Identification of the Interlayer Bond Between Repair Overlay and Concrete Using Nondestructive Testing, an Artificial Neural Network and Principal Component Analysis

Abstract

In construction practice, concrete elements are exposed to adverse environmental influences, and therefore sooner or later require repair. This repair is usually performed by removing the damaged concrete and replacing it with repair overlay. The quality of this repair is evaluated using the destructive pull-off method. In this method, the pull-off adhesion value between the repair overlay and repaired element is measured (fb). Unfortunately, the disadvantage of this method is local damage of the element at every measuring point. It is therefore reasonable to present a reliable nondestructive method of identifying the interlayer pull-off adhesion value. The article presents the results of experimental research, which indicate that such identification is possible using complementary nondestructive methods and an artificial neural network with principal component analysis. © 2019 The Authors. Published by Budapest University of Technology and Economics & Diamond Congress Ltd. Peer-review under responsibility of the scientific committee of the Creative Construction Conference 2019.