Nov 2026· Journal of Structural Design and Construction Practice· Vol 31· 0 citations· 38 references
Abstract
Machine learning is often used to predict concrete properties, but its ability to derive fundamental insights for mix design remains underexplored. This study leverages ensemble methods to predict the split tensile strength (STS) of fiber-reinforced recycled aggregate concrete. The primary novelty of this work lies in building on the results of predictive models to decode the complex interactions governing their performance, which then leads to actionable design recommendations by employing model interpretability techniques. A dataset consisting of 257 samples, representing 11 variables, was analyzed using regression tree, boosted regression tree (BRT), and random forest tree models. The BRT model had the lowest error (root mean square error = 0.49 MPa) among these models. Interpretation of the optimal BRT model revealed that the density of recycled aggregate (RCA) had the highest impact on the model followed by the contents of water, cement, RCA, and superplasticizer (SP). These top five parameters collectively contributed to more than 90% of the prediction variance. Critically, the analysis uncovered nonlinear interactions and optimal design thresholds. The best results (STS of up to 7.7 MPa) were obtained with steel fibers at specific mix parameters, which were identified by the model as cement content (
>
400
kg
/
m
3
), RCA density
>
2,500
kg
/
m
3
, and SP dosage (4%–5%). The results of this study are expected to promote the use of durable and sustainable materials in construction by moving beyond predictive modeling to provide actionable, interpretable insights for composite mix design of concrete with multiple nontraditional materials used in tandem.
Recycled aggregate concrete (RAC) is a sustainable alternative construction material to reduce natural resource exploitation and manage construction and demolition waste. However, predicting the mechanical performance of RAC remains a challenge due to the high variability of recycled aggregate properties. The purpose of this study is to develop a machine learning model to predict the compressive strength of recycled aggregate-based concrete and compare its performance with normal concrete. The dataset used consists of 2165 samples (1600 normal concrete and 565 recycled aggregate concrete) collected from various scientific publications. Three tree-based machine learning algorithms (Random Forest, XGBoost, and LightGBM) were implemented and optimized using RandomizedSearchCV with 5-fold cross-validation. The results showed that LightGBM provided the best performance with R² = 0.92, MAE = 2.45 MPa, and RMSE = 3.52 MPa on the test set. This model is able to predict the compressive strength of normal concrete (R² = 0.92) and recycled aggregate concrete (R² = 0.91) with almost the same accuracy, indicating strong generalization. Feature importance analysis revealed that curing age, cement content, and water content are the most important factors in compressive strength prediction, while for RAC, recycled aggregate water absorption (WRCA) also makes a significant contribution. Error analysis shows that residuals are random and normally distributed without systematic bias. This model can reliably predict concrete compressive strength in the range of 20-60 MPa with an average error of ±3-4 MPa and can be integrated into mix proportioning design software to improve the efficiency of the design process and support the use of sustainable construction materials.
Suji’at, Eko Wahyu Abryandoko, Ocha Silvia Kencana et al.· Journal of Novel Engineering...· 0 citations
Abstract This study develops a robust framework for estimating the compressive strength of self-compacting concrete (SCC) incorporating recycled aggregates using supervised machine learning (ML) techniques. A comprehensive experimental database comprising 582 concrete mix designs was used, encompassing diverse input variables including binder content, water, coarse and fine aggregates, recycled aggregate proportion, superplasticizer dosage, and curing time. Seven ML algorithms—XGBoost, CatBoost, AdaBoost, Extra Trees, Bagging Regressor, K-Nearest Neighbors, and Radius Neighbors—were systematically trained using a stratified 70/15/15 data split and optimized via grid search with five-fold cross-validation. Model performance was evaluated using coefficient of determination (R 2), root mean squared error, and MAE across training, validation, and testing datasets. Among all models, XGBoost demonstrated the highest accuracy, achieving an average R 2 of 0.9799, RMSE of 2.87 MPa, and mean absolute error of 1.97 MPa. The Permutation Feature Importance analysis revealed that binder content, water, and coarse aggregate were the most influential predictors of strength. This study confirms that ensemble ML models, particularly XGBoost, can reliably predict the compressive strength of SCC with recycled aggregates, while offering transparent insights into material behavior. The results provide a valuable tool for sustainable mix design optimization and practical implementation in eco-efficient concrete construction.
A. Khan, M. D. Rasheed, Muhammad Huzaifa Naveed et al.· Data-Centric Engineering· 0 citations
This study provides a robust, interpretable, and generalizable ML framework for optimizing nano-silica concrete mix design and highlights the strong potential of ML, particularly ensemble models combined with explainable AI techniques, to improve prediction reliability, reduce trial-and-error experimentation, and support more cost-efficient and sustainable concrete design.
Yousif J. Bas, Jamal I. Kakrasul, Kamaran S. Ismail et al.· Engineering Research Express· 0 citations
An interpretable machine-learning framework for predicting the splitting strength of asphalt concrete and supporting data-driven mixture design and a GUI platform integrating prediction and SHAP-based explanation was developed to improve the accessibility and practical applicability of the proposed framework.
J. Xing, Xiao Tan, Dongzhan Jin et al.· 0 citations
Alkali-activated recycled aggregate concrete (AARAC) offers a sustainable alternative to traditional concrete but suffers from complex, non-linear mechanical behavior that challenges conventional prediction methods. This study develops and compares five machine learning models, linear regression (LR), M5P, Random Forest (RF), K-Nearest Neighbors (KNN) and XGBoost, for predicting the compressive strength (Cs), flexural strength (Fs), splitting tensile strength (Ss), pull-out bond strength (PT), and water absorption (Wa%) of AARAC. A dataset of 360 experimental samples, incorporating natural aggregate, recycled concrete aggregate (RCA), cement block aggregate (CBA), water-to-cement ratio (W/C), alkaline treatment status, and slump, was used. Models were evaluated via train/test split (80/20) and 10-fold cross-validation using R2, MAE, RMSE, and MAPE. Random Forest achieved the highest test R2 (0.8736) and lowest test MAPE (1.418%) and XGBoost (R2 = 0.8605, MAPE = 1.557%). KNN and M5P performed moderately, while LR was the weakest (R2 = 0.6958, MAPE = 2.147%). All tree-based models exhibited overfitting, with training R2 up to 0.98. Scatter plot analysis revealed systematic underprediction by RF for Cs (constant offset of ~2 MPa) and increasing bias for PT, Ss, and Wa% at higher values. XGBoost gave perfect predictions for PT and Wa% but underpredicted Cs and Fs. K-fold cross-validation confirmed XGBoost as the most robust (mean R2 = 0.9844). Correlation analysis showed W/C strongly increases Wa% (r = 0.80) and decreases PT (r = −0.73); RCA negatively affects mechanical properties, while CBA and alkaline treatment improve them. The study concludes that ensemble tree models, particularly Random Forest, are superior for AARAC prediction, but systematic bias requires post hoc calibration.
Ahmed D. Almutairi, Abd Al-Kader A. Al Sayed· Buildings· 0 citations
The growing need for sustainable construction materials has led to increasing interest in concrete made with recycled and waste-derived components such as polyethylene terephthalate (PET) and rice husk ash (RHA). Accurately predicting the mechanical and durability properties of such materials remains challenging due to the complex interactions among mix constituents. To address this, this study develops a hybrid machine learning framework that combines a TFT-inspired attention-based feature interaction module with XGBoost, optimized using the Aquila Optimizer (AO). The model is further enhanced with Bayesian uncertainty quantification through Monte Carlo Dropout and dual explainable AI techniques (SHAP and LIME) to improve transparency and reliability. The model is trained on an experimental dataset of ten concrete mix designs, which is expanded using Gaussian Noise Injection, SMOTER, and Conditional GAN-based augmentation to improve learning robustness. It predicts six key properties: compressive strength, splitting tensile strength, flexural strength, water absorption, modulus of elasticity, and rapid chloride permeability. The proposed framework achieves high predictive accuracy (R² = 0.9992, RMSE = 0.00205, MAE = 0.0009), outperforming benchmark models. Interpretation results indicate that PET content and cement proportion are the most influential variables, while the uncertainty analysis provides reliable confidence bounds for each prediction. The model is specifically applicable to concrete properties at a curing age of 90 days. The TFT component is used in a non-temporal manner to capture feature interactions in tabular data, making the framework suitable for complex material systems where interpretability and prediction reliability are essential.
Pololy Pradeep Kumar, V. Giridhar, B. Rajini et al.· Discover Civil Engineering· 1 citation