Pith. sign in

REVIEW 2 major objections 5 minor 27 references

An energy-based discontinuous Galerkin method for the wave equation with nonsmooth solutions

T0 review · 2 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper develops an oscillation-free energy-based discontinuous Galerkin method for the second-order wave equation, proves energy stability and a priori error estimates for smooth solutions, and gives numerical evidence that the method…

desk verdict Useful new OF-EDG scheme with a clean linear stability proof, but the nonlinear nonsmooth claims run ahead of the theory. read the letter →

arxiv 2507.01736 v1 pith:634P5YZT submitted 2025-07-02 math.NA cs.NA

Yangxin Fu , Yan Jiang , Siyang Wang This is my paper
classification math.NAcs.NA MSC 65M6065M1265M1535L05
keywords discontinuousGalerkinmethodwaveequationnonsmoothsolutionoscillation-freeenergystabilityapriorierrorestimatesemilinearSSP-RK3
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper aims to give the energy-based discontinuous Galerkin method for second-order wave equations the ability to handle discontinuous solutions. It adds two mechanisms to the standard EDG scheme: cell-wise damping terms, scaled by jumps of derivatives, that activate near discontinuities, and a jump penalty on the solution itself that prevents piecewise-constant initial data from stalling. On a periodic interval and on Cartesian meshes in two dimensions the resulting OF-EDG scheme is shown to dissipate the discrete energy $\int ((\partial_x u_h)^2 + v_h^2)\,dx$ for every choice of the flux parameters with nonnegative coefficients, and to satisfy an a priori error estimate in the energy norm. Numerical tests exhibit optimal convergence rates for smooth solutions and oscillation-free profiles for discontinuous linear and semilinear waves. If correct, the method is a high-order DG option for nonsmooth wave propagation that needs no limiter and keeps its stability proof.

What carries the argument

The workhorse is the semi-discrete OF-EDG scheme (2.11), written in first-order form with $v_h$ approximating $u_t$, using the general EDG flux family that contains the central, alternating, and Sommerfeld fluxes as special choices. Two nonlinear dissipation layers are added per cell: oscillation-free damping terms with coefficients $\sigma_j^l$ and $\tilde\sigma_j^l$ proportional to jumps of the $l$-th derivatives of $u_h$ and $v_h$ at the two cell interfaces, projected through the local $L^2$ projection $P_{l-1}$, and an interior-penalty term with strength $c/h^2$ acting on the jump of $u_h$ itself. The damping terms vanish for smooth cells and switch on near discontinuities, while the penalty term provides the force that piecewise-constant solutions require. The energy identity obtained from testing with $(u_h,v_h)$ reduces all interface and damping contributions to nonpositive squares, which is exactly what produces the stability inequality and the error estimate.

What would settle it

Compute the discrete energy $E_h(t)$ from (4.56) while running Example 4.5 with $g(u)=4u^3$ and piecewise constant initial data; because $G(u)=-u^4<0$ the hypothesis of the energy estimate fails, so any measured growth of $E_h$ would show the nonlinear examples are outside the theorem's coverage, and a comparison spike in the OF-EDG profile against a fine CTCS reference would refute the oscillation-free claim for that test.

Watch

Extended reading notes

Core claim

The central claim is that combining oscillation-free damping with an interior-penalty term inside the EDG framework yields a scheme that is simultaneously energy-stable, high-order accurate, and effectively non-oscillatory for nonsmooth wave solutions. Stability holds semidiscretely for any nonnegative penalty parameter $c$ and any nonnegative flux coefficients $\tau,\beta$, because every added term contributes a nonpositive jump or projection residual to the rate of change of $E_h$. The error analysis gives the energy error bound of order $h^{2\gamma}$ with exponent $\gamma=\min(p',q')$, where $p'$ is $p-1$ or $p-1/2$ depending on whether $\tau=0$ and $q'$ is $q$ or $q+1/2$ depending on whether $\beta=0$, under the polynomial-degree restriction $p-2\le q\le p$. The authors further report that in practice the alternating and Sommerfeld fluxes reach the optimal rate $p+1$ for displacement, while the central flux reaches it for odd polynomial degrees only. For discontinuous data, the penalty term moves piecewise-constant values that pure damping would leave untouched, and the damping term removes the spurious oscillations the penalty alone would create.

Load-bearing premise

For the nonlinear wave tests the stability statement is inherited from a semilinear energy argument that assumes $G(u)=-\int_0^u g(z)\,dz>0$, but the test sources $g(u)=160\sin u$ and $g(u)=4u^3$ violate that condition, and the paper does not acknowledge the mismatch, leaving those oscillation-free results resting on numerical observation rather than on the proved estimate.

Editorial extensions

If this is right

  • Stability of the semidiscrete scheme holds for every member of the flux family with $c,\tau,\beta\ge0$, so users can choose an energy-conserving flux such as the alternating flux or a dissipative Sommerfeld flux without losing the energy bound.
  • For smooth solutions the energy error is $O(h^{2\gamma})$ with $\gamma=\min(p',q')$, and the reported experiments show optimal $p+1$ convergence for the alternating and Sommerfeld fluxes and for the central flux with odd $p$.
  • For piecewise-constant initial data whose jumps lie on cell interfaces, the damping terms alone do nothing and the penalty term is what drives the scheme to the correct solution; the full OF-EDG combination then suppresses the oscillations near the jumps.
  • The same stability and error analysis extends to two-dimensional problems on Cartesian meshes, and the numerical examples show the oscillation-free behavior persists there.
  • For semilinear sources the scheme with $\chi=1$ inherits an energy-stable treatment under $G(u)>0$, and the two-dimensional runs use $\chi=0$ with empirically stable results.

Reading between the lines

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

  • The paper leaves the discrete-in-time analysis untouched; the reported SSP-RK3 step sizes are empirical matchings of the spatial accuracy, so a fully discrete stability proof and CFL condition are a natural next step rather than a proven property.
  • Because the damping coefficients are built from jumps of derivatives, they double as a discontinuity indicator; one could plausibly use their magnitude to drive h-refinement or p-adaptivity, an application the authors do not mention.
  • For nonlinear sources with $G(u)\le0$, such as the two discontinuous tests in Examples 4.4 and 4.5, the stability theorem does not apply; a plausible working hypothesis is that dissipation from the OF terms, not the semilinear energy, is what keeps those runs bounded, which could be tested by switching the damping and penalty terms off.
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

2 major / 5 minor

Summary. This paper proposes an oscillation-free energy-based discontinuous Galerkin (OF-EDG) method for the second-order wave equation. The spatial discretization augments the EDG scheme of Appelö and Hagstrom with an interior-penalty jump term and with projection-based damping terms whose coefficients depend on jumps of the numerical solution. The authors prove a semi-discrete energy-stability estimate for general fluxes in one and two dimensions, derive a priori error estimates under H^{p+1}/H^{q+1} regularity, and present numerical experiments for smooth and discontinuous solutions, including semilinear sine-Gordon and Klein-Gordon source terms. The paper claims optimal convergence for smooth solutions and oscillation-free behavior for nonsmooth solutions, with no fitted parameters used to produce the reported rates.

Significance. If the claims hold, the method is a practically attractive high-order DG scheme for wave propagation with discontinuous data: the stability proof is clean and self-contained, the damping and penalty design is clearly motivated, and the smooth convergence tests against exact solutions support the analysis. The numerical evidence for oscillation suppression is visually convincing. However, the paper's central robustness claim for nonlinear source terms currently rests on examples that violate the stated semilinear stability hypothesis, and the error-analysis proof contains a regularity gap; these issues make the overall significance conditional until repaired.

major comments (2)
  1. [Remark 4.1, Eq. (4.56), Examples 4.4-4.5 and 4.7-4.8] The semilinear stability statement in Remark 4.1 is invoked for numerical tests whose source terms do not satisfy its hypothesis G(u)>0. For g(u)=160 sin(u) one has G(u)=160(cos(u)-1)<=0, and for g(u)=4u^3 one has G(u)=-u^4<=0; thus the energy estimate (4.56) is not available for Examples 4.4, 4.5, 4.7, or 4.8. Moreover, the two-dimensional versions are run with chi=0, for which no stability statement is given anywhere in the paper. Since the abstract advertises robustness for nonsmooth solutions with nonlinear source terms, the paper should either add discontinuous semilinear tests satisfying G>0, extend the analysis to chi=0, or explicitly state that the nonlinear nonsmooth results are empirical only.
  2. [Section 2.4, Eqs. (2.27)-(2.31); Section 3.3, Eq. (3.51)] The proof of Theorem 2.2 contains a step that is not justified under the stated H^{p+1} regularity. In (2.27) the authors bound ||(u_x-P_{l-1}u_x)||_{L2(I_j)} first by h^{max(1,l)}|u|_{H^{max(1,l)+1}(I_j)}, then by a quantity involving |∂^{max(1,l)+1}u|∞, and finally drop that L∞ seminorm, writing the result as O(h^{max(1,l)+1/2}) with no factor containing u. Without an additional W^{p+1,∞} assumption or a careful elementwise Sobolev-embedding argument with constants under control, the chain does not follow from H^{p+1} regularity. The same issue appears in the two-dimensional estimate (3.51). Because this chain feeds directly into (2.31) and hence into the Gronwall argument, the a priori error bound as stated is not fully proven.
minor comments (5)
  1. [Abstract and Section 1] The abstract contains the typo 'apriori error estimates'; it should be 'a priori error estimates'. Also, in the Introduction, 'This rest of the paper' should be 'The rest of the paper'.
  2. [Remark 4.1, Eq. (4.55)] In the third displayed equation of (4.55), the term 'qP' should be a summation symbol or explicitly written as a sum over l=0,...,q; as printed it is not readable.
  3. [Examples 4.4 and 4.5] The initial data are written as 'u(t,0)=...', but they are initial conditions in space and should be 'u(x,0)=...'.
  4. [Examples 4.6-4.8 and Figure 4.6] The comparison with the CTCS reference is only visual; no quantitative error or mesh-convergence data is reported for the nonsmooth nonlinear tests, and the choice of damping/penalty parameters for those runs is not described. Reporting such details would make the robustness claims easier to reproduce and assess.
  5. [Section 2.4 and Section 4.1] Theorem 2.2 is stated in the energy norm, while the convergence tests report the L2 error of u_h; the indirect relationship between the two is only mentioned in Remark 2.2 and is not used to compare theory with the observed rates.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the stability theorem and error bounds are proved directly from the scheme's definition, and the smooth benchmark tests compare against exact solutions.

full rationale

The derivation chain for the paper's central claims is self-contained. The semi-discrete OF-EDG scheme (2.11) is defined with explicit damping coefficients (2.13), and Theorem 2.1 proves dE_h/dt <= 0 by direct substitution of the general numerical fluxes (2.8), using only sigma, sigma-tilde >= 0 and c, tau, beta >= 0. Theorem 2.2 derives the a priori error estimate in the energy norm via projections, approximation estimates, inverse inequalities, and Gronwall's inequality; no convergence rate is obtained by fitting or by assuming the target rate. The smooth numerical tests (Examples 4.1, 4.2, and 4.6) compare against exact solutions, so the reported rates are independent external benchmarks. Citations of the EDG lineage [1,3] are prior independent work and are not used to define or prove the present scheme; the stability proof here does not rest on those citations. The nonlinear extension in Remark 4.1 inherits an energy estimate from [3] under a G>0 hypothesis that is violated by Examples 4.4-4.5 and 4.7-4.8; for instance, g(u)=160 sin u gives G(u)=160(cos u - 1) <= 0. Remark 4.2 also concedes that the chi=1 scheme is sensitive to damping and penalty parameters and requires careful adjustment. These are correctness and robustness concerns, not circularity, because the oscillation-free numerical claims are presented as visual comparisons against a CTCS reference rather than as consequences obtained by construction from fitted parameters. The self-citation in reference [4] is not load-bearing for any theorem or prediction in this paper. Overall circularity score: 0.

Assumptions & free parameters 4 free parameters · 4 assumptions · 1 invented entities

The central claims rest on standard finite element approximation theory, a periodic-boundary setup, and an inherited nonlinear stability condition that is violated by two of the reported nonlinear examples. No new physics is postulated. Method parameters such as c, s, chi, and the time-step prefactor are chosen by hand and not fitted to exact solutions.

free parameters (4)
  • Penalty parameter c = 1 in all numerical tests
    Chosen by hand; theorems only require c >= 0. The h^{-2} scaling is load-bearing for not degrading smooth convergence, but the value itself is not fitted to data.
  • Sommerfeld flux parameter s = not reported; only s > 0 is stated in (2.9)
    Needed to run the S-flux tests, but its numerical value is absent, reducing reproducibility of Figures 4.2, 4.3, 4.4, and 4.8.
  • Nonlinear coupling parameter chi = chi = 1 in 1D, chi = 0 in 2D
    Scheme (4.55) defines two variants. No analysis justifies chi = 0, and Remark 4.2 notes chi = 1 requires careful adjustment of damping and penalty parameters.
  • Time-step prefactor 1/20 and accuracy-matching exponents = dt = h/20 for p=2, h^{4/3}/20 for p=3, h^{5/3}/20 for p=4, h^2/20 for p=5, h^{7/3}/20 for p=6
    Chosen by hand to match spatial accuracy order. This is standard practice, but the observed convergence rates depend on this temporal resolution choice.
assumptions (4)
  • standard math Finite element approximation, inverse, and projection estimates (2.20)-(2.23) hold uniformly on quasi-uniform meshes.
    Used throughout the proofs of Theorems 2.2 and 3.2; cited to Ciarlet [8] and standard FEM theory.
  • domain assumption Periodic boundary conditions are assumed in all stability and error theorems.
    Section 2.1 states periodic conditions and says other conditions are handled readily, but no proof is given. Several numerical examples (for example Example 4.2) use Neumann conditions.
  • domain assumption The semilinear energy estimate from [3] applies to the nonlinear examples, requiring G(u) = -integral_0^u g(z) dz > 0 and lim_{u->0} g(u)/u bounded.
    Remark 4.1 invokes this condition, but Examples 4.4 and 4.5 use source terms for which G is non-positive, so the assumption is silently violated in the reported tests.
  • ad hoc to paper The pointwise derivative seminorm in (2.27) is finite and behaves as written for the exact solution.
    The bound on the damping term in (2.26)-(2.31) relies on |partial^{max(1,l)+1}u|_infty being controlled, which is stronger than the H^{p+1} assumption stated in Theorem 2.2.
invented entities (1)
  • No new physical or mathematical entities
    purpose: The paper introduces a numerical scheme, not a new particle, force, dimension, or conserved quantity.
    The ledger has no invented entities; the OF damping and penalty are algorithmic mechanisms, not added physical degrees of freedom.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An energy-based discontinuous Galerkin method for the wave equation with nonsmooth solutions." pith.science (2026). https://pith.science/paper/634P5YZT

@misc{pith2026250701736,
  author       = {Pith},
  title        = {Pith review of: An energy-based discontinuous Galerkin method for the wave equation with nonsmooth solutions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/634P5YZT}},
  note         = {Machine review of arXiv:2507.01736}
}
read the original abstract

We develop a stable and high-order accurate discontinuous Galerkin method for the second order wave equation, specifically designed to handle nonsmooth solutions. Our approach integrates the energy-based discontinuous Galerkin method with the oscillation-free technique to effectively suppress spurious oscillations near solution discontinuities. Both stability analysis and apriori error estimates are established for common choices of numerical fluxes. We present a series of numerical experiments to confirm the optimal convergence rates for smooth solutions and its robustness in maintaining oscillation-free behavior for nonsmooth solutions in wave equations without or with nonlinear source terms.

Figures

Figures reproduced from arXiv: 2507.01736 by the authors.

Figure 3.1
Figure 3.1. Illustration of the jump term in (3.43), where the red solid point represents a vertex v of Kij , and the jumps on the faces Kij ∩ Ki−1,j and Kij ∩ Ki,j+1 are used in the definition. the vertices of Kij . Let [[w]]|v denote the jump of w on element Kij and its adjacent elements at vertex v. As illustrated in [PITH_FULL_IMAGE:figures/full_fig_p018_3_1.png] view at source ↗
Figure 4.2
Figure 4.2. Example 4.1: L 2 errors with different numerical fluxes on uniform meshes. We have also performed the same convergence test on nonuniform meshes, created by randomly perturbing all internal nodes on a uniform mesh by up to 10% of its mesh size. In [PITH_FULL_IMAGE:figures/full_fig_p025_4_2.png] view at source ↗
Figure 4.3
Figure 4.3. Example 4.1: L 2 errors with different numerical fluxes on nonuniform meshes. g(u) = − sin u on domain Ω = (−40, 40), and apply the homogeneous Neumann bound￾ary conditions with the following initial conditions,    utt = uxx − sin u, u(x, 0) = 4 arctan √ 0.75 0.5 cosh √ 0.75x , ut(x, 0) = 0, ux(−40, t) = ux(40, t) = 0. These conditions correspond to an exact standing breather soliton solution u(x, t… view at source ↗
Figures from the paper (7 more)
Figure 4.4
Figure 4.4. Figure 4.4: Example 4.2: L 2 errors with different numerical fluxes. Example 4.3. (Non-smooth solutions for linear problem) For the problems with non-smooth solutions, we consider equation (2.1) in domain Ω = (−1, 1), whose initial condition is a piecewise constant function, u(x…
Figure 4.5
Figure 4.5. Figure 4.5: Example 4.3: numerical solutions with different schemes and cell numbers at t = 0.25. 29 [PITH_FULL_IMAGE:figures/full_fig_p029_4_5.png]
Figure 4.6
Figure 4.6. Figure 4.6: Example 4.4: numerical results of uh at t = 0.25. and piecewise constant initial data, u(t, 0) =    4, 0.3 ≤ x ≤ 0.425, 2, 0.575 ≤ x ≤ 0.7, 0, otherwise. ut(x, 0) = 0, and final time t = 0.25. We use scheme (4.55) with χ = 1 to treat the nonlinear term. We c…
Figure 4.7
Figure 4.7. Figure 4.7: Example 4.5: numerical results of uh at t = 0.25. final time t = 0.25 for different choices of numerical fluxes and polynomial degrees are presented, demonstrating the same conclusion as the 1D case in the least-squares sense. Example 4.7. (Sine-Gorden equation in 2D…
Figure 4.8
Figure 4.8. Figure 4.8: Example 4.6: L 2 errors of uh with different numerical fluxes. (a) CTCS scheme. (b) OF-EDG scheme [PITH_FULL_IMAGE:figures/full_fig_p032_4_8.png]
Figure 4.9
Figure 4.9. Figure 4.9: Example 4.7: numerical solutions uh at t = 0.25. 32 [PITH_FULL_IMAGE:figures/full_fig_p032_4_9.png]
Figure 4.10
Figure 4.10. Figure 4.10: Example 4.8: numerical solutions uh at t = 0.25. 5 Conclusion In this paper, we develop the OF-EDG method for solving the second order wave equation. Since the original EDG method is designed for smooth problems, the method produces spurious oscillations or even fai…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

27 extracted references · 26 canonical work pages

  1. [3]

    Appel¨ o, T

    D. Appel¨ o, T. Hagstrom, Q. Wang, and L. Zhang. An energy-based discontinuous Galerkin method for semilinear wave equations. Journal of Computational Physics , 418:109608, 2020

    work page 2020

  2. [1]

    Appel¨ o and T

    D. Appel¨ o and T. Hagstrom. A new discontinuous Galerkin formulation for wave equations in second-order form. SIAM Journal on Numerical Analysis , 53(6):2705– 2726, 2015

    work page 2015

  3. [2]

    Appel¨ o and T

    D. Appel¨ o and T. Hagstrom. An energy-based discontinuous Galerkin discretization of the elastic wave equation in second order form. Computer Methods in Applied Mechanics and Engineering , 338:362–391, 2018

    work page 2018

  4. [4]

    Appel¨ o and S

    D. Appel¨ o and S. Wang. An energy-based discontinuous Galerkin method for cou- pled elasto-acoustic wave equations in second-order form. International Journal for Numerical Methods in Engineering , 119(7):618–638, 2019

    work page 2019

  5. [5]

    D. N. Arnold, F. Brezzi, B. Cockburn, and L. D. Marini. Unified analysis of dis- continuous Galerkin methods for elliptic problems. SIAM journal on numerical analysis, 39(5):1749–1779, 2002

    work page 2002

  6. [6]

    J. Cao, B. Li, Y. Lin, and F. Yao. Numerical approximation of discontinuous solutions of the semilinear wave equation. SIAM Journal on Numerical Analysis , 63(1):214–238, 2025

    work page 2025

  7. [7]

    Chou, C.-W

    C.-S. Chou, C.-W. Shu, and Y. Xing. Optimal energy conserving local discontinuous Galerkin methods for second-order wave equation in heterogeneous media. Journal of Computational Physics , 272:88–107, 2014

    work page 2014

  8. [8]

    P. G. Ciarlet. The finite element method for elliptic problems . SIAM, 2002

    2002

Show all 27 references
  1. [9]

    Cockburn

    B. Cockburn. Discontinuous Galerkin methods for convection-dominated problems. In High-Order Methods for Computational Physics , pages 69–224. Springer, 1999

    1999

  2. [10]

    Cockburn, S.-Y

    B. Cockburn, S.-Y. Lin, and C.-W. Shu. TVB Runge-Kutta local projection discon- tinuous Galerkin finite element method for conservation laws III: one-dimensional systems. Journal of computational Physics , 84(1):90–113, 1989. 34

    1989

  3. [11]

    Cockburn and C.-W

    B. Cockburn and C.-W. Shu. TVB Runge-Kutta local projection discontinuous Galerkin finite element method for conservation laws. II. General framework. Math- ematics of computation , 52(186):411–435, 1989

    1989

  4. [12]

    Cockburn and C.-W

    B. Cockburn and C.-W. Shu. Runge–Kutta discontinuous Galerkin methods for convection-dominated problems. Journal of Scientific Computing, 16:173–261, 2001

    2001

  5. [13]

    J. Du, Y. Liu, and Y. Yang. An oscillation-free bound-preserving discontinuous Galerkin method for multi-component chemically reacting flows. Journal of Scien- tific Computing , 95(3):90, 2023

    2023

  6. [14]

    M. J. Grote, A. Schneebeli, and D. Sch¨ otzau. Discontinuous Galerkin finite element method for the wave equation. SIAM Journal on Numerical Analysis , 44(6):2408– 2431, 2006

    2006

  7. [15]

    A. Harten. High resolution schemes for hyperbolic conservation laws. Journal of Computational Physics , 135(2):260–278, 1997

    1997

  8. [16]

    Y. Liu, J. Lu, and C.-W. Shu. An essentially oscillation-free discontinuous Galerkin method for hyperbolic systems. SIAM Journal on Scientific Computing , 44(1):A230–A259, 2022

    2022

  9. [17]

    Y. Liu, J. Lu, and C.-W. Shu. An entropy stable essentially oscillation-free dis- continuous Galerkin method for hyperbolic conservation laws. SIAM Journal on Scientific Computing , 46(2):A1132–A1159, 2024

    2024

  10. [18]

    Y. Liu, J. Lu, Q. Tao, and Y. Xia. An oscillation-free discontinuous Galerkin method for shallow water equations. Journal of Scientific Computing , 92(3):109, 2022

    2022

  11. [19]

    J. Lu, Y. Liu, and C.-W. Shu. An oscillation-free discontinuous Galerkin method for scalar hyperbolic conservation laws. SIAM Journal on Numerical Analysis , 59(3):1299–1324, 2021

    2021

  12. [20]

    M. Peng, Z. Sun, and K. Wu. OEDG: Oscillation-eliminating discontinuous Galerkin method for hyperbolic conservation laws. Mathematics of Computation , 94(353):1147–1198, 2025

    2025

  13. [21]

    W. H. Reed and T. R. Hill. Triangular mesh methods for the neutron transport equation. Technical report, Los Alamos Scientific Lab., N. Mex.(USA), 1973

    1973

  14. [22]

    Riviere and M

    B. Riviere and M. F. Wheeler. Discontinuous finite element methods for acoustic and elastic wave problems. Contemporary Mathematics, 329(271-282):4–6, 2003

    2003

  15. [23]

    Shu and S

    C.-W. Shu and S. Osher. Efficient implementation of essentially non-oscillatory shock-capturing schemes. Journal of computational physics , 77(2):439–471, 1988. 35

    1988

  16. [24]

    Q. Tao, Y. Liu, Y. Jiang, and J. Lu. An oscillation free local discontinuous Galerkin method for nonlinear degenerate parabolic equations. Numerical Methods for Partial Differential Equations, 39(4):3145–3169, 2023

    2023

  17. [25]

    Xing, C.-S

    Y. Xing, C.-S. Chou, and C.-W. Shu. Energy conserving local discontinuous Galerkin methods for wave propagation problems. Inverse Problems & Imaging , 7(3), 2013

    2013

  18. [26]

    Yi and H

    N. Yi and H. Liu. An energy conserving local discontinuous Galerkin method for a nonlinear variational wave equation. Communications in Computational Physics , 23(3), 2018

    2018

  19. [27]

    Zhong and C.-W

    X. Zhong and C.-W. Shu. A simple weighted essentially nonoscillatory limiter for Runge-Kutta discontinuous Galerkin methods. Journal of Computational Physics , 232(1):397–415, 2013. 36

    2013

Pith tools

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