Pith. sign in

REVIEW 2 major objections 6 minor 36 references

Fourth-order Adaptive Mesh Refinement both in space and in time for incompressible Navier-Stokes equations with Dirichlet boundary conditions

T0 review · 2 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The paper proposes a subcycled AMR projection method for the incompressible Navier-Stokes equations with Dirichlet conditions that achieves fourth-order accuracy in space and time by preserving divergence decay at coarse-fine interfaces.

desk verdict A genuine fourth-order subcycled AMR method for incompressible Navier-Stokes, with clean velocity convergence tables, but equation (6.5) has a reversed time-ordering bug that must be fixed before the paper is reproducible. read the letter →

arxiv 2506.02663 v2 pith:XBXGBJPZ submitted 2025-06-03 math.NA cs.NA

Shubo Zhao , Qinghai Zhang This is my paper
classification math.NAcs.NA MSC 76D0565M5076M12
keywords incompressibleNavier-StokesequationsadaptivemeshrefinementsubcyclingintimeprojectionmethodsGePUP-EformulationfinitevolumegeometricmultigridDirichletboundaryconditions
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

This paper sets out to establish the first fourth-order subcycled adaptive mesh refinement method for the incompressible Navier-Stokes equations with Dirichlet boundary conditions: refine in space and take smaller time steps on finer levels, without losing the fourth-order accuracy of the projection. The core difficulty is that the divergence-free constraint and the associated pressure-Poisson solvability conditions must survive across coarse-fine interfaces when coarse data are not yet available at fine times. The authors' answer is to evolve a non-solenoidal velocity $\mathbf{w}$, obtain the physical velocity $\mathbf{u}$ by projection at each stage, and derive interface conditions that make $\nabla\cdot\mathbf{w}$ decay exponentially on every refinement level. Interface values are supplied by fourth-order interpolation using four stored coarse time levels, with solvability corrections enforced to machine precision on each connected subdomain component. In the reported benchmarks, velocity converges at fourth order on static and dynamic hierarchies, and the method runs roughly 9 to 29 times faster than uniform grids.

What carries the argument

The load-bearing mechanism is the GePUP-E reformulation: instead of evolving the solenoidal velocity, the method evolves a velocity $\mathbf{w}$ that is not required to be divergence-free, then obtains the physical velocity $\mathbf{u} = \mathbf{P}\mathbf{w}$ and the auxiliary scalar $q$ from $\mathbf{w}$ at each stage, so that $\nabla\cdot\mathbf{w}$ obeys a heat equation with homogeneous Dirichlet boundary condition and decays exponentially. Around that core, the machinery consists of the coarse-fine interface conditions that preserve this decay on each refinement subdomain; four stored coarse time levels $W_\ell(\tau)$ feeding a fourth-order Taylor interpolation matrix $I^{(4)}(k,\eta)$; a constant correction $\delta_c^w$ per connected component that enforces the discrete solvability condition to machine precision; and a composite projection applied at synchronization without the fine-to-coarse averaging steps of earlier AMR projection methods.

What would settle it

Run Algorithm 6.1 on the Taylor-Green vortex with a three-level hierarchy and refinement ratio $r=4$; if the $L^2$ velocity error between coarsest levels $h_0=1/64$ and $h_0=1/128$ does not shrink by a factor close to 16, the fourth-order claim fails. A second check is to monitor $\max|\nabla\cdot\mathbf{w}|$ on each level of the viscous box test over time; if any level shows divergence growth rather than exponential decay while the mesh is fixed, the interface conditions are not doing their stated job.

Watch

Extended reading notes

Core claim

The authors claim that Algorithm 6.1 is the first subcycled AMR projection method for the incompressible Navier-Stokes equations with Dirichlet or no-slip conditions that is fourth-order accurate in both space and time. The method avoids refluxing at coarse-fine interfaces and never applies fine-to-coarse averaging to the projected velocity; instead, synchronizing levels with the composite projection $\mathbf{P}^{\mathrm{comp}}_{\ell} = \mathbf{I} - \mathbf{G}^{\mathrm{comp}}_{\ell}(\mathbf{L}^{\mathrm{comp}}_{\ell})^{-1}\mathbf{D}^{\mathrm{comp}}_{\ell}$ and omitting the fine-to-coarse replacement step recovers fourth-order velocity convergence. The design rests on the GePUP-E formulation, in which the divergence of the evolutionary velocity solves a heat equation with homogeneous Dirichlet data, giving exponential decay on every level. Interface conditions are computed by fourth-order spatiotemporal Taylor interpolation from four stored coarse time levels, and the discrete compatibility conditions are enforced to machine precision on each connected subdomain. In the reported tests, velocity errors in $L^\infty$, $L^1$, and $L^2$ decrease by roughly the fourth-power factor under refinement, while pressure is observed at third order in $L^2$ in the viscous box test.

Load-bearing premise

The load-bearing premise is that four stored coarse time levels of the velocity capture the local time history of the exact solution well enough for the fourth-order Taylor interpolation to deliver interface values with $O(k^4)$ error, and that the constant correction per connected component does not contaminate that accuracy; the paper offers convergence tables rather than a discrete error estimate for this step.

Editorial extensions

If this is right

  • If the central claim holds, subcycling in time can be combined with fourth-order spatial accuracy for the incompressible Navier-Stokes equations under no-slip or Dirichlet conditions, not just for hyperbolic conservation laws or compressible flows.
  • The exponential divergence-decay mechanism implies that velocity divergence is actively controlled on every level of the AMR hierarchy, not merely on the coarsest level.
  • Because refluxing and fine-to-coarse averaging are absent, the synchronization step is simpler than in prior second-order block-structured AMR projection methods and avoids their third-order error contamination at the coarse-fine interface.
  • The reported speedup of about 9 to 29 over uniform grids at the same finest resolution suggests that the approach makes simulations with localized flow structures substantially cheaper.
  • The use of an implicit-explicit time integrator makes the same AMR framework applicable at both low and high Reynolds numbers.

Reading between the lines

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

  • The paper leaves implicit that the absence of fine-to-coarse averaging may mean exact redundancy of composite data is unnecessary; if so, approximately nested or non-nested patch hierarchies could be used, simplifying regridding and load balancing.
  • A testable extension the paper does not run is replacing the stored four-level history $W_\ell(\tau)$ with dense-output reconstruction inside the implicit-explicit integrator and comparing where the $O(k^4)$ interpolation error concentrates; the authors note this alternative but do not test it.
  • Because pressure converges at third order in the corner tests while velocity is fourth order, a natural next step is a pressure-specific interface correction aiming for fourth-order pressure as well; the paper does not propose one.
  • The construction should transfer to quadtree or octree adaptive grids, as the authors suggest, but the concrete test would be a hanging-node variant of the Taylor-Green run; until that is done, the transplant is a conjecture.
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 presents a fourth-order projection method for the incompressible Navier-Stokes equations with Dirichlet boundary conditions on block-structured AMR hierarchies with subcycling in time. The method is built on the GePUP-E formulation, derives coarse-fine interface conditions intended to preserve exponential decay of velocity divergence on refined subdomains, approximates those conditions by spatiotemporal interpolation from four stored coarse time levels, enforces discrete solvability conditions to machine precision, synchronizes levels through a composite projection without fine-to-coarse averaging of projected velocities, and uses geometric multigrid for the resulting linear systems. Numerical tests cover Taylor-Green vortex, a viscous box with Re=100, four-way vortex merging, single-vortex evolution, dipole vortex-wall collision, and speedup measurements over uniform grids.

Significance. If the reported accuracy and efficiency hold, the paper is a substantial contribution to high-order AMR for incompressible flow: it offers a concrete algorithmic route to fourth-order subcycled projection methods with Dirichlet conditions, a gap that the authors correctly identify. The Taylor-Green test in Table 1 and the static-AMR viscous box test in Table 2(b) show clean fourth-order velocity convergence against independent analytic or reference solutions, and the multigrid residual histories in Figure 6 give useful evidence for the optimal-complexity claim. The paper is also commendably explicit about which design choice is harmful, namely appending fine-to-coarse averaging (COH-4) to the composite projection. The main reservation is that the printed interpolation formula (6.5) is internally inconsistent with the stated data ordering, and no code is provided to confirm that the reported tables were produced with a corrected formula.

major comments (2)
  1. [§6.3, Eq. (6.5)] The interpolation formula in (6.5) is internally inconsistent with the data ordering defined in (6.4). For t* = t_{-0}, i.e., eta = k_{\ell-1}, the first row of the printed I^(4) becomes [0,0,0,1], so (6.5) returns phi(t_{-3}) instead of phi(t_{-0}). The second row at the same point is [-2,9,-18,11]/(6k), which is the derivative stencil for the reversed ordering t_{-3}, t_{-2}, t_{-1}, t_{-0}. Thus the printed matrix is the correct interpolation matrix for the data vector [phi(t_{-3}), phi(t_{-2}), phi(t_{-1}), phi(t_{-0})], not for the vector written in (6.5). Since every coarse-fine interface condition in (6.1a-c) is evaluated through (6.5) at fine stages, a literal implementation would use coarse data from the wrong time levels, and the fourth-order rates in Section 7 could not occur. The authors must correct the ordering in (6.4) or (6.5) (or both), state the intended convention explicitly, and provide a small independent check of the interpolation, for example by applying (6.5) to a cubic polynomial.
  2. [§6.3–§7] The fourth-order claim rests entirely on the convergence tables; no discrete error analysis or explicit regularity assumptions are given for the coupled use of AMRCFI, the temporal interpolation (6.5), and the correction (6.7). In particular, (6.7) adds a constant correction delta_c^w to all interpolated normal fluxes on each connected component, and the paper does not justify that this constant preserves O(k^4) accuracy of the face data in (6.1b,c). Because the interface data are the only coupling between levels in the subcycled recursion, the authors should either supply a proof sketch or explicitly state the regularity and consistency assumptions under which fourth-order velocity convergence is expected, and discuss whether the correction (6.7) changes the local order near the interface.
minor comments (6)
  1. [§6.3] The text says 'Talor expansions' and should read 'Taylor expansions'.
  2. [§7.3 and Table 3] The term 'vertex merging' is used in the table caption and surrounding text; it should be 'vortex merging'.
  3. [Algorithm 6.1] The symbol t0 is used both for the constant initial time (input parameter) and for the current coarsest-level time in the while condition; this overloading is confusing and should be resolved by renaming one of the two quantities.
  4. [§6.2] The text refers to 'line 12' and 'line 10' as algorithmic steps, but line 12 in Algorithm 6.1 is the closing 'end if'; the references should be renumbered or rephrased.
  5. [Conclusion and Table 2(b)] The pressure rates in Table 2(b) are far from fourth order (for example, p L^infinity rate 1.80, p L^2 rates 3.78 and 3.27), and the concluding statement that the pressure is 'third-order accurate in the L^2 norm' is not clearly established by these numbers; additional refinement levels or a more qualified statement are needed.
  6. [General] No code or reproducible data are provided; given the inconsistency in (6.5), a verification artifact or an independent numerical check of the interpolation would substantially strengthen the manuscript.

Circularity Check

1 steps flagged · score 2.0 of 10

No significant circularity: the central fourth-order AMR claim is validated against independent analytic and reference solutions; the only overlap concern is the GePUP-E equivalence and decay theorem cited to the authors' own preprint.

  1. other [Section 2, Definition 2.1 through Theorem 2.2]
    "By the arguments in [19, Section 3], it is straightforward to prove the equivalence of the INSE in (1.1) and the GePUP-E in (2.1). ... Theorem 2.2. ... Proof. See [19, Theorem 4]."

    The exponential decay of velocity divergence is a stated core feature of the method, and the paper proves the analogous single-level result in Theorem 6.2. However, the foundational GePUP-E equivalence and Theorem 2.2 are not proved here; both are imported from reference [19], an overlapping-author preprint that includes the present corresponding author and first author. Thus a load-bearing stability premise is supported by self-citation rather than by a proof inside this manuscript or by an external machine-checked source. This is a minor circularity only, because Section 7 independently tests the complete algorithm against the analytic Taylor-Green solution and reference grids, so the central numerical claim does not reduce to the citation.

full rationale

The central claim of the paper is a fourth-order subcycled AMR projection method for INSE with Dirichlet conditions, and the main algorithmic novelty is tested in Section 7 against exact solutions (Taylor-Green vortex), corner-singularity benchmarks, and reference single-level grids. Those tests are not fitted inputs: the method parameters are fixed, and the reported convergence rates emerge from comparing computed solutions with independent analytic or converged reference data. The projection operator, spatial discretizations, and AMRCFI are cited from the corresponding author's prior work, but they are also exercised through the same external benchmarks, so those citations are not circular in the sense of renaming a fit as a prediction. One minor overlap is the GePUP-E foundation: the equivalence of INSE and GePUP-E and the exponential decay theorem are delegated to reference [19], an unpublished preprint by overlapping authors, rather than proved in this manuscript. This is flagged per the reviewing rule, but it is not fully load-bearing because the full method is independently validated. Separately, a non-circular correctness concern is noted: in Eq. (6.5), with W_ell(τ) ordered as t_{-0}, t_{-1}, t_{-2}, t_{-3}, setting η = k (i.e., t* = t_{-0}) makes the first row of I^(4)(k,η) equal [0,0,0,1], so the printed formula would return φ(t_{-3}) rather than φ(t_{-0}). This appears to be a reversed ordering of the stored time levels and would invalidate the stated O(k^4) interface interpolation if implemented literally; the convergence tables suggest the code uses a corrected ordering, but the manuscript does not state this. That issue concerns correctness, not circularity, and does not raise the circularity score.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The central claim rests on continuous-PDE results adapted from the authors' earlier GePUP-E work, fourth-order finite-volume operators, and conservative AMRCFI interpolation. The key discrete design choices (no COH-4, four-level temporal interpolation, solvability corrections) are justified mainly by numerical experiments rather than by a priori error analysis. No new physical entities are introduced.

free parameters (5)
  • penalty parameter lambda = lambda = 1 in all tests
    Chosen constant in (2.1f) rather than fitted; affects how strongly the normal boundary condition is enforced.
  • refinement criterion thresholds eps_omega = 1 (Sec 7.3 dynamic), 0.15/0.8 (Sec 7.4), 0.05/0.2 (Sec 7.5)
    User-specified thresholds for tagging cells by |curl u|; determine which regions are refined and the measured speedups.
  • Courant number Cr = 0.1 in Tests 7.1 and 7.2; 0.5 in Tests 7.3-7.5
    Sets time step size relative to grid and velocity; standard CFL-like choice, not fitted to convergence rates.
  • multigrid smoother weight and iteration parameters = omega = 2/3; 4 pre/post smoothing; residual tolerances 1e-8 and 1e-12
    Implementation settings for the V-cycle; stability and convergence of linear solvers depend on them but the order claim does not.
  • refinement ratio r and level count lmax = r=2 or 4; lmax=1 or 2 depending on test
    Standard AMR parameters chosen per test; the method's design allows r=2 or 4.
assumptions (6)
  • standard math Heat-equation maximum principle implies exponential decay of div w (Theorem 2.2, Theorem 6.2).
    Used to claim numerical stability from continuous divergence decay; the discrete analogue is not proved.
  • domain assumption Equivalence of the GePUP-E formulation (2.1) to INSE (1.1).
    Taken from [19], a paper by overlapping authors described as 'minor revision submitted'; no proof of equivalence appears in this manuscript.
  • domain assumption Fourth-order accuracy of the finite-volume operators G, D, L and of the projection P in (3.3) and (3.5).
    Cited to [33,34], both by the corresponding author; the AMR composite operators inherit this order.
  • domain assumption AMRCFI coarse-fine interpolation is conservative and fifth-order accurate; its stencil validity follows from triangular lattices [36].
    Heuristic stencil generation is stated to be proved in [36] by Zhang, Zhu and Li; the proof is not repeated.
  • ad hoc to paper Four stored coarse time levels W_ell (6.4) and the interpolation matrix (6.5) give fourth-order temporal interface data.
    The paper assumes 3 initialization steps by synchronized AMR are sufficient and that w has enough temporal regularity for O(k^4) interpolation; no error analysis for the full subcycled algorithm is supplied.
  • ad hoc to paper The composite projection (5.10) without COH-4 preserves global fourth-order accuracy.
    Subsection 6.2 states this is demonstrated numerically (Table 2b) and that COH-4 reduces order to 3; no theoretical proof is given.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fourth-order Adaptive Mesh Refinement both in space and in time for incompressible Navier-Stokes equations with Dirichlet boundary conditions." pith.science (2026). https://pith.science/paper/XBXGBJPZ

@misc{pith2026250602663,
  author       = {Pith},
  title        = {Pith review of: Fourth-order Adaptive Mesh Refinement both in space and in time for incompressible Navier-Stokes equations with Dirichlet boundary conditions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XBXGBJPZ}},
  note         = {Machine review of arXiv:2506.02663}
}
read the original abstract

We present a fourth-order projection method with adaptive mesh refinement (AMR) for numerically solving the incompressible Navier-Stokes equations (INSE) with subcycling in time. Our method features (i) a reformulation of INSE so that the velocity divergence decays exponentially on the coarsest level, (ii) a derivation of coarse-fine interface conditions that preserves the decay of velocity divergence on any refinement level of the AMR hierarchy, (iii) an approximation of the coarse-fine interface conditions via spatiotemporal interpolations to facilitate subcycling in time, (iv) enforcing to machine precision solvability conditions of elliptic equations over each connected component of the subdomain covered by any refinement level, (v) a composite projection for synchronizing multiple levels, and (vi) geometric multigrid for solving linear systems with optimal complexity. Different from current block-structured AMR algorithms, our method never adopts refluxing at the coarse-fine interface, nor is fine-to-coarse averaging applied to projected velocities. Results of numerical tests demonstrate the high accuracy and efficiency of the proposed method.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages

  1. [1]

    A. S. Almgren, J. B. Bell, P. Colella, L. H. Howell, and M. L. Welcome, A conser- vative adaptive projection method for the variable density incompressible Navier–Stokes equations, J. Comput. Phys., 142 (1998), pp. 1–46

    work page 1998

  2. [2]

    J. B. Bell, M. J. Berger, J. Saltzman, and M. Welcome, Three-dimensional adaptive mesh refinement for hyperbolic conservation laws, SIAM J. Sci. Comput., 15 (1994), pp. 127–138

    work page 1994

  3. [3]

    J. B. Bell, L. H. Howell, and P. Colella, An efficient second-order projection method for viscous incompressible flow , in 10th Computational Fluid Dynamics Conference, 1991, p. 1560

    work page 1991

  4. [4]

    M. J. Berger, Data structures for adaptive grid generation , SIAM J. Sci. Stat. Comput., 7 (1986), pp. 904–916

    work page 1986

  5. [5]

    M. J. Berger and P. Colella, Local adaptive mesh refinement for shock hydrodynamics , J. Comput. Phys., 82 (1989), pp. 64–84. FOURTH-ORDER AMR FOR INCOMPRESSIBLE NA VIER-STOKES 25

    work page 1989

  6. [6]

    M. J. Berger and R. J. LeVeque, Implicit adaptive mesh refinement for dispersive tsunami propagtion, SIAM J. Sci. Comput., 46 (2024), pp. B554–B578

    work page 2024

  7. [7]

    M. J. Berger and J. Oliger, Adaptive mesh refinement for hyperbolic partial differential equations, J. Comput. Phys., 53 (1984), pp. 484–512

    work page 1984

  8. [8]

    M. J. Berger and I. Rigoutsos, An algorithm for point clustering and grid generation , IEEE Trans. Syst., Man, Cybern., 21 (1991), pp. 1278–1286

    work page 1991

Show all 36 references
  1. [9]

    W. L. Briggs, V. E. Henson, and S. F. McCormick, A Multigrid Tutorial , SIAM, Philadel- phia, 2nd ed., 2000

    2000

  2. [10]

    Christopher, S

    J. Christopher, S. M. Guzik, and X. Gao, High-order implicit-explicit additive Runge–Kutta schemes for numerical combustion with adaptive mesh refinement , Int. J. Numer. Meth. Fluids, 94 (2022), pp. 1082–1110

    2022

  3. [11]

    Coquel, Q

    F. Coquel, Q. L. Nguyen, M. Postel, and Q. H. Tran, Local time stepping applied to implicit-explicit methods for hyperbolic systems , Multiscale Model. Simul., 8 (2010), pp. 540–570

    2010

  4. [12]

    Cozzi and R

    E. Cozzi and R. L. Pego, On optimal estimates for the Laplace-Leray commutator in planar domains with corners , Proc. Amer. Math. Soc., 139 (2011), pp. 1691–1706

    2011

  5. [13]

    Emmett, E

    M. Emmett, E. Motheau, W. Zhang, M. Minion, and J. Bell, A fourth-order adaptive mesh refinement algorithm for the multicomponent, reacting compressible Navier-Stokes equations, Combust. Theory Model., 23 (2019), pp. 592–625

    2019

  6. [14]

    M. J. Grote, M. Mehlin, and T. Mitkova, Runge–kutta-based explicit local time-stepping methods for wave propagation , SIAM J. Sci. Comput., 37 (2015), pp. A747–A775

    2015

  7. [15]

    L. H. Howell and J. B. Bell, An adaptive mesh projection method for viscous incompressible flow, SIAM J. Sci. Comput., 18 (1997), pp. 996–1013

    1997

  8. [16]

    D. Hu, K. Liang, L. Ying, S. Li, and Q. Zhang, ARMS: Adding and removing markers on splines for high-order general interface tracking under the MARS framework , J. Comput. Phys., 521 (2025), p. 113574

    2025

  9. [17]

    C. A. Kennedy and M. H. Carpenter, Additive Runge–Kutta schemes for convection– diffusion–reaction equations, Appl. Numer. Math., 44 (2003), pp. 139–181

    2003

  10. [18]

    Kramer, H

    W. Kramer, H. J. H. Clercx, and G. J. F. van Heijst, Vorticity dynamics of a dipole colliding with a no-slip wall , Phys. Fluids, 19 (2007), p. 126603

    2007

  11. [19]

    Y. Li, X. Wu, J. Yan, J. Yang, Q. Zhang, and S. Zhao, GePUP-ES: High-order energy-stable projection methods for the incompressible Navier-Stokes equations with no-slip conditions , J. Sci. Comput., minor revision submitted (2025), https://arxiv.org/abs/2409.11255

    2025 arXiv

  12. [20]

    J.-G. Liu, J. Liu, and R. L. Pego, Stability and convergence of efficient Navier-Stokes solvers via a commutator estimate , Commun. Pure Appl. Math., 60 (2007), pp. 1443–1487

    2007

  13. [21]

    D. F. Martin, P. Colella, and D. Graves, A cell-centered adaptive projection method for the incompressible Navier–Stokes equations in three dimensions , J. Comput. Phys., 227 (2008), pp. 1863–1886

    2008

  14. [22]

    McCorquodale and P

    P. McCorquodale and P. Colella, A high-order finite-volume method for conservation laws on locally refined grids , Commun. Appl. Math. Comput. Sci., 6 (2011), pp. 1–25

    2011

  15. [23]

    M. L. Minion, A projection method for locally refined grids , J. Comput. Phys., 127 (1996), pp. 158–178

    1996

  16. [24]

    Osher and R

    S. Osher and R. Sanders, Numerical approximations to nonlinear conservation laws with locally varying time and space grids , Math. Comput., 41 (1983), pp. 321–336

    1983

  17. [25]

    R. R. Rosales, B. Seibold, D. Shirokoff, and D. Zhou, High-order finite element methods for a pressure poisson equation reformulation of the Navier-Stokes equations with electric boundary conditions, Comput. Methods Appl. Mech. Engrg., 373 (2021), p. 113451

    2021

  18. [26]

    Saad, Iterative Methods for Sparse Linear Systems , SIAM, Philadelphia, 2nd ed., 2003

    Y. Saad, Iterative Methods for Sparse Linear Systems , SIAM, Philadelphia, 2nd ed., 2003

    2003

  19. [27]

    Shirokoff and R

    D. Shirokoff and R. R. Rosales, An efficient method for the incompressible Navier-Stokes equations on irregular domains with no-slip boundary conditions, high order up to the boundary, J. Comput. Phys., 230 (2011), pp. 8619–8646

    2011

  20. [28]

    Y. Tan, Y. Qian, Z. Li, and Q. Zhang, The multiphase cubic MARS method for fourth- and higher-order interface tracking of three or more materials with arbitrarily complex topology and geometry, 2025, https://arxiv.org/abs/2506.11897

    2025 arXiv

  21. [29]

    G. I. Taylor and A. E. Green, Mechanism of the production of small eddies from large ones , Proc. R. Soc. Lond. Ser. A, Math. Phys. Sci., 158 (1937), pp. 499–521

    1937

  22. [30]

    J. A. van Hooft and S. Popinet, A fourth-order accurate adaptive solver for incompressible flow problems, J. Comput. Phys., 462 (2022), p. 111251

    2022

  23. [31]

    Zhang, AMRCFI: A matlab package for coarse-fine interpolation in adaptive mesh refine- ment, 2011

    Q. Zhang, AMRCFI: A matlab package for coarse-fine interpolation in adaptive mesh refine- ment, 2011. http://sourceforge.net/projects/amrcfi/

    2011

  24. [32]

    Zhang, High-order, multidimensional, and conservative coarse-fine interpolation for adap- tive mesh refinement , Comput

    Q. Zhang, High-order, multidimensional, and conservative coarse-fine interpolation for adap- tive mesh refinement , Comput. Methods Appl. Mech. Engrg., 200 (2011), pp. 3159–3168. 26 SHUBO ZHAO AND QINGHAI ZHANG

    2011

  25. [33]

    Zhang, A fourth-order approximate projection method for the incompressible Navier-Stokes equations on locally-refined periodic domains , Appl

    Q. Zhang, A fourth-order approximate projection method for the incompressible Navier-Stokes equations on locally-refined periodic domains , Appl. Numer. Math., 77 (2014), pp. 16–30

    2014

  26. [34]

    Zhang, GePUP: Generic projection and unconstrained PPE for fourth-order solutions of the incompressible Navier-Stokes equations with no-slip boundary conditions , J

    Q. Zhang, GePUP: Generic projection and unconstrained PPE for fourth-order solutions of the incompressible Navier-Stokes equations with no-slip boundary conditions , J. Sci. Com- put., 67 (2016), pp. 1134–1180

    2016

  27. [35]

    Zhang, H

    Q. Zhang, H. Johansen, and P. Colella, A fourth-order accurate finite-volume method with structured adaptive mesh refinement for solving the advection-diffusion equation , SIAM J. Sci. Comput., 34 (2012), pp. B179–B201

    2012

  28. [36]

    Zhang, Y

    Q. Zhang, Y. Zhu, and Z. Li, An AI-aided algorithm for multivariate polynomial reconstruc- tion on Cartesian grids and the PLG finite difference method , J. Sci. Comput., 101, 66 (2024)

    2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.