Assessment of Scaling Durability of Concrete with CFBC Ash by Automatic Classification Rules

Abstract

The objective of this investigation was to develop rules for automatic assessment of concrete quality by using selected artificial intelligence methods based on machine learning. The range of tested materials included concrete containing nonstandard waste material—the solid residue from coal combustion in circulating fluidized bed combustion boilers (CFBC ash) used as an additive. Performed experimental tests on the surface scaling resistance provided data for learning and verification of rules discovered by machine learning techniques. It has been found that machine learning is a tool that can be applied to classify concrete durability. The rules generated by computer programs AQ21 and WEKA by using the J48 algorithm provided a means for adequate categorization of plain concrete and concrete modified with CFBC fly ash as materials resistant or not resistant to the surface scaling.