Iteration Techniques in Computer Study of Flooded Cylindrical Radiators

Abstract

The computer study reported at the 65th Meeting of the ASA has been successful in explaining some of the behavior of large, flooded, barrel-stave transducers. This approach has a fundamental weakness in that one does not know a priori the correct source surface-velocity distribution to feed into the computer program. By the time these data are available in a specific design case, it is too late for the computer study to be of value to the designer in the prediction of general behavior. An iteration procedure has been developed, which solves this difficulty by successively modifying the velocity distribution. This is done in such a way that the prescribed source velocity of the active region of the transducer is maintained constant while the inactive region of the model is given nearly any surface shape that is convenient. In general, the result of this procedure is a set of boundary conditions that fails to satisfy the physical requirements. These inconsistencies are used to produce a more reasonable source velocity distribution, which serves as an improved input for a second solution. This process is repeated, each time giving boundary conditions that are more physically acceptable. Program results are compared with experimental model measurements. [Work supported by U. S. Office of Naval Research, Acoustic Programs.]