Incompressible smoothed particle hydrodynamics simulation of multifluid flows

This paper presents an incompressible SPH (ISPH) technique to simulate multifluid flows. The SPH method is a mesh‐free particle modeling approach that can treat free surfaces and multi‐interfaces in a simple and efficient manner. The ISPH method employs an incompressible hydrodynamic formulation to solve the fluid pressure that ensures a stable pressure field. Two multifluid ISPH models are proposed following different interface treatments: the coupled ISPH model does not distinguish the different fluid phases and applies the standard ISPH technique across the interface, whereas the decoupled ISPH model first treats each fluid phase separately and then couples the different phases by applying pressure and shear stress continuities across the interface. The two proposed models were used to investigate a gravity underflow with a low density ratio in a Generalized Reservoir Hydrodynamics (GRH) flume and a horizontal lock exchange flow with a high density ratio. Comparisons with data and relevant numerical error analysis indicated that the decoupled model performed well in cases of both low and high density ratios because of the accurate treatment of interface boundaries. Copyright © 2011 John Wiley & Sons, Ltd.     

对流扩散方程的迎风变换及相应有限差分方法

本文提出所谓迎风变换,将对流扩散方程分解为对流迎风函数和扩散方程,并构造相应的有限差分格式。对流迎风函数以简明的指数解析形式反映对流扩散现象的迎风效应,原则上消除了源于不对称对流算子的困难,能够便利对流扩散方程的数值求解。有限差分格式具有二阶精度和无条件稳定性,算例表明其准确性、收敛速度及对边界层效应的适应能力均明显优于中心差分格式和迎风差分格式。     

Parallel three‐dimensional simulation of the injection molding process

This paper describes the development of a parallel three‐dimensional unstructured non‐isothermal flow solver for the simulation of the injection molding process. The numerical model accounts for multiphase flow in which the melt and air regions are considered to be a continuous incompressible fluid with distinct physical properties. This aspect avoids the complex reconstruction of the interface. A collocated finite volume method is employed, which can switch between first‐ and second‐order accuracy in both space and time. The pressure implicit with splitting of operators algorithm is used to compute the transient flow variables and couple velocity and pressure. The temperature equation is solved using a transport equation with convection and diffusion terms. An upwind differencing scheme is used for the discretization of the convection term to enforce a bounded solution. In order to capture the sharp interface, a bounded compressive high‐resolution scheme is employed. Parallelization of the code is achieved using the PETSc framework and a single program multiple data message passing model. Predicted numerical solutions for several example problems are considered. The first case validates the solution algorithm for moderate Reynolds number flows using a structured mesh. The second case employs an unstructured hybrid mesh showing the capability of the solver to describe highly viscous flows closer to realistic injection molding conditions. The final case presents the non‐isothermal filling of a thick cavity using three mesh sizes and up to 80 processors to assess parallel performance. The proposed algorithm is shown to have good accuracy and scalability. Copyright © 2008 John Wiley & Sons, Ltd.     

Pure axial flow of viscoelastic fluids in rectangular microchannels under combined effects of electro-osmosis and hydrodynamics

This paper presents an analysis of the combined electro-osmotic and pressure-driven axial flows of viscoelastic fluids in a rectangular microchannel with arbitrary aspect ratios. The rheological behavior of the fluid is described by the complete form of Phan-Thien–Tanner (PTT) model with the Gordon–Schowalter convected derivative which covers the upper convected Maxwell, Johnson–Segalman and FENE-P models. Our numerical simulation is based on the computation of 2D Poisson–Boltzmann, Cauchy momentum and PTT constitutive equations. The solution of these governing nonlinear coupled set of equations is obtained by using the second-order central finite difference method in a non-uniform grid system and is verified against 1D analytical solution of the velocity profile with less than 0.06% relative error. Also, a parametric study is carried out to investigate the effect of channel aspect ratio (width to height), wall zeta potential and the Debye–Hückel parameter on 2D velocity profile, volumetric flow rate and the Poiseuille number in the mixed EO/PD flows of viscoelastic fluids with different Weissenberg numbers. Our results show that, for low channel aspect ratios, the previous 1D analytical models underestimate the velocity profile at the channel half-width centerline in the case of favorable pressure gradients and overestimate it in the case of adverse pressure gradients. The results reveal that the inapplicability of the Debye–Hückel approximation at high zeta potentials is more significant for higher Weissenberg number fluids. Also, it is found that, under the specified values of electrokinetic parameters, there is a threshold for velocity scale ratio in which the Poiseuille number is approximately independent of channel aspect ratio.     

On the application of an approximate kinetic equation of heat and mass transfer processes: the effect of body shape

An approximate kinetic equation of heat and mass transfer has been presented. It is an ordinary differential equation which is easy to integrate. The derivation of this approximation was based on an analysis of the available analytical solution of the problem. The proposed equation can be applied to bodies (pellets) in the shape of an infinite slab, infinite cylinder and sphere. A generalization of this equation to cases where the transfer resistance occurs both in the body and in the surrounding fluid has been proposed. The equation has been tested in various conditions both for thermal and diffusive processes. Radiative cooling of bodies has been considered as a thermal process and adsorption in a single pellet—as a diffusive process. All tests showed high accuracy of the approximate equation; in many cases the results were indistinguishable from the results of the exact model. A special feature of the proposed equation is its high accuracy for short times of the process, what significantly differentiates it from the classical approximate kinetic equation Linear Driving Force.     

An analytical interface reconstruction algorithm in the PLIC‐VOF method for 2D polygonal unstructured meshes

Piecewise linear interface calculation (PLIC) schemes have been extensively employed in the volume‐of‐fluid (VOF) method for interface capturing in numerical simulations of multiphase flows. Polygonal unstructured meshes are often adopted because of their geometric flexibility and superiority in gradient calculation. An analytical interface reconstruction algorithm in the PLIC‐VOF method for arbitrary convex polygonal cells has been proposed in this study. The line interface at a given orientation within a polygonal cell is located by an analytical technique. It has been tested successfully for four different geometric shapes that are common in polygonal meshes. The computational efficiency of the present algorithm has been compared with several published schemes in the literature. The proposed algorithm has been shown to yield higher accuracy with reduction in computational complexity. A numerical simulation of a dam‐breaking problem has been performed using the proposed analytical PLIC technique on polygonal meshes. The results are in good agreement with experimental data available in the literature, which serves as a demonstration of its performance in a real multiphase flow.     

Enhanced equilibrium distribution functions for simulating immiscible multiphase flows with variable density ratios in a class of lattice Boltzmann models

This research examines the behavior of a class of lattice Boltzmann (LB) models designed to simulate immiscible multiphase flows. There is some debate in the scientific literature as to whether or not the “color gradient” models, also known as the Rothman–Keller (RK) models, are able to simulate flow with density variation. In this paper, we show that it is possible, by modifying the original equilibrium distribution functions, to capture the discontinuity present in the analytical momentum profile of the two-layered Couette flow with variable density ratios. Investigations carried out earlier were not able to simulate such a flow correctly. Now, with the proposed approach, the new scheme is compatible with the analytical solution, and it is also possible to simulate the two-layered Couette flow with simultaneous density ratios of O(1000) and viscosity ratios of O(100). To test the model in a more complex flow situation, i.e. with non-zero surface tension and a curved interface, an unsteady simulation of an oscillating bubble with variable density ratio is undertaken. The numerical frequency of the bubble is compared to that of the analytical frequency. It is demonstrated that the proposed modification greatly increases the accuracy of the model compared to the original model, i.e. the error can be up to one order of magnitude lower with the proposed enhanced equilibrium distribution functions. The authors believe that this improvement can be made to other RK models as well, which will allow the range of validity of these models to be extended. This is, in fact, what the authors found for the method proposed in this article.     

A Flexible Coupled Level Set and Volume of Fluid (flexCLV) method to simulate microscale two-phase flow in non-uniform and unstructured meshes

An innovative Flexible Coupled Level Set (LS) and Volume of Fluid (VOF) algorithm (flexCLV) to simulate two-phase flows at the microscale on unstructured and non-uniform meshes is proposed. The method combines the advantages of the VOF method in terms of mass conservation and the LS method in terms of accuracy of the surface tension implementation and can handle both 2D and 3D domains discretized by either structured hexaedra or unstructured tetrahedral grids with high aspect ratio elements, thus guaranteeing flexibility and robustness. The method is implemented within the VOF-based OpenFOAM’s solver interFoam, which is retained as the base algorithm for the interface advection, while the surface tension force is calculated by using the level set function reconstructed from the VOF’s fraction. The method is first validated in static flow conditions by simulating a circular bubble at equilibrium and then in dynamic flow conditions by studying a freely bubble rising in both 2D and 3D domains discretized by both structured and unstructured meshes. The proposed flexCLV algorithm is then used to simulate the dynamics of confined bubbles in circular microchannels in the low capillary number regime. 2D and 3D mesh grids with high aspect ratio elements are utilized to discretized the liquid film at the tube’s walls. The numerical results are compared with the available literature and simulations performed with the original interFoam solver in terms of bubble shape and velocity, thickness of the liquid film and amplitude of the bubble tail oscillations. Results compare very well with the experimental measurements and demonstrate the superior accuracy of the coupled flexCLV method with respect to the original VOF method when surface tension and accurate interface representation play a fundamental role. Importantly, the present study also provides a precious insight on the time-dependent patterns appearing on the bubble surface in the visco-inertial regime, which could be here investigated in detail.     

Effects of aspect ratio and orientation on the wake characteristics of low Reynolds number flow over a triangular prism

The wake characteristics of unconfined flows over triangular prisms of different aspect ratios have been numerically analysed in the present work. For this purpose, a fixed Cartesian-grid based numerical technique that involves the porous medium approach to mimic the effect of solid blockage has been utilised. Correspondingly, laminar flow simulations ranging from the sub-critical regime (before the onset of vortex shedding) to the super-critical regime have been considered here within the limits of two-dimensionality. In the sub-critical regime, correlations relating the wake bubble length with Reynolds number (Re) have been proposed for various aspect ratios. Also, the effects of aspect ratio and Reynolds Number on the drag force coefficient (C_D) have been characterised for two different geometrical orientations of the prism (base or apex facing the flow). Subsequently, the critical Reynolds number at the onset of vortex shedding has been predicted for each of the aspect ratio considered, by an extrapolation procedure. The unsteady flow characteristics of the super-critical regime are finally highlighted for different aspect ratios and triangular orientations considered in the study.     

Effect of chamber pressure on spreading and splashing of liquid drops upon impact on a dry smooth stationary surface

Liquid drop impacts on a smooth surface were studied at elevated chamber pressures to characterize the effect of gas pressure on drop spreading and splashing. Five common liquids were tested at impact speeds between 1.0 and 3.5 m/s and pressure up to 12 bars. Based on experiments at atmospheric pressure, a modification to the “free spreading” model (Scheller and Bousfield in AIChE Paper 41(6):1357–1367, 1995) has been proposed that improves the prediction accuracy of maximum spread factors from an error of 15–5%. At high chamber pressures, drop spreading and maximum spread factor were found to be independent of pressure. The splash ratio (Xu et al. in Phys Rev Lett 94:184505, 2005) showed a non-constant behavior, and a power-law model was demonstrated to predict the increase in splash ratio with decreasing impact speed in the low impact speed regime. Also, drop shape was found to affect splash promotion or suppression for an asymmetry greater than 7–8% of the equivalent drop diameter. The observations of the current work could be especially useful for the study of formation of deposits and wall combustion in engine cylinders.     

Orientation tensors in simple flows of dilute suspensions of non-Brownian rigid ellipsoids,comparison of analytical and approximate solutions

General analytical solutions are obtained for the planar orientation structure of rigid ellipsoid of revolutions subjected to an arbitrary homogeneous flow in a Newtonian fluid. Both finite and infinite aspect ratio particles are considered. The orientation structure is described in terms of two-dimensional, time-dependent tensors that are commonly employed in constitutive equations for anisotropic fluids such as fiber suspensions. The effect of particle aspect ratio on the evolution of orientation structure is studied in simple shear and planar elongational flows. With the availability of analytical solutions, accuracies of quadratic closure approximations used for nonhomogeneous flows are analyzed, avoiding numerical integration of orientation distribution function. In general, fourth-order orientation evolution equations with sixth-order quadratic closure approximations yield more accurate representations compared to the commonly used second-order evolution equations with fourth-order quadratic closure approximations. However, quadratic closure approximations of any order are found to give correct maximum orientation angle (i.e., preferred direction) results for all particle aspect ratios and flow cases.     

A mathematical model for the propulsive thrust of the thin elastic wing harmonically oscillating in a flow of non-viscous incompressible fluid

In the present paper there is proposed an analytical approach to study vibration of a rectangular elastic wing in the stationary stream of non-viscous fluid. We first develop a basic two-dimensional integral equation. Then a series expansion along the short coordinate is applied. This reduces the problem to an infinite set of one-dimensional integral equations which is studied asymptotically with respect to the large aspect ratio parameter. An example of optimization of thickness of the wing is demonstrated, to test the efficiency of the proposed method in applications.     

An improved diffuse interface method for three-dimensional multiphase flows with complex interface deformation

An improved diffuse interface (DI) method is proposed for accurately capturing complex interface deformation in simulations of three-dimensional (3D) multiphase flows. In original DI methods, the unphysical phenomenon of interface thickening or blurring can affect the accuracy of numerical simulations, especially for flows with large density ratio and high Reynolds number. To remove this drawback, in this article, an interface-compression term is introduced into the Cahn-Hilliard equation to suppress the interface dispersion. The additional term only takes effect in the interface region and works normal to the interface. The difference of the current method from the previous work is that the compression rate can be adjusted synchronously according to the magnitude of local vorticity, which is strongly correlated to the interface dispersion and changes with the computational time and interface position. Numerical validations of the proposed method are implemented by simulating problems of Laplace law, Rayleigh-Taylor instability, bubble rising in a channel, and binary droplet collision. The obtained results agree well with the analytical solutions and published data. The numerical results show that the phenomenon of interface dispersion is suppressed effectively and the tiny interfacial structures in flow field can be captured accurately.     

Predicting the mechanical behavior of two-phase materials with cellular automata

A new mechanical model is proposed to predict the mechanical behavior of inhomogeneous materials. The simple case of an aggregate of two linearly viscous phases submitted to plane strain is addressed here. Calculations involve the Eshelby localization relationships associated with a Walpole-type averaging procedure. When the grains are equiaxed, derivations are entirely analytical, which allows the proposed method to be simply compared with the classical Hashin-Shtrikman lower and upper bounds and the self-consistent model. When large strains are considered, the model is used to predict numerically the overall stress-strain relationships. Significant morphological hardening or softening is shown to occur with increasing strain, depending on the initially equiaxed or elongated grain shapes. Finally, the local distributions of strains, and thus the development of strain inhomogeneities, are also predicted by the model.     

A new Kirchhoff plate model based on a modified couple stress theory

In this paper a new Kirchhoff plate model is developed for the static analysis of isotropic micro-plates with arbitrary shape based on a modified couple stress theory containing only one material length scale parameter which can capture the size effect. The proposed model is capable of handling plates with complex geometries and boundary conditions. From a detailed variational procedure the governing equilibrium equation of the micro-plate and the most general boundary conditions are derived, in terms of the deflection, using the principle of minimum potential energy. The resulting boundary value problem is of the fourth order (instead of existing gradient theories which is of the sixth order) and it is solved using the Method of Fundamental Solutions (MFS) which is a boundary-type meshless method. Several plates of various shapes, aspect and Poisson’s ratios are analyzed to illustrate the applicability of the developed micro-plate model and to reveal the differences between the current model and the classical plate model. Moreover, useful conclusions are drawn from the micron-scale response of this new Kirchhoff plate model.     

The spatial evolution of velocity and density profiles in an arrested salt wedge

The spatial variation in the properties of an arrested salt wedge have been investigated, both analytically and in the laboratory. In the laboratory particle image velocimetry and laser induced fluorescence were used to obtain flow velocities and the height of the density interface. An analytical solution for the profile of interface height, in the absence of interfacial instabilities, has been developed from two-layer internal hydraulic theory. The evolution of the velocity profile is predicted using a momentum diffusion equation following a Lagrangian frame of reference along the interface of the salt wedge. The centre of the shear layer is predicted to lie above the density interface, with this offset decreasing in the downstream direction. Our theoretical predictions are in good agreement with our laboratory measurements.     

Modeling of Turbulent Natural Convection in Enclosed Tall Cavities

It was shown in our previous work (J. Appl. Mech. Tech. Phys 57 (7), 1159–1171 (2016)) that the eddy-resolving parameter-free CABARET scheme as applied to two-and three-dimensional de Vahl Davis benchmark tests (thermal convection in a square cavity) yields numerical results on coarse (20 × 20 and 20 × 20 × 20) grids that agree surprisingly well with experimental data and highly accurate computations for Rayleigh numbers of up to 10¹⁴. In the present paper, the sensitivity of this phenomenon to the cavity shape (varying from cubical to highly elongated) is analyzed. Box-shaped computational domains with aspect ratios of 1: 4, 1: 10, and 1: 28.6 are considered. The results produced by the CABARET scheme are compared with experimental data (aspect ratio of 1: 28.6), DNS results (aspect ratio of 1: 4), and an empirical formula (aspect ratio of 1: 10). In all the cases, the CABARET-based integral parameters of the cavity flow agree well with the other authors’ results. Notably coarse grids with mesh refinement toward the walls are used in the CABARET calculations. It is shown that acceptable numerical accuracy on extremely coarse grids is achieved for an aspect ratio of up to 1: 10. For higher aspect ratios, the number of grid cells required for achieving prescribed accuracy grows significantly.     

Flexural vibration of a square plate clamped along its contour

Analytical expressions are constructed for calculating the natural frequencies and mode shapes of flexural vibrations of a square homogeneous plate clamped along its contour. An error estimate is given by comparing predicted results with those of known high-precision calculations. Also the results of analytical calculations are compared with experimental data obtained by the author using the resonance method. The analytical and corresponding numerical results coincide with the experimental data to within less than 1%.     

基于等效特征应变原理的复合材料有效模量细观力学分析

基于等效特征应变原理,提出了一种新的复合材料有效模量细观力学分析方法。首先,在等效特征应变原理基础上提出平均等效特征应变原理,它可用于解决有限体下任意形状(无论是凸或凹形)的单个夹杂或多个夹杂的弹性变形问题。其次,将平均等效特征应变原理与细观力学直接均匀法相结合,来分析确定复合材料的有效模量。最后利用复合材料纤维与基体的力学性能参数及纤维的体分比,借助MATLAB编程方法,预测其有效模量。通过将理论预测值与已有的的试验值、其它理论预测值进行对比,验证了新分析方法的合理性和分析精度。     

Two‐dimensional non‐Newtonian injection molding with a new control volume FEM/volume of fluid method

A new method based on volume of fluid for interface tracking in the simulation of injection molding is presented. The proposed method is comprised of two main stages: accumulation and distribution of the volume fraction. In the first stage the equation for the volume fraction with a noninterfacial flux condition is solved. In the second stage the accumulated volume of fluid that arises as a consequence of the application of the first one is dispersed. This procedure guarantees that the fluid fills the available space without dispersion of the interface. The mathematical model is based on two‐phase transport equations that are numerically integrated through the control volume finite element method. The numerical results for the interface position are successfully verified with analytical results and numerical data available in the literature for one‐dimensional and two‐dimensional domains. The transient position of the advance fronts showed an effective and consistent simulation of an injection molding process. The nondispersive volume of fluid method here proposed is implemented for the simulation of nonisothermal injection molding in two‐dimensional cavities. The obtained results are represented as transient interface positions, isotherms and pressure distributions during the injection molding of low density polyethylene. Copyright © 2012 John Wiley & Sons, Ltd.