Prediction of mechanical and durability characteristics of concrete including slag and recycled aggregate concrete with artificial neural networks (ANNs)

This research aims to create an artificial neural networks (ANN) model to forecast mechanical and durability properties (compressive strengths and the durability test of rapid chloride migration test (RCMT)) of waste included concrete according to the features of the reference samples. Moreover, this experimental study evaluated the mechanical characteristics and long-term durability of concrete when using a synergic of ground granulated blast furnace slag (GGBFS) and recycled concrete aggregate. This experimental study evaluated concrete's mechanical characteristics and long-term durability when using a synergic of ground granulated blast furnace slag (GGBFS) and recycled concrete aggregate. Concrete samples were made with different binder content and water to binder ratios, including the RCA at substitution levels of 25 percent and 50 percent as well as slag at substitution levels of 20 percent and 40 percent by weight of Portland cement in order to evaluate the influence of waste materials on concrete mixtures. The mechanical characteristics of concrete deteriorated when the quantities of recycled aggregate increased in all replacement ratios. However, the findings of the mechanical characteristics suggest that slag can reduce the detrimental effects of substituting coarse aggregate with RCA in the early ages of the samples. In the case of durability, applying RCA increases the chloride ion migration rate in concrete. Slag mitigates the detrimental impacts of recycled aggregate by enhancing hydration products, except for early stages.