REVIEW 4 major objections 5 minor 46 references
A posteriori error analysis for the second-order BDF method for the Landau-Lifshitz-Gilbert equation
T0 review · 4 major / 5 minor · reviewed 2026-08-04 · deepseek-v4-flash
Pith's one-line read This paper proves a fully computable a posteriori error bound for the second-order BDF tangent plane discretization of the Landau–Lifshitz–Gilbert equation: the squared gradient error of the magnetization is bounded by an explicit sum of es
desk verdict First rigorous a posteriori bound for the tangent-plane BDF2 discretization of LLG, but the main theorem omits the initial projection error term its own proof requires, so it is not proven as stated. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing construction is the elliptic reconstruction R, defined at each time step by a(R m_h^n, φ)=⟨−Δ_h^n m_h^n,φ⟩ for all φ∈H¹(Ω), which turns the discrete solution into a continuous function whose error to m_h^n is controlled by standard elliptic a posteriori estimators (element and edge residuals). Around this, the paper builds a quadratic three-point reconstruction M_h(t) that matches the BDF(2) operator and a corresponding time-space reconstruction W(t), leading to a parabolic error equation whose residual splits into the listed estimator families. The tangent-plane projection P(m)=I−mmᵀ enters through a Lipschitz bound that lets the projection error be controlled by the projec
What would settle it
Compute the right-hand estimator sum and the exact gradient error for a smooth manufactured solution of the LLG equation on a sequence of uniformly refined meshes; if the ratio of true squared gradient error to the estimator sum is not bounded from above, or if the estimators do not tend to zero as h,τ→0, Theorem 9 is false. The most direct check is to verify Assumption 4 on a nonconvex or polyhedral domain by attempting to construct the L∞ and W^{1,∞} elliptic estimators with uniform constants.
Extended reading notes
Core claim
The central claim is Theorem 9: for the BDF(2) tangent-plane scheme with conforming finite elements, the true gradient error satisfies ‖∇(m−m_h)‖²_{L∞(t0,tN;L2)} ≲ F1+F2+τ⁴ΣE_i+Λ1+Λ2+Λ3+P+Ξ1+Ξ2+Ξ3+C1+C2+Q1+Q2+I1+I2+I3, with a constant depending only on α, ‖M_h‖_{W^{1,∞}}, ‖∂tM_h‖_{W^{1,∞}}, η(m_h^n;W^{1,∞}), and ‖λ_h‖_{L∞}. The terms on the right are defined explicitly in Definition 8: time-error indicators built from discrete second and third differences, a reconstruction-error indicator from the discrete Laplacian, a projection-error indicator measuring how far the time reconstruction is from satisfying the unit-length tangent constraint, elliptic a posteriori space estimators, data-approx
Load-bearing premise
The load-bearing premise is that the elliptic reconstruction lies in W^{1,∞} and that a posteriori estimators for its L∞ and W^{1,∞} error are available and uniformly bounded independent of mesh and time step; if this uniform bound fails on the domains where the method is used, the main error bound is not justified.
Editorial extensions
If this is right
- Theorem 9 yields a computable upper bound for the true gradient error of the BDF(2) tangent-plane scheme, so adaptive refinement can in principle be driven by the estimators rather than by heuristic indicators.
- The error decomposition separates temporal (E1–E6), spatial (Λ), projection (P), predictor (Q), changing-mesh (Ξ), and initial (I) error sources, so a solver can attribute error to specific causes.
- The variable-step analogue (Theorem 19) extends the bound to adaptive time stepping, provided τ1=τ2 and the spatial mesh is fixed; this is the setting used by practical adaptive integrators.
- The bound holds for finite elements of arbitrary polynomial degree, matching the most flexible TPS implementations.
- Since every estimator is defined through discrete differences, projections, and elliptic residuals, no information about the exact solution is needed beyond the data and numerical solution.
Reading between the lines
- One immediate next test is estimator efficiency: checking numerically whether the right-hand side also provides a lower bound up to a constant on the true error; if it does, the indicators become quantitatively reliable, not just safe.
- The uniform W^{1,∞} assumption (Assumption 4) is likely the point to audit first on nonconvex or polyhedral domains; if the cited L∞-type elliptic estimators do not hold uniformly, the theorem's bound as stated is not justified, and a modified estimate would be needed.
- Because the analysis treats the predictor c m_h^n abstractly, the same framework should extend to projection-free and higher-order predictor variants; a natural extension would be to prove an analogous bound for third-order BDF or harmonically mapped predictors.
- If implemented in an adaptive solver, the projection estimator P alone could drive local refinement in regions where the discrete solution fails to satisfy |m|=1, which is exactly where domain walls develop.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper derives a residual-based a posteriori error estimate for the BDF(2) tangent-plane scheme for the Landau-Lifshitz-Gilbert equation, using elliptic reconstruction and a three-point time reconstruction. For constant time steps, the main result (Theorem 9) bounds the squared gradient error ||∇(m - m_h)||^2_{L∞(t0,tN;L2)} by a sum of computable estimators measuring temporal, spatial, projection, mesh-change, conformity, extrapolation, data-approximation, and initial errors. A variable-step version is stated in Theorem 19, with the proofs said to follow the constant-step arguments. The paper is presented as the first rigorous a posteriori analysis for the tangent-plane formulation and as a foundation for fully adaptive algorithms.
Significance. If the main estimate is correct, this is a significant contribution: it provides the first computable, rigorous upper bound for the gradient error of a standard TPS discretization of LLG, and it explicitly separates the spatial, temporal, and constraint-related error components. The constant-time-step analysis is detailed: Lemmas 10-17 provide explicit residual bounds, and the use of Assumptions 3-4 is clearly separated from the derivation of the error equation and the Gronwall argument. The paper does not make any ad-hoc fitting to the target error, and the estimators are genuinely computable from the discrete solution. However, the central theorem has load-bearing gaps: the initial-data projection error is missing, the stated assumptions omit hypotheses used in the initial-step lemma, and the variable-step theorem is not proved in the manuscript. These issues affect the validity of the main claim as stated, so the paper requires substantial revision rather than acceptance.
major comments (4)
- [§3, Lemma 11 and Theorem 9] Lemma 11, whose right-hand side starts with ||∇(P_0^0 m(0) - m(0))||^2, is the only estimate covering the initial interval [0,t1]. Since m_h^0 = P_0^0 m(0) and m(0) is not assumed to lie in V_h, this term is generally nonzero and is not present in Definition 8 or in the bound (3.4). The estimators I1-I3 involve m_h^{-1}, Ψ0, and first-step differences, but no term controls the exact initial projection mismatch. For a stationary exact solution with m0∉V_h (e.g., smooth m0 with f=-Δm0), no listed estimator is forced to control ||∇(P_0^0 m0 - m0)||. Thus the proof of (3.4) cannot account for the initial time interval unless the hidden constant is allowed to depend on the initial projection error, which contradicts the stated dependence in Theorem 9. The theorem and estimator list must either include this term or prove that it is controlled by the existing estimators.
- [§3.1, Lemma 11 and Theorem 9] Lemma 11 assumes V_h^0 = V_h^1 = V_h^2, but Theorem 9 does not state this assumption. The proof of Lemma 11 uses this equality to obtain ∂m_h^1 = P_0^1 ∂m_h^1 in (3.24). If the mesh changes during the first two time steps, additional consistency terms appear and are not covered by the derivation. Since adaptive algorithms typically change the mesh after each step, the main theorem must either explicitly require V_h^0 = V_h^1 = V_h^2, or the analysis must be extended to handle changing initial meshes.
- [§4, Theorem 19] The variable-time-step theorem is stated without proof. The text says 'the proofs follow the same arguments as in the uniform time-step case' and then lists modified estimators, but the actual verification is not provided. This is load-bearing because the abstract and the claimed contribution include adaptive meshes with variable time-step sizes. In particular, the variable-step BDF(2) operator (4.1), the modified reconstruction (4.2), the new estimator eE7, and the handling of the initial interval require a complete proof. The constant-step analysis does not automatically cover the variable-step case, especially for the initial step and for the estimator E2/E7. A theorem whose proof is only sketched cannot support the main contribution.
- [§2.3, Assumption 4 and Theorem 9] The proof depends on the uniform boundedness of η(m_h^n; W^{1,∞}) and on the regularity R m_h^n ∈ W^{1,∞}(Ω). Assumption 4 merely assumes these bounds and notes that 'additional assumptions on the domain' may be needed, but Theorem 9 does not state those assumptions. The hidden constant in (3.4) explicitly depends on η(m_h^n; W^{1,∞}); if this quantity is not uniformly bounded on the polyhedral or nonconvex domains of interest, the estimate is not justified. The paper should state precise conditions on Ω and on the finite element spaces that guarantee (2.14) with a uniform bound, or alternatively include η(m_h^n; W^{1,∞}) explicitly in the a posteriori bound so that it is not hidden.
minor comments (5)
- [Abstract] Typographical: 'rigorousa posteriori' should be 'rigorous a posteriori' in two places.
- [§4 heading] 'V ariable time-steps' should be 'Variable time-steps'.
- [Definition 8] The quantity αm_h^{-1} is introduced in (3.1) but is not an estimator; its role in defining the BDF2 first step should be clarified in the text, especially since m_h^{-1} depends on m_h^1 and thus is not a standard initial value.
- [Theorem 19] The theorem assumes τ = τ_1 = τ_2 (the first two time steps equal) and V_h^n = V_h for all n, but this is not reconciled with the claim of adaptive variable time steps. The condition should be stated in the abstract or introduction so the reader knows the scope of the variable-step result.
- [Lemma 5] In the proof, the estimate ||e||_{L∞} ≤ ||m||_{L∞} + ||f_m||_{L∞} is used; this is correct but should be stated explicitly before it is applied in the gradient estimate.
Circularity Check
No circularity: the a posteriori bound is residual-based and conditional on external estimators; the missing initial-data term is a rigor gap, not a circular reduction.
full rationale
The derivation chain in this paper is not circular. Theorem 9 is a conditional residual-type estimate: the discrete BDF(2) equations (2.20)–(2.23), the elliptic reconstruction (2.11), and the assumed a posteriori bounds (Assumptions 3 and 4) are the only inputs. Every estimator in Definition 8 is explicitly computable from the discrete sequence {m_h^n, cm_h^n, f_h^n, λ_h^n} and the data; none is defined in terms of the target error ∥∇(m−m_h)∥. The proof starts from the parabolic error equation (Lemma 10), derives the differential inequality (3.19)–(3.21), and concludes via Gronwall; the target error appears only on the left-hand side. The cited elliptic estimators [2,44,45], the L∞ estimators [39,22], and the time-reconstruction framework [5,36,15] are independent external tools, not self-citations or assumptions equivalent to the theorem. The Lipschitz-type projection bound (Lemma 5) and the normalization-error bound (Lemma 17) are proved in the paper from the algebraic structure of P(m), not imported as the result. There is no self-citation chain: the author does not invoke any of his own prior results to force the form of the bound. Separate rigor gaps exist, but they are not circularity: Lemma 11's initial-interval estimate contains the term ∥∇(P_0^0m(0)−m(0))∥², which does not appear in Definition 8 or in the right-hand side of Theorem 9; Lemma 11 also assumes V_h^0=V_h^1=V_h^2, a hypothesis not stated in Theorem 9. Additionally, Assumption 4's uniform boundedness of η(m_h^n;W^{1,∞}) and the W^{1,∞} regularity of the reconstruction (3.8) are assumed rather than proved. These are missing hypotheses or unproved regularity assumptions that a correctness review should flag, but they do not make any estimate reduce to its own input by construction. Therefore the circularity score is 0.
Assumptions & free parameters
assumptions (6)
- domain assumption The exact solution (m, λ) of the saddle point formulation (2.16) exists with sufficient regularity, including m ∈ W^{1,∞}(Ω) for the inf-sup condition in Lemma 6 and the required Bochner space memberships.
- domain assumption The elliptic reconstruction R m_h^n satisfies R m_h^n ∈ W^{1,∞}(Ω), and computable a posteriori estimators η(m_h^n; L2), η(m_h^n; H1), η(m_h^n; L∞), η(m_h^n; W^{1,∞}) exist with uniform bound η(m_h^n; W^{1,∞}) ≤ C independent of discretization.
- domain assumption The fully discrete saddle point problem (2.21) satisfies the discrete inf-sup condition so that (m_h^n, λ_h^n) is well defined.
- domain assumption The initial finite element spaces coincide, V^0_h = V^1_h = V^2_h, for the initial time-step estimate.
- domain assumption For variable time steps, the BDF(2) coefficients remain stable; the paper does not state a bound on step ratios κ_n = τ_n / τ_{n-1}.
- standard math Standard mathematical tools: Young, Hölder, trace and interpolation inequalities, Gronwall's inequality, and standard elliptic a posteriori estimates from the cited literature.
Cite this review
Pith. "Pith review of A posteriori error analysis for the second-order BDF method for the Landau-Lifshitz-Gilbert equation." pith.science (2026). https://pith.science/paper/KJ3HMEMZ
@misc{pith2026260802021,
author = {Pith},
title = {Pith review of: A posteriori error analysis for the second-order BDF method for the Landau-Lifshitz-Gilbert equation},
year = {2026},
howpublished = {\url{https://pith.science/paper/KJ3HMEMZ}},
note = {Machine review of arXiv:2608.02021}
}
read the original abstract
The tangent plane scheme (TPS) is a well-established discretization of the Landau-Lifshitz-Gilbert (LLG) equation. However, rigorous a posteriori error estimates have not been established. In this work, we derive a rigorous a posteriori error estimate for the TPS based on the second-order backward differentiation formula (BDF(2)) in time and finite elements of arbitrary polynomial degree in space. The proposed estimators provide computable upper bounds for the temporal and spatial discretization errors on adaptive meshes with variable time-step sizes. This result establishes the mathematical foundation for fully adaptive algorithms for the LLG equation.
Reference graph
Works this paper leans on
-
[1]
Spin-polarized transport in ferromagnetic multilayers: an unconditionally convergent FEM integrator
C. Abert, G. Hrkac, M. Page, D. Praetorius, M. Ruggeri, and D. Suess. “Spin-polarized transport in ferromagnetic multilayers: an unconditionally convergent FEM integrator”. In: Comput. Math. Appl.68.6 (2014), pp. 639–654
2014
-
[2]
A posteriori error estimation in finite element analysis
M. Ainsworth and J. T. Oden. “A posteriori error estimation in finite element analysis”. In: Comput. Methods Appl. Mech. Engrg.142.1-2 (1997), pp. 1–88
1997
-
[3]
Quadratic constraint consistency in the projection- free approximation of harmonic maps and bending isometries
G. Akrivis, S. Bartels, and C. Palus. “Quadratic constraint consistency in the projection- free approximation of harmonic maps and bending isometries”. In:Math. Comp.94.355 (2025), pp. 2251–2269
2025
-
[4]
G. Akrivis, S. Bartels, M. Ruggeri, and J. Wang.Projection-free approximation of flows of harmonic maps with quadratic constraint accuracy and variable step sizes. 2025. arXiv: 2505.05655 [math.NA]
arXiv 2025
-
[5]
A posteriori error estimates for the two-step backward differentiation formula method for parabolic equations
G. Akrivis and P. Chatzipantelidis. “A posteriori error estimates for the two-step backward differentiation formula method for parabolic equations”. In:SIAM J. Numer. Anal.48.1 (2010), pp. 109–132
2010
-
[6]
Higher-order linearly implicit full discretization of the Landau-Lifshitz-Gilbert equation
G. Akrivis, M. Feischl, B. Kovács, and C. Lubich. “Higher-order linearly implicit full discretization of the Landau-Lifshitz-Gilbert equation”. In:Math. Comp.90.329 (2021), pp. 995–1038
2021
-
[7]
M. Aldé, M. Feischl, and D. Praetorius.BDF2-type integrator for Landau–Lifshitz–Gilbert equation in micromagnetics: unconditional weak convergence to weak solutions. 2026. arXiv: 2511.22000 [math.NA]
arXiv 2026
-
[8]
A new finite element scheme for Landau-Lifchitz equations
F. Alouges. “A new finite element scheme for Landau-Lifchitz equations”. In:Discrete Contin. Dyn. Syst. Ser. S1.2 (2008), pp. 187–196
2008
Show all 46 references
-
[9]
Convergence of a finite element discretization for the Landau- Lifshitz equations in micromagnetism
F. Alouges and P. Jaisson. “Convergence of a finite element discretization for the Landau- Lifshitz equations in micromagnetism”. In:Math. Models Methods Appl. Sci.16.2 (2006), pp. 299–316
2006
-
[10]
On global weak solutions for Landau-Lifshitz equations: existence and nonuniqueness
F. Alouges and A. Soyeur. “On global weak solutions for Landau-Lifshitz equations: existence and nonuniqueness”. In:Nonlinear Anal.18.11 (1992), pp. 1071–1084
1992
-
[11]
2D and 3D calculations of micromagnetic wall structures using finite elements
A. Bagneres-Viallix, P. Baras, and J. Albertini. “2D and 3D calculations of micromagnetic wall structures using finite elements”. In:IEEE Transactions on Magnetics27.5 (1991), pp. 3819–3822. REFERENCES 27
1991
-
[12]
On dynamical Micromagnetism with Magnetostriction
L. Baňas. “On dynamical Micromagnetism with Magnetostriction”. PhD thesis. Ghent University, 2005
2005
-
[13]
Adaptive techniques for Landau-Lifshitz-Gilbert equation with magnetostric- tion
L. Baňas. “Adaptive techniques for Landau-Lifshitz-Gilbert equation with magnetostric- tion”. In:J. Comput. Appl. Math.215.2 (2008), pp. 304–310
2008
-
[14]
Error estimates for Landau-Lifshitz-Gilbert equation with magnetostriction
L. Baňas and M. Slodička. “Error estimates for Landau-Lifshitz-Gilbert equation with magnetostriction”. In:Appl. Numer. Math.56.8 (2006), pp. 1019–1039
2006
-
[15]
A posteriori estimates for the two-step backward differentiation formula and discrete regularity for the time-dependent Stokes equations
E. Bänsch and A. Brenner. “A posteriori estimates for the two-step backward differentiation formula and discrete regularity for the time-dependent Stokes equations”. In:IMA J. Numer. Anal. 39.2 (2019), pp. 713–759
2019
-
[16]
The effect of mesh modification in time on the error control of fully discrete approximations for parabolic equations
E. Bänsch, F. Karakatsani, and C. G. Makridakis. “The effect of mesh modification in time on the error control of fully discrete approximations for parabolic equations”. In:Appl. Numer. Math.67 (2013), pp. 35–63
2013
-
[17]
Projection-free approximation of geometrically constrained partial differential equations
S. Bartels. “Projection-free approximation of geometrically constrained partial differential equations”. In:Math. Comp.85.299 (2016), pp. 1033–1049
2016
-
[18]
Error analysis for the numerical approximation of the harmonic map heat flow with nodal constraints
S. Bartels, B. Kovács, and Z. Wang. “Error analysis for the numerical approximation of the harmonic map heat flow with nodal constraints”. In:IMA J. Numer. Anal.44.2 (2024), pp. 633–653
2024
-
[19]
Regular solutions for Landau-Lifschitz equation inR3
G. Carbou and P. Fabrie. “Regular solutions for Landau-Lifschitz equation inR3”. In: Commun. Appl. Anal.5.1 (2001), pp. 17–30
2001
-
[20]
Regular solutions for Landau-Lifschitz equation in a bounded domain
G. Carbou and P. Fabrie. “Regular solutions for Landau-Lifschitz equation in a bounded domain”. In:Differential Integral Equations14.2 (2001), pp. 213–229
2001
-
[21]
A survey on the numerics and computations for the Landau-Lifshitz equation of micromagnetism
I. Cimrák. “A survey on the numerics and computations for the Landau-Lifshitz equation of micromagnetism”. In:Arch. Comput. Methods Eng.15.3 (2008), pp. 277–309
2008
-
[22]
Localized pointwise a posteriori error estimates for gradients of piecewise linear finite element approximations to second-order quasilinear elliptic problems
A. Demlow. “Localized pointwise a posteriori error estimates for gradients of piecewise linear finite element approximations to second-order quasilinear elliptic problems”. In: SIAM J. Numer. Anal.44.2 (2006), pp. 494–514
2006
-
[23]
The extrapolated explicit midpoint scheme for variable order and step size controlled integration of the Landau-Lifschitz-Gilbert equation
L. Exl, N. J. Mauser, T. Schrefl, and D. Suess. “The extrapolated explicit midpoint scheme for variable order and step size controlled integration of the Landau-Lifschitz-Gilbert equation”. In:J. Comput. Phys.346 (2017), pp. 14–24
2017
-
[24]
An adaptive moving mesh method for simulating finite-time blowup solutions of the Landau-Lifshitz-Gilbert equation
Z. Fang and X. Wang. “An adaptive moving mesh method for simulating finite-time blowup solutions of the Landau-Lifshitz-Gilbert equation”. In:East Asian J. Appl. Math.14.3 (2024), pp. 601–635
2024
-
[25]
Existence of regular solutions of the Landau-Lifshitz-Gilbert equation in 3D with natural boundary conditions
M. Feischl and T. Tran. “Existence of regular solutions of the Landau-Lifshitz-Gilbert equation in 3D with natural boundary conditions”. In:SIAM J. Math. Anal.49.6 (2017), pp. 4470–4490
2017
-
[26]
The eddy current–LLG equations: FEM-BEM coupling and a priori error estimates
M. Feischl and T. Tran. “The eddy current–LLG equations: FEM-BEM coupling and a priori error estimates”. In:SIAM J. Numer. Anal.55.4 (2017), pp. 1786–1819
2017
-
[27]
Skyrmions on the track
A. Fert, V. Cros, and J. Sampaio. “Skyrmions on the track”. In:Nat. Nanotechnol.8 (2013), pp. 152–156
2013
-
[28]
Electrical control of magnetism by electric field and current-induced torques
A. Fert, R. Ramesh, V. Garcia, F. Casanova, and M. Bibes. “Electrical control of magnetism by electric field and current-induced torques”. In:Rev. Mod. Phys.96 (1 Mar. 2024), p. 015005. 28 REFERENCES
2024
-
[29]
Adaptive mesh refinement for micromagnetics simulations
C. J. García-Cervera and A. M. Roma. “Adaptive mesh refinement for micromagnetics simulations”. In:IEEE transactions on magnetics42.6 (2006), pp. 1648–1654
2006
-
[30]
Konvergente numerische Integration der Landau-Lifshitz-Gilbert Gleichung
P. Goldenits. “Konvergente numerische Integration der Landau-Lifshitz-Gilbert Gleichung”. PhD thesis. Institute for Analysis and Scientific Computing, Vienna University of Technol- ogy: TU Wien, Nov. 2012
2012
-
[31]
The Landau-Lifshitz equation of the ferromagnetic spin chain and harmonic maps
B. L. Guo and M. C. Hong. “The Landau-Lifshitz equation of the ferromagnetic spin chain and harmonic maps”. In:Calc. Var. Partial Differential Equations1.3 (1993), pp. 311–334
1993
-
[32]
Adaptive finite element mesh refinement techniques in three-dimensional micromagnetic modeling
R. Hertel and H. Kronmuller. “Adaptive finite element mesh refinement techniques in three-dimensional micromagnetic modeling”. In:IEEE Transactions on Magnetics34.6 (1998), pp. 3922–3930
1998
-
[33]
Review on spintronics: Principles and device applications
A. Hirohata, K. Yamada, Y. Nakatani, I.-L. Prejbeanu, B. Diény, P. Pirro, and B. Hille- brands. “Review on spintronics: Principles and device applications”. In:Journal of Mag- netism and Magnetic Materials509 (2020), p. 166711
2020
-
[34]
Recent developments in the modeling, analysis, and numerics of ferromagnetism
M. Kružík and A. Prohl. “Recent developments in the modeling, analysis, and numerics of ferromagnetism”. In:SIAM Rev.48.3 (2006), pp. 439–483
2006
-
[35]
Elliptic reconstruction and a posteriori error estimates for fully discrete linear parabolic problems
O. Lakkis and C. G. Makridakis. “Elliptic reconstruction and a posteriori error estimates for fully discrete linear parabolic problems”. In:Math. Comp.75.256 (2006), pp. 1627–1658
2006
-
[36]
An anisotropic error estimator for the Crank-Nicolson method: application to a parabolic problem
A. Lozinski, M. Picasso, and V. Prachittham. “An anisotropic error estimator for the Crank-Nicolson method: application to a parabolic problem”. In:SIAM J. Sci. Comput. 31.4 (2009), pp. 2757–2783
2009
-
[37]
Elliptic reconstruction and a posteriori error estimates for parabolic problems
C. Makridakis and R. H. Nochetto. “Elliptic reconstruction and a posteriori error estimates for parabolic problems”. In:SIAM J. Numer. Anal.41.4 (2003), pp. 1585–1594
2003
-
[38]
McLean.Strongly elliptic systems and boundary integral equations
W. McLean.Strongly elliptic systems and boundary integral equations. Cambridge Univer- sity Press, Cambridge, 2000, pp. xiv+357
2000
-
[39]
Pointwise a posteriori error estimates for monotone semi-linear equations
R. H. Nochetto, A. Schmidt, K. G. Siebert, and A. Veeser. “Pointwise a posteriori error estimates for monotone semi-linear equations”. In:Numer. Math.104.4 (2006), pp. 515–538
2006
-
[40]
Mesh refinement in FE-micromagnetics for multi- domain Nd2Fe14B particles
W. Scholz, T. Schrefl, and J. Fidler. “Mesh refinement in FE-micromagnetics for multi- domain Nd2Fe14B particles”. In:Journal of Magnetism and Magnetic Materials196-197 (1999), pp. 933–934
1999
-
[41]
Finite elements in numerical micromagnetics. Part I: Granular hard magnets
T. Schrefl. “Finite elements in numerical micromagnetics. Part I: Granular hard magnets”. In: Journal of Magnetism and Magnetic Materials207.1 (1999), pp. 45–65
1999
-
[42]
Numerical methods for dynamic micromagnetics
D. Shepherd. “Numerical methods for dynamic micromagnetics”. PhD thesis. University of Manchester, 2015
2015
-
[43]
Finite element micromagnetic simulations with adaptive mesh refinement
K. Tako, T. Schrefl, M. Wongsam, and R. Chantrell. “Finite element micromagnetic simulations with adaptive mesh refinement”. In:Journal of applied physics81.8 (1997), pp. 4082–4084
1997
-
[44]
Verfürth
R. Verfürth. A review of a posteriori error estimation and adaptive mesh-refinement techniques. Advances in numerical mathematics. Wiley, 1996
1996
-
[45]
Verfürth.A posteriori error estimation techniques for finite element methods
R. Verfürth.A posteriori error estimation techniques for finite element methods. Numer- ical Mathematics and Scientific Computation. Oxford University Press, Oxford, 2013, pp. xx+393. REFERENCES 29
2013
-
[46]
On Landau-Lifshitz’ equations for ferromagnetism
A. Visintin. “On Landau-Lifshitz’ equations for ferromagnetism”. In:Japan J. Appl. Math. 2.1 (1985), pp. 69–84. Institute for Applied and Numerical Mathematics, Karlsruhe Institute of Technology, Englerstr. 2, 76131 Karlsruhe, Germany Email address: stefan.karch@kit.edu
1985
Reviewed August 4, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.