A combined experimental-numerical investigation of crack growth in a carbon–carbon composite

A combined experimental–numerical investigation of crack growth in a carbon–carbon composite is reported. In this material, both matrix fracture and fibre bridging contribute significantly to toughness. Crack growth experiments were performed using side‐notched DCB specimens with doublers. A digital image correlation method was used to measure displacements fields on the specimen surfaces, crack extension and crack opening profiles. An effective cohesive zone law was determined from the experimental data. The effective cohesive zone law is subsequently separated into the individual contributions from matrix cracking and fibre bridging. Numerical simulation of crack growth based on this cohesive zone law and experimental data are in good agreement. Special focus of the numerical study is on the investigation of the discontinuous nature of crack growth.     

Incremental‐secant modulus iteration scheme and stress recovery for simulating cracking process in quasi‐brittle materials using XFEM

In this paper, an incremental‐secant modulus iteration scheme using the extended/generalized finite element method (XFEM) is proposed for the simulation of cracking process in quasi‐brittle materials described by cohesive crack models whose softening law is composed of linear segments. The leading term of the displacement asymptotic field at the tip of a cohesive crack (which ensures a displacement discontinuity normal to the cohesive crack face) is used as the enrichment function in the XFEM. The opening component of the same field is also used as the initial guess opening profile of a newly extended cohesive segment in the simulation of cohesive crack propagation. A statically admissible stress recovery (SAR) technique is extended to cohesive cracks with special treatment of non‐homogeneous boundary tractions. The application of locally normalized co‐ordinates to eliminate possible ill‐conditioning of SAR, and the influence of different weight functions on SAR are also studied. Several mode I cracking problems in quasi‐brittle materials with linear and bilinear softening laws are analysed to demonstrate the usefulness of the proposed scheme, as well as the characteristics of global responses and local fields obtained numerically by the XFEM. Copyright © 2006 John Wiley & Sons, Ltd.     

Experimental validation of large-scale simulations of dynamic fracture along weak planes

A well-controlled and minimal experimental scheme for dynamic fracture along weak planes is specifically designed for the validation of large-scale simulations using cohesive finite elements. The role of the experiments in the integrated approach is two-fold. On the one hand, careful measurements provide accurate boundary conditions and material parameters for a complete setup of the simulations without free parameters. On the other hand, quantitative performance metrics are provided by the experiments, which are compared a posteriori with the results of the simulations. A modified Hopkinson bar setup in association with notch-face loading is used to obtain controlled loading of the fracture specimens. An inverse problem of cohesive zone modeling is performed to obtain accurate mode-I cohesive zone laws from experimentally measured deformation fields. The speckle interferometry technique is employed to obtain the experimentally measured deformation field. Dynamic photoelasticity in conjunction with high-speed photography is used to capture experimental records of crack propagation. The comparison shows that both the experiments and the numerical simulations result in very similar crack initiation times and produce crack tip velocities which differ by less than 6%. The results also confirm that the detailed shape of the non-linear cohesive zone law has no significant influence on the numerical results.     

Cohesive Crack with Rate-Dependent Opening and Viscoelasticity: I. Mathematical Model and Scaling

The time dependence of fracture has two sources: (1) the viscoelasticity of material behavior in the bulk of the structure, and (2) the rate process of the breakage of bonds in the fracture process zone which causes the softening law for the crack opening to be rate-dependent. The objective of this study is to clarify the differences between these two influences and their role in the size effect on the nominal strength of stucture. Previously developed theories of time-dependent cohesive crack growth in a viscoelastic material with or without aging are extended to a general compliance formulation of the cohesive crack model applicable to structures such as concrete structures, in which the fracture process zone (cohesive zone) is large, i.e., cannot be neglected in comparison to the structure dimensions. To deal with a large process zone interacting with the structure boundaries, a boundary integral formulation of the cohesive crack model in terms of the compliance functions for loads applied anywhere on the crack surfaces is introduced. Since an unopened cohesive crack (crack of zero width) transmits stresses and is equivalent to no crack at all, it is assumed that at the outset there exists such a crack, extending along the entire future crack path (which must be known). Thus it is unnecessary to deal mathematically with a moving crack tip, which keeps the formulation simple because the compliance functions for the surface points of such an imagined preexisting unopened crack do not change as the actual front of the opened part of the cohesive crack advances. First the compliance formulation of the cohesive crack model is generalized for aging viscoelastic material behavior, using the elastic-viscoelastic analog (correspondence principle). The formulation is then enriched by a rate-dependent softening law based on the activation energy theory for the rate process of bond ruptures on the atomic level, which was recently proposed and validated for concrete but is also applicable to polymers, rocks and ceramics, and can be applied to ice if the nonlinear creep of ice is approximated by linear viscoelasticity. Some implications for the characteristic length, scaling and size effect are also discussed. The problems of numerical algorithm, size effect, roles of the different sources of time dependence and rate effect, and experimental verification are left for a subsequent companion paper. This revised version was published online in July 2006 with corrections to the Cover Date.     

Quasi‐static crack propagation in plane and plate structures using set‐valued traction‐separation laws

We introduce a numerical technique to model set‐valued traction‐separation laws in plate bending and also plane crack propagation problems. By using of recent developments in thin (Kirchhoff–Love) shell models and the extended finite element method, a complete and accurate algorithm for the cohesive law is presented and is used to determine the crack path. The cohesive law includes softening and unloading to origin, adhesion and contact. Pure debonding and contact are obtained as particular (degenerate) cases. A smooth root‐finding algorithm (based on the trust‐region method) is adopted. A step‐driven algorithm is described with a smoothed law which can be made arbitrarily close to the exact non‐smooth law. In the examples shown the results were found to be step‐size insensitive and accurate. In addition, the method provides the crack advance law, extracted from the cohesive law and the absence of stress singularity at the tip. Copyright © 2007 John Wiley & Sons, Ltd.     

坝基界面在非线性水压力驱动下的非线性断裂过程模拟

基于多边形比例边界有限元法和粘聚裂缝模型提出了混凝土坝坝基界面在随缝宽非线性变化的水压力驱动下的非线性断裂数值模型。混凝土和基岩采用多边形比例边界单元模拟,界面裂缝的断裂过程区采用粘性界面单元模拟。因为界面裂缝总是处于复合断裂模态,故同时引入了法向和切向的界面单元,且考虑了裂纹面作用有法向和切向任意荷载时的应力强度因子求解。以裂尖为原点,裂尖附近的位移场和应力场在径向上解析求解,在环向具有有限元精度。因此无需在裂尖附近加密网格或采用富集技术即可求得高精度的解。对于界面断裂,可模拟出与两种材料差异性相关的非1/2奇异性。断裂过程区的水压力随缝面宽度变化,采用指数函数的形式进行表征,通过参数调整可实现不同分布的水压力的模拟。水压力与粘聚力考虑为与裂缝宽度相关的组合函数,便于非线性迭代的实现。结合多边形网格生成和重剖分技术,可方便地模拟界面裂缝在水力驱动下的扩展过程。算例研究表明了该文模型的有效性,从中也可看出考虑缝内水压及其具体分布形式对研究坝的稳定性具有重要影响。     

Pore pressure cohesive zone modeling of hydraulic fracture in quasi-brittle rocks

Hydraulic fracturing technology has been widely applied in the petroleum industry for both waste injection and unconventional gas production wells. The prevailing analytical solutions for hydraulic fracture mainly depend on linear elastic fracture mechanics. These methods can give reasonable prediction for hard rock, but are ineffective in predicting hydraulic fractures in quasi-brittle materials, such as ductile shale and sandstone. One of the reasons is that the fracture process zone ahead of the crack tip and the softening effect should not be neglected for quasi-brittle materials. In the current work, a set of chevron-notch three point bending tests were performed on sandstone samples from an oil field in Ordos Basin, Shaanxi province, China, and the results were compared with the cohesive zone method based on finite element analysis. The numerical results fit the experimental data well and it shows that the cohesive zone model and the Traction-Separation law used in the model are effective in modeling fracture nucleation and propagation in sandstone without considering the porous effect. A 3D pore pressure cohesive zone model was developed to predict nucleation and propagation of a penny-shaped fluid-driven fracture. The predictions were compared with the analytical asymptotic solutions and a field minifrac test from the literature; it shows that the proposed method can not only predict the length and aperture of hydraulic fracture well, but also predict the bottomhole pressure with reasonable accuracy. Based on analytical asymptotic and computational solutions, parametric studies were conducted to investigate the effects of different parameters on the fracture aperture and fracture length, fracture process zone and bottomhole pressure.     

Eigenvalue analysis of size effect for cohesive crack model

The paper analyses the effect of structure size on the nominal strength of the structure that is implied by the cohesive (or fictitious) crack model proposed for concrete by Hillerborg et al. A new method to calculate the maximum load of geometrically similar structures of different sizes without calculating the entire load-deflection curves is presented. The problem is reduced to a matrix eigenvalue problem, in which the structure size for which the maximum load occurs at the given (relative) length of the cohesive crack is obtained as the smallest eigenvalue. Subsequently, the maximum load, nominal strength and load-point displacement are calculated from the matrix equilibrium equation. The nonlinearity of the softening stress-displacement law is handled by iteration. For a linear softening law, the eigenvalue problem is linear and independent of the matrix equilibrium equation, and the peak load can then be obtained without solving the equilibrium equation. The effect of the shape of the softening law is studied, and it is found that the size effect curve is not very sensitive to it. The generalized size effect law proposed earlier by Baant, which describes a transition between the horizontal and inclined asymptotes of strength theory and linear elastic fracture mechanics, is found to fit the numerical results very well. Finally some implications for the determination of fracture energy from the size effect tests are discussed. The results are of interest for quasibrittle materials such as concrete, rocks, sea ice and modern tough ceramics.     

Cohesive-zone-model formulation and implementation using the symmetric Galerkin boundary element method for homogeneous solids

A new symmetric boundary integral formulation for cohesive cracks growing in the interior of homogeneous linear elastic isotropic media with a known crack path is developed and implemented in a numerical code. A crack path can be known due to some symmetry implications or the presence of a weak or bonded surface between two solids. The use of a two-dimensional exponential cohesive law and of a special technique for its inclusion in the symmetric Galerkin boundary element method allows us to develop a simple and efficient formulation and implementation of a cohesive zone model. This formulation is dependent on only one variable in the cohesive zone (relative displacement). The corresponding constitutive cohesive equations present a softening branch which induces to the problem a potential instability. The development and implementation of a suitable solution algorithm capable of following the growth of the cohesive zone and subsequent crack growth becomes an important issue. An arc-length control combined with a Newton–Raphson algorithm for iterative solution of nonlinear equations is developed. The boundary element method is very attractive for modeling cohesive crack problems as all nonlinearities are located along the boundaries (including the crack boundaries) of linear elastic domains. A Galerkin approximation scheme, applied to a suitable symmetric integral formulation, ensures an easy treatment of cracks in homogeneous media and excellent convergence behavior of the numerical solution. Numerical results for the wedge split and mixed-mode flexure tests are presented.     

Nonlinear fracture of cohesive materials

The cohesive crack is a useful model for describing a wide range of physical situations from polymers and ceramics to fiber and particle composite materials. When the cohesive zone length is of the order of the specimen size, the influence method—based on finite elements—may be used to solve the fracture problem. Here a brief outline of an enhanced algorithm for this method is given. For very large specimen sizes, an asymptotic analysis developed by the authors allows an accurate treatment of the cohesive zone and provides a powerful framework for theoretical developments. Some recent results for the zeroth order and first order asymptotic approaches are discussed, particularly the effective crack concept and the maximum load size effect. These methods are used to analyze the effect of the size and of the shape of the softening curve on the value at the peak load of several variables for three point bent notched beams. The results show, among other things, that for intermediate and very large sizes the size effect curves depend strongly on the shape of the softening curve, and that only the simultaneous use of asymptotic and influence methods may give an adequate estimate of the size effect in the intermediate range.     

On the Fracture Models Determined by the Continuum-strong Discontinuity Approach

The paper focuses on the Continuum Strong Discontinuity Approach (CSDA) to fracture mechanics, and the traction-separation cohesive laws induced from continuum dissipative models as their projections onto the failure interface. They are compared with the cohesive laws commonly used for the fracture simulation in quasi-brittle materials, typically concrete. Emphasis is placed in the analysis of the mechanical stress-strain states induced by the CSDA into the fracture process zone: first when the damage mechanism is initiated and, after, when the cohesive model determines the crack response. The influence of the material parameters, particularly the fracture energy and the initial continuum softening modulus, in the obtained phenomenological responses is also analyzed. Representative numerical solutions of fracture problems are finally presented.     

Asymptotic analysis of a cohesive crack: 2. Influence of the softening curve

This paper presents a numerical method well suited to solve the integral equation governing the asymptotic behavior of a cohesive crack, and uses it to analyze the influence of the softening curve on the cracking response of large specimens. The analysis is performed with two main objectives in mind: (1) providing criteria to determine when a simplified linear elastic fracture mechanics (LEFM) approach can be applied, and (2) providing possible procedures of extracting information on the softening behavior from experimental data. The main conclusion is that the effective crack extension prior to peak is nearly determined by the length of the softening curve (the critical crack opening) and so is the deviation from LEFM. Furthermore, a simplified curve approach is proposed as an approximate alternative to solving the governing integral equation.     

Determination of mixed mode cohesive laws

A novel approach is proposed for the determination of mixed mode cohesive laws for large scale crack bridging problems. The approach is based on a plane, two-dimensional analysis utilizing the J integral applied a double cantilever beam specimens loaded with uneven bending moments. The normal and shear stresses of the cohesive laws are obtained from consecutive values of the fracture resistance, the normal and tangential displacements of the end of the cohesive zone. The data analysis involves fitting and determination of partial differentials. This is done by a numerical method using Chebyshev polynomials. The accuracy of the numerical procedure is investigated by the use of synthetic data. It is found that both the shape and peak stress of the cohesive law can be determined with high accuracy, providing that the data possess low noise and a sufficiently high number of datasets are used. The investigation leads to some practical guidelines for experimental use of the proposed approach.     

A new nonlocal cohesive stress law and its applicable range

In this paper, we attempt to provide a new analytical method to determine the cohesive law in the framework of nonlocal continuum mechanics. Firstly, the equivalence between the cohesive stress and the surface-induced traction (nonlocal surface residual) is established on the basis of the nonlocal stress boundary condition. Then a new cohesive stress law is derived logically from the perspective of rational mechanics, which characterizes the dependence of the cohesive stress on the crack opening displacement (COD) within the cohesive zone. Finally, we apply this new cohesive crack model to two fracture examples with different cohesive zone sizes, and investigate the stress field ahead of the crack tip in detail. The results show that the stress singularity at the crack tip is removed, and the maximum stress occurs within the cohesive zone away from the crack tip. Moreover, the stress in the large-scale cohesive zone drops rapidly to a constant approaching zero, exhibiting a stronger softening behavior.     

A procedure for superposing linear cohesive laws to represent multiple damage mechanisms in the fracture of composites

The relationships between a resistance curve (R-curve), the corresponding fracture process zone length, the shape of the traction-displacement softening law, and the propagation of fracture are examined in the context of the through-the-thickness fracture of composite laminates. A procedure for superposing linear cohesive laws to approximate an experimentally-determined R-curve is proposed. Simple equations are developed for determining the separation of the critical energy release rates and the strengths that define the independent contributions of each linear softening law in the superposition. The proposed procedure is demonstrated for the longitudinal fracture of a fiber-reinforced polymer-matrix composite. It is shown that the R-curve measured with a Compact Tension Specimen test cannot be predicted using a linear softening law, but can be reproduced by superposing two linear softening laws.     

The effects of inter-surface cohesive tractions on linear and penny-shaped cracks

A mesoscopic fracture model of equilibrium slit cracks in brittle solids, including inter-surface cohesive tractions acting near the crack tip, is analyzed and the effects of the cohesive tractions on the in-plane stress fields, crack-opening displacement profiles, and crack driving forces examined quantitatively for linear and penny-shaped cracks. The (numerical) analysis method is described in detail, along with results for four different cohesive forces. The assumed distribution of cohesive tractions were found to suppress the in-plane stress field adjacent to cracks in a homogeneous, isotropic medium when uniformly loaded in mode-I, and the suppression was a function of crack length. The crack-opening displacement profile was also perturbed and a new regime identified between the near-field Barenblatt zone and the far-field continuum zone. The extent of this `cohesive zone' was quantified by use of an interpolating function fit to the calculated profiles and found to be independent of crack size for a given cohesive tractions distribution. Furthermore, the crack-opening displacement at the edge of the cohesive zone was found to be independent of crack size, implying that despite significant perturbations to the stress field, the crack driving force at unstable equilibrium remains unchanged with crack size.     

A single-domain dual-boundary-element formulation incorporating a cohesive zone model for elastostatic cracks

A cracked elastostatic structure is artificially divided into subdomains of simpler topology such that the well-developed classic dual integral equations can be applied appropriately to each domain. Applying the continuity and equilibrium conditions along artificial boundaries and properties of the integral kernels a single-domain dual-boundary-integral equation formulation is derived for a cracked elastic structure. A cohesive zone model is used to model the crack tip processes and is coupled with the single-domain dual-boundary-integral equation formulation; the resulting nonlinear equations are solved using the iterative method of successive-over-relaxation. The constitutive law used for a crack includes three parts: a law relating cohesive force to crack displacement difference when a crack is opening, a characterization of tangential interaction between crack surfaces when the crack surfaces are in contact, and a maximum principal stress criterion of crack advance. Incorporation of local unloading effect of the cohesive zone material has enabled a simulation of fracture with initial damage, partial development of the failure process zone at structural instability and multiple crack interaction. Some of the features of the method are demonstrated by considering three examples. The first problem is a single-edge-cracked specimen that exhibits a snap-back instability. The second example is the development of wing cracks from an angled crack under compression. The last example demonstrates the capability to consider mixed-mode crack growth and interaction of cracks. Thus, the problem of crack growth has been reduced to the determination of the cohesive model for the fracture process. This revised version was published online in July 2006 with corrections to the Cover Date.     

Cohesive Crack Model with Rate-Dependent Opening and Viscoelasticity: II. Numerical Algorithm, Behavior and Size Effect

In the preceding companion paper (Bažant and Li, 1995), the solution of an aging viscoelastic law was structure containing a cohesive crack with a rate-dependent stress-displacement softening law was reduced to a system of one-dimensional integro-differential equations involving compliance functions for points on the crack faces and the load point. An effective numerical algorithm for solving these equations, which dramatically reduces the computer time compared to the general two-dimensional finite element solution, is presented. The behavior of the model for various loading conditions is studied. It is shown that the model can closely reproduce the available experimental data from fracture tests with different loading rates spanning several orders of magnitude, and tests with sudden changes of the loading rate. Influence of the loading rate on the size effect and brittleness is also analyzed and is shown to agree with experiments. This revised version was published online in July 2006 with corrections to the Cover Date.     

Determination of cohesive laws of composite bonded joints under mode II loading

In this work, mode II cohesive laws of carbon–epoxy composite bonded joints were obtained using the direct method applied to the end notched flexure (ENF) test. The direct method is based on the differentiation of the relation between the evolution of the fracture energy (J_II) and the crack tip opening displacement in mode II (CTOD_II) during the test. A data reduction scheme based on equivalent crack length concept was used to obtain the evolution of the fracture energy during the test. The method allows overcoming problems related to identification of crack tip in mode II tests and the presence of a non-negligible fracture process zone (FPZ), which both difficult the right estimate of J_II. The digital image correlation technique (DIC) was used to monitor the CTOD_II, which was synchronized with the load–displacement data. A trapezoidal cohesive law was fitted to the experimental one in order to perform numerical simulations using finite element analysis. The main goal was to validate all the procedure used to get the cohesive laws. The good agreement obtained between the numerical and experimental load-CTOD_II curves and between the cohesive laws demonstrates the adequacy of the proposed procedure concerning the evaluation of the composite bonded joints cohesive laws under mode II loading.     

Smeared-tip superposition method for cohesive fracture with rate effect and creep

A recent formulation of the smeared-tip superposition method presented by Baant [1], which itself was a generalization and modification of an integral equation formulation with an asymptotic series solution derived by Planas and Elices [2], is further improved, generalized and adapted to an efficient finite difference solution scheme. A crack with bridging stresses is modeled as a superposition of infinitely many LEFM cracks with continuously distributed (smeared) tips having infinitely small intensity factors. Knowledge of the stress intensity factor as a function of the location of the crack tip along the crack path is all that is needed to obtain the load-displacement relation. The solution is reduced to a singular integral equation for a function describing the components of applied load associated with crack tips at various locations. The integral equation is complemented by an arbitrary relation between the bridging stress and the crack opening displacement, which can be rate-independent or rate-dependent. Furthermore, using the creep operator method, the equation is extended to aging linearly viscoelastic behavior in the bulk of the specimen. The previously presented finite difference solution is improved and generalized in a form that leads to a system of nonlinear algebraic equations, which can be solved by an optimization method. Application of the smeared-tip method to the analysis of recent measurements of the size effect in three-point-bend fracture specimens of different sizes is presented and a crack opening law that yields the main qualitative characteristics of the test results, particularly an increase of brittleness with a decreasing loading rate, is presented.