流体-结构耦合问题的有限元并行计算研究

流体-结构耦合问题广泛存在于各种工程领域,本文采用ALE显式有限元法求解该类问题,并对该方法的并行性进行讨论。同时根据流体-结构耦合问题与ALE显式有限元的计算特点,在坐标递归分区方法的基础上设计并程序实现了基于流体-结构耦合均衡的分区算法。通过与坐标递归分区方法的计算结果相比较,对于流体-结构耦合问题的求解,耦合均衡并行分区方法具有更好的加速比和并行效率。     

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.     

受谐激励的带压电层充液圆柱容器稳态解

基于线性压电理论和可压缩无粘流场运动方程,推导出无限长带压电圆柱体流固耦合稳态解,由于流固耦合与力电耦合的复杂性,文中只考虑轴对称问题,研究了流固耦合智能结构在不同电压作用下位移、应力、流体压力的分布情况,为振动噪声控制奠定基础。     

基于拓扑优化的声学结构材料分布设计

本文针对结构的声学设计问题进行研究,通过优化两种不同的材料在结构设计域内的拓扑分 布来最小化谐振结构所产生的声场中指定参考面/参考域内的声压。在研究中假定结构为线弹性小变 形结构,材料阻尼为Rayleigh阻尼,声学介质为无粘、可压缩、小扰动流体。对结构响应采用有限 元格式进行计算,对声场采用基于Helmholtz积分的边界元格式进行计算,由于声场在无穷远自由边 界的无反射条件在边界积分中能自动得到满足,该格式特别适合于具有开放边界的声场计算。建立 了结构有限元-声场边界元格式的耦合系统拓扑优化模型,导出了耦合系统敏感度分析的一般格式及 伴随格式。数值算例验证了所提出的结构-声学耦合系统优化方法的有效性和可靠性,并揭示了基于 声学准则的拓扑优化结果的有关特性 关键词边界积分,结构声学耦合系统,拓扑优化,敏感度分析,伴随方法     

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．     

结构物在水下运行过程中的自振特性及动力响应研究

本文运用流体力学、结构动力学及变分原理推导了结构－流体耦合系统的动力学方程，并对水下结构物在运行时的自振特征及动力响应给出了数值计算分析。     

数值流形方法在流固耦合谐振分析中的应用

数值流形方法(流形法)是石根华博士利用现代数学中流形分析的有限覆盖技术建立起来的新的数值分析方法,统一解决了连续和非连续变形的力学问题,具有广阔的应用前景。本文将流形法应用于交界面耦合的流固振动分析,采用平面矩形数学网格,针对无粘、无旋、不可压缩流体和无阻尼的固体结构,提出分析流固耦合系统简谐振动的高阶流形法公式,其中,采用拉格朗日乘子法引入流场的已知边界条件。本文还初步研究了在特殊的无限远流场中采用解析解覆盖函数的实现技术。文中算例体现了流形法网格划分的方便性和计算的高精度,显示出流形法在数值解和解析解联合运用上的优势。     

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.      

流体与结构相互作用问题的FE-SPH耦合模拟

应用有限元(FE)-光滑粒子流体动力学(SPH)耦合法模拟了具有自由表面的不可压流体与结构的相互作用问题.流体和结构分别采用SPH法和有限元法同时求解,两者在交界面处的相互作用通过接触算法进行处理.为了避免隐式计算压力,通过引入人工压缩率,将不可压流体近似为人工可压缩流体.采用FE-SPH耦合法对弹性板在随时间变化的水压作用下的变形以及倒塌水柱冲击弹性结构两个问题进行了模拟.模拟结果与实验结果以及其他已有数值结果符合良好,说明FE-SPH耦合法用于流体与结构相互作用问题的模拟是可行和有效的.     

用Laplace变换分析弹┐粘塑性梁的弯曲

在弹塑性梁弯曲变形理论基础上，本文用Ｌａｐｌａｃｅ变换进一步分析了弹－粘塑性梁的弯曲问题．并以矩形截面梁为例，说明弹－粘塑性梁弯曲时的弹性与粘塑性区的应力，梁的挠度及弹－粘塑性交线的计算     

二维含裂纹结构与声耦合问题研究

应用分形有限元方法结合边界元方法研究了二维含裂纹结构和声耦合问题.采用二级分形有限元方法对含裂纹的弹性结构体进行离散处理,这样可以使得自由度数大大地减少;无限大外域声场的计算使用边界元方法,可以自动满足无穷远辐射条件.数值仿真算例结果表明:结构声耦合系统的共振频率随着裂纹深度的增加而下降;裂纹附近的声场所受的影响较为明显.     

Partitioned subiterative coupling schemes for aeroelasticity using combined interface boundary condition method

The combined interface boundary condition (CIBC) method has been recently proposed for fluid–structure interaction. The CIBC method employs a Gauss–Seidel-like procedure to transform traditional interface conditions into velocity and traction corrections whose effect is controlled by a dimensional parameter. However, the original CIBC method has to invoke the uncorrected traction when forming the traction correction. This process limits its application to fluid–rigid body interaction. To repair this drawback, a new formulation of the CIBC method has been developed by using a new coupling parameter. The reconstruction is simple and the structural traction is removed completely. Two partitioned subiterative coupling versions of the CIBC method are developed. The first scheme is an implicit strategy while the second one is a semi-implicit strategy. Iterative loops are actualised by the fixed-point algorithm with Aitken accelerator. The obtained results agree with the well-documented data, and some famous flow phenomena have been successfully detected.     

Efficient shape optimization for fluid–structure interaction problems

An approach for the shape optimization of fluid–structure interaction (FSI) problems is presented. It is based on a partitioned solution procedure for fluid–structure interaction, a shape representation with NURBS, and sequential quadratic programming approach for optimization within a parallel environment with MPI as direct coupling tool. The optimization procedure is accelerated by employing reduced order models based on a proper orthogonal decomposition method with snapshots and Kriging. After the verification of the FSI optimization, the functionality and efficiency of the reduced order modeling as well as the corresponding optimization procedure are investigated.     

Interaction of a deformable solid with two-phase flows: An Eulerian-based numerical model for fluid-structure interaction using the level contour reconstruction method

We describe the formulation of a method for fluid-structure interaction involving the coupling of moving and/or flexible solid structures with multiphase flows in the framework of the Level Contour Reconstruction Method. We present an Eulerian-based numerical procedure for tracking the motion and interaction of a liquid-gas interface with a fluid-solid interface in the Lagrangian frame together with the evaluation of the fluid transport equations coupled to those for the solid transport, namely the left Cauchy-Green strain tensor field, in the Eulerian frame. To prevent excessive dissipation due to the convective nature of the solid transport equation, a simple incompressibility constraint for the strain field is enforced. A single grid structure is used for both the fluid and solid phases which allows for a simple and natural coupling of the fluid and solid dynamics. Several benchmark tests are performed to show the accuracy of the numerical method and which demonstrate accurate results compared to several of those in the existing literature. In particular we show that surface tension effects including contact line dynamics on the deforming solid phase can be properly simulated. The three-phase interaction of a droplet impacting on a flexible cantilever is investigated in detail. The simulations follow the detailed motion of the droplet impact (and subsequent deformation, breakup, and fall trajectory) along with the motion of the deformable solid cantilever due to its own weight as well as due to the force of the droplet impact.     

Monolithic approach to thermal fluid-structure interaction with nonconforming interfaces

This paper presents a monolithic approach to the thermal fluid-structureinteraction(FSI) with nonconforming interfaces.The thermal viscous flow is governedby the Boussinesq approximation and the incompressible Navier-Stokes equations.Themotion of the fluid domain is accounted for by an arbitrary Lagrangian-Eulerian(ALE)strategy.A pseudo-solid formulation is used to manage the deformation of the fluid do-main.The structure is described by the geometrically nonlinear thermoelastic dynamics.An efficient data transfer strategy based on the Gauss points is proposed to guarantee theequilibrium of the stresses and heat along the interface.The resulting strongly coupledset of nonlinear equations for the fluid,structure,and heat is solved by a monolithicsolution procedure.A numerical example is presented to demonstrate the robustness andefficiency of the methodology.     

Fluid–structure interaction modelling of nonlinear aeroelastic structures using the finite element corotational theory

The application of the finite element corotational theory to model geometric nonlinear structures within a fluid–structure interaction procedure is proposed. A dynamic corotational approximately-energy-conserving algorithm is used to solve the nonlinear structural response and it is shown that this algorithm's application with a four-node flat finite element is more stable than the nonlinear implicit Newmark method. This structural dynamic algorithm is coupled with the unsteady vortex-ring method using a staggered technique. These procedures were used to obtain aeroelastic results of a nonlinear plate-type wing subjected to low speed airflow. It is shown that stable and accurate numerical solutions are obtained using the proposed fluid–structure interaction algorithm. Furthermore, it is illustrated that geometric nonlinearities lead to limit cycle oscillations.     

Computational algorithms for tracking dynamic fluid–structure interfaces in embedded boundary methods

A robust, accurate, and computationally efficient interface tracking algorithm is a key component of an embedded computational framework for the solution of fluid–structure interaction problems with complex and deformable geometries. To a large extent, the design of such an algorithm has focused on the case of a closed embedded interface and a Cartesian computational fluid dynamics grid. Here, two robust and efficient interface tracking computational algorithms capable of operating on structured as well as unstructured three‐dimensional computational fluid dynamics grids are presented. The first one is based on a projection approach, whereas the second one is based on a collision approach. The first algorithm is faster. However, it is restricted to closed interfaces and resolved enclosed volumes. The second algorithm is therefore slower. However, it can handle open shell surfaces and underresolved enclosed volumes. Both computational algorithms exploit the bounding box hierarchy technique and its parallel distributed implementation to efficiently store and retrieve the elements of the discretized embedded interface. They are illustrated, and their respective performances are assessed and contrasted, with the solution of three‐dimensional, nonlinear, dynamic fluid–structure interaction problems pertaining to aeroelastic and underwater implosion applications. Copyright © 2012 John Wiley & Sons, Ltd.     

含液颗粒材料液固耦合分析的离散颗粒模型及特征线SPH法

对含液颗粒材料流固耦合分析建议了一个基于离散颗粒模型与特征线SPH法的显式拉格朗日-欧拉无网格方案。在已有的用以模拟固体颗粒集合体的离散颗粒模型[1]基础上,将颗粒间间隙内的流体模型化为连续介质,对其提出并推导了基于特征线的SPH法。数值例题显示了所建议方案在模拟颗粒材料与间隙流相互作用的能力和性能以及间隙流体对颗粒结构承载能力及变形的影响。     

DEVELOPMENT OF A THREE-DIMENSIONAL VISCOUS AEROELASTIC SOLVER FOR NONLINEAR PANEL FLUTTER

A new three-dimensional (3-D) viscous aeroelastic solver for nonlinear panel flutter is developed in this paper. A well-validated full Navier–Stokes code is coupled with a finite-difference procedure for the von Karman plate equations. A subiteration strategy is employed to eliminate lagging errors between the fluid and structural solvers. This approach eliminates the need for the development of a specialized, tightly coupled algorithm for the fluid/structure interaction problem. The new computational scheme is applied to the solution of inviscid two-dimensional panel flutter problems for subsonic and supersonic Mach numbers. Supersonic results are shown to be consistent with the work of previous researchers. Multiple solutions at subsonic Mach numbers are discussed. Viscous effects are shown to raise the flutter dynamic pressure for the supersonic case. For the subsonic viscous case, a different type of flutter behavior occurs for the downward deflected solution with oscillations occurring about a mean deflected position of the panel. This flutter phenomenon results from a true fluid/structure interaction between the flexible panel and the viscous flow above the surface. Initial computations have also been performed for inviscid, 3-D panel flutter for both supersonic and subsonic Mach numbers.