Equilibrium characteristics of the secondary convection in a vertical fluid layer between two flat plates 1

The nonlinear stability of the natural convection in a vertical fluid layer between two flat plates with different temperatures is investigated by a direct method to find the equilibrium states of the secondary convection. We confine ourselves to two-dimensional flows and assume that the aspect ratio of the fluid layer is very large. Since the Prantl number is assumed to be very small, the buoyancy effect caused by temperature disturbances is negligible. As a result we obtained a neutral surface of the energy of the fundamental mode of the secondary convection. It is concluded that there is no finite amplitude instability below the critical Grashof number derived from linear stability theory, and that both the unstable equilibrium solution (threshold amplitude solution) and the stable equilibrium solution (finite amplitude solution) are found outside the neutral curve of the linear stability. Our results are almost consistent with those of Nagata and Busse (1983), but are more accurate and more thorough.     

Simulation of two‐phase flow with moving immersed boundaries

A two‐dimensional multi‐phase model for immiscible binary fluid flow including moving immersed objects is presented. The fluid motion is described by the incompressible Navier–Stokes equation coupled with a phase‐field model based on van der Waals' free energy density and the Cahn–Hilliard equation. A new phase‐field boundary condition was implemented with minimization of the free energy in a direct way, to specifically improve the physical behavior of the contact line dynamics for moving immersed objects. Numerical stability and execution time were significantly improved by the use of the new boundary condition. Convergence toward the analytical solution was demonstrated for equilibrium contact angle, the Lucas–Washburn theory and Stefan's problem. The proposed model may be used for multi‐phase flow problems with moving boundaries of complex geometry, such as the penetration of fluid into a deformable, porous medium. Copyright © 2010 John Wiley & Sons, Ltd.     

Effect of Coulomb Damping on Buckling of a Two-Rod System

This paper describes a significant influence of a slight Coulomb damping on buckling, using a simple two rods system. Coulomb damping produces equilibrium regions around the well-known stable and unstable steady states under the pitchfork bifurcation which occurs in the case without Coulomb damping. Also, the stability of the states in the equilibrium regions is examined by using the phase portrait. As a consequence, due to the slight Coulomb damping, it is theoretically clarified that the states in the equilibrium regions are locally stable, even in the neighborhood of the unstable steady states under the pitchfork bifurcation in the case without Coulomb damping, i.e., even in the neighborhood of the unstable trivial steady states in the postbuckling and the unstable nontrivial steady states under the subcritical pitchfork bifurcation. Furthermore, the experimental results are in qualitative agreement with the theoretically predicted phenomena.     

Mobility‐dependent bifurcations in capillarity‐driven two‐phase fluid systems by using a lattice Boltzmann phase‐field model

Bifurcations in capillarity‐driven two‐phase fluid systems, due to different mobilities in phase‐field models for such systems, are studied by using a lattice Boltzmann method (LBM). Specifically, two‐dimensional (2D) and three‐dimensional (3D) droplets on a flat wall with given wettability variations are investigated. It is found that the mobility controls the rate of diffusive relaxation of the phase field from non‐equilibrium toward equilibrium, and similar to previous findings on mechanically driven two‐phase systems, the mobility is closely related to the contact line velocity. For the cases investigated, different mobilities across a critical value result in fundamentally different system evolution routes and final stable equilibrium states. These results may provide some implications for phase‐field study of droplet manipulations by surface wettability adjustments in microfluidics. Copyright © 2008 John Wiley & Sons, Ltd.     

A timestepper approach for the systematic bifurcation and stability analysis of polymer extrusion dynamics

We discuss how matrix-free/timestepper algorithms can efficiently be used with dynamic non-Newtonian fluid mechanics simulators in performing systematic stability/bifurcation analysis. The timestepper approach to bifurcation analysis of large-scale systems is applied to the plane Poiseuille flow of an Oldroyd-B fluid with non-monotonic slip at the wall, in order to further investigate a mechanism of extrusion instability based on the combination of viscoelasticity and non-monotonic slip. Due to the non-monotonicity of the slip equation the resulting steady-state flow curve is non-monotonic and unstable steady states appear in the negative-slope regime. It has been known that self-sustained oscillations of the pressure gradient are obtained when an unstable steady state is perturbed [M.M. Fyrillas, G.C. Georgiou, D. Vlassopoulos, S.G. Hatzikiriakos, A mechanism for extrusion instabilities in polymer melts, Polymer Eng. Sci. 39 (1999) 2498–2504].Treating the simulator of a distributed parameter model describing the dynamics of the above flow as an input–output “black-box” timestepper of the state variables, stable and unstable branches of both equilibrium and periodic oscillating solutions are computed and their stability is examined. It is shown for the first time how equilibrium solutions lose stability to oscillating ones through a subcritical Hopf bifurcation point which generates a branch of unstable limit cycles and how the stable periodic solutions lose their stability through a critical point which marks the onset of the unstable limit cycles. This implicates the coexistence of stable equilibria with stable and unstable periodic solutions in a narrow range of volumetric flow rates.     

Numerical simulation of the accumulation of heavy particles in a circular bounded vortex flow

In present work, an Eulerian–Lagrangian CFD model based on the discrete element method (DEM) and immersed boundary method (IBM) has been developed, validated and used to investigate the accumulation of heavy particles in a circular bounded viscous vortex flow. The inter-particle and particle-wall collisions are resolved by a hard-sphere model. Effects of one-way and two-way coupling, Reynolds number, and particle diameter are systematically explored. Results show that, in case of one-way coupling, the majority of particles will spiral into an accumulation point located near the stagnation point of the flow field. The accumulation point represents a stable equilibrium point as the drag created by the flow field balances the destabilizing centrifugal force on the particle. However, in case of two-way coupling, there does not exist a stable accumulation point due to the strong interaction between the particles and fluid dynamics. Instead most particles are expelled from the circular domain and accumulate on the confining wall. The percentage of accumulated particles on the wall increases with increasing Reynolds number and particle diameter. Moreover, influence of three well-known drag models is also studied and they give consistent results on the particle accumulation behavior, although small quantitative differences can still be discerned.     

Assessment of volume of fluid and immersed boundary methods for droplet computations

The volume of fluid (VOF) and immersed boundary (IB) methods are two popular computational techniques for multi‐fluid dynamics. To help shed light on the performance of both techniques, we present accuracy assessment, which includes interfacial geometry, detailed and global fluid flow characteristics, and computational robustness. The investigation includes the simulations of a droplet under static equilibrium as a limiting test case and a droplet rising due to gravity for Re?1000. Surface tension force models are key issues in both VOF and IB and alternative treatments are examined resulting in improved solution accuracy. A refined curvature model for VOF is also presented. With the newly developed interfacial treatments incorporated, both IB and VOF perform comparably well for the droplet dynamics under different flow parameters and fluid properties. Copyright © 2004 John Wiley & Sons, Ltd.     

浸入边界法及其应用

浸入边界法主要用于模拟存在复杂外形结构的流场的运动情况和处理各种动边界问题,目前已广泛应用于计算流体力学领域.浸入边界法既是数学建模方法又是数值离散方法,它将物体边界与流体的相互作用通过在流体运动方程中加体积力项来体现,并在数值计算中采用简单的笛卡尔网格,避免了按照物体边界形状生成贴体网格时所遇到的各种问题.浸入边界法分为连续力法和离散力法:连续力法主要用于处理弹性边界问题,它的力源项满足特定的力学关系式;离散力法主要用于处理固体界面问题,它的力源项由边界条件推导得到.着重阐述了浸入边界法的基本原理和数学构造,对目前已有的几种不同的浸入边界法做了简单地介绍,并给出了一些应用实例,最后提出了浸入边界法未来的发展方向.     

The remarkable nature of radially symmetric equilibrium states of aeolotropic nonlinearly elastic bodies

This paper treats the radially symmetric equilibrium states of aeolotropic nonlinearly elastic solid cylinders and balls under constant normal forces on their boundaries. It is shown that the aeolotropy gives rise to solutions describing both intact and cavitating states, which exhibit an array of remarkable new phenomena, not suggested by the solutions for isotropic bodies. E.g., it is shown that there are materials having a critical pressure such that for applied pressures on the boundary below the critical value, the normal pressures at the center of the body are zero and for applied pressures above the critical value, the normal pressures at the center are infinite. There are also materials for which there is no equilibrium state with center intact when the boundary is subjected to uniform tension. It is also shown that the equilibrium states treated here are the only radially symmetric equilibrium states. Thus the strange phenomena discovered here must be present in such stable equilibrium states.     

基于反馈力浸入边界法模拟复杂动边界流动

浸入边界法是模拟流固耦合的重要数值方法之一。本文采用反馈力浸入边界方法,对旋转圆柱和水轮机活动导叶旋转摆动绕流后的动边界流场进行数值模拟。其中,固体边界采用一系列离散的点近似代替,流体为不可压缩牛顿流体,使用笛卡尔自适应加密网格,利用有限差分法进行求解。固体对流场的作用通过构造适宜的反馈力函数实现。本文首先通过旋转圆柱绕流的计算结果同实验结果进行对比,吻合较好,验证了该计算方法的可靠性。然后针对水电站水力过渡过程中水轮机活动导叶旋转摆动绕流后的动边界流场进行数值模拟,得到导叶动态绕流后的流场分布特性和涡结构的演化特性。     

Experiments on waves induced in the hollow core of vortices

Recent experimental findings with respect to wave activity in the core of a hollow disk-shaped vortex generated in a cylindrical container with a flat disk rotating at the bottom are reported herein. For relatively low viscosity fluid, several stationary states of the core exist within a certain range of disk speeds. This range becomes narrow as the wave number grows. Between the stationary states, mixed, time dependent states were found to occur. Their interval of endurance decrease with the wave number. The disk speed at which the static state first appears and ends increases linearly with the original height. The phase velocity of the stationary waves rises with the angular velocity of the disk. For a liquid with an intermediate value of viscosity, abrupt transitions from one equilibrium state to the other is taking place. Steady vortex core patterns with wave numbers from one to eleven are observed. States were the basic pattern is subharmonically modulated and states where a wave packet encircles periodically the core are also encountered. Hysterisis is clearly evident. The core of a highly viscous liquid is stable. A suppressed form of instability appears at very large liquid heights and disk speeds.     

Unified one‐fluid formulation for incompressible flexible solids and multiphase flows: Application to hydrodynamics using the immersed structural potential method (ISPM)

In this paper, we present a two‐dimensional computational framework for the simulation of fluid‐structure interaction problems involving incompressible flexible solids and multiphase flows, further extending the application range of classical immersed computational approaches to the context of hydrodynamics. The proposed method aims to overcome shortcomings such as the restriction of having to deal with similar density ratios among different phases or the restriction to solve single‐phase flows. First, a variation of classical immersed techniques, pioneered with the immersed boundary method (IBM), is presented by rearranging the governing equations, which define the behaviour of the multiple physics involved. The formulation is compatible with the “one‐fluid” formulation for two‐phase flows and can deal with large density ratios with the help of an anisotropic Poisson solver. Second, immersed deformable structures and fluid phases are modelled in an identical manner except for the computation of the deviatoric stresses. The numerical technique followed in this paper builds upon the immersed structural potential method developed by the authors, by adding a level set–based method for the capturing of the fluid‐fluid interfaces and an interface Lagrangian‐based meshless technique for the tracking of the fluid‐structure interface. The spatial discretisation is based on the standard marker‐and‐cell method used in conjunction with a fractional step approach for the pressure/velocity decoupling, a second‐order time integrator, and a fixed‐point iterative scheme. The paper presents a wide d range of two‐dimensional applications involving multiphase flows interacting with immersed deformable solids, including benchmarking against both experimental and alternative numerical schemes.     

The Duhem-Hadamard hypothesis in thermofluiddynamics

The Duhem–Hadamard hypothesis (DHH) states that in a fluid flow the same equation relates the pressure, the mass density, and the temperature as in equilibrium. The entire theory of bodies concerning flow processes in engineering science and practice rests upon it. Our recent work on fundamentals of non-equilibrium phenomena suggests that the hypothesis are only approximate in some sense, but its mathematical formulation does not make clear what is being approximated. To get a preliminary clarification, the paper deals with two items: first, a theoretical basis is worked out to yield appropriate presuppositions for an efficient experimental program and, second, provisional experimental data are established by a new and precise set-up. Our theoretical examinations admit to derive definite distinguishing features of both non-equilibrium and equilibrium processes. The latter appear as a limiting case of the former and are identified with the hypothetical fluid state at rest. Only for this special state the thermal equation of state is representative of thermodynamic equilibrium. Yet, there is a second kind of equilibrium: the kinetic equilibrium defined for reversible flows. All theoretical considerations lead us to the conclusion that the DHH can only be justified for kinetic equilibria. The first phase of our research program allowed to learn that our high-precision measurement devices need to be improved in order to decide, whether this special kind of reversible flow exists as a real limiting case. Unlike this, our experiments prove that non-equilibrium states induced by the motions of the fluid cannot be described by the DHH.     

An immersed smoothed point interpolation method (IS‐PIM) for fluid‐structure interaction problems

An immersed smoothed point interpolation method using 3‐node triangular background cells is proposed to solve 2D fluid‐structure interaction problems for solids with large deformation/displacement placed in incompressible viscous fluid. In the framework of immersed‐type method, the governing equations can be decomposed into 3 parts on the basis of the fictitious fluid assumption. The incompressible Navier‐Stokes equations are solved using the semi‐implicit characteristic‐based split scheme, and solids are simulated using the newly developed edge‐based smoothed point interpolation method. The fictitious fluid domain can be used to calculate the coupling force. The numerical results show that immersed smoothed point interpolation method can avoid remeshing for moving solid based on immersed operation and simulate the contact phenomenon without an additional treatment between the solid and the fluid boundary. The influence from information transfer between solid domain and fluid domain on fluid‐structure interaction problems has been investigated. The numerical results show that the proposed interpolation schemes will generally improve the accuracy for simulating both fluid flows and solid structures.     

Inhomogeneous deformation of elastomer gels in equilibrium under saturated and unsaturated conditions

A variational method is employed to obtain governing equations and boundary conditions describing finite strain equilibrium configurations of elastomeric gels. Three situations are considered: a liquid saturated gel, an unsaturated gel, and a gel in equilibrium with a vapor of its own liquid. Surface tractions can lead to equilibrium transitions between these cases. The liquid saturated gel is regarded as immersed in a liquid bath. If this bath becomes depleted, then the gel is unsaturated. The degree of unsaturation - a measure of the amount of liquid that would restore a state of saturation - affects the subsequent mechanical behavior. If the unsaturated system is further allowed to condense or evaporate its liquid component at the gel surface, then a new state of equilibrium is achieved. The transition between the unsaturated case and the case of being in equilibrium with the vapor phase corresponds to the chemical potential variable of the gel changing its value from one that is determined by a volume constraint to the value of the chemical potential in the vapor phase. A finite element method is created on the basis of the variational method and demonstrated in the context of eversion, a deformation that imposes very large finite strains. Liquid migration within the gel is not modeled as our focus is on equilibrium states that occur after all such non-equilibrium processes come to rest.     

The transient development of the flow in an impulsively rotated annular container

When a fluid-filled container is spun up from rest to a constant angular velocity the fluid responds in such a way that the fluid–container system is ultimately in a state of rigid-body rotation. The fluid can then be said to have traversed a trajectory in phase space from a simple stable equilibrium state of no motion to another stable equilibrium representing full rigid-body rotation. This simple statement belies the fact that during this process the fluid can undergo a series of transitions, from a laminar through a transient turbulent state, before attaining the stable motion that is rigid-body rotation. Using a combination of analytical and computational methods, we focus on the dynamics resulting from an impulsive change in the rotation rate of a fluid-filled annulus, specifically, the impulsive spin-up of a stationary annulus, or the impulsive spin-down of an annulus already in a state of rigid-body rotation. We explore the initial development of the impulsively generated axisymmetric boundary layer, its subsequent instability, and the larger-scale transient features within this class of flows, allowing us to look at the effect these features have on the time it takes for the system to spin up to a steady state, or spin down to rest.     

On the Fourier representation of elastic immersed boundaries

This paper presents an efficient treatment of fluid/elastic–structure interactions that takes advantage of the Fourier representation of immersed boundaries. We assume that the fluid is incompressible with uniform density and viscosity and that the immersed boundaries have fixed topologies. These elastic bodies can have large deformations and evolve anywhere within the fluid domain. They may be thick and are assumed to be piecewise smooth. We process the fluid–structure coupling with the immersed boundary method of C. S. Peskin. We can take advantage of the Fourier representation of the immersed bodies in many ways. First, the use of Fourier expansions allows us to filter out the high frequencies of the spatial oscillations along the boundary vectors. Second, we can work with a smaller number of boundary points to represent the interface, while preserving the same level of accuracy as long as enough points are used in the force spreading process. Finally, the harmonic information gathered by the Fourier coefficients is useful to control some global properties of the immersed boundaries. For example, we introduce a technique that corrects the volume conservation issue of closed immersed boundaries by performing constrained optimization in the Fourier space. We illustrate our method with two applications: one is a suspension flow with a large number of elastic ‘bubbles’, the other is an interesting case of artificial motion based on inertia rather than on flapping fins or flagella. Copyright © 2009 John Wiley & Sons, Ltd.     

Mechanical Properties of Microcapsules Used in a Self-Healing Polymer

The elastic modulus and failure behavior of poly(urea-formaldehyde) shelled microcapsules were determined through single-capsule compression tests. Capsules were tested both dry and immersed in a fluid isotonic with the encapsulent. The testing of capsules immersed in a fluid had little influence on mechanical behavior in the elastic regime. Elastic modulus of the capsule shell wall was extracted by comparison with a shell theory model for the compression of a fluid filled microcapsule. Average capsule shell wall modulus was 3.7 GPa, regardless of whether the capsule was tested immersed or dry. Microcapsule diameter was found to have a significant effect on failure strength, with smaller capsules sustaining higher loads before failure. Capsule size had no effect on the modulus value determined from comparison with theory.