A phase field approach for multivariant martensitic transformations of stable and metastable phases

A phase field model approach for multivariant martensitic transformations of stable and metastable phases is introduced. The evolution of the microstructure is examined with respect to elastic energy minimization in which one or two martensitic orientation variants are considered. In this context, the martensitic nucleation behavior is simulated for different activation barriers. Furthermore, the influence of time-dependent external loads on the formation of the different phases is studied. The numerical implementation is performed with finite elements and an implicit time integration scheme.     

Preconditioned finite element algorithms for 3D Stokes flows

Preconditioned conjugate gradient algorithms for solving 3D Stokes problems by stable piecewise discontinuous pressure finite elements are presented. The emphasis is on the preconditioning schemes and their numerical implementation for use with Hermitian based discontinuous pressure elements. For the piecewise constant discontinuous pressure elements, a variant implementation of the preconditioner proposed by Cahouet and Chabard for the continuous pressure elements is employed. For the piecewise linear discontinuous pressure elements, a new preconditioner is presented. Numerical examples are presented for the cubic lid-driven cavity problem with two representative elements, i.e. the Q2-PO and the Q2-P1 brick elements. Numerical results show that the preconditioning schemes are very effective in reducing the number of pressure iterations at very reasonable costs. It is also shown that they are insensitive to the mesh Reynolds number except for nearly steady flows (Re_m → 0) and are almost independent of mesh sizes. It is demonstrated that the schemes perform reasonably well on non-uniform meshes.     

Finite element simulations of martensitic phase transitions and microstructures based on a strain softening model

A new approach for modeling multivariant martensitic phase transitions (PT) and martensitic microstructure (MM) in elastic materials is proposed. It is based on a thermomechanical model for PT that includes strain softening and the corresponding strain localization during PT. Mesh sensitivity in numerical simulations is avoided by using rate-dependent constitutive equations in the model. Due to strain softening, a microstructure comprised of pure martensitic and austenitic domains separated by narrow transition zones is obtained as the solution of the corresponding boundary value problem. In contrast to Landau-Ginzburg models, which are limited in practice to nanoscale specimens, this new phase field model is valid for scales greater than 100 nm and without upper bound. A finite element algorithm for the solution of elastic problems with multivariant martensitic PT is developed and implemented into the software ABAQUS. Simulated microstructures in elastic single crystals and polycrystals under uniaxial loading are in qualitative agreement with those observed experimentally.     

Converged solutions of the Newtonian extrudate‐swell problem

Both the axisymmetric and the planar Newtonian extrudate‐swell problems are solved using the standard and the singular finite element methods. In the latter method, special elements that incorporate the radial form of the stress singularity are used around the exit of the die. The convergence of each of the two methods with mesh refinement is studied for various values of the Reynolds and the capillary numbers. The numerical results show that the singular finite elements perform well if coarse or moderately refined meshes are used, and appear to be superior to the standard finite elements only when the Reynolds number is low and the surface tension is not large. The standard finite elements perform better as the surface tension or the Reynolds number are increased. This implies that the effect of the stress singularity on the accuracy of the numerical solution in the neighborhood of the die exit becomes less significant when the Reynolds number is high or the surface tension is large. Copyright © 1999 John Wiley & Sons, Ltd.     

A study of penalty elements for incompressible laminar flows

A finite element model is developed based on the penalty formulation to study incompressible laminar flows. The study includes a number of new quadrilateral and triangular elements for 2-dimensional flows and a number of new hexahedral and tetrahedral elements for 3-dimensional flows. All elements employ continuous velocity approximations and discontinuous pressure approximations respecting the LBB condition of numerical instability. An incremental Newton–Raphson method coupled with the Broyden method is used to solve the non-linear equations. Several numerical examples (colliding flow, cavity flow, etc.) are presented to assess the efficiency of elements.     

Solution of the Incompressible Unsteady Navier–Stokes Equations Only in Terms of the Velocity Components

A method which uses only the velocity components as primitive variables is described for solution of the incompressible unsteady Navier–Stokes equations. The method involves the multiplication of the primitive variable-based Navier–Stokes equations with the unit normal vector of finite volume elements and the integration of the resulting equations along the boundaries of four-node quadrilateral finite volume elements. Therefore, the pressure term is eliminated from the governing equations and any difficulty associated with pressure or vorticity boundary conditions is avoided. The equations are discretized on four-node quadrilateral finite volume elements by using the second-order-accurate central finite differences with the mid-point integral rule in space and the first-order-accurate backward finite differences in time. The resulting system of algebraic equations is solved in coupled form using a direct solver. As a test case, an impulsively accelerated lid-driven cavity flow in a square enclosure is solved in order to verify the accuracy of the present method.     

The intrinsic XFEM for two‐fluid flows

In two‐fluid flows, jumps and/or kinks along the interfaces are present in the resulting velocity and pressure fields. Standard methods require mesh manipulations with the aim that either element edges align with the interfaces or that the mesh is sufficiently refined near the interfaces. In contrast, enriched methods, such as the extended finite element method (XFEM), enable the representation of arbitrary jumps and kinks inside elements. Thereby, optimal convergence can be achieved for two‐fluid flows with meshes that remain fixed throughout the simulation. In the intrinsic XFEM, in contrast to other enriched methods, no more unknowns are present in the approximation than in a standard finite element approximation. In this work, the intrinsic XFEM is employed for the simulation of incompressible two‐fluid flows. Numerical results are shown for a number of test cases and prove the success of the method. Copyright © 2008 John Wiley & Sons, Ltd.     

多边形有限元研究进展

有限元法是数值求解偏微分方程边值问题的重要方法,采用不规则多边形单元网格, 可以方便有效地模拟材料的力学性能, 又使得区域网格剖分变得灵活方便. 特别是对于复杂的几何形状, 多边形单元网格具有更大的优势. 本文对国内外有关多边形有限元法的最新进展作了初步的总结和评述, 主要以基于位移法的多边形有限元为主.论述了多边形有限元的发展历史, 给出了多边形单元上的Wachspress插值、Laplace插值和重心坐标的一些最新研究成果. 与经典有限元法形函数为多项式形式不同, 多边形单元的形函数为有理函数或者无理函数形式. 多边形单元插值形函数满足线性完备性, 可以再现线性位移场, 像经典有限元法一样直接施加本质边界条件; 插值函数在多边形的边界上是线性的,确保不同单元间的自动协调. 不同单元的插值形函数表达公式形式统一, 方便混合单元网格计算的程序编写. 提出了多边形有限元法今后需要研究的问题.      

A numerical study on flow through the spiral casing of a hydraulic turbine

Flow through the spiral casing of a hydraulic turbine was analyzed. Reynolds averaged Navier–Stokes equations were solved using a finite element method. The physical domain was divided into a number of hexahedral elements which are isoparametrically mapped onto standard cubic elements. Numerical integration for the unsteady momentum equation is performed over such hexahedral elements to obtain a provisional velocity field. Compliance with the mass conservation equation and determination of the pressure correction are accomplished through an iterative procedure. The velocity distribution inside the spiral casing corroborates the results available in literature. The static pressure at the midplane generally decreases from the outside wall towards the exit of the spiral casing. © 1998 John Wiley & Sons, Ltd.     

Numerical simulations of the steady Navier–Stokes equations using adaptive meshing schemes

In this paper, we consider an adaptive meshing scheme for solution of the steady incompressible Navier–Stokes equations by finite element discretization. The mesh refinement and optimization are performed based on an algorithm that combines the so‐called conforming centroidal Voronoi Delaunay triangulations (CfCVDTs) and residual‐type local a posteriori error estimators. Numerical experiments in the two‐dimensional space for various examples are presented with quadratic finite elements used for the velocity field and linear finite elements for the pressure. The results show that our meshing scheme can equally distribute the errors over all elements in some optimal way and keep the triangles very well shaped as well at all levels of refinement. In addition, the convergence rates achieved are close to the best obtainable. Extension of this approach to three‐dimensional cases is also discussed and the main challenge is the efficient implementation of three‐dimensional CfCVDT generation that is still under development. Copyright © 2007 John Wiley & Sons, Ltd.     

Finite Element Analysis of Elasto-plastic Plate Bending Problems using Transition Rectangular Plate Elements

In this work, the finite element analysis of the elasto-plastic plate bending problems is carried out using transition rectangular plate elements. The shape functions of the transition plate elements are derived based on a practical rule. The transition plate elements are all quadrilateral and can be used to obtain efficient finite element models using minimum number of elements. The mesh convergence rates of the models including the transition elements are compared with the regular element models. To verify the developed elements, simple tests are demonstrated and various elasto-plastic problems are solved. Their results are compared with ANSYS results.The English text was polished by Keren Wang.     

??????????????????????

将有限元方法引入到塑性极限分析中,采用刚体有限元离散挡土墙后土体计算区域, 同时构造运动许可速度场,在满足屈服条件、流动法则、虚功方程以及相应的边界条件的基 础上,建立约束方程,引入数学规划方法求解挡土墙在不同变位模式下极限土压力分布. 算 例说明了该方法的正确性和有效性.     

沉管隧道结构的空间受力性态分析

目前沉管隧道的结构设计计算大我和一维弹性地基梁模型,或在此基础上再取若干个横截面进行平面分析。本文以甬江水底沉管隧道为工程背景,采用墙单元和板壳单元建立沉和结构的三维有限元分析模型,用水土分算原理确定沉管结构的水土荷载,对沉管隧道结构水土荷载的空间分布及沉管结构的空间受力性态进行了数值计算与分析,探讨了水位对沉管结构水土荷载值的影响以及沉管板底与基础的部分脱空和沉管顶部覆土厚度等因素对沉管结构内力     

新型p型板单元及其在结构振动分析中的应用

利用Legendre正交多项式作为形函数基底函数,开发了两种新型的通用p型板单元.单元矩阵的解析积分保证了p型有限单元解的精确性及单调收敛性,计算实例表明所开发的p型有限单元计算结果随基底函数中附加项数量的增加而快速收敛,且它们的计算精度远高于一般线性单元.另外,p型板单元不使用缩减积分也能分析薄板的振动问题,利用它们收敛率高的特点,分析了结构破坏的时频特性.p型有限单元仿真结果与实测结果良好的吻合证明了它们用于结构振动响应分析的有效性.     

A singular finite element for Stokes flow: The stick–slip problem

Abrupt changes in boundary conditions in viscous flow problems give rise to stress singularities. Ordinary finite element methods account effectively for the global solution but perform poorly near the singularity. In this paper we develop singular finite elements, similar in principle to the crack tip elements used in fracture mechanics, to improve the solution accuracy in the vicinity of the singular point and to speed up the rate of convergence. These special elements surround the singular point, and the corresponding field shape functions embody the form of the singularity. Because the pressure is singular, there is no pressure node at the singular point. The method performs well when applied to the stick–slip problem and gives more accurate results than those from refined ordinary finite element meshes.     

Mixed-mode thermal stress intensity factors from the finite element discretized symplectic method

A finite element discretized symplectic method is introduced to find the thermal stress intensity factors (TSIFs) under steady-state thermal loading by symplectic expansion. The cracked body is modeled by the conventional finite elements and divided into two regions: near and far fields. In the near field, Hamiltonian systems are established for the heat conduction and thermoelasticity problems respectively. Closed form temperature and displacement functions are expressed by symplectic eigen-solutions in polar coordinates. Combined with the analytic symplectic series and the classical finite elements for arbitrary boundary conditions, the main unknowns are no longer the nodal temperature and displacements but are the coefficients of the symplectic series after matrix transformation. The TSIFs, temperatures, displacements and stresses at the singular region are obtained simultaneously without any post-processing. A number of numerical examples as well as convergence studies are given and are found to be in good agreement with the existing solutions.     

Determination of stress intensity factors by the finite element discretized symplectic method

When rewriting the governing equations in Hamiltonian form, analytical solutions in the form of symplectic series can be obtained by the method of separation of variable satisfying the crack face conditions. In theory, there exists sufficient number of coefficients of the symplectic series to satisfy any outer boundary conditions. In practice, the matrix relating the coefficients to the outer boundary conditions is ill-conditioned unless the boundary is very simple, e.g., circular. In this paper, a new two-level finite element method using the symplectic series as global functions while using the conventional finite element shape functions as local functions is developed. With the available classical finite elements and symplectic series, the main unknowns are no longer the nodal displacements but are the coefficients of the symplectic series. Since the first few coefficients are the stress intensity factors, post-processing is not required. A number of numerical examples as well as convergence studies are given.     

A New Flexible Multibody Beam Element Based on the Absolute Nodal Coordinate Formulation Using the Global Shape Function and the Analytical Mode Shape Function

Several techniques for the reduced dimensionality of finite elementformulations were considered as component mode reduction methods in themiddle sixties. These techniques are widely used in flexiblemultibody simulations for solving small deformation problems. Theabsolute nodal coordinate formulation for solving large rotation anddeformation problems has been established as a full finite elementmethod instead of using similar kinds of reduction techniques. In thispaper, a reduced order absolute nodal coordinate formulation is newlyestablished by introducing the global beam shape function and theanalytical deformation modes as a full finite element. This formulationleads to a constant and symmetric mass matrix as the conventionalabsolute nodal coordinate formulation, and makes it possible to reducethe number of elements and system coordinates of the beam structurewhich undergoes large rotations and large deformations. Numericalexamples show that the excellent agreements between thepresent formulation and the conventional absolute nodal coordinateformulation using a large number of elements are examined. These results demonstratethat the present formulation has high accuracy in the sense that thepresent solutions are similar to the conventional ones with fewersystem coordinates, and high efficiency in computation.     

Finite difference and finite element methods for mhd channel flows

A Galerkin finite element method and two finite difference techniques of the control volume variety have been used to study magnetohydrodynamic channel flows as a function of the Reynolds number, interaction parameter, electrode length and wall conductivity. The finite element and finite difference formulations use unequally spaced grids to accurately resolve the flow field near the channel wall and electrode edges where steep flow gradients are expected. It is shown that the axial velocity profiles are distorted into M-shapes by the applied electromagnetic field and that the distortion increases as the Reynolds number, interaction parameter and electrode length are increased. It is also shown that the finite element method predicts larger electromagnetic pinch effects at the electrode entrance and exit and larger pressure rises along the electrodes than the primitive-variable and streamfunction–vorticity finite difference formulations. However, the primitive-variable formulation predicts steeper axial velocity gradients at the channel walls and lower axial velocities at the channel centreline than the streamfunction–vorticity finite difference and the finite element methods. The differences between the results of the finite difference and finite element methods are attributed to the different grids used in the calculations and to the methods used to evaluate the pressure field. In particular, the computation of the velocity field from the streamfunction–vorticity formulation introduces computational noise, which is somewhat smoothed out when the pressure field is calculated by integrating the Navier–Stokes equations. It is also shown that the wall electric potential increases as the wall conductivity increases and that, at sufficiently high interaction parameters, recirculation zones may be created at the channel centreline, whereas the flow near the wall may show jet-like characteristics.