International Journal of Fracture最新文献

The fracture mode of AA 5083 Al–Mg alloy is predominantly intergranular. For this, a three-dimensional Crystal Plasticity Finite Element (CPFE) model is developed to investigate Grain Boundary (GB) fracture behavior under uniaxial tensile loading. Following parameter calibration, the simulated tensile strength demonstrates consistency with experimental data, with a high accuracy of 97%. Furthermore, by integrating CPFE model incorporating the dislocation slips with a GB model, a two-dimensional tri-junction crystal plasticity framework is established, and cohesive elements are implemented within the three grains to simulate GB fracture process under uniaxial tension. This tri-junction model shows a good convergence. The effect of grain orientation, location and GB thickness on GB fracture is investigated. The results reveal that grain orientation and location significantly influences GB fracture behavior. It is found that the Goss-Brass-P orientation combination demonstrates the optimal comprehensive mechanical properties, achieving a strain energy density of 7.09 MJ/m³. Furthermore, there exists a threshold effect about GB thickness on comprehensive property. The optimal performance is obtained when the cohesive that determines GB thickness is set to 2.

We formulate a multiscale model of adhesive layers undergoing impurity-dependent cohesive fracture. The model contemplates three scales: i) at the atomic scale, fracture is controlled by interatomic separation and the thermodynamics of separation depends on temperature and impurity concentration; ii) the mesoscale is characterized by the collective response of a large number of interatomic planes across the adhesive layer, resulting in a thickness-dependence strength; in addition, impurities are uptaken from the environment and diffuse through the adhesive layer; and iii) at the macroscale, we focus on lap joints under the action of static loads and aggressive environments. Within this scenario, we obtain closed form analytical solutions for: the dependence of the adhesive layer strength on thickness; the length of the edge cracks, if any, as a function of time; the lifetime of the joint; and the dependence of the strength of the joint on time of preexposure to the environment. Overall, the theory is found to capture well the experimentally observed trends. Finally, we discuss how the model can be characterized on the basis of atomistic calculations, which opens the way for the systematic exploration of new material specifications.

Once shear fractures begin to propagate in relatively thin weak layers underneath dry snow slabs, avalanche release is imminent. The evolution of the dynamic weak layer fracture likely determines the size or destructive potential once release takes place. In this paper, 54 values of mode II fracture speed collected from field tests are analyzed. In addition, values of speed collected with connection with avalanche release which involves both mode II and mode III shear fracture are analyzed. The Rayleigh wave speed is the normal maximum speed limit for a mode II fracture based on continuum formulations and the shear wave speed is the companion speed limit for mode III fracture. The 54 values of speed data from field tests are shown to have a median value of 12% of the predicted Rayleigh wave speed with a range between 2 and 56%. The data collected during avalanche release do not exceed 59% of the Rayleigh wave speed. The test data were collected from weak layers composed of highly porous crystal forms which are known to exhibit dynamic porosity loss during fracture propagation. Such events imply viscous dissipation and high friction which help to explain the low speeds. For snow slab avalanche release, which occurs for slope angles greater than (25^{0}), field observations show that the mixed mode (II, III) weak layer fracture terminates with mode I fracture through the slab following dynamic propagation. Analysis using the avalanche speed data with respect to terminal mode I fracture suggests that the shear stress drop behind the fracture process zone can contribute to produce the mode I fracture to promote avalanche release but dynamic contraction of the fracture process zone is a minor effect at best.

In this article, an isotropic eikonal gradient-enhanced damage modeling technique for studying time-dependent fracture mechanisms in brittle/quasi-brittle solids is presented. As is consistent with the classical gradient damage theory, the eikonal-type gradient damage modeling framework is mesh-independent and simple to implement, as it does not require any numerical tracking algorithm for discontinuities in the displacement field during crack propagation. With the evolving interaction known as the damage-based transient internal length scale, eikonal damage formulations could overcome the drawbacks of spurious damage diffusion and incorrect damage initiation by the traditional gradient theory. More realistic failure behavior could be predicted by the eikonal-based damage models. In the work, the derived governing equations of the momentum and a smooth Rankine strain-based eikonal gradient-enhanced evolution for the motion of the body and evolution of the nonlocal equivalent strain (time-dependent fracture driving term) are effectively solved using the standard finite element method (FEM). For temporal discretization, a robust explicit solver of the central difference method integrated with the row-sum technique for mass lumping is adopted. The advantages and abilities of the present dynamic model are shown by considering a number of time-dependent crack propagation problems in two-dimensional (2D) and three- dimensional (3D) solids.

The main objectives of this paper are to simulate 3-D fatigue crack propagation using a Generalized Finite Element Method (GFEM) and to validate the results against experimental data. This GFEM adopts a high-order p-hierarchical basis and explicit representations of crack surfaces. Both h-refinement around the fracture fronts and non-uniform p-enrichment of the analysis domain are used to control discretization errors. A systematic validation of this GFEM applied to 3-D fatigue crack propagation has not been reported in the literature. The Displacement Correlation Method (DCM) is used to extract stress intensity factors. The effect of material parameters adopted in the DCM on the crack growth rate and fracture shape is investigated. Three increasingly complex fatigue crack propagation problems are solved. The first involves mixed-mode loading in a modified compact tension specimen. The second one involves the transition from 2-D to 1-D crack surfaces and interactions between the crack front and the corners of the domain boundary. The final problem simulates the growth of a circumferential surface crack in a steel pipe subjected to fatigue bending with overloading, where interactions between the crack and the pipe’s inner surface result in the splitting of the crack front. Another contribution is an algorithm designed to manage cyclic load histories featuring variable loading ranges and ratios between minimum and maximum load magnitudes.

Unfolding failure is a sudden delamination occurring in highly curved fiber-reinforced polymer composite laminates when they are subjected to opening bending moments. Depending on the stacking sequences involved, unfolding failure includes intralaminar damage, interlaminar damage and their complex interactions. The use of these laminates in primary structural components of commercial aircrafts has intensified interest in understanding this failure mechanism. To tackle this, a simple phase field formulation to model the onset and propagation of transverse damage in highly curved laminates is presented. Firstly, the Abaqus implementation of the phase field to model intralaminar damage in composite laminates is established, exploiting the analogy between phase field and the heat transfer model to use Abaqus temperature as the phase field damage parameter. In addition, since residual stresses developed during manufacturing processes affect damage propagation, they are incorporated into the formulation as constant residual strains. Secondly, the proposed implementation is validated by performing tests on benchmark problems. Finally, a preliminary analysis of the unfolding failure test is provided to highlight the importance of including residual stresses in the transverse damage evolution.

The purpose of this work is to develop a computational tool for the analysis of ice fracture. A new numerical approach to simulate the crushing and spalling behavior of ice using the Material Point Method (MPM) is presented. Ice behavior under general triaxial load involves complex processes such as pressure softening, crack formation, fragmentation, and large deformations of crushed particles, which are challenging to capture with traditional simulation methods. The new approach leverages MPM’s ability to handle large strains and complex failure mechanisms without the mesh entanglement issues common in Finite Element Methods. The elliptical failure surface model for ice is used, accommodating both tensile and compressive failure criteria within a unified framework. This model allows for an accurate representation of ice behavior under a variety of loading conditions and supports the simulation of spalling and crushing phenomena. We validate our numerical model against experimental data obtained in large-scale indentation tests, demonstrating its efficacy in replicating the observed variability in oscillations, load ranges, and average load values. The results indicate that MPM is a promising tool for numerical investigation of ice failure, capturing key features such as fragmentation patterns and load response more effectively than conventional methods. The presented methodology offers significant potential for advancing the modeling of ice behavior, with implications for engineering applications in cold regions and ice-structure interactions.

Crack initiation in flattened disks under compression containing either a central or eccentric circular hole is investigated through the Finite Fracture Mechanics (FFM) approach. An implementation of the FFM criterion based on digital image correlation (DIC) full-field measurement is proposed. The coupling between FFM and DIC is provided through boundary conditions taken from the measured displacement fields. This approach offers a more accurate representation of the actual loading conditions compared to the use of idealized prescribed force or displacement in standard FFM implementations. Furthermore, by exploiting the value of the critical energy release rate obtained from the stable crack growth phase analysis, this method enables precise estimations of the inherent material strength and critical crack advance.