Pith. sign in

REVIEW 4 major objections 6 minor 91 references

Multi-Head Neural Operator for Modelling Interfacial Dynamics

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that a neural operator with time-specific output heads and step-to-step coupling can predict the full time evolution of phase-field PDEs in one forward pass, with lower error and cost than FNO-2d and FNO-3d.

desk verdict The MHNO architecture idea is genuinely useful, but the headline comparison is compromised by an unspecified and likely broken FNO-2d protocol under Approach III. read the letter →

arxiv 2507.17763 v1 pith:WGYDN35C submitted 2025-07-09 physics.comp-ph

Mohammad Sadegh Eshaghi , Navid Valizadeh , Cosmin Anitescu , Yizheng Wang , Xiaoying Zhuang , Timon Rabczuk This is my paper
classification physics.comp-ph
keywords Multi-HeadNeuralOperatorphase-fieldmodelingtime-dependentPDEAllen-CahnequationCahn-HilliardSwift-Hohenbergmolecularbeamepitaxy
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper introduces the Multi-Head Neural Operator (MHNO), an extension of neural operators for time-dependent PDEs. Its central proposal is to replace the single global projection operator at the network's output with one projection network per time step and to couple consecutive time steps with small message-passing networks, so the model emits all future snapshots in one forward pass. The authors argue this removes the error accumulation of step-by-step (FNO-2d-style) rollouts and the parameter inflation of space-time (FNO-3d-style) convolutions. They test MHNO against FNO-2d and FNO-3d on Allen-Cahn, Cahn-Hilliard, Swift-Hohenberg, phase-field crystal, and molecular beam epitaxy equations and report lower L2 errors, fewer parameters, and shorter training times in nearly every configuration. If that result holds, MHNO is a faster and more accurate surrogate for stiff interfacial dynamics over long horizons.

What carries the argument

The load-bearing object is the pair $\{Q_n\}_{n=1}^{N_t}$ and $\{H_n\}_{n=2}^{N_t}$: time-specific projection operators replace the fixed output projection $Q$, and temporal coupling operators implement a message-passing-style recurrence between adjacent outputs. The shared Fourier backend (lift $P$, kernel layers $W_\ell + K_\ell$, activation) is evaluated once; each $Q_n$ maps the terminal latent representation to the field at time $t_n$, and each $H_n$ adds a learned contribution from the previous predicted field, so information flows forward in a way that mirrors causality. This design is what lets a single forward pass cover the full horizon while keeping parameter growth roughly linear rather than multiplicative in $N_t$.

What would settle it

Re-run the Cahn-Hilliard Approach III experiment with FNO-2d under an expanded hyperparameter search or longer training and check whether its test L2 error stays near 90%; if a reasonably tuned FNO-2d closes most of the gap to MHNO's roughly 2.5% error, the superiority claim is not sustained.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that temporal structure in neural operators is best handled at the output, not by unrolling in time or by inflating the input into space-time. MHNO writes the solution map as $G_\theta(x,t_n)(a) = Q_n \circ (W_L + K_L) \circ \cdots \circ \sigma(W_1 + K_1) \circ P(a) + H_n \circ G_\theta(x,t_{n-1})(a)$, where each $Q_n$ is a projection neural network specialized to time step $n$ and each $H_n$ carries the previous predicted field into the current step. Because the shared Fourier layers are evaluated once and the heads are cheap, one forward pass yields all $N_t$ snapshots; because $H_n$ couples adjacent steps, the model has an explicit temporal prior without a 3D tensor. The paper further shows that setting all $H_n$ to zero recovers a standard neural operator, so MHNO inherits the universal-approximation guarantee for neural operators, and derives a bound $\|G_\theta(a;t_n)\|_U \le 4M$ when the output heads have norm at most 1 and the coupling maps have norm below 1. Empirically, on five phase-field benchmarks, MHNO reports the lowest test L2 errors across the three training approaches in most configurations, with particularly large gaps in the full-trajectory regime where FNO-2d sometimes exceeds 70-90% error.

Load-bearing premise

The comparison assumes FNO-2d and FNO-3d were tuned and trained as diligently as MHNO, so the huge baseline errors (about 90% on Cahn-Hilliard, 77% on MBE under Approach III) reflect the methods rather than their training setup; it also assumes the spectral reference solutions are accurate enough to serve as ground truth.

Editorial extensions

If this is right

  • Inference cost for a full trajectory is nearly independent of horizon: one pass over the shared backbone plus cheap per-step heads, instead of $N_t$ sequential network evaluations or one $N_t$-times-larger convolution.
  • Training on full trajectories (Approach III) becomes usable and is reported as the best regime for Allen-Cahn, Cahn-Hilliard, phase-field crystal, and molecular beam epitaxy, whereas FNO-2d degrades sharply in that regime.
  • Parameter count grows only modestly with the number of time steps, since no time dimension is replicated through the whole network.
  • Because standard neural operators are a special case (all $H_n=0$ and $Q_n=Q$), any approximation guarantee for standard neural operators applies to MHNO.
  • The same architecture extends beyond 2D: the paper demonstrates it on 3D Allen-Cahn where the FNO baselines become computationally prohibitive.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural stress test would swap MHNO's message-passing step for a single shared $Q$ head with $N_t$ output channels; if accuracy stays similar, the temporal coupling, not the multiple heads, may not be the source of gains.
  • The architecture is generic enough that it should transfer to other stiff time-dependent PDEs, such as reaction-diffusion systems or turbulent flows, but the paper only demonstrates phase-field benchmarks, so that transfer is unverified.
  • The reported failures of FNO-2d under Approach III are so large (roughly 90% L2 error on Cahn-Hilliard) that they raise the question of whether the baselines were given a comparable tuning budget; a controlled re-run would clarify whether the gap is architectural or operational.
  • One could hybridize MHNO with a physics-informed loss by adding residual PDE terms in the $H_n$ connections, but the paper does not explore this; the architecture's explicit step coupling makes such a test natural.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 6 minor

Summary. The paper proposes the Multi-Head Neural Operator (MHNO), a modification of the Fourier Neural Operator for time-dependent PDEs. Instead of a single global projection operator Q after the final latent representation, MHNO uses time-step-specific projection networks Q_n and explicit temporal coupling networks H_n between consecutive output heads, so that the model outputs all N_t snapshots in one forward pass while retaining sequential dependence. The authors claim an approximation theorem and a stability bound for the architecture, and they benchmark MHNO against FNO-2d and FNO-3d on five phase-field problems: Allen-Cahn, Cahn-Hilliard, Swift-Hohenberg, phase-field crystal, and molecular beam epitaxy, under three temporal training strategies (windowed, two-window, and full-trajectory). The reported results show MHNO generally achieving lower L2 test errors, fewer parameters, and shorter training times than the two baselines.

Significance. If the claims are correct, MHNO is a practically useful contribution: it offers a simple way to extend operator learning to long temporal horizons without the error accumulation of iterated one-step models or the parameter inflation of space-time FNO-3d. The paper has several concrete strengths: it uses an independent Fourier spectral solver to generate ground-truth data, so the benchmarking is not circular; it covers five distinct phase-field equations with different mathematical structure; it compares three training protocols; and it makes code and data publicly available. However, the theoretical support for MHNO is currently incomplete, and the most striking empirical advantage—especially under Approach III—rests on baseline results that are not adequately explained. The central architectural idea is plausible and the evidence base is broad, but the paper needs substantial clarification and correction before the superiority claim can be accepted.

major comments (4)
  1. [Section 3, Eq. (7)] The bound in Eq. (7) is asserted without a proof. The assumptions ||Q_n|| <= 1 and ||H_n|| <= gamma < 1 control only the projection and coupling operators; they do not control the norm of the latent representation v_T, which enters the recurrence G_n = Q_n v_T + H_n G_{n-1}. Without a bound on ||v_T||, or an argument that the approximation error in Eq. (6) can be used to control it, the claimed uniform bound ||G_theta(a;t_n)||_U <= 4M does not follow. The authors should either provide a complete proof or state a corrected theorem with the additional hypothesis needed.
  2. [Section 3, Theorem 1 and Eq. (5)] The containment argument that derives the approximation theorem is not established as written. Setting H_n = O and all Q_n = Q makes the MHNO output N_t identical copies of the same field, which cannot approximate a nontrivial trajectory in the product space U^{N_t}. To inherit the universal approximation theorem from standard neural operators, the authors must specify the target space and norm for the joint map a -> (u_1,...,u_{N_t}) and prove componentwise approximation (for example, by applying [70, Thm. 11] to each time-specific component with its own Q_n). As it stands, the theorem overstates what the containment argument shows.
  3. [Section 4, Approach III; Tables 2 and 5; Appendix B] The protocol for FNO-2d under Approach III is unspecified. Approach III is defined as the direct mapping phi_0 -> (phi_1,...,phi_Nt), but Section 2 defines FNO-2d as a one-step-ahead iterator. The paper never states whether FNO-2d was given N_t output heads, unrolled with teacher forcing, or modified in some other way. The train L2 errors of 90.30 +/- 1.11% for Cahn-Hilliard (Table 2) and 76.12 +/- 10.70% for MBE (Table 5) are signatures of a model that failed to train, and no implementation detail rules out an ill-posed output head or a train/inference mismatch. The authors must specify the exact FNO-2d architecture, loss, teacher-forcing scheme, and inference procedure used for Approach III, or remove these rows from the comparison. Appendix B does not resolve the issue: Table 6 is captioned 'Approach I' while the surrounding text says the table covers Approach III, so the hyperparameter documentation for the crucial setting is internally inconsistent.
  4. [Tables 1-5] The reported parameter counts for FNO-3d are inconsistent with the stated architecture. FNO-3d's final output channel dimension must equal the number of predicted time slices, which differs across approaches (n_w = 10 for Approach I, N_t/2 for Approach II, and N_t = 90 or 100 for Approach III), yet the parameter count is identical across all three approaches for each equation (e.g., 4,197,937 in all rows of Table 1). This is not plausible unless the implementation shares output weights across approaches in a way that is not described. Please report the actual parameter counts or clarify how the three approaches are realized in FNO-3d, since the parameter-efficiency claim depends on these numbers.
minor comments (6)
  1. [Section 3, Eq. (8)] The notation |H_n| and |Q_n| in the definition of R_H^{(n)} is undefined; please specify whether these are operator norms, Frobenius norms, or absolute values of the network parameters, and justify the use of the exponential softmax-style weighting.
  2. [Table 2, Approach II row] The FNO-2d entry '8,00±5,47' uses a decimal comma while all other entries in the paper use decimal points; please unify the formatting.
  3. [Figures 3, 4, 6, 7, 9, 11, 13] Several figure captions and colorbar labels are garbled, including strings like '1-101-10101-101-1010' and the repeated 'Reference Predicted' text in Figure 8; these need to be regenerated or cleaned before publication.
  4. [Section 2-3, notation] The notation in Eq. (4), 'G_theta(x,t_n)(a(x))', mixes the input function and the spatial variable awkwardly; write G_theta(a)(x,t_n) or an equivalent consistently throughout.
  5. [Introduction, Section 1] The statement that FNO-3d has 'no explicit connection between the solution fields at successive time steps' is inaccurate: the 3D convolution kernels in FNO-3d mix neighboring time slices in the hidden layers before the output projection. Please rephrase this motivation to avoid mischaracterizing the baseline.
  6. [Throughout] Minor typos include 'anthiphase' in Table 1's caption, 'Distibution' in Figure 5, 'T emp. Appr.' in the table headers, and 'we explored' in the Introduction; these should be corrected in a final pass.

Circularity Check

0 steps flagged · score 0.0 of 10

No substantive circularity: MHNO's empirical claims rest on external spectral-reference benchmarks, and the approximation theorem is imported from external prior work, so no load-bearing step reduces to its own input.

full rationale

The central claim is that replacing the global projection Q in Eq. (3) with time-specific projections {Q_n} and temporal couplings {H_n} in Eq. (4) lets a single neural operator output the full trajectory in one forward pass. This is an architectural proposal whose validation is external: the reference data are generated by a Fourier-spectral solver (Section 4, Appendix A), and the accuracy comparisons in Tables 1-5 are made against FNO-2d and FNO-3d on those independently generated trajectories. No fitted parameter is renamed as a prediction; the R_H metric of Eq. (8) is a post-hoc diagnostic of trained weights, not a forecast. The approximation theorem is explicitly inherited from Kovachki et al. [70], which is outside the author group, and it is not used as a self-citation chain to forbid alternatives. The paper's self-citations (e.g., [8], [11], [14], [40]-[42], [62]) appear in the literature background and are not load-bearing for the MHNO design or the numerical results. The weaknesses that do exist are non-circular: Appendix B says Table 6 concerns Approach III while the table caption says Approach I, and the paper never specifies how FNO-2d was configured under Approach III, whose reported near-90% train errors are likely a benchmarking or setup artifact. Similarly, the containment argument for Theorem 1 is logically questionable because a restricted subclass with all Q_n equal would emit identical time slices, but it relies on an external theorem and does not make the empirical claim equivalent to its inputs. Overall, the derivation chain is self-contained against independent benchmarks and does not exhibit circularity.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central claim rests on the standard neural operator approximation result, the accuracy of the spectral reference solver, the Gaussian-random-field training distribution, and the unverified norm bounds in Theorem 1. The model's hyperparameters are tuned per problem, and the output-branch count is a design choice that scales with trajectory length. MHNO introduces no new physical entities, particles, forces, or conserved quantities; it is purely an architectural modification of neural operators.

free parameters (2)
  • Per-model and per-approach hyperparameters (learning rate, widths M, W, W_Q, W_H, Fourier layers, modes, batch size… = Examples: Allen-Cahn MHNO best config uses lr=0.001, M=12, W_Q=32, W_H=8, N_l=4, N_Q=2; full grid in Table 7
    Tuned independently for every model, equation, and training approach, and only best results are reported (Appendix B). This selection can inflate measured advantages and is not accompanied by repeated-seed statistics.
  • Number of predicted output heads N_t or window length n_w = 10 (Approach I), 45 (Approach II), 90 or 100 (Approach III)
    MHNO parameter count grows linearly with the number of predicted time steps because each step gets its own Q_n and H_n; this design choice directly affects the efficiency comparison.
assumptions (4)
  • standard math Universal approximation of standard neural operators (Kovachki et al., Theorem 11 and Assumptions 9-10)
    Invoked to prove Theorem 1; the paper inherits the theorem by restricting H_n to zero, so it is an accepted external result rather than a new derivation.
  • domain assumption Spectral reference solutions are accurate enough to serve as ground truth
    All training and test labels come from a Fourier-spectral semi-implicit solver described in Appendix A; no convergence study with respect to solver time step and grid size is reported.
  • domain assumption Gaussian random field initial conditions cover the target application regime
    All problems are trained and tested on i.i.d. draws from the same initial-condition generator; no out-of-distribution or resolution-transfer tests are included.
  • ad hoc to paper Norm bounds ||Q_n|| less than or equal to 1 and ||H_n|| less than gamma with gamma less than 1 in Theorem 1
    These bounds are introduced specifically to derive the 4M stability estimate in Eq. (7), but no construction, verification, or evidence is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multi-Head Neural Operator for Modelling Interfacial Dynamics." pith.science (2026). https://pith.science/paper/WGYDN35C

@misc{pith2026250717763,
  author       = {Pith},
  title        = {Pith review of: Multi-Head Neural Operator for Modelling Interfacial Dynamics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WGYDN35C}},
  note         = {Machine review of arXiv:2507.17763}
}
read the original abstract

Interfacial dynamics underlie a wide range of phenomena, including phase transitions, microstructure coarsening, pattern formation, and thin-film growth, and are typically described by stiff, time-dependent nonlinear partial differential equations (PDEs). Traditional numerical methods, including finite difference, finite element, and spectral techniques, often become computationally prohibitive when dealing with high-dimensional problems or systems with multiple scales. Neural operators (NOs), a class of deep learning models, have emerged as a promising alternative by learning mappings between function spaces and efficiently approximating solution operators. In this work, we introduce the Multi-Head Neural Operator (MHNO), an extended neural operator framework specifically designed to address the temporal challenges associated with solving time-dependent PDEs. Unlike existing neural operators, which either struggle with error accumulation or require substantial computational resources for high-dimensional tensor representations, MHNO employs a novel architecture with time-step-specific projection operators and explicit temporal connections inspired by message-passing mechanisms. This design allows MHNO to predict all time steps after a single forward pass, while effectively capturing long-term dependencies and avoiding parameter overgrowth. We apply MHNO to solve various phase field equations, including antiphase boundary motion, spinodal decomposition, pattern formation, atomic scale modeling, and molecular beam epitaxy growth model, and compare its performance with existing NO-based methods. Our results show that MHNO achieves superior accuracy, scalability, and efficiency, demonstrating its potential as a next-generation computational tool for phase field modeling. The code and data supporting this work is publicly available at https://github.com/eshaghi-ms/MHNO.

Figures

Figures reproduced from arXiv: 2507.17763 by the authors.

Figure 1
Figure 1. Architectures of Multi Head Neural Operator. Start from input a, 1. Lift to higher dimension channel space by neural network P, 2. Apply L layers Fourier operator and activation function, 3. Convert to n branches by neural networks {Qn}, 4. Finalize by adding neural networks {Hn}. 6 [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Hn contributions over time, showing the evolving ratio R (n) H that highlights the dynamic interplay between the two terms, Hn and Qn. The periodic pattern indicates their changing importance throughout the time steps. The plot starts from the first time step at the top left of the figure and progresses to the 100th time step at the bottom right. 8 [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Phase-field approximation of the antiphase boundary motion from different initial conditions. The [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figures from the paper (11 more)
Figure 4
Figure 4. Figure 4: Phase Field Predictions and Errors for 2D Allen-Cahn Equation. Visualization of the ref￾erence, predicted, and error fields for the 2D Allen-Cahn equation over time steps (∆t, 5∆t, 10∆t, . . . , 90∆t). (a) Results for a sample with an error close to the mean L 2 error …
Figure 5
Figure 5. Figure 5: L 2 Error Distribution for 2D Allen-Cahn Equation. (a) L 2 error distribution across different samples, showing variability in model accuracy. (b) L 2 error distribution over time steps, highlighting temporal evolution of prediction errors. To further analyze the perfo…
Figure 6
Figure 6. Figure 6: Phase Field Predictions and Errors for the 3D Allen-Cahn Equation. Visualization of the reference, predicted, and error fields for the 3D Allen-Cahn equation over time steps (5∆t, 10∆t, . . . , 100∆t) for a sample with an error close to the mean L 2 error of the test d…
Figure 7
Figure 7. Figure 7: Phase Field Predictions and Errors for Cahn-Hilliard Equation. Visualization of the reference, predicted, and error fields for the Cahn-Hilliard equation over time steps (∆t, 5∆t, 10∆t, . . . , 90∆t). (a) Results for a sample with an error close to the mean L 2 error o…
Figure 8
Figure 8. Figure 8: The Reference and MHNO predictions for 10 random samples from the dataset for spinodal [PITH_FULL_IMAGE:figures/full_fig_p018_8.png]
Figure 9
Figure 9. Figure 9: Phase Field Predictions and Errors for Swift–Hohenberg Equation. Visualiza￾tion of the reference, predicted, and error fields for the Swift–Hohenberg equation over time steps (∆t, 5∆t, 10∆t, . . . , 90∆t). (a) Results for a sample with an error close to the mean L 2 er…
Figure 10
Figure 10. Figure 10: Performance comparison of FNO2D, FNO3D, and MHNO on the pattern formation problem. (a) Average L 2 error over time steps, computed across test samples. MHNO demonstrates the lowest error throughout the trajectory. (b–d) Training loss convergence for Approaches I (wind…
Figure 11
Figure 11. Figure 11: Phase Field Predictions and Errors for 2D Phase Field Crystal Equation. Visual￾ization of the reference, predicted, and error fields for the Phase Field Crystal equation over time steps (∆t, 5∆t, 10∆t, . . . , 90∆t). (a) Results for a sample with an error close to the…
Figure 12
Figure 12. Figure 12: Box plots of L 2 error over the first ten and last ten time steps for the PFC equation, evaluated across different test samples. Each row corresponds to one of the three training approaches, with comparisons between FNO2D, FNO3D, and MHNO. All methods show some increa…
Figure 13
Figure 13. Figure 13: Phase Field Predictions and Errors for 2D Molecular Beam Epitaxy Equation. Vi￾sualization of the reference, predicted, and error fields for the Molecular Beam Epitaxy equation over time steps (∆t, 5∆t, 10∆t, . . . , 90∆t). (a) Results for a sample with an error close …
Figure 14
Figure 14. Figure 14: Comparison of L2 error statistics over time steps for different methods and different approaches [PITH_FULL_IMAGE:figures/full_fig_p027_14.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

91 extracted references · 31 canonical work pages

  1. [1]

    Pore-scale modelling and sensitivity analyses of hydrogen- brine multiphase flow in geological porous media

    L. Hashemi, M. Blunt, and H. Hajibeygi. “Pore-scale modelling and sensitivity analyses of hydrogen- brine multiphase flow in geological porous media”. In:Scientific reports11.1 (2021), p. 8348.doi: 10.1038/s41598-021-87490-7

    work page doi:10.1038/s41598-021-87490-7 2021

  2. [2]

    An introduction to phase-field modeling of microstructure evolution

    N. Moelans, B. Blanpain, and P. Wollants. “An introduction to phase-field modeling of microstructure evolution”. In:Calphad32.2 (2008), pp. 268–294.doi:10.1016/j.calphad.2007.11.003

    doi:10.1016/j.calphad.2007.11.003 2008

  3. [3]

    Phase-field models in materials science

    I. Steinbach. “Phase-field models in materials science”. In:Modelling and simulation in materials science and engineering17.7 (2009), p. 073001.doi:10.1088/0965-0393/17/7/073001

    doi:10.1088/0965-0393/17/7/073001 2009

  4. [4]

    Interfacial dynamics with soluble surfactants: A phase-field two-phase flow model with variable densities

    G. Zhu and A. Li. “Interfacial dynamics with soluble surfactants: A phase-field two-phase flow model with variable densities”. In:Advances in Geo-Energy Research4.1 (2020), pp. 86–98.doi:10.26804/ ager.2020.01.08. 31

    2020

  5. [5]

    Quantitative phase-field modeling of solute trapping in rapid solidifi- cation

    S. Kavousi and M. A. Zaeem. “Quantitative phase-field modeling of solute trapping in rapid solidifi- cation”. In:Acta Materialia205 (2021), p. 116562.doi:10.1016/j.actamat.2020.116562

    arXiv 2021

  6. [6]

    Phase field modeling and computer implementation: A review

    X. Zhuang et al. “Phase field modeling and computer implementation: A review”. In:Engineering Fracture Mechanics262 (2022), p. 108234.doi:10.1016/j.engfracmech.2021.107881

    arXiv 2022

  7. [7]

    Phase field simulation of liquid phase separation with fluid flow

    G. Tegze, T. Pusztai, and L. Gr´ an´ asy. “Phase field simulation of liquid phase separation with fluid flow”. In:Materials Science and Engineering: A413 (2005), pp. 418–422.doi:10.1016/j.msea.2005. 09.045

    work page doi:10.1016/j.msea.2005 2005

  8. [8]

    Isogeometric analysis of hydrodynamics of vesicles using a mono- lithic phase-field approach

    N. Valizadeh and T. Rabczuk. “Isogeometric analysis of hydrodynamics of vesicles using a mono- lithic phase-field approach”. In:Computer Methods in Applied Mechanics and Engineering388 (2022), p. 114191.doi:10.1016/j.cma.2021.114191

    work page arXiv 2022

Show all 91 references
  1. [9]

    Phase-field Navier–Stokes model for vesicle doublets hydro- dynamics in incompressible fluid flow

    M. Ashour, N. Valizadeh, and T. Rabczuk. “Phase-field Navier–Stokes model for vesicle doublets hydro- dynamics in incompressible fluid flow”. In:Computer Methods in Applied Mechanics and Engineering 412 (2023), p. 116063.doi:10.1016/j.cma.2023.116063

    2023

  2. [10]

    Hydrodynamics of multicomponent vesicles: A phase-field approach

    Z. Wen et al. “Hydrodynamics of multicomponent vesicles: A phase-field approach”. In:Computer Methods in Applied Mechanics and Engineering432 (2024), p. 117390.doi:10.1016/j.cma.2024. 117390

    2024 doi

  3. [11]

    A monolithic finite element method for phase-field mod- eling of fully Eulerian fluid–structure interaction

    N. Valizadeh, X. Zhuang, and T. Rabczuk. “A monolithic finite element method for phase-field mod- eling of fully Eulerian fluid–structure interaction”. In:Computer Methods in Applied Mechanics and Engineering435 (2025), p. 117618.doi:10.1016/j.cma.2024.117618

    2025

  4. [12]

    Phase-field models for microstructure evolution

    L.-Q. Chen. “Phase-field models for microstructure evolution”. In:Annual review of materials research 32.1 (2002), pp. 113–140.doi:10.1146/annurev.matsci.32.112001.132041

    2002

  5. [13]

    Interface-tracking and interface-capturing techniques for finite element computation of moving boundaries and interfaces

    T. E. Tezduyar. “Interface-tracking and interface-capturing techniques for finite element computation of moving boundaries and interfaces”. In:Computer Methods in Applied Mechanics and Engineering 195.23-24 (2006), pp. 2983–3000.doi:10.1016/j.cma.2004.09.018

    2006 doi

  6. [14]

    Isogeometric analysis for phase-field models of geometric PDEs and high-order PDEs on stationary and evolving surfaces

    N. Valizadeh and T. Rabczuk. “Isogeometric analysis for phase-field models of geometric PDEs and high-order PDEs on stationary and evolving surfaces”. In:Computer Methods in Applied Mechanics and Engineering351 (2019), pp. 599–642.doi:10.1016/j.cma.2019.03.043

    2019 doi

  7. [15]

    Accelerating phase-field-based mi- crostructure evolution predictions via surrogate models trained by machine learning methods

    D. Montes de Oca Zapiain, J. A. Stewart, and R. Dingreville. “Accelerating phase-field-based mi- crostructure evolution predictions via surrogate models trained by machine learning methods”. In:npj Computational Materials7.1 (2021), p. 3.doi:10.1038/s41524-020-00471-8

    2021 doi

  8. [16]

    A spectral element-based phase field method for incompressible two-phase flows

    Y. Xiao et al. “A spectral element-based phase field method for incompressible two-phase flows”. In: Physics of Fluids34.2 (2022).doi:10.1063/5.0077372. 32

    2022 doi

  9. [17]

    A comparative review of XFEM, mixed FEM and phase-field models for quasi- brittle cracking

    M. Cervera et al. “A comparative review of XFEM, mixed FEM and phase-field models for quasi- brittle cracking”. In:Archives of Computational Methods in Engineering29.2 (2022), pp. 1009–1083. doi:10.1007/s11831-021-09604-8

    2022 doi

  10. [18]

    Learning two-phase microstructure evolution using neural operators and autoen- coder architectures

    V. Oommen et al. “Learning two-phase microstructure evolution using neural operators and autoen- coder architectures”. In:npj Computational Materials8.1 (2022), p. 190.doi:10.1038/s41524-022- 00876-7

    2022 doi

  11. [19]

    R. C. Smith.Uncertainty quantification: theory, implementation, and applications. SIAM, 2024.doi: 10.1137/1.9781611973228

    2024 doi

  12. [20]

    Non-linear model reduction for uncertainty quantification in large-scale inverse problems

    D. Galbally et al. “Non-linear model reduction for uncertainty quantification in large-scale inverse problems”. In:International journal for numerical methods in engineering81.12 (2010), pp. 1581– 1608.doi:10.1002/nme.2746

    2010 doi

  13. [21]

    A dual mesh method with adaptivity for stress-constrained topology optimization

    D. A. White, Y. Choi, and J. Kudo. “A dual mesh method with adaptivity for stress-constrained topology optimization”. In:Structural and Multidisciplinary Optimization61.2 (2020), pp. 749–762. doi:10.1007/s00158-019-02393-6

    2020 doi

  14. [22]

    Accelerating Phase Field Simulations Through a Hybrid Adaptive Fourier Neural Operator with U-Net Backbone

    C. Bonneville et al. “Accelerating Phase Field Simulations Through a Hybrid Adaptive Fourier Neural Operator with U-Net Backbone”. In:arXiv preprint arXiv:2406.17119(2024).doi:arXiv : 2406 . 17119v2

    2024 arXiv

  15. [23]

    An efficient algorithm for 3D adaptive meshing

    P. George, H. Borouchaki, and P. Laug. “An efficient algorithm for 3D adaptive meshing”. In:Advances in Engineering Software33.7-10 (2002), pp. 377–387.doi:10.1016/S0965-9978(02)00065-0

    2002 doi

  16. [24]

    Machine-learning-based data-driven discovery of nonlinear phase-field dynamics

    E. Kiyani et al. “Machine-learning-based data-driven discovery of nonlinear phase-field dynamics”. In: Physical Review E106.6 (2022), p. 065303.doi:10.1103/PhysRevE.106.065303

    2022 doi

  17. [25]

    Bayesian deep convolutional encoder-decoder networks for surrogate modeling and uncertainty quantification

    Y. Zhu and N. Zabaras. “Bayesian deep convolutional encoder-decoder networks for surrogate modeling and uncertainty quantification”. In:Journal of Computational Physics366 (2018), pp. 415–447.doi: 10.1016/j.jcp.2018.04.018

    2018 doi

  18. [26]

    Deep convolutional encoder-decoder networks for uncertainty quantification of dynamic multiphase flow in heterogeneous media

    S. Mo et al. “Deep convolutional encoder-decoder networks for uncertainty quantification of dynamic multiphase flow in heterogeneous media”. In:Water Resources Research55.1 (2019), pp. 703–728.doi: 10.1029/2018WR023528

    2019 doi

  19. [27]

    Predicting CO2 plume migration in heterogeneous formations using conditional deep convolutional generative adversarial network

    Z. Zhong, A. Y. Sun, and H. Jeong. “Predicting CO2 plume migration in heterogeneous formations using conditional deep convolutional generative adversarial network”. In:Water Resources Research 55.7 (2019), pp. 5830–5851.doi:10.1029/2018WR024592. 33

    2019 doi

  20. [28]

    A deep-learning-based surrogate model for data assimilation in dynamic subsurface flow problems

    M. Tang, Y. Liu, and L. J. Durlofsky. “A deep-learning-based surrogate model for data assimilation in dynamic subsurface flow problems”. In:Journal of Computational Physics413 (2020), p. 109456.doi: 10.1016/j.jcp.2020.109456

    2020

  21. [29]

    Multiphase flow prediction with deep neural networks

    G. Wen, M. Tang, and S. M. Benson. “Multiphase flow prediction with deep neural networks”. In: arXiv preprint arXiv:1910.09657(2019).doi:10.48550/arXiv.1910.09657

    2019 doi arXiv

  22. [30]

    CCSNet: a deep learning modeling suite for CO2 storage

    G. Wen, C. Hay, and S. M. Benson. “CCSNet: a deep learning modeling suite for CO2 storage”. In: Advances in Water Resources155 (2021), p. 104009.doi:10.1016/j.advwatres.2021.104009

    2021

  23. [31]

    Accelerating phase-field predictions via recurrent neural net- works learning the microstructure evolution in latent space

    C. Hu, S. Martin, and R. Dingreville. “Accelerating phase-field predictions via recurrent neural net- works learning the microstructure evolution in latent space”. In:Computer Methods in Applied Me- chanics and Engineering397 (2022), p. 115128.doi:10.1016/j.cma.2022.115128

    2022

  24. [32]

    Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders

    K. Lee and K. T. Carlberg. “Model reduction of dynamical systems on nonlinear manifolds using deep convolutional autoencoders”. In:Journal of Computational Physics404 (2020), p. 108973.doi: 10.1016/j.jcp.2019.108973

    2020

  25. [33]

    Gplasdi: Gaussian process-based interpretable latent space dynamics identification through deep autoencoder

    C. Bonneville et al. “Gplasdi: Gaussian process-based interpretable latent space dynamics identification through deep autoencoder”. In:Computer Methods in Applied Mechanics and Engineering418 (2024), p. 116535.doi:10.1016/j.cma.2023.116535

    2024

  26. [34]

    A Comprehensive Review of Latent Space Dynamics Identification Algorithms for Intrusive and Non-Intrusive Reduced-Order-Modeling

    C. Bonneville et al. “A Comprehensive Review of Latent Space Dynamics Identification Algorithms for Intrusive and Non-Intrusive Reduced-Order-Modeling”. In:arXiv preprint arXiv:2403.10748(2024). doi:10.48550/arXiv.2403.10748

    2024 doi arXiv

  27. [35]

    Domain adaptation based transfer learning approach for solving PDEs on complex geometries

    A. Chakraborty et al. “Domain adaptation based transfer learning approach for solving PDEs on complex geometries”. In:Engineering with Computers38.5 (2022), pp. 4569–4588.doi:10 . 1007 / s00366-022-01661-2

    2022

  28. [36]

    Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations

    M. Raissi, P. Perdikaris, and G. E. Karniadakis. “Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations”. In:Journal of Computational physics378 (2019), pp. 686–707.doi:10.1016/j....

    2019 doi

  29. [37]

    Physics-constrained deep learning for high-dimensional surrogate modeling and uncer- tainty quantification without labeled data

    Y. Zhu et al. “Physics-constrained deep learning for high-dimensional surrogate modeling and uncer- tainty quantification without labeled data”. In:Journal of Computational Physics394 (2019), pp. 56– 81.doi:10.1016/j.jcp.2019.05.024

    2019 doi

  30. [38]

    SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks

    E. Haghighat and R. Juanes. “SciANN: A Keras/TensorFlow wrapper for scientific computations and physics-informed deep learning using artificial neural networks”. In:Computer Methods in Applied Mechanics and Engineering373 (2021), p. 113552.doi:10.1016/j.cma.2020.113552. 34

    2021

  31. [39]

    Variational energy based XPINNs for phase field analysis in brittle fracture

    A. Chakraborty et al. “Variational energy based XPINNs for phase field analysis in brittle fracture”. In:arXiv preprint arXiv:2207.02307(2022).doi:10.48550/arXiv.2207.02307

    2022 doi arXiv

  32. [40]

    Adaptive fourth-order phase field analysis using deep energy minimization

    S. Goswami, C. Anitescu, and T. Rabczuk. “Adaptive fourth-order phase field analysis using deep energy minimization”. In:Theoretical and Applied Fracture Mechanics107 (2020), p. 102527.doi: 10.1016/j.tafmec.2020.102527

    2020

  33. [41]

    Transfer learning enhanced physics informed neural network for phase-field modeling of fracture

    S. Goswami et al. “Transfer learning enhanced physics informed neural network for phase-field modeling of fracture”. In:Theoretical and Applied Fracture Mechanics106 (2020), p. 102447.doi:10.1016/j. tafmec.2019.102447

    2020

  34. [42]

    Deepnetbeam: A Framework for the Analysis of Functionally Graded Porous Beams

    M. S. Es-haghi et al. “Deepnetbeam: A Framework for the Analysis of Functionally Graded Porous Beams”. In:Available at SSRN 4846935().doi:10.2139/ssrn.4846935

    doi

  35. [43]

    Deep residual U-net convolution neural networks with autore- gressive strategy for fluid flow predictions in large-scale geosystems

    Z. Jiang, P. Tahmasebi, and Z. Mao. “Deep residual U-net convolution neural networks with autore- gressive strategy for fluid flow predictions in large-scale geosystems”. In:Advances in Water Resources 150 (2021), p. 103878.doi:10.1016/j.advwatres.2021.103878

    2021

  36. [44]

    Deep-learning-based surrogate flow modeling and geologi- cal parameterization for data assimilation in 3D subsurface flow

    M. Tang, Y. Liu, and L. J. Durlofsky. “Deep-learning-based surrogate flow modeling and geologi- cal parameterization for data assimilation in 3D subsurface flow”. In:Computer Methods in Applied Mechanics and Engineering376 (2021), p. 113636.doi:10.1016/j.cma.2020.113636

    2021

  37. [45]

    Physics-constrained deep learning for data assimilation of subsurface transport

    H. Wu and R. Qiao. “Physics-constrained deep learning for data assimilation of subsurface transport”. In:Energy and AI3 (2021), p. 100044.doi:10.1016/j.egyai.2020.100044

    2021

  38. [46]

    A deep learning method for the dynamics of classic and conservative Allen-Cahn equa- tions based on fully-discrete operators

    Y. Geng et al. “A deep learning method for the dynamics of classic and conservative Allen-Cahn equa- tions based on fully-discrete operators”. In:Journal of Computational Physics496 (2024), p. 112589. doi:10.1016/j.jcp.2023.112589

    2024

  39. [47]

    Rethinking materials simulations: Blending direct numerical simulations with neural operators

    V. Oommen et al. “Rethinking materials simulations: Blending direct numerical simulations with neural operators”. In:npj Computational Materials10.1 (2024), p. 145.doi:10.1038/s41524-024-01319-1

    2024 doi

  40. [48]

    U-FNO—An enhanced Fourier neural operator-based deep-learning model for multi- phase flow

    G. Wen et al. “U-FNO—An enhanced Fourier neural operator-based deep-learning model for multi- phase flow”. In:Advances in Water Resources163 (2022), p. 104180.doi:10.1016/j.advwatres. 2022.104180

    2022

  41. [49]

    Simulating fluid flow in complex porous materials by integrating the governing equations with deep-layered machines

    S. Kamrava, M. Sahimi, and P. Tahmasebi. “Simulating fluid flow in complex porous materials by integrating the governing equations with deep-layered machines”. In:NPJ Computational Materials 7.1 (2021), p. 127.doi:10.1038/s41524-021-00598-2

    2021 doi

  42. [50]

    A physics-informed and hierarchically regularized data-driven model for predicting fluid flow through porous media

    K. Wang et al. “A physics-informed and hierarchically regularized data-driven model for predicting fluid flow through porous media”. In:Journal of Computational Physics443 (2021), p. 110526.doi: 10.1016/j.jcp.2021.110526. 35

    2021

  43. [51]

    Physics-Informed Neural Networks for 2nd order ODEs with sharp gradients

    M. De Florio et al. “Physics-Informed Neural Networks for 2nd order ODEs with sharp gradients”. In:Journal of Computational and Applied Mathematics436 (2024), p. 115396.doi:10.1016/j.cam. 2023.115396

    2024

  44. [52]

    Characterizing possible failure modes in physics-informed neural networks

    A. Krishnapriyan et al. “Characterizing possible failure modes in physics-informed neural networks”. In:Advances in neural information processing systems34 (2021), pp. 26548–26560.doi:10.48550/ arXiv.2109.01050

    2021 doi arXiv

  45. [53]

    Physics-informed neural networks with residual/gradient-based adaptive sam- pling methods for solving partial differential equations with sharp solutions

    Z. Mao and X. Meng. “Physics-informed neural networks with residual/gradient-based adaptive sam- pling methods for solving partial differential equations with sharp solutions”. In:Applied Mathematics and Mechanics44.7 (2023), pp. 1069–1084.doi:10.1007/s10483-023-2994-7

    2023 doi

  46. [54]

    Nonlocal kernel network (NKN): A stable and resolution-independent deep neural network

    H. You et al. “Nonlocal kernel network (NKN): A stable and resolution-independent deep neural network”. In:Journal of Computational Physics469 (2022), p. 111536.doi:10.1016/j.jcp.2022. 111536

    2022 doi

  47. [55]

    Solving Allen-Cahn and Cahn-Hilliard equations using the adaptive physics informed neural networks

    C. L. Wight and J. Zhao. “Solving Allen-Cahn and Cahn-Hilliard equations using the adaptive physics informed neural networks”. In:arXiv preprint arXiv:2007.04542(2020).doi:10.48550/arXiv.2007. 04542

    2020 doi arXiv

  48. [56]

    Physics based deep learning for nonlinear two-phase flow in porous media

    O. Fuks and H. Tchelepi. “Physics based deep learning for nonlinear two-phase flow in porous media”. In:ECMOR XVII. Vol. 2020. 1. European Association of Geoscientists & Engineers. 2020, pp. 1–10. doi:10.3997/2214-4609.202035147

    2020

  49. [57]

    Prediction of porous media fluid flow using physics informed neural networks

    M. M. Almajid and M. O. Abu-Al-Saud. “Prediction of porous media fluid flow using physics informed neural networks”. In:Journal of Petroleum Science and Engineering208 (2022), p. 109205.doi:10. 1016/j.petrol.2021.109205

    2022

  50. [58]

    Physics informed deep learning for flow and transport in porous media

    C. G. Fraces and H. Tchelepi. “Physics informed deep learning for flow and transport in porous media”. In:SPE Reservoir Simulation Conference?SPE. 2021, D011S006R002.doi:10.2118/203934-MS

    2021 doi

  51. [59]

    Model reduction and neural networks for parametric PDEs

    K. Bhattacharya et al. “Model reduction and neural networks for parametric PDEs”. In:The SMAI journal of computational mathematics7 (2021), pp. 121–157.doi:10.5802/smai-jcm.74

    2021 doi

  52. [60]

    Neural operator: Graph kernel network for partial differential equations

    Z. Li et al. “Neural operator: Graph kernel network for partial differential equations”. In:arXiv preprint arXiv:2003.03485(2020).doi:10.48550/arXiv.2003.03485

    2020 doi arXiv

  53. [61]

    Learning the solution operator of parametric partial differential equations with physics-informed DeepONets

    S. Wang, H. Wang, and P. Perdikaris. “Learning the solution operator of parametric partial differential equations with physics-informed DeepONets”. In:Science advances7.40 (2021), eabi8605.doi:10. 1126/sciadv.abi8605

    2021

  54. [62]

    Variational Physics-informed Neural Operator (VINO) for Solving Partial Dif- ferential Equations

    M. S. Eshaghi et al. “Variational Physics-informed Neural Operator (VINO) for Solving Partial Dif- ferential Equations”. In:arXiv preprint arXiv:2411.06587(2024).doi:10.48550/arXiv.2411.06587. 36

    2024 doi arXiv

  55. [63]

    Fourier neural operator for parametric partial differential equations

    Z. Li et al. “Fourier neural operator for parametric partial differential equations”. In:arXiv preprint arXiv:2010.08895(2020).doi:10.48550/arXiv.2010.08895

    2020 doi arXiv

  56. [64]

    Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators

    L. Lu et al. “Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators”. In:Nature machine intelligence3.3 (2021), pp. 218–229.doi:10.1038/s42256- 021- 00302-5

    2021 doi

  57. [65]

    LNO: Laplace neural operator for solving differential equations

    Q. Cao, S. Goswami, and G. E. Karniadakis. “LNO: Laplace neural operator for solving differential equations”. In:arXiv preprint arXiv:2303.10528(2023).doi:10.1038/s42256-024-00844-4

    2023 arXiv

  58. [66]

    Convolutional neural operators for robust and accurate learning of PDEs

    B. Raonic et al. “Convolutional neural operators for robust and accurate learning of PDEs”. In:Ad- vances in Neural Information Processing Systems36 (2024).doi:10.48550/arXiv.2302.01178

    2024 doi arXiv

  59. [67]

    Solving parametric PDE problems with artificial neural networks

    Y. Khoo, J. Lu, and L. Ying. “Solving parametric PDE problems with artificial neural networks”. In: European Journal of Applied Mathematics32.3 (2021), pp. 421–435.doi:10.1017/S0956792520000182

    2021 doi

  60. [68]

    POD-DL-ROM: Enhancing deep learning-based reduced order models for nonlinear parametrized PDEs by proper orthogonal decomposition

    S. Fresca and A. Manzoni. “POD-DL-ROM: Enhancing deep learning-based reduced order models for nonlinear parametrized PDEs by proper orthogonal decomposition”. In:Computer Methods in Applied Mechanics and Engineering388 (2022), p. 114181.doi:10.1016/j.cma.2021.114181

    2022

  61. [69]

    Phase-Field DeepONet: Physics-informed deep operator neural network for fast simulations of pattern formation governed by gradient flows of free-energy functionals

    W. Li, M. Z. Bazant, and J. Zhu. “Phase-Field DeepONet: Physics-informed deep operator neural network for fast simulations of pattern formation governed by gradient flows of free-energy functionals”. In:Computer Methods in Applied Mechanics and Engineering(2023), p. 116299.doi...

    2023

  62. [70]

    Neural operator: Learning maps between function spaces with applications to pdes

    N. Kovachki et al. “Neural operator: Learning maps between function spaces with applications to pdes”. In:Journal of Machine Learning Research24.89 (2023), pp. 1–97

    2023

  63. [71]

    Fourier-spectral method for the phase-field equations

    S. Yoon et al. “Fourier-spectral method for the phase-field equations”. In:Mathematics8.8 (2020), p. 1385.doi:10.3390/math8081385

    2020 doi

  64. [72]

    Energy-dissipative evolutionary deep operator neural networks

    J. Zhang et al. “Energy-dissipative evolutionary deep operator neural networks”. In:Journal of Com- putational Physics498 (2024), p. 112638.doi:10.1016/j.jcp.2023.112638

    2024

  65. [73]

    Mean curvature flow and low energy solutions of the parabolic Allen-Cahn equation on the three-sphere

    J. Chen and P. Gaspar. “Mean curvature flow and low energy solutions of the parabolic Allen-Cahn equation on the three-sphere”. In:The Journal of Geometric Analysis33.9 (2023), p. 283.doi:10. 1007/s12220-023-01347-1

    2023

  66. [74]

    A second-order maximum bound principle preserving operator splitting method for the Allen–Cahn equation with applications in multi-phase systems

    X. Xiao and X. Feng. “A second-order maximum bound principle preserving operator splitting method for the Allen–Cahn equation with applications in multi-phase systems”. In:Mathematics and Comput- ers in Simulation202 (2022), pp. 36–58.doi:10.1016/j.matcom.2022.05.024. 37

    2022 doi

  67. [75]

    A wavelet-Laplace variational technique for image decon- volution and inpainting

    J. A. Dobrosotskaya and A. L. Bertozzi. “A wavelet-Laplace variational technique for image decon- volution and inpainting”. In:IEEE Transactions on Image Processing17.5 (2008), pp. 657–663.doi: 10.1109/TIP.2008.919367

    2008

  68. [76]

    A microscopic theory for antiphase boundary motion and its application to antiphase domain coarsening

    S. M. Allen and J. W. Cahn. “A microscopic theory for antiphase boundary motion and its application to antiphase domain coarsening”. In:Acta metallurgica27.6 (1979), pp. 1085–1095.doi:10.1016/ 0001-6160(79)90196-2

    1979

  69. [77]

    Free energy of a nonuniform system. I. Interfacial free energy

    J. W. Cahn and J. E. Hilliard. “Free energy of a nonuniform system. I. Interfacial free energy”. In: The Journal of chemical physics28.2 (1958), pp. 258–267.doi:10.1063/1.1744102

    1958 doi

  70. [78]

    Numerical simulation of a binary alloy of 2D Cahn–Hilliard model for phase sepa- ration

    R. Abazari et al. “Numerical simulation of a binary alloy of 2D Cahn–Hilliard model for phase sepa- ration”. In:Computational and Applied Mathematics41.8 (2022), p. 389.doi:10.1007/s40314-022- 02109-5

    2022 doi

  71. [79]

    Continuum-scale modelling of polymer blends using the Cahn–Hilliard equa- tion: transport and thermodynamics

    P. K. Inguva et al. “Continuum-scale modelling of polymer blends using the Cahn–Hilliard equa- tion: transport and thermodynamics”. In:Soft matter17.23 (2021), pp. 5645–5665.doi:10.1039/ D1SM00272D

    2021

  72. [80]

    Two-dimensional Cahn–Hilliard simulations for coarsening kinetics of spinodal decomposition in binary mixtures

    B. K¨ onig, O. J. Ronsin, and J. Harting. “Two-dimensional Cahn–Hilliard simulations for coarsening kinetics of spinodal decomposition in binary mixtures”. In:Physical Chemistry Chemical Physics23.43 (2021), pp. 24823–24833.doi:10.1039/D1CP03229A

    2021 doi

  73. [81]

    Turbulent superstructures in Rayleigh-B´ enard convec- tion

    A. Pandey, J. D. Scheel, and J. Schumacher. “Turbulent superstructures in Rayleigh-B´ enard convec- tion”. In:Nature communications9.1 (2018), p. 2118.doi:10.1038/s41467-018-04478-0

    2018 doi

  74. [82]

    Hyperbolic chaos of Turing patterns

    P. V. Kuptsov, S. P. Kuznetsov, and A. Pikovsky. “Hyperbolic chaos of Turing patterns”. In:Physical Review Letters108.19 (2012), p. 194101.doi:10.1103/PhysRevLett.108.194101

    2012 doi

  75. [83]

    Hydrodynamic fluctuations at the convective instability

    J. Swift and P. C. Hohenberg. “Hydrodynamic fluctuations at the convective instability”. In:Physical Review A15.1 (1977), p. 319.doi:10.1103/PhysRevA.15.319

    1977 doi

  76. [84]

    Nucleation and bulk crystallization in binary phase field theory

    L. Gr´ an´ asy, T. B¨ orzs¨ onyi, and T. Pusztai. “Nucleation and bulk crystallization in binary phase field theory”. In:Physical review letters88.20 (2002), p. 206105.doi:10.1103/PhysRevLett.88.206105

    2002 doi

  77. [85]

    Phase-field crystal modeling and classical density functional theory of freezing

    K. R. Elder et al. “Phase-field crystal modeling and classical density functional theory of freezing”. In: Physical Review B—Condensed Matter and Materials Physics75.6 (2007), p. 064107.doi:10.1103/ PhysRevB.75.064107

    2007

  78. [86]

    A phase field crystal theory of the kinematics of dislocation lines

    V. Skogvoll et al. “A phase field crystal theory of the kinematics of dislocation lines”. In:Journal of the Mechanics and Physics of Solids166 (2022), p. 104932.doi:10.1016/j.jmps.2022.104932

    2022

  79. [87]

    Modeling elasticity in crystal growth

    K. Elder et al. “Modeling elasticity in crystal growth”. In:Physical review letters88.24 (2002), p. 245701.doi:10.1103/PhysRevLett.88.245701. 38

    2002 doi

  80. [88]

    Semiconductor molecular-beam epitaxy at low temperatures

    D. Eaglesham. “Semiconductor molecular-beam epitaxy at low temperatures”. In:Journal of applied physics77.8 (1995), pp. 3597–3617.doi:10.1063/1.358597

    1995 doi

  81. [89]

    Selective area epitaxy of GaN nanostructures: MBE growth and morphological analysis

    S. M. Hasan et al. “Selective area epitaxy of GaN nanostructures: MBE growth and morphological analysis”. In:Crystal Growth & Design23.6 (2023), pp. 4098–4104.doi:10.1021/acs.cgd.2c01506

    2023 doi

  82. [90]

    An adaptive time-stepping strategy for the molecular beam epitaxy models

    Z. Qiao, Z. Zhang, and T. Tang. “An adaptive time-stepping strategy for the molecular beam epitaxy models”. In:SIAM Journal on Scientific Computing33.3 (2011), pp. 1395–1414.doi:10 . 1137 / 100812781

    2011

  83. [91]

    Regularized linear schemes for the molecular beam epitaxy model with slope selection

    L. Chen, J. Zhao, and X. Yang. “Regularized linear schemes for the molecular beam epitaxy model with slope selection”. In:Applied Numerical Mathematics128 (2018), pp. 139–156.doi:10.1016/j. apnum.2018.02.004. 39

    2018 doi

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.