Application of Machine Learning for Prediction of Mechanical Properties of Mortars and Concretes

Abstract

This paper deals with the application of machine learning (ML) in the field of concrete technology. Two databases of test mortars and concretes were created from selected academic theses, which include mechanical properties in relation to their composition. These databases were used to develop two ML models that predict the mechanical properties of mortars and concretes depending on their composition. The mortar test database contains a total of 242 mechanical property records and the concrete test database contains 111 records. The materials in the database are CEM I, CEM II and CEM III cements combined with additives such as ground granulated blast furnace slag, high temperature fly ash and micro-ground limestone.