Qualification of anechoic chambers is intended to demonstrate that the chamber supports the intended free-field environment within some permissible tolerance bounds. Key qualification issues include the method used to obtain traverse data, the analysis method for the data, and the use of pure tone or broadband noise as the chamber excitation signal. This paper evaluates the relative merits of continuous versus discrete traverses, of fixed versus optimal reference analysis of the traverse data, and of the use of pure tone versus broadband signals. The current practice of using widely space discrete sampling along a traverse is shown to inadequately sample the complexity of the sound field extant with pure tone traverses, but is suitable for broadband traverses. Continuous traverses, with spatial resolution on the order of 15% of the wavelength at the frequency of interest, are shown to be necessary to fully resolve the spatial complexity of pure tone qualifications. The use of an optimal reference method for computing the deviations from inverse square law is shown to significantly improve the apparent performance of the chamber for pure tone qualifications. Finally, the use of broadband noise as the test signal, as compared to pure tone traverses over the same span, is demonstrated to be a marginal indicator of chamber performance.

Results of a research program to develop computational methods to minimize noise transmission into aircraft fuselage interiors are discussed. A design tool to perform a constrained optimization of the acoustic environment within a vibrating structure is developed utilizing nite element methods and boundary element methods (FEM/BEM), and its application to aircraft cabin noise problems is studied. The results of a study to optimize the cross section shapes of frames and stringers of an idealized aircraftlike stiffened cylinder are reviewed. The structure is optimized for minimum noise at speci ed points in the interior, as a result of a single frequency (tonal) exterior acoustic disturbance. For the cylinder and excitation frequency studied, it has been found that spatially varying the stiffener sizes over the cylinder is more important than optimizing the shape of the cross sections. Because FEM/BEM methods are only reliable for lower frequencies, the problems studied are applicable to low-frequency tonal noise such as seen in turboprop aircraft. Nomenclature vectors of lower and upper bounds F(b)= objective function f 0 = initial value of objective function p i (b) = complex acoustic pressure at node i S = scale factor design variable v = normal velocity of shell structure, or acoustic velocity at shell wall W = structural weight W max = constrained maximum weight x i , y i = coordinate design variables at node i

No abstract

No abstract

A large underwater acoustic tank facility located in the Woodruff School of Mechanical Engineering at Georgia Tech has recently been completed. The facility includes a rectangular concrete water tank 25 feet deep, 25 feet wide, and 34 feet long containing around 160,000 gallons of water. There are three computer-controlled positioners: an x-y-z-θ positioner and a z-θ positioner mounted on carriages and a bottom mounted rotator. The facility has a large rectangular nearfield array which can be used either as a receiver or a transmitter. A single vertical nearfield line array can be translated by the x-y positioner to synthesize a cylindrical nearfield receiving array. The rectangular nearfield transmitting array and the synthesized cylindrical receiving array were designed to be used with the bottom mounted rotator to measure the true farfield bistatic target strength of any target up to one meter in length as a function of the target aspect angle. Such measurements can be done from 2 kHz to over 10 kHz. The tank is being used for transducer development, materials, and flow noise studies in addition to structural acoustics. Several available multichannel data acquisition systems will be described. [Work supported, in part, by a DURIP grant from ONR.]