Use of acoustic emission data clustering to identify damage mode in concrete structures

Abstract

Acoustic Emission (AE) technique gained increasing interest in the last two decades as monitoring methodology and as assessment tool for safety and reliability evaluation of reinforced concrete structures, historical and masonry buildings. However a widely accepted analytical instrument for AE data handling and interpretation is still missing. Cluster and discriminant analysis have been recently applied to classify AE patterns and to identify damage modes. Aim of this paper was to develop a cluster analysis procedure devoted to identify cracking mechanisms in concrete structures. Unsupervised methods, k-means as well as Principal Component Analysis and Self Organizing Map, have been used as analytical instruments. A procedure aimed to remove environmental AE noise has been also proposed. © RILEM 2013.