Cluster-based identification algorithm for in-line recycled concrete aggregates characterization using Laser-Induced Breakdown Spectroscopy (LIBS)

More Info
expand_more

Abstract

To upcycle End-of-Life (EoL) concrete from demolished buildings, it is essential to efficiently identify the different materials that may contaminate it. The precise identification and classification of materials and contaminants are vital processes for in-line quality inspection of recycled concrete aggregates transported on a conveyor belt. In this study, a total of eight potential contaminants are considered as target contaminant materials in the streams made of coarse and fine aggregates resulting from the upcycling of EoL concrete. These contaminants degrade the quality of the aggregates even at low concentrations, so it is essential to identify the presence of such contaminants along with the main products of recycling which are recycled coarse aggregates (RCA) and recycled fine aggregates (RFA). An efficient method is proposed to identify and classify EoL concrete waste along with RCA and RFA in motion on conveyor belts via laser-induced breakdown spectroscopy (LIBS) coupled with a cluster-based identification algorithm. The model is verified with an accuracy of 0.97, a precision (weighted average) of 0.98, a recall (weighted average) of 0.97, and an F1-score (weighted average) of 0.98 for the validation set, under the optimal conditions. This study suggests that LIBS may be well suited for fast and in-line analysis of recycled concrete aggregates in industrial applications. This approach presents an innovative approach for the quality characterization of secondary materials produced from EoL concrete being transported on conveyor belts, and therefore can be of great value for the processing and high-end utilization of EoL concrete.