Prediction of self-healing characteristics of GGBS admixed concrete using Artificial Neural Network

Abstract Concrete has become a significant part of our lives; the utilization of concrete is increasing at a high rate. One of the most constituents of concrete is Ordinary Portland Cement (OPC). The manufacturing process of OPC leads to the emission of huge amounts of CO 2 . Thus the researchers have started finding alternatives for the replacement of cement. The primary objective of this paper is to investigate the behavior of M40 grade concrete when partially replaced with Ground Granulated Blast Furnace slag (GGBS) at the same time using SAP and study the self-healing behavior of partially replacement concrete. In the self-healing process, the healing agent absorbs the moisture content within the atmosphere to heal the crack. Superabsorbent Polymers (SAPs) are materials that will absorb and retain an oversized volume of water and aqueous solutions. In this investigation, 51 samples of cubes are prepared for compressive strength test and self-healing test, the specimen is pre-cracked on the 28th day for healing purposes. Further, this article aims to predict the self-healing characteristics of the M40 grade of concrete using Neural Networks by incorporating different proportions of GGBS (0%, 40% and 60%) and SAPs. The predicted results obtained from the ANN model were in good agreement with experimental values.