Phase-sensitive and phase-insensitive ultrasonic measurements of rough interfaces and anisotropic media with predictions using angular spectral decomposition

This thesis presents phase-sensitive and phase-insensitive measurements of ultrasonic fields transmitted through randomly rough surfaces and seeks to contribute to the understanding of the physical principles leading to phase distortion from these rough interfaces between two media or anisotropy within a medium. To accomplish this, an isotropic specimen with a specially prepared rough surface was measured ultrasonically in an immersion tank using a transmitting transducer and a hydrophone scanned in a two-dimensional pseudo-array. Signals from the arrays were analyzed with phase-insensitive (a method that reduces phase-cancellation) and phase-sensitive (a method currently in wide use) processing. Phase-sensitive measurements yielded signal losses in the bandwidth from 2 to 8 MHz which could be described as frequency-dependent transmission coefficients for the rough surface. Receiver operating characteristic (ROC) analysis revealed that phase-insensitive measurements provide a more robust and effective detection of material abnormalities in a specimen with a rough surface. Computer simulations based on the method of angular spectral decomposition provided further insights into the distortions which contribute to phase-cancellation across the face of piezoelectric receivers. The method of angular spectral decomposition was generalized to simulate propagation in anisotropic media. The resulting simulations demonstrated distortion effects such as energy flux deviation in unidirectional graphite-epoxy. To simulate fields which have been transmitted through rough surfaces the isotropic angular spectral decomposition was extended to simulate the effects of transmission across an interface. The phase-sensitive and phase-insensitive detection of the simulated fields showed good agreement with the experimental measurements.