Holographic projection of sound fields based on spatially-limited data sets

Yong Thung Cho, Purdue University

Abstract

In the past, it has been demonstrated that nearfield acoustical holography (NAH) is a useful tool for visualizing noise sources. At present, planar and cylindrical holography are the most widely used among the various nearfield acoustical holography techniques. However, to avoid spatial Fourier transform-related truncation effects in conventional implementations of the latter procedures, the measurement aperture (i.e., the hologram surface) must typically extend well beyond the source to a region where the sound pressure level drops to a level significantly lower than the peak level within the measurement aperture. In contrast, statistically optimized nearfield acoustical holography (SONAH), does not suffer from the effects of spatial truncation and can be used to visualize limited areas of a source without compromising the accuracy of the projection. Here the implementation of SONAH in cylindrical coordinate is described, and its performance is compared with "patch" holography, an extrapolation procedure. It was found that SONAH is faster and more accurate than patch holography, but significantly slower than conventional, DFT-based holography. In addition to the identification of acoustical properties on the source plane, farfield radiation properties of sources were also investigated. Farfield prediction beyond the measurement region requires extrapolation in addition to forward projection of measurement pressure. Further, in contrast to conventional NAH procedures, which cannot be implemented in non-regular geometries, SONAH was here implemented in conical geometry to facilitate the measurement of aeroacoustic sources. Acoustical holography usually requires a large number of measurements that cover the entire source area. Here, a technique for estimating the sound field radiated by composite sources where sub-sources have fixed component directivity is introduced. This procedure allows sound fields to be reconstructed by using only a very small number of measurements when certain conditions are satisfied. Finally, various acoustical holography procedures have also been verified experimentally in measurements using loudspeakers and simulated multipole noise sources. In both cases, reliable results were obtained if appropriate procedures were followed. Thus it was concluded that acoustical holography can be a flexible and robust tool for noise source identification and for farfield projection based on nearfield measurements.

Degree

Ph.D.

Advisors

Bolton, Purdue University.

Subject Area

Mechanical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS