Geometric Crack Modeling Module

• Geometric crack modeling modules are computational frameworks that simulate crack propagation using explicit geometric representations and lattice-based algorithms.
• They integrate elastic stress solvers, threshold criteria, and wavelet-based roughness analyses to quantify crack front morphology and regime transitions.
• The module supports both displacement-control and gradient loading to calibrate material properties and match experimentally observed scaling exponents.

A geometric crack modeling module is a computational framework for simulating, analyzing, and predicting the growth and morphology of cracks in solids through explicit geometric or lattice-based representations. These modules encode crack topology, discretization of material interactions, and algorithms for crack advance, branching, and statistical characterization. The goal is to capture both the physical mechanisms (elastic interactions, threshold disorder) and the geometric/statistical features (front roughness, scaling exponents) of interfacial crack propagation.

1. Model Geometry and Physical Setup

The essential geometric framework, as presented in "A model for stable interfacial crack growth" (Gjerden et al., 2012), consists of a constrained configuration where two blocks—one perfectly rigid, one with finite Young's modulus $E$ and Poisson ratio $\nu$—are separated by a square lattice of $N=L^2$ linear elastic fibers (discrete springs), each spanning a patch $a \times a$. The crack propagates along the interfacial plane between the blocks, with the lattice discretization indexed in two dimensions: $i \equiv (x_i, y_i)$.

Boundary conditions are periodic in the direction parallel to the crack front ($x$), simulating an infinite system. Depending on the system's stiffness, bi-periodic boundary conditions (for $x$ and $y$) are employed for numerical stability in conjugate-gradient solvers for soft regimes.

Loading is implemented either by imposing a global displacement $D$ (displacement-control), where intact fiber $i$ carries force $\nu$0, or by assigning fiber breaking thresholds $\nu$1 (threshold-gradient loading), simulating an external gradient of driving force.

2. Governing Equations and Failure Criteria

The crack modeling module is constructed from the following field equations and key physical assumptions:

• Stress-strain relations: Each fiber is linear elastic, with force law

$\nu$2

where $\nu$3 is the local stiffness ($\nu$4 for intact, $\nu$5 for broken fibers). Displacement $\nu$6 is not local but coupled: the elastic block transmits force via a discretized Green's function,

$\nu$7

which, in practice, is implemented via Love's formula for a square patch.

• Fracture criterion: For each fiber, local tensile strain is

$\nu$8

which is compared to the breaking threshold $\nu$9. In computation, the fiber with largest $N=L^2$0 is chosen for irreversible failure (extremal-dynamics).

• Crack advance: There is no explicit calculation of a stress-intensity factor; crack evolution occurs by local failures, triggered when $N=L^2$1 meets the threshold $N=L^2$2 at a site.

3. Geometric and Statistical Front Analysis

To extract and quantify crack morphology, the module tracks a front height function $N=L^2$3 by computing the upper envelope of the largest connected cluster of broken fibers. This mapping removes overhangs using a solid-on-solid (SOS) criterion ($N=L^2$4), yielding a well-defined single-valued front.

The module then measures self-affine scaling properties: $N=L^2$5 Alternatively, wavelet-based roughness analysis computes the mean modulus of wavelet coefficients $N=L^2$6 over scale $N=L^2$7.

Typical roughness exponents—$N=L^2$8 (stiff regime), $N=L^2$9 (soft regime)—are observed, matching but below the corresponding experimental values ($a \times a$0, $a \times a$1).

4. Scaling Mechanism and Physical Regimes

The geometric crack modeling module reveals two distinct self-affine regimes, each arising from different physical mechanisms:

• Small-scale regime ($a \times a$2): Dominated by fracture coalescence. Microcrack islands nucleate ahead of the front and merge (coalescence model). Roughness exponent $a \times a$3 experimentally, $a \times a$4 in the model.
• Large-scale regime ($a \times a$5): The crack front behaves as a fluctuating elastic line driven through a random medium (line model). $a \times a$6 (model); $a \times a$7 (experiment).

The crossover length $a \times a$8 between regimes increases with block stiffness $a \times a$9 and decreases with gradient $i \equiv (x_i, y_i)$0, approximately following

$i \equiv (x_i, y_i)$1

where $i \equiv (x_i, y_i)$2 are non-universal parameters dependent on system details.

5. Computational Architecture and Algorithms

Implementation is structured around efficient data handling and algorithmics:

Core data structures:

• $i \equiv (x_i, y_i)$3: status array (intact = 1, broken = 0)
• $i \equiv (x_i, y_i)$4: threshold array
• $i \equiv (x_i, y_i)$5: current fiber forces
• $i \equiv (x_i, y_i)$6: local displacements
• $i \equiv (x_i, y_i)$7: front height extraction

Algorithm pseudocode:

$x$9

Stress solver: Fast conjugate-gradient (CG) with FFT-based acceleration. Stress transfer is diagonal in Fourier space due to the Green's function structure. Preconditioning is required at high stiffness; double precision is recommended to control round-off.

Crack-advance logic: The extremal-dynamics rule—breaking the fiber of largest $i \equiv (x_i, y_i)$8 at each step—naturally advances the crack front without explicit computation of $i \equiv (x_i, y_i)$9.

6. Regime Calibration and Practical Integration

For robust quantitative modeling, system parameters are selected according to physical and computational considerations:

• System size: $x$0 recommended to reduce finite-size effects, especially in extracting roughness exponents.
• Elastic modulus $x$1 and Poisson ratio $x$2: Wide sampling across $x$3 to $x$4, typical $x$5–$x$6.
• Threshold gradient $x$7: Chosen so that $x$8–$x$9.
• CG tolerance: $x$0–$x$1 per solve.
• Computational complexity: Each CG iteration is $x$2; total cost $x$3. Preconditioning required for numerical stability in high-$x$4 regimes.

Module extensions include:

• Calibration of $x$5 and threshold noise against experimental data on fracture toughness.
• Substitution of non-uniform $x$6 for modeling material heterogeneity.
• Introduction of anisotropic $x$7 for layered media.
• Extraction and fitting of $x$8 via wavelet or height function analysis.

7. Scientific Impact and Comparison to Experiments

The geometric crack modeling module systematically unifies mechanisms for stable interfacial crack growth, matching experimentally observed dual-regime self-affine scaling. Its algorithms are suitable for direct implementation, numerical parameter scans, and quantitative comparison to physical measurements. Notably, the module resolves the regime crossover and provides a computational laboratory for testing statistical properties, scaling laws, and the influence of elasticity and disorder on crack front morphology (Gjerden et al., 2012).

This framework's modularity and explicit geometrization make it adaptable to broader classes of interfacial fracture problems, including layered composites, anisotropic systems, and real materials with calibrated disorder. Its statistical characterization tools (SOS mapping, wavelet analysis) are directly transferable to experimental imaging data, facilitating rigorous theory–experiment comparison.