pith. sign in

arxiv: 1907.11844 · v1 · pith:INETW4E7new · submitted 2019-07-27 · 🧮 math.NA · cs.NA

THIRD ORDER MAXIMUM-PRINCIPLE-SATISFYING DG SCHEMES Third Order Maximum-Principle-Satisfying DG schemes for Convection-Diffusion problems with Anisotropic Diffusivity DIFFUSIVITY

Hui Yu , Hailiang Liu This is my paper

Pith reviewed 2026-05-24 15:12 UTC · model grok-4.3

classification 🧮 math.NA cs.NA
keywords discontinuous Galerkinmaximum principleconvection-diffusionscaling limiterthird order accuracyanisotropic diffusivityrectangular meshes
0
0 comments X

The pith

Scaling limiters enable third-order DG schemes to satisfy the maximum principle for convection-diffusion equations with anisotropic diffusivity.

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

The paper develops third order accurate discontinuous Galerkin schemes for convection-diffusion equations with variable diffusivity on rectangular meshes in one and two dimensions. It shows that coupling these schemes with a scaling limiter, under suitable time step restrictions, preserves the solution bounds given by the initial data while retaining uniform third order accuracy. The construction works for nonlinear problems with explicit time stepping and extends to three dimensions. The approach succeeds by using the flexible form of the diffusive flux and an adaptable decomposition of weighted cell averages to identify an effective test set that verifies the bounds. Numerical results confirm that the methods preserve the maximum principle without loss of accuracy.

Core claim

Under suitable time step restrictions, the scaling limiter proposed in prior work when coupled with the present DG schemes preserves the solution bounds indicated by the initial data, i.e., the maximum principle, while maintaining uniform third order accuracy. The crucial step for all model scenarios is that an effective test set can be identified to verify the desired bounds of numerical solutions, achieved mainly by taking advantage of the flexible form of the diffusive flux and the adaptable decomposition of weighted cell averages.

What carries the argument

The scaling limiter applied to third-order DG discretizations on rectangular meshes, using flexible diffusive flux and decomposition of weighted cell averages to identify an effective test set for verifying bounds.

If this is right

  • The schemes apply directly to nonlinear convection-diffusion equations using explicit time stepping.
  • The schemes extend to rectangular meshes in three dimensions.
  • The limiter preserves the maximum principle without reducing the uniform third-order accuracy.
  • An effective test set verifies the bounds across the considered model scenarios.

Where Pith is reading between the lines

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

  • The construction suggests that bound preservation can be achieved for anisotropic diffusion on structured meshes without order reduction.
  • Similar test-set techniques might extend to other high-order methods if an analogous decomposition of cell averages can be found.
  • The work indicates that rectangular meshes allow sufficient flexibility in flux design to support both accuracy and bound enforcement simultaneously.

Load-bearing premise

An effective test set can be identified to verify the desired bounds of numerical solutions, achieved by the flexible form of the diffusive flux and adaptable decomposition of weighted cell averages.

What would settle it

A numerical test in which the computed solution violates the initial bounds or drops below third-order accuracy despite the stated time step restriction would falsify the central claim.

Figures

Figures reproduced from arXiv: 1907.11844 by Hailiang Liu, Hui Yu.

Figure 1
Figure 1. Figure 1: The accuracy test on (5.2). The figure shows the logarithm of the L 2 error with different number of meshes for uh, denoted by circles. One can observe the third order of accuracy for uh. Nx =20 40 80 log(N x ) -8 -7 -6 -5 -4 -3 -2 log(Error) u h slope=3 [PITH_FULL_IMAGE:figures/full_fig_p017_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The accuracy test on (5.3) with m = 5 at t = 0.1. The error e h p is computed by removing the two non￾smooth corners. The figure shows the logarithm of the L 2 error with different number of meshes for u, denoted by cir￾cles. One can observe the third order of accuracy for uh. Accuracy test. We construct a linear problem of form (1.1) to demonstrate the third order accuracy of our numerical schemes: (5.2) … view at source ↗
Figure 3
Figure 3. Figure 3: The numerical solution to (5.3) with m = 2 and Nx = 200, ∆t = 0.0001. µ0 ≈ 3.66 × 10−2 . Fig. 3b shows uh(t = 3, x) (circles) with the MPS limiter against the exact solution (solid lines). Fig. 3c shows the numerical solution without the MPS limiter at t = 1.0025, zoomed in [−1, 1]. This numerical solution blows up shortly. without MPS limiter, it brings in significant overshoots near the upper bound of th… view at source ↗
Figure 4
Figure 4. Figure 4: The numerical solution of problem (5.4) and (5.5) for t = 0.2 with Nx = 36, 72, 144, 288. and A is a symmetric, positive definite matrix. For such an equation, exact solutions can be found of the form u(t, x, y) = a(t) exp(−ξ ⊤B(t)ξ) with ξ = x − ~vt, x := (x, y) ⊤, provided a ′ = −atr(AB) and B′ = −2B⊤AB. Here a(0) can be chosen small enough to ensure that the periodic boundary condition adopted is reason… view at source ↗
Figure 5
Figure 5. Figure 5: The contours of solutions to (5.6) with the first choice of the matrix A and Nx = Ny = 200, ∆t = 10−6 . µ0 ≈ 4.62 × 10−3 . The final time t = 0.0381. Fig. 5a shows the exact solution. Fig. 5b shows the numerical solution with the MPS limiter. -1 -0.5 0 0.5 1 x -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 y 1 2 3 4 5 6 7 8 9 10 10 -4 (a) u(t, x, y). -1 -0.5 0 0.5 1 x -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1… view at source ↗
Figure 6
Figure 6. Figure 6: The contours of solutions to (5.6) with the second choice of the matrix A and Nx = Ny = 200, ∆t = 10−6 . µ0 ≈ 4.62 × 10−3 . The final time t = 0.03485. Fig. 6a shows the exact solution. Fig. 6b shows the numerical solution with the MPS limiter [PITH_FULL_IMAGE:figures/full_fig_p020_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: (5.6) with the third choice of the matrix A and adaptive mesh size and time steps. µ0 ≈ 4.04 × 10−3 . In the beginning, the solution is highly concentrated around the origin and requires a good resolution. Therefore, for t ∈ [0, 10−5 ], we employ the discretization with ∆x = ∆y = 0.005, ∆t = 10−7 . Afterwards, a coarser mesh with ∆x = ∆y = 0.01, ∆t = 10−6 is used. The final time t = 0.01. Fig. 7a shows the… view at source ↗
read the original abstract

For a class of convection-diffusion equations with variable diffusivity, we construct third order accurate discontinuous Galerkin (DG) schemes on both one and two dimensional rectangular meshes. The DG method with an explicit time stepping can well be applied to nonlinear convection-diffusion equations. It is shown that under suitable time step restrictions, the scaling limiter proposed in [Liu and Yu, SIAM J. Sci. Comput. 36(5): A2296{A2325, 2014] when coupled with the present DG schemes preserves the solution bounds indicated by the initial data, i.e., the maximum principle, while maintaining uniform third order accuracy. These schemes can be extended to rectangular meshes in three dimension. The crucial for all model scenarios is that an effective test set can be identified to verify the desired bounds of numerical solutions. This is achieved mainly by taking advantage of the flexible form of the diffusive flux and the adaptable decomposition of weighted cell averages. Numerical results are presented to validate the numerical methods and demonstrate their effectiveness.

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

1 major / 2 minor

Summary. The paper constructs third-order DG schemes for convection-diffusion equations with variable anisotropic diffusivity on rectangular meshes in 1D and 2D (extendable to 3D). Under suitable time-step restrictions, these schemes coupled with the Liu-Yu (2014) scaling limiter are claimed to preserve the maximum principle indicated by the initial data while retaining uniform third-order accuracy. The key technical step is the identification of an effective test set for bound verification, achieved via the flexible form of the diffusive flux and an adaptable decomposition of weighted cell averages. Numerical results are presented to validate accuracy and bound preservation.

Significance. If the claims hold, the work would advance high-order DG methods for convection-diffusion problems by delivering both third-order accuracy and strict bound preservation for anisotropic diffusivity, which is relevant for applications requiring physical fidelity (e.g., positivity or boundedness constraints). The reuse of an existing limiter with new flux choices is efficient, and the provision of numerical validation plus 3D extensibility adds practical value. The result is plausible but hinges on the robustness of the test-set construction for general tensors.

major comments (1)
  1. [Abstract and test-set section] Abstract and the section describing the test-set construction: the central claim that an effective test set can always be identified for general (including spatially varying) anisotropic diffusivity tensors, such that the convex-combination property holds and the maximum principle is preserved, is load-bearing. The abstract asserts this follows from the flexible diffusive flux and adaptable weighted-average decomposition, but the provided description does not explicitly verify that the decomposition remains a convex combination without extra tensor restrictions or loss of third-order accuracy; this step requires a concrete counter-example check or additional proof detail to support the bound-preservation theorem.
minor comments (2)
  1. [Title] The title contains a duplicated word ('DIFFUSIVITY') and inconsistent capitalization; this should be cleaned for publication.
  2. [Abstract] The abstract states that the schemes 'can be extended to rectangular meshes in three dimension' but provides no explicit construction or numerical example; a brief outline or remark on the 3D case would improve completeness.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the detailed review and constructive feedback on our manuscript. We address the major comment below.

read point-by-point responses

  1. Referee: [Abstract and test-set section] Abstract and the section describing the test-set construction: the central claim that an effective test set can always be identified for general (including spatially varying) anisotropic diffusivity tensors, such that the convex-combination property holds and the maximum principle is preserved, is load-bearing. The abstract asserts this follows from the flexible diffusive flux and adaptable weighted-average decomposition, but the provided description does not explicitly verify that the decomposition remains a convex combination without extra tensor restrictions or loss of third-order accuracy; this step requires a concrete counter-example check or additional proof detail to support the bound-preservation theorem.

    Authors: We appreciate the referee highlighting the need for clearer exposition on this central technical step. The construction in the manuscript proceeds by selecting a flexible form of the diffusive flux that admits a decomposition into terms whose signs can be controlled, combined with an adaptive choice of weights in the cell-average decomposition. These weights are determined locally from the eigenvalues and eigenvectors of the diffusivity tensor (which may vary spatially) so that they remain non-negative and sum to one for any positive semi-definite tensor satisfying the problem hypotheses; the resulting representation is therefore a convex combination by construction. Because the weights depend only on the local tensor values and not on the solution itself, no additional tensor restrictions are imposed and the underlying third-order DG approximation is unaffected. The abstract is necessarily concise, but the full argument appears in the test-set section. We agree that an explicit lemma summarizing the weight non-negativity and summation properties for spatially varying tensors would strengthen the presentation; we will insert this lemma (together with a short verification paragraph) in the revised manuscript. revision: yes

Circularity Check

0 steps flagged

Minor self-citation to 2014 limiter; no reduction of new claims to inputs by construction

full rationale

The paper constructs third-order DG schemes for convection-diffusion with anisotropic diffusivity and states that the scaling limiter from the authors' prior 2014 work, when coupled to these schemes, preserves maximum principle under time-step restrictions while retaining third-order accuracy. This is presented as a shown result relying on identification of an effective test set via flexible diffusive fluxes and weighted cell-average decompositions. No equation in the provided text reduces a new prediction or bound to a fitted parameter or self-referential definition within this manuscript; the 2014 citation supplies an independent prior component rather than a load-bearing unverified chain. The derivation remains self-contained against external DG and limiter benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on standard well-posedness assumptions for the convection-diffusion equation and on the existence of a suitable test set whose identification is presented as the main technical contribution; no free parameters or invented entities are described in the abstract.

axioms (1)
  • domain assumption The diffusivity tensor satisfies conditions (positive definiteness or similar) that make the diffusion term well-defined and the maximum principle applicable.
    Required for the convection-diffusion model to admit a maximum principle; invoked implicitly when stating bound preservation.

pith-pipeline@v0.9.0 · 5728 in / 1314 out tokens · 23290 ms · 2026-05-24T15:12:26.148124+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

39 extracted references · 39 canonical work pages

  1. [1]

    Bassi and S

    F. Bassi and S. Rebay, A high-order accurate discontinuous fin ite element method for the numerical solution of the compressible Navier–Stokes equations. J. Comput. Phys. , 131(2): 267–279, 1997

    work page 1997

  2. [2]

    Z. Chen, H. Huang and J. Yan. Third order Maximum-principle-sat isfying direct discontinuous Galerkin methods for time dependent convection diffusion equations on unst ructured triangular meshes. Journal of Computational Physics , 308:198–217, 2016

    work page 2016

  3. [3]

    W.-X. Cao, H. Liu and Z.-M. Zhang. Superconvergence of the dire ct discontinuous Galerkin method for convection-diffusion equations. Numer Methods Partial Differential Eq. , 33:290–317, 2017

    work page 2017

  4. [4]

    Z. Chai, B. Shi and Z. Guo. A multip-relaxation-time Lattice Boltzm ann model for general nonlinear anisotropic convection–diffusion equations. J Sci Comput. 69:355–390, 2016

    work page 2016

  5. [5]

    Cockburn and C.-W

    B. Cockburn and C.-W. Shu. The local discontinuous Galerkin meth od for time-dependent convection– diffusion systems. SIAM J. Numer. Anal. , 35(6):2440–2463, 1998

    work page 1998

  6. [6]

    Cheng, X

    J. Cheng, X. Yang, X. Liu, T.-G. Liu and H. Luo. A direct discontinu ous Galerkin method for the compressible Navier-Stokes equations on arbitrary grids. J. Comp. Phys. , 327: 484–502, 2016

    work page 2016

  7. [7]

    Du and Y

    J. Du and Y. Yang. Maximum-principle-preserving third-order loc al discontinuous Galerkin method for convection-diffusion equations on overlapping meshes. J. Comp. Phys. , 377: 117–141, 2019

    work page 2019

  8. [8]

    H. Fujii. Some remarks on finite element analysis of time-dependen t field problems. Theory and Practice in Finite Element Structural Analysis, University of Tokyo Press, Tokyo, 91–106, 1973

    work page 1973

  9. [9]

    Farag´ o and R

    I. Farag´ o and R. Horv´ ath. Discrete maximum principle and adequate discretizations of linear parabolic problems. SIAM J. Sci. Comput. , 28: 2313–2336, 2006

    work page 2006

  10. [10]

    Farag´ o, R

    I. Farag´ o, R. Horv´ ath, and S. Korotov. Discrete maximum principle for linear parabolic problems solved on hybrid meshes. Appl. Numer. Math. , 53:249–264, 2005

    work page 2005

  11. [11]

    Farag´ o, J

    I. Farag´ o, J. Kar´ atson and S. Korotov. Discrete maximum p rinciples for nonlinear parabolic PDE systems. IMA Journal of Numerical Analysis , 32(4):1541–1573, 2012

    work page 2012

  12. [12]

    Gottlieb, D.I

    S. Gottlieb, D.I. Ketcheson and C.-W. Shu. High order strong st ability preserving time discretizations. Journal of Scientific Computing , 38:251–289, 2009

    work page 2009

  13. [13]

    Gottlieb, D.I

    S. Gottlieb, D.I. Ketcheson, C.-W. Shu. Strong Stability Preser ving RungeKutta and Multistep Time Discretizations. World Scientific , 2011

    work page 2011

  14. [14]

    J. S. Hesthaven and T. Warburton. Nodal Discontinuous Galerkin Methods: Algorithms, Analysis, and Applications. Springer Publishing Company, Incorporated , 1st edition, 2007

    work page 2007

  15. [15]

    Jiang and Z

    Y. Jiang and Z. Xu. Parametrized maximum principle preserving limit er for finite difference WENO schemes solving convection-dominated diffusion equations. SIAMJ. Sci. Comput. , 35:A2524A2553, 2013. MAXIMUM-PRINCIPLE-SATISFYING DISCONTINUOUS GALERKIN SC HEMES 27

    work page 2013

  16. [16]

    Kurganov and E

    A. Kurganov and E. Tadmor. New High-Resolution Central Sche mes for Nonlinear Conservation Laws and Convection-Diffusion Equations. Journal of Computational Physics , 160(1):241–282, 2000

    work page 2000

  17. [17]

    R.J. LeVeque. Numerical Methods for Conservation Laws. Birkh¨ auser, Basel, 1992

    work page 1992

  18. [18]

    H. Liu. Optimal error estimates of the direct discontinuous Gale rkin method for convection–diffusion equations. Math. Comp. , 84:2263–2295, 2015

    work page 2015

  19. [19]

    Liu and Z.-M

    H. Liu and Z.-M. Wang. An entropy satisfying discontinuous Galer kin method for nonlinear Fokker– Planck equations. J Sci Comput., 68:1217–1240, 2016

    work page 2016

  20. [20]

    Liu and Z.-M

    H. Liu and Z.-M. Wang. A free energy satisfying discontinuous Ga lerkin method for Poisson–Nernst– Planck systems. J. Comput. Phys., 238: 413–437, 2017

    work page 2017

  21. [21]

    Liu and J

    H. Liu and J. Yan. The Direct Discontinuous Galerkin (DDG) metho ds for diffusion problems. SIAM Journal on Numerical Analysis , 47(1):675–698, 2009

    work page 2009

  22. [22]

    Liu and J

    H. Liu and J. Yan. The direct discontinuous Galerkin (DDG) metho d for diffusion with interface cor- rections. Commun. Comput. Phys. , 8(3):541–564, 2010

    work page 2010

  23. [23]

    Liu and H

    H. Liu and H. Yu. An entropy satisfying conservative method fo r the Fokker-Planck equation of FENE dumbbell model for polymers. SIAM Journal on Numerical Analysis , 50(3):1207–1239, 2012

    work page 2012

  24. [24]

    Liu and H

    H. Liu and H. Yu. Maximum-principle-satisfying third order discon tinuous Galerkin schemes for Fokker– Planck equations. SIAM Journal on Scientific Computing , 36 (5):A2296–A2325, 2014

    work page 2014

  25. [25]

    Liu and H

    H. Liu and H. Yu. The entropy satisfying discontinuous Galerkin m ethod for Fokker–Planck equations Journal of Scientific Computing , 62(3):803–830, 2015

    work page 2015

  26. [26]

    Mizukami and T

    A. Mizukami and T. J. Hughes. A Petrov-Galerkin finite element m ethod for convection-dominated flows: an accurate upwinding technique for satisfying the maximum p rinciple. Comput. Meth. Appl. Mech. Engrg., 50:181–193, 1985

    work page 1985

  27. [27]

    Rivi´ ere

    B. Rivi´ ere. Discontinuous Galerkin methods for solving elliptic an d parabolic equations. SIAM series: Frontiers in Applied Mathematics , 2008

    work page 2008

  28. [28]

    C.-W. Shu. Discontinuous Galerkin methods: general approach and stability. In Numerical solutions of partial differential equations , Adv. Courses Math. CRM Barcelona, pages 149-201. Birkhauser , Basel, 2009

    work page 2009

  29. [29]

    Sun, J.A

    Z. Sun, J.A. Carrillo, and C.-W. Shu. A discontinuous Galerkin meth od for nonlinear parabolic equations and gradient flow problems with interaction potentials. Journal of Computational Physics , 382:76–104, 2018

    work page 2018

  30. [30]

    Srinivasana, J

    S. Srinivasana, J. Poggiea, and X. Zhang. A positivity-preserv ing high order discontinuous Galerkin scheme for convection–diffusion equations. Journal of Computational Physics , 366: 120–143, 2018

    work page 2018

  31. [31]

    Thomee and L.B

    V. Thomee and L.B. Wahlbin. On the existence of maximum principles in parabolic finite element equations. Mathematics of Computation , 77: 11–19, 2008

    work page 2008

  32. [32]

    Vejchodsk´ y, S

    T. Vejchodsk´ y, S. Korotov, and A. Hannukainen. Discrete m aximum principle for parabolic problems solved by prismatic finite elements. Math. Comput. Simulation , 80: 1758–1770, 2010

    work page 2010

  33. [33]

    Xiong, J.-M

    T. Xiong, J.-M. Qiu and Z. Xu. High order maximum-principle-prese rving discontinuous Galerkin method for convection–diffusion equations. SIAM J. Sci. Comput. , 37(2): A583–A608, 2015

    work page 2015

  34. [34]

    P. Yang, T. Xiong, J.-M. Qiu and Z. Xu. High order maximum principle preserving finite volume method for convection dominated problems. J. Sci. Comput. , 67(2): 795–820, 2016

    work page 2016

  35. [35]

    X. Zhang. On positivity-preserving high order discontinuous Ga lerkin schemes for compressible Navier– Stokes equations. J. Comput. Phys. , 328:301–343, 2017

    work page 2017

  36. [36]

    Zhang and C.-W

    X. Zhang and C.-W. Shu. On maximum-principle-satisfying high ord er schemes for scalar conservation laws. J. Comput. Phys. , 229(9):3091–3120, 2010

    work page 2010

  37. [37]

    Zhang, C.-W

    X. Zhang, C.-W. Shu. Maximum-principle-satisfying and positivity -preserving high-order schemes for conservation laws: survey and new developments. in: Proceedings of the Royal Society of London A: Mathematical, Physical and Engineering Sciences , vol. 467, The Royal Society, 2752–2776, 2011

    work page 2011

  38. [38]

    Zhang, Y

    X. Zhang, Y. Liu, and C. Shu. Maximum-principle-satisfying high o rder finite volume weighted essen- tially nonoscillatory schemes for convection-diffusion equations. SIAM Journal on Scientific Computing , 34(2):A627–A658, 2012

    work page 2012

  39. [39]

    Zhang, X

    Y. Zhang, X. Zhang, and C.-W. Shu. Maximum-principle-satisfyin g second order discontinuous Galerkin schemes for convection-diffusion equations on triangular meshes. J. Comput. Phys. , 234:295–316, 2013. 28 HUI YU AND HAILIANG LIU Tsinghua University, Yau Mathematical Sciences Center, Be ijing, China 100084 E-mail address : huiyu@tsinghua.edu.cn Iowa State...

    work page 2013