Symmetric smoothed particle hydrodynamics (SSPH) method and its application to elastic problems

We discuss the symmetric smoothed particle hydrodynamics (SSPH) method for generating basis functions for a meshless method. It admits a larger class of kernel functions than some other methods, including the smoothed particle hydrodynamics (SPH), the modified smoothed particle hydrodynamics (MSPH), the reproducing kernel particle method (RKPM), and the moving least squares (MLS) methods. For finding kernel estimates of derivatives of a function, the SSPH method does not use derivatives of the kernel function while other methods do, instead the SSPH method uses basis functions different from those employed to approximate the function. It is shown that the SSPH method and the RKPM give the same value of the kernel estimate of a function but give different values of kernel estimates of derivatives of the function. Results computed for a sine function defined on a one-dimensional domain reveal that the L ², the H ¹ and the H ² error norms of the kernel estimates of a function computed with the SSPH method are less than those found with the MSPH method. Whereas the L ² and the H ² norms of the error in the estimates computed with the SSPH method are less than those with the RKPM, the H ¹ norm of the error in the RKPM estimate is slightly less than that found with the SSPH method. The error norms for a sample problem computed with six kernel functions show that their rates of convergence with an increase in the number of uniformly distributed particles are the same and their magnitudes are determined by two coefficients related to the decay rate of the kernel function. The revised super Gauss function has the smallest error norm and is recommended as a kernel function in the SSPH method. We use the revised super Gauss kernel function to find the displacement field in a linear elastic rectangular plate with a circular hole at its centroid and subjected to tensile loads on two opposite edges. Results given by the SSPH and the MSPH methods agree very well with the analytic solution of the problem. However, results computed with the SSPH method have smaller error norms than those obtained from the MSPH method indicating that the former will give a better solution than the latter. The SSPH method is also applied to study wave propagation in a linear elastic bar.     

SSPH basis functions for meshless methods, and comparison of solutions with strong and weak formulations

We propose a new and simple technique called the Symmetric Smoothed Particle Hydrodynamics (SSPH) method to construct basis functions for meshless methods that use only locations of particles. These basis functions are found to be similar to those in the Finite Element Method (FEM) except that the basis for the derivatives of a function need not be obtained by differentiating those for the function. Of course, the basis for the derivatives of a function can be obtained by differentiating the basis for the function as in the FEM and meshless methods. These basis functions are used to numerically solve two plane stress/strain elasto-static problems by using either the collocation method or a weak formulation of the problem defined over a subregion of the region occupied by the body; the latter is usually called the Meshless Local Petrov–Galerkin (MLPG) method. For the two boundary-value problems studied, it is found that the weak formulation in which the basis for the first order derivatives of the trial solution are derived directly in the SSPH method (i.e., not obtained by differentiating the basis function for the trial solution) gives the least value of the L²-error norm in the displacements while the collocation method employing the strong formulation of the boundary-value problem has the largest value of the L²-error norm. The numerical solution using the weak formulation requires more CPU time than the solution with the strong formulation of the problem. We have also computed the L²-error norm of displacements by varying the number of particles, the number of Gauss points used to numerically evaluate domain integrals appearing in the weak formulation of the problem, the radius of the compact support of the kernel function used to generate the SSPH basis, the order of complete monomials employed for constructing the SSPH basis, and boundary conditions used at a point on a corner of the rectangular problem domain. It is recommended that for solving two-dimensional elasto-static problems, the MLPG formulation in which derivatives of the trial solution are found without differentiating the SSPH basis function be adopted.     

The intrinsic XFEM: a method for arbitrary discontinuities without additional unknowns

A new method for treating arbitrary discontinuities in a finite element (FE) context is presented. Unlike the standard extended FE method (XFEM), no additional unknowns are introduced at the nodes whose supports are crossed by discontinuities. The method constructs an approximation space consisting of mesh‐based, enriched moving least‐squares (MLS) functions near discontinuities and standard FE shape functions elsewhere. There is only one shape function per node, and these functions are able to represent known characteristics of the solution such as discontinuities, singularities, etc. The MLS method constructs shape functions based on an intrinsic basis by minimizing a weighted error functional. Thereby, weight functions are involved, and special mesh‐based weight functions are proposed in this work. The enrichment is achieved through the intrinsic basis. The method is illustrated for linear elastic examples involving strong and weak discontinuities, and matches optimal rates of convergence even for crack‐tip applications. Copyright © 2006 John Wiley & Sons, Ltd.     

Heat transfer analysis of composite slabs using meshless element Free Galerkin method

This paper deals with three dimensional heat transfer analysis of composite slabs using meshless element free Galerkin method. The element free Galerkin method (EFG) method utilizes moving least square (MLS) approximants to approximate the unknown function of temperature Tx). These approximants are constructed by using a weight function, a basis function and a set of coefficients that depends on position. Penalty and Lagrange multiplier techniques have been used to enforce the essential boundary conditions. MATLAB codes have been developed to obtain the EFG results. Two new basis functions namely trigonometric and polynomial have been proposed. A comparison has been made among the results obtained using existing (linear) and proposed (trigonometric and polynomial) basis functions for three dimensional heat transfer in composite slabs. The effect of penalty parameter on EFG results has also been discussed. The results obtained by EFG method are compared with those obtained by finite element method     

A collocation mesh-free method based on multiple basis functions

Two dominant shape functions are used to approximate scattered points in mesh-free methods, e.g. the interpolating radial basis function (RBF) and the approximating moving least squares (MLS). In the present paper, a new shape function is developed as a linear interpolating function of both MLS and RBF. This function inherits the properties of both MLS and RBF and is regularized by a control parameter μ, which takes different values in the domain [0,1]. Based on the proposed shape function, the collocation method is applied to solve initial and boundary value problems in one and two dimensions. The present method gives good results and achieves good convergence trends for different values of μ, compared with MLS and RBF individually, for a large number of nodes.     

A three‐dimensional element‐free Galerkin elastic and elastoplastic formulation

A small strain, three‐dimensional, elastic and elastoplastic Element‐Free Galerkin (EFG) formulation is developed. Singular weight functions are utilized in the Moving‐Least‐Squares (MLS) determination of shape functions and shape function derivatives allowing accurate, direct nodal imposition of essential boundary conditions. A variable domain of influence EFG method is introduced leading to increased efficiency in computing the MLS shape functions and their derivatives. The elastoplastic formulations are based on the consistent tangent operator approach and closely follow the incremental formulations for non‐linear analysis using finite elements. Several linear elastic and small strain elastoplastic numerical examples are presented to verify the accuracy of the numerical formulations. Copyright © 1999 John Wiley & Sons, Ltd.     

一种基于修正SSPH近似的无网格局部Petrov-Galerkin 法

首先采用奇异权函数对对称光滑粒子流体动力学(SSPH)近似进行了修正,使其构造的形函数近似满足d函数性质,方便无网格法中本质边界条件施加;然后应用修正的SSPH 近似法构造试函数,结合以Heaviside 函数为权函数的局部弱形式,提出了一种新的求解弹性静力问题的无网格局部Petrov-Galerkin 法;最后应用新的无网格法计算了一系列数值算例,结果表明:该方法具有良好的精度和收敛性。     

MLS‐based variable‐node elements compatible with quadratic interpolation. Part II: application for finite crack element

Two‐dimensional finite ‘crack’ elements for simulation of propagating cracks are developed using the moving least‐square (MLS) approximation. The mapping from the parental domain to the physical element domain is implicitly obtained from MLS approximation, with the shape functions and their derivatives calculated and saved only at the numerical integration points. The MLS‐based variable‐node elements are extended to construct the crack elements, which allow the discontinuity of crack faces and the crack‐tip singularity. The accuracy of the crack elements is checked by calculating the stress intensity factor under mode I loading. The crack elements turn out to be very efficient and accurate for simulating crack propagations, only with the minimal amount of element adjustment and node addition as the crack tip moves. Numerical results and comparison to the results from other works demonstrate the effectiveness and accuracy of the present scheme for the crack elements. Copyright © 2005 John Wiley & Sons, Ltd.     

Fatigue Crack Propagation in cylindrical bars

Crack propagation calculations were conducted to describe a developing surface crack in a solid round bar. The theoretical basis is a stress intensity factor solution in weighted average for the center and the surface points of a so called “almond-shaped” crack under Mode 1 loading. This solution was obtained by use of approximate weight functions. With these weight functions stress intensity factors for tension and bending were calculated. Predictions for the developing crack shape and the crack growth rate can be made utilizing a da/dN-ΔK relation. Experimental results for a 42 CrMo4 steel are in excellent agreement with these predictions.     

On elimination of shear locking in the element‐free Galerkin method

In this study, a method for completely eliminating the presence of transverse shear locking in the application of the element‐free Galerkin method (EFGM) to shear‐deformable beams and plates is presented. The matching approximation fields concept of Donning and Liu has shown that shear locking effects may be prevented if the approximate rotation fields are constructed with the innate ability to match the approximate slope (first derivative of displacement) fields and is adopted. Implementation of the matching fields concept requires the computation of the second derivative of the shape functions. Thus, the shape functions for displacement fields, and therefore the moving least‐squares (MLS) weight function, must be at least C¹ continuous. Additionally, the MLS weight functions must be chosen such that successive derivatives of the MLS shape function have the ability to exactly reproduce the functions from which they were derived. To satisfy these requirements, the quartic spline weight function possessing C² continuity is used in this study. To our knowledge, this work is the first attempt to address the root cause of shear locking phenomenon within the framework of the element‐free Galerkin method. Several numerical examples confirm that bending analyses of thick and thin beams and plates, based on the matching approximation fields concept, do not exhibit shear locking and provide a high degree of accuracy for both displacement and stress fields. Copyright © 2001 John Wiley & Sons, Ltd.     

A generalised solution for the crack surface displacement of mode I two-dimensional part-elliptical crack

A generalised approximate crack surface displacement solution for the two-dimensional part-elliptical mode I crack was developed. This solution includes the surface crack, corner crack and embedded crack, which is subjected to the arbitrary crack surface pressure. The crack surface displacement is derived from stress intensity factor solution and corresponding crack surface pressure distribution. Comparisons of the solution with accurate solutions showed that rather high accuracy has been achieved with the developed solution for various surface, embedded and corner crack problems. This solution can be used to derive three-dimensional weight functions as long as the stress intensity factor and the corresponding crack surface pressure are available for arbitrary mode I problems.     

Natural frequencies of functionally graded plates by a meshless method

We use the global collocation method, the first and the third-order shear deformation plate theories, the Mori–Tanaka technique to homogenize material properties, and approximate the trial solution with multiquadric radial basis functions to analyze free vibrations of functionally graded plates. Frequencies computed by the present method are found to agree well with those from the analytical solution of Vel and Batra, and the numerical solution of Qian et al. based on the meshless local Petrov–Galerkin formulation.     

Comparison of local weak and strong form meshless methods for 2-D diffusion equation

A comparison between weak form meshless local Petrov-Galerkin method (MLPG) and strong form meshless diffuse approximate method (DAM) is performed for the diffusion equation in two dimensions. The shape functions are in both methods obtained by moving least squares (MLS) approximation with the polynomial weight function of the fourth order on the local support domain with 13 closest nodes. The weak form test functions are similar to the MLS weight functions but defined over the square quadrature domain. Implicit timestepping is used. The methods are tested in terms of average and maximum error norms on uniform and non-uniform node arrangements on a square without and with a hole for a Dirichlet jump problem and involvement of Dirichlet and Neumann boundary conditions. The results are compared also to the results of the finite difference and finite element method. It has been found that both meshless methods provide a similar accuracy and the same convergence rate. The advantage of DAM is in simpler numerical implementation and lower computational cost.     

Weight functions and strip yield solution for two equal-length collinear cracks in an infinite sheet

The problem of two equal-length collinear cracks in an infinite sheet is treated using the weight function method. Exact weight functions for the inner and outer crack tips are derived based on the crack opening displacement solution for a reference load case. These weight functions are used to calculate stress intensity factors for different load cases, plastic zone sizes and crack tip opening displacements of the strip yield model. The approach is validated by the perfect agreement between the present strip yield model solutions and Collins and Cartwright’s analytical results based on the direct complex stress function formulation.     

Spiral weight for modeling cracks in meshless numerical methods

In the last decade several different approaches have been developed to study arbitrary static and dynamic cracks. Among these methods meshless techniques play an important role. These methods provide an accurate solution of a wide range of fracture mechanics problems while traditional methods such as finite element and boundary element have limitations. We wish to increase the accuracy of the meshless approximations without increasing the nodal density. This is done by an appropriate modification of the weight function near crack tips. Earlier attempts still had limitations that result in a lack of accuracy, especially in the case when a linear basis is used. In this work a new technique, the spiral weight, is introduced that minimizes the drawbacks of existing methods. Numerical examples show that the spiral weight method is more efficient than existing methods, when using a linear basis, for the solution of crack problems.     

Explicit form and efficient computation of MLS shape functions and their derivatives

This work presents a general and efficient way of computing both diffuse and full derivatives of shape functions for meshless methods based on moving least‐squares approximation (MLS) and interpolation. It is an extension of the recently introduced consistency approach based on Lagrange multipliers which provides a general framework for constrained MLS along with robust algorithms for the computation of shape functions and their diffuse derivatives. The particularity of the proposed algorithms is that they do not involve matrix inversion or linear system solving. The previous approach is limited to diffuse derivatives of the shape functions and not their full derivatives which are usually much more expensive to obtain. In the present paper we propose to efficiently compute the full derivatives by a new algorithm based on the formal differentiation of the previous one. In this way, we obtain a unified low‐cost consistent methodology for evaluating the shape functions and both their diffuse and full derivatives. In the second part of the paper we introduce explicit forms of MLS shape functions in 1D, 2D and 3D for an arbitrary number of nodes. These forms are especially useful for comparing finite element and MLS approximations. Finally we present a general architecture of an MLS program. Copyright © 2000 John Wiley & Sons, Ltd.     

Two-dimensional stress analysis of functionally graded solids using the MLPG method with radial basis functions

The meshless local Petrov–Galerkin (MLPG) method is used for analysing two-dimensional (2D) static and dynamic deformations of functionally graded materials (FGMs) with material response modelled as either linear elastic or as linear viscoelastic. The multiquadric radial basis function (RBF) is employed to approximate the trial solution. Results are computed with two different choices of test functions, namely a fourth-order spline weight function, and a Heaviside step function, each having a compact support. No background mesh is used to numerically evaluate integrals appearing in the weak formulation of the problem, thus the method is truly meshless. A benefit of using RBFs is that they possess the Kronecker delta property; thus it is easy to satisfy essential boundary conditions. For five problems, the computed results are found to match well with those either from their analytical solutions or numerical solutions of other researchers who employed different algorithms. For a dynamic problem, the Laplace-transform technique is utilised. The numerical examples illustrate that displacements and stress distributions in a structure made of an FGM differ considerably from those at the corresponding points in the same structure made of a homogeneous material. Thus, the inhomogeneity in material properties can be exploited to optimise stress distribution, minimise deflection and reduce the maximum stress.     

Analyzing interaction between coplanar square cracks using an efficient boundary element‐free method

This paper examines the interaction between coplanar square cracks by combining the moving least‐squares (MLS) approximation and the derived boundary integral equation (BIE). A new traction BIE involving only the Cauchy singular kernels is derived by applying integration by parts to the traditional boundary integral formulation. The new traction BIE can be directly applied to a crack surface and no displacement BIE is necessary because all crack boundary conditions (both upper and lower ones) are incorporated. A boundary element‐free method is then developed by combining the derived BIE and MLS approximation, in which the crack opening displacement is first expressed as the product of weight functions and the characteristic terms, and the unknown weight is approximated with the MLS approximation. The efficiency of the developed method is tested for isotropic and transversely isotropic media. The interaction between two and three coplanar square cracks in isotropic elastic body is numerically studied and the case of any number of coplanar square cracks is deduced and discussed. Copyright © 2012 John Wiley & Sons, Ltd.     

Simulation of second order singularly-perturbed boundary-value problems using improved meshless method

Abstract

In this study, numerical integration based on Block-Pulse functions and Chebyshev wavelets are employed for Element Free Galerkin approximation. Moving Least Squares (MLS) approach is used to construct shape functions with optimized weight functions and basis. The proposed techniques are implemented on singularly-perturbed boundary-value problems with two-point boundary conditions. Numerical results for two examples are presented in this article to show the pertinent features for the proposed technique. Comparison with existing techniques shows that our proposed method based on integration technique provides better approximation at reduced computational cost. Moreover the effect of perturbation parameter on solution of test problems has been studied.     

Analysis of cylindrical shell panels. A meshless solution

The Meshless Analog Equation Method, a purely meshless method, is applied to the static analysis of cylindrical shell panels. The method is based on the concept of the analog equation of Katsikadelis, which converts the three governing partial differential equations in terms of displacements into three substitute equations, two of second order and one fourth order, under fictitious sources. The fictitious sources are represented by series of radial basis functions of multiquadric type. Thus the substitute equations can be directly integrated. This integration allows the representation of the sought solution by new radial basis functions, which approximate accurately not only the displacements but also their derivatives involved in the governing equations. This permits a strong formulation of the problem. Thus, inserting the approximate solution in the differential equations and in the associated boundary conditions and collocating at a predefined set of mesh-free nodal points, a system of linear equations is obtained, which gives the expansion coefficients of radial basis functions series that represent the solution. The minimization of the total potential of the shell results in the optimal choice of the shape parameter of the radial basis functions. The method is illustrated by analyzing several shell panels. The studied examples demonstrate the efficiency and the accuracy of the presented method.