A Data-Driven Investigation of Factors Influencing Compressive Strength in Ultra-High-Performance Concrete
DOI:
https://doi.org/10.69980/spp2aa17Keywords:
Ultra-high-performance concrete, compressive strength,, Random ForestAbstract
Ultra-high-performance concrete (UHPC) exhibits exceptional mechanical performance, durability, and long-term serviceability; however, its compressive strength is governed by numerous interacting mixture-design and curing variables. Conventional experimental optimization is often time-consuming and resource-intensive, creating a need for interpretable data-driven approaches capable of supporting material design and performance prediction. This study investigated the factors influencing UHPC compressive strength and developed a predictive framework for identifying the dominant variables affecting strength development. A cleaned secondary dataset containing 626 UHPC mixtures, 24 predictor variables, and one compressive-strength target variable was analyzed. Exploratory analysis and Pearson correlation were used to examine variable relationships. A Random Forest regression model was developed using an 80:20 train-test split and evaluated using the coefficient of determination (R²), mean absolute error (MAE), root mean square error (RMSE), and five-fold cross-validation. Feature importance and partial dependence analyses were subsequently employed to interpret the contribution of individual variables. The Random Forest model achieved strong predictive performance, with a test R² of 0.9041, MAE of 7.79 MPa, RMSE of 11.09 MPa, and a mean five-fold cross-validation R² of 0.9178. Curing time emerged as the most influential variable, contributing 57.4% of the total feature importance. Sand content showed the strongest negative influence, while silica fume content, cement content, steel fiber content, and superplasticizer content also contributed to prediction accuracy. The findings demonstrate that interpretable machine learning can accurately predict UHPC compressive strength while revealing the relative influence of key mixture-design variables, providing useful guidance for future UHPC design and optimization.
References
1. Abdellatief, M., Abd Elrahman, M., Abadel, A. A., Wasim, M., & Tahwia, A. (2023). Ultra-high performance concrete versus ultra-high performance geopolymer concrete: mechanical performance, microstructure, and ecological assessment. Journal of Building Engineering, 79, 107835.
2. Bahmani, H., & Mostofinejad, D. (2022). Microstructure of ultra-high-performance concrete (UHPC)–a review study. Journal of Building Engineering, 50, 104118.
3. Beskopylny, A. N., Stel’makh, S. A., Shcherban’, E. M., Mailyan, L. R., Meskhi, B., Razveeva, I., ... & Beskopylny, N. (2022). Concrete strength prediction using machine learning methods CatBoost, k-nearest neighbors, support vector regression. Applied Sciences, 12(21), 10864.
4. Bu, Y., Cai, Z., Peng, J., Yu, Y., & Xu, J. (2025). Data-driven compressive strength estimation and mix optimization for ultra-high-performance concrete. Journal of Building Engineering, 114253.
5. Feng, D. C., Liu, Z. T., Wang, X. D., Chen, Y., Chang, J. Q., Wei, D. F., & Jiang, Z. M. (2020). Machine learning-based compressive strength prediction for concrete: An adaptive boosting approach. Construction and Building Materials, 230, 117000.
6. Huang, J., He, G., Liao, Z., & Hou, M. (2026). Machine Learning-Based Strength Prediction of Fiber-Reinforced UHPC: A Data-Driven Framework with Feature Engineering and Uncertainty Quantification. Symmetry, 18(5), 710.
7. Katlav, M., & Ergen, F. (2024, January). Data-driven moment-carrying capacity prediction of hybrid beams consisting of UHPC-NSC using machine learning-based models. In Structures (Vol. 59, p. 105733). Elsevier.
8. Kumar, R., Rai, B., & Samui, P. (2025). Prediction of mechanical properties of high‐performance concrete and ultrahigh‐performance concrete using soft computing techniques: a critical review. Structural Concrete, 26(2), 1309-1337.
9. Li, J., Wu, Z., Shi, C., Yuan, Q., & Zhang, Z. (2020). Durability of ultra-high performance concrete–A review. Construction and Building Materials, 255, 119296.
10. Li, Y., Shen, J., Li, Y., Wang, K., & Lin, H. (2024). The data-driven research on the autogenous shrinkage of ultra-high performance concrete (UHPC) based on machine learning. Journal of Building Engineering, 82, 108373.
11. Mahjoubi, S. (2021). The key material properties of ultra-high-performance concrete (UHPC) (Version 1) [Data set]. Mendeley Data. https://doi.org/10.17632/DD62D5HYZR.1
12. Mahmoodzadeh, A., Kewalramani, M., Alghamdi, A., Ahmed, A., Alsubai, S., Alqahtani, A., ... & Palani, S. (2025). Machine learning-based prediction of crack mouth opening displacement in ultra-high-performance concrete. Scientific Reports, 15(1), 39930.
13. Mohaisen, K. O., Ahmad, S., Adekunle, S. K., Maslehuddin, M., & Al-Dulaijan, S. U. (2023). Effect of Curing Methods on the Performance of UHPC. Arabian Journal for Science and Engineering, 48(10), 13791-13805.
14. Mu, X., Zhang, S., Xu, D., Li, Y., Ni, W., & Jin, F. Influence of Water-to-Binder Ratio on the Properties and Hydration of Solid Waste-Based Ultra-High Performance Concrete. Available at SSRN 6447283.
15. Park, S., Wu, S., Liu, Z., & Pyo, S. (2021). The role of supplementary cementitious materials (SCMs) in ultra high performance concrete (UHPC): A review. Materials, 14(6), 1472.
16. Rong, H., Sun, W., Ma, H., Luo, M., You, Z., Zhang, G., ... & Gómez-Zamorano, L. Y. (2025). Machine Learning-Driven Strength Prediction and Sustainability Analysis of Ultra-High-Performance Concrete. Materials, 18(22), 5116.
17. Sun, C., Wang, K., Liu, Q., Wang, P., & Pan, F. (2023). Machine-learning-based comprehensive properties prediction and mixture design optimization of ultra-high-performance concrete. Sustainability, 15(21), 15338.
18. Tabani, A., & Biswas, R. (2025). Assessment of compressive strength of ultra‐high‐performance concrete using advanced machine learning models. Structural Concrete, 26(6), 7269-7283.
19. Teoh, K. B., Hiew, S. Y., Banthia, N., & Yoo, D. Y. (2026). Machine Learning in Ultra-High-Performance Concrete (UHPC): A Multiscale Review of Applications, Challenges, and Prospects. Archives of Computational Methods in Engineering, 1-56.
20. Wang, Z., Sun, Z., Yin, H., Liu, X., Wang, J., Zhao, H., ... & Yu, X. F. (2022). Data‐driven materials innovation and applications. Advanced Materials, 34(36), 2104113.
21. Xue, J., Briseghella, B., Huang, F., Nuti, C., Tabatabai, H., & Chen, B. (2020). Review of ultra-high performance concrete and its application in bridge engineering. Construction and Building Materials, 260, 119844.
22. Zhou, Z., Chakma, J., Hoque, M. A., Chakma, V., & Ahmed, A. (2025). Prediction of UHPC mechanical properties using optimized hybrid machine learning model with robust sensitivity and uncertainty analysis. Materials Research Express, 12(8), 085703.
23. Zuo, Z., Zhang, J., Li, B., Shen, C., Xin, G., & Chen, X. (2022). Effect of curing regime on the mechanical strength, hydration, and microstructure of ecological ultrahigh-performance concrete (EUHPC). Materials, 15(5), 1668
