Mathematical modeling techniques to predict the compressive strength of high-strength concrete incorporated metakaolin with multiple mix proportions

Abstract

Many environmental and health problems are raised from cement manufacturing processes and other factories because those factories emit large amounts of carbon dioxide (CO2) into the atmosphere. Using supplementary cementitious material reduces the amount of cement used, then the amount of CO2 is also reduced. Metakaolin (MK) is used as a mineral admixture to improve the mechanical properties of concrete because higher strength requires a more densified concrete mixture. Mineral admixtures have a pozzolanic reaction and lower thickness of the transition zone. In this study, four different models, specifically linear regression (LR), nonlinear regression (NLR), multi-logistic regression (MLR), and M5P-tree models, are proposed to predict the compressive strength (compressive strength) of high-strength concrete (HSC) modified with metakaolin admixture. These models are useful since they serve as an effective way to predict compressive strength without waiting for a long time theoretically and are cost-effective and time-saving. For this purpose, 197 experimental data of HSC mixtures with MK from previous research are collected, statistically analyzed, and used to propose the prediction models. The curing time of the tested sample ranged between 1 and 360 days, and water to binder ratios was from 100-206 kg/m3. The highest compressive strength of the HSC samples was 105 MPa at 28 days. The proposed models are assessed by the coefficient of determination (R2), root mean squared error (RMSE), scatter index (SI). And OBJ. The M5P-tree model was the best model among the proposed models with R2, RMSE, SI of 0.77, 6.67 MPa, 0.096, respectively.

Keywords

• Compressive strength
• High-strength concrete
• Metakaolin
• Mineral admixtures
• Modeling
• Sensitivity analysis