pith. sign in

arxiv: 2606.03833 · v1 · pith:7CNOR376new · submitted 2026-06-02 · 🧮 math.NA · cs.NA

Three-term recurrence iterations for energy-based models

R. Altmann , J. Ramme , P. Schulze This is my paper

Pith reviewed 2026-06-28 08:52 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords energy-based modelsmidpoint rulethree-term recurrencedissipation preservationiteration matrixvariable scalingblock diagonalporoelasticity
0
0 comments X

The pith

Scaling state variables makes the symmetric part of the iteration matrix positive definite, enabling three-term recurrence schemes for energy-based models.

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

For energy-based models where the midpoint rule already preserves the dissipation inequality, an appropriate scaling of the state variables is introduced. This scaling guarantees that the symmetric part of the resulting iteration matrix is positive definite. The property then permits direct application of three-term iteration schemes such as the methods of Widlund and Rapoport. When the symmetric part is block diagonal the linear systems decouple, producing efficient dissipation-preserving discretizations. The approach is demonstrated on the biharmonic heat equation and linear poroelasticity.

Core claim

We introduce an appropriate scaling of the state variables such that the symmetric part of the resulting iteration matrix is guaranteed to be positive definite. This allows the application of three-term iteration schemes such as the methods of Widlund and Rapoport. Special emphasis is put on examples where the symmetric part is block diagonal such that the computations decouple. This then leads to efficient dissipation-preserving numerical schemes as illustrated in two numerical examples, namely the biharmonic heat equation and linear poroelasticity.

What carries the argument

Scaled state variables that force the symmetric part of the midpoint-rule iteration matrix to be positive definite.

If this is right

  • Three-term recurrence methods such as Widlund and Rapoport become applicable to the scaled systems.
  • Block-diagonal structure of the symmetric part yields fully decoupled linear solves.
  • The resulting schemes remain dissipation-preserving by construction.
  • The same scaling produces efficient integrators for the biharmonic heat equation and linear poroelasticity.

Where Pith is reading between the lines

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

  • The scaling construction might extend to other linear or mildly nonlinear dissipative systems that admit a midpoint discretization.
  • Decoupling in the block-diagonal case suggests possible domain-decomposition or parallel-in-time variants.
  • The technique could be combined with existing structure-preserving time-stepping methods to enlarge the set of solvable models.

Load-bearing premise

The models belong to the class where the midpoint rule preserves the dissipation inequality and a scaling exists that makes the symmetric part positive definite without destroying other model properties.

What would settle it

A concrete model in the admissible class for which no scaling produces a positive definite symmetric part while still preserving the dissipation inequality under the midpoint rule.

Figures

Figures reproduced from arXiv: 2606.03833 by J. Ramme, P. Schulze, R. Altmann.

Figure 4.1
Figure 4.1. Figure 4.1: Relative residual norms for the four methods and different time step sizes τ and system dimensions η. We would like to emphasize that none of the methods is actually minimizing the Euclidean norm. Instead, as discussed previously, the method of Rapoport minimizes the residual in the H−1 -norm and preconditioned GMRES minimizes the residual in the H−2 -norm, while the method of Widlund does not minimize t… view at source ↗
Figure 4.2
Figure 4.2. Figure 4.2: Comparison of the three different norms of the relative residual when solving system (4.3) with η = τ −1 = 105 . 4.3. Linear poroelasticity. Going back to [Bio41, Sho00], the (spatially discretized) equa￾tions of linear poroelasticity are given by the DAE " 0 0 D C# "u˙ p˙ # = " −A DT 0 −B # "u p # + " f g # (4.5) with initial conditions u(0) = u 0 ∈ R n and p(0) = p 0 ∈ R m. In applications, u models th… view at source ↗
Figure 4.3
Figure 4.3. Figure 4.3: Computation times vs. energy error for a single time step of the midpoint rule using different iteration schemes. The gray dashed line in￾dicates the 3.15 seconds which were needed to compute the reference solution with Matlab’s backslash operator applied to the overall system. and a maximum number of inner iteration steps K. It was shown that convergence is only guaranteed if K grows for decreasing time… view at source ↗
Figure 4.4
Figure 4.4. Figure 4.4: Comparison of different decoupling time stepping methods of order two in terms of computation time. The fixed stress scheme is shown for L = 0 and K = 3 (dotted), K = 6 (dashed), and K = 9 (solid). Widlund’s method is run with TOL = τ 2 (dotted), TOL = 0.01 τ 2 (dashed), and TOL = 0.001 τ 2 (solid). its block structure, we show that the method of Widlund is compatible with well-established (but possibly … view at source ↗
read the original abstract

It is well-known that the midpoint rule preserves the dissipation inequality if applied to a certain class of energy-based models. We introduce an appropriate scaling of the state variables such that the symmetric part of the resulting iteration matrix is guaranteed to be positive definite. This allows the application of three-term iteration schemes such as the methods of Widlund and Rapoport. Special emphasis is put on examples where the symmetric part is block diagonal such that the computations decouple. This then leads to efficient dissipation-preserving numerical schemes as illustrated in two numerical examples, namely the biharmonic heat equation and linear poroelasticity.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

0 major / 2 minor

Summary. The manuscript proposes scaling the state variables of a certain class of energy-based models so that, after midpoint-rule discretization, the symmetric part of the iteration matrix is guaranteed positive definite. This enables three-term recurrence solvers (Widlund, Rapoport) while preserving the dissipation inequality; special attention is given to block-diagonal cases that permit decoupling. The claims are illustrated on the biharmonic heat equation and linear poroelasticity.

Significance. If the explicit scaling construction is correct and preserves the model properties, the work supplies a practical route to efficient, structure-preserving iterative solvers for dissipation-preserving discretizations, especially when block structure survives. The numerical examples are chosen precisely to exploit the decoupling, which strengthens the practical relevance.

minor comments (2)
  1. [Abstract / Section introducing the scaled iteration] The abstract asserts that an appropriate scaling exists and guarantees positive definiteness, but the manuscript should include an explicit derivation or algorithm for constructing the scaling (e.g., in the section introducing the iteration matrix) together with a short verification that the scaled system remains equivalent to the original model.
  2. [Numerical examples section] The two numerical examples are named but the manuscript should report at least one quantitative indicator (iteration counts, residual histories, or CPU times) comparing the three-term scheme against a standard solver to substantiate the efficiency claim.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive assessment of the manuscript, the accurate summary of its contributions, and the recommendation for minor revision. No major comments were provided in the report.

Circularity Check

0 steps flagged

No significant circularity in derivation chain

full rationale

The paper presents an explicit construction of a scaling for state variables in energy-based models to which the midpoint rule applies, guaranteeing that the symmetric part of the resulting iteration matrix is positive definite and thereby permitting three-term recurrences such as Widlund and Rapoport while preserving the dissipation inequality. This construction is introduced directly for the targeted model class and illustrated on concrete examples (biharmonic heat equation, linear poroelasticity) where block-diagonal structure survives; no step reduces a claimed prediction to a fitted quantity defined by the result itself, no load-bearing premise rests solely on self-citation, and no uniqueness theorem or ansatz is smuggled in. The derivation chain is therefore self-contained against external benchmarks and receives score 0.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the existence of a scaling that renders the symmetric part positive definite for the target class of models. No free parameters, invented entities, or non-standard axioms are mentioned in the abstract.

axioms (1)
  • domain assumption The midpoint rule preserves the dissipation inequality for the considered class of energy-based models.
    Stated as well-known in the abstract; forms the baseline for the new scaling.

pith-pipeline@v0.9.1-grok · 5622 in / 1089 out tokens · 16453 ms · 2026-06-28T08:52:57.131596+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

177 extracted references · 37 canonical work pages

  1. [1]

    2021 , doi =

    Semi-explicit discretization schemes for weakly-coupled elliptic-parabolic problems , journal =. 2021 , doi =

    2021

  2. [2]

    and Maier, R

    Altmann, R. and Maier, R. and Unger, B. , title =. PAMM , volume =. doi:10.1002/pamm.202000061 , year =

    work page doi:10.1002/pamm.202000061

  3. [3]

    and Mehrmann, V

    Altmann, R. and Mehrmann, V. and Unger, B. , title =. Math. Comput. Model. Dyn. Sys. , year =. doi:10.1080/13873954.2021.1975137 , volume =

    work page doi:10.1080/13873954.2021.1975137 2021

  4. [4]

    and Deiml, M

    Altmann, R. and Deiml, M. , title =. SIAM J. Sci. Comput. , volume =

  5. [5]

    and Schulze, P

    Altmann, R. and Schulze, P. , Year =. A port-. Systems Control Lett. , Volume =

  6. [6]

    and Schulze, P

    Bansal, H. and Schulze, P. and Abbasi, M. H. and Zwart, H. and Iapichino, L. and Schilders, W. H. A. and van de Wouw, N. , Year =. Port-. Systems Control Lett. , Volume =

  7. [7]

    and Mehrmann, V

    Beattie, C. and Mehrmann, V. and Xu, H. and Zwart, H. , Year =. Linear port-. Math. Control Signal Syst. , Volume =

  8. [8]

    and Reichel, L

    Beckermann, B. and Reichel, L. , Year =. The. SIAM J. Matrix Anal. Appl. , Volume =

  9. [9]

    2017 , Booktitle =

    Model Order Reduction for Differential--Algebraic Equations: A Survey , Author =. 2017 , Booktitle =

    2017

  10. [10]

    Biot, M. A. , title =. J. Appl. Phys. , fjournal =

  11. [11]

    , edition =

    Braess, D. , edition =. Finite Elements - Theory, Fast Solvers, and Applications in Solid Mechanics , year =

  12. [12]

    2013 , publisher =

    Mixed finite element methods and applications , author =. 2013 , publisher =

    2013

  13. [13]

    Both, J. W. and Borregales, M. and Nordbotten, J. M. and Kumar, K. and Radu, F. A. , TITLE =. Appl. Math. Lett. , FJOURNAL =. 2017 , PAGES =

    2017

  14. [14]

    and Alazard, D

    Brugnoli, A. and Alazard, D. and Pommier-Budinger, V. and Matignon, D. , Year =. Port-. Appl. Math. Model. , Volume =

  15. [15]

    and Rashad, R

    Brugnoli, A. and Rashad, R. and Stramigioli, S. , Year =. Dual field structure-preserving discretization of port-. J. Comput. Phys. , Volume =

  16. [16]

    and Mura, J

    Caiazzo, A. and Mura, J. , title =. Multiscale Model. Simul. , fjournal =. 2014 , doi =

    2014

  17. [17]

    Campbell, S. L. and Marszalek, W. , TITLE =. J. Comput. Appl. Math. , FJOURNAL =. 1997 , NUMBER =. doi:10.1016/S0377-0427(97)00084-8 , URL =

    work page doi:10.1016/s0377-0427(97)00084-8 1997

  18. [18]

    Campbell, S. L. and Marszalek, W. , TITLE =. Math. Comput. Model. Dyn. Syst. , VOLUME =. 1999 , NUMBER =. doi:10.1076/mcmd.5.1.18.3625 , URL =

    work page doi:10.1076/mcmd.5.1.18.3625 1999

  19. [19]

    Cardoso-Ribeiro, F. L. and Matignon, D. and Pommier-Budinger, V. , Year =. A port-. J. Fluids Struct. , Volume =

  20. [20]

    Celledoni, E. and H. Energy-preserving and passivity-consistent numerical discretization of port-. 2017 , Type =

    2017

  21. [21]

    and Beattie, C

    Chaturantabut, S. and Beattie, C. and Gugercin, S. , Year =. Structure-Preserving Model Reduction for Nonlinear Port-. SIAM J. Sci. Comput. , Volume =

  22. [22]

    Ciarlet, P. G. , TITLE =

  23. [23]

    Ciarlet, P. G. , year =. Mathematical elasticity. Vol

  24. [24]

    and Strehmel, K

    Debrabant, K. and Strehmel, K. , TITLE =. Appl. Numer. Math. , FJOURNAL =. 2005 , NUMBER =. doi:10.1016/j.apnum.2004.08.023 , URL =

    work page doi:10.1016/j.apnum.2004.08.023 2005

  25. [25]

    Mixed Finite Elements, Compatibility Conditions, and Applications , pages =

    Mixed finite element methods , author =. Mixed Finite Elements, Compatibility Conditions, and Applications , pages =. 2008 , publisher =

    2008

  26. [26]

    and Maschke, B

    Eberard, D. and Maschke, B. , Year =. Port. IFAC Proceedings Volumes , Volume =

  27. [27]

    2018 , Journal =

    On Structure-Preserving Model Reduction for Damped Wave Propagation in Transport Networks , Author =. 2018 , Journal =

    2018

  28. [28]

    2022 , Journal =

    Order theory for discrete gradient methods , Author =. 2022 , Journal =

    2022

  29. [29]

    and Meunier, S

    Ern, A. and Meunier, S. , TITLE =. ESAIM: Math. Model. Numer. Anal. , FJOURNAL =. 2009 , NUMBER =

    2009

  30. [30]

    and H\'elie, T

    Falaize, A. and H\'elie, T. , Year =. Passive simulation of the nonlinear port-. J. Sound Vib. , Volume =

  31. [31]

    and G\"unther, M

    Frommer, A. and G\"unther, M. and Liljegren-Sailer, B. and Marheineke, N. , Year =. Operator splitting for port-. Progress in Industrial Mathematics at ECMI 2023 , Publisher =

    2023

  32. [32]

    and Haller, F

    Gernandt, H. and Haller, F. E. and Reis, T. and van der Schaft, A. J. , Year =. Port-. J. Geom. Phys. , Volume =

  33. [33]

    and Karsai, A

    Giesselmann, J. and Karsai, A. and Tscherpel, T. , title =. 2025 , no-issn =

    2025

  34. [34]

    and Kumar, K

    Girault, V. and Kumar, K. and Wheeler, M. F. , TITLE =. Comput. Geosci. , FJOURNAL =. 2016 , NUMBER =

    2016

  35. [35]

    , Year =

    Gonzalez, O. , Year =. Time integration and discrete. J. Nonlinear Sci. , Volume =

  36. [36]

    and Paige, C

    Greif, C. and Paige, C. C. and Titley-Peloquin, D. and Varah, J. M. , Year =. Numerical equivalences among. SIAM J. Matrix Anal. Appl. , Volume =

  37. [37]

    and \"Ottinger, H

    Grmela, M. and \"Ottinger, H. C. , Year =. Dynamics and thermodynamics of complex fluids. Phys. Rev. E , Volume =

  38. [38]

    and Polyuga, R

    Gugercin, S. and Polyuga, R. V. and Beattie, C. and van der Schaft, A. J. , Year =. Structure-preserving tangential interpolation for model reduction of port-. Automatica , Volume =

  39. [39]

    and Marheineke, N

    Hauschild, S.-A. and Marheineke, N. and Mehrmann, V. and Mohring, J. and Moses Badlyan, A. and Rein, M. and Schmidt, M. , Year =. Port-. Progress in Differential--Algebraic Equations II , Pages =

  40. [40]

    1974 , Publisher =

    Differential Equations, Dynamical Systems, and Linear Algebra , Author =. 1974 , Publisher =

    1974

  41. [41]

    and Couenne, F

    Hoang, H. and Couenne, F. and Jallut, C. and Le Gorrec, Y. , Year =. The port. J. Process Contr. , Fjournal =

  42. [42]

    and Kraus, J

    Hong, Q. and Kraus, J. , TITLE =. Electron. Trans. Numer. Anal. , FJOURNAL =. 2018 , PAGES =. doi:10.1553/etna_vol48s202 , URL =

    work page doi:10.1553/etna_vol48s202 2018

  43. [43]

    and Kraus, J

    Hong, Q. and Kraus, J. and Lymbery, M. and Philo, F. , title =. Numer. Linear Algebr. , fjournal =. 2019 , number =. doi:10.1002/nla.2242 , url =

    work page doi:10.1002/nla.2242 2019

  44. [44]

    2014 , Publisher =

    Engineering Vibration , Author =. 2014 , Publisher =

    2014

  45. [45]

    and Tchelepi, H

    Kim, J. and Tchelepi, H. A. and Juanes, R. , TITLE =. Comput. Methods Appl. Mech. Engrg. , FJOURNAL =. 2011 , NUMBER =

    2011

  46. [46]

    and Tchelepi, H

    Kim, J. and Tchelepi, H. A. and Juanes, R. , TITLE =. Comput. Methods Appl. Mech. Engrg. , FJOURNAL =. 2011 , NUMBER =. doi:10.1016/j.cma.2011.02.011 , URL =

    work page doi:10.1016/j.cma.2011.02.011 2011

  47. [47]

    SPE Journal , volume =

    Stability, accuracy, and efficiency of sequential methods for coupled flow and geomechanics , author =. SPE Journal , volume =. 2011 , publisher =

    2011

  48. [48]

    Kinon, P. L. and Thoma, T. and Betsch, P. and Kotyczka, P. , Year =. Discrete nonlinear elastodynamics in a port-. Proc. Appl. Math. Mech. , Volume =

  49. [49]

    2019 , Publisher =

    Numerical Methods for Distributed Parameter Port-Hamiltonian Systems , Author =. 2019 , Publisher =

    2019

  50. [50]

    and Schiffer, J

    Krishna, A. and Schiffer, J. , Year =. A port-. IFAC-PapersOnLine , Volume =

  51. [51]

    Lamour, R. and M. Differential-Algebraic Equations: A Projector Based Analysis , pages =. doi:10.1007/978-3-642-27555-5 , isbn =

    work page doi:10.1007/978-3-642-27555-5

  52. [52]

    and Mascia, C

    Lattanzio, C. and Mascia, C. and Serre, D. , TITLE =. Hyperbolic problems: theory, numerics, applications , PAGES =. 2008 , DOI =

    2008

  53. [53]

    and Kotyczka, P

    Lohmayer, M. and Kotyczka, P. and Leyendecker, S. , Year =. Exergetic port-. Math. Comput. Model. Dyn. Syst. , Volume =

  54. [54]

    and Strehmel, K

    Lucht, W. and Strehmel, K. and Eichler-Liebenow, C. , TITLE =. BIT , FJOURNAL =. 1999 , NUMBER =. doi:10.1023/A:1022370703243 , URL =

    work page doi:10.1023/a:1022370703243 1999

  55. [55]

    2014 , Journal =

    Towards a port-based formulation of macro-economic systems , Author =. 2014 , Journal =

    2014

  56. [56]

    and Melchiorri, C

    Macchelli, A. and Melchiorri, C. , Year =. Modeling and control of the. SIAM J. Control Optim. , Volume =

  57. [57]

    A generalized finite element method for linear thermoelasticity , journal =

    M. A generalized finite element method for linear thermoelasticity , journal =

  58. [58]

    2021 , Journal =

    A two-level iterative scheme for general sparse linear systems based on approximate skew-symmetrizers , Author =. 2021 , Journal =

    2021

  59. [59]

    and van der Schaft, A

    Maschke, B. and van der Schaft, A. J. , Year =. Port-controlled. IFAC Proceedings Volumes , Volume =

  60. [60]

    1999 , Journal =

    Geometric integration using discrete gradients , Author =. 1999 , Journal =

    1999

  61. [61]

    and van der Schaft, A

    Mehrmann, V. and van der Schaft, A. J. , Year =. Differential--algebraic systems with dissipative. Math. Control Signals Syst. , Fjournal =

  62. [62]

    and Unger, B

    Mehrmann, V. and Unger, B. , Year =. Control of port-. Acta Numer. , Volume =

  63. [63]

    and Zwart, H

    Mehrmann, V. and Zwart, H. , Year =. Abstract dissipative. DAE Panel , Volume =

  64. [64]

    2011 , Journal =

    A gradient structure for reaction--diffusion systems and for energy-drift-diffusion systems , Author =. 2011 , Journal =

    2011

  65. [65]

    Convergence of iterative coupling for coupled flow and geomechanics , JOURNAL =

    Mikeli\'. Convergence of iterative coupling for coupled flow and geomechanics , JOURNAL =. 2013 , NUMBER =

    2013

  66. [66]

    , Year =

    Morandin, R. , Year =. Modeling and numerical treatment of port-

  67. [67]

    Mora, L. A. and Le Gorrec, Y. and Matignon, D. and Ramirez, H. and Yuz, J. I. , Year =. On port-. Phys. Fluids , Volume =

  68. [68]

    Mora, L. A. and Yuz, J. I. and Ramirez, H. and Le Gorrec, Y. , Year =. A port-. IFAC-PapersOnLine , Volume =

  69. [69]

    , title =

    Mujahid, A. , title =. GAMM Archive for Students , year =. doi:10.14464/gammas.v4i1.500 , number =

    work page doi:10.14464/gammas.v4i1.500

  70. [70]

    2001 , Journal =

    The geometry of dissipative evolution equations: the porous medium equation , Author =. 2001 , Journal =

    2001

  71. [71]

    and Matignon, D

    Payen, G. and Matignon, D. and Haine, G. , Year =. Modelling and structure-preserving discretization of. IFAC-PapersOnLine , Volume =

  72. [72]

    Polyuga, R. V. and van der Schaft, A. J. , Year =. Effort- and flow-constraint reduction methods for structure preserving model reduction of port-. Systems Control Lett. , Volume =

  73. [73]

    and Maschke, B

    Ramirez, H. and Maschke, B. and Sbarbaro, D. , Year =. Irreversible port-. Chem. Eng. Sci. , Volume =

  74. [74]

    and Califano, F

    Rashad, R. and Califano, F. and van der Schaft, A. J. and Stramigioli, S. , Year =. Twenty years of distributed port-. IMA J. Math. Control Inf. , Volume =

  75. [75]

    and Califano, F

    Rashad, R. and Califano, F. and Schuller, F. P. and Stramigioli, S. , Year =. Port-. J. Geom. Phys. , Volume =

  76. [76]

    van der Schaft, A. J. , Year =. Port-. Advanced Dynamics and Control of Structures and Machines , Publisher =

  77. [77]

    van der Schaft, A. J. and Jeltsema, D. , Year =. Port-. Found. Trends Syst. Control , Volume =

  78. [78]

    van der Schaft, A. J. , Year =. Port-. Proceedings of the international congress of mathematicians , Pages =

  79. [79]

    van der Schaft, A. J. , Year =. Port-. Surveys in Differential-Algebraic Equations I , Publisher =

  80. [80]

    , Year =

    Schulze, P. , Year =. Structure-preserving time discretization of port-. Progress in Industrial Mathematics at ECMI 2023 , Editor =

    2023

Showing first 80 references.