pith. sign in

arxiv: 2405.11063 · v2 · submitted 2024-05-17 · ⚛️ physics.flu-dyn · astro-ph.IM· astro-ph.SR· cs.NA· math.NA

Spectral Difference method with a posteriori limiting: II- Application to low Mach number flows

D. A. Velasco-Romero , R. Teyssier This is my paper

Pith reviewed 2026-05-24 01:02 UTC · model grok-4.3

classification ⚛️ physics.flu-dyn astro-ph.IMastro-ph.SRcs.NAmath.NA
keywords spectral difference methodlow Mach number flowsstellar convectionwell-balanced schemea posteriori limitingturbulent kinetic energynumerical fluid dynamics
0
0 comments X

The pith

Fourth-order Spectral Difference scheme emerges as optimal for low-Mach stellar convection.

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

This paper applies the Spectral Difference method with a posteriori limiting to low-Mach number flows and minuscule perturbations over steeply stratified hydrostatic equilibria in stellar convection. It adds a low-Mach adapted HLLC Riemann solver and a well-balanced scheme to reduce truncation errors that plague most methods. Tailored tests show the high-order SD approach handles very subsonic flows without the modified solver, yet the well-balanced framework proves essential for accurate small-amplitude convective and acoustic modes. Temporal and spatial analysis of turbulent kinetic energy identifies the fourth-order variant as the optimal choice for this numerical problem.

Core claim

The high-order SD method deals with very subsonic flows without necessarily using the modified Riemann solver, but the well-balanced framework is unavoidable to capture accurately small amplitude convective and acoustic modes. Analysis of the temporal and spatial evolution of the turbulent kinetic energy shows that the fourth-order SD scheme emerges as an optimal variant to solve this difficult numerical problem.

What carries the argument

Spectral Difference method with a posteriori limiting, incorporating L-HLLC Riemann solver and well-balanced properties for stratified equilibria.

If this is right

  • High-order SD schemes solve very subsonic flows without the modified Riemann solver.
  • Well-balanced framework is required to capture small amplitude convective and acoustic modes accurately.
  • Fourth-order SD emerges as optimal based on temporal and spatial turbulent kinetic energy evolution.
  • The combination of a posteriori limiting and well-balancing succeeds on stellar convection test problems.

Where Pith is reading between the lines

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

  • The same well-balanced modifications could transfer to other high-order discontinuous schemes used in astrophysical flows.
  • Reduced numerical diffusion may enable longer-time simulations of stellar convection zones before artificial damping dominates.
  • Testing the method on fully three-dimensional, rotating stellar models would check whether the fourth-order optimum persists outside the paper's suite.

Load-bearing premise

The tailored test suite sufficiently represents the full range of low-Mach and stratified perturbation challenges in stellar convection without introducing biases that favor the SD method.

What would settle it

A direct comparison in which a different SD order or another high-order method produces higher accuracy in turbulent kinetic energy levels across the same stellar convection test cases would falsify the claimed optimality of fourth order.

Figures

Figures reproduced from arXiv: 2405.11063 by D. A. Velasco-Romero, R. Teyssier.

Figure 1
Figure 1. Figure 1: Flux blending: On the left in red the troubled sub-cells. On the right the values for the blending coefficient θ admissible. This procedure, although robust, can result to be too computationally expensive. As described in Hennemann et al. (2021); Vilar & Abgrall (2024), a less expensive solution, although not perfectly ro￾bust, is to blend the SD and FV2 fluxes at trouble-adjacent sub-cells in a convex com… view at source ↗
Figure 2
Figure 2. Figure 2: Gresho vortex test: Color maps for the control volume average of M “ vϕ{cs . Results at t “ 1 for three values of the Mach number (Mmax “ 10´1 , 10´2 and 10´3 ) and a background velocity v0 “ 5. On the first row the results for FV2 (with HLLC), on the second row FV2L (with L-HLLC), and and on the third row SD3 (p “ 2). All of them making use of 962 DOF. MNRAS 000, 1–17 (2022) [PITH_FULL_IMAGE:figures/full… view at source ↗
Figure 3
Figure 3. Figure 3: Gresho vortex test: 1-dimensional slices of the control volume average of M “ vϕ{cs at t “ 1, corresponding to five laps over the domain (v0 “ 5). This test is interesting because we can change the mid￾plane pressure P0 without changing the velocity perturbation ∆v. The corresponding Mach number will be inversely propor￾tional to the midplane sound speed cs “ a γPc{ρ1. In princi￾ple, we expect our solution… view at source ↗
Figure 5
Figure 5. Figure 5: Rayleigh Taylor instability test. Iso-contours for the den￾sity at t “ 1 for 2nd, 4th and 8th order of the numerical approx￾imation for P0 “ 100. On the first row the results with 96 ˆ 384 DOF, on the second row the results with 192 ˆ 768 DOF. move the sound speed dependence in the numerical flux, al￾lowing us by construction to recover the proper low-Mach number limit. We now adopt our high-order SD schem… view at source ↗
Figure 4
Figure 4. Figure 4: Rayleigh Taylor instability test. Colormaps for the density at t “ 1.95 for 2nd, 4th and 8th order of the numerical approxima￾tion for 48 ˆ 192 DOF. On the first row the results for P0 “ 1, on the second row the results for P0 “ 10 and on the third row the results for P0 “ 100. In [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: Perturbation over hydrostatic equilibrium test. Iso￾contours at t “ 0.25 of δP for different values of the perturbation amplitude (η “ 10´2 , 10´4 , 10´8 and 10´12), with a total of 962 DOF. We repeat now this test making use of our well-balanced scheme (see subsection 2.2) [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Small perturbation of hydrostatic equilibrium test: Re￾sults using FV and SD with a well-balanced scheme. Iso-contours at t “ 0.25 of δP with initial perturbation amplitude η “ 10´12 , with 962 DOF for the 2nd-, 3rd, 4th- and 6th-order. waves, with many applications in astro-seismology (Rogers et al. 2013; Bhattacharya & Hanasoge 2023). In order to test our numerical techniques, we consider a simplified se… view at source ↗
Figure 8
Figure 8. Figure 8: Temperature perturbation for the buoyantly rising bubble test at t “ 100. Results for FV2, SD4B and SD8B with and without L-HLLC. tom row left panels) where the flow appears more chaotic. In comparison, the high-order schemes develop chaotic features much earlier in terms of resolution. It is interesting to see how large amplitude compressive waves have now entered the top stable region. These acous￾tic wa… view at source ↗
Figure 9
Figure 9. Figure 9: Temperature perturbation for the buoyantly rising bubble test at t “ 1750. Results for FV2, SD4B and SD8B with and without L-HLLC. turbulence slowly dissipates. It is striking to see the kinetic energy almost entirely dissipated by t “ 1000 for FV2 (expect for the lowest resolution, which exhibits an oscillatory be￾haviour), while other schemes combining the low-Mach num￾ber L-HLLC Riemann solver and high-… view at source ↗
Figure 10
Figure 10. Figure 10: Time evolution for the total kinetic energy of the buoyant bubble test. The top row shows (from left to right) our different schemes using similar number of DOF. The bottom row on the other hand, shows the effect of increasing the number of DOF for a given numerical scheme. 10 13 10 11 10 9 10 7 10 5 E(k) NDOF = 96 2 k 3 FV2 FV2L SD4BL SD8BL NDOF = 192 2 k 3 FV2 FV2L SD4BL SD8BL NDOF = 384 2 k 3 FV2 FV2L … view at source ↗
Figure 11
Figure 11. Figure 11: Kinetic energy power spectra Eˆpkq at late time, averaged over 5 snapshots between time t “ 1750 and t “ 2000. The top row shows (from left to right) our different schemes using similar number of DOF. The bottom row on the other hand, shows the effect of increasing the number of DOF for a given numerical scheme. The dashed line shows the predicted scaling for 2D subsonic turbulence based on the conservati… view at source ↗
read the original abstract

Stellar convection poses two main gargantuan challenges for astrophysical fluid solvers: low-Mach number flows and minuscule perturbations over steeply stratified hydrostatic equilibria. Most methods exhibit excessive numerical diffusion and are unable to capture the correct solution due to large truncation errors. In this paper, we analyze the performance of the Spectral Difference (SD) method under these extreme conditions using an arbitrarily high-order shock capturing scheme with a posteriori limiting. We include both a modification to the HLLC Riemann solver adapted to low Mach number flows (L-HLLC) and a well-balanced scheme to properly evolve perturbations over steep equilibrium solutions. We evaluate the performance of our method using a series of test tailored specifically for stellar convection. We observe that our high-order SD method is capable of dealing with very subsonic flows without necessarily using the modified Riemann solver. We find however that the well-balanced framework is unavoidable if one wants to capture accurately small amplitude convective and acoustic modes. Analyzing the temporal and spatial evolution of the turbulent kinetic energy, we show that our fourth-order SD scheme seems to emerge as an optimal variant to solve this difficult numerical problem.

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

2 major / 1 minor

Summary. The manuscript applies the Spectral Difference (SD) method with a posteriori limiting to low-Mach stellar convection, incorporating an L-HLLC Riemann solver and a well-balanced discretization. Through a series of tests tailored for stellar convection, the authors analyze performance on subsonic flows and small-amplitude perturbations over steep stratification, concluding from the temporal and spatial evolution of turbulent kinetic energy that the fourth-order SD variant emerges as optimal.

Significance. If the optimality claim is substantiated, the work would contribute a high-order shock-capturing approach suitable for the dual challenges of low-Mach advection and stratified perturbations in astrophysical flows. The explicit inclusion of well-balancing and low-Mach modifications addresses known difficulties, but the absence of quantitative error norms, baseline comparisons, or sensitivity studies in the presented results limits the immediate significance.

major comments (2)
  1. [Abstract] Abstract: The claim that the fourth-order SD scheme 'seems to emerge as an optimal variant' rests on TKE temporal/spatial evolution in tailored tests, yet no quantitative error metrics (e.g., L2 norms of velocity or density perturbations), baseline comparisons to other orders or methods, or data on test-parameter sensitivity are provided to support optimality.
  2. [Abstract] Abstract: The statement that well-balancing is 'unavoidable' for small modes and that L-HLLC is 'sometimes unnecessary' is presented without accompanying quantitative evidence (e.g., error tables or convergence rates) showing when each component is required or dispensable across the test suite.
minor comments (1)
  1. [Abstract] The abstract refers to 'a series of test tailored specifically for stellar convection' without naming the specific test problems or citing their original sources in the literature.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful review and constructive feedback. We respond point-by-point to the major comments below.

read point-by-point responses

  1. Referee: [Abstract] Abstract: The claim that the fourth-order SD scheme 'seems to emerge as an optimal variant' rests on TKE temporal/spatial evolution in tailored tests, yet no quantitative error metrics (e.g., L2 norms of velocity or density perturbations), baseline comparisons to other orders or methods, or data on test-parameter sensitivity are provided to support optimality.

    Authors: We acknowledge that the optimality statement relies on TKE evolution as the primary diagnostic in the tailored stellar convection tests. While this metric directly reflects the capture of convective and acoustic modes, we agree that the claim would be strengthened by supplementary quantitative measures. In the revised manuscript we will add L2 error norms for velocity and density perturbations (where definable), direct comparisons across SD orders, and sensitivity results for key test parameters. revision: yes

  2. Referee: [Abstract] Abstract: The statement that well-balancing is 'unavoidable' for small modes and that L-HLLC is 'sometimes unnecessary' is presented without accompanying quantitative evidence (e.g., error tables or convergence rates) showing when each component is required or dispensable across the test suite.

    Authors: The statements are based on comparative runs performed during the study, but we concur that explicit quantitative support (error tables, convergence rates) is needed to demonstrate necessity or dispensability across the full test suite. We will incorporate such tables and rates in the revised version to clarify the role of each component. revision: yes

Circularity Check

0 steps flagged

No circularity; optimality from numerical experiments on external tests

full rationale

The paper's central claim—that fourth-order SD emerges as optimal—is derived from direct numerical experiments measuring TKE evolution on a suite of tailored stellar convection tests. No equations, parameters, or results reduce by construction to fitted inputs, self-definitions, or self-citation chains. The method description, well-balancing, and L-HLLC modifications are presented as independent algorithmic choices whose performance is then evaluated against the test problems. Self-citations (e.g., to part I) are not load-bearing for the optimality conclusion, which rests on observable simulation outcomes rather than imported uniqueness theorems or ansatzes. This matches the default case of a self-contained numerical study with no circular reduction.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract supplies no information on free parameters, background axioms, or newly postulated entities.

pith-pipeline@v0.9.0 · 5747 in / 1067 out tokens · 23761 ms · 2026-05-24T01:02:02.022370+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

1 extracted references · 1 canonical work pages · 1 internal anchor

  1. [1]

    A numerical scheme for the compressible low-Mach number regime of ideal fluid dynamics

    Baraffe I., et al., 2023, MNRAS, 519, 5333 Barranco J. A., Marcus P. S., 2006, Journal of Computational Physics, 219, 21 Barsukow W., Edelmann P. V. F., Klingenberg C., Miczek F., Roepke F. K., 2016, arXiv e-prints, p. arXiv:1612.03910 Bhattacharya J., Hanasoge S. M., 2023, ApJS, 264, 21 Birken P., Meister A., 2005a, PAMM, 5, 759 Birken P., Meister A., 20...

    work page internal anchor Pith review Pith/arXiv arXiv 2023