Neural prediction of the pull-off adhesion of the concrete layers in floors on the basis of nondestructive tests

Abstract

This paper presents the application of artificial neural networks to the identification of the pull-off adhesion of the concrete layers in floors on the basis of parameters evaluated on the structural layer surface by the nondestructive optical method and on the floor surface by the nondestructive of impulse-response and impact-echo acoustic methods. The tests were carried out on specially prepared model specimens of concrete floor. In order to vary pull-off adhesion in the specimens the surface of the base layer was prepared in four versions. The aim of the investigations was the neural prediction of the pull-off adhesion of the concrete layers in floors on the basis of nondestructive tests. This is of practical importance since in this way the condition of the surface of the tested floor is not impaired.