INTERPRETING CONCRETE COMPRESSIVE STRENGTH THROUGH MIX PROPORTIONS AND CURING AGE
DOI:
https://doi.org/10.69980/645dt871Keywords:
Concrete compressive strength, Mix proportions, Curing age, Regression analysis, Grouped cross-validationAbstract
The relationship between the compressive strength of concrete and the proportion of the mix and curing age of the concrete allows for understanding and a transparent interpretation of the concrete performance in the laboratory. These relationships were investigated in this study through experimental concrete data with 188 observations. Three repeats were removed and 185 observations were arranged into 47 operational mix groups. Descriptive analysis, correlation assessment, multiple linear regression and ridge regression techniques have been used and the degree of predictiveness assessed by grouped cross validation. The positive adjusted associations of total amount of binder with strength, and of the log of curing age with strength, and the negative associations of water/binder ratio and fly ash replacement fraction with strength were significant. The confidence intervals for GGBS, metakaolin and recycled aggregate fractions were found to contain zero. These values were 0.843 for ordinary regression, and 0.844 for ridge regression in the validation set, respectively. Logarithmic age gave better prediction than did untransformed age; there was no improvement provided by the tested composition–age interaction. Similar predictive outcomes were generated using sensitivity analyses. These results justify the use of interpretable regression for estimating the strength variation under specific conditions in the laboratory. However, since the predictions are dependent on one another, and the composition is limited, there is a lack of interpretability until it is separately validated prior to general use.
References
[1]. Abbas, Y. M., & Alsaif, A. (2025). Explainable data-driven modeling for optimized mix design of 3D-Printed concrete: interpreting nonlinear synergies among binder components and proportions. CMES-Computer Modeling in Engineering and Sciences, 145(2), 1789-1819.
[2]. Adamu, M., Jibril, M. M., Gora, A. U. M., Ibrahim, Y. E., & Alanazi, H. (2026). Exploring Variable Influences on the Compressive Strength of Alkali-Activated Concrete Using Ensemble Tree, Deep Learning Methods and SHAP-Based Interpretation. Eng, 7(5), 192.
[3]. Adiguzel Tuylu, A. N., Tuylu, S., Adiguzel, D., & Demir, I. (2026). Quantifying the Conditional Contribution of Cement Content to Concrete Strength Using Interpretable Causal Machine Learning. Buildings, 16(5), 1059.
[4]. Akber, M. Z. (2024). Improving the experience of machine learning in compressive strength prediction of industrial concrete considering mixing proportions, engineered ratios and atmospheric features. Construction and Building Materials, 444, 137884.
[5]. Badeleh, A., Seyedkazemi, A., Rahimi, S., Soleimani Kutanaei, S., & Malidarreh, N. R. (2026). GMDH-based comparative modeling and reliability assessment provide interpretable prediction of compressive strength in natural-zeolite concrete. Scientific Reports.
[6]. Biswal, U. S., Pasla, D., & Mishra, M. (2022). Experimental data set for concrete compressive strength prediction from IIT Bhubaneswar concrete laboratory (Version 2) [Data set]. Mendeley Data. https://doi.org/10.17632/5wkxzmzwnz.2
[7]. Chen, X., Chen, N., Cheng, S., Li, N., Wu, Q., Liu, Z., ... & Zhang, Z. (2026). Machine learning models for predicting the compressive strength of sustainable recycled aggregate concrete incorporating supplementary cementitious materials. Sustainable Materials and Technologies, e01907.
[8]. Ezz, H., & Bakr, S. M. (2026). Comparative evaluation of machine learning and deep learning approaches for compressive strength prediction of geopolymer concrete. Scientific Reports, 16(1), 14652.
[9]. George, S., Hussain, S. A., Alamiery, A. A., Pathan, M. G., Sabihuddin, S., Thakur, N., ... & Smerat, A. (2026). Enhancing compressive strength prediction of sustainable concrete using MEP and GEP models with SHAP-based interpretation. Asian Journal of Civil Engineering, 27(1), 245-262.
[10]. Ghrici, A. A., Benzaamia, A., Mezzoudj, F., Medjahed, C., & Ghrici, M. (2025). SHAP-enhanced tree-based regression for predicting the compressive strength of high performance concrete. Construction and Building Materials, 495, 143602.
[11]. Guo, J., Wang, K., Zhang, P., & Xu, H. (2023). Effect of internal curing on early-age properties of concrete under simulative natural environment in arid regions. Construction and Building Materials, 362, 129697.
[12]. Hamed, A. K., Alsaadawi, M. M., & Elshaarawy, M. K. (2026). Interpretable deep neural network deployment for concrete compressive strength prediction. Journal of Building Pathology and Rehabilitation, 11(2), 98.
[13]. Hao, Y., Wei, S., Guan, J., He, S., Xie, X., & He, W. (2026). Interpretable data-driven prediction of UHPC compressive strength: A GWO-BP hybrid model with SHAP analysis and experimental validation. Developments in the Built Environment, 101012.
[14]. Hari, R., & Zhuge, Y. (2024). Performance assessment of pervious concrete incorporated with calcium silicate hydrates (CSH) cultivated on rice husk ash substrates–A trend surface analysis interpretation. Construction and Building Materials, 446, 138050.
[15]. Kabbo, M. K. I., Sobuz, M. H. R., Khatun, M., Ghalla, M., Jameel, M., Hasan, N. M. S., & Abubakar, S. A. (2025). Analyzing the influence of chemical components of incinerated bottom ash on compressive strength of magnesium phosphate cement using machine learning analysis. Geoenvironmental Disasters, 12(1), 35.
[16]. Kassa, S. M., Kacprzak, G., Wubineh, B. Z., Demlew, M. B., & Sewunet, S. D. (2026). Explainable ensemble machine learning for reliable concrete compressive strength prediction: a validation and uncertainty assessment framework. Scientific Reports.
[17]. Liu, G., & Sun, B. (2023). Concrete compressive strength prediction using an explainable boosting machine model. Case Studies in Construction Materials, 18, e01845.
[18]. Liu, Q., Qian, C., Xu, W., & Li, B. (2026). Study of Curing Regime and Mix Proportion Optimization for Precast High-Strength Concrete Considering Temperature Difference and Early-Age Strength. Journal of Building Engineering, 115911.
[19]. Lv, Q., Zhang, J., Zhang, L., Zhao, H., & Ren, J. (2025). Machine learning-based optimization of concrete strength using interpretable models. Materials Today Communications, 47, 112872.
[20]. Nikoopayan Tak, M. S., Feng, Y., & Mahgoub, M. (2025). Advanced machine learning techniques for predicting concrete compressive strength. Infrastructures, 10(2), 26.
[21]. Palanisamy, A. K., Jegatheeswaran, D., Ananth, C., Bhogayata, A., Juneja, B., Buribay, R., & Mezgebe, T. T. (2025). Prediction of early age compressive strength of concrete using machine learning. Scientific Reports, 15(1), 45293.
[22]. Peng, X., Zhuang, Z., & Yang, Q. (2022). Predictive modeling of compressive strength for concrete at super early age. Materials, 15(14), 4914.
[23]. Saxena, A., Sabillon-Orellana, C., & Prozzi, J. (2024). Prediction of compressive strength in sustainable concrete using regression analysis. Journal of Material Cycles and Waste Management, 26(5), 2896-2909.
[24]. Sharma, S., Burman, A., & Samui, P. (2026). Data-driven prediction of ultra-high-performance concrete compressive strength using ensemble and neural network models. Asian Journal of Civil Engineering, 27(5), 2189-2206.
[25]. Wang, Y., Jia, Y., Wang, C., He, W., Ding, Q., Wang, F., ... & Fang, K. (2025). Interpretation of dominant features governing compressive strength in one-part geopolymer. Buildings, 15(20), 3661.
