🔬 physics

A comprehensive anisotropic phase-field fracture formulation with a novel positive/negative projection scheme

This paper presents a thermodynamically consistent, monolithically solved phase-field framework for anisotropic brittle fracture that utilizes a novel positive/negative projection scheme to accurately model orientation-dependent crack propagation under complex loading conditions, validated through extensive numerical benchmarks and implemented via Abaqus user subroutines.

Original authors: Shank Kulkarni, Mohammad Safi, Timothy Truster, Khalid A. Alshibli

Published 2026-07-16

📖 3 min read☕ Coffee break read

CC BY 4.0

Original authors: Shank Kulkarni, Mohammad Safi, Timothy Truster, Khalid A. Alshibli

Original paper licensed under CC BY 4.0 (https://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 you are watching a piece of glass shatter. In the real world, cracks are sharp, jagged lines that appear out of nowhere and race across the surface, changing direction in unpredictable ways. For a long time, trying to predict exactly where those cracks would go on a computer was like trying to draw a map of a storm while the wind was still blowing. Engineers had to guess where the crack would start, and if it decided to turn a corner, the computer program would often crash or give up, requiring a complete restart. This is a big deal because broken materials cost the world billions of dollars every year, and understanding how things break helps us build safer bridges, planes, and phones.

To solve this, scientists invented a clever trick called the "phase-field" method. Instead of trying to draw a sharp, thin line for a crack, they imagine the crack is a fuzzy, blurry cloud that spreads out a little bit. Think of it like a drop of ink spreading in water; the center is dark (fully broken), the edges are light (partially damaged), and the rest is clear (perfectly strong). This "fuzzy crack" approach lets computers handle cracks that twist, turn, branch, and merge without needing to redraw the map every time the crack moves. However, there was a catch: this method sometimes got confused. It would think that squeezing a material (compression) was just as likely to break it as pulling it apart (tension), which doesn't make sense in the real world. You can squeeze a rock all you want, and it won't crack open like a soda can, but pull it, and it snaps.

This paper introduces a new, smarter way to handle that confusion, specifically for materials that act differently depending on which way you pull them, like wood or certain crystals. The researchers, Shank Kulkarni and his team from the University of Tennessee, developed a "Positive/Negative Projection" scheme. You can think of this as a special filter or a bouncer at a club. When the computer calculates the energy trying to break the material, this filter checks the direction. If the energy is trying to pull the material apart (positive/tension), the bouncer lets it through to cause damage. But if the energy is trying to squeeze the material together (negative/compression), the bouncer stops it cold, ensuring the material stays strong. The team tested this new filter on a computer simulation of a material with a crystal structure (trigonal symmetry) and found that it correctly predicted how cracks would start, grow, and even change direction based on the material's internal grain, all while refusing to let the material break under pure squeezing. They verified this by running hundreds of tests on a computer, showing that their new method is a reliable way to simulate how complex, direction-sensitive materials fail.

Technical Summary: A Comprehensive Anisotropic Phase-Field Fracture Formulation with a Novel Positive/Negative Projection Scheme

Problem Statement
Phase-field (PF) theory has emerged as an efficient approach for simulating complex crack patterns without predefined paths. However, extending these models to anisotropic brittle materials presents significant challenges. While isotropic PF models often rely on strain energy decompositions (such as spectral splits or volumetric-deviatoric splits) to prevent unphysical crack growth under compression, these methods frequently fail to maintain thermodynamic consistency or energy orthogonality when applied to anisotropic elasticity. Specifically, existing positive/negative (PN) projection schemes often assume that stress and strain are coaxial—a valid assumption for isotropic materials but one that does not generally hold for anisotropic solids where principal stress and strain directions diverge. Consequently, there is a lack of a rigorously consistent framework that simultaneously handles elastic anisotropy, ensures thermodynamic balance, and suppresses damage under compressive states without heuristic approximations.

Methodology
The authors developed a comprehensive phase-field framework for anisotropic brittle materials, implemented within the commercial finite element software Abaqus via user element (UEL) and user material (UMAT) subroutines. The core of the methodology involves three primary components:

Variationally Consistent Framework: The formulation adheres to fundamental balance laws and the first and second laws of thermodynamics. It utilizes a monolithic solution strategy to solve the coupled displacement and phase-field equations simultaneously, ensuring robust convergence.
Anisotropic Elasticity: The model incorporates trigonal material symmetry (representative of quartz-type crystals) by rotating the stiffness tensor from the crystal coordinate system to the global finite element frame. This allows for the capture of orientation-dependent stiffness and mechanical response.
Novel Anisotropic Positive/Negative (PNP) Projection Scheme: To address the non-coaxiality of stress and strain in anisotropic materials, the authors propose a novel PNP scheme based on the complementary energy norm.
- Mapping Strategy: The method maps the real anisotropic stress and strain tensors to an auxiliary, fictitious isotropic space using transformation tensors derived from energy equivalence.
- Projection: In this auxiliary isotropic space, where stress and strain are coaxial, a Linear Complementarity Problem (LCP) is solved to decompose the strain into positive (tensile) and negative (compressive) parts. This ensures the decomposition is orthogonal in the energy norm and thermodynamically consistent.
- Back-Transformation: The projected quantities are mapped back to the real anisotropic space. The phase-field degradation function is applied only to the tensile (positive) contribution, while the compressive (negative) contribution remains intact, effectively suppressing crack growth under compression.

Key Contributions

Thermodynamically Consistent Anisotropic Split: The paper introduces a rigorous extension of the Wu and Cervera PNP scheme to anisotropic elasticity. By utilizing an energy-equivalent mapping to an auxiliary isotropic space, the formulation guarantees that the crack-driving energy is consistently associated with tensile deformation, even when stress and strain are not coaxial.
Monolithic Abaqus Implementation: The authors provide a practical implementation using a UEL for the coupled displacement-damage system and a UMAT for post-processing visualization. This approach leverages the robustness of the Abaqus solver while allowing for complex anisotropic constitutive behavior.
Comprehensive Verification: The framework is verified through a hierarchy of tests, including single-element tests under tension, compression, and shear; patch tests with distorted meshes to ensure element formulation robustness; and multiple quasi-static fracture benchmarks (single-edge-notched tension, compression, shear, and three-point bending).

Results
The numerical results demonstrate the efficacy of the proposed formulation:

Isotropic Benchmarks: The model reproduces analytical solutions and literature results for standard isotropic benchmarks (e.g., single-edge-notched tension and three-point bending) with high accuracy. It successfully suppresses damage under pure compression and exhibits symmetric responses for positive and negative shear loading.
Anisotropic Behavior: For trigonal materials (silica sand crystals), the model captures distinct orientation-dependent responses.
- Stiffness and Strength: Simulations show higher peak loads and stiffer responses when loading aligns with the crystallographic "strong" direction (c-axis) compared to the "weak" direction (a-axis).
- Crack Trajectories: The model predicts orientation-dependent crack paths. In the strong orientation, cracks deviate from the initial notch direction due to anisotropic coupling terms (e.g., $C_{14}$ ), bending toward an energetically preferred direction before reorienting near free boundaries. In the weak orientation, cracks propagate more linearly.
- Compression and Shear: The LCP-based split successfully prevents damage evolution under compressive loading for both orientations. Under shear, the model captures the asymmetry in response between positive and negative shear loading for the strong orientation, a direct consequence of the trigonal symmetry and constitutive coupling, while maintaining symmetry for the weak orientation.
Parametric Studies: The study confirms that mesh refinement, regularization length scale, and critical energy release rate ( $G_c$ ) influence the peak load and post-peak softening behavior, consistent with phase-field theory expectations.

Significance
The paper claims that this work provides a "comprehensive and novel framework" for quantifying the fracture behavior of anisotropic brittle materials. Its primary significance lies in bridging the gap between rigorous thermodynamic consistency and the practical modeling of directionally dependent fracture. By extending the PNP scheme to anisotropic elasticity through an energy-equivalent mapping, the formulation overcomes the limitations of previous heuristic splits that assumed stress-strain coaxiality. The authors assert that the model is capable of representing fracture phenomena in directionally dependent brittle solids, capturing not only orientation-dependent stiffness and peak loads but also complex, anisotropy-driven crack trajectories. The work is presented as a step toward more accurate simulations of materials like quartz and other crystalline structures, though the authors note current limitations to trigonal symmetry, 2D geometries, and orientation-independent fracture energy, suggesting these as targets for future extension.

More like this