线性与非线性流固耦合动力学数值方法的进展及应用（英文）

本文综述了线性与非线性流固耦合问题数值方法的进展及工程应用.讨论了四种数值分析方法:(1)混合有限元–子结构–子区域数值模型,以求解有限域线性流固耦合问题,如流体晃动,声腔–结构耦合,流体中的压力波,化工容器的地震响应,坝水耦合等;(2)混合有限元–边界元数值模型,以求解涉及无限域的线性流固耦合问题,如大型浮体承受飞机降落冲击,船舰的炮击回应等;(3)混合有限元–有限差分(体积)数值模型,以求解不涉及破浪和两相分离的非线性流固耦合问题;(4)混合有限元–光滑粒子数值模型,以求解涉及破浪和两相分离的非线性流固耦合问题.文中推荐分区迭代求解过程,以便应用现有的固体及流体求解器,于毎一时间步长分别求解固体及流体的方程,通过耦合迭代收敛,向前推进以达问题求解.文中选用的工程应用例子包含气–液–壳三相耦合,液化天然气船水晃动,人体步行冲击引起的声腔–建筑结构耦合,大型浮体承受飞机降落冲击的瞬态动力回应,涉及破浪和两相分离的气–翼耦合及结构于水上降落的冲击.数值分析结果与可用的实验或计算结果作了比较,以说明所述方法的精度及工程应用价值.文中列出了基于流固耦合的波能采积装置模型,以应用线性系统的共振及非线性系统的周期解原理,有效地采积波能.本文列出了231篇参考文献,以便读者进一步研讨所感兴趣方法.     

Developments of numerical methods for linear and nonlinear fluid-solid interaction dynamics with applications

本文综述了线性与非线性流固耦合问题数值方法的进展及工程应用. 讨论了四种数值分析方法: (1) 混合有限元-子结构-子区域数值模型, 以求解有限域线性流固耦合问题, 如流体晃动, 声腔-结构耦合, 流体中的压力波, 化工容器的地震响应,坝水耦合等; (2) 混合有限元-边界元数值模型, 以求解涉及无限域的线性流固耦合问题, 如大型浮体承受飞机降落冲击, 船舰的炮击回应等; (3) 混合有限元-有限差分(体积) 数值模型, 以求解不涉及破浪和两相分离的非线性流固耦合问题; (4) 混合有限元-光滑粒子数值模型, 以求解涉及破浪和两相分离的非线性流固耦合问题. 文中推荐分区迭代求解过程, 以便应用现有的固体及流体求解器, 于毎一时间步长分别求解固体及流体的方程, 通过耦合迭代收敛, 向前推进以达问题求解. 文中选用的工程应用例子包含气-液-壳三相耦合, 液化天然气船水晃动, 人体步行冲击引起的声腔-建筑结构耦合, 大型浮体承受飞机降落冲击的瞬态动力回应, 涉及破浪和两相分离的气-翼耦合及结构于水上降落的冲击. 数值分析结果与可用的实验或计算结果作了比较, 以说明所述方法的精度及工程应用价值. 文中列出了基于流固耦合的波能采积装置模型, 以应用线性系统的共振及非线性系统的周期解原理, 有效地采积波能. 本文列出了231 篇参考文献, 以便读者进一步研讨所感兴趣方法.     

虚拟区域法在流固耦合问题中的应用

将多相流领域内的虚拟区域法引入到流固耦合问题的分析中，将固体视为应变率为 零的虚拟流体，对流体和虚拟流体均以速度和压强作为基本变量，采用Navier-Stokes方 程作为控制方程，同时求解流体域和虚拟流体域, 得到整个计算域的流场分布，应用分布式拉 格朗日乘子法在虚拟流体域上施加刚体约束, 以保持虚拟流体的刚体外形和运动形式，最终建 立一种流固耦合模型及其数值求解方法. 通过对粒子流问题和流固耦合问题进行数值模拟， 验证了此模型的正确性和求解大变形/运动流固耦合问题的有效性.     

液饱和多孔介质中三维应力波的传播

为了深入研究液饱和多孔介质中应力波的传播，提出了三维两相细观计算模型．基于此模型。应用Galerkin余量法并计及流-固耦合界面的耦合效应，利用直接耦合的技术，开发了三维流-固混合显式动力有限元计算程序．在此基础上对冲击载荷作用下液饱和多孔介质中三维应力波的传播现象进行了数值模拟，并详细讨论了孔隙率，孔隙形状等因素对应力波传播主导波形的影响．     

非线性流体-刚体结构相互作用问题的一种数值模拟方法

给出了一种模拟非线性流体-刚体结构相互作用问题的数值方法．文中假定结构承受大的刚体运动,流体流动受非线性有粘或无粘的场方程支配并满足自由表面和两相耦合界面上的非线性边界条件,利用任意拉氏-欧氏（ALE）网格系统构造了数值模型．采用所探讨的多块数值格式,允许可动重造网格间有独立的相对运动,从而克服了流体网格与固体大运动匹配的困难．通过数值离散化,导出了描述非线性流固耦合动力学的数值方程并应用耦合迭代过程对其作了求解．通过算例,说明了所提出数值模型的应用．     

流固耦合破坏分析的多分辨率PD-SPH方法

流固耦合破坏是一类涉及结构变形与破坏以及复杂自由表面现象的强非线性力学问题. 结合近场动力学(peridynamics, PD)与光滑粒子流体动力学(smoothed particle hydrodynamics, SPH)各自的优势并考虑其计算效率问题, 提出一种适用于分析流?固耦合破坏问题的多分辨率PD-SPH混合方法. 分别采用SPH和PD方法以不同的空间和时间分辨率对流体和结构进行离散与求解, 利用具有与流体粒子相同光滑长度的虚粒子处理流?固界面, 以高精度满足界面边界条件. 通过两个经典算例: 液柱静压力下弹性板的变形和溃坝流体冲击弹性闸门的变形问题, 表明提出的多分辨率PD-SPH方法兼具较高的计算精度和计算效率; 对含裂缝的Koyna重力坝水力劈裂问题进行模拟计算, 所得裂缝扩展路径与文献结果吻合, 说明该方法适用于涉及结构破坏的流固耦合问题仿真. 最后尝试采用该方法进行流体冲击作用下含裂纹混凝土板崩塌过程数值仿真, 准确描述混凝土板的断裂破坏和全过程中的流体运动. 多分辨率PD-SPH混合方法或可为流?固耦合作用下的结构损伤破坏仿真提供一种新选择.      

生物芯片压电微流体泵液-固耦合系统模态分析

对压电微流体泵粘性流体周期流动进行厚度积分平均近似,得到包含粘性的,非线性浅水波动方程,并采用有限元法得到微泵液体压强矩阵方程．液体压强矩阵方程和压电硅片振动有限元方程耦合,得到一个包含微泵进出口扩散管的液-固耦合系统振动方程．液-固耦合系统的模态分析结果表明,做泵液-固耦合系统的自然频率比不耦合的硅片振动自然频率低很多．随着微泵厚度的减少,液体附加质量和粘性阻尼对耦合系统自然频率的影响更加明显．同时发现,对应的压电片振型函数在液-固耦合前后没有明显变化,还给出硅片-阶模态的振幅-频率特征曲线,对薄型无阀压电微流体泵,浅水波模型合理地表达了微泵液体流动和压电硅片振动的相互作用,以及液体附加质量和粘性阻尼对微泵液-固耦合系统动力特征的影响。     

面向对象的土石坝参数随机反演程序设计

将储液容器流固耦合系统中的液体和容器分别视为理想可压缩流体和线弹 性固体，采用流体压力单元和固体壳单元对流固耦合系统进行有限元离散，得到一个非对称 的大型流固耦合有限元方程. 采用Arnoldi方法求解上面这个大型有限元方程的非对称特征 值问题，以得到储液容器的动力特性. 通过移频技术避免了处理零频问题，并构造了迭代格 式计算Arnoldi向量. 数值算例表明所用解法对于流固耦合系统都是非常有效的.     

孔隙流体耦合效应对射流冲击应力分布的影响分析

运用全解耦流固耦合理论,建立了水射流冲击岩石介质流固耦合数值分析模型,给出了数值算法,计算分析了考虑和不考虑孔隙流体耦合效应对射流冲击岩石时应力分布的影响规律。结果表明,在射流冲击作用下,如不考虑孔隙流体耦合作用,最大拉应力位于冲击面,离冲击中心径向距离与喷距成正比,最大剪切应力位于岩石冲击中心下部约0.5倍喷嘴直径位置;如考虑孔隙流体耦合作用,最大拉应力位于岩石冲击中心下部约0.4倍喷嘴直径位置。数值分析结果可为水射流破岩机理研究中岩石破坏准则的选择提供依据。     

求解三维流固耦合问题的一种全隐全耦合区域分解并行算法

三维流固耦合问题的非结构网格数值算法在很多工程领域都有重要应用, 目前现有的数值方法主要基于分区算法, 即流体和固体区域分别进行求解, 因此存在收敛速度较慢以及附加质量导致的稳定性问题, 此外, 该类算法的并行可扩展性不高, 在大规模应用计算方面也受到一定限制．本文针对三维非定常流固耦合问题, 提出一种基于区域分解的全隐全耦合可扩展并行算法．首先基于任意拉格朗日?欧拉框架建立流固耦合控制方程, 然后时间方向采用二阶向后差分隐式格式、空间方向采用非结构稳定化有限元方法进行离散．对于大规模非线性离散系统, 构造一种结合非精确Newton法、Krylov子空间迭代法与区域分解Schwarz预条件子的Newton-Krylov-Schwarz (NKS) 并行求解算法, 实现流体、固体和动网格方程的一次性整体求解．采用弹性障碍物绕流的标准测试算例对数值方法的准确性进行了验证, 数值性能测试结果显示本文构造的全隐全耦合算法具有良好的稳定性, 在不同的物理参数下具有良好的鲁棒性, 在“天河二号”超级计算机上, 当并行规模从192增加到3072个处理器核时获得了91%的并行效率．性能测试结果表明本文构造的NKS算法有望应用于复杂区域流固耦合问题的大规模数值模拟研究中．      

A new numerical technique for the analysis of parametrically excited nonlinear systems

A new computational scheme using Chebyshev polynomials is proposed for the numerical solution of parametrically excited nonlinear systems. The state vector and the periodic coefficients are expanded in Chebyshev polynomials and an integral equation suitable for a Picard-type iteration is formulated. A Chebyshev collocation is applied to the integral with the nonlinearities reducing the problem to the solution of a set of linear algebraic equations in each iteration. The method is equally applicable for nonlinear systems which are represented in state-space form or by a set of second-order differential equations. The proposed technique is found to duplicate the periodic, multi-periodic and chaotic solutions of a parametrically excited system obtained previously using the conventional numerical integration schemes with comparable CPU times. The technique does not require the inversion of the mass matrix in the case of multi degree-of-freedom systems. The present method is also shown to offer significant computational conveniences over the conventional numerical integration routines when used in a scheme for the direct determination of periodic solutions. Of course, the technique is also applicable to non-parametrically excited nonlinear systems as well.     

Numerical investigation of ice breaking by a high-pressure bubble based on a coupled BEM-PD model

In this paper, by combining the boundary element method (BEM) and peridynamics (PD), a bubble-ice interaction model is established, which can investigate the dynamic interactions between a high-pressure bubble and an ice plate with particular focus on the mechanical behaviors of ice breaking. The bubble dynamics are solved by BEM based on the potential flow theory. Ice cracks initiation and propagation are simulated by the bond-based peridynamics which is validated by a three-point bending test. The fluid–structure interaction (FSI) is achieved by matching the normal velocity and hydrodynamic loads at the fluid–structure interface. To validate the proposed FSI model, an experiment is carried out in which an oscillating bubble is generated under an ice plate by underwater discharge system. The whole interaction process is captured by a Phantom V711 high-speed camera. Qualitative agreements are achieved between the numerical and experimental results. The underlying mechanism of cracks initiation, propagation, branching, and coalescence of the ice plate is found to highly depend on three parameters, i.e., bubble–ice distance, ice thickness and bubble size. The present study is expected to provide further assists in the understanding of ice breaking problems.     

Sloshing simulation of standing wave with time-independent finite difference method for Euler equations

The numerical solutions of standing waves for Euler equations with the nonlinear free surface boundary condition in a two-dimensional(2D) tank are studied.The irregular tank is mapped onto a fixed square domain through proper mapping functions. A staggered mesh system is employed in a 2D tank to calculate the elevation of the transient fluid.A time-independent finite difference method,which is developed by Bang-fuh Chen,is used to solve the Euler equations for incompressible and inviscid fluids.The numer...     

Internal waves in a two‐layer system using fully nonlinear internal‐wave equations

In order to understand the nonlinear effect in a two‐layer system, fully nonlinear strongly dispersive internal‐wave equations, based on a variational principle, were proposed in this study. A simple iteration method was used to solve the internal‐wave equations in order to solve the equations stably. The applicability of the proposed numerical computation scheme was confirmed to agree with linear dispersion relation theoretically obtained from variational principle. The proposed computational scheme was also shown to reproduce internal waves including higher‐order nonlinear effect from the analysis of internal solitary waves in a two‐layer system. Furthermore, for the second‐order numerical analysis, the balance of nonlinearity and dispersion was found to be similar to the balance assumed in the KdV theory and the Boussinesq‐type equations. Copyright © 2009 John Wiley & Sons, Ltd.     

Numerical study of flow-induced vibrations of cylinders under the action of nonlinear energy sinks (NESs)

An isolated two-dimensional circular cylinder with two linear degrees of freedom, parallel and perpendicular to the free-stream direction, and owning a nonlinear energy sink (NES) is investigated by fluid–structure interaction (FSI) simulations to assess vortex-induced vibrations (VIV) at moderate Reynolds numbers. Subsequently, the wake-induced vibration (WIV) of a pair of identical cylinders under the action of two NES in a tandem arrangement and in a proximity–wake interference regime is explored using the same approach. The NES parameters (mass, nonlinear stiffness and damping) are investigated to determine their effects on the dynamic response of a single degree of freedom (in transverse flow direction) coupled system by a reduced-order model based on an experimentally validated van der Pol oscillator. The CFD model coupled with FSI method is also validated against VIV experimental data for an isolated cylinder in a uniform flow. The study is aimed to investigate the effect of the passive suppression NES device on VIV and WIV. The amplitude response, trajectories of cylinder motion and temporal evolutions of vortex shedding are obtained by conducting a series of numerical simulations. It is found that placing a tuned NES in the cylinders can provide good suppression effect; however, the effectiveness is function of the reduced velocity.     

二维定常不可压缩粘性流动N-S方程的数值流形方法

将流形方法应用于定常不可压缩粘性流动N-S方程的直接数值求解,建立基于Galerkin加权余量法的N-S方程数值流形格式,有限覆盖系统采用混合覆盖形式,即速度分量取1阶和压力取0阶多项式覆盖函数,非线性流形方程组采用直接线性化交替迭代方法和Nowton-Raphson迭代方法进行求解.将混合覆盖的四节点矩形流形单元用于阶梯流和方腔驱动流动的数值算例,以较少单元获得的数值解与经典数值解十分吻合.数值实验证明,流形方法是求解定常不可压缩粘性流动N-S方程有效的高精度数值方法.     

Soil stress and deformation determination under a landing airplane on an unsurfaced airfield

This paper presents a multidisciplinary approach to a problem of soil–wheel interaction during the landing of an airplane on a grass airfield, with terramechanical analysis of the forces acting between a wheel and the surface. The experiment on stress and deformation state in soil under a wheel on touchdown was performed in the field. The soil stress state (major stress S1, minor stresses S2 and S3, stress invariant in an octahedral system, MNS and OCTSS) was determined with the use of an SST (stress state transducer), which was installed at 15 cm depth. Soil deformation was determined with an inertial measuring system, integrated with the SST. These soil data were captured exactly at the moment touchdown. In the experiment, a four-passenger, STOL multirole airplane was used and pilots performed two landing patterns: normal landing and emergency landing (without propeller thrust).     

基于ALE方法的飞机水面降落过程

为了研究飞机水面降落过程中的动态响应规律,采用arbitrary Lagrangian-Eulerian(ALE)方法开展了飞机水面降落的动力学分析。研究了飞机水面降落过程的速度变化规律,结果表明,降落速度和飞行速度在飞机入水的初始阶段变化较快,随后变化幅度趋于平缓。分析了不同降落速度、飞行速度和降落仰角下的机身结构响应,得到了飞机水面降落时结构响应随时间的变化规律。机身结构应力在入水的初始阶段达到最大值,随后迅速下降,最后保持稳定。飞机结构的最大变形也出现在入水的初始阶段,随后迅速回复到初始状态。对比了降落速度、飞行速度和降落仰角对飞机结构响应的影响程度,结果表明降落速度对结构响应的影响程度最大,降落仰角次之,飞行速度的影响最小。 更多还原