Theoretical and Experimental Study of Composite Materials Using Longitudinal Ultrasonic Spectroscopy

Abstract

Composite materials are nowadays involved in various designs [1] in areas like aero- nautics, road, rail or sea transportation, and chemical industries, among other. Using these material brings a number of advantages : the lowering of overall mass, the increase in the applied mechanical loads before breakdown occurs, the resistance to contamination by chemical agents, the ease of material forming using molding techniques, and the matcha- bility of the material to the particular mechanical function required for a given part in a structure. Among composite materials, the carbon-epoxy ones are costly but also very ef- ficient. They find important applications in aeronautics and we will focus our attention here to them. Like for any machined product, defects are induced during the realization of compo- site material parts. However, the nature of these defects is very specific (delamination, porosities ... ) when compared to metals. The heterogeneous and anisotropic nature [2,3] of these materials make them very difficult to control and specialized testing procedures must be elaborated for this purpose. Specifically, ultrasonic techniques appear well suited for the detection of bulk defects [4]. The classical ultrasonic techniques rely on the obser- vation of the time diagram of the echoes reflected on or transmitted through the defects in the sample under study. The heterogeneous structure of composite materials may however induce extraneous acoustic noise [5] and make the interpretation of the previous echo- gram very questionnable and often impossible. Spectral analysis [6] appears then as a complementary tool useful for overcoming these difficulties. The study reported in this paper will make use of the two techniques, all the tests being performed with bulk longi- tudinal ultrasonic waves impiging under normal incidence on unidirectionnal or crossed- ply carbon-epoxy composites.