Journal of Computational Acoustics

A parallelized algorithm based on an existing 3D wide-angle parabolic equation model is developed to perform numerical simulations of underwater acoustic propagation on a massively parallel computer. The parallelization method used is a suitable two-level procedure: A frequency decomposition and a spatial decomposition of the calculations, which are respectively dedicated to reduce CPU times for broadband and cw signal propagation. The high-performance of the parallelized algorithm is examined for the 3D extension of the classical ASA wedge benchmark. CPU times are reported and both speedup and efficiency are analyzed. An investigation of significant 3D effects at higher frequencies and at longer propagation ranges than in earlier works [F. Sturm, J. Acoust. Soc. Am.117(3) (2005) 1058–1079] is performed with reasonable CPU times by using the new parallel algorithm. Further, the feasability of the procedure applied to a realistic oceanic environment problem involving both real sound speed profiles and bathymetry data sets is also illustrated.

This study obtains wind noise variations by experimental data and simulated results to describe meteorological and oceanic effects. The ambient noise data were measured by a vertical line array in the 2001 ASIAEX South China Sea experiment. An acoustic propagation model is used in noise modeling for calculating the sound pressure of noises at receiving sites, including the effects of ocean environmental changes, bottom interactions, and noise fluctuations at different depths. Both range-independent and range-dependent sound speed profiles are generated with in-situ water temperature data. Results show fluctuating noise levels with variations in ocean environments. But the fluctuations are small such that only weak correlation exists in the acoustic noise data and ocean data. Results also indicate that using range-independent sound speed profiles can simulate noise field in range-dependent ocean environments with nonlinear internal waves for shallow regions with flat bottoms.

This paper models three-dimensional wave propagation around two-dimensional rigid acoustic screens, with minimal thickness (approaching zero), and placed in a fluid layer. Rigid or free boundaries are prescribed for the flat fluid surfaces.

The problem is computed using the Traction Boundary Element Method (TBEM), which is appropriate for modeling thin-body inclusions, overcoming the difficulty posed by the conventional direct Boundary Element Method (BEM).

The problem is solved as a summation of two-dimensional problems for different wave numbers along the direction for which the geometry does not vary. The source in each problem is a spatially sinusoidal harmonic line load. The influence of the horizontal boundaries of the fluid medium on the final wave field is computed analytically using appropriate 2.5D Green's functions for each model developed. Thus, only the boundary of the rigid acoustic screen needs to be discretized by boundary elements.

The computations are performed in the frequency domain and are subsequently inverse Fourier transformed to obtain time domain results. Complex frequencies are used to avoid aliasing phenomena in the time domain results.

Having available efficient global optimization methods is of high importance when going to a practical application of geo-acoustic inversion, where fast processing of the data is an essential requirement. A series of global optimization techniques are available and have been described in literature. In this paper three optimization techniques are considered, being a genetic algorithm (GA), differential evolution (DE), and the downhill simplex algorithm (DHS). The performance of these three methods is assessed using a test function, demonstrating superior performance of DE. Additionally, the DE optimal setting is determined. As a next step DE is applied for determining the geo-acoustic properties of the upper seabed sediments from simulated seabed reflection loss, indicating good DE performance also for real geo-acoustic inversion problems.

A coupled-mode formalism based on complex Airy layer mode solutions is presented. It is an extension into the complex horizontal wavenumber plane of the companion article [Stotts, J. Acoust. Soc. Am.111 (2002) 1623–1643], referred to here as the real horizontal wavenumber version, which accounted for general ocean environments but was restricted to normal modes on the real horizontal wavenumber axis. A formulation of the expressions for both trapped and continuum complex coupling coefficients is developed to dramatically reduce computer storage requirements and to make the calculation more practical. The motivation of this work is to demonstrate the numerical implementation of the derivations and to apply the method to an example benchmark. Differences from the real horizontal wavenumber formalism are highlighted. The coupled equations are solved using the Lanczos method [Knobles, J. Acoust. Soc. Am.96 (1994) 1741–1747]. Comparisons of the coupled-mode solution to a parabolic equation solution for the ONR shelf break benchmark validate the results.

Complex structures often include hollow parts. These hollow parts constitute the skeleton of the structure and are largely responsible for its global behavior, hence the importance of analyzing them precisely. The method we propose in this paper is a modal method based on the modal analysis of elements of hollow parts. This method does not require nodal degrees of freedom on the boundaries between the elements: "modal" elements are created, and these elements can be assembled through modal mass and stiffness matrices in the same way as the finite element method. Thus, it is possible to choose the precision of the analysis by choosing the quantity of modes used in the modal analysis of the elements. We will study not only structural systems but also coupled fluid–structure systems, and our results will be compared with experimental tests.

In many long-range ocean propagation experiments the source and receiver are placed close to the depth of the waveguide axis. In this case, rays emerging from the source at sufficiently small angles intersect the sound-channel axis many times and form in its vicinity a large number of caustics with caustics cusps located repeatedly along the axis. In neighborhoods of cusped caustics there exists a very complicated interference pattern. Neighborhoods of interference grow with range and at long ranges they overlap. As a result, a complex interference wave (axial wave), that propagates along the waveguide axis, appears. The goal of this paper is to obtain the representation for the axial wave in the time domain and calculate its magnitude for a realistic model of a three-dimensional range-independent medium. Numerical computations are done for the average profile from the Acoustic Engineering Test (AET) experiment. The pulse center frequency of 75 Hz with 30-Hz bandwidth is used that corresponds to broadband acoustic signals which were transmitted during November 1994 in the eastern North Pacific Ocean as a part of the AET. The propagation range is 3250 km. The sound source is located on the waveguide axis, and the receiver is placed close to the depth of the axis. Through numerical simulation the dependencies of the magnitude of the axial wave on depth of the receiver and propagation range are studied.

Compared to the traditional boundary element method (BEM), the single level fast multipole boundary element method (SLFMBEM) or the multilevel fast multipole boundary element method (MLFMBEM) reduces the computational complexity of a job from O(n²) to O(n^3/2) or O(n log²n), respectively with n being the number of unknowns; this means a dramatical reduction in terms of CPU-time and storage requirement. Large scale problems, unsolvable with the traditional BEM, can be solved by using the FMBEM. In this paper, the traditional BEM, SLFMBEM, and MLFMBEM are formulated within the framework of the Burton–Miller Collocation BEM for acoustic radiation and scattering from 3D structures. Attention is especially paid to the practical aspects of the method in order to get a reliable and efficient computation code. The performance of the method is tested with practical examples, including one for computing the head-related transfer function (HRTF) between 1000 and 18 000 Hz.