Studies on Green Concrete Pavement: Experimental and SVM Modelling

Abstract

In this study, the thickness of the rigid pavement slab is optimized by replacing cement with Ground Granulated Blast Furnace Slag (GGBFS) and Polypropylene (PP) fibers. In this regard, the flexural strength of concrete samples is studied with 40, and 50% GGBFS and 0.2 to 0.6 % PP-fibers. From the results, it is observed that the flexural strength of concrete with 50% GGBFS yields higher strength and with PP fibers improves the crack resistance. Later from the obtained experimental data, the support vector machines (SVM) with different kernel functions are developed to predict the flexural strength of concrete. The SVM model shows a good correlation with the experimental data regarding the correlation coefficient (CC) and root mean square error (RMSE). SVM with a spline kernel yields good prediction for both flexural strength of concrete with CC of 0.9996 and 0.9756, RMSE of 0.0070 and 0.0554 for training and testing respectively. From the results, it can be concluded that GGBFS and PP fiber could be used as alternative materials to reduce the carbon footprint and SVM as an alternative tool to predict the flexural strength of concrete.