A crack-length control technique for phase field fracture in FFT homogenization
Pedro Aranda, Javier Segurado · 2024
Read A crack-length control technique for phase field fracture in FFT homogenization by Pedro Aranda, Javier Segurado (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-17445.txt.
Introduction
Pedro Aranda and Javier Segurado present a simulation method for cracking materials in a 2024 paper. To learn how microstructure shapes fracture, models must follow cracks at the tiny scale. Crack growth there is jumpy, and standard solvers stall or skip over the sudden snaps.
Their method controls the crack's length directly, forcing it to grow in steady steps. That allows a single-pass solver to trace even the snap-back parts of the curve. Versions were built for two solver families, and both matched older staggered results where growth was stable. Energy checks confirmed the crack released energy at a steady effective rate. Tests on composites and porous materials yielded toughness estimates, showing the method's reach.
Catalog record
| CREATOR | Pedro Aranda, Javier Segurado |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 93 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-17445
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A crack-length control technique for phase field fracture in FFT homogenization Pedro Aranda, Javier Segurado Abstract Modeling the propagation of cracks at the microscopic level is fundamental to understand the effect of the microstructure on the fracture process. Nevertheless, microscopic propagation is often unstable and when using phase field fracture poor convergence is found or, in the case of using staggered algorithms, leads to the presence of jumps in the evolution of the cracks. In this work, a novel method is proposed to perform micromechanical simulations with phase field fracture imposing monotonic increases of crack length and allowing the use of monolithic implementations, being able to resolve all the snap-backs during the unstable propagation phases. The method is derived for FFT based solvers in order to exploit its very high numerical performance in micromechanical problems, but an equivalent method is also developed for Finite Elements (FE) showing the equivalence of both implementations . It is shown that the stress-strain curves and the crack paths obtained using the crack control method are superposed in stable propagation regimes to those obtained using strain control with a staggered scheme. J-integral calculations confirm that during the propagation process in the crack control method, the energy release rate remains constant and equal to an effective fracture energy that has been determined as function of the discretization for FFT simulations. Finally, to show the potential of the method, the technique is applied to simulate crack propagation through the microstructure of composites and porous materials providing an estimation of the effective fracture toughness . 1 Introduction There is a strong interest on studying fracture of heterogeneous materials at the microscale in order to understand the effect of the microstructure on the fracture processes at microscale and macroscale. These microscopic simulations are fundamental for the estimation of effective properties in heterogeneous materials, which relate the microstructure with the macroscopic crack propagation [ 1 , 2 ] . Many studies can be found focused on microscopic fracture simulations in different heterogeneous materials including concrete [ 3 , 4 ] , biological materials [ 5 , 6 ] , composites [ 7 , 8 ] or polycrystals [ 9 , 10 ] . These works are based on the numerical resolution of a mechanical problem at the microscale using a representative volume elements (RVE) of the microstructure, and the simulations can be focused on the behavior of a single material point to extract some macroscopic fracture property or to be used as the local response in some concurrent multiscale framework [ 11 , 12 ] . Different approaches have been used to incorporate degradation at the microscale into micromechanical simulations. Some models rely on discontinuous techniques to model cracks, such as the introduction of cohesive elements [ 13 , 8 ] , X-FEM [ 6 ] or embedded strong discontinuities [ 11 , 14 ] . These techniques require the use of the Finite Element Method (FEM) to solve the mechanical problem, either using the connection between elements or special purpose elements, and their extension to other numerical framework becomes unnatural. Another option to model degradation in micromechanics are continuum damage (CDM) approaches, e.g. Gurson’s [ 15 , 16 , 17 ] or Lemaitrie’s [ 18 , 19 ] , which enable the representation of material damage using continuum fields. As these models do not consider explicitly the crack discontinuity, but continuous damage fields, they can be solved using alternative numerical methods. However, CDM models have limitations in accurately representing actual cracks and are prone to strong mesh dependency and artificial damage localization in bands of zero thickness [ 20 ] . To avoid this pathological discretization dependency, non-local versions of CDM models are generally used to regularize the damaged region by the introduction of some material characteristic length, as the so-called gradient damage models [ 21 , 22 ] . These approaches allow a variational formulation since they can be established by energy minimization considering damage and its gradient [ 23 ] , yet do not explicitly consider cracks but damaged regions. An alternative approach sharing properties with the two previous methods is the phase-field fracture model (PFF), which has become a breakthrough in the simulation of fracture. The method was initiated in the seminal work of Francfort and Marigo [ 24 ] and later improved by Ambrosio-Tortorelli [ 25 , 26 , 27 ] and Miehe [ 28 ] . This model relies on a variational approach and resembles gradient damage models in many aspects, as in the mathematical definition of energy functional and the use of a characteristic length [ 29 ] . Nevertheless, the physics behind PFF models is closer to discontinuous fracture approaches, since the model aims at simulating actual cracks and their evolution. For this purpose, the model represents cracks with a continuous field highly concentrated around the actual crack. Its evolution depends directly on the availability of elastic energy to propagate the crack, which depends on the constitutive model and the heterogeneity of the domain. This allows the calculation of multiple and topologically complex fracture patterns with low discretization dependence [ 30 , 31 ] . As gradient CDM, PFF can be implemented in any numerical solver for boundary value problems. In particular, Fast Fourier Transform (FFT) based solvers represent a great alternative to FEM for the implementation of PFF in micromechanical problems. FFT based solvers have the advantage of their high computational efficiency (the cost of the method with the number of degrees of freedom is proportional to n is nlog(n) ) and the natural incorporation of periodic boundary conditions, typical of micromechanical simulations. The method is mature nowadays and extensive reviews of its use in micromechanical problems can be found in the literature [ 32 , 33 ] . Indeed, the PFF model has already been implemented in FFT-based methods. Some examples of are the work of Chen et al. [ 34 ] or the works of Ernesti et al. [ 35 , 36 ] which study brittle fracture with fast and memory-efficient FFT-based implicit schemes. The application of periodic boundary conditions in problems including fracture or softening is controversial. This is because the crack evolution at the macroscale is a localized phenomena, so the stress-strain behavior of a periodic model considering microgeometries that are infinitely repetitive is not representative of the macroscopic response which includes the effect of damage areas and pristine ones. Nevertheless, periodic boundary conditions can be used when the response of a periodic cell is used as constitutive equation in a multiscale problem [ 11 , 12 ] or when the objective is estimating the effective fracture properties of a heterogeneous material considering clear scale separation, as proposed in the works [ 1 , 2 ] . In these cases, the use of periodic cells with periodic boundary conditions becomes an advantage to minimize cell boundary effects. Focusing on homogenizing the fracture properties, the concept of the effective toughness of an heterogeneous media has been boarded in many works, and presents differences to the homogenization of the elastic properties. This problem can be stated as a path minimization problem in which the calculation of the path with minimum fracture energy is decoupled of the mechanical problem associated