Air-coupled ultrasonic testing of diffusion bonds

Abstract

The diffusion bond between two steel plates can be ultrasonically evaluated, at normal incidence in an immersion experiment, by analyzing the frequency dependence of the echo reflected from the imperfect bond. The interfacial stiffness, derived from the echo amplitude, correlates well with the bond-strength. However, a non-contact method is desirable for applications where immersion or contact is not wanted or even dangerous for damaging the material. This above mentioned bond-echo technique would not work in the situation of air-coupling as the reflected echo becomes then too weak due to the high impedance mismatch at the air–solid interface. Therefore we propose a theoretical method based on the study of two neighbouring resonance frequencies of the diffusion bonded plate–plate structure. In this way the physical signal sensitive to the adhesion status is not the (too weak) echo reflected from the bond, but the resonance frequency of the whole plate–plate system, and this frequency is detectable as working at resonance ensures high enough signal levels. It was shown that the odd resonance is as well sensitive to the plate thickness as to the interfacial bond parameter, whereas the even resonance feels only the plate thickness. On the basis of a theoretical formula, it is possible to extract, from a single point measurement, out of these two resonance frequencies both the plate thickness and the interfacial stiffness. In this way bond information is separated from geometrical information. Finally it is shown that thickness differences between the plates did not affect the reliability of the bonding-strength predictions.