Pith. sign in

REVIEW 4 major objections 5 minor 73 references

Bifurcation curve detection with deflation for multiparametric PDEs

T0 review · 4 major / 5 minor · reviewed 2026-08-02 · deepseek-v4-flash

Pith's one-line read A hybrid deflation–arclength method reconstructs complete bifurcation diagrams and traces bifurcation curves and surfaces for multiparametric PDEs, without spectral analysis.

desk verdict A useful zigzag extension of deflated continuation for multiparametric bifurcation detection, with real benchmark checks but a few addressable flaws; worth refereeing, not accepting as-is. read the letter →

arxiv 2602.12940 v2 pith:DF22FYTO submitted 2026-02-13 math.NA cs.NA

classification math.NAcs.NA MSC 35B3265H1765N35
keywords bifurcationdetectionmultiparametricPDEsdeflationarclengthcontinuationcurvessurfacesBratuequationAllen-Cahn
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

This paper aims to establish a general numerical methodology for multiparametric PDEs that both constructs complete bifurcation diagrams and detects the curves (two parameters) or surfaces (three parameters) where solution multiplicity changes. The proposed deflated arclength continuation algorithm combines adaptive arclength continuation along user-prescribed parameter paths with deflation, which modifies the residual so Newton's method avoids previously found solutions and discovers coexisting branches. A zigzag path-following strategy then crosses the bifurcation boundary back and forth, reconstructing the critical curve or surface. The authors show the approach reproduces known bifurcation values for the Bratu and Allen–Cahn equations in 1D/2D and recovers both linear and nonlinear bifurcation curves, arguing that no comparable general methodology currently exists. If correct, this gives engineers and scientists a way to map stability boundaries in multi-parameter models without expensive eigenvalue computations.

What carries the argument

The load-bearing object is the deflated arclength continuation step combined with the zigzag path. Arclength continuation treats the parameter λ as an unknown and steps along the arclength of the solution branch, with step size adaptively chosen by the solver; in multiparametric settings the step follows a prescribed path g(λ)=0. Deflation multiplies the residual G(u,λ) by an operator M(u,u*) = (||u−u*||^{-p}+α)I that blows up near previously found solutions, steering Newton away from them and toward new coexisting solutions. The zigzag detector changes the continuation direction through an angle θ whenever deflation indicates crossing the boundary between unique and multiple-solution region

What would settle it

Take a one-parameter or two-parameter PDE with an analytically known bifurcation curve and a deliberately hidden disconnected branch that deflation cannot reach (e.g., a solution branch not connected to the trivial path and with a tiny basin of attraction). Run the zigzag algorithm and compare the detected transition points to the true boundary; if deflation misses the hidden branch inside the true multiplicity region, the zigzag curve will deviate from the analytic one, showing that the detected boundary depends on deflation's reach rather than on the system's actual solution count.

Watch

Extended reading notes

Core claim

The central claim is that a continuation framework can explore p-dimensional parameter spaces by tracing a smooth curve g(λ) in the parameter plane and, at each arclength step, using deflation to enumerate all coexisting solutions. Branch switching then requires no eigenvector computation. The zigzag strategy classifies a parameter point as inside or outside the multiple-solution region solely from deflation outcomes: when deflation produces additional solutions, the path has entered the bifurcating region; when it stops, the path has left it. Repeating this across parameter space yields the bifurcation curve or surface. The paper validates this on two benchmarks — the Bratu equation (saddle

Load-bearing premise

The zigzag detector treats deflation as a complete oracle: it infers that a parameter point lies in the uniqueness region solely from deflation's failure to find additional solutions, and infers boundary crossings from that binary outcome; a missed branch or a divergence caused by a near-singular Jacobian would shift every detected crossing point.

Editorial extensions

If this is right

  • For the Bratu equation, the method reproduces the known saddle-node values λ₁*≈3.5 (1D) and λ₁*≈6.8 (2D), and the detected bifurcation curves match across different continuation paths.
  • For the Allen–Cahn equation, the method recovers the pitchfork cascade λ₁*≈1,4,9 in 1D and ≈2,5,10 in 2D, including a multiple pitchfork with three emerging branches at the third point.
  • The zigzag strategy detects a nonlinear bifurcation curve for a modified Allen–Cahn problem with quadratic diffusion, demonstrating applicability beyond linear boundaries.
  • For p=3, the method produces bifurcation surfaces for both benchmarks, separating uniqueness from multiplicity regions in the three-parameter space.
  • The framework requires no spectral analysis: branch switching and boundary detection rely solely on deflation outcomes and arclength continuation.

Reading between the lines

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

  • If deflation is a complete oracle for coexisting solutions, the zigzag method implicitly computes the boundary of the solution-multiplicity set; this suggests a formal connection to degree-theoretic or topological-count methods that could be made rigorous.
  • The angle θ is currently a fixed hyperparameter; an adaptive schedule based on the local curvature of the detected boundary should reduce cost and improve accuracy on sharply curved bifurcation surfaces.
  • The method's success on path-independent reconstructions suggests it could be combined with reduced-order models to produce cheap surrogates of the stability boundary for engineering design, though this extension is not explored in the paper.
  • A natural stress test is a PDE with analytically known bifurcation locus, e.g., a one-dimensional problem with a prescribed nonlinear coefficient, to measure the zigzag's error as a function of θ and step size.
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 / 5 minor

Summary. The paper proposes a numerical strategy for constructing multiparametric bifurcation diagrams and detecting bifurcation curves/surfaces in nonlinear PDEs. The method combines arclength continuation (generalized to parameter spaces of dimension p≥2 via prescribed paths g(λ)=0) with deflation, which is used to discover multiple coexisting solutions at each parameter sample. A 'zigzag' path-following strategy is introduced: the algorithm marches along a horizontal path until deflation detects multiple solutions, then changes direction along a slanted line, crossing the bifurcation curve back and forth, thereby tracing the boundary between uniqueness and multiplicity. The method is demonstrated on the Bratu equation (saddle-node bifurcations) and the Allen–Cahn equation (pitchfork bifurcations) in one and two spatial dimensions, including a modified Allen–Cahn problem with a nonlinear bifurcation curve, and on two- and three-dimensional parameter spaces. The reported bifurcation values (Bratu λ1*≈3.5/6.8; Allen–Cahn sequences ≈1,4,9 for d=1 and ≈2,5,10 for d=2) agree with classical linearized-eigenvalue predictions, providing independent ground truth for the examples.

Significance. If the framework is as robust as claimed, it would be a convenient, black-box-type tool for exploring multiparametric bifurcation structures without performing spectral analysis or explicit eigenvector computations. The numerical validation against known analytical values is a genuine strength, as are the breadth of test problems (saddle-node, pitchfork, nonlinear bifurcation curves, p=3 surfaces). The method is built from standard, well-understood components (pseudo-arclength continuation and deflation), which lowers the risk of fundamental algorithmic errors. However, the central novelty claim—that no general methodology exists for both diagram construction and curve detection in this multiparametric setting—is overstated in light of existing continuation packages (e.g., pde2path, MATCONT), and the robustness of the zigzag detector rests on an unproven completeness assumption about deflation. The present results are promising and point to a useful practical tool, but the 'robustly tracks' claim needs additional support.

major comments (4)
  1. [§4.2.3, §3.2] The zigzag detector classifies a parameter point as being in the multiple-solution region iff deflation discovers additional coexisting solutions, and it infers crossing of the bifurcation curve from a change in this outcome. However, §3.2 only states that deflation 'possibly' discovers n distinct solutions, and Algorithm 2 terminates when Newton's method diverges. Divergence is not a certificate of nonexistence, especially because §2 notes that Newton's method 'may perform poorly and diverge in these regions due to near-singular jacobians.' A missed branch or a divergence near a near-singular point therefore shifts the detected crossing point, and since the zigzag path changes direction exactly at these crossings, the entire detected curve/surface is affected. The manuscript gives no completeness proof, no sensitivity analysis with respect to the deflation parameters (α,p) or the zigzag
  2. [§4.2.1, Eq. (16)] The unit-tangent normalization for p≥2 is written as ∥˙x∥² = ∥˙u∥² + ∑_{i=1}^{p} ˙λ_i = 1, omitting the squares on the parameter components. The correct condition should be ∥˙u∥² + ∑ |˙λ_i|² = 1, as given in the p=2 case in Eq. (11). If implemented literally, the tangent vector would not have unit norm, invalidating the hyperplane constraint in the extended system (17). Please correct the equation and confirm that the numerical implementation uses the squared norm.
  3. [§5.2.2, §5.1.2] There is a reproducibility contradiction in the Allen–Cahn settings. §5.1.2 defines the parameter range as λ3 ∈ [π,3.8], where λ3 is the length of the domain [0,λ3]^d. Yet §5.2.2 fixes λ3=1, which lies outside that range, and the reported bifurcation values λ1*≈1,4,9 (d=1) and ≈2,5,10 (d=2) are consistent with λ3=1 (or equivalently λ3=π for d=1, but not both). Additionally, §5.2.4 refers to 'Equation (20)' (the Bratu equation) when describing the Allen–Cahn problem, which should be Eq. (21). These inconsistencies must be fixed to make the experiments reproducible and to clarify the parameter ranges used.
  4. [§1, Abstract] The abstract as provided at the top of the manuscript promises 'three benchmark problems of increasing complexity' including '2D/3D Rayleigh–Benard convection,' but the body's abstract and Section 5 present only the Bratu and Allen–Cahn benchmarks. The Rayleigh–Bénard problem does not appear anywhere in the results. Moreover, the novelty claim in §1—'there is currently no general methodology in the literature capable of both constructing bifurcation diagrams and detecting bifurcation curves in such multiparametric context'—is too strong. Existing tools such as pde2path (cited in §3) and MATCONT can track fold and branch-point curves in two-parameter systems, albeit with different techniques. The abstract should be reconciled with the actual content, and the novelty claim should be qualified to avoid overstatement.
minor comments (5)
  1. [§4.2.1, Eq. (14)] The expression for |˙λ1| contains malformed absolute-value bars and parentheses; please clean up the typesetting.
  2. [Algorithm 3] The set notation in lines 5, 10, and 11 is hard to parse (e.g., `{u^k_i}_{k<j}`, `{{u^k_{i+1}}_{k<r}, u^j_{i+1}}`). Rewrite with explicit index sets or prose to make the branch-update logic clear.
  3. [§5.3.1] For the modified Allen–Cahn example, the text states that the diffusion coefficient is ρ(λ2)=-(λ2-1)^2+3, but does not explicitly derive the resulting bifurcation curve (λ1* = ρ(λ2)·(π/2?)²). Stating this formula would make the test more transparent.
  4. [Figures 7–15] Several figures combine solution profiles and bifurcation diagrams in a single panel without subfigure labels, making it difficult to distinguish the left and right plots (e.g., Figure 7). Please add (a)/(b) labels or separate panels.
  5. [§4.2.3] The description of the crossing back from the multiplicity region P2 to the uniqueness region P1 is vague. Specify the stopping criterion used to detect the second crossing and state what data are recorded for the bifurcation curve at each crossing.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: central outputs are validated against independent bifurcation values and known linear theory.

full rationale

The paper's derivation chain is not circular. Deflated arclength continuation combines two established ingredients: arclength continuation (external refs) and deflation (Farrell et al. [22], external). The zigzag detector infers crossing of the bifurcation curve from deflation success/failure, but this is an operational surrogate, not a definition of the bifurcation curve; the detected values are checked against independent ground truth (§5.2.2: Allen–Cahn λ1*≈1,4,9 for d=1 and ≈2,5,10 for d=2; §5.1.1: Bratu fold λ1*≈3.5/6.8), so the method is anchored externally rather than fitting its own inputs. The custom nonlinear-curve example (§5.3.1) chooses a quadratic diffusion coefficient ρ(λ2) a priori and then recovers the PDE's bifurcation locus from the deflated zigzag runs; this is a constructed benchmark with a known family of critical values, not a fitted parameter renamed as a prediction. The unproven completeness of deflation (possible missed branches or divergence near singular Jacobians, noted in §3.2 and §2) is a correctness/robustness limitation of the oracle used by the zigzag detector, not a circularity: the paper does not define the bifurcation curve as the deflation-success boundary, and the numerical results on standard benchmarks match known theory. Authors cite several of their own papers (e.g., [25],[50],[55],[57],[58],[59],[67]), but none supplies a load-bearing uniqueness theorem or unverified premise; the deflation convergence theorem cited is the external Deflated Rall–Rheinboldt theorem from [22]. No reduction of the claimed result to its own inputs was found.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The framework contributes methodology, not new physics, so no invented entities. The central claim rests on standard continuation theory plus several hand-set hyperparameters (θ, k, ds, and the unreported α, p) and on two unproven heuristics: that deflation finds all coexisting solutions and that divergence of deflated Newton signals the uniqueness boundary. The benchmarks' ground truth comes from classical linear-stability theory, which is standard math. Discretization fidelity of the wavelet collocation at N=32/64 is asserted rather than verified by a convergence study.

free parameters (5)
  • deflation shift parameter α = not reported
    Hand-set; §3.2 says the choice is 'essential' for stability/convergence, but §5 never states the value used, so results cannot be reproduced or tested for sensitivity.
  • deflation power parameter p = not reported
    Hand-set; controls how strongly previously found solutions are penalized (Eq. 7). Never reported in the numerical section.
  • zigzag angle θ = π/20 (Bratu); π − π/20 (Allen–Cahn)
    User-set hyperparameter of Eq. (19); §4.2.3 concedes large θ can skip parts of the bifurcation curve, and no adaptive or sensitivity analysis is given.
  • zigzag persistence steps k = 5
    Number of continuation steps taken before rotating the path (§5.3.1); chosen without justification or sensitivity study.
  • base arclength step ds = 0.2 (Bratu); 0.01 (Allen–Cahn)
    Chosen per benchmark; the claimed adaptivity is inside the arclength predictor–corrector, but all reported results use these fixed base steps.
assumptions (5)
  • standard math Implicit function theorem / smooth branch assumption: solutions of G(u,λ)=0 form smooth curves parameterizable by arclength except at isolated bifurcation points.
    Invoked in §2 ('existence and uniqueness ... guaranteed by the implicit function theorem') and used throughout §3.1 to justify tangent-based continuation.
  • domain assumption The prescribed continuation path g(λ)=0 is smooth, transverse to the bifurcation boundary (g_{λ₂}≠0 in Eq. 13), and free of sharp turns.
    The multiparametric tangent formula divides by g_{λ₂} (Eq. 13) and the Remark in §4.2.2 concedes sharp directional changes can break convergence; this restricts admissible paths.
  • ad hoc to paper Deflation completeness: deflated Newton iterations find all coexisting solutions, and their divergence marks the boundary of the multiple-solution region.
    Load-bearing for the zigzag curve detector (§4.2.3), which classifies P1/P2 regions purely from deflation outcomes; §3.2 hedges ('possibly having discovered n distinct solutions'), and no completeness proof is given.
  • domain assumption Wavelet collocation at fixed N=32/64 faithfully reproduces the continuous bifurcation structure (no spurious or missed bifurcations).
    All experiments use one fixed discretization level (§5); the paper asserts discretization-independence of the framework but never demonstrates convergence of detected curves with respect to N.
  • standard math Classical linear-stability/eigenvalue theory provides the benchmark bifurcation values used as ground truth.
    The reported Allen–Cahn bifurcations ≈1,4,9 (d=1) and ≈2,5,10 (d=2) and the linear Bratu curves match λ₁* = λ₂(nπ/λ₃)² / eigenvalue sums; used in §5 to judge the method's output.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Bifurcation curve detection with deflation for multiparametric PDEs." pith.science (2026). https://pith.science/paper/DF22FYTO

@misc{pith2026260212940,
  author       = {Pith},
  title        = {Pith review of: Bifurcation curve detection with deflation for multiparametric PDEs},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DF22FYTO}},
  note         = {Machine review of arXiv:2602.12940}
}
abstract

This work presents a comprehensive framework for capturing bifurcating phenomena and detecting bifurcation curves in nonlinear multiparametric partial differential equations, where the system exhibits multiple coexisting solutions for given values of the parameters. Traditional continuation methods for one-dimensional parameterizations employ the previously computed solution as the initial guess for the next parameter value. These are usually very inefficient, since small step sizes increase computational cost, while larger steps could jeopardize the method convergence jumping to a different solution branch or missing the bifurcation point. To address these challenges, we propose a novel framework that combines: (i) arclength continuation, adaptively selecting new parameter values in higher dimension, and (ii) the deflation technique, discovering multiple branches to construct complete bifurcation diagrams without requiring a costly spectral analysis of the system. In particular, the arclength continuation method is designed to handle multiparametric scenarios, where the parameter vector $\lambda \in \mathbb{R}^p$ traces a curve $g(\lambda)$ within a $p$-dimensional parameter space. In addition, we introduce a zigzag path-following strategy to robustly track the bifurcation curves and surfaces, respectively, for two- and three-dimensional parametric spaces. Finally, we demonstrate its performance on three benchmark problems of increasing complexity: from the 1D/2D Bratu and Allen--Cahn equations to the 2D/3D Rayleigh--Benard convection problem.

Figures

Figures reproduced from arXiv: 2602.12940 by the authors.

Figure 1
Figure 1. depicting the case of pitchfork and saddle node bifurcations, Figures 1a and 1b, respectively with q denoting the quantity of interest used to generate the bifurcation diagram. λ q U 1 U 2 U 3 (a) Pitchfork bifurcation λ q U 1 U 2 (b) Saddle node bifurcation [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Visualization of the arclength continuation method. The solution of the extended system in Equation (6) with the reparametrization given by s requires the computation of ( ˙u, λ˙). From Equation (3), ˙u can be obtained as u˙ = −G −1 u Gλλ. ˙ Moreover, since λ˙ is a scalar quantity, it can be obtained from Equation (4) as follows ∥G −1 u Gλλ˙ ∥ 2 + |λ˙ | 2 = 1 ⇒ |λ˙ | = (1 + ∥G −1 u Gλ∥ 2 ) − 1 2 . The final step is … view at source ↗
Figure 3
Figure 3. Sketch of pitchfork bifurcation discovery. λ q λi (a) Multiple solutions λ q λi (b) Arclength continuation λ q λi (c) Deflation strategy [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figures from the paper (15 more)
Figure 4
Figure 4. Figure 4: Sketch of saddle-node bifurcation discovery. Algorithm 3 Deflated arclength continuation algorithm Input: Initial guess u 1 0 , arclength step ds, maximum value λmax, nonlinear operator G(u, λ) Output: Solution branches {Uj} n j=1 1: u 1 1 ← Solve Equation (2) with ini…
Figure 5
Figure 5. Figure 5: Paths for the multiparametric arclength continuation with a = 1, b = 10, c = 1, and d = 5. this setting, continuation is performed in the extended (u,λ)-space, where the parameter vector λ evolves along the chosen path. In this way, deflated arclength continuation prov…
Figure 6
Figure 6. Figure 6: Bifurcation curve and the approximating zigzag tracking path (red) on the (λ1, λ2)-domain with p = 2 with the zoom-in showing the local bifurcation diagram. path may skip portions of the bifurcation curve, particularly in regions where the curve exhibits sharp turns or…
Figure 7
Figure 7. Figure 7: Lower and upper Bratu solutions with d = 1 for λ1 = 3.24, and saddle￾node bifurcation diagram. 0.002 0.004 0.006 0.008 0.010 2 4 6 8 0 1 2 3 4 5 6 7 1 0 2 4 6 8 ||u|| Upper branch Lower branch [PITH_FULL_IMAGE:figures/full_fig_p017_7.png]
Figure 8
Figure 8. Figure 8: Lower and upper Bratu solutions with d = 2 for λ1 = 0.14, and saddle￾node bifurcation diagram. The bifurcation behavior of the Bratu equation possesses the same qualitative properties between the one- and two-dimensional settings, with some notable differences. In one …
Figure 9
Figure 9. Figure 9: Bifurcating Allen–Cahn solutions with d = 1 for λ1 = 13, and pitchfork bifurcation diagram. remark that, due to the intrinsic symmetry of the Allen–Cahn equation, the lower solution branches consist of symmetric solutions, which appear in pairs related by sign symmetry…
Figure 10
Figure 10. Figure 10: Five solutions corresponding to five upper branches of the two￾dimensional Allen–Cahn equation at λ1 = 12, and pitchfork bifurcation diagram. 0 4 8 12 16 24 20 32 28 36 1 12 3 4 5 6 7 8 9 10 2 0 1 2 3 4 5 6 7 8 ||u|| (a) Horizontal lines 0 4 2 8 6 12 10 16 14 20 18 1 …
Figure 11
Figure 11. Figure 11: Bifurcation diagrams for Bratu with different paths and d = 1. the (λ1, λ2)-plane, where the saddle-node bifurcation curves are clearly visible. This further demon￾strates the robustness of the proposed technique in exploring complex solution landscapes even in the mu…
Figure 12
Figure 12. Figure 12: Bifurcation diagrams for Bratu with different paths and d = 2. 0 5 10 15 20 25 30 35 1 2 4 6 8 10 2 (a) d = 1 0 5 10 15 20 25 30 35 1 1 2 3 4 5 6 7 2 (b) d = 2 [PITH_FULL_IMAGE:figures/full_fig_p020_12.png]
Figure 13
Figure 13. Figure 13: Paths in the parameter space with Bratu bifurcation curve. pitchfork phenomena, with the methodology being able to identify multiple bifurcation curves from which several coexisting solutions originate. Exploiting the same paths described before, we explore the parame…
Figure 14
Figure 14. Figure 14: Bifurcation diagrams for Allen–Cahn with different paths and d = 1. 0 2 4 6 8 10 12 14 16 1 1 2 3 4 5 6 7 8 2 4 3 2 1 0 1 2 3 4 S(u) [PITH_FULL_IMAGE:figures/full_fig_p021_14.png]
Figure 15
Figure 15. Figure 15: Bifurcation diagrams for Allen–Cahn with different paths and d = 2. Allen–Cahn problems defined before as illustrative examples for p = 2 and p = 3, highlighting the effectiveness of the proposed approach in identifying transition regions in multiparametric systems. 5…
Figure 16
Figure 16. Figure 16: Bifurcation curve detection for different benchmarks and spatial di￾mension, i.e. d = 1 (red) and d = 2 (blue). given as the quadratic term in the parameter ρ(λ2) = −(λ2 − 1)2 + 3, and the boundary conditions are modified as ux(0) = 0 and u(1) = 0 for d = 1. In this w…
Figure 17
Figure 17. Figure 17: Bifurcation surface for Bratu problem with p = 3 and different spatial dimension, d = 1 and d = 2, left and right respectively. 2 4 6 8 10 1 1 2 3 4 5 6 7 8 9 10 2 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3 [PITH_FULL_IMAGE:figures/full_fig_p023_17.png]
Figure 18
Figure 18. Figure 18: Bifurcation surface for Allen–Cahn problem with p = 3 and different spatial dimension, d = 1 and d = 2, left and right respectively. sharp changes occur, and larger steps in smoother regions, thus making the method more robust and reliable. Moreover, we extended the a…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

73 extracted references

  1. [1]

    S. M. Allen and J. W. Cahn. A microscopic theory for antiphase boundary motion and its application to antiphase domain coarsening.Acta Metallurgica, 27(6):1085–1095, 1979

  2. [2]

    S. M. Allen and J. W. Cahn. Phase diagram features associated with multicritical points in alloy systems.Bulletin of Alloy Phase Diagrams, 3(3):287–295, 1982

  3. [3]

    E. L. Allgower and K. Georg.Introduction to numerical continuation methods. SIAM, 2003

  4. [4]

    E. L. Allower and K. Georg. Detection of bifurcation points along a curve. InNumerical Continuation Methods: An Introduction, pages 75–90. Springer, 1990

  5. [5]

    Bauer and E

    L. Bauer and E. L. Reiss. Nonlinear buckling of rectangular plates.Journal of the Society for Industrial and Applied Mathematics, 13(3):603–626, 1965

  6. [6]

    M. S. Berger. On Von K´ arm´ an’s equations and the buckling of a thin elastic plate, i the clamped plate.Commu- nications on Pure and Applied Mathematics, 20(4):687–719, 1967

  7. [7]

    Boull´ e, V

    N. Boull´ e, V. Dallas, and P. E. Farrell. Bifurcation analysis of two-dimensional Rayleigh-B´ enard convection using deflation.Physical Review E, 105(5):055106, 2022

  8. [8]

    J. P. Boyd. An analytical and numerical study of the two-dimensional Bratu equation.Journal of Scientific Computing, 1(2):183–206, 1986

Show all 73 references
  1. [9]

    Caloz and J

    G. Caloz and J. Rappaz. Numerical analysis for nonlinear and bifurcation problems.Handbook of Numerical Analysis, 5:487–637, 1997

  2. [10]

    T. F. Chan. Newton-like pseudo-arclength methods for computing simple turning points.SIAM Journal on Scientific and Statistical Computing, 5(1):135–148, 1984

  3. [11]

    T. F. C. Chan and H. B. Keller. Arc-length continuation and multigrid techniques for nonlinear elliptic eigenvalue problems.SIAM Journal on Scientific and Statistical Computing, 3(2):173–194, 1982

  4. [12]

    E. G. Charalampidis, P. G. Kevrekidis, and P. E. Farrell. Computing stationary solutions of the two-dimensional Gross–Pitaevskii equation with deflated continuation.Communications in Nonlinear Science and Numerical Simulation, 54:482–499, 2018

  5. [13]

    Y. Choi, T. Ha, J. Han, S. Kim, and D. S. Lee. Turing instability and dynamic bifurcation for the one-dimensional Gray–Scott model.Studies in Applied Mathematics, 154(1):e12786, 2025

  6. [14]

    S. N. Chow and J. K. Hale.Methods of bifurcation theory, volume 251. Springer Science & Business Media, 2012

  7. [15]

    P. G. Ciarlet.Linear and nonlinear functional analysis with applications. SIAM, 2025

  8. [16]

    J. A. Dahlke and R. A. Bettinger. Practical implementation of pseudo-arclength continuation to ensure consistent path direction.Acta Astronautica, 215:205–216, 2024

  9. [17]

    E. A. de Souza Neto and Y. T. Feng. On the determination of the path direction for arc-length methods in the presence of bifurcations and snap-backs’.Computer Methods in Applied Mechanics and Engineering, 179(1- 2):81–89, 1999

  10. [18]

    N. Deng, B. R. Noack, M. Morzy´ nski, and L. R. Pastur. Low-order model for successive bifurcations of the fluidic pinball.Journal of Fluid Mechanics, 884:A37, 2020

  11. [19]

    Dhooge, W

    A. Dhooge, W. Govaerts, and Y. A. Kuznetsov. MATCONT: a MATLAB package for numerical bifurcation analysis of ODEs.ACM Transactions on Mathematical Software (TOMS), 29(2):141–164, 2003

  12. [20]

    K. I. Dickson, C. T. Kelley, I. C. F. Ipsen, and I. G. Kevrekidis. Condition estimates for pseudo-arclength continuation.SIAM Journal on Numerical Analysis, 45(1):263–276, 2007

  13. [21]

    Fabiani, F

    G. Fabiani, F. Calabr` o, L. Russo, and C. Siettos. Numerical solution and bifurcation analysis of nonlinear partial differential equations with extreme learning machines.Journal of Scientific Computing, 89(2):44, 2021

  14. [22]

    P. E. Farrell, A. Birkisson, and S. W. Funke. Deflation techniques for finding distinct solutions of nonlinear partial differential equations.SIAM Journal on Scientific Computing, 37(4):A2026–A2045, 2015

  15. [23]

    Feng and A

    X. Feng and A. Prohl. Numerical analysis of the Allen-Cahn equation and approximation for mean curvature flows.Numerische Mathematik, 94:33–65, 2003

  16. [24]

    Gobat, A

    G. Gobat, A. Opreni, S. Fresca, A. Manzoni, and A. Frangi. Reduced order modeling of nonlinear microstructures through proper orthogonal decomposition.Mechanical Systems and Signal Processing, 171:108864, 2022. DETECTING BIFURCATIONS FOR MULTIPARAMETRIC PDES 25

  17. [25]

    I. C. Gonnella, M. Khamlich, F. Pichi, and G. Rozza. A stochastic perturbation approach to nonlinear bifurcating problems. (arXiv:2402.16803), 2024

  18. [26]

    Guevel, T

    Y. Guevel, T. Allain, G. Girault, and J. M. Cadou. Numerical bifurcation analysis for 3-dimensional sudden expansion fluid dynamic problem.International Journal for Numerical Methods in Fluids, 87(1):1–26, 2018

  19. [27]

    Hao and C

    W. Hao and C. Zheng. An adaptive homotopy method for computing bifurcations of nonlinear parametric systems. Journal of Scientific Computing, 82(3):53, 2020

  20. [28]

    Hao and C

    W. Hao and C. Zheng. Learn bifurcations of nonlinear parametric systems via equation-driven neural networks. Chaos: An Interdisciplinary Journal of Nonlinear Science, 32(1), 2022

  21. [29]

    M. W. Hess, A. Quaini, and G. Rozza. Reduced basis model order reduction for Navier–Stokes equations in domains with walls of varying curvature.International Journal of Computational Fluid Dynamics, 34(2):119– 126, 2020

  22. [30]

    V. R. Hosseini, A. A. Mehrizi, A. Gungor, and H. H. Afrouzi. Application of a physics-informed neural network to solve the steady-state Bratu equation arising from solid biofuel combustion theory.Fuel, 332:125908, 2023

  23. [31]

    J. L. Hueso, E. Mart ´ ınez, and J. R. Torregrosa. Modified Newton’s method for systems of nonlinear equations with singular Jacobian.Journal of Computational and Applied Mathematics, 224(1):77–83, 2009

  24. [32]

    H. B. Keller. Numerical solution of bifurcation and nonlinear eigenvalue problems.Applications of Bifurcation Theory., 1977

  25. [33]

    C. T. Kelley.Solving nonlinear equations with Newton’s method. SIAM, 2003

  26. [34]

    Khamlich, F

    M. Khamlich, F. Pichi, and G. Rozza. Model order reduction for bifurcating phenomena in Fluid-Structure Interaction problems.International Journal for Numerical Methods in Fluids, 94(10):1611–1640, 2022

  27. [35]

    Kielh¨ ofer.Bifurcation theory

    H. Kielh¨ ofer.Bifurcation theory. Springer, 2012

  28. [36]

    D. A. Knoll and D. E. Keyes. Jacobian-free Newton–Krylov methods: a survey of approaches and applications. Journal of Computational Physics, 193(2):357–397, 2004

  29. [37]

    C. Kuehn. Numerical continuation and SPDE stability for the 2d cubic-quintic Allen–Cahn equation.SIAM/ASA Journal on Uncertainty Quantification, 3(1):762–789, 2015

  30. [38]

    Kuehn, C

    C. Kuehn, C. Piazzola, and E. Ullmann. Uncertainty quantification analysis of bifurcations of the Allen–Cahn equation with random coefficients.Physica D: Nonlinear Phenomena, 470:134390, 2024

  31. [39]

    Kumar and M

    N. Kumar and M. Mehra. Generalized fractional-order legendre wavelet method for two dimensional distributed order fractional optimal control problem.Journal of Vibration and Control, 30(7-8):1690–1705, 2024

  32. [40]

    Y. A. Kuznetsov.Elements of applied bifurcation theory. Springer, 1998

  33. [41]

    X. Li. Numerical solution of fractional differential equations using cubic b-spline wavelet collocation method. Communications in Nonlinear Science and Numerical Simulation, 17(10):3934–3946, 2012

  34. [42]

    S. Lopez. Detection of bifurcation points along a curve traced by a continuation method.International Journal for Numerical Methods in Engineering, 53(4):983–1004, 2002

  35. [43]

    Mehra.Wavelets theory and its applications

    M. Mehra.Wavelets theory and its applications. Springer, 2018

  36. [44]

    H. D. Mittelmann. A pseudo-arclength continuation method for nonlinear eigenvalue problems.SIAM Journal on Numerical Analysis, 23(5):1007–1016, 1986

  37. [45]

    C. J. Naudet and A. E. Lindsay. Numerical bifurcation analysis of post-contact states in mathematical models of Micro-Electromechanical systems.Mathematics and Computers in Simulation, 220:296–308, 2024

  38. [46]

    R. D. Nussbaum. A global bifurcation theorem with applications to functional differential equations.Journal of Functional Analysis, 19(4):319–338, 1975

  39. [47]

    R. J. Olsen and I. R. Epstein. Bifurcation analysis of chemical reaction mechanisms. I. steady state bifurcation structure.The Journal of Chemical Physics, 94(4):3083–3095, 1991

  40. [48]

    Opreni, A

    A. Opreni, A. Vizzaccaro, A. Frangi, and C. Touz´ e. Model order reduction based on direct normal form: appli- cation to large finite element MEMS structures featuring internal resonance.Nonlinear Dynamics, 105(2):1237– 1272, 2021

  41. [49]

    J. M. Ortega and W. C. Rheinboldt.Iterative solution of nonlinear equations in several variables. SIAM, 2000

  42. [50]

    Pichi, F

    F. Pichi, F. Ballarin, G. Rozza, and J. S. Hesthaven. An artificial neural network approach to bifurcating phe- nomena in computational fluid dynamics.Computers & Fluids, 254:105813, 2023

  43. [51]

    Pichi, B

    F. Pichi, B. Moya, and J. S. Hesthaven. A graph convolutional autoencoder approach to model order reduction for parametrized PDEs.Journal of Computational Physics, 501:112762, 2024

  44. [52]

    Pichi, A

    F. Pichi, A. Quaini, and G. Rozza. A reduced order modeling technique to study bifurcating phenomena: Appli- cation to the Gross–Pitaevskii equation.SIAM Journal on Scientific Computing, 42(5):B1115–B1135, 2020

  45. [53]

    Pichi and G

    F. Pichi and G. Rozza. Reduced basis approaches for parametrized bifurcation problems held by non-linear Von K´ arm´ an equations.Journal of Scientific Computing, 81:112–135, 2019

  46. [54]

    Pichi and G

    F. Pichi and G. Rozza. Reduced Order Models for the Buckling of Hyperelastic Beams. In G. Rozza, G. Stabile, M. Gunzburger, and M. D’Elia, editors,Reduction, Approximation, Machine Learning, Surrogates, Emulators and Simulators: RAMSES, pages 199–240. Springer Nature Switzerla...

  47. [55]

    Pichi and M

    F. Pichi and M. Strazzullo. Deflation-based certified greedy algorithm and adaptivity for bifurcating nonlinear PDEs.Communications in Nonlinear Science and Numerical Simulation, 149:108941, 2025

  48. [56]

    Pichi, M

    F. Pichi, M. Strazzullo, F. Ballarin, and G. Rozza. Driving bifurcating parametrized nonlinear PDEs by optimal control strategies: application to Navier–Stokes equations with model order reduction.ESAIM: Mathematical Modelling and Numerical Analysis, 56(4):1361–1400, 2022. 26 ...

  49. [57]

    Pintore, F

    M. Pintore, F. Pichi, M. Hess, G. Rozza, and C. Canuto. Efficient computation of bifurcation diagrams with a deflated approach to reduced basis spectral element method.Advances in Computational Mathematics, 47:1–39, 2021

  50. [58]

    Pitton, A

    G. Pitton, A. Quaini, and G. Rozza. Computational reduction strategies for the detection of steady bifurcations in incompressible fluid-dynamics: Applications to Coanda effect in cardiology.Journal of Computational Physics, 344:534–557, 2017

  51. [59]

    Pitton and G

    G. Pitton and G. Rozza. On the application of reduced basis methods to bifurcation problems in incompressible fluid dynamics.Journal of Scientific Computing, 73(1):157–177, 2017

  52. [60]

    Quarteroni, R

    A. Quarteroni, R. Sacco, and F. Saleri.Numerical mathematics, volume 37. Springer Science & Business Media, 2006

  53. [61]

    S. S. Ray and A. K. Gupta.Wavelet methods for solving partial differential equations and fractional differential equations. Chapman and Hall/CRC, 2018

  54. [62]

    Seydel.Practical bifurcation and stability analysis, volume 5

    R. Seydel.Practical bifurcation and stability analysis, volume 5. Springer Science & Business Media, 2009

  55. [63]

    M. L. Shahab and H. Susanto. Neural networks for bifurcation and linear stability analysis of steady states in partial differential equations.Applied Mathematics and Computation, 483:128985, 2024

  56. [64]

    Shiralashetti and A

    S. Shiralashetti and A. Deshi. An efficient Haar wavelet collocation method for the numerical solution of multi- term fractional differential equations.Nonlinear Dynamics, 83(1):293–303, 2016

  57. [65]

    T. X. Sun, Z. C. Xue, and H. T. Zhang. Spatiotemporal dynamics of a diffusive si model in the regions of Turing-Hopf bifurcation point.Nonlinear Dynamics, 113(9):10681–10703, 2025

  58. [66]

    M. I. Syam and A. Hamdan. An efficient method for solving Bratu equations.Applied Mathematics and Compu- tation, 176(2):704–713, 2006

  59. [67]

    Tomada, M

    L. Tomada, M. Khamlich, F. Pichi, and G. Rozza. Sparse Identification for bifurcating phenomena in Computa- tional Fluid Dynamics.Computers & Fluids, 302:106841, 2025

  60. [68]

    H. Uecker. Continuation and bifurcation in nonlinear PDEs – algorithms, applications, and experiments.Jahres- bericht der Deutschen Mathematiker-Vereinigung, 124(1):43–80, 2022

  61. [69]

    Uecker, D

    H. Uecker, D. Wetzel, and J. D. M. Rademacher. pde2path - A Matlab package for continuation and bifurcation in 2d elliptic systems.Numerical Mathematics: Theory, Methods and Applications, 7(1):58–106, 2014

  62. [70]

    Venturi, X

    D. Venturi, X. Wan, and G. E. Karniadakis. Stochastic bifurcation analysis of Rayleigh–B´ enard convection. Journal of Fluid Mechanics, 650:391–413, May 2010. Publisher: Cambridge University Press

  63. [71]

    T. E. Woolley. Bespoke Turing patterns with specific nonlinear properties.Proceedings of the Royal Society A, 481(2312):20250030, 2025

  64. [72]

    J. Xia, P. E. Farrell, and S. G. P. Castro. Nonlinear bifurcation analysis of stiffener profiles via deflation techniques. Thin-Walled Structures, 149:106662, 2020

  65. [73]

    Yu and Y

    X. Yu and Y. Fu. An analytic derivation of the bifurcation conditions for localization in hyperelastic tubes and sheets.Zeitschrift f¨ ur Angewandte Mathematik und Physik, 73(3):116, 2022

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.