No abstract

The benefits of using highly directive parabolic reflectors for acoustics and electromagnetics work is well known. The primary difficulty for underwater acoustics work is making a highly reflective surface, one that is acoustically very hard or very soft, in the shape of a parabola. Very hard would require a great deal of machined metal and would most likely be cost and weight prohibitive. Therefore, for underwater acoustics work, parabolas are usually made to be acoustically soft. Although soft foam is a good choice, it will crush at even shallow depths. Other solutions include compliant tube designs with and without pressurized air compensation. These tend to be very expensive. Here, an inexpensive and rugged parabolic receiver/transmitter with virtually no depth limitation is described. It consists of a fiberglass parabola backed by a fiberglass dome. The air cavity between them results in a soft reflector. Vibration isolation is provided by steel rings and rubber gaskets. To prevent crushing, the air chamber can be pressurized either by compressed air fed by hose from the surface or an attached tank and regulator. Both transmit and receive directivity patterns will be shown. The device is currently being used for target strength measurements at the NUWC Lake Seneca Test Facility.

The radiated noise goals for various underwater vehicles are becoming more stringent. As these goals are approached, novel acoustic test and evaluation schemes will be required to overcome ambient noise limitations and confirm performance. Customized directive arrays will be required in order to make the measurements. Large multielement volumetric arrays will be cost prohibitive. Here, a unique baffled vertical line array deployed at the AUTEC ocean haul down facility is described. This device was designed to overcome shrimp noise from a nearby reef and make acoustic measurement of buoyant test vehicles and weapons under shallow-water conditions. This array meets the following specialized criteria: (1) very high backside rejection to shield the array from shrimp noise, (2) very wide (120°) frontside horizontal beam pattern to ensure capturing the vehicle as it passes the array, (3) beam patterns stable to within 1 dB over the beamwidth for accurate measurements, and (4) the ability to discriminate against surface noise. Diffraction problems from the edge of the baffle were solved by randomizing the edges thereby creating what is called ‘‘diffraction wings.’’ Plate vibration problems were solved by constructing a baffle with blocks welded to the back of a relatively thin plate (thereby having mass with no stiffness) and the addition of damping treatment. Receiver standoff problems were solved by the use of conformal PVDF sensors.

For the next generation of quiet underwater vehicles, high-array gain noise measurements systems, with accuracy on the order of 1 dB, are necessary. The design of these systems requires evaluation of the propagation stability over the short ranges normally encountered during radiated noise trials. Intensity fluctuations were measured at AUTEC at frequencies 1 kHz<f<20 kHz, at ranges r<1 km, at source depths SD = 90 m, 240 m, and 580 m, and at several grazing angles 0° < θ < 40°. Measured TL often differed from spherical spreading by greater than 5 dB for f > 2 kHz, θ near 0°, and r > 200 m at all source depths. This difference, with a few exceptions, fell off rapidly to less than 1 dB for grazing angles greater than 4°. The mean and standard deviation of the TL difference from spherical spreading were estimated in 1° grazing angle bands. The standard deviation was lower at the deeper source depths and always fell off rapidly for θ > 4°. The mean value was not always lower at deeper source depths and, although it fell off rapidly for θ > 4° and f < 8 kHz, it did not always do so at the higher frequencies. These higher frequency and higher angle measurements, however, were associated with longer range measurements (r > 500 m). Theoretical predictions of the TL using a fine structure CTD measurement as input to the SAFARI propagation routine explain most of the results seen and suggest that the mean errors seen at higher grazing angles will go away for f < 500 m. All of these results indicate that intensity accuracy on the order of 1 dB can be achieved for f < 20 kHz and r < 500 m (or f < 8 kHz and r < 1 km) by making radiated noise measurements at θ > 4° or by restricting ranges to less than 200 m.