Two-dimensional stress intensity factor computations using the boundary element method

A multidomain boundary element formulation for the analysis of general two-dimensional plane strain/stress crack problems is presented. The numerical results were accurate and efficient. The analyses were performed using traction singular quater-point boundary elements on each side of the crack tip(s) with and without transition elements. Traction singular quarter-point boundary elements contain the correct √r displacement and 1/√r traction variations at the crack tip. Transition elements are appended to the traction singular elements to model the √r displacement variation. The 1/√r traction singularity is not represented with these elements. Current research studies for the crack propagation analysis of quasi-static and fatigue fracture problems are discussed.     

Time domain boundary element method for dynamic stress intensity factor computations

This paper presents a procedure for transient dynamic stress intensity factor computations using traction singular quarter-point boundary elements in combination with the direct time domain formulation of the Boundary Element Method. The stress intensity factors are computed directly from the traction nodal values at the crack tip. Several examples of finite cracks in finite domains under mode-I and mixed mode dynamic loading conditions are presented. The computed stress intensity factors are represented versus time and compared with those obtained by other authors using different methods. The agreement is very good. The results are reliable and little mesh dependent. These facts allow for the analysis of dynamic crack problems with simple boundary discretizations. The versatile procedure presented can be easily applied to problems with complex geometry which include one or several cracks.     

Comparison of methods for calculating stress intensity factors with quarter-point elements

The quarter-point quadrilateral element is employed with various methods for calculating the stress intensity factor in order to provide guidelines for a best method. These methods include displacement extrapolation, J-integral and Griffith's energy calculations, and the stiffness derivative technique. Three geometries are considered: a central crack, a single edge crack and double edge cracks in a rectangular sheet. For these cases, it is observed that the stiffness derivative method yields the most accurate results, whereas displacement extrapolation is the easiest method to implement and still yields reasonable accuracy.

---

Résumé On utilise les éléments en quadrilatère quart point dans diverses méthodes de calcul du facteur d'intensité de contrainte, afin de servir de guide pour le choix de la meilleure méthode. Il s'agit notamment des méthodes par extrapolation des déplacements, par calcul d'intégrale J ou d'énergie de Griffith, et par dérivée de la raideur. On considère trois géométries: une fissure centrale, une fissure de bord simple et deux fissures de bord dans une feuille rectangulaire. On observe pour ces trois cas que la méthode de la dérivée de la raideur conduit aux résultats les plus précis; par ailleurs, la méthode d'extrapolation des déplacements est la plus aisée à mettre en oeuvre et conduit néanmoins à une raisonnable exactitude.

     

Hypersingular quarter-point boundary elements for crack problems

The present paper deals with the study and effective implementation for Stress Intensity Factor computation of a mixed boundary element approach based on the standard displacement integral equation and the hypersingular traction integral equation. Expressions for the evaluation of the hypersingular integrals along general curved quadratic line elements are presented. The integration is carried out by transformation of the hypersingular integrals into regular integrals, which are evaluated by standard quadratures, and simple singular integrals, which are integrated analytically. The generality of the method allows for the modelling of curved cracks and the use of straight line quarter-point elements. The Stress Intensity Factors can be computed very accurately from the Crack Opening Displacement at collocation points extremely close to the crack tip. Several examples with different crack geometries are analyzed. The computed results show that the proposed approach for Stress Intensity Factors evaluation is simple, produces very accurate solutions and has little dependence on the size of the elements near the crack tip.     

The use of quarter-point crack-tip elements for T-stress determination in boundary element method analysis

A simple formula for obtaining the elastic T-stress in boundary element method fracture mechanics analysis is presented in this communication. This formula is obtained by comparing the variation of the displacements along the quarter-point crack-tip element with the classical field solution for the crack-tip. Its validity is tested with four example problems for a range of crack sizes and good agreement with solutions in the literature is generally obtained.     

Stress-intensity factor computation in three dimensions with quarter-point elements

A consistent method for computing stress-intensity factors from three-dimensional quarter-point element nodal displacements is presented. The method is generalized to permit functional evaluation of stress-intensity factors along the crack front. Embedded, surface, and corner crack problems are solved using the proposed technique. Results are compared to previous finite element and boundary element solutions. The comparison shows that use of the functional evaluation technique allows a dramatic decrease in problem size while still maintaining engineering accuracy. Next, a three-dimensional stress-intensity factor calibration of an unusual specimen configuration is presented. By taking advantage of the proposed technique, the calibration was performed with little difference in cost over the more usual two-dimensional approach. Moreover, the three-dimensional solution revealed intersting behaviour that would have been undetected by a two-dimensional solution. Finally, the results of a study on optimum size of the quarter-point element are presented. Surprisingly, Poisson ratio is shown to have marked effect on optimum element size.     

梯度复合材料应力强度因子计算的梯度扩展单元法

推导了一种适用于梯度复合材料断裂特性分析的梯度扩展单元,采用细观力学方法描述材料变化的物理属性,通过线性插值位移场给出了4节点梯度扩展元随空间位置变化的刚度矩阵,并建立了结构的连续梯度有限元模型.通过将梯度单元的计算结果与均匀单元以及已有文献结果进行对比,证明了梯度扩展有限元(XFEM)的优越性,并进一步讨论了材料参数对裂纹尖端应力强度因子(SIF)的影响规律.研究结果表明:随着网格密度的增加,梯度单元的计算结果能够迅速收敛于准确解,均匀单元的计算误差不会随着网格细化而消失,且随着裂纹长度和属性梯度的增大而增大;属性梯度和涂层基体厚度比的增大导致涂覆型梯度材料的SIF增大;裂纹长度的增加和连接层基体厚度比的减小均导致连接型梯度材料的SIF增大.     

梯度复合材料应力强度因子计算的梯度扩展单元法

推导了一种适用于梯度复合材料断裂特性分析的梯度扩展单元, 采用细观力学方法描述材料变化的物理属性, 通过线性插值位移场给出了4节点梯度扩展元随空间位置变化的刚度矩阵, 并建立了结构的连续梯度有限元模型。通过将梯度单元的计算结果与均匀单元以及已有文献结果进行对比, 证明了梯度扩展有限元(XFEM)的优越性, 并进一步讨论了材料参数对裂纹尖端应力强度因子(SIF)的影响规律。研究结果表明: 随着网格密度的增加, 梯度单元的计算结果能够迅速收敛于准确解, 均匀单元的计算误差不会随着网格细化而消失, 且随着裂纹长度和属性梯度的增大而增大; 属性梯度和涂层基体厚度比的增大导致涂覆型梯度材料的SIF增大; 裂纹长度的增加和连接层基体厚度比的减小均导致连接型梯度材料的SIF增大。     

On quarter-point three-dimensional finite elements in linear elastic fracture mechanics

Use of the finite element method for calculating stress intensity factors of two-dimensional cracked bodies has become commonplace. In this study, the more difficult task of applying finite elements to three-dimensional cracked bodies is investigated. Since linear elastic material is considered, square root singular stresses exist along the edge of an embedded crack. To deal with this numerical difficulty, twenty noded, isoparametric, serendipity, quarter-point, singular, solid elements are employed. Examination of these elements is carried out in order to determine the extent of the singular behavior.In addition, the stiffness derivative technique is explored, together with quarter-point elements, to determine an accurate procedure for computing stress intensity factors in three-dimensions. The problem of chosing a proper virtual crack extension is addressed. To this end, the disturbance in the square root singular stresses is examined and compared with a similar disturbance which occurs in two-dimensions. As a numerical example, a pennyshaped crack in a finite height cylinder is considered with various meshes. It is found that stress intensity factors can be calculated to an accuracy within 1 percent when quarter-point cylindrical elements are employed with the stiffness derivative technique such that the crack extension is one in which one corner node is not moved, the other corner node is moved a small distance, and the midside node is moved one-half that distance. This crack extension is analogous to that of a straight crack advance for a brick element. Both of these crack advances disturb the square root singular stresses in a manner similar to that which occurs with the two-dimensional eight noded element in which the crack has been advanced a small distance.     

The performance of isoparametric finite elements in stress intensity factor determination

Analysis of a compact compression specimen used for fracture toughness evaluation of cementitious materials is carried out by the finite element method using isoparametric elements. Both triangular and rectangular elements were used with those surrounding the crack tip being of the quarter point type. Solutions were obtained for different mesh subdivisions and convergenece curves for the stress intensity factor were obtained by several methods based on extrapolation and energy techniques. It is found that monotonic convergence was obtained for all cases considered. Employing uniformly graded rectangular element representations converged solutions for the stress intensity factor (assuming a 1 percent convergence criterion) were obtained by the energy methods using a total of 720 degrees of freedom for solving half the structure.Tests on modified 100 mm cubes with symmetrical notches were conducted to determine the fracture toughness. The fracture toughness was calculated from the stress intensity factor and the maximum load obtained from the tests which were conducted in a stiff Instron testing machine. The fracture toughness is found to be independent of the size of the notch.     

A warning against misuse of quarter-point elements

Quarter-point elements are used very frequently for fracture mechanics computations, because the quarter-point technique yields the required singular interpolation without any modification to existing software. This advantage is particularly significant for three-dimensional stress intensity factor computations because of the difficulty of implementing other techniques. However, in practical 3-D applications, the crack front is usually curved, and this note proves that a crack front distortion leads to a negative Jacobian in the region surrounding the crack front. The numerical difficulties to be expected depend on the aspect ratio of the elements.     

The stress intensity factor for a double edge cracked plate by boundary collocation method

A set of complex functions for the double edge cracked plate is proposed. The boundary collocation method is used for estimating the stress intensity factors and the results obtained by this method compare very favorably with existing solutions for many cases.

---

Résumé On propose un jeu de fonctions complexes pour représenter une plaque fissurée sur ses deux bords. On utilise la méthode de collocation des limites pour estimer les facteurs d'intensité de contraintes. Dans de nombreux cas, les résultats obtenus par cette méthode supportent très favorablement la comparaison avec les solutions existantes.

     

On the computation of two-dimensional stress intensity factors using the boundary element method

General two-dimensional linear elastic fracture problems are investigated using the boundary element method. The √r displacement and 1/√r traction behaviour near a crack tip are incorporated in special crack elements. Stress intensity factors of both modes I and II are obtained directly from crack-tip nodal values for a variety of crack problems, including straight and curved cracks in finite and infinite bodies. A multidomain approach is adopted to treat cracks in an infinite body. The body is subdivided into two regions: an infinite part with a finite hole and a finite inclusion. Numerical results, compared with exact solution whenever possible, are accurate even with a coarse discretization.     

Calculation of the stress intensity factor for arbitrary finite cracked body by using the boundary weight function method

The state of thermal stresses for a periodic two-layered elastic space weakened by an interface thermally insulated Griffith crack and loaded by concentrated line heat sources is investigated. The analysis is performed within the framework of the homogenized model with microlocal parameters. The stress intensity factors at the crack tips are determined. Two examples of the application of the results obtained are detailed.     

The optimum size of quarter-point crack tip elements

Ingraffea and Manu¹ and Lynn and Ingraffea² have shown that the size of the quarter-point elements can affect the computed elastic stress intensity factor. The nature of the effect is such that, all other details remaining constant, there is a particular crack tip element size which minimizes the error in the computed stress intensity factor. Here, size of element means the radial edge length. The reasons for this size dependence are discussed below. It will be seen that the discussion is in terms of the need to simultaneously represent the singular and finite stress terms in a given problem. The discussion has relevance to other formulations of crack tip elements.     

Calculation of stress intensity factor from stress concentration factor

Using the general formulas of stress concentration factor, methods for calculating stress intensity factor are mentioned.These methods make use of the several known values of stress concentration and radius of curvature at the point of stress concentration to form expression of stress concentration factor. Values of stress concentration from handbooks or experiments and others can be used.This paper deals with plane elastic, longitudinal shearing and thin plate bending problem.     

On the uniqueness of the stress intensity factor — crack velocity relationship

Experimental results due to several different investigators on the K _I – a relationship are reviewed and the apparent differences in results leading to questions regarding the uniqueness of this relationship are discussed. The influence of the errors due to the three dimensional state of stress at the crack tip, the effects of non-singular stresses, velocity, transient loading and velocity measurement is presented. These errors have obscured resolution of the uniqueness question and an experiment is described to resolve the issue.

---

Résumé On passe en revue les résultats expérimentaux obtenus par différents chercheurs sur la relation K _I – a et l'on discute des différences apparentes dans les résultats qui conduisent à des questions en ce qui regarde l'unicité de cette relation. On présente l'influence des erreurs dues à l'état tridimensionnel des tensions à l'extrémité de la fissure, les effets des contraintes non singulières de la vitesse de la mise en charge transitoire et de la mesure de la vitesse. Ces erreurs ne contribuent pas à éclaircir la question de l'unicité et, en vue de la résoudre, on décrit une expérience possible.