| 周益清,骆文于,吴双林.双层介质声传播问题的标准解及有限元解[J].声学技术,2022,41(2):149~159 |
| 双层介质声传播问题的标准解及有限元解 |
| Benchmark solution and finite element solution of sound propagation in a two-layer medium |
| 投稿时间：2020-12-17 修订日期：2021-01-25 |
| DOI：10.16300/j.cnki.1000-3630.2022.02.001 |
| 中文关键词: 波数积分方法 有限元方法 水下声传播 简正波方法 |
| 英文关键词: wavenumber integration method finite element method underwater sound propagation normal mode method |
| 基金项目:国家重点研发计划项目资助（2018YFC0308600），国家自然科学基金（11774374）资助项目。 |
|
| 摘要点击次数: 713 |
| 全文下载次数: 394 |
| 中文摘要: |
| 文章通过双层介质中的声传播问题，研究了有限元方法在水下声场计算中的应用。基于传统的Galerkin方法推导出水下声场的有限元方程，采用四节点四边形单元离散求解物理域，可选择辐射边界条件、DtN (Dirichlet toNeumann)非局部算子、完美匹配层来处理出射声场，得到有限元解。为了验证该有限元模型，需要高精度的参考解。水平不变均匀介质中的声传播问题存在解析解，但双层介质问题不存在解析解。因此，对于双层介质声传播问题，使用波数积分法推导出标准解。分别考虑了有限深度和无限深度双层介质两种情况，并进行了数值模拟。数值结果表明，文章所提的有限元模型与参考解非常吻合。此外，还发现当某号简正波的本征值非常接近割线时，简正波模型KRAKEN难以准确计算该号简正波的本征值，从而声场计算结果存在明显误差；但是有限元方法不需要计算本征值，所以当KRAKEN模型出现此类问题时，有限元方法仍能给出准确的声场计算结果，表明有限元方法在普适性方面优于简正波方法。 |
| 英文摘要: |
| In this paper, the sound propagation problem in a two-layer medium is studied and a finite element model for un-derwater sound field calculation is proposed. The finite element equation is derived by the traditional Galerkin method, and the four-node quadrilateral element is used to discretize the computation domain. Then the traditional radiation boundary condition, the DtN (Dirichlet to Neumann) non-local operator and perfectly matched layer are adopted to process the outgoing sound field and obtain the finite element solutions, and the differences between them are compared. In order to verify the finite element model, a benchmark solution is needed. Analytical solutions of sound propagation exist in a homogeneous medium. However, in a two-layer medium, there is no analytical solu-tion. Therefore, for the sound field calculation in a two-layer medium, the wavenumber integration method is used to provide benchmark solutions. The numerical simulations of sound field in a two-layer medium with limited and unlimited depths are conducted respectively. The results show that the finite element model is in good agreement with the benchmark solution. In addition, it is found that the normal mode model KRAKEN may be difficult to calculate the eigenvalue of the normal mode accurately when the eigenvalue of a normal mode is very close to the secant, so the error of sound field calculation result is significant; but the finite element method does not need to calculate the eigenvalues. Thus, the finite element model proposed in this paper can solve the sound field calculation problem where the KRAKEN model is hard to solve accurately, which means that the practicability of finite element method is better than that of the normal mode method. |
|
查看全文
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
