Determination of the underwater acoustic properties of a towing tank for source radiated sound power measurement

Abstract

Sound generated by marine vessels and structures, radiates into the marine environment. It is of great importance regarding the acoustic signature of naval vessels, and can have a dramatic impact on the behaviour and well-being of marine fauna. Additionally there is a need to make autonomous and remotely operated underwater vehicles as quiet as possible such that they can record the ocean soundscape, and not influence the behaviour of the marine life which they may be researching. Towing tanks offer a practical and controlled environment to measure and optimise the noise radiated by a ship at model scale. Additionally most underwater robotic vehicles are small enough to operate in a towing tank. However, sound measurements in towing tanks suffer from the effects of reverberation. This means sources acoustical properties measured in the tank can be significantly different to a measurement made in free field conditions. Current methods to overcome this method involve taking a measurement close to the source such the direct field is dominant over the reverberant field. However the measurement of reverberation distance is achieved with classical room acoustic equations which use an assumption of a diffuse field in their derivation. It is shown in this thesis, that the acoustic field inside a towing tank is not diffuse, and as such these classical room acoustic equations are not applicable to a towing tank. Additionally, limiting measurements to be within the direct field may cause the measurement to be in the near field of the source for low frequencies. In order to characterise the acoustic environment of a towing tank, and understand its impact on the acoustical properties of a source, an expression for the exact acoustical field inside a rectangular room with the boundaries and geometry of a towing tank is derived. To improve the computational cost of numerically evaluating this expression, an approximate form is derived, known commonly as the image source model. A spherical wave reflection coefficient is derived to improve the accuracy of the image source model. Impulse response measurements are undertaken in the Boldrewood towing tank, which are used to calibrate the damping in the image source model, through the tuning of the shear and compressional wave speeds in the plane wave reflection coefficient of the concrete wall of the tank. This method has produced an indirect way of measuring the shear and compressional wave speeds in the concrete walls of a water tank. A non-dimensional coefficient relating the power radiated by a point monopole source in a towing tank, to the surface averaged mean square pressure over the side wall of the tank is defined. This power coefficient can be used to predict the radiated power of a source in a tank, through measurement of the mean square pressure on the wall of the tank. The calibrated image source model is used to build a dataset of source radiated power and surface averaged mean square pressure for a large number of source locations. This enables the power coefficient to become statistically valid, and enable a prediction of power to be made with the measurement of surface averaged mean square pressure, with a quantifiable random error. Using the statistical analysis it is found that only 15 hydrophones are required to achieve a 98% probability of being within 3dB of the tanks mean value. This is a significant reduction in the number of hydrophone measurement locations used to measure power using the method of volume averaged mean square pressure in a diffuse field. The radiated power of an omnidirectional ITC1032 transducer is measured and validated against manufacturers specifications.

More information

Identifiers

ORCID for Stephen Turnock: orcid.org/0000-0001-6288-0400

ORCID for Victor Humphrey: orcid.org/0000-0002-3580-5373

ORCID for Dominic Hudson: orcid.org/0000-0002-2012-6255

Catalogue record

Export record