Search

This study addresses the challenge of designing hot-mix asphalt by evaluating the impact of aggregate surface area (ASA) and the number of aggregates (NA) in machine learning (ML) models. A dataset of 107 asphalt mixtures containing 0–50 % reclaimed asphalt pavement (RAP) was analyzed. Virgin aggregates and RAP particles were counted and measured via digital photography to calculate NA and ASA, with specific surface areas determined in a physics engine environment. Then, measured average aggregate particle weights were calibrated using 13 specimens. Various ML models were developed with the random forest algorithm, incorporating different input feature sets (IFS), including NA, ASA, and other basic features of the mixtures. Results revealed that including NA and ASA did not significantly improve model performance compared to using only gradation percentage inputs. Consequently, IFS-4, which includes only gradation inputs, was recommended for simplicity. The most crucial features were found to be gradation-related, with R² values around 0.90 and above achieved for stiffness modulus (ITSM), air voids, Marshall stability (MS), and theoretical maximum density (G mm). Specifically, the test R² values for air voids, ITSM, and MS were 0.96, 0.89, and 0.87, with a mean absolute percentage error (MAPE) of 5.8 %, 5.4 %, and 5.6 %, respectively. Predictions for G mm demonstrated the highest performance across all metrics with an R² value of 0.99. Air void content predictions performed better than those for ITSM, MS, and MF regarding R² and mean squared error (MSE) values, although their MAPE values were similar. These findings suggest that while NA and ASA provide additional details, gradation features are the most critical inputs for accurate ML model predictions in asphalt mixture design. [Display omitted] • The number and surface area of aggregates were calculated in different asphalt mixtures. • The number and surface area of aggregates did not make a significant difference in ML models. • Aggregate gradation had the highest feature importance in the ML model. • R2 values around 0.90 and above were obtained for air voids, ITSM, Marshall stability, and G mm. [ABSTRACT FROM AUTHOR]

Freeze-thaw cycle (FTC) deterioration poses significant durability challenges to concrete structures. However, the detailed quantitative relationship between FTC-induced internal damage pattern development and mechanical degradation remains underexplored, limiting the accuracy of FTC deterioration assessments. This study aims to provide a comprehensive analysis of microstructural damage progression and its quantitative relationship with mechanical degradation in mortar specimens throughout FTC process. Through establishing an experimental setup applied to specimens at two scales and facilitated by micro-computed tomography (micro-CT), this research concurrently evaluates the microstructural damage progression, mechanical degradation, and expansion of the specimens under various FTC severity levels. The quantification of multiple crack indices reveals a stronger correlation between compressive strength reduction and crack ingression than with increases in crack quantity or width. The findings further highlight the imperative to incorporate both expansion-induced damage and time-dependent fatigue damage in evaluating FTC deterioration, thus laying a foundation for refining FTC prediction models. • Developed a framework for synchronous evaluation of microstructural damage, mechanical reduction and expansion during FTC • Quantitatively evaluated effects of various crack indices on compressive strength reduction. • Identified a strong correlation between crack ingression and compressive strength reduction. • Emphasized the importance of integrating expansion-induced damage and time-dependent fatigue for precise simulation of FTC. [ABSTRACT FROM AUTHOR]

Corrosion greatly affects the bond behaviour of reinforced concrete. In the case of reinforcing low-corrosion reinforced concrete, the bond strength of steel bars and concrete is at an enhanced stage. Accurate prediction of bond behaviour at this stage is crucial for the rational reinforcement. This study designed ten groups, totalling 30 pull-out specimens, were accelerated corroded with various current densities and electrification times. Through 3D scanning and image analysis technology, we observed that increasing the current density yielded less homogeneous corrosion, creating pit-like rust marks. Meanwhile, we developed a model for predicting the corrosion level. The model is applicable to the corrosion level design of accelerated corrosion. Pull-out tests showed that as corrosion increased, the specimen failure mode changed from pull-out to splitting and back to pull-out, with the bond strength first increasing and then decreasing. A bond-slip prediction model for corroded reinforced concrete was also established, considering the effects of corrosion on bond strength in different stages, especially the enhancement stage. The model performs well in predicting the bond behaviour of low-corrosion reinforced concrete and can be used in conjunction with corrosion prediction models and provides guidance for reinforcement. • 3D scanning visualizes steel bar corrosion. • Predictive model for corrosion level in concrete developed. • Low-corrosion bond-slip model aids reinforcement strategies. [ABSTRACT FROM AUTHOR]

This study aimed to develop models assessing 26 machine-learning algorithms in regression analysis to predict the properties of terminal blend crumb rubber-modified bitumen (TB-CRMB) made with crosslinking additives. During the data collection, the properties of the modified binders prepared at 6, 10 and 14% of crumb rubber (CR), considering three types of modifications and eighteen blending scenarios with different interaction factors, were assessed in terms of penetration, softening point, rotational viscosity, storage stability, rheological parameters, and rutting and fatigue factors. Results showed that the Matern 5/2 Gaussian Process Regression (GPR) model demonstrated efficient performance in predicting physical, viscoelastic, rutting, and fatigue properties whereas wide artificial neural networks exhibited enhanced accuracy in predicting storage stability and rotational viscosity. The results also suggest that it is feasible to implement a single type of model developed using the Matern 5/2 GPR algorithm for accurately predicting all the TB-CRMB properties considered. The best models demonstrated that crosslinking additives significantly influenced TB-CRMB production and performance. In TB-CRMB production, sulfur as a crosslinking additive showed better compatibility than trans-polyoctenamer-rubber and significantly reduced interaction temperatures at lower CR content, leading to energy savings compared to the traditional TB production. • Fifty-four interaction seniors were explored to develop the terminal blend (TB) with crumb rubber (CR) and crosslinking additive (CA). • 26 machine-learning algorithms were examined to develop models to predict TB properties. • A single type of model for accurately predicting multiple properties of TB-CRMB was proposed. • Sulfur as a CA showed better compatibility than trans-polyoctenamer rubber and reduced interaction temperatures at lower CR content. [ABSTRACT FROM AUTHOR]

To overcome the poor energy-absorbing efficiency of existing layered panels, three kinds of novel steel-glass fiber reinforced polymer (GFRP)-foam sandwich structures composed of two steel face sheets and GFRP-foam composite cores were proposed in this paper. The effect of the composite core thickness and GFRP lattice configuration on the energy absorption performance for sandwich structures was studied through quasi-static crushing tests. Considering the role of steel face sheets, a theoretical prediction model was derived to calculate the initial absorbed energy for all types of sandwich panels. Parametric studies were carried out to investigate the influence of five different variables (axial compressive load, steel sheet thickness, GFRP web thickness, GFRP and foam compressive modulus) on the overall energy dissipation for each kind of composite panel. Test results indicated that both the thicker core component and double-layered dislocation GFRP layout could effectively promote the energy consumption property of all the steel-GFRP-foam structures, and the steel face sheet played a crucial role in the proposed sandwich panels due to its excellent absorbed-energy capacity per volume and exceptional ductility. Theoretical results demonstrated that the equivalent energy-absorbing formula was desirable to be used to assess the crushing resistant behavior of each sandwich structure. Parametric analysis results showed that a larger axial compressive load was able to improve the structural initial energy absorption, but the increase of four other structural parameters, especially the thickness of the front (rear) steel sheet, could weaken such an identical performance. • Three types of innovative steel-GFRP-foam sandwich structures were proposed to promote the overall energy-absorbing efficiency. • Quasi-static crushing experiments were taken to investigate the energy absorption performance of steel-GFRP-foam structures. • A theoretical model was firstly established to predict the initial absorbed energy of steel-GFRP-foam sandwich panels. • Parametric studies were conducted to discuss the effect of various structural indicators on the energy dissipation of steel-GFRP-foam panels. [ABSTRACT FROM AUTHOR]

Hydraulic concrete is often scoured by sediment-laden flow, resulting in wear damage on the surface of concrete. This study, used a homemade scouring device to explore mainly the effect of sand characteristics on mortar surface erosion. A 3D scanner was used to digitally reconstruct the damage contours on the specimen surface. The effects of various influencing factors on the scouring damage were investigated by analyzing the morphology and the characteristics of the scouring pits. Results indicate that impact angle and sand particle size greatly influence the morphology of the scour pit. Smaller sand particles cause more scouring damage to the mortar than larger ones. The abrasion volume is greatest at 90° impact when the sand particle size is large (0.4–2 mm) and at 20° impact when the sand particle size is small (0.1–0.2 mm). Gray correlation analysis shows that impact angle exerts the greatest effect on abrasion volume, whereas the sand content has the least effect. In addition, a mathematical model was proposed for predicting mortar abrasion volume taking into account the characteristics of sand particles. This study can provide a reference for the life prediction and durability design of concrete runners. • The effects of sediment characteristics on the wear damage of mortar were investigated in detail. • A new wear measurement method is proposed. A homemade test setup was used for the scouring test, and a 3D scanner was used to digitally reconstruct the damage profile on the specimen surface. • By analyzing the morphology and characteristics of the scour pits, the scour damage mechanism is investigated. • A mathematical prediction model for the abrasion volume prediction is deduced on the basis of experimental data. The model also takes into account the effect of sand properties. [ABSTRACT FROM AUTHOR]

The complex modulus (E*) and phase angle (δ) are fundamental viscoelastic (VE) properties of asphalt material, crucial for pavement performance analysis and modeling. This study presents a novel approach to predict the fundamental VE properties of asphalt mixtures using 20 study mixes with disparate mix design information including different Performance Graded (PG grade) binders, recycled asphalt (RAP) contents, with/without polymer modifications, and mix volumetric properties. By employing the XGBoost model for feature importance analysis, integrating the Bailey method to streamline gradation information, and utilizing master curve shape parameters for modeling, this study has effectively reduced data redundancy and dimensionality. Among four machine learning (ML) models compared, the Multi-Hidden-Layer Backpropagation Neural Network (Multi BP) model demonstrates superior applicability to the asphalt mixture dataset, providing accurate predictions for VE properties at any temperature and frequency combinations. Notably, this predictive model relies solely on the mixture design information as inputs, minimizing the need for laborious laboratory experiments. The developed model can be integrated with Mechanistic-Empirical (ME) design, predicting the variations in rutting and cracking performance of asphalt mixtures over the entire service life. This provides essential foundational input support for long-lasting pavement design. • 20 study mixes with disparate mix design information were considered in this study. • Four ML models were employed to model |E*| and δ, and the model accuracies were compared. • Feature importance analysis effectively selects feature variables. • The Bailey method efficiently reduces data redundancy and dimensionality. • The developed model can be integrated with Mechanistic-Empirical (ME) design to predict pavement performance. [ABSTRACT FROM AUTHOR]

Fiber-reinforced recycled aggregate concrete, with its remarkable mechanical properties and sustainable characteristics, has gained widespread application in building engineering as an eco-friendly building material. In the context of green development and the increasing emphasis on sustainable and environmentally friendly materials, this study investigated the variability of mechanical parameters and factors influencing the spatial distribution of fibers within high-toughness recycled aggregate concrete (HTRAC) using in-situ 4D CT experiments and four-phase meso-scale simulations. Specifically, this study considered copper-plated micro-straight steel fibers. A predictive mathematical model was developed through the integration of experimental constitutive damage parameters and computational modeling to characterize the constitutive properties of HTRAC, drawing from uniaxial compression tests and simulation data. The porosity and fiber spatial distribution characteristics of HTRAC were analyzed employing computational models that utilized CT technology and simulation methods. [Display omitted] • 3D simulation and in-situ 4D CT integrate to enhance HTRAC mesoscopic model accuracy. • Novel predictive model optimizes mechanical properties of HTRAC for sustainability. • Giving the fiber spatial distribution model, improving fiber reinforcing insights. [ABSTRACT FROM AUTHOR]

Prediction models using machine learning assume an important role in supporting decisions in asphalt road construction, such as the scheduling of tasks and the control of compaction operations. The development of prediction models for physical properties can be informed by insights from the adoption of specific machine learning techniques. However, the available evidence has not yet been synthesized. To address this deficit, we systematically selected and analyzed 30 eligible studies published in peer-reviewed journals between 2011 and 2023 for data collection and preprocessing as well as training and evaluation of prediction models. The results establish a comprehensive picture of the adoption of machine learning techniques for predicting physical properties in asphalt road construction. Specifically, the review revealed the following findings: (1) a large range of input variables and sensors used; (2) pre-specified models using few input variables that made feature selection unnecessary; (3) an emphasis on Artificial Neural Networks although empirical evidence for their higher performance is yet ambiguous; (4) low adoption rates of unitless performance metrics, which are necessary for the integration of evidence from different studies; and (5) the need for greater completeness and clarity in the reporting of training and test data used. • Prediction models address a variety of physical properties and tasks in road construction. • High levels of reported prediction performance even for models with few features. • Focus on ANN models although evidence for their higher performance is weak. • Integration of evidence undermined by low adoption rates of unitless metrics. • Extended reporting on training and test data required to assess generalizability of models. [ABSTRACT FROM AUTHOR]

Recycled aggregates (RAs) are characterized by large porosity and complex interfacial transition zones, leading to a more complex drying shrinkage mechanism for recycled aggregate concrete (RAC). In reporting on the mechanism of drying shrinkage of RAC, the internal relative humidity (RH) and drying shrinkage were investigated considering replacement ratios of recycled coarse aggregate (RCA) and recycled fine aggregate (RFA) of 0 %, 30 %, 50 %, 70 % and 100 %, respectively. Using empirical comparisons, the research identifies the link between the internal RH and drying shrinkage in RAC, offering an analysis of the shrinkage mechanism caused by the internal RH. Following that, the prediction model of RAC was proposed based on the internal RH, which will be implemented for evaluating the drying shrinkage of RAC considering the parameters corresponding to the RCA and RFA replacement ratios. Results showed that (1) A coupling effect between the RCA and RFA on the internal relative humidity is observed in the case of single-sided drying; (2) When all natural aggregates are replaced by RAs, the relative humidity loss (RH loss) of concrete at by 36.4 % compared to normal concrete, while drying shrinkage increases by 94.8 %; (3) An excellent linear correlation is achieved between internal RH and drying shrinkage of the RAC. The model proposed in this study achieves accurate prediction for the drying shrinkage of RAC with the change of the internal RH, providing the theoretical basis for the quantitative analysis of the long-term performance of RAC. • The internal relative humidity and drying shrinkage tests of RAC with different RCA and RFA contents were monitored. • The linear relationship between drying shrinkage and internal relative humidity of RAC was found. • The drying shrinkage mechanism of RAC based on internal relative humidity was revealed. • The drying shrinkage prediction model was established to predict the shrinkage of RAC based on internal relative humidity. [ABSTRACT FROM AUTHOR]

To account for the random orientation of ellipsoidal inhomogeneities, the averaging process over two Euler angles has been widely used in existing micromechanics models. However, the interaction among various inhomogeneity orientations is often overlooked and may also affect the prediction of composite elastic modulus. In this work, to consider the interaction among randomly oriented aggregates and fibers in cement concrete, a micromechanics model, the orientation interaction model (OIM), is proposed. An orientation average model (OAM) is also proposed, and by comparing OIM and OAM predictions, the effect of inhomogeneity orientation interaction on the predicted elastic modulus is shown. In both models, geometries of coarse aggregates and fibers are approximated using ellipsoids with three different semiaxes and cylinders, respectively, and their orientations are described using three Euler angles. Compared with existing random orientation models using two Euler angles, the proposed models can capture the isotropy of composites with randomly oriented asymmetric ellipsoidal inhomogeneities. By comparing the predictions of OIM, OAM, composite sphere models and the Mori-Tanaka method with spherical inhomogeneities, it is found that with the increase of the stiffness contrast between the matrix and inhomogeneities, the effects of the inhomogeneity geometry and orientation interaction become more apparent. The model predictions agree well with the experimental data. By changing the value range of three Euler angles based on the inhomogeneity orientation distribution, both models are also suitable for transversely isotropic and anisotropic composites containing several types of inhomogeneities. • The proposed models apply to composites with asymmetric ellipsoidal inhomogeneities. • Every parameter in the proposed models has a physical meaning, and model calibration is not needed. • The proposed models apply to transversely isotropic and anisotropic composites with several types of inhomogeneities. [ABSTRACT FROM AUTHOR]

Recycled Glass Concrete (RGC) exhibits low mechanical strength due to the loose structure of its interfacial transition zone. Adding supplementary cementitious materials is an effective approach to address the issues of low mechanical performance and durability in RGC. In order to improve the performance of recycled glass concrete, three series of tests were prepared for mechanical testing, drying shrinkage testing and microstructure analysis: single-doped limestone powder (RGC-L), single-doped metakaolin (RGC-M), and limestone-metakaolin mix (RGC-LM). The mechanical results demonstrate that RGC, with the addition of 5 % limestone powder and 5 % metakaolin, exhibits optimal strength improvement. Specifically, compressive strength increased by 23.3 %, splitting tensile strength by 38.1 %, and flexural strength by 7.3 %. Additionally, a mechanical prediction model for RGC was proposed, yielding relational equations for splitting tensile strength (f t), flexural strength (f r), and compressive strength (f c): f t = 0.14 f c 0.93 , f r = 0.72 f c 0.56 . The drying shrinkage results indicate that RGC containing 10 % limestone powder and 20 % metakaolin exhibits the best performance, with a 32.5 % reduction in shrinkage rate. Microstructural analysis revealed that the reaction between Al₂O₃ in metakaolin and limestone powder produces CO₃-AFm, which results in a denser structure in the interfacial transition zone of the recycled glass concrete. Ultimately, this study elucidates the micro-mechanisms and drying shrinkage mechanisms of limestone powder and metakaolin in RGC, offering a promising approach for the effective utilization of construction waste. • The reaction mechanisms of Limestone Powder and Metakaolin within Recycled Glass Concrete were explored. • The auxiliary cementing material is used to improve the interfacial transition zone of Recycled Glass Concrete. • A mechanical model for Recycled Glass Concrete was proposed. [ABSTRACT FROM AUTHOR]

Understanding the tensile strength of fibre-reinforced polymer (FRP) at high temperatures is crucial for structural safety. FRPs exposed to elevated temperatures show minimal changes in mechanical properties until reaching the resin softening temperature (T s), followed by a decline to a residual value (P residual) at the resin melting temperature (T m), with the decline past T s indicating the resin's glass transition range (T g). This review analyses 790 experiments on Tensile Strength Retention (TSR) of FRPs exposed to elevated temperatures. Various research studies have investigated the impact of parameters such as fibre and resin types, sample thickness, fibre orientation, exposure duration, cooling-off periods, and fibre-to-resin ratio on the TSR of FRPs under elevated temperatures. The results revealed that while factors like exposure time and fibre-to-resin ratio have marginal effects, others exhibit significant impacts. Researchers have developed predictive models, which are effective in accurately predicting the TSR but are limited to specific datasets, requiring frequent recalibrations when applied to different composites due to overlooking various influential factors. This ongoing need for calibration limits the generalisability of existing predictive models to other composite types. Developing a comprehensive predictive model considering these factors could reduce the need for extensive calibration when using new composite materials. • Evaluates FRP behaviour at high temperatures, revealing key insights for material resilience and structural integrity. • Examines factors affecting FRP performance at high temperatures, improving design strategies for structural safety. • Assesses TSR predictive models to improve understanding and reliability of FRP behaviour. [ABSTRACT FROM AUTHOR]

The embedment strength and stiffness of wood products are key parameters in the design of timber structures using dowel-type connections. The embedment behaviour of layered wood products such as hybrid cross-laminated timber (HCLT) will result from a combination of the behaviours of the multiple layers. In this paper, half-hole embedment tests according to ASTM D5764–2013 are presented, evaluating the embedment performance of HCLT made of fast-growing Chinese fir and OSB with self-tapping screws (STS). Six prediction models proposed by current standards and other scholars are compared with the measured data for strength, and two modified models are proposed to predict embedment strength for HCLT. Specimens loaded parallel to the grain of longitudinal layers had a higher bearing capacity, yield and ultimate embedment strength. An increase of STS diameter improved the bearing capacity of specimens but had a negative influence on the embedment strength. The modified models proposed here achieved more accurate prediction of the experimental characteristic values for HCLT than the existing methods, both for yield and ultimate embedment strength. The findings of this study represent progress towards the safe and economical structural design and wider application of innovative CLT products in the construction industry. • Fabrication and investigation of hybrid CLT. • Summary of current embedment models. • Proposal of two modified prediction models. • Comparison between current models and our proposed models. [ABSTRACT FROM AUTHOR]

This study focuses on accurately predicting rutting depth (RD) through a probabilistic prediction model, which is crucial for timely road maintenance and traffic safety. Natural Gradient Boosting (NGBoost) technology is employed to achieve this goal as it provides accurate predictions and estimates the uncertainty of predictions. Furthermore, to explain the decision-making process of the proposed prediction model, Shapley Additive Explanations (SHAP) analysis is applied to identify and quantify significant factors affecting RD. Research results show that the NGBoost model surpasses the traditional Gaussian Process and Random Forest model regarding prediction accuracy and performs better regarding certainty of prediction intervals. Meanwhile, SHAP analysis improves the transparency of model decision-making and further validates the model. Explanatory ability elucidates the impact of other characteristics on RD. The two-stage probabilistic prediction model not only enhances the reliability of predictions but also provides an in-depth understanding of the key influencing factors of pavement performance, improving the efficiency and effectiveness of maintenance measures. • Use NGBoost to predict rutting depth and give an accurate probability interval. • SHAP value is introduced to identify the key factors of rutting changes. • Relation between seasonal rutting and pavement performance indicators is analyzed. [ABSTRACT FROM AUTHOR]

Fatigue damage of asphalt binder is a notable distress at the intermediate temperature. Several empirical and mechanistic models can be utilized to predict the fatigue life under different loading modes based on limited laboratory data. The prediction accuracy of utilizing different models remains a question. This work utilized the S-VECD based approaches to analyze the effect of failure definitions on prediction accuracy. Different fatigue failure criteria are applied to linear amplitude sweep (LAS) and time sweep (TS) results, inclusive of highly-aged asphalt binders. The candidate methods include original S-VECD approach of S at failure (S f), GR, and DR failure criteria. The GR failure criterion is found to be effective in predicting fatigue failure within three candidate models by comparing measured and predicted results. Various factors influencing the prediction accuracy of fatigue life prediction are discussed. The purpose of this work is to compare the currently available prediction models based on S-VECD approaches and further investigate the potential effect of the different factors for determining the variables within the models. • Compare the prediction accuracy of practical fatigue failure criterion applied to LAS and TS • Evaluate the various factors influencing the prediction accuracy of fatigue life prediction • Verify whether the DR framework in mixture can be directly applied into asphalt binder [ABSTRACT FROM AUTHOR]

This paper investigates the influence of secondary vibration and the time interval of secondary vibration on the shrinkage properties of freshly placed concrete. In this study, four sets of cylindrical specimens with a diameter of 30 cm and a height of 100 cm were designed with different time intervals for secondary vibration. By continuously monitoring the shrinkage strain of concrete for 17 weeks, a calculation model for the early-age shrinkage strain of concrete considering the effects of secondary vibration was proposed. Concrete compressive strength tests and water expansion tests were also designed to investigate the effects of secondary vibration on compressive strength, compactness and other properties under different environments. The study results showed that secondary vibration mainly reduced concrete shrinkage during the first three days, with a reduction ratio of 10%-20%, and then stabilized. The longer the time interval for vibration, the smaller the shrinkage strain of concrete, and the more significant the improvement in compressive strength and compactness, with an increased ratio of 15%-30%. The proposed prediction model can consider the effects of secondary vibration and reflect the early-age shrinkage variation of concrete during the first seven days. After the cast concrete undergoes water absorption and expansion, it can shrink again under natural air-drying conditions, and the shrinkage rate is not affected by the time interval of secondary vibration. • The effect of secondary vibration and time intervals on the shrinkage properties of freshly placed concrete was investigated. • A calculation model for early-age shrinkage strain was proposed considering the effects of secondary vibration. • A new test device was designed to study the improvement of concrete shrinkage properties through secondary vibration. • A new test method was devised to test concrete compactness and water expansion properties. [ABSTRACT FROM AUTHOR]

The carbonation behavior of alkali-activated materials (AAMs) over time is crucial for structural durability. In this study, accelerated carbonation tests were conducted on alkali-activated slag-metakaolin (AASM) composite pastes. The effects of various slag content at 0%, 20%, and 40%, and Na 2 O content at 15%, 17%, 19%, 21%, and 23% on the pore structure, carbonation depth, and material alkalinity of AASM composite pastes were studied. The research results indicated that with the increase in slag content and Na 2 O content, the pore structure of AASM composite pastes tended to become denser. The influence of slag content and Na 2 O content on the early carbonation depth was significant. In the 0% slag content group, the carbonation depth at 3d of accelerated carbonation decreased by 61.4% when the Na 2 O content increased from 15% to 23%. With the increase in slag content, the effect of Na 2 O content on carbonation depth tended to weaken. In the 40% slag content group, the carbonation depth at 3d of accelerated carbonation remained below 10 mm. As accelerated carbonation progressed, the measured pH value rapidly decreased before 7d of accelerated carbonation and gradually stabilized thereafter. Based on pore structure parameters, predictive models for carbonation depth and material alkalinity of AASM composite pastes were constructed. The results of the accelerated carbonation test were used to validate the model calculations. The results show that the model based on pore structure parameters can effectively predict the carbonation depth and material alkalinity of AASM composite pastes. • The most probable pore diameter and porosity of alkali-activated slag-metakaolin (AASM) composite pastes are investigated. • Combined with the theoretical carbonation coefficient of AASM composite pastes, a predictive model for carbonation depth is established. • The alkalinity prediction model is proposed based to the equilibrium relationship between the ions in the pore solution. • Compared with the experimental results, the proposed carbonation depth and alkalinity models have high applicability. [ABSTRACT FROM AUTHOR]

The BBR test has received growing concerns regarding its applicability to field studies and special materials due to the large amount of long-term aged asphalt binder used in preparing asphalt binder beams and the time-consuming beam preparation and testing process. In recent years, the application of a dynamic shear rheometer (DSR) to investigate the low-temperature performance of asphalt binder has attracted widespread attention because it uses significantly less asphalt binder for testing and allows for easy specimen preparation. The objective of this study was to develop an efficient and robust method to evaluate the low-temperature performance of asphalt binder based on DSR testing. Frequency sweep tests were performed on twelve asphalt binders at temperatures ranging from 0 to 25 °C using the 8-mm parallel plate geometry. Two conversions were applied to frequency sweep test data in order to predict the flexural creep stiffness and m -value of asphalt binder, namely the conversion from frequency domain to time domain and the conversion from shear loading to bending loading. The 1S2P1D model was employed to construct the master curves of storage modulus and loss modulus. Afterwards, the flexural creep stiffness was predicted using the shear creep compliance master curve model, which was derived from the 1S2P1D model using the inverse Laplace transform, and the Poisson's ratio, which was assumed to be constant. It was found that the predicted values of S (60) and m (60) were comparable to the counterparts measured by the BBR, which well demonstrated the feasibility of the proposed DSR-BBR conversion method. In comparison to existing analysis methods, the proposed conversion method strictly complied with the linear viscoelastic theory, and it enabled efficient and robust predictions because the prediction model of flexural creep stiffness had the same parameters as those of the 1S2P1D model, thus avoiding multiple data fittings. • Prediction of flexural creep stiffness and m -value based on DSR frequency sweep tests. • Conversion of DSR test data from frequency to time domain and from shear to bending loading. • Derivation of shear creep compliance master curve model from 1S2P1D model using inverse Laplace transform. • Verified feasibility of proposed conversion method by comparing predicted and measured values. [ABSTRACT FROM AUTHOR]

This paper proposed a time series prediction method for concrete shrinkage and creep based on deep learning and the accuracy of the method was verified by field tests. Based on the experiment data in the Northwestern University (NU database), a deep learning model considering the material, environment, and history was constructed by integrating Gated Recurrent Units and Self-Attention mechanism (GSA model). To validated the model, short-term indoor and outdoor field tests of concrete shrinkage and creep were conducted. The proposed GSA model outperformed the LSTM model for both creep and shrinkage predictions. Moreover, the comparison between the GSA model prediction and the outdoor test data indicated that the model can accurately capture the real-time outdoor temperature and humidity variation through historical data. To predict the ten-year shrinkage and creep of concrete, a recursive multi-step method was proposed to extrapolate the short-term predictions of the GSA model. The effectiveness of the recursive multi-step method was validated against the long-term test data in the NU database. Overall, the proposed GSA model demonstrated high accuracy in both short-term and long-term predictions. • The proposed GSA model outperforms the LSTM model in predicting concrete shrinkage and creep. • The GSA model demonstrates superior accuracy over the code specifications in both indoorand outdoor environments. • A recursive multi-step approach for long-term prediction is proposed and validated with the tests from the NU database. [ABSTRACT FROM AUTHOR]

High-performance concrete (HPC) shows obvious autogenous shrinkage in the early stage. In actual engineering, the interior of HPC undergoes a variable temperature evolution, and the impact of variable temperature conditions on the autogenous shrinkage of HPC is unclear. This paper examines the impact of different curing temperatures on the material properties of HPC, as well as the evolution of HPC autogenous shrinkage. Three measured temperature histories and four constant temperatures were used as curing conditions. Additionally, an HPC autogenous shrinkage prediction model was established considering the impact of curing temperature. The test results indicate that the compressive strength and elastic modulus of HPC increase with increasing curing temperature. Furthermore, there is a significant increase in the autogenous shrinkage rate of HPC as the curing temperature increases. Existing maturity methods overestimate the impact of temperature on the autogenous shrinkage process of HPC. The proposed autogenous shrinkage prediction model enhances the accuracy in predicting the autogenous shrinkage of HPC under constant and variable curing temperature conditions. • The autogenous shrinkage of high-performance concrete (HPC) was studied under variable curing temperatures. • The compressive strength and elastic modulus of HPC both increased with higher curing temperatures. • The autogenous shrinkage of HPC increased as the curing temperature increases. • An autogenous shrinkage of HPC prediction model considering the influence of temperature was established. [ABSTRACT FROM AUTHOR]

California Bearing Ratio (CBR) values of the subgrade soils are key parameters in highway route survey and project design. However, in order to calculate the CBR value, time-consuming and specialized tests must be performed in the laboratory or in the field. In this study, a model based on Fuzzy set theory is proposed to predict the CBR value of soils. The geological characteristics of the fine-sized aggregate in soils are considered as the key parameter in the prediction model.For this purpose, novel Fuzzy membership functions are defined. Logarithmic interpolation approach, which takes into account the Atterberg limits and Proctor parameters of the soils, was used to determine the membership degrees of the gravel and sand sized aggregates forming the soil. The membership degrees of fine-sized aggregates were determined by basic sigmoid membership function according to the type of clay they contain. In order to verify the accuracy of the model, the calculated results were compared with the soaked CBR values obtained from laboratory tests and it was found that there was a high correlation (R2=0.86) between each other. Considering the obtained results together, it was determined that the effect of variables such as the geological class to which the aggregates forming the soil belong, which have the potential to affect the CBR value of the soils, on the results of the CBR prediction models can be eliminated by means of membership functions. • Novel Fuzzy membership functions are defined. • The geological characteristics of the fine-sized aggregate in soils are considered as the key parameter. • High correlation was obtained between Fuzzy calculations and CBR test results. [ABSTRACT FROM AUTHOR]

Estimation of the screw pull-out strength of composite railway sleepers is a complex problem due to the wide variability of their material properties and the availability of different screw geometries used to fasten the rail track. This study investigated the effect of screw geometry (screw diameter and embedded length) and sleeper material (timber, synthetic composites, recycled plastics and particulate filled resins) on the pull-out strength and developed a new prediction model for polymer composite railway sleepers. The two-way analysis of variance showed that the material shear strength has a higher influence than the screw geometries on the pull-out strength. Increasing the thread embedded length has a significant effect on all tested sleeper types due to the increased thread engaging area while the screw major diameter affects timber rather than composites owing to timber's continuously high shear strength. Based on the failure behaviour and type of fibre reinforcement, an analytical model was developed to estimate the pull-out resistance and was also verified by the results from the available literature and reports. The proposed model is found over 50% more reliable than other existing theoretical models. [Display omitted] • Evaluating influence level of thread embedded length on pull-out strength. • Evaluating influence level of thread major diameter on pull-out strength. • Pull-out strength of railway sleepers correlated to screw geometrical parameters. • Analysing failure mechanics of different sleeper materials. • Developing analytical model to predict pull-out strength of composite sleepers. [ABSTRACT FROM AUTHOR]

Presently, there is a significant increase in studies dedicated to the reuse of Reclaimed Asphalt Pavement (RAP) material obtained from the recycling of asphalt coatings. This procedure aims to reduce production costs and minimize environmental impacts. To carry out the reuse of this material, it is crucial to understand the interaction mechanisms between the aged binder of the RAP and the new binder. These mechanisms can be assessed through the Degree of Binder Activity (DoA) and the Degree of Blending (DoB). DoA represents the quantity of aged binder available for the new mixture, being an intrinsic property of each RAP. On the other hand, DoB indicates the degree of blending between the aged and new binders, being influenced by external factors such as mixing parameters, thus presenting a challenge in quantification. This study sought to establish a predictive model to quantify DoB based on Dynamic Modulus and Fourier-Transform Infrared Spectroscopy (FTIR) tests. These tests were conducted on mixtures containing four different RAPs, varying in different content levels and mixing temperatures. In each RAP, DoA was quantified, and a statistical analysis was performed to verify the significance of independent variables in dynamic modulus results. After this step, four modeling attempts were made, choosing the one that not only presented significant independent variables but also obtained the highest predicted R² value. The chosen model was validated and demonstrated small variations compared to DoB obtained by the generated equation. Thus, it is concluded that the DoB modeling conducted in this study highlights that mixing temperatures and RAP content influence factors such as DoA, dynamic modulus, FTIR, and consequently, DoB. Additionally, the practicality of the model was emphasized, as it successfully predicted DoB using a single mechanical and chemical test. • Predictive Model Redefines Asphalt Blending. • Sustainable Pavement: DoB Breakthrough. • Innovative Model Enhances Recycling Impact. [ABSTRACT FROM AUTHOR]

Nickel–iron slag, a byproduct of industrial processes in China with an annual production exceeding 400,000 tons, is considered an industrial waste material. This study focuses on the rational utilization of nickel–iron slag by investigating its mechanical properties and road performance as a roadbed fill material. Initially, a detailed analysis of the grading curve of pure nickel–iron slag was conducted, leading to the proposal of various modification schemes for nickel–iron slag. Subsequently, static triaxial tests were performed on nickel–iron slag-clay mixtures to explore the impact of different factors on the stress–strain curve of nickel–iron slag-modified soil. Utilizing these discoveries, a formula for the molar Coulomb shear strength of nickel–iron slag-modified soil was derived. In addition, a numerical simulation study of a nickel–iron slag-reinforced embankment was conducted, integrating field tests. This aimed to investigate the variations in the compression layer sedimentation-thickness ratio and settlement factor of nickel–iron slag-modified soil reinforced embankment under different filling heights and slope rates. The results informed the development of a prediction model for the settlement ratio of nickel–iron slag-modified soil-reinforced embankment. Key findings indicate that pure nickel–iron slag exhibits poorly graded gravel sand characteristics, and optimal gradation is achieved when clay doping ranges from 30% to 40%. As the clay content increases, the stress–strain curve of nickel–iron slag-clay transitions from strain-hardening to strain-softening. Furthermore, the stress–strain curve of nickel–iron slag-cement-clay exhibits strain-softening, and the shear strength fitting formula demonstrates high computational accuracy with a small error range. Numerical simulations reveal that the sink-thickness ratio and settlement factor are minimally affected by the slope rate. The sink-thickness ratio increases with the elevation of filling height, while the settlement factor fluctuates within a small range. The proposed sink-thickness ratio prediction model exhibits high accuracy and strong generalization capabilities. This comprehensive study provides valuable insights into the efficient utilization of nickel–iron slag in construction and road engineering. • Nickel–iron slag, a major industrial waste in China, can be efficiently repurposed for roadbed fill material. • Optimal gradation is achieved with 30%–40% clay doping, impacting the stress–strain curve of nickel–iron slag. • Study unveils formula for molar Coulomb strength of nickel–iron slag. • Numerical simulations reveal minimal impact of slope rate on sink-thickness ratio in reinforced embankments. • Accurate prediction model for settlement in nickel–iron slag-modified soil. [ABSTRACT FROM AUTHOR]

In order to clarify the deformation effect of fly ash concrete members after loading and unloading subsequently, experiments on creep and creep recovery of fly ash concrete were carried out on consideration of three fly ash contents and two concrete loading ages. After comparing the available prediction models for concrete creep, the B4 model was modified by multiplying the influence coefficient and creep terms, and the fly ash concrete creep prediction model was established correspondingly. The comparison shows that the modified B4 model is in good agreement with the experimental data. Furthermore, based on the experimental phenomena and the development law of creep recovery of fly ash concrete, the prediction model for creep recovery of fly ash concrete was proposed in combination with the factors affecting creep recovery which include loading age, fly ash content, etc. Finally, the calculated values of the proposed creep recovery model were compared with the experimental results and the applicability of the model was discussed. The creep and creep recovery model of fly ash concrete proposed is of great significance for revealing the time-varying law of fly ash concrete and improving the performance of concrete structures. • Creep and creep recovery of fly ash concrete. • Establishing a modified B4 model for fly ash concrete creep. • Proposing a creep recovery model for fly ash concrete considering load holding time, loading age and fly ash content. [ABSTRACT FROM AUTHOR]

Porous asphalt concrete (PAC) generates significant less noise during the loading of vehicles compared to conventional pavements. However, it might be clogged during lifespan, which results in a reduced noise reduction function. To evaluate the noise reduction function of PAC during service, the change in sound absorption coefficients of PAC under different clogging conditions was investigated by the means of laboratory tests. The standing wave tube method was used in this paper, with considering the changes in void ratios, clogging materials, and cleaning methods. The attenuation parameters of PAC sound absorption coefficients under continuous clogging and cyclic clogging-cleaning conditions were investigated. Based on this, the prediction model of the sound absorption coefficient under two clogging states was proposed. The results showed that aeolian sand had the most negative impact on the noise reduction function, with a 56 % decrease in the sound absorption coefficient after clogging. The sound absorption coefficient of the PAC remained between 0.3 and 0.4 after completely clogged. The sound absorption coefficient of the PAC can be partially even fully restored after cyclic clogging cleaning. High-pressure water washing is the most effective approach, recovering up to 80 % of the original value before clogging. Therefore, regular cleaning and maintenance of the PAC during its survice life is necessary. Predictive modeling of the sound absorption coefficient was highly correlated with measured values, allowing accurate prediction of changes after clogging and cleaning. • PAC sound absorption coefficient changes during continuous and cyclic clogging cleaning. • Clogging times has the most significant on the PAC sound absorption coefficient. • A prediction model of PAC sound absorption coefficient based on mix parameters and plugging parameters was developed. [ABSTRACT FROM AUTHOR]

Steel fibre, which can exist in the form of mono (single) or hybrid method, exhibits a distinctive fibre reinforcing effect in ultra-high-performance concrete (UHPC). Motivated by the intricate interplay of fibre effects and the ambiguous findings reported in past literature, as well as their limited comparability, the primary objective of this study is to systematically and comprehensively investigate the individual influences of fibre length, fibre geometry, fibre content, hybrid combination type, and hybridisation ratio on the flowability, compressive, and flexural performance of mono and hybrid fibre-reinforced UHPC. In addition, recognising the limited models available to predict the properties of hybrid fibre-reinforced UHPC, we developed two novel non-linear models capable of accurately predicting the compressive and flexural strength of both mono and hybrid fibre-reinforced UHPC together. These models were derived by integrating data from an extensive database encompassing both experimental and literature data. The comprehensive evaluation of the proposed models revealed their excellent accuracy in predicting the mechanical performance of both mono and hybrid fibre-reinforced UHPC. • Insights into the effects of fibre geometry, length, content, and hybridisation. • Fibre usage recommendations for optimal fresh and hardened properties. • Generalised models for mechanical properties of mono/hybrid fibre-reinforced UHPC. • Proposed regression models provide robust and accurate predictions. [ABSTRACT FROM AUTHOR]

Flexible pavements are susceptible to rutting, fatigue, and moisture damage failures. As a sustainable approach, waste plastic modifiers have been used to improve the moisture damage resistance of asphalt mixtures. However, assessing moisture damage indices such as tensile strength ratio (TSR), rut depth (RD), and loss in Marshal stability (LI) for waste plastic modified asphalt mixtures (WPMAM) via traditional lab methods is time-consuming and resource intensive. As a solution, this study suggests a gene expression programming (GEP) approach for predicting moisture damage indices of WPMAM. The models were constructed based on the outcomes from qualitative lab tests on loose coated asphalt specimens, including the static water immersion test (SWIT), rolling bottle test (RBT), and boiling water test (BWT). The proposed models demonstrated robust performance with high R values and favorable low error metrics. TSR achieved 0.99 (training) and 0.98 (testing) R values, while RD and LI displayed strong predictive accuracy with 0.95, 0.93 (training), and 0.94, 0.77 (testing) R values. Notably, lower root means square error (RMSE), mean absolute error (MAE), and relatively squared error (RSE) values reinforce the models' reliability. Sensitivity analysis identified RBT as the most influential factor for TSR, RD, and LI of WPMAM. Finally, parametric analysis verified the alignment of proposed models with underlying physical processes. Thus, these models can be successfully implemented for predicting moisture damage indices of WPMAM, offering a streamlined alternatives to labor-intensive laboratory procedures, making them valuable for optimizing material choices and implementing preventive measures in construction. • Moisture damage indices were calculated of different waste materials modified asphalt mixtures using GEP Programming. • Predictive models were developed for rut depth, TSR and loss in stability based on SWIT, BWT and RBT test results. • The proposed models were validated through various error metrices such as R, RMSE, MAE and RSE. • Sensitivity analysis identified RBT test is the most influential factor for performance of compacted asphalt mixtures. [ABSTRACT FROM AUTHOR]

To examine the dynamic characteristics of clay compounded with rubber particles, a series of resonance column experiments has been designed and performed on the rubber particle-clay mixtures (RPC) with varying rubber particle contents, rubber particle sizes, and confining pressures. The analytical method for the skeleton void ratio of RPC was proposed according to its micro-contact structure, and a prediction model was suggested to anticipate the maximum dynamic shear modulus relying on the skeleton void ratio. Furthermore, empirical prediction models were established to anticipate the dynamic shear modulus ratio and damping ratio of RPC. The findings indicated that the maximum shear modulus and shear modulus ratio will decrease with increasing rubber content, and decreasing rubber particle size and confining pressure. The findings also show that the damping ratio will increase with increasing rubber content, and decreasing rubber particle size and confining pressure. Considering good damping ratio of the RPC compared with pure clay, the RPC has good application prospects in the field of vibration isolation and reduction, and the obtained results here will provide appropriate and scientific guidance for the application of RPC. • The skeleton void ratio was developed based on the micro-contact structure of rubber particle -clay mixtures. • Models were proposed to predict the maximum shear modulus, dynamic shear modulus ratio and damping ratio of mixtures. • The maximum shear modulus and shear modulus ratio decreased as rubber content increased and rubber particle size decreased. • The damping ratio increased as rubber content increased and rubber particle size decreased. [ABSTRACT FROM AUTHOR]

Pavement deflections are critical indicators in pavement design and acceptance programs. Falling weight deflectometer (FWD) technique has been widely adopted to measure the dynamic deflections in pavement engineering. Although the dynamic effect of FWD impulse loading on deflections is significant, FWD deflections are still commonly interpreted and examined using the static analysis programs in acceptance projects because of their ease of implementation and high computational efficiency. This may lead to an inability to effectively control the construction quality of pavements. Accordingly, this study aims to propose a simple and accurate approach for converting static deflection to dynamic deflection. Layered elastic theory (LET) and spectral element method (SEM) were adopted to generate the theoretical static and dynamic deflection databases, respectively. Based on the databases, a simple approach is proposed to predict the dynamic deflection only using four static deflections. The proposed approach shows lower computational workload and higher accuracy than other traditional formula forms which include equivalent layer moduli and thicknesses. In the field engineering cases, the calculated static deflections can be 2 ∼ 3 times higher than the measured ones, while the predicted dynamic deflections from the proposed approach match the theoretical dynamic deflections well and are comparable to the measured deflections for both asphalt pavements with asphalt binder or inorganic binder bases and granular bases. This indicates that dynamic deflections can more effectively control the construction quality of asphalt pavements in the FWD deflection-based design program. Meanwhile, the drawbacks of traditional prediction methods are investigated and evaluated. The proposed approach exhibits simpler form and higher accuracy, while avoiding the drawbacks of traditional methods. Combining with the common static LET-based programs, the proposed approach provides a more reasonable and practical way for the construction quality acceptance of asphalt pavements using FWD data. • Dynamic deflection is more effective in controlling pavement construction quality. • Drawbacks of traditional prediction method of dynamic deflection are evaluated. • A simpler and more accurate approach is proposed to predict the dynamic deflection. • The proposed approach is validated and applied in the field engineering cases. [ABSTRACT FROM AUTHOR]

Shear stress prevalence in in-situ cemented paste backfill (CPB) makes understanding mode-II fracture behaviour and properties vital for the mechanical stability of the backfill structure. This study delves into the effect of cement content on CPB's shear fracture behaviour and properties. Results reveal that higher cement content leads to notable changes in load-displacement characteristics under mode-II loading, including steeper pre-peak slopes, higher peak loads, smaller peak-load displacements, and minimal post-peak resistance. However, increased cement content does not linearly enhance fracture properties. Furthermore, this study proposed predictive models to capture the influence of cement content on the evolutive fracture properties. • Pre-peak fracture behaviour of CPB is sensitive to changes in cement content. • Post-peak resistance in CPB is subtle, regardless of cement content. • Increasing cement content nonlinearly boots mode-II fracture properties. • Predictive models are developed to capture CPB's fracture properties. [ABSTRACT FROM AUTHOR]

Abrasion damage of concrete induced by sediment-laden flow has been a significant durability problem for hydraulic structures. Due to the complexity of dynamic fields and concrete properties as well as the interaction effects between them, many aspects are still subjected to debate despite the tremendous efforts that have been made. In light of this and in order to further advance our understanding, this work aims to present a state-of-the-art review on critical aspects of the abrasion behavior of concrete. Important theoretical basics and mechanisms behind abrasion-induced concrete damage are presented first and different performance indicators that have been used for abrasion damage characterization are discussed. Following that, relevant abrasion test methods are comparatively reviewed and the effects of important hydraulic parameters and concrete properties on concrete abrasion behavior are analyzed in depth. Moreover, representative prediction models proposed for the estimation of concrete abrasion damage are discussed. In addition, an overview of recent development strategies for improving concrete abrasion resistance is presented. Based on the critical analysis of the research status, some important research gaps and challenges are highlighted. • Theoretical basis and mechanisms of abrasion-induced concrete damage are presented. • Relevant abrasion test methods are comparatively reviewed. • Effects of key hydraulic parameters and concrete properties are analyzed in depth. • Important strategies for improving concrete abrasion resistance are summarized. • Existing research gaps and challenges are highlighted. [ABSTRACT FROM AUTHOR]

In the actual service process of asphalt pavement, the dynamic response of the pavement structure layer is induced by vehicle loads. The dynamic modulus is a crucial parameter for analyzing the mechanical response of asphalt pavement. Due to significant influences from temperature and loading frequency on test measurements, along with its time-temperature equivalence, it becomes imperative to investigate the behavior of dynamic modulus under different temperature frequencies. Furthermore, laboratory loading modes also exert substantial effects on dynamic modulus. Employing different loading modes during measurement often leads to significantly varied stress states in specimens, yielding distinct values for dynamic modulus. Therefore, this study conducts dynamic modulus tests using direct tension, uniaxial compression, and indirect tension under five test temperatures and six loading frequencies while examining its variations under different stress states and temperature frequencies. The corresponding master curves and prediction models of dynamic modulus are derived, providing a valuable reference for predicting the dynamic modulus under wide-area test conditions. Through comparative analysis of dynamic modulus values at different stress levels, the stress level is incorporated into the master curve to establish corresponding master curves of dynamic modulus. By applying relevant theories and research on time-temperature equivalence, we obtained dynamic modulus master curves for three loading modes by fitting and analyzing dynamic modulus test data at various temperatures and loading frequencies. Combining the Sigmoid model and Williams-Landel-Ferry(WLF) equation, we derived prediction models for dynamic modulus under three loading modes. Comparative analysis of dynamic modulus values under these three loading modes allowed us to establish transformation equations for the prediction models in different loading scenarios. Finally, representative dynamic modulus values for permanent deformation in typical pavement structure layers were predicted based on laboratory test results and the derived models. The research findings presented in this paper have significant implications in eliminating the influence of loading modes on dynamic modulus predictions, establishing a unified prediction model, and providing essential parameters for predicting permanent deformation and fatigue life equivalent temperature. • Introducing stress level variables into the dynamic modulus master curve. • Extending the two-dimensional dynamic modulus master curve to three-dimensional master surface. • Derived the conversion equations between prediction models for different loading methods. [ABSTRACT FROM AUTHOR]

This study investigates the development of pavement performance models, focusing on the predictive analysis of pavement distress using ensemble machine learning methods. We employ the Research Institute of Highway Ministry of Transport track dataset to construct predictive models for pavement distress, comparing various methodologies including Random Forest (RF) regression, Extreme Gradient Boosting (XGB) machines, Snake Optimizer-Random Forest (SO-RF), and Snake Optimizer-Extreme Gradient Boosting (SO-XGB). The approach is conducted by multiple experiments and a comprehensive comparative analysis leveraging the repeated K -Fold cross-validation method to ensure the robust assessment for these models. Our findings show that the SO-XGB method is particularly effective in predicting rutting depth. The strength of the SO-XGB model lies in its excellent performance in several performance metrics, such as coefficient of determination (R 2), the mean absolute error (MAE), root mean square error (RMSE), the mean absolute percentage error (MAPE), performance index (PI), index of agreement (IA), variance accounted for (VAF) and objective function (OBJ). These results not only illustrate the model's accuracy but also the practical applicability of SO-XGB model in automatically forecasting pavement damage. Thus, the SO-XGB method stands out as a direct, powerful, and efficient tool in the field of pavement performance prediction. [Display omitted] • Developing ML algorithms for rutting depth prediction in asphalt pavements. • Integrating snake optimization to enhance prediction accuracy of RF and XGB. • Evaluating ML models using eight performance metrics. • Utilizing repeated K-Fold cross-validation for precise model assessment. [ABSTRACT FROM AUTHOR]

Cement-based foam has emerged as a strong contender in sustainable construction owing to its superior thermal and sound insulation properties, fire resistance, and cost-effectiveness. To effectively use cement-based foam as a thermal insulation material, it is important to accurately predict its thermal conductivity. The current study aims at coining an accurate methodology for predicting the thermal conductivity of cement-based foam using state-of-the-art machine learning techniques. A comprehensive experimental dataset of 504 data points was developed and used for training ensemble learning models including XGBoost, CatBoost, LightGBM and Random Forest. The independent variables of this dataset affecting the thermal conductivity are the cast density, percentage of pozzolan, porosity, percentage of moisture, and duration of hydration in days. Using the Isolation Forest algorithm proved effective in detecting and eliminating outliers in the dataset. All the ensemble learning techniques explored in this study achieved superior predictive accuracy with a coefficient of determination greater than 0.98 on the test dataset. The influence of the input features on the thermal conductivity was visualized using the SHapley Additive exPlanations (SHAP) approach and individual conditional expectation (ICE) plots. The cast density had the greatest effect on thermal conductivity. The explainable machine learning models demonstrated superior accuracy, efficiency, and reliability in estimating the thermal insulation of cement-based foam, opening the door for wider acceptance of this material in sustainable energy efficient construction. • Comprehensive database comprising 504 experimental datasets on thermal conductivity of cement foam created. • Isolation Forest algorithm proved effective in outlier detection, enhancing model accuracy. • XGBoost, CatBoost, LightGBM and Random Forest attained robust thermal conductivity prediction. • SHapley Additive exPlanations (SHAP) identified most influential input parameters. • Accurate thermal conductivity prediction advances foamed concrete in sustainable construction. [ABSTRACT FROM AUTHOR]

Assessing the long-term durability of concrete structures subjected to external sulfate attack (ESA) poses a significant challenge, primarily due to the lack of a well-defined quantitative metric for durability limit states. Traditional performance indicators such as compressive strength, mass loss, and expansion strain are insufficient for providing a comprehensive understanding of the extent of damage and are often difficult to measure accurately in practice. The primary objective of this study is to introduce a novel durability failure indicator, namely the failure thickness of concrete under ESA exposure, designed to quantitatively assess the ultimate limit state of concrete durability. Experiments were conducted and the results confirmed that the failure thickness can reliably reflect changes in both the elastic modulus and compressive strength of concrete exposed to ESA. A prediction model for estimating the failure thickness was subsequently developed, integrating key influencing factors like sulfate concentration, initial aluminate content, and sulfate ion diffusion coefficient. The established model exhibited a high degree of correlation between the predicted and experimental failure thickness values, with a low margin of error. The findings of this research contribute to serve as a foundation for both lifespan prediction and durability design of concrete structures exposed to sulfate-rich environments. • We propose a durability failure indicator based on critical sulfate concentration. • Concrete performance deterioration can be well characterized by failure thickness. • The validity of the prediction model of failure thickness was confirmed. [ABSTRACT FROM AUTHOR]

Predicting the shrinkage deformation of the partially enclosed steel reinforced concrete (PESRC) column is important for super high-rise buildings. However, due to the hindrance effect of section steel on moisture diffusion, it is difficult to predict the shrinkage deformation of PESRC columns by the existing models. To address this issue, this study proposes a modified CEB-FIP(1990) model to precisely predict the drying shrinkage deformation of PESRC columns. Firstly, the compositions of shrinkage prediction models are analyzed based on the physical significance of each parameter. It reveals that most of the existing models can be summarized as a product of three parts, including the ultimate shrinkage coefficient, the humidity effect coefficient, and the development function of shrinkage, and each part denotes a special characteristic of shrinkage prediction models. Then, the hindrance effect of section steel is analyzed to comprehend the influence of section steel on moisture distribution and drying shrinkage in concrete. The analysis results show that the moisture diffusion of steel reinforced concrete columns is significantly influenced by the shape of the section steel. Specifically, the enclosed section steel shortens the shrinkage time and the drying shrinkage deformation, while the open section steel only slows down the shrinkage rate without affecting the drying shrinkage deformation. Besides, the finite element method verified by drying shrinkage experiment is employed to investigate the moisture distribution law and the drying shrinkage strain of PESRC columns. Finally, a modified CEB-FIP(1990) model is proposed by introducing several correction parameters to adjust the ultimate shrinkage coefficient and the development function of shrinkage. The comparison results demonstrate that the modified CEB-FIP(1990) model can be used to predict the drying shrinkage of PESRC columns. • Existing shrinkage models overestimate the deformation of PESRC columns. • Moisture distribution and drying shrinkage are greatly influenced by section steel. • The drying shrinkage behavior of the PC, RC and PESRC columns is quite different. • A modified CEB-FIP (1990) model for PESRC columns is proposed. • The modified CEB-FIP (1990) model is successfully applied in the Shanghai Tower. [ABSTRACT FROM AUTHOR]

Ultra-high performance concrete (UHPC) has super high strength, ductile behavior, and excellent durability, which is quite different from conventional concrete. Many aspects of UHPC (mixture design, mechanical properties, shrinkage behavior, engineering applications, etc.) have been widely investigated. However, the information about creep properties of UHPC is not sufficient, which is vital to engineers for safe structure design as well as for predicting the structure's long-term performance. Thus, this study systematically brings together the published research on UHPC creep behavior, and major contents are as follows: 1) introduce basic terms for UHPC creep and give comparisons with conventional concrete; 2) summarize the various influencing factors on UHPC creep (curing regime, loading age, stress level, load type, w/b, fibers, mixture compositions, humidity, specimen size, etc.); 3) explain the mechanism of UHPC creep, relaxation performance, creep prediction models or equations in specifications; 4) give engineering applications considering UHPC creep (long-term deformation, prestress loss, early-age restrained stress, etc.); 5) summarize creep coefficient data from literature and propose future perspectives. This review can give researchers and designers a comprehensive and better understanding of UHPC creep, and facilitate the development of safe provisions for UHPC structures considering creep effect. • The compressive creep coefficients of UHPC are between 0.17 to 2.47. • The tensile creep coefficients of UHPC are between 0.78 to 2.38. • Heat treatment has the most pronounced effect on reducing UHPC creep. • Earlier loading age and higher stress levels increase UHPC creep. • Studies report that UHPC's compressive and tensile creep differ a lot or are similar. [ABSTRACT FROM AUTHOR]

The evaporation rate of asphalt emulsion is critical to the strength development of asphalt emulsion materials. Its value is dependent on environmental factors, i.e. relative humidity, temperature, and wind speed. However, the quantitative relation of evaporation rate with these environmental factors are unclear. To quantificationally study the effect of environmental factors on the water evaporation rate of asphalt emulsion, a computational fluid dynamics (CFD) simulation model of water evaporation is firstly proposed and validated by experiment. Then, wide range of environmental factors are studied in the CFD simulation model, and genetic programming-based symbolic regression is used to train the simulation results of CFD samples. Finally, a relatively accurate prediction equation of the evaporation rate of asphalt emulsion versus environmental factor is proposed, and the sensitivity of the water evaporation rate under the three environmental factors is analyzed. Results show that the evaporation rate can be accurately simulated by CFD. All the three environmental factors can greatly affect the evaporation rate of asphalt emulsion, while their effect on the sensitivity of the water evaporation rate is ranked as: temperature > relative humidity > wind speed. To ensure an acceptable curing period, asphalt emulsion-based materials should be avoided to be paved in the conditions of low temperature and high relative humidity. [Display omitted] • A CFD simulation model of water evaporation of asphalt emulsion was proposed and validated by experiment. • A prediction equation of the evaporation rate of asphalt emulsion versus environmental factor was proposed. • The effect of environmental factors on the evaporation rate of asphalt emulsion was quantificationally studied. [ABSTRACT FROM AUTHOR]

Reliable prediction model of CO 2 corrosion is essential for the oil-well cement sheath integrity. This study constructed a complete database containing 301 records of published literature for machine learning modeling. By comparing four standard algorithms (artificial neural networks (ANN), radial product function neural network (RBF), random forest (RF), and support vector machine (SVM)), an effective method is proposed to predict the carbonation depth of oil-well cement under different conditions. The results showed that the coefficient of determination was 0.9732, the root mean square error was 0.6973, the mean absolute error was 0.4231, and the mean square error was 0.0765, indicating that the SVM model has excellent prediction ability. In addition, the RF model characteristic importance analysis reveals that corrosion time, corrosion-resistant additives, and partial pressure of CO 2 are the most important factors for the corrosion depth. This work provides a reliable prediction method for the CO 2 corrosion rate of oil-well cement. [Display omitted] • A Machine Learning (ML) algorithm to interpret corrosion depth is presented. • Results of four machine learning models, ANN, RF, RBF, and SVM, were compared. • SHAP visualizes the process in which all features affect the output results. • SVM achieved high prediction ability in contrast to ANN, RF and RBF. • Corrosion-resistant additive, time and CO 2 pressure are the most important factors. [ABSTRACT FROM AUTHOR]

3D concrete printing (3DCP) technology has gained significant attention in the construction industry due to its potential to revolutionize the way buildings are designed and constructed. However, successful application of 3DCP requires robust standard test procedures to characterize the printability of concrete. Printability is a measure of workability for a printing material in the 3DCP process, which is characterized by pumpability, extrudability, and buildability. This paper presents a comprehensive review of the test methods that could be used for material characterization during various printing steps, including pumping, extrusion, and layer deposition. The underlying mechanisms relevant to each step, key governing factors, and prediction models are highlighted. The influence of the variation of material and process-induced parameters on the properties of fresh concrete are explained, and the optimum parameters that will lead to an extrudable yet shape-stable printing material are discussed. The recent research on the pumpability, extrudability, and buildability of printed concrete is critically discussed, and future research needs for testing and evaluating 3DCP are identified. • Limited shearing during pumping is preferred to minimize the changes in rheology. • Extrusion energy and filament inspection are required for assessing extrudability. • Exploring the interaction between process and material parameters is essential. • Formation of the LL is key to achieving plug flow during the pumping. • Higher filtrate helps LL but excessive bleeding reduces hydrodynamic interaction. [ABSTRACT FROM AUTHOR]

Viscoelastic characterization of asphalt binders is critical for selecting suitable binders for a given climatic region and for improving long-term performance prediction using mechanistic-empirical pavement analysis and design methods, such as the Mechanistic-Empirical-Pavement-Design-Guide (MEPDG). A-VTS information (viscosity data) was used in the past and existing databases have this type of information. Nowadays, Dynamic Shear Modulus (|G*|) and Phase Angle (δ) determination using the Dynamic Shear Rheometer (DSR) is common practice. This resulted in the existence of old and new databases in highway agencies. The goal is to find a way to combine these separate data sets and merge them into a unified national database based on |G*| and δ. The work discussed in this paper is critical to make this merger possible and to use this in future binder to mixture predictive model developments and in MEPDG. In this study, |G*| and δ prediction models were developed. A machine learning technique named Artificial Neural Network (ANN) was employed in this process. 8940 data points from 41 binders were utilized in the model development process. After robust training, validation, and testing process, a |G*| prediction model with seven neurons and a δ prediction model with three neurons were reported. The models were compared with previous Bari-Witczak and Onifade-Birgisson models. ANN-based models performed better in every statistical parameter. Two equations were also extracted from the ANN models, which can be used reliably to predict |G*| and δ from A-VTS data. These two models with closed-form equations can be incorporated into the MEPDG for |G*| and δ predictions. • New models for Binder dynamic shear modulus and phase angle using binder viscosity data with temperature and frequency. • Artificial Neural Network (ANN) is utilized in the model development process. • An equation is extracted from each ANN-based model. • The prediction performance of new models is better than existing models in every statistical parameter. [ABSTRACT FROM AUTHOR]

The dynamic modulus (E*) of hot-mix asphalt mixtures is one of the most laborious and time-consuming material parameters to measure in the laboratory. It involves expensive, specialized equipment and expertize that are not readily available in most laboratories. Consequently, several efforts have been devoted to E* prediction models. Unfortunately, developing these prediction models is complex because of the numerous contributory factors and their non-linear influence on E* values. Moreover, such models are not able to prioritize or screen the major factors influencing the E* values. This study presents a new framework for analyzing the dynamic modulus influencing factors by adopting two modeling approaches. First, deep residual neural networks (DRNNs) for non-parametric approaches are used to improve the E* prediction capabilities and derive deep insight into the contributory parameters' effect on the E* value. Second, the well-known Witczak 1–40D prediction equation is used as a representative of the classical statistical modeling approach. In the validation of the models, a comprehensive laboratory database is utilized to account for all significant contributory parameters, such as binder characteristics, volumetric properties, mixture gradation, and testing circumstances parameters. Then, the performance is assessed using typical performance metrics. Lastly, intensive global sensitivity analysis (GSA) is undertaken with the assistance of Latin Hypercube Simulation. Three distinct GSA methods are used to emphasize the influence of each contributory factor on the value of E* in actual practice while reducing the possibility for result distortion owing to correlations between contributory variables. Performance metrics of the DRNNs and the Witczak 1–40D prediction models give the GSA conclusions high credibility. The GSA reveals that, among all possible inputs, the binder content, shear modulus, voids in the mineral aggregates, and temperature are the most significant factors in determining the E* value. • A robust framework for analyzing the HMA dynamic modulus prediction problem (E*). • Two different modeling approaches are used within the proposed framework. • Three powerful global sensitivity analysis methods are implemented. • Deep insights are driven about the significance of each input factor. • The input parameters are ranked according to their impact with high confidence. [ABSTRACT FROM AUTHOR]

Timber structures are susceptible to sound propagation problems. The issue can be mitigated by decoupling structural components at connections level by means of polyurethane soundproofing interlayers. These layers are much softer than timber but, at the same time, they are characterized by higher friction coefficients. In this paper, an experimental investigation is carried out to characterize the mechanical behavior of insulated screw connections assessing the implications of the soundproofing layer use. Different interlayers, screw inclination angles and setups are considered. The load–slip behavior is studied, and a proper empirical model, suitable for performing push-over analysis of buildings involving soundproofed connections, is derived. The essential design parameters, namely the slip modulus, the strength and ductility, are determined and compared with existing models predictions, thus testing their accuracy. Substantial reductions in terms of stiffness, load-carrying capacity and ductility are observed in soundproofed connections. • Timber-to-timber connections with soundproofing interlayers are tested. • Soundproofing interlayers substantially reduce mechanical performances. • An alternative definition of the ultimate slip modulus is suggested. • An empirical model for the load–slip behavior is formulated and validated. • Withdrawal failure mechanism can be considered additive to the transversal mechanisms. [ABSTRACT FROM AUTHOR]

The improved damping capacity of concrete materials is crucial for many applications, especially the ones exposed to dynamic external forces. This study explores the optimal circumstances of copolymer composite binder, ratio of fine to coarse aggregate, curing time, size and content of tyre crumb rubber on enhancing damping ratio of hybrid polymer concrete. Owing to the limitation of conventional models and experimental methods in effectively handling the multiple and complicated relationship between the input parameters of the hybrid polymer concrete and its damping capacity, two robust machine leaning models including extreme gradient boosting (XGB), and artificial neural network (ANN) algorithms were developed to predict the values of damping ratio from an experimental database of 196 samples. Multiple linear regression (MLR) was employed in this study as a benchmark for comparing with the ANN and XGB methods. The results indicate that the considered algorithms yield accurate models for predicting the damping ratio of HPC composites. However, ANN and XGB outperform MLR with an impressive R-square of 0.985 and 0.981, respectively versus 0.875. Analysing the importance of input features using the XGB algorithm also reveals that the curing time has the highest impact on predicting the damping ratio. The volume fractions of resin matrix and crumb rubber are the next most important features. The crumb rubber size, within the studied range, appears to be the least impactful one. • A large experimental dataset on damping capacity of the hybrid polymer concrete has been produced. • Dataset feature effects of optimised ratio of fine-to-coarse aggregate and curing behaviour of the novel copolymer binder used for the hybrid polymer concrete. • Damping ratio of the hybrid polymer concrete predicted using machine learning techniques. • The importance of the input variables analysed using the extreme gradient boosting algorithm. [ABSTRACT FROM AUTHOR]

Coarse aggregates are recently used to address the high production costs and severe autogenous shrinkage of ultra-high performance concrete (UHPC), however, the influence of coarse aggregates size on the dynamic performances of UHPC remains poorly comprehended. Herein, the impact of coarse aggregates on the dynamic performances of UHPC is explored through a multiscale method with the focus on hydration products, micromechanical property, meso-scale fracture and macro-scale dynamic mechanical characteristics. Results show that the increasing maximum coarse aggregates size slightly enhances both the total amount and the elastic modulus of hydration products. The rise of the maximum coarse aggregate size enhances the percentage of fracture across coarse aggregates, thus, improving the impact resistance of UHPC. Regardless of the strain rates, the toughness and dynamic compressive strength enhance with the expanding maximum coarse aggregate size, while the dynamic increase factor and peak strain decrease with the rising coarser aggregates size. The damage degree of UHPC enhances with strain rate but reduces with the increasing maximum aggregate size. A conceptual enhancement mechanism of increasing maximum coarse aggregate size on the impact resistance is proposed. The present results promote the wider engineering application of UHPC containing coarse aggregates. ● Increasing coarser aggregates size enhances the quasi-static compressive strength. ● A new prediction model of dynamic compressive strength is established. ● Fracture of larger coarse aggregate contributes to increased energy absorption capacity. ● Impact resistance of UHPC enhances with the increased maximum coarse aggregate size. [ABSTRACT FROM AUTHOR]

The influence of pumping on the rheological properties of self-consolidating concrete (SCC) is widely recognized, but there is still limited understanding of its behavior within the pipeline. In this study, we propose a novel approach using equilibrium distances to investigate the evolution of rheological properties along the pipeline. Full-scale horizontal pumping circuits spanning 328 m were established to assess the axial development of SCC's rheological properties. Thirteen mixtures, with water-to-cement (w/c) ratios ranging from 0.22 to 0.27, were pumped at discharge rates varying from 5.77 to 12.69 L/s. Our findings show that for pumping setups and material studied, both the w/c ratio of SCC and the discharge rate significantly affected the influence of rheological equilibrium distances on pressure loss gradient changes. In all tested samples, viscosity values required the entire length of the pipeline to reach post-pumping levels, while yield stress values rapidly achieved their fully developed state upon entering the pipeline. Importantly, the predicted pressure loss displayed a strong correlation with experimental measurements, with an accuracy of 89.42% across all tested mixtures. Our predictive model showed an average improvement of 4.6% in accuracy compared to the linear approach, with the most significant enhancement observed in samples with higher w/c ratios. This research sheds new light on the behavior of SCC during pumping and provides valuable insights for optimizing the pumping process. • The equilibrium distance of rheology is introduced to describe SCC flow behavior along the pipeline. • Horizontal pumping circuits spanning 328 m is used. • Factors affecting the equilibrium distance of rheology are discussed. • The prediction accuracy of the model is validated. [ABSTRACT FROM AUTHOR]

The drying shrinkage caused by rapid water loss will pose a great threat to the safety of concrete structures, especially for those serviced in an extremely dry environment. However, the majority of reported research primarily focuses on drying shrinkage behavior of concrete under relative humidity (RH) ranging from 50% to 100%, and the corresponding drying shrinkage prediction models have poor applicability, leading to an unclear understanding of it under ultra-low RH (RH<40%). In this study, a numerical model, finite element model with cohesive elements, was proposed to better predict the drying shrinkage behavior of concrete, even under the RH of 20%. Initially, geometric models of concrete with varying coarse aggregate size ranges and volume fractions were established. Subsequently, based on the relationship between RH and humidity diffusion coefficient of concrete with different water-cement ratio (w/c), a drying shrinkage model was proposed to examine the corresponding shrinkage strain of concrete even under ultra-low RH conditions, and its validity was verified through comparison with reported experimental research. Utilizing the developed model, the effects of size range and volume fraction of coarse aggregate, and the w/c, on the drying shrinkage of concrete across a wide range of RH levels were systematically investigated. The results indicated that increasing the aggregate size range can only slightly inhibit drying shrinkage, while there is a marked inhibition effect on it (decrease by about 45% at most) for increasing the aggregate vol% or reducing w/c, even under ultra-low RH of 20%. Finally, grey relation analysis was employed to assess the inhibition weight of each parameter on the drying shrinkage, following this sequence in order: RH > w/c > coarse aggregate vol% > coarse aggregate range size. We also envision that the proposed model could provide feasible guidelines for the inhibition of concrete in ultra-dry regions. • A numerical drying shrinkage prediction model was proposed. • Effect of aggregate, w/c and RH on drying shrinkage strain was studied. • The inhibition weight of each parameter on the drying shrinkage was explored. [ABSTRACT FROM AUTHOR]

The chloride ion diffusion coefficient (CIDC) is an important index to evaluate the durability performance of machined sand concrete. In this study, a database of CIDC is established for 82 sets of experimental data for 4 influence parameters. Meanwhile, different models with 4 input variables (Water to binder ratio, fly ash ratio, relative humidity and temperature) and 1 output variable (CIDC) were constructed to predict CIDC using extreme gradient boosting algorithm (XGBoost), backpropagation algorithm (BP), support vector regression algorithm (SVR), and bp algorithm optimized by Bayesian formula. Meanwhile, the SHapley Additive exPlanations (SHAP) method in python is introduced to characterize the established models to clarify the principles behind the model prediction. Based on the findings, it appears that the Bayesian-BP model is the one that performs the best in terms of making predictions. Furthermore, based on the prediction of the Bayesian-BP model, the percentage contribution of various variable factors in the model is evaluated by the SHAP analysis method, and the corresponding variables are characterized. It is found that the obtained characteristic variables with the CIDC have the same trend as the reality that verifies the reliability of the proposed optimum model. Moreover, CIDC was predicted by controlling variables, and flexible adjustments to CIDC were achieved based on this. • A machine learning-based method for chloride diffusion prediction was proposed. • The effect of different variable factors on the diffusion of chloride ions was studied. • CIDC was predicted by controlling variables, and flexible adjustments to CIDC were achieved. [ABSTRACT FROM AUTHOR]