Experimental investigation and prediction of strength development of GGBFS‑, LFS‑ and SCBA‑based green concrete using soft computing techniques

Abstract

The present research article is focusing on the utilization of ground granulated blast furnace slag (GGBFS), ladle furnace slag (LFS) and sugarcane bagasse ash (SCBA) as the partial substitution of ordinary Portland cement (OPC) in concrete mix in order to create a sustainable environment and enhance the engineering performance. The primary purpose of the study is to predict the optimum percentage of different additives in varying proportions, i.e. individually and in a combined manner also to construct a sustainable rigid pavement. Therefore, accurate estimation of strength evaluation is required to minimize the effort. In the current investigation, different modeling analysis techniques have been attempted with different soft computing tools, namely random forest (RF), random tree (RT), M5P, reduced-error pruning (REP) tree and linear regression (LR) to estimate the compressive strength of the concrete using the experimental data. In the present study, all the mentioned additives were added in the concrete mix as the replacement of OPC up to 35%. On the basis of the findings, it was observed that 20% of all the additives in individual form might be used as the partial substitute of OPC. While, in a combined form, concrete mix having 5% GGBFS, 10% LFS and 15% of SCBA was showing the optimum strength value. However, it was also observed that the greater percentage of all the additives can be utilized with an increment in the curing time period. RF approach was found most permissible approach to predict the strength value for green concrete as it was exhibiting higher value of coefficient of correlation, low value of mean absolute error and the root mean square error as revealed by outcomes of the models and statistical assessments methods. Sensitivity analysis is carried out and found that the curing time in days is the utmost effective input variable for estimating the compressive strength of concrete using this data set.