Cheng Lei, White Robert D, Grosh Karl
Department of Mechanical Engineering, 2350 Hayward Avenue, University of Michigan, Ann Arbor, MI 48109-2125, USA.
Comput Methods Appl Mech Eng. 2008 Sep 15;197(49-50):4160-4172. doi: 10.1016/j.cma.2008.04.016.
A three dimensional viscous finite element model is presented in this paper for the analysis of the acoustic fluid structure interaction systems including, but not limited to, the cochlear-based transducers. The model consists of a three dimensional viscous acoustic fluid medium interacting with a two dimensional flat structure domain. The fluid field is governed by the linearized Navier-Stokes equation with the fluid displacements and the pressure chosen as independent variables. The mixed displacement/pressure based formulation is used in the fluid field in order to alleviate the locking in the nearly incompressible fluid. The structure is modeled as a Mindlin plate with or without residual stress. The Hinton-Huang's 9-noded Lagrangian plate element is chosen in order to be compatible with 27/4 u/p fluid elements. The results from the full 3d FEM model are in good agreement with experimental results and other FEM results including Beltman's thin film viscoacoustic element [2] and two and half dimensional inviscid elements [21]. Although it is computationally expensive, it provides a benchmark solution for other numerical models or approximations to compare to besides experiments and it is capable of modeling any irregular geometries and material properties while other numerical models may not be applicable.
本文提出了一种三维粘性有限元模型,用于分析声流体结构相互作用系统,包括但不限于基于耳蜗的换能器。该模型由与二维平面结构域相互作用的三维粘性声流体介质组成。流场由以流体位移和压力为自变量的线性化纳维-斯托克斯方程控制。为了减轻近不可压缩流体中的锁死现象,在流场中采用了基于混合位移/压力的公式。结构被建模为有或无残余应力的明德林板。选择辛顿-黄的9节点拉格朗日板单元,以便与27/4 u/p流体单元兼容。全三维有限元模型的结果与实验结果以及其他有限元结果(包括贝尔特曼的薄膜粘声单元[2]和二维半无粘单元[21])吻合良好。尽管其计算成本高昂,但它为除实验之外的其他数值模型或近似方法提供了一个基准解,并且能够对任何不规则几何形状和材料特性进行建模,而其他数值模型可能并不适用。