology of coral sand result in inferior mechanical properties. This paper first incorporated supplementary cementitious materials (SCM), specifically 10% silica fume (SF) and 15% fly ash (FA), to alleviate the materials' defects and guarantee the 3D printability. Subsequently, four modifiers (nano-SiO2, nanoTiO2, glacial acetic acid, and pre-wetting) for the coral sand were utilized to explore the mechanism of properties enhancement and microstructural evolution. Compared with the control design, 2% nano-SiO2 achieved the optimal enhancement, with the compressive and flexural strengths increased by 7.6% and 13.99%, respectively. This enhancement is attributed to the high pozzolanic reactivity and specific surface area of nano-SiO2, which promoted the formation of additional Calcium Silicate Hydrate (C-S-H), and both the C-S-H gel and nano-SiO2 particles contributed to densifying the matrix by filling pores and the Interfacial Transition Zone (ITZ). Similarly, the 3% glacial acetic acid soaking techniques enhanced the compressive and flexural strength by 2.1% and 4.41% via the creation of micro-voids from removed surface impurities, which accommodated additional C-S-H. However, 2% nano-TiO2 and pre-wetting led to a 7.86% and 12.03% reduction in mechanical performance, respectively. The low-activity TiO2 and excessive water impede the cement hydration reaction and diminish the mixture's compactness. These findings establish a foundation for future research on 3D printed modified coral sand and fill the gap for 3D printed coral sand mortar within infrastructure construction. By validating the feasibility of utilizing local coral sand resources, this study provides a sustainable and efficient technical pathway to mitigate raw material shortages, thereby accelerating the construction of offshore infrastructure in resourcescarce island environments.

3D printed concrete technology has demonstrated great potential in transforming construction methods, improving efficiency, and reducing environmental impacts. However, the current quality control and identification of 3D printed concrete mainly rely on manual experience and traditional non-real-time measurements, enabling the printed quality to face major challenges. Although an increasing number of studies have investigated automated quality monitoring and defect detection in 3D printed concrete, a dedicated review that systematically synthesizes these methods is still lacking. This paper provides a comprehensive review of automated quality monitoring methods for 3D printed concrete, focusing on current techniques, challenges, and future applications. Optical image processing and machine learning have been successfully used to detect defects in 3D printed concrete, although these methods have limitations in real-time performance, automation, and data quality. Further, deep learning-based methods have shown great potential in improving the accuracy and automation of defect detection, although data annotation, model generalization, large-scale construction projects, and real-time integration still face challenges. Finally, the integration of quality monitoring with building information modeling and further developments in multi-source data fusion, data augmentation, real-time adaptive control, and active quality control are recommended to address current challenges.

Engineered cementitious composite (ECC) is a high-performance strain-hardening material widely used in durable infrastructure, yet its complex multi-parameter interactions make accurate mixture design and performance prediction challenging. This study aims to establish an EDFrame, which is an integrated prediction framework for engineered cementitious composite (ECC). First, two original datasets of ECC's tensile stress and strain are collected from the comprehensive and authoritative literature, comprising 18 features and 10 categories of single or hybrid fibers. Data augmentation is then performed using a constraints-modified Conditional Tabular Generative Adversarial Network (Tuned-CTGAN), with two traditional methods for comparison. A One-Dimensional Convolutional Neural Network with a residual module (1D-Residual CNN) is developed to predict tensile stress and strain, and its performance was compared against five popular machine learning models. The interpretability of the proposed model has been achieved through Partial Dependence Plot (PDP) and Kernel SHAP analyses. The results demonstrate that Tuned-CTGAN effectively generates reliable synthetic data, significantly improving the R2 of 1D-Residual CNN from 0.8658 to 0.9128 for tensile stress and from 0.8433 to 0.9378 for tensile strain, outperforming all compared models. PDP analysis identifies optimal fiber content (1.5-2%) and fiber length (12-20 mm) ranges for enhanced tensile performance, while SHAP analysis reveals fiber length and diameter as the most critical features influencing tensile stress and strain, respectively. The proposed EDFrame provides a robust and interpretable solution for ECC performance prediction, supporting efficient and accurate mixture design in engineering practice.

Accurate prediction of compressive strength is essential for improving the performance and durability of Engineered Cementitious Composites (ECC) in construction applications. Traditional methods often fall short in accounting for the complex interactions between material properties, such as fiber type, matrix composition, and curing conditions. To address this challenge, this study presents an advanced ensemble learning framework based on a dataset of 313 ECC samples characterized by 18 key features. The ensemble model integrates three base learners, namely Random Forest (RF), Gradient Boosting Decision Trees (GBDT), and Support Vector Regression (SVR), along with a meta-learner selected from ten candidate models. The proposed ensemble model demonstrates significantly higher prediction accuracy compared to conventional approaches. The results show that the ensemble model achieves a coefficient of determination (R2) of 0.896, a root mean square error (RMSE) of 5.734, and a mean absolute error (MAE) of 4.505, substantially outperforming individual models. Among the evaluated meta-learners, Lasso Regression was identified as the optimal choice. Its regularization capability effectively mitigated overfitting and enhanced generalization, leading to a notable improvement in the final predictive performance of the stacking framework. Furthermore, SHapley Additive exPlanations (SHAP) and Partial Dependence Plots (PDP) were employed for model interpretability and visualization. The analysis reveals that factors such as fiber elastic modulus, silica fume content, and fiber volume fraction significantly contribute to the enhancement of ECC compressive strength. This model provides practical insights for optimizing the design and application of ECC materials.

Solid waste generation from industrial production and construction activities has increased sharply in recent years. Improper disposal and abandonment of solid waste have led to significant resource loss and pose serious risks to soil, water, and air quality. As a result, incorporating solid waste into concrete production presents a promising strategy to reduce environmental pressure and promote sustainable development. This review investigates the feasibility and performance of various solid waste materials, such as fly ash, limestone powder, stone sludge, silica fume, waste glass, rubber particles, and recycled concrete sand, that are incorporated into 3D printed concrete (3DPC). These materials demonstrate the potential to enhance rheological behavior, mechanical strength, interfacial bonding, and dimensional stability of printable composites. Additionally, their incorporation reduces carbon emissions, diverts waste from landfills, and promotes resource circularity. The benefits and limitations of solid waste are critically evaluated, including printability challenges arising from particle morphology and water absorption. The review emphasizes the need for standardized cost-effectiveness analysis frameworks to quantify economic and environmental balance. Finally, this study highlights the importance of optimizing mix design, surface treatment techniques, and large-scale production strategies to enable widespread adoption of waste-based 3DPC in sustainable construction.

We study an optimal control problem in which both the objective function and the dynamic constraint contain an uncertain parameter. Since the distribution of this uncertain parameter is not exactly known, the objective function is taken as the worst-case expectation over a set of possible distributions of the uncertain parameter. This ambiguity set of distributions is, in turn, defined by the first two moments of the random variables involved. The optimal control is found by minimizing the worst-case expectation over all possible distributions in this set. If the distributions are discrete, the stochastic minimax optimal control problem can be converted into a conventional optimal control problem via duality, which is then approximated as a finite-dimensional optimization problem via the control parametrization. We derive necessary conditions of optimality and propose an algorithm to solve the approximation optimization problem. The results of discrete probability distribution are then extended to the case with one dimensional continuous stochastic variable by applying the control parametrization methodology on the continuous stochastic variable, and the convergence results are derived. A numerical example is present to illustrate the potential application of the proposed model and the effectiveness of the algorithm.

The internal micro-defects of 3D printed concrete (3DPC) play a pivotal role in influencing its mechanical properties. Nonetheless, the acquisition of representative internal micro-defect information is hindered by computational inefficiencies and quantification limitations of the current equipment system. This paper proposes a deep learning based system to assist SEM equipment in automatically quantifying micro-defects of 3DPC for indepth microstructural analysis that surpasses traditional SEM methods. Through optimal resizing approach and model enhancement tactics, the proposed micro-defect segmentation model leverages advantages of both convolutional neural networks and transformer. This improvement segmentation capability achieves higher accuracy and faster speed than current algorithms, enabling system to achieve accurate quantitative analyses of micro-defects. Using this automated analysis system, the relationship among micro-defect areas in 3DPC, mechanical properties, and printer parameters is investigated. Therefore, the proposed system reduces labour and computational time, demonstrating significant potential for applications in analyzing concrete microstructure.

3D concrete printing (3DCP) offers formwork-free construction, design flexibility, and potential for solid waste utilization. Red mud (RM), an industrial waste, contains active oxides that enhance hydration and strengthen concrete. However, the variations in RM treatment, dosage, and high-temperature performance remain major challenges for the effective application of RM in 3DCP. This study investigated the influence of manufacturing procedures containing Red Mud Sintering Method (RMSM)and Red Mud Bayer Method (RMBM) in 3D-printed concrete at 10-20 %replacement fraction. Meanwhile, the mechanical performance was evaluated under both ambient conditions and elevated temperatures of 300 degrees C and 600 degrees C. In addition, X-ray microtomography (XCT), scanning electron microscopy (SEM), and Digital Image Correlation (DIC) analyses were conducted to investigate the mechanical behavior and microstructure of 3DCP mixtures containing RM. The results indicate that RMSMbased samples outperformed RMBM-based specimens, reaching 57.8 MPa of compressive strength at 10 % replacement. SEM analysis revealed that RMSM mixtures exhibited a denser and more homogeneous microstructure than RMBM mixtures. In terms of high-temperature performance, the RM-0-1 sample achieved the highest compressive strength of 57.8 MPa at 300 degrees C. XCT analysis confirmed that RMSM-based concrete exhibited a significantly lower porosity of 0.144 % at elevated temperature conditions, indicating enhanced thermal stability. Furthermore, DIC analysis demonstrated that RMSM incorporation improved deformation resistance under ambient conditions and partially mitigated strength reduction at elevated temperatures.

Urban roads and infrastructure, vital to city operations, face growing threats from subsurface anomalies like cracks and cavities. Ground Penetrating Radar (GPR) effectively visualizes underground conditions employing electromagnetic (EM) waves; however, accurate anomaly detection via GPR remains challenging due to limited labeled data, varying subsurface conditions, and indistinct target boundaries. Although visually image-like, GPR data fundamentally represent EM waves, with variations within and between waves critical for identifying anomalies. Addressing these, we propose the Reservoir-enhanced Segment Anything Model (Res-SAM), an innovative framework exploiting both visual discernibility and wave-changing properties of GPR data. Res-SAM initially identifies apparent candidate anomaly regions given minimal prompts, and further refines them by analyzing anomaly-induced changing information within and between EM waves in local GPR data, enabling precise and complete anomaly region extraction and category determination. Real-world experiments demonstrate that Res-SAM achieves high detection accuracy (>85 and outperforms state-of-the-art. Notably, Res-SAM requires only minimal accessible non-target data, avoids intensive training, and incorporates simple human interaction to enhance reliability. Our research provides a scalable, resource-efficient solution for rapid subsurface anomaly detection across diverse environments, improving urban safety monitoring while reducing manual effort and computational cost.

Residential buildings significantly impact global energy consumption. Appropriate residential fa & ccedil;ade designs can considerably reduce energy consumption in maintaining indoor comfort. Current research on residential energy- efficient fa & ccedil;ade design primarily focuses on single-objective studies and exploring parameter boundaries of isolated fa & ccedil;ade elements, neglecting holistic perspectives. It results in a research deficiency in multi-objective optimisation and integral design approaches. This study presents an innovative AI-aided methodology integrating Building Information Modelling (BIM) and Generative Design (GD) to automate multi-objective optimisation and energy-efficient compliance assurance in Australian residential fa & ccedil;ade design. Through the developed BIM-based GD program, well-founded and compliant integral energy-efficient fa & ccedil;ade designs can be generated and modelled automatically and efficiently. A practical case study on a self-contained dwelling using the developed program validates the feasibility and effectiveness of the proposed approach. The program can successfully generate multiple fa & ccedil;ade designs within a significantly short time, only taking an average of 2-3 s per design. The generated fa & ccedil;ade designs are verified energy efficient, reducing around 6.7 % in heating loads and 3.5 % in cooling loads compared to a reference building. These results demonstrate that the proposed approach enables the efficient generation of residential integral fa & ccedil;ade designs while ensuring the designs' energy efficiency. This study's contributions include advancing multi-objective optimisation, streamlining compliance processes, and demonstrating a practical method for AI-aided integral fa & ccedil;ade design. The paper also discusses limitations and future directions. The findings and methodologies provide valuable insights for advancing AI-aided energy-efficient building design.