ON THE IDENTIFICATION OF RESIDUAL IMPACT PROPERTIES OF MATERIALS BY ACOUSTO-ULTRASONICS — A PATTERN RECOGNITION APPROACH

摘要

This paper deals with the application of Acousto-ultrasonics, in conjunction with Pattern Recognition and Classification techniques, to the identification of residual impact properties of a class of polymeric material, namely, Polyvinylchlo-ride (PVC). PVC specimens of different low-energy repeated impact damage states are processed by Acousto-ultrasonics (AU) to retrieve AU signals in the form of dig-italized records. These AU signals are grouped as distinct classes, each pertaining to a known level of repeated impact damage. Describing features of these AU signals are used to build Pattern Recognition (PR) Classifiers. These classifiers are used to identify unknown damage states in other PVC specimens by classifying the retrieved AU signals as belonging to one of the classes. The obtained results indicate that Acousto-ultrasonics in combination with Pattern Recognition and Classification techniques can be used for the quantitative non-destructive identification of damage states in PVC specimens of unknown low-energy repeated impact conditions.