On the characterization of impact properties of engineering materials: An acousto-ultrasonics pattern recognition approach.

Abstract

This thesis describes the experimental work carried out at the University of Ottawa for the characterization of the residual impact properties of polymeric material systems; namely, Polyvinylchloride and graphite/polycarbonate composite. In this context, the nondestructive Acousto-Ultrasonics (AU) technique in conjunction with Pattern Recognition and Classification methods were used. The material specimens, which present different levels of controlled low-energy repeated impact damage, were also evaluated destructively by Falling Weight Impact and Uniaxial Tensile tests. The results of these destructive tests were used to correlate and verify the AU measurements. A review of nondestructive evaluation methods, with a focus on the Acousto-Ultrasonics technique used in this thesis is presented. Analysis techniques of AU data are introduced. The Pattern Recognition and Classification methodology is reviewed for its application to identify AU signals. Fundamentals of the Falling Weight Impact Test used are also presented. The values of the Mean Failure Energy for the samples of the mentioned two materials, as obtained from Falling Weight Impact Test, were found to correlate with the damage states of these samples, as measured by the applied number of controlled repeated-impacts. Also, the values of the Acousto Ultrasonics Parameter, calculated from the retrieved AU signals, were found to correlate with the Mean Failure Energy and Ultimate Tensile Strength values that were obtained from the destructive tests. Identification of the AU signals pertaining to the considered levels of impact damage was performed by means of Pattern Recognition and Classification techniques. The results obtained indicate that a characterization of low-energy repeated impact conditions in the mentioned two polymeric materials can be based on measurements of the Mean Failure Energy, obtained from Falling Weight Impact testing of samples of specimens. Further, the predictions of the residual impact properties of the two materials can be made by Acousto Ultrasonics, in conjunction with Pattern Recognition and Classification techniques.