🔢 mathematics

Quasistatic evolution of cohesive-type fracture

This paper establishes the existence of globally stable quasistatic evolutions for cohesive fracture models in arbitrary dimensions with unprescribed crack paths by introducing a novel convergence notion for memory variables and a modified proof strategy that prioritizes energy balance over the preservation of global stability during the limit process.

Original authors: Vito Crismale, Manuel Friedrich

Published 2026-07-03

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Vito Crismale, Manuel Friedrich

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a piece of material, like a sheet of rubber or a block of glass, sitting on a table. Over time, you pull on it. Eventually, it might crack.

In the world of physics and math, there are two main ways materials break:

Brittle Fracture: Think of a dry twig snapping. It breaks suddenly, and once it's broken, it's broken. The energy required to make the crack is constant.
Cohesive Fracture: Think of peeling a sticker off a wall or stretching a piece of chewing gum. Before it snaps completely, the material stretches, thins out, and resists the break with a "sticky" force. This resistance changes depending on how wide the crack gets. This is called a cohesive zone.

This paper is a mathematical proof that shows such a "sticky" crack can evolve in a predictable, stable way over time, even if we don't know exactly where the crack will go or what shape it will take.

Here is a breakdown of their journey, using simple analogies:

1. The Problem: The "Unpredictable" Crack

The authors wanted to prove that if you slowly pull on a material with these "sticky" properties, you can mathematically describe exactly how it will break step-by-step.

The tricky part is that the crack isn't just a line; it's a surface that can wiggle, branch, and change shape. In the past, mathematicians could only prove this for simple, straight cracks or in one dimension (like a line). They couldn't handle the messy, branching reality of 3D materials without making up rules about where the crack had to go.

2. The Hurdle: The "Oscillation" Trap

To prove the crack behaves well, the authors used a common math trick: they imagined the crack moving in tiny, discrete steps (like a flipbook animation) and then tried to smooth those steps into a continuous movie.

Usually, in simpler problems (brittle fracture), if you smooth out the steps, the final picture looks exactly like the sum of the steps. But in this "sticky" fracture problem, something weird happens:

The Wiggle: The tiny steps might show the crack wiggling back and forth wildly (oscillation).
The Branch: The tiny steps might show two tiny cracks merging into one big crack.

If you just look at the final shape, you might think the energy is low. But if you look at the history of how it got there (the wiggles and branches), the energy was actually much higher. It's like trying to flatten a crumpled piece of paper: the flat paper looks small, but the crumpled one took up a lot of space and energy to get that way.

Because of these wiggles, the standard math rules for proving stability failed. The "smoothed" version didn't seem stable enough to be the real answer.

3. The Solution: The "E-S" Detour

The authors realized they had to change their strategy. Instead of trying to prove the crack is stable first and then check the energy, they did it in reverse. They call this the E–S approach (Energy first, Stability second).

Here is the analogy:

The Old Way: "I know this car is safe (Stable), so I'll check if it uses the right amount of gas (Energy)."
The New Way: "I know this car uses the exact right amount of gas (Energy), so I can prove it must be safe (Stable)."

By first proving that the total energy of the system behaves perfectly (it doesn't magically disappear or appear), they could use that information to prove that the crack is indeed stable, even with all those wiggles and branches.

4. The New Tool: "σcf-Convergence"

To handle the messy crack shapes, they invented a new way of measuring how one crack shape turns into another. They call it σcf-convergence (think of it as a "Cohesive Fracture" convergence).

Imagine you are trying to describe a cloud.

Old method: You try to match the outline of the cloud exactly. If the cloud wiggles, you fail.
New method: You look at the "density" of the cloud. Even if the edges wiggle, if the total amount of water vapor and the way it sticks together remains consistent, you say the clouds are "converging."

This new tool allowed them to ignore the tiny, high-energy wiggles in the math and focus on the big picture, proving that the crack evolves smoothly in the long run.

5. The Result: A Stable Story

The paper concludes that for any material with these "sticky" properties (in any number of dimensions, not just 1D or 2D), there exists a mathematically perfect, stable path for the crack to follow.

It starts with the material unbroken.
It evolves by minimizing energy at every single moment.
It remembers its history (how wide the crack got before), which prevents it from snapping back together unrealistically.
It balances the energy perfectly, accounting for the work done by pulling and the energy lost to the crack.

Summary

The authors built a new mathematical bridge to cross a gap that previously seemed impossible. They showed that even when a material breaks in a complex, "sticky" way with unpredictable shapes, the laws of physics still hold up, and the process is stable and predictable. They did this by inventing a new way to measure cracks and by flipping the order of their proof to focus on energy conservation first.

Technical Summary: Quasistatic Evolution of Cohesive-Type Fracture

Problem Statement
The paper addresses the mathematical existence of globally stable quasistatic evolutions for cohesive-type fracture models in arbitrary dimensions ( $d \geq 2$ ). Unlike brittle fracture, where surface energy is constant, cohesive fracture involves a surface energy density $g$ that depends on the crack opening $[u]$ . The authors focus on a specific model where $g$ is concave, exhibits an activation threshold ( $g(0+) > 0$ ), and is defined on an unprescribed crack path without topological restrictions.

The core difficulty lies in the fact that standard existence proofs for rate-independent systems (energetic solutions) rely on the closedness of the stability set. In brittle fracture, this is often handled via jump-transfer lemmas. However, in cohesive fracture, the presence of an activation threshold and the dependence of energy on the crack opening amplitude introduce oscillation and branching phenomena in approximating sequences. These phenomena prevent the direct preservation of global stability under the passage to the limit, a challenge that has previously restricted existence results to one dimension or models with prescribed crack paths.

Methodology
The authors employ a time-discretization approach based on incremental minimization problems, a standard technique for rate-independent systems. However, they deviate from the classical scheme for proving energetic solutions, which typically establishes global stability (S) first and then the energy balance (E). Instead, they adopt an E–S approach:

First, they prove the energy balance and the convergence of surface energies.
Only subsequently do they recover the global stability condition.

Key methodological innovations include:

$\sigma_{cf}$ -convergence: The authors devise a new notion of convergence for memory variables (pairs of crack sets and maximal opening functions) supported on evolving crack sets. Inspired by $\sigma$ -convergence in brittle fracture, this notion ensures compactness and lower semicontinuity of the dissipation.
Integral Representation: They establish a $\Gamma$ -convergence result for functionals with precracks, characterizing the limit surface energy density $h^-$ which depends on the crack opening.
Localization and Induction: To overcome the failure of global stability preservation, the authors use a localization argument on small cubes. They identify "starting times" for crack appearance where the crack is locally flat. By exploiting the activation threshold $g(0+) > 0$ , they control oscillation and branching through an inductive procedure in time, ensuring that surface energies converge.

Key Contributions and Results

Existence Theorem: The paper proves the existence of an irreversible quasistatic crack evolution $t \mapsto (u(t), K(t), \psi(t))$ satisfying:
- Initial Condition: The state at $t=0$ minimizes the initial energy.
- Irreversibility: The crack set $K(t)$ and the dissipated energy density $\psi(t)$ are non-decreasing in time.
- Global Stability: The state is a global minimizer of the energy functional among all admissible competitors with compatible history.
- Energy Balance: The total energy satisfies the exact balance equation involving the work of external forces.
New Convergence Notion: The introduction of $\sigma_{cf}$ -convergence allows for the handling of memory variables in cohesive fracture. The authors demonstrate that while $\sigma_{cf}$ -convergence guarantees lower semicontinuity, it does not automatically preserve global stability due to potential energy excess caused by oscillation (mechanism A) or branching (mechanism B) in the approximating sequences.
Resolution of the Stability Gap: The paper identifies that the gap between "weak" stability (minimality with respect to the $\Gamma$ -limit) and "strong" stability (minimality with respect to the original energy) corresponds precisely to situations where surface energies fail to converge. By proving the convergence of surface energies a priori to the stability step, the authors bridge this gap and recover the strong stability condition.

Significance and Claims
The authors claim that this work serves as a "proof of concept" demonstrating that the theory of energetic solutions is sufficiently flexible to encompass free-discontinuity evolutions in cohesive fracture, provided the proof strategy is suitably adapted.

Methodological Impact: The proposed E–S approach (establishing energy balance before stability) is highlighted as a potentially useful strategy for other rate-independent problems where stability sets are difficult to close directly.
Generality: The result holds in arbitrary dimensions and within the weak setting of Special functions of Bounded Variation (SBV), removing the topological constraints and dimensional limitations of previous cohesive fracture existence results.
Scope: The paper focuses on a scalar model with isotropic energy densities to illustrate the effectiveness of the approach, noting that more complicated settings (e.g., vectorial elasticity) could be treated with additional technical effort.

The work does not propose new experimental applications or future physical models but rather provides a rigorous mathematical foundation for the existence of solutions in a class of cohesive fracture models that were previously analytically intractable in higher dimensions.