Experimental Validation of UTDefect: Scattering in Anisotropic Media and Near-field Behavior

Theoretical models that simulate measurements of ultrasonic waves undergoing scattering by material defects have been developed by Prof. Bostroem and co-workers at Chalmers Univ. of Tech. for a variety of experimental configurations and defects. A software program named UTDefect has been developed at the same time, which gathers the theoretical results obtained so far in a single package. A discussion of the motivations behind such an effort and details concerning UTDefect can be found in articles by Bostroem. Following an initial effort to validate some of the theoretical predictions available at the time, the present project has been conceived as a support to the on-going theoretical work. In fact, the goal of the project described in this report has been the experimental validation of two aspects of the above theory that have not yet been tested: the scattering of a finite ultrasonic beam by a surface-breaking crack in an anisotropic medium, and an improved model of the behaviour of a finite ultrasonic beam in the near-field region of the source. In the last case, the supporting medium is supposed to be isotropic. To carry out the first task, a single crystal, silicon sample was employed. A surface-breaking notch with a depth of approximately 1.8 mm was introduced by means of a wire-cutting saw to simulate a scattering defect. Two kinds of measurements were performed of this sample. The first one considered the signal amplitude as a function of the transducer position. To this end, three wedges generating beams propagating in different directions were used. The second series of measurements concerned the frequency content of the backscattered signals at the position where the amplitude was maximum. All three wedges mentioned above were used also in this part of the work. The experimental results were compared to the values of the physical quantities of interest as predicted by UTDefect, with the only difference that UTDefect was run for a sub-surface rectangular crack instead of a strip-like, surface-breaking crack. Though differences between the theoretical predictions and the experimental results were found, they could be explained in terms of the high sensitivity of the measured quantities to the uncertainties associated to the orientation of the crystallographic axes of the material. Such a high sensitivity of the scattered waves to the axes orientation cast reasonable doubt on whether the position of crack can be accurately deduced from ultrasonic measurements if the anisotropic structure of the material is not known within a very high degree of precision. To carry out the second task, a series of 19 steel samples of different thickness were prepared. The thickness range and transducer frequency were chosen so that the path travelled by the beam upon reflection from the back surface of each sample spanned the near-field region of the transducer. For each sample the amplitude of the component at 5 MHz was measured three times. Such an experimental configuration was simulated by considering the backscattering of an incident beam of longitudinal waves by a large crack with stress-free surfaces. Even in this case, a reasonable agreement between theory and experiment was found, with the maximum difference between the two of the order of 10 percent at a distance of 0.5 near-field lengths from the transducers