Preliminary Study of the Properties of Recycled Concrete Aggregates for Future Machine Learning Algorithms

One of the most widely used materials in the world of construction is concrete. By analyzing the time phases of this material, we can identify two critical periods: the production step and the demolition step. The overuse of raw materials can cause immoderate exploitation of the subsoil which has a negative impact on the environment. Simultaneously, demolished concrete could be seen as simply waste or as a real source. For these reasons, the use of recycled concrete aggregates (RCA) may offer a considerable advantage in environmental conservation. There are numerous benefits to be achieved; the most pertinent are the reduction of raw material extraction and the consciousness of recycling resources, but the use of Natural Aggregates is still preferable because RCA is too variable and this lead to impact concrete mix performance. Since the goal is to use RCA as a new aggregate in concrete mixes, it is essential to find an effective method to define its properties and correlate them with their potential performance in concrete mixes. This study begins this process by examining the RCA essential characteristics, particularly the physical features and the chemical composition. Analyzing the difficulties that lead to the definition of the unique properties of the RCA material. The study further proposes to develop an artificial neural network to recognize RCA features from porosity, chemical composition, and shape of aggregates.