Adaptive finite element convergence analysis of AT1 phase-field model for quasi-static fracture in strain-limiting solids Dedicated to the cherished memory of Professor K. R. Rajagopal (Professor of Mechanical Engineering, Texas A&M University, College Station, Texas-USA). His foundational work and

Ram Manohar, S. M. Mallikarjunaiah · 2025

Read Adaptive finite element convergence analysis of AT1 phase-field model for quasi-static fracture in strain-limiting solids Dedicated to the cherished memory of Professor K. R. Rajagopal (Professor of Mechanical Engineering, Texas A&M University, College Station, Texas-USA). His foundational work and by Ram Manohar, S. M. Mallikarjunaiah (2025) 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-2507-00376.txt.

Introduction

Manohar and Mallikarjunaiah studied adaptive simulations of cracking in strain-limiting solids in twenty twenty-five. Phase-field models smear a crack over a small zone, using a damage variable. This avoids direct tracking of crack surfaces. The AT1 variant is popular for quasi-static brittle fracture. But proving that adaptive mesh refinement converges is hard.

The authors study an AT1 model inside elasticity theories with algebraically nonlinear stress-strain laws. They propose two refinement algorithms driven by local error indicators. Minimizer sequences from the first converge to a set tolerance. Sequences from the second are proven inherently convergent, needing no stopping rule. Simulations of an edge crack under anti-plane shear loading checked bulk, surface, and total energies. They showed both algorithms outperform standard approaches.

Catalog record

CREATORRam Manohar, S. M. Mallikarjunaiah
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE112 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-00376

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Adaptive finite element convergence analysis of AT1 phase-field model for quasi-static fracture in strain-limiting solids Dedicated to the cherished memory of Professor K. R. Rajagopal (Professor of Mechanical Engineering, Texas A&M University, College Station, Texas-USA). His foundational work and invaluable mentorship left an indelible mark on this research and its contributing authors. Ram Manohar, S. M. Mallikarjunaiah Abstract This research rigorously investigates the convergence of adaptive finite element methods for regularized variational models of quasi-static brittle fracture in elastic solids. We specifically examine a novel Ambrosio-Tortorelli (AT1) phase-field model within the framework of elasticity theories, particularly for material models characterized by an algebraically nonlinear stress-strain relationship. Two distinct and novel adaptive mesh refinement algorithms, underpinned by robust local error indicators, were introduced to efficiently solve the underlying nonlinear energy minimization problem. A detailed convergence analysis was conducted on the sequences of minimizers produced by these strategies. Our findings rigorously demonstrate that the minimizer sequences from the first adaptive algorithm achieve convergence to a predefined tolerance. Crucially, the second algorithm is proven to generate inherently convergent sequences, thereby eliminating the need for an explicit stopping criterion. The practical effectiveness of this proposed adaptive framework is thoroughly validated through extensive numerical simulations. A case study involving an edge crack in an elastic body, governed by an algebraically nonlinear strain-limiting relationship and subjected to anti-plane shear-type loading, is presented. Critical comparisons of the energy components—bulk, surface, and total—showcase the superior performance of both adaptive algorithms. Key words. Galerkin finite element method; Adaptive algorithms; Convergence analysis, Quasi-static functional. AMS subject classifications. 65\mathrm{N}12 ; 65\mathrm{N}15 ; 65\mathrm{N}22 ; 65\mathrm{N}30 ; 65\mathrm{N}50 ; 65\mathrm{R}10 . 1 Introduction to brittle fracture models A foundational mathematical framework for modeling brittle fracture in materials is provided by the Francfort-Marigo model [ 17 ] . The total energy of the system within this model is comprised of three fundamental components: elastic energy , which represents the energy stored due to material deformation; bulk energy , about the intrinsic material properties and internal structure; and surface energy , associated with the formation of new fracture surfaces. While a material’s resistance to deformation is quantified by its elastic energy, the energy required for the creation of new fracture surfaces is captured by the surface energy. These components collectively define the system’s energy landscape. 1.1 Governing equations for strain-limiting elasticity Consider a smooth, open, connected, and bounded domain \mathcal{D}:=\mathcal{D}(t)\subset\mathbb{R}^{d},\,d=2,\,3 , possessing a given boundary \partial\mathcal{D} , with \Gamma(t) representing a crack set. This domain is considered to be occupied by an elastic body whose mechanical response is governed by a specialized nonlinear constitutive relation between the Cauchy stress tensor \bm{T}\colon\mathcal{D}\to\mathbb{R}^{d\times d}_{sym} and the linearized elasticity tensor \bm{\epsilon}\colon\mathcal{D}\to\mathbb{R}^{d\times d}_{sym} . Within the framework of strain-limiting theories of elasticity [ 34 ] , this constitutive class of relations is expressed in the form: (1.1) \mathcal{F}(\bm{T},\;\bm{B})=\bm{0}, where \mathcal{F}\colon\mathbb{R}^{d\times d}_{sym}\times\mathbb{R}^{d\times d}_{sym}\to\mathbb{R}^{d\times d}_{sym} denotes a nonlinear, tensor-valued function, while \bm{B} represents the left Cauchy-Green strain tensor. Through linearization, under the assumption of small displacement gradients, nonlinear constitutive relations such as the following can be derived [ 29 , 24 , 25 , 28 , 35 , 37 ] : (1.2) \bm{\epsilon}:=\mathcal{F}(\bm{T}). The following system of equations governs the material behavior [ 30 , 36 , 42 ] : (1.3) \displaystyle\begin{cases}\mathrm{div}\,\bm{T}=\bm{f},\quad\bm{T}=\bm{T}^{T}&\mbox{in}\quad\mathcal{D},\\ \bm{\epsilon}:=\mathcal{F}(\bm{T})&\mbox{in}\quad\mathcal{D}.\end{cases} In Equation ( 1.3 ), the first part of the first equation represents the balance of linear momentum , while the second part implies the symmetry of the stress tensor in the absence of body couples. For the specific case of anti-plane shear loading, the displacement vector is characterized by only one nonzero component (i.e., in the z -direction), which results in planar stress and strain tensors with only a single nonzero component. If u(x,y) is the Airy’s stress function , satisfying the relations: (1.4) \bm{T}_{13}=\dfrac{\partial u}{\partial y},\quad\bm{T}_{23}=-\dfrac{\partial u}{\partial x}, then the equilibrium equation is automatically satisfied. Consequently, the compatibility condition is reduced to: (1.5) \dfrac{\partial\bm{\epsilon}_{13}}{\partial y}-\dfrac{\partial\bm{\epsilon}_{23}}{\partial x}=0. The current study is specifically devoted to analyzing a subclass of the aforementioned general models. In this work, a material model is considered in which the linearized strain is expressed as a nonlinear function of stress, as given by: (1.6) \bm{\epsilon}:=\dfrac{\bm{T}}{2\mu\left(1+\beta^{\alpha}\,|\bm{T}|^{\alpha}\right)^{1/\alpha}}. Here, \beta\geq 0 and \alpha>0 are defined as modeling parameters, and \mu>0 represents the shear modulus. A similar constitutive class has been investigated in previous studies [ 20 , 21 , 26 , 29 , 41 , 42 , 44 , 45 ] . Substituting equations ( 1.4 ) into ( 1.6 ), and subsequently utilizing the strain components within the compatibility condition ( 1.5 ), yields a second-order quasi-linear elliptic partial differential equation. 1.2 Evolution of fracture mechanics and phase-field models The theoretical aspects of quasi-static and brittle fracture mechanics were first considered by Francfort and Marigo [ 17 ] , who also successfully conducted experiments on a two-dimensional example involving a domain with a straight initial crack. By rephrasing Griffith’s concept [ 23 ] of balancing energy release rate with a fictitious surface energy as an energy minimization problem; Francfort and Marigo developed a model that circumvented the standard constraints of classical fracture mechanics, such as a predetermined and piecewise smooth crack path. The Francfort-Marigo model approach is predicated on the minimization of a highly irregular energy functional, also known as the Mumford-Shah function in the context of image segmentation [ 32 ] . To render this problem amenable to numerical simulation, numerous regularization techniques have been introduced in the literature [ 6 , 11 , 33 ] . These techniques typically construct approximation functionals whose minimizers are demonstrated to converge to those of the original functional through the notion of \Gamma -convergence [ 9 ] . Furthermore, the Ambrosio-Tortorelli approximation [ 2 , 3 ] possesses particularly attractive properties among regularization methods, notably that its minimization can be reduced to the solution of elliptic boundary value problems, which are straightforward to discretize, for instance, using a finite element approach. This strategy has been effectively employed by Bourdin, Francfort, and Marigo [ 6 ] and Bourdin [ 7 , 8 ] for simulating problems that typically fall outside the scope of conventional techniques. The Ambrosio-Tortorelli approximation can be conceptualized as a phase-field model for the fracture set. To accurately resolve the phase-field variable, the mesh in the vicinity of the fracture must be significantly finer than that required for determining the elastic deformation alone. Consequently, the utilization of an adaptively refined mesh becomes a natural consideration, particularly as the fracture path is not known a priori . 1.3 Our Contribution. The entire paper is arranged as follows: In Section 2 , function spaces, basic definitions, and the nonlinear setting of the Francfort-Marigo model of brittle fracture are introduced. Using these fundamental arguments and explanations, the Fréchet differentiability of the minimization functional is demonstrated. Furthermore, it is proven that every phase-field minimizer satisfies the condition 0\leq v(x)\leq 1 for almost all x\in\mathcal{D} . Section 3 focuses on the finite element setup and discretization of the minimization problem. Section 4 is divided into several parts, where the minimization algorithm is developed, residual-based estimates are derived, and the adaptive algorithm, which utilizes these estimates to improve the mesh, is presented. Additionally, convergence results are proven in Theorems 4.6 –

Read the complete text