Introduction Conventional concrete suffers from its inherently low tensile strength.To address this drawback,researchers propose steel fiber reinforced concrete(SFRC)as a novel material that significantly improves tensile properties and exhibits outstanding performance in cracking resistance,toughness,and durability.However,the incorporation of steel fibers makes the mechanical behavior of SFRC more complex.To ensure its safe application in structural engineering,an accurate prediction of its mechanical properties is essential. Artificial intelligence technologies are widely applied to accurately evaluate the mechanical properties of concrete.However,the existing machine learning models for SFRC are typically trained on a limited number of samples(i.e.,fewer than 300),which restricts their generalization capability.Moreover,these models often lack interpretability,undermining their credibility and limiting their practical application in engineering projects. This study was to establish a database for the compressive and splitting tensile strengths of SFRC,develop corresponding machine learning prediction models,and evaluate their predictive performance.Furthermore,the interpretability of the models was analyzed.The findings of this study could offer some insights into the development of high-precision machine learning models for SFRC and their interpretability analysis,thereby promoting their application in engineering practices. Methods In this study,a total of 636 experimental data points on SFRC were collected from 24 independent studies to establish a comprehensive database,including 419 data points for compressive strength and 217 for splitting tensile strength.Each data entry contains a mix-related information such as the water-to-binder ratio,sand ratio,aggregate-to-binder ratio,maximum size of coarse aggregate,fiber shape factor,fiber aspect ratio,fiber volume fraction,and the corresponding compressive or splitting tensile strength. Based on this database,five machine learning models(i.e.,Random Forest(RF),Gradient Boosting Regression Tree(GBRT),Extreme Gradient Boosting(XGB),Light Gradient Boosting Machine(LGBM)and Bayesian Neural Network(BNN))were proposed.The predictive performance of these models was evaluated using the coefficient of determination(R²),mean absolute percentage error(MAPE),and root mean square error(RMSE),in order to identify the most effective prediction model. For the selected model,the SHAP(SHapley Additive exPlanations)method was employed to analyze the trends and contribution of each input parameter on the compressive and splitting tensile strengths of SFRC.In addition,the Partial Dependence Plot(PDP)and Individual Conditional Expectation(ICE)methods were also used to quantitatively investigate the variation patterns of predicted values with respect to individual input parameters. Results and discussion All the models proposed demonstrate a great predictive performance.The R2 for the test set ranges from 0.84 to 0.90,while for the training set it ranges from 0.80 to 0.90.A small difference between them(i.e.,both being>0.75)indicates a good predictive accuracy and a generalization ability.For the overall performance across R²,MAPE and RMSE metrics,the LGBM model shows the optimum prediction performance. The SHAP analysis reveals that the compressive and splitting tensile strengths of SFRC decrease with increasing water-to-binder ratio and aggregate-to-binder ratio,while the strengths both increase at a higher sand ratio and a greater fiber reinforcement factor.In addition,as the maximum size of coarse aggregate increases,the compressive strength decreases,whereas the splitting tensile strength increaseds.These results are consistent with the practical observations.Based on the mean absolute SHAP values,the input parameters influencing the compressive strength in a descending order are water-to-cement ratio,sand ratio,fiber reinforcement factor,aggregate-to-cement ratio,and maximum coarse aggregate size.For splitting tensile strength,the order is fiber reinforcement factor,sand ratio,water-to-binder ratio,aggregate-to-binder ratio,and maximum coarse aggregate size. The influence and parameter importance rankings obtained through the PDP and ICE analysis are in a reasonable agreement with those derived from SHAP,further validating the interpretability and reliability of the model. Conclusion A large-scale database containing 636 data entries was established for the compressive and splitting tensile strengths of SFRC.Based on this database,five machine learning prediction models(i.e.,RF,GBRT,XGB,LGBM and BNN)were proposed.All the models exhibited good predictive performance and generalization ability,significantly outperforming the existing prediction methods.Among them,the LGBM model showed the optimum overall performance.The SHAP,PDP and ICE techniques were employed to analyze the influence and importance of each input parameter on the compressive and splitting tensile strengths.Based on the variable contribution analysis,the main factors affecting the compressive strength were the water-to-cement ratio,sand ratio,fiber reinforcement factor,aggregate-to-cement ratio,and maximum coarse aggregate size.In contrast,the key determinants of splitting tensile strength were the fiber reinforcement coefficient,sand ratio and water-to-cement ratio.

A reliable bond between steel reinforcement and recycled aggregate concrete is essential for ensuring composite action and structural integrity. However, flexural pull-out test data that realistically capture this interfacial behavior remain limited, and existing bond strength models often neglect several key parameters. Research on anchorage length design is also relatively scarce. To address these limitations, this study adopts a data-driven framework to compile and analyze available experimental evidence, aiming to develop a reliable predictive model for peak bond strength and to propose an anchorage length formula with appropriate safety margins. A total of 158 test records from ten representative studies were collected under comparable specimen preparation and loading conditions to establish a baseline database. To mitigate data sparsity, advanced data augmentation techniques were employed to expand the dataset to 500 samples while preserving physical boundary constraints, statistical characteristics, and intrinsic correlations, thereby compensating for missing variables and enhancing data diversity. Based on the enriched dataset, five ensemble learning algorithms combined with interpretability tools were used to predict and interpret bond strength, and a reliability analysis was conducted to evaluate anchorage performance. The results indicate that bond strength is primarily governed by the ratio of bar diameter to bond length, cover thickness, recycled coarse aggregate content and quality, matrix strength, stirrup parameters, and recycled fine aggregate properties. All variables exhibit positive correlations except those related to recycled aggregates. Among the models, gradient boosting achieved the highest predictive accuracy without overfitting, while reliability analysis confirmed representative models provide adequate safety margins.

Recycled aggregate concrete (RAC) offers clear environmental benefits, yet its heterogeneous internal structure limits broader application. Conventional tests cannot isolate key factors such as the amount and spatial distribution of adhered (aged) mortar. To address this, a mesoscale simulation framework is developed to quantify their influence on RAC's mechanical behavior. RAC is idealized as a six-phase composite and modeled with a microplane formulation capable of capturing damage evolution, crack initiation and growth, and tension-compression responses while reducing mesh sensitivity. After validation against experiments, parametric analyses were conducted to clarify the roles of adhered-mortar characteristics. The main conclusions are: (a) The model accurately reproduced RAC's tensile and compressive behavior, including damage patterns, crack evolution and axial-lateral deformation. (b) When new and old mortar had equal strengths, increasing adhered mortar content (0-100%) reduced compressive strength, elastic modulus and tensile strength, with the smallest drop in tensile strength. (c) A uniform coating of old mortar around recycled aggregates lowered uniaxial strength compared with a scattered distribution, with little effect on elastic modulus. (d) With the same average aged-mortar strength, a wider strength variation among aged-mortar phases further decreased RAC strength.

Basalt fiber reinforced recycled aggregate concrete (BFRAC) is an emerging sustainable material that integrates the superior strength and durability of basalt fibers with the environmental benefits of recycled aggregates (RAs). Although its low-carbon potential has been widely recognized, the inherent heterogeneity of BFRAC introduces considerable complexity in its mechanical behavior, especially under elevated temperatures, a topic that remains insufficiently explored. The present study addresses this challenge through an extensive experimental program involving 600 specimens exposed to temperatures up to 800 degrees C, followed by systematic evaluations of their compressive and tensile properties. The investigated parameters include the water-to-cement ratio, RA replacement level, fiber dosage, incorporation method and heating temperature. The results show that: (1) RAs reduce mechanical strength at ambient temperature but effectively alleviate fire-induced spalling at 600-800 degrees C; (2) Basalt fibers significantly improve tensile strength and ductility below 400 degrees C, though their reinforcing effect weakens beyond 600 degrees C due to interfacial degradation; (3) The combined use of RAs and fibers alters failure modes and enhances residual performance after moderate thermal exposure; and (4) A stress-strain model and three predictive equations for key mechanical properties are established, enabling rapid evaluation of BFRAC's post-fire mechanical behavior. These findings provide theoretical support and modeling foundations for assessing the residual load-bearing capacity and mechanical performance of BFRAC structures after fire exposure.

Recycled rubber aggregate concrete (RRAC), a sustainable composite in which end-of-life tire rubber replaces natural aggregates, aids waste reduction, conserves resources, lowers structural weight and enhances acoustic and energy-dissipation performance. Yet its wider deployment remains constrained by limited accuracy in predicting mechanical properties and by the inefficiency of conventional mix-design practices. This study thus introduces an integrated framework that unifies property prediction, sustainability assessment and mix-design optimization. A dataset of 1382 experiments was used to train compressive strength (fc) and elastic modulus (E) models using random forest, gradient-boosted regression trees, extreme gradient boosting (XGB), light gradient boosting machine and a Bayesian neural network, from which the top-performing model was identified. Model transparency was achieved through Shapley additive explanations, partial dependence plots and individual conditional expectation analysis. Life-cycle carbon emissions of RRAC were quantified, and particle swarm optimization was employed to balance fc, E and carbon footprint, yielding optimized mixture formulations. Key findings include: (a) Predictive models attained R2 values of 0.584-0.759 for fc and 0.674-0.842 for E, with traintest gaps <= 0.05, demonstrating solid accuracy and generalization, with XGB performing best. (b) Featureimportance analysis showed that fc was governed primarily by recycled fine-aggregate substitution, water-tocement ratio, recycled coarse aggregate substitution, sand ratio and aggregate-to-cement ratio, with E following a similar hierarchy. (c) Particle swarm optimization produced mix designs that reconcile strength, stiffness and emissions. Relative to unoptimized mixtures, optimized RRAC lowered carbon emissions by 20 %- 55 % without sacrificing mechanical performance, offering a robust pathway toward sustainable concrete design.

Using recycled brick aggregates (RBAs) in concrete provides an eco-friendly and cost-effective alternative to natural aggregates. However, higher RBA replacement ratios can compromise mechanical performance, posing a challenge to sustainability without sacrificing structural integrity. To tackle this, a multi-objective mix design approach is necessary. This study thus introduces an intelligent optimization framework for recycled brick aggregate concrete (RBAC), which first develops predictive models for compressive strength (fc) and elastic modulus (Ec) using Markov Chain Monte Carlo sampling, based on 703 experimental data points. A life cycle assessment (LCA) module then evaluates production costs and carbon emissions, considering the effects of RBA replacement ratio and water absorption, while maintaining mechanical performance. To find optimal mix designs, a cuckoo search algorithm balances mechanical, economic and environmental goals. The results show: (i) The predictive models capture the influence of key variables and outperform traditional empirical equations in accuracy. (ii) When optimizing forfc, Ec and cost, the framework achieves up to 36 % cost savings; for fc, Ec and carbon emissions, reductions reach 49 %. A combined optimization results in 25 % cost savings and 38 % emissions reduction. (iii) RBAs are unsuitable for concrete with compressive strength above 50 MPa but work well for 20 MPa mixes, with water absorption limited to 9.4 % for strengths >= 30 MPa. (iv) Overall, the framework enhances the effective utilization of RBAs and offers practical strategies for sustainable and performance-oriented concrete mix design.

Recycled lump-aggregate concrete (RLAC) is produced by combining large-scale recycled lumps (RLs) with newly mixed recycled aggregate concrete, serving as a new type of construction material for effectively recycling construction and demolition waste. A new type of RLAC precast beam with inclined-crossed (I-C) stirrup was proposed in previous research, which successfully resolves the production inefficiencies arising from the challenges when applying RLAC in conventional precast beams. This study experimentally investigates its flexural behaviour, focusing on stirrup type, spacing, compressive bar diameter, and RL substitution rate, and further analyzes compressive bar buckling to develop an optimized I-C stirrup layout. The findings are as follows: (a) For the similar consumption of stirrup steel, precast beams with I-C stirrups exhibit flexural capacity and ductility equal to or slightly greater than those with traditional vertical stirrups, while the former also show smaller average crack spacing and maximum crack width; (b) Increasing stirrup spacing reduces the flexural ductility of beams with I-C stirrups, but their flexural capacity either remains unchanged (with no premature buckling of the compressive longitudinal bar) or decreases by less than 10 % (with premature buckling); (c) The use of RLs within the precast part has a limited effect on the flexural capacity and ductility of the beams; (d) The proposed method for judging premature buckling of the compressive longitudinal bar is reasonable, and the maximum allowable stirrup spacing derived from it could be employed as a reference for practical.

Predicting concrete behavior under high temperatures and optimizing fire-resistant mix designs remain key challenges in civil engineering. To address the issues to a certain extent, this paper integrates Bayesian prediction with Cuckoo search optimization at material scale. A database of 822 high-temperature compressive strength tests was used and Bayesian model updating developed a predictive equation considering factors like water-to-binder ratio (0.27-0.90), fly ash replacement (0-0.55), slag content (0-0.61), aggregate-to-binder ratio (2.64-9.85) and fire temperature (24-700 degrees C). The Cuckoo search algorithm was then employed to optimize mix designs, balancing high-temperature strength, cost and sustainability. Key findings include: (i) The multiplicative form of the Bayesian model demonstrates high prediction accuracy (R2=0.805), with clear physical interpretation and concise expression form. (ii) At ambient temperature, the incorporation of fly ash and slag significantly mitigates concrete production costs and carbon emissions; nevertheless, these reductions are less substantial than those realized through an increased aggregate-to-binder ratio. (iii) An elevated aggregate-to-binder ratio and increased slag content improve concrete's high-temperature mechanical properties and sustainability, with optimal slag replacement and aggregate-to-binder ratio approximately 40 % and 3.6, respectively. (iv) The integrated mix design, which achieves a reduction of up to 40 % in carbon emissions and 30 % in costs, while preserving mechanical strength, comes highly recommended.

Understanding the damage mechanisms and residual mechanical properties of concrete after high-temperature exposure is essential for evaluating and repairing fire-damaged structures. However, research on recycled aggregate concrete (RAC), particularly those incorporating fine recycled aggregates (RAs) or a combination of coarse and fine RAs, remains limited. This study addresses this gap by performing both macro- and micro-level analyses of RAC. At the macro level, 225 cylindrical specimens were tested after exposure to temperatures of 20, 200, 400, 600 and 800 degrees C. The study examined how factors such as the water-to-binder ratio, RA substitution rates and the compressive strength of the parent concrete affected key compression performance metrics, including spalling, failure modes, compressive strength, elastic modulus, Poisson's ratio, peak strain and stressstrain curves. At the micro level, mercury intrusion porosimetry was used to provide further insights into the pore structure of the mortar, complementing the macro-level findings. Key findings include: (i) Due to their porous structure, RAs reduce spalling frequency between 600 and 800 degrees C without significantly affecting crack propagation or failure morphology. (ii) At room temperature, coarse and fine RAs significantly lower the compressive strength and elastic modulus of hardened concrete while increasing lateral expansion, particularly when the RA quality is poor. (iii) After high-temperature exposure, RAC demonstrates higher relative strength and elastic modulus than natural aggregate concrete (NAC) with the same water-to-binder ratio, suggesting that the porous microstructure of RAs helps preserve mechanical properties, with the residual strength and elastic modulus of both RAC and NAC converging between 600 and 800 degrees C. (iv) Poor-quality or high-content RAs lead to a sharper decline in nominal stress during the descending phase of the dimensionless stress-strain curve, especially when fine RAs are used. (v) The study proposes uniaxial compressive constitutive models and conversion relationships for residual mechanical properties, tailored to various exposure temperatures and RA contents.

Limestone calcined clay cement (LC3) and recycled aggregates (RAs) are important innovations in civil engineering, offering solutions to the industry’s energy and carbon challenges. However, limited research on their combined use has hindered widespread application. This study systematically evaluates LC3-based concrete with both coarse and fine RAs, using material testing, cost analysis and carbon footprint assessments to examine the effects of water-to-binder ratio, metakaolin-limestone powder synergy, RA content and particle quality on workability, compressive strength, elastic modulus, splitting-tensile strength, unit cost and environmental impact, while establishing clear quantitative relationships. Key findings include: (i) Using up to 50% metakaolin-limestone powder by mass, 30% coarse RA and 20% recycled sand by volume results in minimal reductions in 28-day strength and elastic modulus. However, exceeding 70%, 50% and 40%, respectively, leads to significant mechanical deterioration. (ii) LC3 concrete has 4-15% lower strength than Portland cement concrete at 14 days due to slower pozzolanic reactions, but strength levels equalize by 28 days. Similarly, recycled aggregate concrete shows lower early-age strength, with extended curing further highlighting performance differences, likely due to the residual mortar on recycled particles. (iii) Lifecycle analysis confirms that replacing 50% of the binder with metakaolin-limestone powder reduces carbon emissions by 29.9% and costs by 12.4% for strength-equivalent concrete. In contrast, RAs offer only modest sustainability benefits, with 20% recycled sand and 30% coarse RA increasing emissions by 5.4% and 2.8%, respectively, while keeping costs stable. (iv) The quality of RA particles plays a key role in their sustainable application. Additionally, transportation distance significantly affects the circular benefits of RA use but has a limited impact on the sustainability of metakaolin-limestone powder.

Utilizing large pieces of crushed concrete lumps (CCLs) in combination with freshly poured concrete to fabricate recycled lump concrete (RLC) has emerged as an eco-friendly and cost-effective method for advancing sustainable development in construction industry. Nevertheless, existing research predominantly focuses on RLC using CCLs from a single origin, while the mechanical properties of RLC containing multi-source CCLs have been largely overlooked. This neglect impedes the broader application of CCLs in engineering practices owing to uncertainties in design and safety implications. To address this gap, the current study undertakes two key tasks. First, seventy-two prisms incorporating two types of CCLs with varying mix ratios were produced and tested to evaluate the impact of CCL property discrepancies on the uniaxial mechanical behavior of RLC. Subsequently, discrete element simulations were conducted to explore the effects of the new concrete's water-to-binder ratio, CCL source number, replacement ratio, mix ratio and spatial distribution on RLC's strength and deformation characteristics. Furthermore, the simulations analyzed the internal damage evolution and final crack patterns of RLC prisms under compression. The study yielded the following key findings: (a) The inclusion of multi-source CCLs in RLC has a negligible impact on the compressive mechanical properties of the concrete. Two modified empirical models were formulated to estimate the elastic modulus and compressive strength of the composite. (b) The compressive failure pattern of RLC prisms with multi-source CCLs, marked by segmentation into smaller columns via primary cracks, mirrors that of conventional concrete and single-source RLC, though internal damage progression is strongly influenced by the strength ratio between new and old concrete. (c) RLC prisms with vertically stratified CCLs demonstrate elastic moduli comparable to those with uniformly distributed or horizontally stratified CCLs, albeit with a noticeable reduction in compressive strength. (d) The mechanical variability of RLC containing multi-source CCLs exceeds that of RLC with single-source CCLs, with compressive strength and elastic modulus reductions ranging from 2 % to 8 %.

Recycled lump-aggregate concrete (RLAC), composed of recycled lumps (RLs) and fresh recycled aggregate concrete (RAC), offers a highly effective pathway for the large-scale reutilization of construction and demolition waste, enabling the incorporation of up to approximately 80 % waste content. In marine environments, the chloride penetration resistance of RLAC emerges as a critical determinant of its structural durability. However, research addressing the chloride diffusion behavior of RLAC remains notably deficient. To bridge this research lacuna, this study conducted rapid chloride diffusion tests on RLAC, and backscattered electron imaging across various interfacial zones, quantifying the chloride diffusivities at different locales and establishing microstructure-diffusion property correlations. A mesoscale chloride diffusion model was further developed, incorporating parametric sensitivity analysis to elucidate the dominant mechanisms influencing chloride diffusion. Subsequently, a predictive framework for chloride property of RLAC were proposed. Key findings include: (a) When the RL source concrete exhibits markedly higher electric flux and chloride diffusivity than the fresh RAC, increasing the RL substitution from 0 % to 30 % raises RLAC's electric flux and chloride diffusivity by up to 17.5 % and 37.0 %, respectively. (b) Increasing the recycled sand substitution ratio from 0 % to 100 % has little effect on the electric flux and chloride diffusivity of RLAC, changing them by + 1.8 % and -1.7 %, respectively. (c) Under identical water-to-cement ratios, the interfacial transition zone (ITZ) between new mortar and natural stone demonstrates a marginally higher porosity-and hence greater chloride diffusivity-than that between new and old mortar. Existing quantitative relationships between porosity and chloride diffusivity in mortar matrix are extendable to the new mortar-natural stone ITZ and the new mortar-old mortar ITZ. (d) The suggested predictive methods for electric flux and chloride diffusivity of RLAC yield commendable accuracy: the mean value and standard deviation of the predicted-to-measured ratios are 1.00 and 0.03 for electric flux, and 0.95 and 0.03 for chloride diffusivity, respectively.

Recycled lump-aggregate concrete (RLAC), formed by mixing large-sized demolished concrete lumps (DCLs) with fresh recycled aggregate concrete, can achieve a construction and demolition waste content of up to about 80%. However, when applying RLAC to existing precast concrete laminated beams, the difficulty in placing DCLs leads to low production efficiency. Therefore, a novel precast RLAC laminated beam using inclined-crossed stirrups is proposed, effectively solving the above problem. Shear tests are conducted on such beams, checking the effects of shear span-to-depth ratio, stirrup type, stirrup diameter, stirrup spacing, and DCL replacement ratio in the precast portion on the shear properties. In addition, existing researches usually employ shear capacity prediction methods based on a separate shear model for concrete and stirrups. These methods often underestimate beams' shear capacity, due to the lack of consideration for the coupling effect between concrete and stirrups. Therefore, a shear capacity prediction method based on an integrated shear model for concrete and stirrups, applicable to beams using traditional vertical stirrups and those using inclined-crossed stirrups, is developed. The results show that: (a) With the same steel consumption of stirrups, the precast laminated beam with inclined-crossed stirrups generally exhibits comparable or slightly higher shear capacity, smaller maximum diagonal crack width, and slightly better shear ductility compared to that with traditional vertical stirrups; (b) With the same steel consumption of stirrups, increasing stirrup diameter in the beam with inclined-crossed stirrups reduces shear capacity and increases maximum diagonal crack width, both by up to 10%; (c) Using DCLs in the precast portion has a limited impact on the shear capacity, maximum diagonal crack width and shear ductility of the beam but can effectively reduce longitudinal slide along the interface during shear process; (d) The developed shear capacity prediction method demonstrates good accuracy.

Existing semi-empirical formulas for predicting punching shear capacity in FRP bar reinforced concrete flat slabs without shear reinforcement often prove inaccurate and unstable. This is primarily due to limited modeling data, inadequate consideration of key variables and neglect of complex nonlinear relationships. To address these challenges, this study delves into the utilization of advanced machine learning (ML) algorithms to offer precise and dependable estimates of punching shear capacity in such structural components. The study initially compiled a comprehensive database comprising 165 sets of test data, integrating eight crucial variables for model development. Subsequently, four data-driven models including back propagation artificial neural network (BPANN), support vector regression (SVR), random forest (RF) and gradient boosting regression tree (GBRT) were formulated to estimate the shear capacity. The efficacy of these models was assessed in comparison to existing prediction formulas. To interpret the models, this study also introduced shapley additive explanation (SHAP) and partial dependence plot (PDP) to quantitatively evaluate the influence of variables on predicted results. Research findings suggest that: (a) Among 25 existing formulas, Ju et al.’s approach performs notably well, with R2, Pre/Exp, MAPE and RMSE values at 0.76, 1.02, 22.2% and 142.8 kN, respectively. (b) ML models surpass traditional formulas in predictive accuracy, with R2, Pre/Exp, MAPE and RMSE values ranging from 0.89 to 0.93, 1.03 to 1.09, 4.8% to 9.5% and 55.4 kN to 69.0 kN, respectively. The GBRT model demonstrates the highest precision. (c) SHAP analysis of the GBRT model reveals that effective slab height and column section aspect ratio are pivotal variables influencing punching shear capacity. (d) PDP analysis quantitatively illustrates how punching shear capacity varies with each key variable.

Corrosion of steel rebar causes cracks in concrete, significantly impacting the durability of reinforced buildings. While various computational methods aim to simulate this process, only a few successfully handle both internal corrosion pressure and surface crack width evolution. To address this limitation, the study introduces an innovative mesoscale fracture-contact coupled computational method that seamlessly integrates with ABAQUS. Unlike current methods, it creatively utilizes interface elements to model stress transfer and crack propagation under tension in concrete. Additionally, it includes an additional contact model to represent the material's response to compression and shear forces. These innovations allow for accurate simulation of plain concrete mechanical properties under both tension and compression, effectively illustrating the development and failure patterns of cracks in the reinforced concrete cover. After calibrating and validating the method, the study thoroughly analyzed the process of concrete cover cracking, considering factors such as water to cement ratio, spatial location, geometry and size of coarse aggregate, rebar diameter and cover thickness. The modeling results reveal the following findings: (i) Changes in the shape and size of coarse aggregates have minimal effects on internal pressure and crack width. (ii) Higher concrete strength correlates with increased pressure on the outer surface of the rebar in an almost linear fashion, but it has little impact on surface crack width. (iii) Increasing the bar diameter decreases rust pressure but results in more cracks and wider surface cracks. (iv) Augmenting the cover thickness intensifies rust-induced pressure while simultaneously widening surface cracks.

This study delivers an in-depth analysis of predicting the compressive strength and elastic modulus of recycled brick aggregate concrete (RBAC). It assembles an extensive database of 633 compression test results and develops five ensemble learning models-random forest (RF), gradient boosting regression trees (GBRT), extreme gradient boosting (XGB), light gradient boosting machine (LGBM) and stacking (St). These models are evaluated against existing empirical formulas with respect to their effectiveness. The findings reveal that machine learning (ML) models outperform existing formulas: for compressive strength prediction, traditional models achieve a maximum determination coefficient R2 of 0.38, whereas ML models attain an R2 range of 0.91-0.94. For elastic modulus, the highest R2 values are 0.44 for traditional models and 0.97 for ML models. Notably, the LGBM and St models excel in predicting compressive strength and elastic modulus, respectively. The study also identifies critical parameters influencing RBAC's compressive behavior and highlights their impact trends. Remarkably, while the mass-weighted water absorption of coarse aggregates and the replacement ratio of recycled brick aggregates (RBAs) have less impact on compressive strength compared to the effective water-to-cement ratio, they become more influential for elastic modulus. Additionally, the decrease in elastic modulus due to higher RBA replacement ratios or increased mass-weighted water absorption of coarse aggregates exceeds the corresponding reduction in compressive strength. This research not only deepens the understanding of RBAC's mechanical properties but also provides valuable predictive tools for civil engineering applications.

Recycled aggregate concrete (RAC) is widely recognized as a promising approach for recycling construction and demolition waste. However, its practical application remains limited. A contributing factor to this limitation is the incomplete understanding of RAC’s durability characteristics, particularly its permeability, which is closely tied to the transport of harmful ions within the concrete matrix. Despite the critical importance of this issue, research in this area is still relatively sparse. This knowledge gap has motivated the current study, which seeks to thoroughly investigate the water permeability properties of RAC. In this comprehensive study, 84 specimens were fabricated for permeability testing. This study explores the influence of various factors, including the sources of coarse and fine recycled aggregates (RAs), the RA replacement ratios, and the water-to-binder ratio. The results indicate that the inclusion of both coarse and fine RAs reduces the impermeability of concrete. However, enhancing the quality of these RAs—specifically by increasing the compressive strength of the source concrete—can mitigate the reduction in impermeability. A statistical relationship is established between the average and maximum water permeation depths measured during permeability testing, leading to the formulation of a correlation between the permeability coefficient and the impermeability grade of RAC. Furthermore, mercury intrusion porosimetry testing was conducted to quantitatively analyze the pore structure of the mortar in RAC, providing a microscopic perspective that explains the macroscopic permeability behavior observed. Finally, a predictive model for estimating the permeability coefficient of RAC is proposed, demonstrating a high level of accuracy.

In recent years crushing waste brick to produce recycled brick aggregates (RBAs) has become a viable solution for reducing environmental pollution and addressing the natural resource shortage in civil engineering. To promote the widespread use of the recycled brick aggregate concrete (RBAC) in construction, this study analyzes existing test results on the attributes of RBAs and the compressive mechanical behaviors of RBAC. The review results indicate significant differences and variabilities in the characteristics of RBAs compared to natural coarse aggregates and recycled concrete coarse aggregates. RBAs have the highest absorption capacity and crushing index among the three aggregates, leading to changes in the compressive failure mechanism and a decline in the mechanical properties of RBAC. Additionally, it is also observed that existing formulas do not adequately account for the deterioration of the compressive mechanical properties of RBAC. To tackle this problem, artificial intelligence (AI) approaches including artificial neural network and multigene genetic programming are utilized to develop precise models for predicting the compressive strength and elastic modulus of RBAC. It is found that RBAC's these two mechanical indexes are mainly influenced by the standard strength of cement paste, water-to-cement ratio, sand-to-aggregate mass ratio, RBA replacement ratio and mass-weighted water absorption ratio of coarse aggregates. The AI models developed in this study accurately capture the trends of these factors and offer desirable predictive results.

Simultaneously reusing demolished concrete lumps (DCLs), recycled coarse aggregates (RCAs) and sand extracted from soil (SEFS) in concrete to form recycled lump-aggregate concrete (RLAC) can effectively improve the recycling ratio of construction and demolition waste. Fracture and flexural behaviors of RLAC were studied experimentally in laboratory tests. To reveal the mechanism of the increasing effect of the DCLs on the fracture energy of RLAC, RLAC plate fracture specimens were tested and their crack propagations were observed using digital image correlation technique. The results show that: (a) the mass DCL replacement ratio can be used in place of the average DCL area replacement ratio in the fracture section of RLAC; (b) when the compressive strength of the source concrete of DCLs is lower than that of the fresh recycled aggregate concrete (RAC), increasing the RCA replacement reduces the flexural strength, compressive strength, initiation fracture toughness and unstable fracture toughness of RLAC; increasing the DCL replacement also reduces the flexural strength, compressive strength, initiation and unstable fracture toughness of RLAC but the values of the reductions of the flexural strength, initiation and unstable fracture toughness are lower than those of the compressive strength; (c) when the fracture energy of the source concrete of DCLs is lower than that of the fresh RAC, increasing the RCA replacement reduces the fracture energy of RLAC, but increasing the DCL replacement increases it markedly by inducing tortuosity and branching of the cracks; and (d) using SEFS (the soil is the Alluvial-Proluvial soil) in place of river sand has hardly any influence on the fracture and flexural properties of RLAC. Finally, formulas are suggested which usefully predict the flexural strength, the initiation and unstable fracture toughness, and the fracture energy of RLAC.

Water permeability behavior of recycled lump/aggregate concrete (RLAC), which simultaneously contains demolished concrete lumps (DCLs), recycled coarse aggregates and sieved Alluvial-Proluvial (A-P) soil, were experimentally investigated. Scanning electron microscope and mercury intrusion pomsimetry were conducted to explain the macroscopic test results. The results showed that: (a) the RLAC's impermeability grade was not less than P8, even when the waste content in the RLAC reached about 80%; (b) the DCLs deteriorated the RLAC's impermeability when the source concrete of DCLs showed poorer impermeability; (c) the sieved A-P soil hardly influenced the RLAC's permeability; (d) the interfacial transition zone between the fresh concrete and DCLs was not a weak zone for the impermeability of RLAC; and (e) the DCLs had little influence on the pore structure of the surrounding fresh concrete. A method was proposed to predict the RLAC's permeability coefficient.