pith. sign in

arxiv: 2604.25672 · v2 · pith:PU5EF353new · submitted 2026-04-28 · 🧮 math.NA · cs.NA· physics.flu-dyn

A bound-preserving oscillation-eliminating discontinuous Galerkin scheme for compressible two-phase flow

Jia-Jun Zou , Fan Zhang , Yu-Chang Liu , Qi Kong , Yun-Long Liu , A-Man Zhang This is my paper

Pith reviewed 2026-05-07 15:16 UTC · model grok-4.3

classification 🧮 math.NA cs.NAphysics.flu-dyn
keywords bound-preserving schemesoscillation-eliminating limitersdiscontinuous Galerkin methodscompressible two-phase flowsoperator splittingKapila modelAbgrall conditionstiff source terms
0
0 comments X

The pith

Operator splitting with implicit discretization yields a bound-preserving oscillation-eliminating DG scheme for Kapila two-phase flows that removes stiffness-induced CFL restrictions.

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

This paper develops a high-order discontinuous Galerkin method for compressible gas-gas and gas-liquid flows governed by the Kapila five-equation model with Tammann equation of state. The stiff source term in the volume fraction equation normally forces tiny time steps in explicit schemes. The authors decouple the problem via operator splitting, advance the transport part with a quasi-conservative DG discretization, and treat the source part with an adaptive implicit integrator that hybridizes backward Euler and SDIRK2. They prove the implicit step is unconditionally bound-preserving and that the full scheme satisfies the Abgrall condition. An oscillation-eliminating limiter and a bound-preserving limiter are added to control oscillations and enforce physical bounds on densities, pressure, and volume fraction without characteristic decomposition.

Core claim

The proposed BP-OEDG scheme, combined with the splitting strategy, is unconditionally bound-preserving and strictly satisfies the Abgrall condition, as established by rigorous proofs; the implicit treatment of the stiff source term removes the CFL restriction while a velocity-divergence reconstruction improves accuracy inside the implicit solver.

What carries the argument

The operator-splitting strategy that separates the transport model from the stiff kappa-source term, with the source advanced by an adaptive implicit method (backward Euler hybridized with SDIRK2) that is proven unconditionally bound-preserving.

If this is right

  • Time steps become independent of the stiffness parameter, removing the dominant computational bottleneck in explicit treatments.
  • Physical bounds on partial densities, pressure, and volume fraction are preserved at every stage without additional post-processing.
  • Spurious oscillations are controlled without requiring characteristic-variable decomposition.
  • The scheme satisfies the Abgrall condition, so uniform flow remains uniform after discretization.

Where Pith is reading between the lines

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

  • The same splitting-plus-implicit approach may apply directly to other multiphase models that contain stiff algebraic source terms.
  • Larger time steps could enable three-dimensional or long-time engineering simulations that remain out of reach for conventional explicit DG codes.
  • The velocity-divergence reconstruction inside the implicit solver suggests a route to higher-order accuracy for source terms in other hyperbolic systems.

Load-bearing premise

The operator splitting accurately decouples transport and stiff source terms without introducing splitting errors large enough to violate the bound-preserving or Abgrall properties.

What would settle it

A computed solution on the water-air shock-bubble problem in which any volume fraction, partial density, or pressure falls outside its physical interval when the implicit source step is replaced by an explicit step or when the time step is increased beyond the usual explicit CFL limit.

Figures

Figures reproduced from arXiv: 2604.25672 by A-Man Zhang, Fan Zhang, Jia-Jun Zou, Qi Kong, Yu-Chang Liu, Yun-Long Liu.

Figure 3.1
Figure 3.1. Figure 3.1: Illustration of the 4x4 Gauss-Lobatto points and the discontinuous char￾acteristic on element’s interface. We can write (2.10) in the following compact form: ∂W ∂t + ∇ · F(W, z1) = 0, ∂z1 ∂t + u · ∇z1 = 0. (3.29) For the conservative variables W, we directly present the spatial discretization in element Ii,j as follows: Z Ii,j ∂W ∂t φ m i,j dS = Z Ii,j F(W, z1) · ∇φ m i,jdS − Z ∂Ii,j Fˆ(W, z1) · nφ m i,j… view at source ↗
Figure 3.2
Figure 3.2. Figure 3.2: Illustration of the boundary and inner boundary of element Ii,j . The integration of inner boundary of element Ii,j is discretized by: Z ∂Iin i,j φ m i,j (u · nz1) dl = Z y j+ 1 2 y j− 1 2 view at source ↗
Figure 5.3
Figure 5.3. Figure 5.3: Numerical results computed by the DG (P 2 ) in case 5.2. Top left: mixture density ρ. Top right: pressure p. Bottom left: velocity u. Bottom right: volume fraction z1. 5.3 double rarefaction problem The test case is given to verify the BP property of of the BP-OEDG scheme combined with the operator splitting method. The initial conditions are given as follows: (ρ1, ρ2, u, p, z1) = (2.0, 2.0, −1.0, 0.2, 1… view at source ↗
Figure 5.4
Figure 5.4. Figure 5.4: Numerical results computed by the DG (P 2 ) in case 5.3. Top left: pressure p. Top right: internal energy ρe. Bottom left: velocity u. Bottom right: mixture density ρ. 23 view at source ↗
Figure 5.5
Figure 5.5. Figure 5.5: Volume fraction z1 in case 5.3. 5.4 Gas-liquid Riemman problem This test case is a gas-liquid Riemann problem. Similar test cases had been studied in [16, 33]. The initial conditions are given as follows: (ρ1, ρ2, u, p, z1) = (1.27, 1.0, 0.0, 8000.0, 1.0 − 10−10), −5 < x ≤ 0, (ρ1, ρ2, u, p, z1) = (1.27, 1.0, 0.0, 1.0, 10−10), 0 < x < 5, with γ1 = 1.4, pw,1 = 0.0 and γ2 = 7.15, pw,2 = 3309. The computatio… view at source ↗
Figure 5.6
Figure 5.6. Figure 5.6: Numerical results computed by the DG (P 2 ) in the gas-liquid Riemann problem. Top left: pressure p. Top right: mixture density ρ. Bottom left: energy density ρe. Bottom right: volume fraction z1. 5.5 gas-liquid shock tube This test case is a gas-liquid shock tube problem with both high ratios of pressure and density, the initial conditions are given as follows: (ρ1, ρ2, u, p, z1) = (1.0, 200.0, 0.0, 105… view at source ↗
Figure 5.7
Figure 5.7. Figure 5.7: Numerical results computed by the DG (P 2 ) in case 5.3. Top left: logarithm of pressure log10(p). Top right: mixture density ρ. Bottom left: velocity u. Bottom right: volume fraction z1. 5.6 Air shock hit helium bubble In this case, we simulate Hass and Sturtevant’s experiment [13], where a 1.22 Mach number shock wave hits a helium bubble. This study has been widely used [34, 35] to investigate the perf… view at source ↗
Figure 5.8
Figure 5.8. Figure 5.8: Initial configuration for shock-helium bubble interaction. The Fig.5.9 shows the comparison between the BP-OEDG (P 2 ) numerical results of density and the experiment at t = 62, 245, 427, and 983 µs (measured relative to the time where the shock first hits the bubble [34, 24]). It can be seen that the shape of helium bubble agree with the experiment at the same instance, and the shock and contact discont… view at source ↗
Figure 5.9
Figure 5.9. Figure 5.9: Air shock hitting helium bubble: BP-OEDG (P 2 ) vs experiment 28 view at source ↗
Figure 5.10
Figure 5.10. Figure 5.10: Initial configuration of shock-bubble interaction in water. 29 view at source ↗
Figure 5.11
Figure 5.11. Figure 5.11: Water shock hitting air bubble at different times 6 Conclusion In this paper, we propose a BP-OEDG scheme combined with operator splitting method for solving Kapila five-equation model to simulate compressible two-phase flow. The operator splitting method is used to address the stiffness-induced instability of the κ source term. Implicit solver is used in subsystem 3.23 to maintain the BP property of th… view at source ↗
read the original abstract

This paper presents a high-order bound-preserving oscillation-eliminating discontinuous Galerkin (BP-OEDG) scheme for simulating gas-gas and gas-liquid two-phase flows governed by the Kapila five-equation model with the Tammann equation of state (EOS). The primary computational bottleneck arises from the severe CFL restriction imposed by the stiff $\kappa$-source term in the volume fraction equation. To circumvent this, we propose a novel operator-splitting strategy that decouples the system into a transport model and a stiff $\kappa$-source term. The former is discretized via a quasi-conservative DG method \cite{cheng2020quasi}, while the latter is resolved by an adaptive implicit strategy hybridizing the backward Euler and SDIRK2 methods. We rigorously prove that this implicit treatment is unconditionally BP, effectively removing the stiffness-induced stability constraints inherent in traditional explicit schemes. To further enhance precision, a velocity divergence reconstruction inspired by the Local Discontinuous Galerkin (LDG) method is integrated into the implicit solver. Furthermore, an OE limiter is employed to suppress spurious oscillations without characteristic decomposition, complemented by a BP limiter to ensure the BP property of partial densities, pressure, and volume fraction. Crucially, we prove that the proposed BP-OEDG scheme, integrated with the splitting strategy, strictly satisfies the Abgrall condition. Extensive numerical experiments, including challenging water-air shock-bubble interactions, demonstrate the superior robustness and efficiency of the method.

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 / 2 minor

Summary. The manuscript presents a high-order bound-preserving oscillation-eliminating discontinuous Galerkin (BP-OEDG) scheme for the Kapila five-equation model of compressible two-phase flows with Tammann EOS. It introduces a novel operator-splitting strategy that decouples the transport equations (discretized by a quasi-conservative DG method) from the stiff kappa-source term in the volume-fraction equation (treated by an adaptive implicit solver hybridizing backward Euler and SDIRK2). The paper claims rigorous proofs that the implicit treatment is unconditionally bound-preserving and that the full BP-OEDG scheme with splitting strictly satisfies the Abgrall condition; an OE limiter and BP limiter are added, and numerical tests on water-air shock-bubble interactions are shown.

Significance. If the central proofs hold, the work removes the severe CFL restriction imposed by stiff source terms while preserving physical bounds and uniform velocity/pressure states, offering a practical efficiency gain for high-order simulations of gas-liquid flows without sacrificing robustness.

major comments (2)
  1. [operator-splitting strategy and Abgrall proof] The proof that the operator-splitting strategy preserves the Abgrall condition exactly (i.e., that the composed transport-plus-implicit-source operator maps uniform states to uniform states without residual splitting error in velocity or pressure) is load-bearing for the main claim, yet the abstract and high-level description provide no explicit steps showing how the implicit update on the kappa-source interacts with the quasi-conservative transport discretization to maintain exact uniformity.
  2. [implicit solver and BP proof] The rigorous proof of unconditional bound-preservation for the hybrid implicit solver (backward Euler + SDIRK2) on the stiff source term is central, but the manuscript must supply the full derivation, including the specific update formulas and the argument that positivity of partial densities, pressure, and volume fraction is guaranteed independently of the time-step size.
minor comments (2)
  1. [implicit solver description] The velocity-divergence reconstruction inspired by LDG is mentioned but its precise incorporation into the implicit step (e.g., which variables are reconstructed and at which quadrature points) should be stated explicitly to allow reproducibility.
  2. [numerical results] Numerical experiments would benefit from a table reporting maximum CFL numbers achieved and a direct comparison of wall-clock time against a standard explicit scheme on the same meshes.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the positive assessment of the work's significance and for the constructive comments. We respond point-by-point to the major comments below, clarifying the location of the proofs in the manuscript and indicating revisions to improve their presentation and accessibility.

read point-by-point responses

  1. Referee: [operator-splitting strategy and Abgrall proof] The proof that the operator-splitting strategy preserves the Abgrall condition exactly (i.e., that the composed transport-plus-implicit-source operator maps uniform states to uniform states without residual splitting error in velocity or pressure) is load-bearing for the main claim, yet the abstract and high-level description provide no explicit steps showing how the implicit update on the kappa-source interacts with the quasi-conservative transport discretization to maintain exact uniformity.

    Authors: We thank the referee for this observation. The proof that the splitting strategy preserves the Abgrall condition is given in Section 4.2: the quasi-conservative DG discretization of the transport equations is shown to map uniform velocity/pressure states to uniform states, after which the implicit update on the volume-fraction equation (which leaves momentum and energy unchanged) preserves uniformity exactly. To address the concern that these steps are not visible at a high level, we will add a brief outline of the argument to the abstract and the introductory description in the revised manuscript. revision: partial

  2. Referee: [implicit solver and BP proof] The rigorous proof of unconditional bound-preservation for the hybrid implicit solver (backward Euler + SDIRK2) on the stiff source term is central, but the manuscript must supply the full derivation, including the specific update formulas and the argument that positivity of partial densities, pressure, and volume fraction is guaranteed independently of the time-step size.

    Authors: We agree that the full derivation must be supplied clearly. Section 3.3 contains the proof, including the explicit update formulas for the hybrid backward-Euler/SDIRK2 solver and the analysis showing that positivity of partial densities, pressure (via the Tammann EOS), and volume fraction holds for any time-step size. We will revise the manuscript to present this material in a dedicated, self-contained subsection with all intermediate steps written out explicitly. revision: yes

Circularity Check

0 steps flagged

No significant circularity; proofs and splitting strategy are independent of inputs

full rationale

The derivation relies on a novel operator-splitting strategy whose effect on Abgrall preservation is established by new proofs rather than reduction to prior fitted quantities or self-citations. The quasi-conservative DG discretization is imported from an external reference (cheng2020quasi) treated as an independent building block, while the implicit source treatment and overall BP/Abgrall properties are proved directly from the split operators without self-definitional loops or renaming of known results. No load-bearing step collapses to its own inputs by construction.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The scheme rests on the Kapila five-equation model and Tammann EOS as given; no new free parameters, ad-hoc axioms, or invented entities are introduced beyond standard numerical analysis assumptions for DG methods.

axioms (2)
  • domain assumption The Kapila five-equation model with Tammann EOS accurately describes the target gas-gas and gas-liquid flows.
    Invoked in the abstract as the governing system; no derivation provided.
  • ad hoc to paper The operator splitting decouples transport and stiff source without order reduction or loss of conservation properties.
    Central to the proposed strategy; assumed to hold for the proofs.

pith-pipeline@v0.9.0 · 5581 in / 1464 out tokens · 58639 ms · 2026-05-07T15:16:23.514019+00:00 · methodology

discussion (0)

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