基于反射系数估算的半空间边界阻抗和声源直接辐射重构

半空间中声源直接辐射声场重构的实施需要构造以边界声阻抗为参量的半空间基函数,边界声阻抗的获取则通常需要借助原位测量方法.基于半空间球面波基函数叠加的声场重构方法,通过在声源近场布置全息测量面和一支参考传声器采集声压,并以参考传声器声压重构误差取得最小值为准则,估算各全息测点的声压反射系数,就能在边界阻抗未知条件下实现声源直接辐射声压的重构,从而摆脱了常规方法对声阻抗原位测量技术的依赖.本文的目的是对这一方法进行详细的参数讨论,并在估算声压反射系数的基础上,进一步对边界声阻抗加以重构,提出一种基于近场声全息的声阻抗测量方法.以球形声源为例,对声阻抗和声源直接辐射声压的重构进行了仿真,定量地分析参考传声器坐标、边界有效流阻率和边界孔隙度随深度的降低率等参数对重构精度的影响.     

单层传声器阵列信号空间重采样的声波分离方法

为了重构非自由声场中目标声源的声场响应,提出单层传声器阵列信号空间重采样的声波分离方法.以球面波函数为基函数,建立由系列球面波函数叠加表达的声场数学模型.基于近场声全息原理,利用单层传声器阵列面上空间重采样形成的两组声压测量信号,求解基函数系数,并重构出传声器阵列两侧声源各自的声场响应,实现声波分离.使用脉动球和振动球共同作用的非自由声场,检验了数学模型以及传声器信号信噪比、传声器阵列形状和面积、声源中心位置、频率等关键参数对声波分离精度的影响,并在全消声室内进行了实验验证.最后,对单层传声器阵列重采样的声波分离方法的实施给出了建议.     

单层传声器阵列重建相干声场的近场声全息方法研究

针对传声器阵列两侧存在相干声源的非自由声场重建问题,提出基于球面谐波函数扩展近场声全息理论的相干声场重建方法。该方法在已知测量面两侧声源几何位置时,使用单层传声器阵列获取测量面处的声压分布,通过最小二乘法获得与目标声源和干扰噪声源响应对应的最优球波函数扩展项数和最优系数向量,结合测点位置的空间坐标进行声波分解,并分别重建出各声源在测量面上的声压分布。为了验证方法的有效性,分别给出了相干噪声源为球形声源和非球形声源的仿真验证,并在全消声室内对双扬声器产生的相干声场的重建进行了实验验证。结果表明:该方法对球形声源和非球形声源干扰下的声场重建都具有较好的效果,球形声源干扰下的重建精度更高。      

单层传声器阵列重建相干声场的近场声全息方法研究

针对传声器阵列两侧存在相干声源的非自由声场重建问题,提出基于球面谐波函数扩展近场声全息理论的相干声场重建方法。该方法在已知测量面两侧声源几何位置时,使用单层传声器阵列获取测量面处的声压分布,通过最小二乘法获得与目标声源和干扰噪声源响应对应的最优球波函数扩展项数和最优系数向量,结合测点位置的空间坐标进行声波分解,并分别重建出各声源在测量面上的声压分布。为了验证方法的有效性,分别给出了相干噪声源为球形声源和非球形声源的仿真验证,并在全消声室内对双扬声器产生的相干声场的重建进行了实验验证。结果表明:该方法对球形声源和非球形声源干扰下的声场重建都具有较好的效果,球形声源干扰下的重建精度更高。     

双声源驱动热声系统声场调制研究

本文在回热器边界声场调制理论的基础上,考虑了两端换热器和谐振管变径的影响,推导出热声系统(谐振管和回热器)内声场与双声源复声压相关的声场调制关系式。理论和实验分析了双声源复声压对谐振管中的行波比率以及回热器中的声阻抗的调制,并给出通过复声压调制后的系统内声场分布图。对比结果表明该声场调制关系式的适用性,通过双声源复声压的调节,能调制出谐振管及回热器所期望的工作声场。     

一种预测扩散声场在非规则形状刚性壁面附近干涉图样的方法

扩散声场会在反射边界附近形成干涉图样,研究方法包括平面波模型、简正模态分析、渐进模态分析等,但仅适用于尺度远大于声波波长的矩形声腔。提出一种预测扩散声场在非规则刚性壁面结构附近形成的干涉图样的数值方法,表明结构附近“受挡”声压的互谱矩阵取决于:(1)假定该结构在自由空间中振动辐射声音时其表面法向振速到表面及场点声压的边界元系数矩阵;(2)假定结构置于自由空间中且表面刚性时,点声源辐射声波入射到结构表面上产生的散射声场的边界元系数矩阵;(3)扩散声场均方声压。仿真表明,该途径预测的干涉图样与理论值完全吻合。该预测方法还可用于混响环境下声源附近直达声压均方值的空间分布估计,为混响环境下设备的声源定位提供帮助。      

声辐射问题中的模态分析:Ⅲ.声场重构

声辐射模态是定义在振动表面上的一组互相独立的基函数,描绘了振动表面的多极子辐射模式。声场分布模态是定义在声场中的一组互相独立的基函数,描绘了多极子辐射的声场分布模式。振动表面的声辐射可以唯地用声辐射模态和声场分布模态展开表示。对于声场重构,关键问题就是确定展开系数。在声场中选择N个测量点,测量声场的有关声学信息,求斛展开系数,从而重构整个声场。针对平板辐射和球形声源辐射的声场进行重构,数值计算的结果与理论分析结果一致。     

HELS法在循环平稳声场全息重建中的理论与实验研究

Helmholtz 方程最小二乘法利用一组球面波基函数拟合声源产生的声场,根据重建和实际声压的误差最小原则,利用最小二乘法确定基函数展开的项数以及对应的权重系数,该方法具有计算效率高和需要测点少的优点,在实际工程中有很大的实用性.Helmholtz 方程最小二乘法和其他近场声全息方法一样都是针对平稳声场,对非平稳声场的分析很少.对于实际工程中经常遇到的一类特殊非平稳声场——循环平稳声场,现有的技术多以单通道信号分析为主,其高阶统计量在故障诊断领域应用较广.分析了循环平稳声场中Helmholtz方程最小二乘 关键词： 声全息 循环平稳 Helmholtz 方程 球面波     

板结构的声辐射模态分析及其高分辨率法向振动速度重建

提出了一种基于声辐射模态的速度基向量构建方法,该速度基向量不受网格划分的影响,可用于高分辨率的板结构法向振动速度重建。首先对板表面稀疏网格的声辐射模态进行计算,再以声辐射模态和模态系数构建板法向振动速度分布的基向量,然后由声场测量声压求解基向量系数,最后由该系数和加密网格的速度基向量重建高分辨率的板法向振动速度分布。以简支板声源进行仿真计算,当测量声压信噪比为30 dB时,低频的法向振动速度重建误差最低可达3.7%;以固支板声源在消声室中进行实验验证,131.5 Hz振动频率下的重建误差低于7%。该方法实现了只需要少量声压测量点即可精确重建板声源更高分辨率的法向振动速度分布。      

封闭空间声场重构的多层等效源法

对于封闭空间内的多途反射声,传统的等效声源法将其等效为距离边界一定距离的单层等效声源体进行声场重构,然而等效源与边界的距离选取依据不确定。因此,为获得等效声源配置的最优距离,在等效声源法(ESM)的基础上构建多层等效声源,提出一种适用于封闭环境声场重构的多层等效声源法(MESM),并依据等效声源的空间分布的稀疏性来获得等效声源强度信息。首先给出多层等效源法的理论依据,其次通过数值计算以及实验测试两种方式对比验证了所提方法。数值结果表明:MESM相比于ESM可在600 Hz以上频段获得低5~10 dB左右的重构误差,但是200 Hz以下的低频重构误差会增加5 dB左右。实验结果表明:MESM可比ESM获得更低的重构误差。文章最后基于数值计算研究了所提方法的主要影响因素。研究表明:虽然MESM会比ESM耗费2倍的计算时间,但在整体频率范围内,MESM可在ESM基础上提升600 Hz以上的重构性能。另外,等效声源的层数和层内数目的改变不会影响声场重构性能,而当传声器数目较多、阵列位置随机、空间边界的吸声系数不是很大时,MESM可获得比ESM更低重构误差,特别是600 Hz以上的中频段区间。      

Structure normal velocity reconstruction withsparse measurement points

To achieve normal velocity reconstruction of a vibrating surface with sparse measurement points, a reconstruction method is proposed by exploiting of acoustic radiation modes as expansion functions, which are capable of describing the geometric shape of a vibrating surface. Firstly, acoustic radiation modes of the vibrating surface are calculated and the relationship between normal velocity and acoustic radiation modes is built. Then actual measured normal velocity values are expressed by corresponding acoustic radiation modes and the expansion coefficients are calculated. Subsequently, all normal velocity values can be reconstructed by the obtained expansion coefficients. Experimental validations have been performed by a double-layer steel cylindrical shell with enclosed ends in.an anechoic water tank. Two cases with different wavenumber components distribution were designed by a vibration shaker and a rotor device respectively. Two experimental results both show that actual vibration distribution cannot be revealed exactly by the sparse measurement points, which corresponds to severe loss of vibration related wavenumber components. On the other hand, normal velocity and corresponding wavenumber components can be restored accurately in both two wavenumber components distribution cases according to the proposed method, which demonstrates obvious effectiveness of the proposed method.     

On the sound field of a shallow spherical shell in an infinite baffle

A method is presented for calculating the far field sound radiation from a shallow spherical shell in an acoustic medium. The shell has a concentrated ring mass boundary condition at its perimeter representing a loudspeaker voice coil and is excited by a concentrated ring force exerted by the end of the voice coil. A Green's function is developed for a shallow spherical shell, which is based upon Reissner's solution to the shell wave equation [Q. Appl. Math. 13, 279-290 (1955)]. The shell is then coupled to the surrounding acoustic medium using an eigenfunction expansion, with unknown coefficients, for its deflection. The resulting surface pressure distribution is solved using the King integral together with the free space Green's function in cylindrical coordinates. In order to eliminate the need for numerical integration, the radiation (coupling) integrals are solved analytically to yield fast converging expansions. Hence, a set of simultaneous equations is obtained which is solved for the coefficients of the eigenfunction expansion. These coefficients are finally used in formulas for the far field sound radiation.     

Axial acoustic radiation force on an elastic spherical shell near an impedance boundary for zero-order quasi-Bessel–Gauss beam

Shell structures have increasingly widespread applications in biomedical ultrasound fields such as contrast agents and drug delivery,which requires the precise prediction of the acoustic radiation force under various circumstances to improve the system efficiency.The acoustic radiation force exerted by a zero-order quasi-Bessel-Gauss beam on an elastic spherical shell near an impedance boundary is theoretically and numerically studied in this study.By means of the finite series method and the image theory,a zero-order quasi-Bessel-Gauss beam is expanded in terms of spherical harmonic functions,and the exact solution of the acoustic radiation force is derived based on the acoustic scattering theory.The acoustic radiation force function,which represents the radiation force per unit energy density and per unit cross-sectional surface,is especially investigated.Some simulated results for a polymethyl methacrylate shell and an aluminum shell are provided to illustrate the behavior of acoustic radiation force in this case.The simulated results show the oscillatory property and the negative radiation force caused by the impedance boundary.An appropriate relative thickness of the shell can generate sharp peaks for a polymethyl methacrylate shell.Strong radiation force can be obtained at small half-cone angles and the beam waist only affects the results at high frequencies.Considering that the quasi-Bessel-Gauss beam possesses both the energy focusing property and the non-diffracting advantage,this study is expected to be useful in the development of acoustic tweezers,contrast agent micro-shells,and drug delivery applications.     

Radiation impedance loading of a spherical source in a two-dimensional perfect acoustic waveguide

We employ the classical method of separation of variables in combination with the method of images and the translational addition theorems for spherical wave functions to investigate the acoustic radiation load on a finite-sized spherical source which is submerged at an arbitrary point within a plane-parallel sound channel with absolutely reflecting boundaries. To simulate an idealized situation, it is supposed that the fluid layer is homogeneous and is bounded below by an acoustically hard surface and above by a rigid/compliant boundary. Subsequently, the resistive and the reactive components of the modal acoustic radiation impedance load as a function of source position for a pulsating (n = 0) and an oscillating (n = 1) sphere at selected nondimensional frequencies are calculated and discussed. The presented benchmark solution can lead to a better understanding of the acoustics of waveguide sources (transducers) that are of practical interest in underwater acoustics and ocean engineering. It could eventually be used to validate those found by numerical approximation techniques. The text was submitted by the authors in English.     

AN ANALYTICAL INVESTIGATION OF THE DIRECT MEASUREMENT METHOD OF ESTIMATING THE ACOUSTIC IMPEDANCE OF A TIME-VARYING SOURCE

This paper presents an analytical investigation of the direct method of measurement of the source impedance of a linear time-variant source. The direct method yields a frequency-dependent effective source impedance which is routinely used in a time-invariant analysis to determine the insertion loss of two different acoustic loads applied to the same source. In such an analysis the strength of the source is assumed to be invariant with load. It is shown here that there is generally no precise correspondence between the effective source impedance as given by the direct method and the characteristics of the actual source. Furthermore, it is shown that the effective source impedance values given by the direct method are functions of the acoustic load and the location of the injected signal as used in the measurement. However, the effective source resistance is always found to be positive, in accordance with experimental measurements. In this regard the direct method is an improvement on the indirect method, where physically implausible negative resistance values are often found. Finally, it is shown that the effective impedance values as given by the direct method when used with a time-invariant analysis give rise to very accurate predictions of insertion loss, even when the strength of the actual time-variant source is allowed to vary with the acoustic load.     

Visualization of acoustic radiation from a vibrating bowling ball.

This paper presents visualization of acoustic radiation from a vibrating bowling ball using the Helmholtz equation least squares (HELS) method. In conducting the experiments, the ball is excited by a vibration shaker using stationary random signals. The radiated acoustic pressures are measured using two microphones and taken as input to the HELS formulations. The reconstructed acoustic pressures on the bowling ball surface are compared with those measured at the same locations. Also shown are comparisons of the reconstructed and measured acoustic pressure spectra at various locations on the bowling ball surface. Results demonstrate that the accuracy of reconstruction based on measurements over a conformal surface is much higher than that over a finite planar surface. This is because the latter often extends beyond the near-field region, making the accuracy of measurements inconsistent. Nevertheless, satisfactory reconstruction of acoustic pressure fields over the entire bowling ball surface can still be obtained based on the measurements taken over a finite planar surface on one side of the source. In a similar manner, the normal component of the surface velocity is reconstructed. Once these acoustic quantities are determined, the time-averaged acoustic intensity is calculated. Also presented are the formulations for estimating a priori the numbers of expansion functions and measurements required by the HELS method and the guidelines for determining the reconstruction error and optimum measurement locations, given the overall dimensions of the source and the highest frequency of interest in reconstruction.     

On the choice of expansion functions in the Helmholtz equation least-squares method

This paper examines the performance of Helmholtz equation least-squares (HELS) method in reconstructing acoustic radiation from an arbitrary source by using three different expansions, namely, localized spherical waves (LSW), distributed spherical waves (DSW), and distributed point sources (DPS), under the same set of measurements. The reconstructed acoustic pressures are validated against the benchmark data measured at the same locations as reconstruction points for frequencies up to 3275 Hz. Reconstruction is obtained by using Tikhonov regularization or its modification with the regularization parameter selected by error-free parameter-choice methods. The impact of the number of measurement points on the resultant reconstruction accuracy under different expansion functions is investigated. Results demonstrate that DSW leads to a better-conditioned transfer matrix, yields more accurate reconstruction than both LSW and DPS, and is not affected as much by the change in measurement points. Also, it is possible to obtain optimal locations of the auxiliary sources for DSW, LSW, and DPS by taking an independent layer of measurements. Use of these auxiliary sources and an optimal combination of regularization and error-free parameter choice methods can yield a satisfactory reconstruction of acoustic quantities on the source surfaces as well as in the field in the most cost-effective manner.     

稀疏测点条件下的结构法向速度重建

为了实现稀疏测点条件下结构表面法向速度的准确重建,利用声辐射模态包含结构表面几何形状信息的性质,以声辐射模态作为基函数提出了一种稀疏测点条件下的结构法向振速重建方法。首先对结构表面的声辐射模态进行计算,并建立结构表面法向振动速度与声辐射模态之间的关系;在此基础上,由实际布置情况形成测点位置处的振速与其声辐射模态值的关系,并通过最小二乘法求得展开系数;最后由展开系数重建出结构表面的全部法向振速。利用两端封闭的双层钢质圆柱壳体在消声水池中进行了试验验证,分别开启激振器和转子台进行激励,两个试验的结果均表明,当测点数目较少时,所测的结果不能准确地表示结构的实际振动情况,在波数域内就表现为与振动相关的波数成份的丢失;利用所提出的方法,可以较为准确地重建和恢复结构的表面法向速度及其波数成份,由此验证了所提出方法的有效性。      

高斯波束对界面附近离轴球形粒子的轴向声辐射力

从声波的散射理论出发，利用级数展开法得到高斯波束的波束因子，推导其对阻抗边界下离轴球形粒子声辐射力.针对刚性球与液体球两种球形粒子进行数值模拟，与自由空间的情况进行比较.讨论边界反射系数、粒子与边界距离、束腰半径以及离轴角度与距离等对声辐射力的影响.仿真结果表明：边界反射系数的增大会引起声辐射力的增加，但不改变峰值的位置；在合适的频率处，可以产生负向声辐射力；声辐射力随粒子与边界距离呈周期性变化；束腰半径的影响主要体现在中高频；随着粒子偏离传播轴的距离和角度增大，声辐射力明显衰减.该研究为利用高斯波束实现对粒子的操纵提供理论基础.