多边形有限单元形函数的比较研究

多边形有限单元形函数有wachspress插值、Laplace插值和平均值插值三种类型.本文对三种多边形有限单元形函数的性质作了比较研究,给出了三种形函数各自的优点和局限性.Waclaspress和Laplace形函数是有理函数形式,而平均值形函数是无理函数形式.三种形函数均满足单位分解性、线性完备性,且在单元边界上呈线性.在三角形单元上,它们都等价于三角形面积坐标插值.在矩形单元上,Wachspress和Laplace形函数等价于双线性多项式插值形函数.Wachspress和平均值形函数适用于任意凸多边形单元,Laplace形函数更适用于圆内接多边形单元.Wachspress形函数不能推广到含有边节点的单元,平均值形函数可以直接推广到含有边节点的单元.数值试验,验证了本文理论分析的结论.     

The improvement of numerical modeling in the solution of incompressible viscous flow problems using finite element method based on spherical Hankel shape functions

In this paper, the finite element method with new spherical Hankel shape functions is developed for simulating 2‐dimensional incompressible viscous fluid problems. In order to approximate the hydrodynamic variables, the finite element method based on new shape functions is reformulated. The governing equations are the Navier‐Stokes equations solved by the finite element method with the classic Lagrange and spherical Hankel shape functions. The new shape functions are derived using the first and second kinds of Bessel functions. In addition, these functions have properties such as piecewise continuity. For the enrichment of Hankel radial basis functions, polynomial terms are added to the functional expansion that only employs spherical Hankel radial basis functions in the approximation. In addition, the participation of spherical Bessel function fields has enhanced the robustness and efficiency of the interpolation. To demonstrate the efficiency and accuracy of these shape functions, 4 benchmark tests in fluid mechanics are considered. Then, the present model results are compared with the classic finite element results and available analytical and numerical solutions. The results show that the proposed method, even with less number of elements, is more accurate than the classic finite element method.     

Role of Prehistories in the Initial Value Problems of Fractional Viscoelastic Equations

The fractional viscoelastic equation (FVE), which is a second-order differential equation with fractional derivatives describing the dynamical behavior of a single-degree-of-freedom viscoelastic oscillator, is considered. Some viscoelastic damped mechanical systems may be described by FVEs. However, FVEs with conventional nonzero initial values cannot generally be solved. In this paper, the prehistories of the unknown functions before the initial times, referred to as the initial functions, are taken into account to solve FVEs. Mathematically, appropriate initial functions are essential for unique solutions of FVEs. Physically, the initial functions reflect the processes of giving the initial values. FVEs are solved for some initial functions both by analytical and numerical methods. The initial functions affect the solutions of FVEs. It is discussed how the solutions depend on the initial functions. Implication of the solutions to viscoelastic materials will be discussed.     

Role of Prehistories in the Initial Value Problems of Fractional Viscoelastic Equations

The fractional viscoelastic equation (FVE), which is a second-order differential equation with fractional derivatives describing the dynamical behavior of a single-degree-of-freedom viscoelastic oscillator, is considered. Some viscoelastic damped mechanical systems may be described by FVEs. However, FVEs with conventional nonzero initial values cannot generally be solved. In this paper, the prehistories of the unknown functions before the initial times, referred to as the initial functions, are taken into account to solve FVEs. Mathematically, appropriate initial functions are essential for unique solutions of FVEs. Physically, the initial functions reflect the processes of giving the initial values. FVEs are solved for some initial functions both by analytical and numerical methods. The initial functions affect the solutions of FVEs. It is discussed how the solutions depend on the initial functions. Implication of the solutions to viscoelastic materials will be discussed.     

Selection of kernel functions used in a new constitutive equation for viscoelastic fluids

A selection of kernel functions is given to be used in a new integral constitutive equation proposed by Piau whereby the deviatoric stress is calculated from the integral of the history of the past intrinsic rate of rotation and rate of deformation tensors through a representation theorem. Piau has demonstrated the objectivity of a frame moving with a given particle whose axis are directed along the eigenvectors of the rate of deformation tensor. The use of such a framework provides a new approach in the attempt to reduce the computational difficulties associated with conventional constitutive equations written in co-deformational or co-rotational reference frames.The shear and primary normal-stress material functions and the extensional (elongational) stress growth function are defined for the proposed integral constitutive equation. These material functions are used to calculate the kernel functions using steady state, stress relaxation and stress growth data of Attané in simple shear flow for monodisperse polystyrene solutions. The shear and extensional stress growth data of Meissner for a polyethylene melt are also used to show the flexibility of the rheological model.The material functions are first written in terms of five monotonically decreasing functions of the time lag between the past and the present time. Then kernel functions are chosen such that when substituted in the new integral constitutive equation they yield the functions used to describe the data. A further condition imposed on the normalized kernel functions is that they be decreasing functions of time lag.     

Experimental validation of the constitutive equations of thermoelastoplasticity including the third invariant of the stress deviator

The equations describing thermoelastoplastic deformation along nonstraight paths and taking into account the third invariant of the stress deviator are experimentally validated. The equations contain two scalar functions that are specified in base tests on tubular specimens. The test data are tabulated. The values of the scalar functions for strains, temperature, and stress mode are found by using nonlinear interpolation of the data and the temperature similarity of the functions. The stresses in elements of the body are calculated from given strains by the method of successive approximations     

Fractional Trigonometry and the Spiral Functions

This paper develops two related fractional trigonometries based on the multi-valued fractional generalization of the exponential function, the R-function. The trigonometries contain the traditional trigonometric functions as proper subsets. Also developed are relationships between the R-function and the new fractional trigonometric functions. Laplace transforms are derived for the new functions and are used to generate solution sets for various classes of fractional differential equations. Because of the fractional character of the R-function, several new trigonometric functions are required to augment the traditional sine, cosine, etc. functions. Fractional generalizations of the Euler equation are derived. As a result of the fractional trigonometry a new set of phase plane functions, the Spiral functions, that contain the circular functions as a subset, is identified. These Spiral functions display many new symmetries.     

Generalized solutions of beams with jump discontinuities on elastic foundations

Summary  The bending solutions of the Euler–Bernoulli and the Timoshenko beams with material and geometric discontinuities are developed in the space of generalized functions. Unlike the classical solutions of discontinuous beams, which are expressed in terms of multiple expressions that are valid in different regions of the beam, the generalized solutions are expressed in terms of a single expression on the entire domain. It is shown that the boundary-value problems describing the bending of beams with jump discontinuities on discontinuous elastic foundations have more compact forms in the space of generalized functions than they do in the space of classical functions. Also, fewer continuity conditions need to be satisfied if the problem is formulated in the space of generalized functions. It is demonstrated that using the theory of distributions (i.e. generalized functions) makes finding analytical solutions for this class of problems more efficient compared to the traditional methods, and, in some cases, the theory of distributions can lead to interesting qualitative results. Examples are presented to show the efficiency of using the theory of generalized functions. Received 6 June 2000; accepted for publication 24 October 2000     

On the Calculation of Stationary Solutions of Multi-Dimensional Fokker–Planck Equations by Orthogonal Functions

In this paper, nonlinear stochastic systems are investigatedvia associated Fokker–Planck equations. Their stationary solutions arecalculated by expansions into orthogonal functions, e.g. especiallyadjusted polynomials and Fourier series. The weighting functions of thenew polynomials are obtained by the application of the stochasticaveraging method. The proposed analysis is demonstrated with severalexamples. The first one is a two-dimensional problem of nonlinearoscillators driven by white noise. The second one describes two-massoscillators with independent coloured noise excitations leading tosix-dimensional probability density functions. The next example ispresenting a system driven by both harmonic and stochastic excitationleading to three-dimensional probability density functions. Finally,oscillators with dry friction characteristics are examined.     

New approximate solution for N-point correlation functions for heterogeneous materials

Statistical N-point correlation functions are used for calculating properties of heterogeneous systems. The strength and the main advantage of the statistical continuum approach is the direct link to statistical information of microstructure. Two-point correlation functions are the lowest order of correlation functions that can describe the morphology and the microstructure-properties relationship. Experimentally, statistical pair correlation functions are obtained using SEM or small x-ray scattering techniques. Higher order correlation functions must be calculated or measured to increase the precision of the statistical continuum approach. To achieve this aim a new approximation methodology is utilized to obtain N-point correlation functions for non-FGM (functional graded materials) heterogeneous microstructures. Conditional probability functions are used to formulate the proposed theoretical approximation. In this approximation, weight functions are used to connect subsets of (N?1)-point correlation functions to estimate the full set of N-point correlation function. For the approximation of three and four point correlation functions, simple weight functions have been introduced. The results from this new approximation, for three-point probability functions, are compared to the real probability functions calculated from a computer generated three-phase reconstructed microstructure in three-dimensional space. This three-dimensional reconstruction was based on an experimental two-dimensional microstructure (SEM image) of a three-phase material. This comparison proves that our new comprehensive approximation is capable of describing higher order statistical correlation functions with the needed accuracy.     

Fractional Trigonometry and the Spiral Functions

This paper develops two related fractional trigonometries based on the multi-valued fractional generalization of the exponential function, the R-function. The trigonometries contain the traditional trigonometric functions as proper subsets. Also developed are relationships between the R-function and the new fractional trigonometric functions. Laplace transforms are derived for the new functions and are used to generate solution sets for various classes of fractional differential equations. Because of the fractional character of the R-function, several new trigonometric functions are required to augment the traditional sine, cosine, etc. functions. Fractional generalizations of the Euler equation are derived. As a result of the fractional trigonometry a new set of phase plane functions, the Spiral functions, that contain the circular functions as a subset, is identified. These Spiral functions display many new symmetries.     

A uniform asymptotic analysis of dispersive wave motion

The signaling problem for the one dimensional Klein-Gordon equation with spatially varying coefficients is analyzed. A formal, uniformly valid, asymptotic expansion of the solution is obtained with the help of two families of rays, and involving four functions : two successive Bessel functions of integer order and two new functions which we call the diffraction functions. The validity of the expansion is established when the coefficients in the Klein-Gordon equation are constants, and the results are applied to a signaling problem for a class of acoustic wave guides.     

薄板小波有限元理论及其应用

利用样条小波尺度函数构造了常用的三角形和矩形薄板单元的位移函数,得到了利用小波函数表示的形函数。采用合理的局部坐标,对单元进行压缩,使单元在局部坐标区间上有其值,成功地推导出了分域的三角形和矩形薄板小波有限元列式。在此基础上,提出了弹性地基薄板的小波有限元求解方法。通过两个算例对薄板的挠度和弯矩进行了计算,数值结果表明,求解结果具有收敛快、精度高的特点。     

On Nonlinear Stability of Isotropic Models in Stellar Dynamics

Consider an aggregation of mass particles in space which attract each other according to Newton's law of attraction. This system can be described by distribution functions satisfying the Vlasov equation and the Poisson laws. We obtain the nonlinear stability of certain stationary states, including those obeying modified Emden's laws. A priori estimates of the energy-Casimir functions around stationary states are established for distribution functions for which the L^5/3L^{5/3}-norms of the density functions are uniformly bounded. A uniform bound on the kinetic energy of the system readily implies that these norms of the density functions are indeed uniformly bounded. In this way we prove nonlinear stability.     

Determination of relaxation and retardation spectrum by inverse functional filtering

The article is devoted for the determination of the relaxation and retardation spectrum (RRS) from monotonic time- and frequency-domain material functions by the inverse functional filters executing discrete convolution algorithms for geometrically spaced data. It is shown that the problem of RRS determination from a wide variety of material functions leads to the three inverse filtering tasks on a logarithmic time or frequency scale with the three specific frequency responses concerning: (i) the time-domain functions, (ii) the real parts and (iii) the imaginary parts of the frequency-domain functions, and three algorithms (having the versions with even and odd number of coefficients) are to be applied to: (i) time-domain compliance and modulus functions, (ii) their derivatives, and (iii) frequency-domain functions. It is demonstrated that ill-posedness of an inverse filter manifests as large sampling-rate-dependent noise amplification coefficients. A novel regularization strategy allowing to ensure the desired noise immunity is proposed based on choosing sampling rate for geometrically spaced data. The performance of the algorithms is investigated. Optimal sampling rates are disclosed for specific material functions. The frequency range of 2–3 decades is established to be optimal for the recovery of a single RRS point estimate ensuring maximum accuracy with reasonable noise immunity. Practical algorithms are proposed for recovering RRS from the real and imaginary parts of frequency-domain functions. Some known non-parametric methods are compared with the suggested functional filters.     

Numerical solution of singular integral equations in stress concentration problems under longitudinal shear loading

Summary The paper deals with numerical solutions of singular integral equations in stress concentration problems for longitudinal shear loading. The body force method is used to formulate the problem as a system of singular integral equations with Cauchy-type singularities, where unknown functions are densities of body forces distributed in the longitudinal direction of an infinite body. First, four kinds of fundamental density functions are introduced to satisfy completely the boundary conditions for an elliptical boundary in the range 0≤φ _k ≤2π. To explain the idea of the fundamental densities, four kinds of equivalent auxiliary body force densities are defined in the range 0≤φ _k ≤π/2, and necessary conditions that the densities must satisfy are described. Then, four kinds of fundamental density functions are explained as sample functions to satisfy the necessary conditions. Next, the unknown functions of the body force densities are approximated by a linear combination of the fundamental density functions and weight functions, which are unknown. Calculations are carried out for several arrangements of elliptical holes. It is found that the present method yields rapidly converging numerical results. The body force densities and stress distributions along the boundaries are shown in figures to demonstrate the accuracy of the present solutions. Received 26 May 1998; accepted for publication 27 November 1998     

Modal Representation of Stress in Flexible Multibody Simulation

An application of the floating frame of reference formulation together with the nodal approach using quasi-comparison functions as shape functions allows an efficient analysis of stress in the flexible bodies of a multibody system. This is demonstrated using two simple examples. They are chosen to demonstrate the effects of various choices of shape functions and associated body reference frames. In the floating frame of reference formulation the equations of motion are linearized assuming the deformations to be small. The quasi-comparison functions, i.e. shape functions, can be selected in ways to increase the range of validity of the linearized equations of motion. The latter goal is achieved as well by so-called substructuring techniques. Combining both of the methodologies, one obtains efficient models for flexible multibody simulation.     

Calculation of stresses in a multiply-connected anisotropic plate using stress functions of multiple complex variables

On the basis of anisotropic mathematical elasticity, using multiple conformal representations, the stress functions of multiple complex variables for an infinite multiply-connected anisotropic plate are derived. The functions are developed in Fourier series on unit circles, and the unknown coefficients of the functions are determined by undetermined coefficients method. Then the stresses in the plate can be calculated. A plate containing multiple elliptical holes or cracks is discussed, and the corresponding FORTRAN77 program is developed. Five examples are given. The results show that this method is very effective and convenient. The project supported by Aeronautical Science Foundation of China     

Numerical simulation of density distribution during compaction of iron powders

Summary The compaction process of iron powder is considered. Due to negligible elastic strains the rigid-plastic model is applied. A yield condition containing the first stress invariant is used. All material functions depend on the relative density of the powder, which changes during the compaction process. Siebel friction law is applied, and the friction factor is considered to be depending on the relative density. Various material functions are applied in the numerical simulation, and the results are compared with experimentally obtained data. The best fitting material functions and friction factors are obtained. Accepted for publication 18 July 1996     

On the derivatives of a subclass of isotropic tensor functions of a nonsymmetric tensor

This paper studies a subclass of isotropic tensor-valued functions of a nonsymmetric tensor, which satisfy the commutative condition, and their derivatives. This subclass of tensor functions includes tensor power series, exponential tensor function, etc., and is more general than those investigated before. In the case of three distinct eigenvalues, the derivatives of these tensor functions are constructed by solving a tensor equation, which is acquired by differentiating the commutative condition. By taking limits, the results are extended to the cases of repeated eigenvalues.