A unified variational framework for phase-field fracture and third-medium contact in finite deformation hyperelasticity

Jaemin Kim, Gukheon Kim, Sungmin Yoon, Dong-Hwa Lee · 2026

Read A unified variational framework for phase-field fracture and third-medium contact in finite deformation hyperelasticity by Jaemin Kim, Gukheon Kim, Sungmin Yoon, Dong-Hwa Lee (2026) 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-2603-15635.txt.

Introduction

Kim and co-authors join two hard simulation tasks in one variational framework. Phase-field fracture turns sharp cracks into a smooth damage field. Third-medium contact replaces a contact surface with a soft fake medium. Both tricks remove the need for crack tracking or contact search. The team tests two-dimensional bending and a three-dimensional Brazilian disk. The disk test shows crushing zones near contact points, as seen in lab tests.

Catalog record

CREATORJaemin Kim, Gukheon Kim, Sungmin Yoon, Dong-Hwa Lee
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE72763
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.15635

Browse the full library · About the Tower of Babel

Full text

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

A unified variational framework for phase-field fracture and third-medium contact in finite deformation hyperelasticity Jaemin Kim, Gukheon Kim, Sungmin Yoon, Dong-Hwa Lee Abstract This paper presents a unified variational framework that integrates phase-field fracture (PFF) and third-medium contact (TMC) within finite deformation hyperelasticity. The key idea is that both crack and contact are treated through regularization: the sharp crack topology is regularized into a diffuse damage field, while the discrete contact interface is regularized by a compliant fictitious medium with auxiliary fields. This strategy eliminates the need for explicit contact detection or crack tracking algorithms. The framework is validated through two-dimensional three-point bending and three-dimensional Brazilian disk test simulations, demonstrating the interplay between contact-induced stress concentration and crack nucleation/propagation. In particular, the Brazilian disk simulation naturally reproduces secondary crushing-type fracture zones near the contact regions—a phenomenon consistently observed in experiments yet inaccessible to simplified loading models. These results pave the way for predictive simulation of coupled contact–fracture phenomena without recourse to explicit interface tracking. Nomenclature Symbol Definition Symbol Definition \mathbf{X},\mathbf{x} Reference/Current position vectors \mathbf{u} Displacement field \bm{\varphi} Deformation mapping \mathbf{F} Deformation gradient J,\widetilde{J} Jacobian determinant/Regularized J \mathbf{C} Right Cauchy–Green tensor \mathbf{I} Identity tensor I_{1},I_{3} Principal invariants of \mathbf{C} \bm{\varepsilon} Infinitesimal strain tensor \varepsilon_{i},\mathbf{n}_{i} Principal strains and eigenvectors \mathbf{P},\bm{\sigma} First Piola–Kirchhoff/Cauchy stress \mathbf{N} Unit outward normal \mathbb{A} Fourth-order elasticity tensor \mathbf{B},\mathbf{T} Body force/Surface traction \lambda,\mu,\kappa Lamé parameter, shear and bulk moduli \Omega_{0},\Omega_{1},\Omega_{2},\Omega_{3} Reference configuration and subdomains \Gamma Crack surface S_{u},S_{t} Dirichlet/Neumann boundaries \Pi Total potential energy \Psi Strain energy density d,\ell Phase-field damage and length scale g(d),w(d) Degradation/Crack geometric functions G_{c} Fracture toughness c_{w} Crack density normalization k_{\ell} Residual stiffness \Psi_{1}^{\pm} Tensile/Compressive strain energies \varepsilon Jacobian regularization parameter \gamma Third-medium scaling parameter \omega,\bm{\xi} Scalar/Vector microstresses \mathbf{W} Skew-symmetric part of \mathbf{F} p,q Auxiliary regularization fields \alpha_{r} Gradient regularization \beta_{1},\beta_{2} Penalty parameters 1 Introduction Contact and fracture phenomena are widely encountered in engineering applications across multiple length scales, from nanoindentation testing for thin-film characterization ( 44 ; 24 ; 54 ; 30 ) to impact-induced delamination in laminated composites ( 40 ; 10 ) , interfacial failure in thermal barrier coatings ( 18 ; 26 ) , crack formation in battery electrodes ( 33 ; 14 ; 9 ) , and rupture of biological tissues under physiological loading ( 22 ; 19 ; 17 ; 29 ) . The simultaneous occurrence of contact and fracture is particularly prevalent in indentation-induced failure scenarios, where the concentrated stresses beneath a rigid indenter can nucleate cracks that subsequently propagate through the substrate ( 23 ) . Despite significant advances in computational mechanics, the unified treatment of these phenomena within a computational framework remains an open challenge because of the inherent complexity of contact algorithms themselves. Classical contact formulations require a sequence of non-trivial algorithmic steps: (i) contact detection through closest-point projection to identify active contact pairs on evolving surfaces, (ii) master–slave surface designation, which introduces asymmetry and can bias the solution, and (iii) enforcement of the non-penetration constraint. The penalty method, while straightforward to implement, introduces a user-defined stiffness parameter whose choice governs a trade-off between residual penetration and ill-conditioning; the Lagrange multiplier method enforces the constraint exactly but enlarges the algebraic system with a saddle-point structure and requires satisfaction of the inf-sup (Ladyzhenskaya–Babuška–Brezzi) condition for stable discretization; the augmented Lagrangian approach combines elements of both but at the cost of an iterative update loop ( 50 ) . This challenge is even greater when finite deformation is taken into account, as large sliding and evolving surface topology demand frequent updates of the contact search and projection operators. Coupling such contact formulations with fracture introduces a further layer of complexity. Crack propagation continuously creates new internal surfaces that may themselves come into contact—for instance, crack-face closure under compressive loading or debris interaction during fragmentation. This means the contact detection and constraint enforcement must operate on a dynamically evolving and a priori unknown topology, a requirement that is difficult to satisfy within conventional node-to-segment or segment-to-segment frameworks without substantial algorithmic augmentation. For these reasons, numerous computational approaches have been developed to study fracture without explicit surface tracking. Early approaches based on remeshing tracked crack surfaces explicitly and required costly mesh updates at each propagation step ( 13 ) . The extended/generalized finite element method (XFEM/GFEM) alleviated this by enriching the approximation space with discontinuous and near-tip functions, enabling cracks to propagate independently of the mesh ( 38 ; 8 ) ; however, robust implementation in three dimensions and for multiple interacting cracks remains challenging. Cohesive zone models insert traction–separation relations along predefined element interfaces ( 56 ; 39 ) , but their accuracy depends on element alignment with the (unknown) crack path. Phase-field fracture (PFF) methods have emerged as an alternative paradigm that circumvents these difficulties, enabling the simulation of crack initiation, propagation, branching, and coalescence without explicit crack tracking ( 20 ; 12 ; 37 ) . The approach regularizes the sharp crack topology through a continuous damage field d\in[0,1] that interpolates between intact ( d=0 ) and fully broken ( d=1 ) material states. This regularization transforms the free-discontinuity problem of Griffith fracture into a coupled boundary value problem amenable to standard finite element discretization. A key advantage is that crack path and topology emerge naturally from energy minimization, eliminating the need for crack tracking algorithms ( 11 ; 31 ) . The variational structure also provides a thermodynamically consistent framework for incorporating history-dependent irreversibility and various energy decompositions to prevent crack healing under compression ( 3 ) . The model has been extended to the case of finite deformations in soft materials ( 45 ; 58 ; 43 ; 28 ) , anisotropic fracture ( 52 ; 42 ; 57 ) , and length-scale insensitive formulations ( 46 ) . The approach has also been applied to thin-film and film-substrate systems ( 6 ; 21 ; 23 ; 34 ; 30 ) . On the contact mechanics, the third-medium contact (TMC) method has been developed as an elegant alternative to classical contact formulations based on penalty methods, Lagrange multipliers, or mortar approaches ( 49 ; 48 ) . The central idea is to introduce a compliant fictitious medium—the “third medium”—that fills the gap between potentially contacting bodies. As the bodies approach each other, the third medium compresses and naturally transmits contact forces, eliminating the need for contact detection algorithms and the associated algorithmic complexity. Notably, the resulting contact tractions have been shown to closely reproduce classical Hertzian contact solutions ( 27 ; 51 ) , confirming the physical fidelity of the approach. The method is particularly attractive for large-deformation problems where contact surfaces undergo significant changes in geometry. However, the severe compression of the third medium can lead to mesh distortion and element inversion. To address this issue, 16 proposed an auxiliary-field regularization, in which additional fields track local rotation and volumetric deformation to maintain mesh quality. Extensions to thermo-mechanical coupling ( 51 ) and three-dimensional hyperelastic contact ( 55 ) further demonstrate the versatility of the TMC framework. Despite these parallel advances in PFF and TMC, a unified formulation that couples both methods within a single variational framework has not yet been established; as a result, a deep understanding of the interplay between

Read the complete text