arxiv: 2605.03731 · v1 · submitted 2026-05-05 · ❄️ cond-mat.mtrl-sci

Recognition: unknown

A Correction Method for Crack Area Overestimation in Phase-Field Fracture

I. Romero, J. Segurado, M. Castill\'on

Authors on Pith no claims yet

Pith reviewed 2026-05-07 15:49 UTC · model grok-4.3

classification ❄️ cond-mat.mtrl-sci

keywords phase-field fracturecrack area correctionenergy equipartitionmesh-independent methoddiffuse crack representationthree-dimensional fracturenumerical artifacts

0 comments

The pith

Crack area in phase-field fracture models can be recovered as twice the gradient energy term.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Phase-field fracture models represent cracks as smooth transitions rather than sharp cuts, which causes them to overestimate the actual crack surface area. This overestimation arises from both the diffuse nature of the model and numerical issues like the phase-field variable reaching its maximum value inside single elements. The paper proposes a fix that relies on the fact that the model's energy splits equally between the phase-field contribution and its gradient contribution when the length-scale parameter becomes small. Because numerical problems mainly distort the phase-field part, the true crack area is recovered simply by doubling the gradient energy. This approach requires no special mesh handling and works directly in three dimensions.

Core claim

The central claim is that the crack area can be accurately approximated as twice the gradient-dependent energy. This follows from energy equipartition in the limit of vanishing length-scale parameter, combined with the observation that numerical artifacts primarily impact the phase-field term while leaving the gradient term largely unaffected. The resulting correction is mesh-independent and applicable over the entire computational domain, including in three-dimensional settings with complex crack geometries.

What carries the argument

Energy equipartition between the phase-field and gradient energy terms, allowing the crack area to be computed as twice the gradient energy.

If this is right

The corrected crack area matches analytical solutions in benchmark problems more closely than uncorrected values.
Established methods like skeletonization are no longer necessary for accurate area measurement.
The method applies seamlessly to unstructured meshes and three-dimensional domains.
Fracture simulations can incorporate reliable crack area calculations without additional post-processing steps.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

If the correction holds, it could improve predictions of energy release rates in dynamic fracture simulations.
Similar energy-based corrections might apply to other diffuse models in materials science.
Implementation in existing phase-field codes would be straightforward since only the gradient energy needs to be doubled.

Load-bearing premise

Numerical artifacts distort the phase-field energy term but leave the gradient energy term accurate, while the two terms become equal as the length-scale parameter goes to zero.

What would settle it

Running the method on a benchmark problem with a known exact crack length or area, such as a straight crack in a plate, and checking if the corrected area converges to the true value independently of mesh size.

Figures

Figures reproduced from arXiv: 2605.03731 by I. Romero, J. Segurado, M. Castill\'on.

**Figure 1.** Figure 1: Bar with a central crack: (a) Sharp crack representation as a discontinuity at the center. (b) Phase-field approximation showing the smooth transition of ϕ(x), regularized by the length scale l. For the purpose of analysis, it is useful to decompose Γ[ϕ] into two distinct energy contributions: a term dependent explicitly on the phase-field variable, Γϕ, and a term dependent on its gradient, Γ∇ϕ: Γϕ[ϕ] = Z … view at source ↗

**Figure 2.** Figure 2: Comparison of analytical solutions for the diffuse crack representation. (a) Phase-field profile ϕ(x) and (b) its derivative ϕ ′ (x) for various length scales l view at source ↗

**Figure 3.** Figure 3: Energy as a function of the l/a ratio. As l/a decreases, the total energy approaches one, and the contributions from the phase-field and its gradient become equal. of ϕ = 1 applied at the symmetry plane to represent the central crack. We consider parameters l = 0.1 mm and a = 1.0 mm, resulting in a ratio of a/l = 10. In this configuration, boundary effects are negligible, and the analytical energy values d… view at source ↗

**Figure 4.** Figure 4: Relative error in the total crack surface energy Γ, phase-field energy component Γϕ, and gradient energy component Γ∇ϕ in a finite element one dimensional simulation as a function of the mesh resolution ratio l/h. 9 view at source ↗

**Figure 5.** Figure 5: Illustration of the phase-field profile distortion caused by strain localization. The ideal profile (dashed line) is artificially flattened to ϕ = 1 over an element of size h (solid line), leading to an overestimation of the crack surface energy. Within the localized element, the phase-field is constant (ϕ = 1), and its gradient is zero view at source ↗

**Figure 6.** Figure 6: (a) and 6(b) illustrate this for the phase-field and its gradient, respectively. This alters the energy calculation. The total energy, taking the strain localization effect, is denoted as Γsl, and can be obtained by modifying the integral in Eq. (11) to account for the localized region: −1.0 −0.5 0.0 0.5 1.0 x/a 0.0 0.2 0.4 0.6 0.8 1.0 φ(x) (a) Phase-field profile ϕ(x) with strain localization. −1.0 −0.5 0… view at source ↗

**Figure 7.** Figure 7: Effect of strain localization on energy contributions for different element sizes h. As h decreases, both the total energy (a) and the phase-field energy (b) converge to the ideal theoretical solution (solid black line). 4 Crack surface overestimation correction Accurately measuring the crack surface area is a critical challenge in phase-field modeling. The standard approach of integrating the crack surfac… view at source ↗

**Figure 8.** Figure 8: Schematic of the center-cracked tension test specimen. (a) Full geometry, showing dimensions and loading configuration. (b) Half-symmetry finite element model. (c) Quarter-symmetry finite element model. This analysis investigates the influence of the length scale parameter l and the mesh size h. A critical aspect of this study is the ratio l/h, which can be realized through various combinations of these pa… view at source ↗

**Figure 9.** Figure 9: Comparison of numerical results with the analytical LEFM solution for the center-cracked specimen (Simulation 8). (a) Force-displacement response. (b) Stiffness degradation as a function of crack length. Since these simulations share the same l/h ratio, the Bourdin correction factor remains constant across all cases; this is represented by a solid horizontal line on the graph. As the length scale is reduce… view at source ↗

**Figure 10.** Figure 10: Convergence analysis for a constant mesh resolution ratio l/h = 2.5. (a) Stiffness degradation as a function of applied force for decreasing length scales l, illustrating convergence toward the LEFM solution. (b) Evolution of the DGCM correction factor with crack length compared to the constant Bourdin factor. Having analyzed the general trends of the correction factors, we now examine the convergence beh… view at source ↗

**Figure 11.** Figure 11: (b) illustrates the convergence of the crack length for a fixed stiffness value of 38.52514 kN/mm. The theoretical crack length corresponding to this stiffness is 0.9 mm. Since this measurement point is well beyond the initial crack nucleation phase and the associated force overshoot, both the Bourdin and DGCM corrections yield similar results. As the length scale parameter decreases, both methods converg… view at source ↗

**Figure 12.** Figure 12: Convergence of the DGCM correction factor as a function of the length scale parameter l for different crack lengths. 5.1.2 Convergence analysis with respect to mesh size h at constant length scale l In this case, the length scale parameter is fixed at l = 0.0025 mm, while the mesh size h is varied to study the influence of mesh refinement. The simulations corresponding to this analysis are detailed in row… view at source ↗

**Figure 13.** Figure 13: Convergence analysis with respect to mesh size h for a constant length scale l = 0.0025 mm. (a) Peak force convergence. (b) Crack length convergence for a specific stiffness value. 0.20 0.25 0.30 0.35 0.40 h/l 1.20 1.25 1.30 1.35 1.40 F Bourdin DGCM view at source ↗

**Figure 14.** Figure 14: Convergence of the Bourdin and DGCM correction factors as a function of mesh size h, evaluated at a crack length of 0.9 mm. model shown in view at source ↗

**Figure 15.** Figure 15: Crack propagation in the 3D center-cracked tension specimen. The images show the isosurface of the phase-field variable for ϕ > 0.95 at three different stages of the simulation, illustrating the evolution of the crack front view at source ↗

**Figure 16.** Figure 16: Configuration of the compact tension specimen. The complete geometry is shown with all dimensions labeled in blue, relative to the specimen width W = 40.0 mm. Boundary conditions, labeled in red, indicate the fixed constraint at the bottom loading hole and the applied forces at the top loading hole. Three configurations are analyzed: the standard CT specimen without additional holes (Specimen 1) and two m… view at source ↗

**Figure 17.** Figure 17: Final crack paths represented by the phase-field variable ϕ at the end of the simulation for: (a) Specimen 1 (H = 0.60W), (b) Specimen 2 (H = 0.56W), and (c) Specimen 4 (H = 0.64W). a peak force of 21.31 kN. Applying the corrections significantly modifies these values: the Bourdin method results in 41.88 mm and 19.45 kN, whereas the skeletonization and DGCM methods provide much closer estimates of 39.99 m… view at source ↗

**Figure 18.** Figure 18: Specimen 1: (a) Comparison of crack length correction factors from the Bourdin, Skeletonization, and DGCM methods against the uncorrected result. (b) Force-displacement curves corresponding to each correction method. Turning to Specimen 2, the correction factors and force-displacement responses are depicted in Figures 19(a) and 19(b), respectively. In the uncorrected simulation, the final crack length re… view at source ↗

**Figure 19.** Figure 19: Specimen 2: (a) Comparison of crack length correction factors from the Bourdin, Skeletonization, and DGCM methods against the uncorrected result. (b) Force-displacement curves corresponding to each correction method. 27.22 mm (15.67 kN) and 26.83 mm (15.87 kN), respectively. 10 15 20 25 30 Crack Length (a) [mm] 1.0 1.2 1.4 1.6 1.8 F Reference Bourdin Skeleton DGCM (a) Evolution of correction factors. 0.0… view at source ↗

**Figure 20.** Figure 20: Specimen 4: (a) Comparison of crack length correction factors from the Bourdin, Skeletonization, and DGCM methods against the uncorrected result. (b) Force-displacement curves corresponding to each correction method. In all three cases, the uncorrected results exhibit the highest overestimation of the crack length. While the Bourdin correction reduces this overestimation, a significant discrepancy remain… view at source ↗

read the original abstract

Phase-field fracture models are known to overestimate the crack area, a discrepancy that compromises the accuracy of fracture predictions. This issue stems from the diffuse crack representation and numerical artifacts, such as strain localization, where the phase-field variable artificially saturates across finite elements. Existing correction strategies, including mesh-dependent factors and skeletonization algorithms, have significant limitations. Mesh-based corrections are often unreliable for unstructured meshes, while skeletonization can be complex and inaccurate for intricate crack topologies, especially in three dimensions. This paper introduces a novel and robust framework to correct this overestimation. Our approach is founded on the principle of energy equipartition, where the energy contributions from the phase-field and its gradient are equal as the length-scale parameter approaches zero. Since numerical artifacts primarily affect the phase-field term while leaving the gradient term largely unperturbed, we propose that the crack area can be accurately approximated as twice the gradient-dependent energy. This method is inherently mesh-independent and readily applicable to the entire domain, including 3D simulations. The proposed methodology is validated against benchmarks with analytical solutions and compared with established methods like skeletonization to demonstrate its accuracy. It is then applied to complex geometries with curvilinear crack paths and evaluated in a three-dimensional simulation.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper claims a mesh-independent crack area correction via twice the gradient energy in phase-field models, but the key assumption about unperturbed gradients needs stronger support from the discrete results.

read the letter

The main point is that the authors propose correcting the crack area overestimation by approximating it as twice the gradient-dependent energy term. This comes from energy equipartition in the small length-scale limit, with the idea that numerical artifacts hit the phase-field energy more than the gradient part. What is new is this energy-based approach that avoids the drawbacks of mesh-dependent factors and skeletonization algorithms. It is straightforward to apply across the entire domain and should work in three dimensions without tracing crack paths. The paper does a decent job laying out the problem and the limitations of prior methods. It validates the idea on benchmarks with analytical solutions and tests it on curvilinear cracks and 3D cases, comparing against skeletonization. The soft spots are around the central assumption and the lack of hard numbers. The claim that the gradient term stays largely unperturbed by discretization effects like element saturation is not obviously true, since saturating the phase-field to 1 over elements will change the gradients inside them too. The paper needs to show why or how much this affects the result. The abstract describes the validation but supplies no error metrics or specific improvements, which makes it hard to assess how well it actually performs. This is for computational materials scientists running phase-field fracture simulations who want more accurate crack area estimates for their predictions. Readers who work on engineering applications would find it worth looking at if the method holds up. It deserves a serious referee because it engages with a real limitation in a common modeling tool using a principle from the model itself. I recommend sending it for peer review, but the referees should focus on verifying the assumption in discrete settings and seeing the quantitative results.

Referee Report

2 major / 1 minor

Summary. The manuscript proposes a correction for crack-area overestimation in phase-field fracture models. It rests on the energy-equipartition principle that the phase-field and gradient energy contributions become equal as the length-scale parameter l approaches zero. The authors argue that numerical artifacts (strain localization, element saturation) primarily perturb only the phase-field term, so the true crack area can be recovered as twice the gradient-dependent energy. The resulting estimator is asserted to be mesh-independent and directly applicable to the full domain, including three-dimensional simulations. Validation against benchmarks with analytical solutions, comparison with skeletonization, and application to curvilinear cracks and 3D geometries are described.

Significance. If the central approximation is rigorously justified, the method would supply a simple, parameter-free, mesh-independent correction that avoids the well-known drawbacks of ad-hoc mesh factors and topology-sensitive skeletonization algorithms. This would be particularly valuable for three-dimensional fracture simulations where skeletonization becomes impractical.

major comments (2)

[Abstract (principle of energy equipartition and numerical-artifact argument)] The central claim that numerical artifacts affect only the phase-field energy term while leaving the gradient term “largely unperturbed” is load-bearing yet unsupported by derivation or quantitative evidence. In any finite-element discretization the phase-field variable saturates to 1 over whole elements; this directly modifies both the (1−ϕ)² contribution and the computed |∇ϕ| inside those elements, so the gradient energy is also perturbed. No analysis is supplied showing that the discretization error in the gradient term remains negligible relative to the phase-field error across the tested meshes, length-scale ratios, or 3D topologies.
[Abstract (validation statement)] The validation paragraph states that the method is “validated against benchmarks with analytical solutions and compared with established methods like skeletonization,” yet supplies no quantitative error metrics, tables of relative crack-area errors, or convergence plots with respect to mesh size or l. Without these data it is impossible to assess whether the proposed estimator actually recovers the analytical crack area more accurately than skeletonization or uncorrected phase-field measures.

minor comments (1)

[Abstract] The abstract refers to “the entire domain, including 3D simulations” without clarifying whether the correction is applied element-wise, integrated over the whole mesh, or restricted to a crack band; a brief clarifying sentence would improve readability.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed comments. We address each major point below and will revise the manuscript to strengthen the presentation and evidence.

read point-by-point responses

Referee: [Abstract (principle of energy equipartition and numerical-artifact argument)] The central claim that numerical artifacts affect only the phase-field energy term while leaving the gradient term “largely unperturbed” is load-bearing yet unsupported by derivation or quantitative evidence. In any finite-element discretization the phase-field variable saturates to 1 over whole elements; this directly modifies both the (1−ϕ)² contribution and the computed |∇ϕ| inside those elements, so the gradient energy is also perturbed. No analysis is supplied showing that the discretization error in the gradient term remains negligible relative to the phase-field error across the tested meshes, length-scale ratios, or 3D topologies.

Authors: We agree that the manuscript would be strengthened by a more explicit analysis of discretization errors in each term. While the numerical benchmarks show the estimator recovers analytical crack areas accurately, we acknowledge the absence of a dedicated quantitative comparison of the separate energy contributions under refinement. In the revised version we will add plots and tables tracking the phase-field and gradient energies individually across mesh sizes, length-scale ratios, and 3D cases to demonstrate the relative perturbation levels. revision: yes
Referee: [Abstract (validation statement)] The validation paragraph states that the method is “validated against benchmarks with analytical solutions and compared with established methods like skeletonization,” yet supplies no quantitative error metrics, tables of relative crack-area errors, or convergence plots with respect to mesh size or l. Without these data it is impossible to assess whether the proposed estimator actually recovers the analytical crack area more accurately than skeletonization or uncorrected phase-field measures.

Authors: The referee correctly notes that the abstract itself contains no numerical error values or plots. We will revise the abstract to include key quantitative results (relative errors, convergence rates) drawn from the benchmark studies and will ensure the results section explicitly presents the corresponding tables and figures for direct comparison with skeletonization and uncorrected measures. revision: yes

Circularity Check

0 steps flagged

No circularity: central claim applies established equipartition principle without reduction to inputs or self-citations

full rationale

The paper grounds its correction on the standard energy-equipartition property of phase-field models (phase-field and gradient contributions become equal as the length scale l approaches zero), then argues that discretization artifacts perturb the phase-field term more than the gradient term, yielding the approximation of crack area as twice the gradient energy. This chain does not reduce any claimed result to a fitted parameter renamed as prediction, a self-defined quantity, or a load-bearing self-citation; the equipartition is invoked as an external mathematical limit rather than derived from the present work's own equations or prior author results. No enumerated circularity pattern is exhibited, and the derivation remains self-contained against the stated assumptions.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the energy-equipartition principle as a domain assumption in phase-field theory; no free parameters or new entities are introduced in the abstract description.

axioms (1)

domain assumption Energy contributions from the phase-field and its gradient are equal as the length-scale parameter approaches zero.
Invoked as the foundational principle that enables the factor-of-two approximation.

pith-pipeline@v0.9.0 · 5526 in / 1081 out tokens · 49200 ms · 2026-05-07T15:49:55.035259+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

21 extracted references · 13 canonical work pages

[1]

Journal of the Mechanics and Physics of Solids , author =

G.A. Francfort and J.-J. Marigo. “Revisiting brittle fracture as an energy minimization problem”. In:Journal of the Mechanics and Physics of Solids46.8 (1998), pp. 1319–1342. doi:https://doi.org/10.1016/S0022-5096(98)00034-9. 30

work page doi:10.1016/s0022-5096(98)00034-9 1998
[2]

Journal of the Mechanics and Physics of Solids 48, 797–826

B. Bourdin, G.A. Francfort, and J.-J. Marigo. “Numerical experiments in revisited brittle fracture”. In:Journal of the Mechanics and Physics of Solids48.4 (2000), pp. 797–826.doi: https://doi.org/10.1016/S0022-5096(99)00028-9

work page doi:10.1016/s0022-5096(99)00028-9 2000
[3]

Miehe, M

C. Miehe, M. Hofacker, and F. Welschinger. “A phase field model for rate-independent crack propagation: robust algorithmic implementation based on operator splits”. In:Computer Methods in Applied Mechanics and Engineering199.45-48 (2010), pp. 2765–2778.doi: https://doi.org/10.1016/j.cma.2010.04.011

work page doi:10.1016/j.cma.2010.04.011 2010
[4]

Miehe, F

C. Miehe, F. Welschinger, and M. Hofacker. “Thermodynamically consistent phase-field models of fracture: Variational principles and multi-field FE implementations”. In:Inter- national Journal for Numerical Methods in Engineering83.10 (2010), pp. 1273–1311.doi: https://doi.org/10.1002/nme.2861. eprint:https://onlinelibrary.wiley.com/doi/ pdf/10.1002/nme.28...

work page doi:10.1002/nme.2861 2010
[5]

Thermodynamically con- sistent linear-gradient damage model in Abaqus

G. Molnár, A. Doitrand, A. Jaccon, B. Prabel, and A. Gravouil. “Thermodynamically con- sistent linear-gradient damage model in Abaqus”. In:Engineering Fracture Mechanics266 (2022), p. 108390.issn: 0013-7944.doi: https://doi.org/10.1016/j.engfracmech.2022. 108390.url: https://www.sciencedirect.com/science/article/pii/S001379442200145X

work page doi:10.1016/j.engfracmech.2022 2022
[6]

A phase-field approach to fatigue analysis: Bridging theory and simulation

M. Castillón, I. Romero, and J. Segurado. “A phase-field approach to fatigue analysis: Bridging theory and simulation”. In:International Journal of Fatigue205 (2026), p. 109397. issn: 0142-1123.doi: https : / / doi . org / 10 . 1016 / j . ijfatigue . 2025 . 109397.url: https://www.sciencedirect.com/science/article/pii/S0142112325005948

2026
[7]

The Variational Approach to Fracture

B. Bourdin, G.A. Francfort, and J.-J. Marigo. “The Variational Approach to Fracture”. In: Journal of Elasticity91.1 (2008), pp. 5–148.issn: 1573-2681.doi:10.1007/s10659-007- 9107-3.url:https://doi.org/10.1007/s10659-007-9107-3

work page doi:10.1007/s10659-007- 2008
[8]

Higher order phase- field modeling of brittle fracture via isogeometric analysis

L. Greco, A. Patton, M. Negri, A. Marengo, U. Perego, and A. Reali. “Higher order phase- field modeling of brittle fracture via isogeometric analysis”. In:Engineering with Computers 40.6 (2024), pp. 3541–3560.doi:https://doi.org/10.1007/s00366-024-01949-5

work page doi:10.1007/s00366-024-01949-5 2024
[9]

A crack-length control technique for phase-field fracture in FFT homogenization

P. Aranda and J. Segurado. “A crack-length control technique for phase-field fracture in FFT homogenization”. In:International Journal for Numerical Methods in Engineering 126.2 (2025), e7664.doi: https : / / doi . org / 10 . 1002 / nme . 7664. eprint: https : / / onlinelibrary.wiley.com/doi/pdf/10.1002/nme.7664.url: https://onlinelibrary. wiley.com/doi/a...

work page doi:10.1002/nme.7664.url: 2025
[10]

A robust monolithic solver for phase-field fracture integrated with fracture energy based arc-length method and under- relaxation

R. Bharali, S. Goswami, C. Anitescu, and T. Rabczuk. “A robust monolithic solver for phase-field fracture integrated with fracture energy based arc-length method and under- relaxation”. In:Computer Methods in Applied Mechanics and Engineering394 (2022), p. 114927.doi:https://doi.org/10.1016/j.cma.2022.114927

work page doi:10.1016/j.cma.2022.114927 2022
[11]

An arc-length control technique for solving quasi-static fracture problems with phase field models and a staggered scheme

J. Zambrano, S. Toro, P.J. Sánchez, F.P. Duda, C.G. Méndez, and A.E. Huespe. “An arc-length control technique for solving quasi-static fracture problems with phase field models and a staggered scheme”. In:Computational Mechanics73.4 (2024), pp. 751–772. doi:https://doi.org/10.1007/s00466-023-02388-7

work page doi:10.1007/s00466-023-02388-7 2024
[12]

T. J. R. Hughes.The finite element method. Englewood Cliffs, New Jersey: Prentice-Hall Inc., 1987.isbn: 0-13-317025-X

1987
[13]

Ern and J

A. Ern and J. L. Guermond.Theory and practice of finite elements. Springer, 2004

2004
[14]

Phase-field modeling of fracture with physics-informed deep learning

M. Manav, R. Molinaro, S. Mishra, and L. De Lorenzis. “Phase-field modeling of fracture with physics-informed deep learning”. In:Computer Methods in Applied Mechanics and Engineering429 (2024), p. 117104.issn: 0045-7825.doi: https://doi.org/10.1016/ j.cma.2024.117104 .url: https://www.sciencedirect.com/science/article/pii/ S0045782524003608. 31

work page arXiv 2024
[15]

A finite element-based adaptive energy response function method for 2D curvilinear progressive fracture

D. Wagner, M.J. Garcia, A. Montoya, and H. Millwater. “A finite element-based adaptive energy response function method for 2D curvilinear progressive fracture”. In:International Journal of Fatigue127 (2019), pp. 229–245.issn: 0142-1123.doi:https://doi.org/10. 1016/j.ijfatigue.2019.05.036 .url: https://www.sciencedirect.com/science/ article/pii/S0142112319302300

2019
[16]

A finite element-based adaptive energy response function method for curvilinear progressive fracture

D. Wagner. “A finite element-based adaptive energy response function method for curvilinear progressive fracture”. PhD thesis. United States – Texas: The University of Texas at San Antonio, 2018, p. 181

2018
[17]

Castillón.Github Repository to be added upon acceptance

M. Castillón.Github Repository to be added upon acceptance
[18]

Castillón.Zenodo DOI to be added upon acceptance

M. Castillón.Zenodo DOI to be added upon acceptance
[19]

PhaseFieldX: an open-source framework for advanced phase-field simulations

M. Castillón. “PhaseFieldX: an open-source framework for advanced phase-field simulations”. In:Journal of Open Source Software10.108 (2025), p. 7307.doi: https://doi.org/10. 21105/joss.07307

2025
[20]

Castillón.PhaseFieldX: An Open-Source Framework for Advanced Phase-Field Simula- tions

M. Castillón.PhaseFieldX: An Open-Source Framework for Advanced Phase-Field Simula- tions. Version v0.2.0. Oct. 2025.doi:10.5281/zenodo.17454862.url: https://doi.org/ 10.5281/zenodo.17454862

work page doi:10.5281/zenodo.17454862.url: 2025
[21]

DOLFINx: the next generation FEniCS problem solving environment

I.A. Baratta, J.P. Dean, J.S. Dokken, M. Habera, J.S. Hale, C.N. Richardson, M.E. Rognes, M.W. Scroggs, N. Sime, and G.N. Wells.DOLFINx: the next generation FEniCS problem solving environment. 2023.doi:https://doi.org/10.5281/zenodo.10447666. 32

work page doi:10.5281/zenodo.10447666 2023