AI-Assisted optimisation of green concrete mixes incorporating recycled concrete aggregates

Abstract

Maximising the content of supplementary cementitious materials as a partial replacement for Portland cement and using recycled concrete aggregates as a full or partial replacement for coarse aggregates are among the widely adopted strategies in the design of green concrete. However, the synergetic implications of adopting both approaches on concrete properties add complexity to the design of green concrete mixes. This paper proposes a novel AI-based multi-objective optimisation framework addressing the lack of a systematic method to formulate these concretes and solves the common multi-objective mix design problems. The framework is developed for the mix proportioning of green concrete mixes containing recycled concrete aggregates and commonly used supplementary cementitious materials while considering the desired compressive strength. The proposed optimisation model leverages a dataset of about 2,120 mixes and employs an extreme gradient boosting machine to model the compressive strength requirements as a mechanical constraint throughout the optimsiation process. Moreover, a set of key environmental objectives, including minimising global warming, acidification, fossil fuel depletion potentials, and the cost of production, are targeted. The effectiveness of the framework is validated through case studies across different strength ranges. The results indicate a significant improvement in environmental sustainability indicators, including a 19.6% and 21.6% reduction in global warming and acidification potentials, besides considerable cost savings of up to 23.7% without compromising the mechanical performance of concrete.