ABSTRACT
A set of tank experiments was performed to investigate acoustic transmission across a roughened fluid-fluid interface with the intention to test heuristic Bragg scattering predictions used to explain observations of anomalous transmission in field experiments. In the tank experiments, two immiscible fluids (vegetable oil floating on glycerin) formed the layers. Small polystyrene beads were floated at the interface to simulate roughness. An array of hydrophones placed in the bottom layer (glycerin) was used to measure the acoustic levels transmitted across the interface. This array was also employed as a beamformer to determine the apparent angle and sound speed of the scattered signals. Data were acquired at subcritical grazing angles in the frequency range of 100-200 kHz for three different bead diameters and for various configurations in which the locations of the beads floating on the interface were varied. Results of these measurements demonstrated that a significant amount of acoustic energy can be scattered into the bottom layer by beads floating at the interface. The scattered levels increased with increasing bead diameter. However, discrepancies occurred between observed propagation properties and the Bragg predictions. By comparing the processed tank data to a computer simulation of the same it was determined that these discrepancies are a consequence of near-field reception of the scattering by the bead array and ignoring the directionality of the scattering by the beads. Consequences to observations made in field experiments are discussed.
