Figure 8 - uploaded by Con Doolan
Content may be subject to copyright.
Acoustic beamformer installed in the AWT, with the microphones arranged in the Arcondoulis Spiral configuration.
Source publication
Acoustic beamforming is an experimental tool that can be used to locate and quantify aeroacoustic noise sources. Much of the available aeroacoustic beamforming literature presents beamforming results of noise at relatively high frequencies. There are few experimental acoustic beamforming results for acoustic frequencies between 1 kHz and 5 kHz, alt...
Context in source publication
Similar publications
Microphone arrays methods are useful for determining the location and magnitude of rotating acoustic sources. This work presents an approach to calculating a discrete directivity pattern of a rotating sound source using inverse microphone array methods. The proposed method is divided into three consecutive steps. Firstly, a virtual rotating array m...
Citations
... The polymer fibres used were 80 mm in length and 750 μm in diameter, and were arranged in a five-branch spiral array with logarithmic spacing. The design developed has been based on the key features of designs for aero-acoustic microphone arrays, which seek to minimise co-linearity for better beamforming performance [19]. Despite how closely packed the fibres appeared (as visible in Fig. 2), they were out of plane from one another, positioned roughly 0.35 whisker lengths apart. ...
... This number does not account for the internal processing time of the tracking camera or the FBG interrogator (tracking the shift in wavelength of the FBG sensors), but conveniently the delay from both of these components was extremely small and could be ignored, compared to the Fig. 7. Array layouts performance assessed using large numbers of visually tracked plastic sensors spaced 200 mm from the disturbance. This spiral array is inspired by similar array designs for acoustic microphone arrays, used for acoustic source localisation [19]. DoA computation time. ...
Previous research has identified the capabilities of artificial whisker sensor arrays as an innovative method of tracking and evaluating flow events and disturbances. In this work, a new approach is put forward, focusing on achieving enhanced sensitivity and improved performance of fibre-optic whisker sensors, based on the principle of fibre Bragg Grating (FBG) based optical stress sensing. Its performance is evaluated against more simplistic approaches, and previously demonstrated methods of optically tracking the tips of whisker sensors. The study has found the performance and sensitivity of these FBG-based sensors to be very satisfactory, with sufficient sensitivity to bending stresses to enable using Cross-Correlation (CC) and multilateration techniques (as evaluated in prior studies) to produce reliable Direction of Arrival (DoA) and velocity estimations, at a typical SNR of around 2 dB. The system has shown the capability for correction of potentially disruptive variations in temperature, allowing for effective measurement of the key hydrodynamic disturbances under study, irrespective of the local environmental conditions.
... Typically logarithmic spiral patterns, multiarm spiral patterns and their variations are successful at producing quality beamforming image source maps with relatively few microphones [24,[28][29][30][31][32], as compared to other designs such as randomized [33] and grid pattern arrays [34,35]. ...
Acoustic beamforming array design methods are typically suited for circular and rectangular areas. A comparison of three array design methods is presented in this paper over irregular shaped areas, including L-shapes and arches. Partial-logarithmic spiral arrays that possess their geometric center either at the origin of the array area or the centroid of the irregular shaped area are compared against randomized array designs based on Maximum Sidelobe Level (MSL) parameters and arrays generated using a recently published array design method named the Adaptive Array Reduction Method (AARM). In the AARM a large array is reduced to a smaller array by seeking the removed microphone that possesses the minimum value of the MSL, the Main Lobe Width (MLW) and a lobe-distortion term. The AARM method is also tested in two practical cases against a partial-spiral array design used at the NASA Langley Low-Turbulence Pressure Tunnel and a hypothetical rectangular wall case. In both cases the AARM showed superior performance to the logarithmic spiral arrays in all cases based on MSL and MLW criteria. Of the three methods compared, the AARM best utilizes the full potential array aperture of an irregular area and therefore produces the best MSL, MLW and lobe-distortion values.
... Using simulated monopole sources and measuring performance in both the near and far field, a comparison of popular spiral array types was performed [20] showing that the Underbrink spiral design [14] possessed the best all-round performance in the acoustic near and far field. Small improvements in sidelobe characteristics have been achieved by modifying a typical logarithmic array pattern by focussing more of the microphones near the array centre while still spanning the same array area [21,22] for sources that are located near the centre of the scanning grid [20]. While the logarithmic spiral arrays and their variations present a well-balanced MSL and MLW in the CSB source map, the array design process for a specific frequency range, array size, scanning grid size and expected distance of source to the array, can be highly iterative and indeterminate. ...
... the size of the main lobe (true signal) [1][2][3][4][5]. Typically spiral-based arrays and their variations are successful at producing high quality beamforming image source maps with relatively few microphones [3,[6][7][8][9][10], as compared to other designs such as randomized [11] and grid pattern arrays [12,13]. An alternative array design method proposed by Arcondoulis and Liu [14] uses an iterative microphone removal scheme, labeled the Adaptive Array Reduction Method (AARM). ...
Experimental facilities for aeroacoustic tests can contain large apparatus and supporting structures. This can impair the design and/or installation of an acoustic beamforming array that requires significant space and, in most cases, a clear line of sight between the array and the acoustic source. To alleviate this issue, previous studies have populated portions of spiral-based arrays over usable areas to best utilize the available space. This paper aims to numerically
compare the performance of this array design method with a recently published method, called the Adaptive Array Reduction Method (AARM). The AARM simulates a large initial array and iteratively reduces the array to a desired number of channels by removing the channel based on sidelobe and main lobe criteria. The AARM and a spiral-based array design method were applied to a hypothetical room containing several obstructions, and a non-square area at NASA Langley Low-Turbulence Pressure Tunnel. In both cases the AARM outperformed the spiral based array designs in terms of main lobe width and maximum sidelobe level and in some cases the main lobe distortion. To experimentally verify the design of AARM arrays over irregular areas, experiments were conducted at the Southern University of Science and Technology in an anechoic wind tunnel, using both a speaker source in zero-flow conditions and a NACA 0012 airfoil placed in uniform flow. The performance of the AARM arrays created over irregular areas were compared against an AARM array designed over a square area, revealing little to no decrease in performance in terms of source localization and sidelobe distributions in the source maps. The AARM has been shown both numerically and experimentally to be a suitable method for designing acoustic beamforming arrays in a spatially restricted experimental facility.
... A well-designed array minimizes the contributions of spatial aliasing images called sidelobes and also minimizes the size of the main lobe (assuming a single source configuration) [4,5]. Typically logarithmic spiral patterns and their variations have been successful at producing quality beamforming image source maps with relatively few microphones [6][7][8][9][10], as compared to other designs, such as randomized [11] and grid pattern arrays [12]. Array optimization algorithms that aim to minimize the Maximum Sidelobe Level (MSL) [3,[12][13][14] yield MSL improvements relative to an initial array design yet none conclusively display a superior combination of MSL and Main Lobe Width (MLW) compared to logarithmic spiral arrays and their variations for a range of frequencies and acoustic source locations. ...
New methods of beamforming array design have been recently proposed, namely the Array Reduction Method (ARM) and its improvement, the Adaptive Array Reduction Method (AARM). For both of these methods, a large grid array design possessing many more microphones than desired is simulated. By calculating the product of the Maximum Sidelobe Level (MSL) and the Main Lobe Width (MLW) (defined as Φ), a microphone is selected that possesses the minimum value of Φ or a variation of Φ. This microphone is removed and the process is then repeated until a desired number of microphones remains. The AARM was derived that factors in the relative change of MSL and MLW with respect to the number of microphones removed, to ensure that the algorithm produced the best overall image source map, rather than simply attempting to minimize Φ. Furthermore, a penalty is introduced during the microphone selection stage that is based on the distortion of the main lobe, such that the final array produces a source image map with a main lobe that is symmetric as possible. In this paper, ARM, AARM and AARM (without factoring in lobe distortion effects) arrays are presented, that were designed for specific acoustic source locations and frequencies. They are numerically tested using a single source at the center and off-center relative to the array center and a dual-source configuration. The experimental conditions were conducted in an anechoic wind tunnel using the same conditions as the numerical simulations, with the exception of airfoil self-noise tests. The AARM is further tested over a challenging irregular array area, such that a typical logarithmic spiral cannot be fitted, revealing the capability of using the AARM in challenging environments.
... Using simulated monopole sources and measuring performance in both the near and far field, a comparison of popular spiral array types was performed [19] showing that the Underbrink spiral design [18] possessed the best all-round performance in the acoustic near and far field. Small improvements in sidelobe characteristics have been achieved by modifying a typical logarithmic array pattern by focussing more of the microphones near the array centre while still spanning the same array area [20,21] for sources that are located near the centre of the scanning grid [19]. While the logarithmic spiral arrays and their variations present a well-balanced MSL and MLW in the CSB source map, the array design process for a specific frequency range, array size, scanning grid size and expected distance of source to the array, can be highly iterative and indeterminate. ...
... This results in a decrease of MSL to below −30 dB upon removal of the first approximate 100-channels. Some redundant microphones in a densely spaced grid array often contribute to strong grating lobes at 90 • angles to the main lobe centre [15,21] and their removal decreases the strength of the grating lobes. Progressing to Fig. 7 (Fig. 7(d)) to M f = 48 (Fig. 7(e)). ...
... 6 A well-designed beamforming array minimizes the contributions of spatial aliasing images called sidelobes and also minimizes the size of the main lobe (assuming a single source configuration). 7,8 Typically, logarithmic spiral patterns and their variations are successful at producing quality beamforming source image maps with relatively few microphones, [8][9][10][11][12] as compared to other designs, such as randomized 13 and grid pattern arrays. 14 Array optimization algorithms that aim to minimize the Maximum Side Level (MSL) 6,14-17 yield MSL improvements relative to an initial array design; however, none conclusively display a superior combination of MSL, Main Lobe Width (MLW), and main lobe distortion compared to logarithmic spiral arrays and their variations for a range of frequencies and acoustic source locations. ...
This letter presents a modification to a very recent iterative microphone removal array design process, called the Array Reduction Method (ARM). The ARM iteratively reduces an array based on the product of sidelobe and main lobe criteria, yet the main lobe may become distorted and the beamformer output can disproportionately favor one of the criteria. The Adaptive Array Reduction Method (AARM) presented here uses the derivative estimates of the maximum sidelobe level and main lobe width with respect to the number of microphones removed, and the distortion of the main lobe during iteration, to produce an improved source image map.
... Using simulated monopole sources and measuring performance in both the near and far field, a comparison of popular spiral array types was performed [17] showing that the Underbrink spiral design [16] possessed the best all-round performance in the acoustic near and far field. Small improvements in sidelobe characteristics have been achieved by modifying a typical logarithmic array pattern by focussing more of the microphones near the array centre while still spanning the same array area [18,19] for sources that are located near the centre of the scanning grid [17]. Efforts have been made to optimise array designs by using cost functions to minimise MSL [20][21][22][23] that yield significant improvements of MSL relative to an initial array design yet none of these designs conclusively display a superior combination of MSL and MLW compared to logarithmic spiral arrays and their variations. ...
... This results in an a decrease of MSL to below -30 dB upon removal of the first approximate 100-channels. Some redundant microphones in a densely spaced grid array often contribute to strong grating lobes at 90°angles to the main lobe centre [19,23] and their removal decreases the strength of the grating lobes. Progressing to Figs. 7b and 7c, there do not appear to be any clear discernible patterns of array removal and only small changes in MSL, all of which are less than -30 dB. ...
A well-designed acoustic beamforming array aims to simultaneously minimise the Maximum Sidelobe Level (MSL) and Main Lobe Width (MLW) within a desired acoustic frequency range, yet these properties tend to be mutually exclusive. This paper describes an iterative microphone removal method. By starting from a large initial array stencil, the microphone within the array that results in the smallest product of frequency-averaged MLW and MSL of the point spread function is identified and removed. This process is then repeated until a predetermined desired number of microphones is reached. Using the cross-spectral beamforming algorithm, several 841 and 961-channel equi-and-non-equispaced grid arrays are reduced to unique 48-channel arrays. The reduction method is implemented in two ways, i.e., one-by-one microphone removal and an accelerated procedure that removes several microphones per iteration , where the removed number of microphones decreases in an exponential manner with increasing iterations. The reduced arrays are compared against several logarithmic spiral and randomised 48-channel pattern arrays using a non-dimensional metric, evaluated as the product of the MSL normalised by main lobe magnitude, and the ratio of the MLW area to the total scanning grid area. The frequency range of investigation is 2000 Hz to 8000 Hz to replicate typical conditions in small scale anechoic facilities. The array reduction methods reveal improved array designs according to the aforementioned metric.
Many workers are exposed daily to excessive noise levels. In order to reduce the noise exposure at the workstation, the main noise sources have to be detected. This task can be done with a microphone array and a source localization technique. Previous studies have shown promising results with time domain beamforming based on the generalized cross-correlation technique. The objective of this work is to propose an optimal spherical microphone array geometry dedicated to this technique. A cost function based on the symmetry of the aperture angle maps is proposed and is maximized using a Nonlinear Optimization by Mesh Adaptive Direct Search. Numerical results show that the optimized geometry improves the noise source map by reducing the side lobe amplitude without increasing the main lobe surface. Experimental measurements are carried out in a semi-anechoic chamber with prototyped spherical microphone arrays confirming that the optimized microphone array improves the quality of the noise source map.
Some aspects of wind turbine noise generation can be best observed and understood by taking measurements in wind tunnels of noise from models of wind turbines complete with wind turbine blades and turbine blade sections. These tests are referred to as fixed-blade testing. Wind tunnel testing can remove some of the complicating factors related to atmospheric conditions. Also, scaling requirements result in higher rotor speeds and hence the frequency range of interest is not so low as to require specialised low-frequency microphones. The chapter focuses on wind tunnel testing and shows how to take acoustic measurements in wind tunnels, as aerodynamic and acoustic measurements are important when evaluating rotor blade sections. This includes single-microphone and acoustic-array measurements plus a summary of some recent wind tunnel measurements relevant to wind turbines. The chapter also focuses on recent acoustic-array testing of full-scale turbines.
