Pith. sign in

REVIEW 2 major objections 6 minor 2 cited by

Entropy Stable Nodal Discontinuous Galerkin Methods via Quadratic Knapsack Limiting

T0 review · 2 major / 6 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 Replacing the linear knapsack solver in entropy-stable discontinuous Galerkin methods with a quadratic one yields a continuous limiter output, which restores second-order accuracy in time and sharply reduces adaptive timesteps in…

desk verdict Quadratic knapsack limiting is a solid, modest contribution; the temporal-regularity story is plausible but needs a sharper mechanism argument and code/data before it fully lands. read the letter →

arxiv 2507.14488 v1 pith:RVSWT5ZO submitted 2025-07-19 math.NA cs.NA

Brian Christner , Jesse Chan This is my paper
classification math.NAcs.NA MSC 68Q2568R1068U0565M60
keywords entropystabilitydiscontinuousGalerkinquadraticknapsacksubcelllimitingfluxcorrectedtransportadaptivetimesteppingcompressibleEulerequationspositivitypreservation
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

High-order discontinuous Galerkin methods for compressible flow need additional stabilization near shocks, and a recent framework enforces a cell entropy inequality by blending the high-order scheme with a low-order positivity-preserving scheme through coefficients chosen by a knapsack optimization problem. This paper proposes replacing the linear knapsack objective with a quadratic one. The authors prove that the resulting quadratic knapsack problem reduces to a scalar quasi-Newton root-finding iteration that converges in finitely many steps, and they show numerically that its solution is continuous in the data, whereas the linear knapsack solution is discontinuous. That continuity is presented as the reason the blended scheme attains second-order accuracy in time and requires far fewer adaptive timesteps in shock-type problems, while preserving spatial accuracy, entropy stability, positivity, and local linear stability.

What carries the argument

The central object is the continuous quadratic knapsack problem (3.9), whose solution is the output of the limiter. The load-bearing representation is $\theta^*(\lambda^*) = \mathrm{clip}_{0,1-\ell_c}(\lambda^* a)$: the optimal blending coefficients are the componentwise clipping of a scalar multiple of the algebraic data vector $a$, making the limiter output a continuous function of $a$ for fixed bounds. The quasi-Newton iteration (3.10) is the mechanism that makes the approach practical, reducing the constrained optimization to scalar root-finding of the monotone concave function $f$ with at most $L+1$ iterations and $O(L)$ cost per iteration. Continuity of this map in $a$ is what the paper argues restores smoothness in time of the blended semi-discretization, in contrast to the sorting-induced discontinuity of the linear knapsack solution.

What would settle it

Take the quadratic knapsack limiter and deliberately make its output discontinuous in the data (for example, by adding a sorting-based tie-break or switching to an $\ell^\infty$ objective), then measure the time-convergence order and adaptive timestep count on the modified Sod shock tube; if the scheme still shows second-order time convergence and few timesteps, continuity of the limiter is not the controlling factor. Alternatively, smooth the linear knapsack output near its sorting discontinuity and check whether second-order time convergence is restored.

Watch

Extended reading notes

Core claim

The central discovery is that the blending coefficients enforcing a cell entropy inequality in an entropy-stable nodal DG method can be chosen as the solution of a continuous quadratic knapsack problem, $\min_{\theta} \theta^{\mathsf T}\theta$ subject to $a^{\mathsf T}\theta \ge b$ and $0 \le \theta \le 1-\ell_c$, and that this solution has the explicit form $\theta^* = \mathrm{clip}_{0,1-\ell_c}(\lambda^* a)$, where $\lambda^*$ is the unique nonnegative root of the monotone concave function $f(\lambda)=a^{\mathsf T}\mathrm{clip}_{0,1-\ell_c}(\lambda a)-b$. The authors prove that $f$ is nondecreasing and concave, with right-derivative $\delta f(\lambda)=\sum_{i:\,0<a_i,\ \lambda a_i<1-\ell_{c,i}} a_i^2$, so that a scalar quasi-Newton iteration, $\lambda_{k+1}=\lambda_k - f(\lambda_k)/\delta f(\lambda_k)$, converges to $\lambda^*$ in at most $L+1$ iterations. Unlike the linear knapsack solution, which relies on sorting and is therefore discontinuous as a function of the data $a$, this quadratic solution is continuous in $a$. Because the limiter output enters directly into the blended right-hand side, the continuity is argued to translate into higher temporal regularity; numerically, this appears as $O(\Delta t^2)$ time convergence in shock-type problems instead of $O(\Delta t)$ for linear knapsack limiting, together with large reductions in adaptive timestep counts at high polynomial order.

Load-bearing premise

The reported gains in temporal accuracy and timestep count rest on the assumption that the limiter output's continuity in the algebraic data is what controls the temporal smoothness of the blended scheme, so that the linear knapsack solution's discontinuity is the actual cause of its first-order time convergence.

Editorial extensions

If this is right

  • Quadratic knapsack limiting can replace linear knapsack limiting in the Lin-Chan entropy-stable DG framework while matching spatial accuracy, entropy stability, and positivity, and improving temporal convergence order for shock-type problems.
  • Adaptive timesteppers on shock-dominated problems should take far fewer steps with quadratic knapsack limiting at high polynomial order, reducing overall simulation cost.
  • The measured $O(\Delta t^2)$ temporal convergence brings quadratic knapsack limiting on par with entropy-stable flux differencing for the modified Sod problem, while avoiding the strong oscillations that accompany the linearly unstable ESFD scheme.
  • Because the blended scheme is a convex combination of two locally linearly stable schemes, knapsack limiting appears to sidestep the local linear instability documented for entropy-stable flux differencing, making it usable for under-resolved and long-time simulations.

Reading between the lines

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

  • If the continuity mechanism is correct, any limiter whose output depends discontinuously on the data (e.g., solvers that sort or switch active sets) may similarly degrade temporal accuracy in shock-dominated regimes; this can be tested in the same DG framework by swapping only the solver.
  • The scalar root-finding reduction holds for arbitrary convex box constraints, not just the particular subcell operators used here, so the quadratic knapsack construction could be ported to other convex-limiting and invariant-domain-preserving schemes to restore temporal smoothness.
  • The at-most-$L+1$ iteration bound gives a deterministic per-evaluation complexity that could support a priori cost estimates in adaptive DG codes; the paper observes that only about four iterations occur in practice even at high order.
  • The same continuity benefit should extend to any strictly convex objective with a differentiable square norm, such as smoothed $\ell^p$ objectives with $1<p<\infty$, which would generalize the mechanism beyond the exact quadratic case.
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 / 6 minor

Summary. The paper proposes replacing the linear knapsack subcell limiter of Lin and Chan with a quadratic knapsack problem for entropy stable nodal discontinuous Galerkin methods. The main theoretical contribution is Theorem 3.3, which states that the continuous quadratic knapsack solution can be obtained by a scalar quasi-Newton iteration that terminates in at most L+1 iterations, and Algorithm 3.1 implements this O(L) solver. The numerical section reports spatial convergence for a smooth density wave, shock-tube results for modified Sod and Shu-Osher problems, temporal self-convergence rates and adaptive timestep counts for shock-type problems, local linear stability spectra, and positivity-preserving benchmarks (Leblanc, Sedov, Kelvin-Helmholtz). The headline numerical claim is that quadratic knapsack limiting attains roughly O(dt^2) temporal self-convergence versus roughly O(dt) for linear knapsack limiting, while requiring substantially fewer adaptive timesteps in high-order shock-type simulations.

Significance. If the temporal-regularity claim holds, the paper offers a simple, cheap modification of an existing entropy-stable limiter that preserves entropy stability, positivity, and spatial accuracy while improving time accuracy and robustness under adaptive timestepping. The theoretical part is a genuine contribution: the reduction of the quadratic knapsack problem to finite scalar root-finding is derived from first principles, with no fitted constants other than the explicitly reported positivity parameter alpha, and the algorithm has a provable worst-case iteration count. The numerical evidence is broad and internally consistent, including validation against analytic and WENO reference solutions. The main weakness is that the central advertised benefit, improved temporal regularity, rests on an asserted continuity-to-convergence mechanism that is not demonstrated, and the supporting experiments are narrow and not independently reproducible from the manuscript alone.

major comments (2)
  1. [§4.2.2, Fig. 5, Table 3] The measured temporal convergence rates may be contaminated by the discontinuity itself. In the modified Sod problem, the global L2 difference between S(dt_ref) and S(dt) includes errors in shock location and errors generated at the discontinuity, which scale differently from the temporal order of a smooth-flowing method. Because the reference solution is also discontinuous, the reported slopes of approximately 1.0 for LK and 2.1 for QK may reflect how the shock is tracked rather than the regularity of the knapsack solution. I suggest computing temporal self-convergence on a smooth test or within a region excluding the shock, and reporting rates over a longer range of dt with more sample points. A demonstration that the RHS itself is more regular for QK than for LK would directly support the proposed mechanism.
  2. [§4.3, Table 4] The conclusion that LK and QK are 'locally linearly stable' is stronger than the evidence. Table 4 reports the maximal real part of the linearized Jacobian at the initial state of the modified Sod problem only. The knapsack limiter is state-dependent and nonsmooth (LK) or merely continuous/Lipschitz (QK), so an instantaneous linearization at t=0 does not imply that perturbations remain bounded along a trajectory. The values near 1e-13 show that the initial linearization has no positive spectral growth beyond roundoff, but they do not establish local linear stability in the usual sense. Please either add a trajectory-level test (e.g., time evolution of small perturbations) or restrict the wording to what the computation actually shows.
minor comments (6)
  1. [Theorem 3.3] The proof states 'it can be shown' that the iterates lambda_k are nondecreasing and lie in [0, lambda*]; please write out this short argument using concavity and the right derivative, and clarify why each iteration must reach at least one new breakpoint, so that the L+1 bound is fully justified.
  2. [§4.2.2, Fig. 5, Table 3] Please clarify whether Figure 5 and the 'same study' used for Table 3 use fixed timesteps or adaptive timesteps; the text says 'fixed timestep dt' for Figure 5 but the table description is ambiguous.
  3. [§4.4.3] The case alpha = 0 in the Kelvin-Helmholtz experiment should be explained: with alpha = 0, the 'relative positivity constraint' (3.12) collapses to the base positivity bound, so it is not actually a relative constraint in that run.
  4. [Reproducibility] No code or data availability statement is included. Since the paper advertises specific adaptive timestep counts and convergence slopes, a reproducibility statement with the exact commit hashes and solver settings would be very helpful.
  5. [Figure 4 caption] The caption of Figure 4 does not identify the flux used in panel (c); the text says it is the LxF version of ESFD, and this should be reflected in the caption.
  6. [Title] In the provided manuscript text, the word 'Stable' is split as 'St able' in the title and abstract; please correct the spacing.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the quadratic knapsack solver is derived from first principles, and the numerical comparisons rest on independent benchmarks and explicitly reported parameters.

full rationale

The paper's central theoretical contribution is the reduction of the quadratic knapsack problem (3.9) to scalar root-finding. The derivation in Section 3.2 defines the dual residual f(lambda) and proves, in Theorem 3.3, that the clipped vector clip(0,1-lc)(lambda a) solves the constrained problem whenever f(lambda*)=0; this is a self-contained KKT/clipping argument rather than an equivalence imposed by definition. The baseline linear knapsack method is quoted from Lin and Chan [16], and positivity coefficients follow Lin, Chan, and Tomas [17], but those prior works are auxiliary foundations and are not used to prove the paper's headline claim about quadratic knapsack. The temporal-regularity claim is an empirical measurement: Table 2 reports adaptive timestep counts and Table 3 reports measured slopes from fixed-timestep convergence studies, so no fitted parameter is renamed as a prediction. The positivity parameter alpha is explicitly reported (e.g., alpha = .55 for Leblanc, alpha = .6/.9 for Sedov), and the timesteps are stated to be within the positivity-preserving CFL bound, so the benchmark results are not manufactured by hidden fits. No equation in the paper reduces to its own input, and no load-bearing argument collapses into a self-citation. The only caveats are the unproven continuity-to-temporal-accuracy mechanism and the narrow benchmark set, which are correctness or evidence questions, not circularity.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

No new physical entities are introduced. The method introduces algorithmic knobs (alpha, tolerance) rather than fitted parameters; the entropy-stability properties are inherited from the prior low-order and flux formulations.

free parameters (1)
  • alpha (relative positivity parameter) = 0.55 (Leblanc), 0.6 (Sedov), 0.9 (Sedov comparison), 0 (KHI)
    User-chosen parameter in the relative positivity constraint (3.12); not derived from first principles, affects the amount of limiting, and is set per benchmark.
assumptions (4)
  • domain assumption The low-order finite volume scheme (2.13) is entropy stable and positivity preserving for the compressible Euler equations under the LxF flux with a suitable wavespeed.
    Invoked in Sections 2.3 and 3.3; relies on the external result in [17]. The knapsack problem's feasibility and the blended scheme's positivity rest on this.
  • domain assumption The high-order scheme (2.8) violates the cell entropy inequality by O(h^{N+1}) in smooth regions, so blending preserves high order.
    Used to justify that the blended scheme retains high order accuracy; cited from Vilar [29] and Chan [4].
  • domain assumption The surface flux f* is entropy stable, so the simplification from (3.2) to (3.3) is valid.
    The derivation of the cell entropy inequality constraint a^T theta >= b depends on this cancellation.
  • standard math The cell entropy inequality (2.17) with Gauss-Lobatto collocation is formally high order accurate.
    Section 2.4 states this using chain rule and quadrature accuracy.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Entropy Stable Nodal Discontinuous Galerkin Methods via Quadratic Knapsack Limiting." pith.science (2026). https://pith.science/paper/RVSWT5ZO

@misc{pith2026250714488,
  author       = {Pith},
  title        = {Pith review of: Entropy Stable Nodal Discontinuous Galerkin Methods via Quadratic Knapsack Limiting},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RVSWT5ZO}},
  note         = {Machine review of arXiv:2507.14488}
}
read the original abstract

Lin, Chan (High order entropy stable discontinuous Galerkin spectral element methods through subcell limiting, 2024) enforces a cell entropy inequality for nodal discontinuous Galerkin methods by combining flux corrected transport (FCT)-type limiting and a knapsack solver, which determines optimal limiting coefficients that result in a semi-discrete cell entropy inequality while preserving nodal bounds. In this work, we provide a slight modification of this approach, where we utilize a quadratic knapsack problem instead of a standard linear knapsack problem. We prove that this quadratic knapsack problem can be reduced to efficient scalar root-finding. Numerical results demonstrate that the proposed quadratic knapsack limiting strategy is efficient and results in a semi-discretization with improved regularity in time compared with linear knapsack limiting, while resulting in fewer adaptive timesteps in shock-type problems.

Figures

Figures reproduced from arXiv: 2507.14488 by the authors.

Figure 1
Figure 1. First and second components of the solution to a two-dimensional continuous [PITH_FULL_IMAGE:figures/full_fig_p009_1.png] view at source ↗
Figure 2
Figure 2. First and second components of the solution to a two-dimensional quadratic [PITH_FULL_IMAGE:figures/full_fig_p010_2.png] view at source ↗
Figure 3
Figure 3. Density of the modified Sod shock tube solution at final time [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: Density of the Shu Osher shock tube solution at final time [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]
Figure 5
Figure 5. Figure 5: L 2 error of the modified Sod shock tube simulations against reference solutions at final time T = .2 against fixed timestep. Timestepper N, M Linear Knapsack Quadratic Knapsack ESFD RK4 3, 64 0.99 2.11 2.11 7, 32 1.01 1.99 2.63 SSPRK43 3, 64 1.00 2.26 2.46 7, 32 1.09 …
Figure 6
Figure 6. Figure 6: L 2 error of the Sod shock tube simulations against analytic solution over time. simulation which amplifies nonphysical oscillations over time, as seen in Figures (3c) and (4c). We see in the same plots that the knapsack limiting solutions do not seem to undergo the sa…
Figure 7
Figure 7. Figure 7: HLLC quadratic knapsack limiting solution to the Leblanc shock tube with [PITH_FULL_IMAGE:figures/full_fig_p020_7.png]
Figure 8
Figure 8. Figure 8: Density of the Sedov blast wave solution at the final time [PITH_FULL_IMAGE:figures/full_fig_p021_8.png]
Figure 9
Figure 9. Figure 9: Quadratic knapsack limiting solution to the long-time Kelvin-Helmholtz in [PITH_FULL_IMAGE:figures/full_fig_p022_9.png]

Discussion (0). Sign in to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Entropy correction artificial viscosity for high order DG methods using multiple artificial viscosities

    math.NA 2026-04 unverdicted novelty 7.0 of 10

    Multiple artificial viscosities with analytical optimal parameters enable more flexible entropy-stable DG simulations than monolithic viscosity for 1D and 2D problems.

  2. Entropy stable finite difference methods via entropy correction artificial viscosity and knapsack limiting

    math.NA 2025-08 conditional novelty 6.0 of 10

    Two finite difference schemes, one using entropy-correction artificial viscosity and one using knapsack limiting, satisfy discrete entropy inequalities; the knapsack variant also provably preserves positivity for the ...

Reference graph

Works this paper leans on

29 extracted references · 19 canonical work pages · cited by 2 Pith papers

  1. [1]

    Batten, N

    P. Batten, N. Clarke, C. Lambert, and D. M. Causon , On the choice of wavespeeds for the hllc riemann solver , SIAM Journal on Scientific Computing, 18 (1997), pp. 1553–1570

    work page 1997

  2. [2]

    M. H. Carpenter, T. C. Fisher, E. J. Nielsen, and S. H. Frankel , Entropy Stable Spectral Collocation Schemes for the Navier–Stokes Equations: Discontinuous Interfaces , SIAM Journal on Scientific Computing, 36 (2014), pp. B835–B867, https://doi.org/10.1137/ 130932193

    work page 2014

  3. [3]

    Chan, On discretely entropy conservative and entropy stable discontinuous Galerkin meth- ods, Journal of Computational Physics, 362 (2018), pp

    J. Chan, On discretely entropy conservative and entropy stable discontinuous Galerkin meth- ods, Journal of Computational Physics, 362 (2018), pp. 346–374, https://doi.org/https: //doi.org/10.1016/j.jcp.2018.02.033, https://www.sciencedirect.com/science/article/pii/ S0021999118301153

    doi:10.1016/j.jcp.2018.02.033 2018

  4. [4]

    Chan , An artificial viscosity approach to high order entropy stable discontinuous galerkin methods, 2025, https://arxiv.org/abs/2501.16529, https://arxiv.org/abs/2501.16529

    J. Chan , An artificial viscosity approach to high order entropy stable discontinuous galerkin methods, 2025, https://arxiv.org/abs/2501.16529, https://arxiv.org/abs/2501.16529

    arXiv 2025

  5. [5]

    J. Chan, H. Ranocha, A. M. Rueda-Ram ´ırez, G. Gassner, and T. W arburton , On the Entropy Projection and the Robustness of High Order Entropy Stable Discontinu- ous Galerkin Schemes for Under-Resolved Flows , Frontiers in Physics, Volume 10 - 2022 (2022), https://doi.org/10.3389/fphy.2022.898028, https://www.frontiersin.org/journals/ physics/articles/10....

    work page arXiv 2022

  6. [6]

    P. Chandrashekar , Kinetic Energy Preserving and Entropy Stable Finite Volume Schemes for Compressible Euler and Navier-Stokes Equations , Communications in Computational Physics, 14 (2013), p. 1252–1286, https://doi.org/10.4208/cicp.170712.010313a

    work page arXiv 2013

  7. [7]

    Chen and C.-W

    T. Chen and C.-W. Shu, Entropy stable high order discontinuous Galerkin methods with suit- able quadrature rules for hyperbolic conservation laws , Journal of Computational Physics, 345 (2017), pp. 427–461, https://doi.org/https://doi.org/10.1016/j.jcp.2017.05.025, https: //www.sciencedirect.com/science/article/pii/S002199911730400X

    doi:10.1016/j.jcp.2017.05.025 2017

  8. [8]

    Cominetti, W

    R. Cominetti, W. F. Mascarenhas, and P. J. Silva , A newton ’s method for the continuous quadratic knapsack problem, Mathematical Programming Computation, 6 (2014), pp. 151– 169

    work page 2014

Show all 29 references
  1. [9]

    Crean, J

    J. Crean, J. E. Hicken, D. C. Del Rey Fern ´andez, D. W. Zingg, and M. H. Carpenter, Entropy-stable summation-by-parts discretization of the Euler equa- tions on general curved elements , Journal of Computational Physics, 356 (2018), pp. 410–438, https://doi.org/https://doi.or...

    2018 doi

  2. [10]

    G. J. Gassner, M. Sv ¨ard, and F. J. Hindenlang , Stability issues of entropy-stable and/or split-form high-order schemes: analysis of linear stability , Journal of Scientific Computing, 90 (2022), p. 79

    2022

  3. [11]

    G. J. Gassner, A. R. Winters, and D. A. Kopriva, Split form nodal discontinuous Galerkin schemes with summation-by-parts property for the compressible Euler equations , Journal of Computational Physics, 327 (2016), pp. 39–66, https://doi.org/https://doi.org/10.1016/ j.jcp.2016...

    2016

  4. [12]

    Godlewski and P.-A

    E. Godlewski and P.-A. Raviart, Numerical approximation of hyperbolic systems of conser- vation laws , vol. 118, Springer Science & Business Media, 2013

    2013

  5. [13]

    Harten, P

    A. Harten, P. D. Lax, and B. v. Leer , On Upstream Differencing and Godunov-Type Schemes for Hyperbolic Conservation Laws , SIAM Review, 25 (1983), pp. 35–61, https: //doi.org/10.1137/1025002

    1983 doi

  6. [14]

    J. S. Hesthaven and T. W arburton , Nodal discontinuous Galerkin methods: algorithms, analysis, and applications , Springer, 2008

    2008

  7. [15]

    D. A. Kopriva and G. Gassner , On the Quadrature and Weak Form Choices in Colloca- 24 B. CHRISTNER AND J. CHAN tion Type Discontinuous Galerkin Spectral Element Methods , J. Sci. Comput., 44 (2010), p. 136–155, https://doi.org/10.1007/s10915-010-9372-3

    2010 doi

  8. [16]

    Lin and J

    Y. Lin and J. Chan , High order entropy stable discontinuous Galerkin spectral el- ement methods through subcell limiting , Journal of Computational Physics, 498 (2024), p. 112677, https://doi.org/https://doi.org/10.1016/j.jcp.2023.112677, https:// www.sciencedirect.com/scienc...

    2024

  9. [17]

    Y. Lin, J. Chan, and I. Tomas , A positivity preserving strategy for entropy stable dis- continuous Galerkin discretizations of the compressible Euler and Navier-Stokes equa- tions, Journal of Computational Physics, 475 (2023), p. 111850, https://doi.org/https: //doi.org/10.10...

    2023

  10. [18]

    Pazner, Sparse invariant domain preserving discontinuous Galerkin methods with sub- cell convex limiting , Computer Methods in Applied Mechanics and Engineering, 382 (2021), p

    W. Pazner, Sparse invariant domain preserving discontinuous Galerkin methods with sub- cell convex limiting , Computer Methods in Applied Mechanics and Engineering, 382 (2021), p. 113876, https://doi.org/https://doi.org/10.1016/j.cma.2021.113876, https:// www.sciencedirect.com...

    2021

  11. [19]

    Peyvan, K

    A. Peyvan, K. Shukla, J. Chan, and G. Karniadakis , High-order methods for hyper- sonic flows with strong shocks and real chemistry , Journal of Computational Physics, 490 (2023), p. 112310, https://doi.org/https://doi.org/10.1016/j.jcp.2023.112310, https: //www.sciencedirect....

    2023

  12. [20]

    Rackauckas and Q

    C. Rackauckas and Q. Nie , Differentialequations.jl – a performant and feature-rich ecosys- tem for solving differential equations in julia , The Journal of Open Research Software, 5 (2017), https://doi.org/10.5334/jors.151, https://app.dimensions.ai/details/publication/ pub.1...

    2017 doi

  13. [21]

    H. Ranocha , Entropy conserving and kinetic energy preserving numerical methods for the euler equations using summation-by-parts operators , Spectral and high order methods for partial differential equations ICOSAHOM 2018, 134 (2020), pp. 525–535

    2020

  14. [22]

    Ranocha and G

    H. Ranocha and G. J. Gassner, Preventing Pressure Oscillations Does Not Fix Local Linear Stability Issues of Entropy-Based Split-Form High-Order Schemes , Communications on Applied Mathematics and Computation, 4 (2021), p. 880–903, https://doi.org/10.1007/ s42967-021-00148-z

    2021

  15. [23]

    Ranocha, M

    H. Ranocha, M. Schlottke-Lakemper, J. Chan, A. M. Rueda-Ram ´ırez, A. R. Winters, F. Hindenlang, and G. J. Gassner , Efficient Implementation of Modern Entropy Stable and Kinetic Energy Preserving Discontinuous Galerkin Methods for Conservation Laws , ACM Transactions on Mathe...

    2023

  16. [24]

    Ranocha, M

    H. Ranocha, M. Schlottke-Lakemper, A. R. Winters, E. F aulhaber, J. Chan, and G. Gassner , Adaptive numerical simulations with Trixi.jl: A case study of Julia for scientific computing , Proceedings of the JuliaCon Conferences, 1 (2022), p. 77, https: //doi.org/10.21105/jcon.00...

    2022 arXiv

  17. [25]

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

    1973

  18. [26]

    Revels, M

    J. Revels, M. Lubin, and T. Papamarkou, Forward-mode automatic differentiation in Julia, arXiv preprint arXiv:1607.07892, (2016)

    2016 arXiv

  19. [27]

    A. M. Rueda-Ram´ırez, B. Bolm, D. Kuzmin, and G. J. Gassner, Monolithic convex limiting for legendre-gauss-lobatto discontinuous galerkin spectral-element methods , Communica- tions on Applied Mathematics and Computation, 6 (2024), pp. 1860–1898

    2024

  20. [28]

    Schlottke-Lakemper, G

    M. Schlottke-Lakemper, G. J. Gassner, H. Ranocha, A. R. Winters, J. Chan, and A. Rueda-Ram´ırez, Trixi.jl: Adaptive high-order numerical simulations of hyperbolic PDEs in Julia . https://github.com/trixi-framework/Trixi.jl, 2025, https://doi.org/10. 5281/zenodo.3996439

    2025

  21. [29]

    Vilar , Local subcell monolithic DG/FV convex property preserving scheme on un- structured grids and entropy consideration , Journal of Computational Physics, 521 (2025), p

    F. Vilar , Local subcell monolithic DG/FV convex property preserving scheme on un- structured grids and entropy consideration , Journal of Computational Physics, 521 (2025), p. 113535, https://doi.org/10.1016/j.jcp.2024.113535, https://www.sciencedirect. com/science/article/pi...

    2025

Pith tools

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