On the uses of special crack tip elements in numerical rock fracture mechanics

Numerical methods such as boundary element methods are widely used for the stress analysis in solid mechanics. These methods are also used for crack analysis in rock fracture mechanics. There are singularities for the stresses and displacements at the crack tips in fracture mechanics problem, which decrease the accuracy of the numerical results in areas very close to the crack ends. To overcome this, higher order elements and isoperimetric higher order elements have been used. Recently, special crack tip elements have been proposed and used in most of the numerical fracture mechanics models. These elements can drastically increase the accuracy of the results near the crack tips, but in most of the models only one special crack tip element has been used for each crack end. In this study the uses of higher order crack tip elements are discussed and a higher order displacement discontinuity method is used to investigate the effect of these elements on the accuracy of the results in some crack problems. The useful shape functions for two special crack tip elements, are derived and given in the text and appendix for both infinite and semi-infinite plane problems. In this analysis both Mode I and Mode II stress intensity factors are computed . Some example problems are solved and the computed results are compared with the results given in the literature. The numerical results obtained here are in good agreement with those cited in the literature. For the curved crack problem, the strain energy release rate, G can be calculated accurately in the vicinity of the crack tips by using the higher order displacement discontinuity method with a quadratic variation of displacement discontinuity elements and with two special crack tip elements at each crack end.     

Verification of a mixed high‐order accurate DNS code for laminar turbulent transition by the method of manufactured solutions

This paper presents results on a verification test of a Direct Numerical Simulation code of mixed high‐order of accuracy using the method of manufactured solutions (MMS). This test is based on the formulation of an analytical solution for the Navier–Stokes equations modified by the addition of a source term. The present numerical code was aimed at simulating the temporal evolution of instability waves in a plane Poiseuille flow. The governing equations were solved in a vorticity–velocity formulation for a two‐dimensional incompressible flow. The code employed two different numerical schemes. One used mixed high‐order compact and non‐compact finite‐differences from fourth‐order to sixth‐order of accuracy. The other scheme used spectral methods instead of finite‐difference methods for the streamwise direction, which was periodic. In the present test, particular attention was paid to the boundary conditions of the physical problem of interest. Indeed, the verification procedure using MMS can be more demanding than the often used comparison with Linear Stability Theory. That is particularly because in the latter test no attention is paid to the nonlinear terms. For the present verification test, it was possible to manufacture an analytical solution that reproduced some aspects of an instability wave in a nonlinear stage. Although the results of the verification by MMS for this mixed‐order numerical scheme had to be interpreted with care, the test was very useful as it gave confidence that the code was free of programming errors. Copyright © 2009 John Wiley & Sons, Ltd.     

On the Use of Implicit Integration Methods and the Absolute Nodal Coordinate Formulation in the Analysis of Elasto-Plastic Deformation Problems

In the general theory of continuum mechanics, the state of rotation and deformation of material points can be uniquely defined from the displacement field by using the nine independent components of the displacement gradients. For this reason, the use of the absolute rotation parameters as nodal coordinates, without relating them to the displacement gradients, leads to coordinate redundancy that leads to numerical and fundamental problems in many existing large rotation finite element formulations. Because of this fundamental problem, special measures that require modifications of the numerical integration methods were proposed in the literature in order to satisfy the principle of work and energy. As demonstrated in this paper, no such measures need to be taken when the finite element absolute nodal coordinate formulation is used since the principle of work and energy are automatically satisfied. This formulation does not suffer from the problem of coordinate redundancy and ensures the continuity of stresses and strains at the nodal points. In this study, the use of the implicit integration methods with the consistent Lagrangian elasto-plastic tangent moduli is examined when the absolute nodal coordinate formulation is used. The performance of different numerical integration methods in the dynamic analysis of large elasto-plastic deformation problems is investigated. It is shown that all these methods, in the case of convergence, yield a solution that satisfies the principle of work and energy without the need of taking any special measures. Semi-implicit integration methods, however, can lead to numerical difficulties in the case of very stiff problems due to the linearization made in these methods in order to avoid the iterative Newton--Raphson procedure. It is also demonstrated that the use of the consistent Lagrangian-plastic tangent moduli derived in this investigation using the absolute nodal coordinate formulation leads to better convergence of the iterative Newton--Raphson procedure used in the implicit integration methods.     

THE NUMERICAL SOLUTION OF A SINGULARLY PERTURBED PROBLEM FOR SEMILINEAR PARABOLIC DIFFERENTIAL EQUATION

The numerical solution of a singularly perturbed problem for the semilinear parabolicdifferential equation with parabolic boundary layers is discussed.A nonlinear two-leveldifference scheme is constructed on the special non-uniform grids.The uniform convergenceof this scheme is proved and some numerical examples are given.     

Efficient quadrature for a boundary element method to compute three-dimensional Stokes flow

A collocation-type boundary element method based on bilinear B-splines is used for the numerical solution of the Stokes Dirichlet problem in bounded domains D ? R³. The computation of the influence matrix requires the numerical evaluation of weakly singular integrals on the domain boundary if the usual double-layer potential ansatz is chosen. Here mostly standard methods with disjoint grids for collocation and integration are used. We develop a special integration scheme based on triangular co-ordinates near the singularity and show its efficiency compared with the method mentioned above.     

UNSTEADY HYDROELASTICITY OF FLOATING PLATES

Asymptotic and numerical analyses of unsteady hydroelastic behaviour of a floating plate due to given external loads are presented. The main parameters are the plate length and duration of the external loads. For very long plates (VLFS) the problem is decoupled and its approximate solution is given by the method of matched asymptotic expansions. For a short duration of the external loads and small length of the plate (impact onto a floating plate) the problem is coupled, but gravity effects can be neglected in determining the maximum of both the plate deflection and bending stresses in the plate. In this case, the problem is solved numerically by the method of normal modes. If the plate is short but the load duration is moderate, the rigid-body motion of the plate and its elastic vibrations can be approximately separated. In the general case, it is suggested that the coupled problem can be treated numerically by the method of normal modes. In order to construct an appropriate numerical algorithm, ideas inspired by the asymptotic analysis are used.     

Fluid flow in an arbitrary cascade on an axisymmetric stream surface in a variable thickness layer

A solution is given for the problem of flow past a cascade on an axisymmetric stream surface in a layer of variable thickness, which is a component part of the approximate solution of the three-dimensional problem for a three-dimensional cascade. Generalized analytic functions are used to obtain the integral equation for the potential function, which is solved via iteration method by reduction to a system of linear algebraic equations. An algorithm and a program for the Minsk-2 computer are formulated. The precision of the algorithm is evaluated and results are presented of the calculation of an example cascade.In the formulation of [1, 3] the problem of flow past a three-dimensional turbomachine cascade is reduced approximately to the joint solution of two-dimensional problems of the averaged axisymmetric flow and the flow on an axisymmetric stream surface in an elementary layer of variable thickness.In the following we solve the second problem for an arbitrary cascade with finite thickness rotating with constant angular velocity in ideal fluid flow: the solution is carried out on a Minsk-2 computer.Many studies have been devoted to this problem. A method for solving the direct problem for a cascade of flat plates in a hyperbolic layer was presented in [2]. Methods were developed in [1, 3] for constructing the flow for the case of a channel with variable thickness; these methods are approximately applicable for dense cascades but yield considerable error for small-load turbomachine cascades. The solution developed in [4], somewhat reminiscent of that of [2], is applicable for thin, slightly curved profiles in a layer with monotonically varying thickness. A solution has been given for a circular cascade for layers varying logarithmically [5] and linearly [6]. Approximate methods for slightly curved profiles in a monotonically varying layer with account for layer variability only in the discharge component were examined in [7–9]. A solution is given in [10] for an arbitrary layer by means of the relaxation method, which yields a roughly approximate flow pattern. The general solution of the problem by means of potential theory and the method of singularities presented in [11] is in error because of neglect of the crossflow through the skeletal line. The computer solution of [12] contains an unassessed error for the calculations in an arbitrary layer. The finite difference method is used in [13] to solve the differential equation of flow, which is illustrated by numerical examples for monotonie layers of axial turbomachines. The numerical solution of [13] is very complex.The solution presented below is found in the general formulation with respect to the geometric parameters of the cascade and the axisymmetric surface and also in terms of the layer thickness variation law.The numerical solution requires about 15 minutes of machine time on the Minsk-2 computer.     

Optimal Design of Thin-Walled Cylinders of Variable Wall-Thickness Subject to Flexural and Torsional Stiffness Constraints

ABSTRACT

This paper considers the problem of determining center-line shape and wall-thickness distribution of a thin-walled cylinder of given center-line length that uses the minimum possible amount of material to achieve prescribed minimum stiffnesses in torsion and bending in a given plane. Necessary optimality conditions are derived and the solution is found partly in closed form and partly by numerical methods. Optimal solutions are presented for various stiffness ratios and compared with other admissible designs.     

均布荷载作用下功能梯度悬臂梁弯曲问题的解析解

采用弹性力学半逆解法,假设所有材料常数沿梁厚度方向按同一函数规律变化,求得了功能梯度悬臂梁在均布载荷作用下的解析解.该解退化到各向同性均匀弹性情况时与已有的理论解相一致.对弹性模量按指数函数梯度变化的算例进行了分析.所得到的解对任意梯度函数均成立,可作为数值解以及简化理论的检验依据.     

A VORTEX LATTICE METHOD FOR HIGH-SPEED PLANNING

A three-dimensional numerical model using vortex lattice methods (VLMs) is developed to solve the steady planning problem. Planing hydrodynamics have similarities to the aerodynamic swept wing problem—the fundamental difference being the existence of a free surface. Details of the solution scheme are discussed, including the special features of the VLM used here in obtaining accurate flows at the leading and side edges. Computational results are presented and compared with existing theories and experiments.     

The convergent condition and united formula of step reduction method

The step reduction method was first suggested by Prof. Yeh Kai-yuan^[1]. This method has more advantages than other numerical methods. By this method, the analytic expression of solution can he obtained for solving nonuniform elastic mechanics. At the same time, its calculating time is very short and convergent speed very fast. In this paper. the convergent condition and united formula of step reduction method are given by mathematical method. It is proved that the solution of displacement and stress resultants obtained by this method can converge to exact solution uniformly. when the convergent condition is satisfied. By united formula, the analytic solution can be expressed as matrix form, and therefore the former complicated expression can be avoided. Two numerical examples are given at the end of this paper which indicate that by the theory in this paper, a right model can be obtained for step reduction method.Project Supported by Science and Technic Fund of the National Education Committee.The author would like to thank Prof. Yeh Kai-yuan for his directing.     

Solution of simultaneous equations of cosine law arising from subjectivity geometry

This paper discusses the solution of a group of two-order six elements rootedalgebraic simultaneous equations set up by cosine law arising from the application example of subjectivity geometry^[1]. By means of the implicit function theorem, this paper proves that there exists a unique real solution of those equations. Transforming this problem into an unconstrained nonlinear optimization problem, the solution can be found by known methods. A numerical example by descent method is given.Supported by Scientific Foundation of South China Unviersity of Technology. Ben Xiu-ming took part in the calculation.     

内聚力模型裂纹问题分析的解析奇异单元

内聚力模型已经被广泛应用于需要考虑断裂过程区的裂纹问题当中,然而常用的数值方法应用于分析内聚力模型裂纹问题时还存在着一些不足,比如不能准确的给出断裂过程区的长度、需要网格加密等。为了克服这些缺点,论文构造了一个新型的解析奇异单元,并将之应用于基于内聚力模型的裂纹分析当中。首先将虚拟裂纹表面处的内聚力用拉格拉日插值的方法近似表示为多项式的形式,而多项式表示的内聚力所对应的特解可以被解析地给出。然后利用一个简单的迭代分析,基于内聚力模型的裂纹问题就可以被模拟出来了。最后,给出二个数值算例来证明本文方法的有效性。     

Tomographic diagnostics of technical materials and biological tissues using electric current

A mathematical model for impedance computer tomography methods is considered. The continuum formulation of the main problem is studied. Resolving integral equations are derived. A solution algorithm based on the Bubnov—Galerkin method with linearization of nonlinear resolving equations is developed. A numerical example is given, and numerical results are analyzed. Some drawbacks of the model are considered together with methods for avoiding them.     

非光滑方程组方法在求解非匹配网格接触问题中的应用

将非光滑方程组方法与Mortar StS接触模型(Mortar Segment-to-Segment)相结合,来求解接触面网格非匹配时的弹性接触问题.其中,非光滑方程组方法是求解弹性摩擦接触问题的有效方法,具有精确满足接触条件、迭代算法收敛性有理论保证的优点,但目前仅用于求解网格匹配的接触问题.Mortar StS接触模型可以较为方便地处理网格非匹配接触问题,其特点是不引入过多约束,满足接触分片检验条件,但目前大都采用“试验-误差”迭代方法求解控制方程,对于复杂接触问题,其收敛性不易保证.因此,将二者结合来处理网格非匹配接触问题,既可以提高求解精度,又能使得算法的收敛性得到理论保证.数值算例对接触分片检验和算法的计算精度进行了验证.     

Finite difference scheme for the solution of fluid flow problems on non‐staggered grids

A new finite difference methodology is developed for the solution of computational fluid dynamics problems that do not require the use of staggered grid systems. Previous successful and robust non‐staggered methods, which used primitive variables and mass conservation in order to solve the pressure field, either interpolate cell‐face velocities or interpolate the pressure gradients in a special way, usually with an upwind‐bias to avoid the problem of odd–even coupling between the velocity and pressure fields. The new methodology presented does not detail a ‘special interpolation procedure for a primitive variable’, however, it manages to avoid the problem of odd–even coupling. The odd–even coupling is avoided by applying fourth‐order dissipation to the pressure field. It is shown that this approach can be regarded as a modified Rhie and Chow scheme. The method is implemented using a SIMPLE‐type algorithm and is applied to two test problems: laminar flow over a backward‐facing step and laminar flow in a square cavity with a driven lid. Good agreement is obtained between the numerical solutions and the corresponding benchmark solutions. The pressure dissipation term was found to successfully suppress wiggles in the pressure field. Copyright © 2000 John Wiley & Sons, Ltd.     

A combined experimental and numerical method for the solution of generalized elasticity problems

This work presents the method for the investigation of three-dimensionally stressed bodies with arbitrary shape which are under the action of an outside system of arbitrary forces. The combined method is based on syntheses of photoelastic experimental methods (other experimental methods may also be used) and digital methods of discrete analysis. Experimental procedures are used for defining superfluous boundary conditions. The boundary-value problem with such boundary conditions is solved by numerical methods. This approach qualitatively changes the very essence of experimental methods and essentially widens their range. It reduces the amount of measurements required and, at the same time, allows one to obtain complete stress fields throughout a body in a short time. In comparison with numerical methods, the combined method increases the accuracy of problem solutions and, at the same time, reduces the time required for complete investigations.     

Numerical stress-strain analysis of a nonthin elastic plate

A numerical solution to elastic-equilibrium problems for nonthin plates is proposed. The solution is obtained by using the curvilinear-mesh method in combination with Vekua’s method. The efficiency (rapid convergence and accuracy) of this approach is demonstrated by solving test problems for thick plates that can also be solved exactly or approximately by other methods. A numerical solution is obtained to the bending problem for orthotropic nonthin plates of constant and varying thickness __________ Translated from Prikladnaya Mekhanika, Vol. 42, No. 3, pp. 119–126, March 2006.     

A boundary-value problem for a nonlinear heat-conduction equation

Methods for the numerical solution of the boundary-value problem are comidered. One method is related to the method of successive approximations and the other employs the collocation method [1]. A relationship between the latter method and the Ritz and Galerkin methods [2] is shown. An application of the collocation method to the nonstationary problem is given. The approximate solution is represented in analytic form. A way of finding the absolute error of the approximate solution is given.