Sand/Cement ratio evaluation on mortar using neural network and ultrasonic transmission inspection

Abstract

The quality and degradation state of building materials can be determined by means of nondestructive testing (NDT). These materials are composed by a cementitious matrix and several aggregates, generally sand and gravel. Sand/cement ratio (s/c) provides the final material quality; however, the sand content can mask the matrix properties in a nondestructive measurement. This work presents an ultrasonic evaluation of the mortar sand content. This evaluation is carried out with several mortar probes which have been varied both, in the quality of the cementitious matrix, and in the s/c ratio. This ratio has been determined by the ultrasonic transmission inspection information with different transducers. Statistical principal component analysis (PCA) in order to reduce the dimension of the captured traces is applied. Feed-forward neural networks have been trained using a reduced PCA and the outputs of the network allow displaying a false color map that shows the mortar probes (s/c) distribution.