ANN approach to evaluate the effects of supplementary cementitious materials on the compressive strength of recycled aggregate concrete

Abstract

The growing awareness of environmental issues has triggered significant research into the impacts generated on the environment during concrete production. To improve sustainability in this process, ordinary Portland cement (OPC) and natural aggregates (NA) have been replaced by recycled concrete aggregates (RCAs) and supplementary cementitious materials (SCMs) to reduce the negative issues to the environment. However, the use of SCMs can be challenging due to their complex chemical composition, and RCAs are often of inferior quality compared to NAs, leading to reduced concrete strength. To address these issues, an artificial neural networks (ANN) model has been used for calculating the compressive strength of concrete mixes that oodórate SCMs and RCAs. To train the model, 1181 concrete mixtures from 116 sources were included, considering parameters such as hydraulic reactivity, silica, and alumina modulus. In this research was found that the ANN model is an effective instrument to predict concrete strength even with the use of SCMs and RCAs. The analysis of input variables using the Connection Weight Approach (CWA) concludes that silica fume is the external material that, when mixed with RCA, significantly improves the resistance of concrete, followed by the reactivity modulus of cementitious materials, cement content, silica modulus, fine NA content, and coarse NA dosage. Conversely, excessive ood dosage, high ood/cement ratio, and excessive RCA content (fine and coarse) negatively affect concrete strength.