首页> 外文学位 >Numerical methods for solving the wave equation in large enclosures with application to room acoustics.
【24h】

Numerical methods for solving the wave equation in large enclosures with application to room acoustics.

机译:求解大型机壳中波动方程的数值方法,并应用于室内声学。

获取原文
获取原文并翻译 | 示例

摘要

Many acoustical events in our everyday lives occur in enclosures, in which echoes and reverberation impact auditory perception in many ways. Numerical models play an important role in analyzing the acoustic behavior in reverberant environments, as they allow systematic control of physical parameters that affect human perception. Most numerical models used in architectural and room acoustics are based on the ray acoustics approximation, which leads to reasonable computational costs, at the expense of excluding certain important wave phenomena such as diffraction. In order to obtain more accurate results, an approach based on solving the wave equation is appropriate. However, standard numerical methods for solving the wave equation are not practical for room acoustics applications, since these problems are acoustically large and incur prohibitively large computational costs. Even in a small room, a sound wave with audible high-frequency content must propagate for about 10,000 wavelengths before it decays to an inaudible level. To address this, three new, efficient ways of simulating acoustic responses in a room are developed in this study. First, a method for calculating the resonant frequencies and normal modes in a rectangular room with arbitrary wall impedance is developed that uses the interval Newton/generalized bisection (IN/GB) method for solving the acoustic eigenvalue equation. The second approach applies the finite element method in the frequency domain, but uses a Dirichlet-to-Neumann (DtN) map to model empty, rectangular portions of the room, thereby truncating the effective computational domain. Finally, a finite difference method with minimal dispersion and dissipation errors is developed in the time domain. The parameters in the discretization in both space and time are optimized to minimize these errors. This method has been implemented on the IBM Blue Gene platform at Boston University, and allows for the calculation of the impulse response in a practical size of room (3 m x 3 m x 3 m) at relatively high frequencies (5 kHz) for the entire duration of the reverberation (1--3 seconds). Initial results indicate that this methodology has the ability to serve as a tool for conducting psychoacoustic experiments in reverberant spaces.
机译:我们日常生活中的许多声音事件都发生在封闭空间中,其中回声和混响会以多种方式影响听觉。数值模型在分析混响环境中的声学行为中起着重要作用,因为它们可以对影响人类感知的物理参数进行系统控制。建筑和室内声学中使用的大多数数值模型都是基于射线声学近似,这导致合理的计算成本,但要以排除某些重要的波现象(例如衍射)为代价。为了获得更准确的结果,基于求解波动方程的方法是合适的。但是,用于求解波动方程的标准数值方法在室内声学应用中不切实际,因为这些问题在声学上很大,并且会导致计算成本过高。即使在很小的房间中,具有可听高频成分的声波也必须传播大约10,000个波长,然后才能衰减到听不见的水平。为了解决这个问题,本研究开发了三种新的,有效的模拟房间声学响应的方法。首先,开发了一种用于计算具有任意壁阻抗的矩形房间中的共振频率和法向模的方法,该方法使用区间牛顿/广义二等分(IN / GB)方法求解声学特征值方程。第二种方法在频域中应用了有限元方法,但是使用Dirichlet-to-Neumann(DtN)映射对房间的空矩形部分建模,从而舍弃了有效的计算域。最后,在时域中开发了一种色散和耗散误差最小的有限差分方法。优化了离散化在时间和空间上的参数,以最大程度地减少这些误差。此方法已在波士顿大学的IBM Blue Gene平台上实现,并且允许在整个持续时间内以相对高的频率(5 kHz)在实际房间大小(3 mx 3 mx 3 m)中计算脉冲响应。混响的时间(1--3秒)。初步结果表明,该方法具有作为在混响空间中进行心理声学实验的工具的能力。

著录项

  • 作者

    Naka, Yusuke.;

  • 作者单位

    Boston University.;

  • 授予单位 Boston University.;
  • 学科 Engineering Mechanical.; Physics Acoustics.
  • 学位 Ph.D.
  • 年度 2007
  • 页码 155 p.
  • 总页数 155
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类 机械、仪表工业;声学;
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号