Pith. sign in

REVIEW 5 major objections 5 minor 67 references

A preprocessing tool locally refines hybrid unstructured meshes around immersed bodies so that immersed-boundary flow predictions match body-fitted computations.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-01 05:15 UTC pith:KBFZV4TM

load-bearing objection Useful integration paper with a real gap: CODA's handling of the exported hanging-node meshes is never stated. the 5 major comments →

arxiv 2607.27580 v1 pith:KBFZV4TM submitted 2026-07-30 physics.flu-dyn

Mesh Adaptation on Hybrid Unstructured Meshes for Immersed Boundary Methods

Jonatan N\'u\~nez-de la Rosa , Esteban Ferrer , Eusebio Valero This is my paper
classification physics.flu-dyn
keywords mesh adaptationhybrid meshesimmersed boundariesvolume penalizationhanging nodescomputational fluid dynamicslift optimizationunstructured mesh
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper describes a preprocessing tool that locally refines hybrid unstructured meshes—made of tetrahedra, hexahedra, prisms, and pyramids—around a geometry treated as immersed rather than body-fitted. The central idea is to split the full geometry into a fixed part that keeps a body-fitted mesh and a changing part, say a flap, that is immersed and locally refined. The authors establish that this hybrid body-fitted/immersed approach can reproduce force coefficients from full body-fitted simulations and experiments, as shown for flow past a cylinder, a NACA0012 airfoil, a multi-element airfoil, and a flap-position optimization. A sympathetic reader would care because it offers a path to reuse a background mesh while optimizing or varying only a component, avoiding full remeshing.

Core claim

On its own terms, the paper claims that the new mesh-refinement tool makes immersed-boundary volume penalization accurate and efficient on hybrid unstructured meshes. The tool takes a conforming hybrid background mesh, defines a refinement box around where the immersed body will sit, refines elements that overlap the body, and balances refinement so neighboring elements differ by at most a 2:1 level ratio. The resulting locally refined meshes, with hanging nodes and nonconforming faces, are exported for the CFD solver. Validation cases match body-fitted and experimental pressure and force coefficients when the mesh is sufficiently refined: the drag for cylinder and NACA0012 is within a few p

What carries the argument

The load-bearing mechanism is a set of element-splitting conventions plus master–slave interface arrays (TRI4TRI for triangular faces, QUAD4QUAD for quadrilateral faces). Each tetrahedron splits into 8 tetrahedra, each hexahedron into 8 hexahedra, each prism into 8 prisms, and each pyramid into 6 pyramids plus 4 tetrahedra; hanging nodes created by neighboring elements at different refinement levels are encoded through the master/slave arrays. Hash tables and kd-trees make the geometry-overlap detection and face-matching efficient. These interfaces drive the 2:1 balancing step that keeps the mesh locally refined without producing more than one refinement level difference between neighbors. T

Load-bearing premise

The pipeline depends on the CFD solver correctly handling hanging-node, nonconforming faces between neighboring elements after refinement; the paper demonstrates refined meshes but does not show the solver's treatment of nonconforming interfaces.

What would settle it

Take a problem with a known smooth solution and run the same pipeline on a hybrid mesh with a deliberately introduced hanging-node face; if the solver does not converge to the analytical solution at the expected rate, or if a refined mesh with hanging nodes fails to reproduce the body-fitted force coefficient on the same geometry, the central claim fails.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • A single body-fitted background mesh can serve a family of configurations: only the changing part, such as a flap, is placed as an immersed body and locally refined.
  • Local refinement around immersed surfaces can replace uniform Cartesian refinement, keeping element counts and runtime practical for industrial cases.
  • The reported force coefficients (drag for cylinder and NACA0012, lift for the multi-element airfoil) approach full body-fitted and experimental values as the refinement level increases.
  • Optimization loops can reuse a pre-refined region that covers all possible positions of the moving component, avoiding a full remesh at each design evaluation.
  • Both finite volume and discontinuous Galerkin spatial discretizations can be coupled with the volume-penalization immersed boundary method on the refined hybrid meshes.

Where Pith is reading between the lines

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

  • Beyond the paper: if the CFD solver truly accepts nonconforming faces, this pipeline could be applied to any hybrid mesh with local refinement, not just two-dimensional extrusions; the same splitting conventions should extend to 3-D refinement of complex aircraft components.
  • The 2:1 balancing rule is a design choice; one could test whether relaxing it to 4:1 or allowing graded refinement reduces element counts without hurting force predictions, though accuracy would need checking near the immersed surface.
  • The optimization example suggests a practical workflow: pre-refine a region large enough to contain all possible flap positions, then move the immersed geometry inside it without remeshing—an implicit cost-saving claim the paper does not quantify directly.
  • Because the immersed geometry is refined by surface-overlap, the tool's accuracy will depend on the facet size of the STL representation relative to the mesh; a sensitivity study on facet density would be a natural next test.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

5 major / 5 minor

Summary. This manuscript presents a preprocessing tool that refines hybrid unstructured (tetrahedral, hexahedral, prismatic, pyramidal) body-fitted background meshes locally around immersed geometries for use with the CODA CFD solver and an immersed-boundary volume-penalization method. The tool constructs nonconforming hanging-node interfaces, balances refinement to a 2:1 level ratio, and exports an HDF5 mesh that is claimed to be directly usable by the solver. The authors validate the approach on 2D-extruded cases: laminar flow past a cylinder (Re=40), laminar flow past an NACA0012 airfoil (Re=5000), and RANS flow past an MDA30P30N multi-element airfoil (Re=9e6), reporting pressure coefficients and force coefficients in reasonable agreement with body-fitted and experimental references. A further example optimizes flap position in a two-element airfoil with GEMSEO. The central claim is that the refined meshes allow the IBM-equipped CODA solver to simulate flows around immersed components accurately and efficiently by avoiding full remeshing.

Significance. If substantiated, the tool would enable a practically useful workflow: a body-fitted, wall-resolved mesh for a fixed geometry, with changing components (flaps, control surfaces) treated as immersed and locally refined, avoiding costly re-meshing. The paper gives a detailed algorithmic description and validates against independent published data without curve-fitting; the reported coefficient matches (e.g., cylinder Cd≈1.53–1.55 vs. 1.56–1.57 in the literature; fine-mesh MDA30P30N Cl=2.776 vs. 2.876 experiment) are encouraging. However, the strongest claims extend to true three-dimensional hybrid meshes, while every validation case is extruded 2D. Moreover, the pipeline's compatibility with CODA hinges on an unstated assumption about nonconforming-interface support, and key numerical parameters (notably η) are not reported. These gaps weaken the paper in its current form but are addressable.

major comments (5)
  1. [§4.1, §4.2, Algorithm 1] The refinement deliberately creates hanging nodes and nonconforming faces (Fig. 4, Section 4.1), and Algorithm 1 exports 'the resulting mesh in HDF5 format, compatible with the CFD solver.' Nowhere is it stated or demonstrated that CODA's finite-volume or discontinuous-Galerkin discretizations accept nonconforming interfaces, nor is any conforming step described before export. The reported successful runs are indirect evidence, but the central compatibility claim ('compatible with the CFD solver') is load-bearing and rests on an unverified external capability. Please state explicitly whether CODA supports hanging-node/nonconforming faces, or add a conforming/interface-treatment step, with a reference or a test.
  2. [§4.2, Algorithms 2 and 4] The element-flagging logic is internally inconsistent. The text preceding Algorithm 4 says an element is flagged when the number of intersecting geometry facets is 'greater than zero,' but Algorithm 4 uses 'if Elem.nFacets > 1.' Algorithm 2 appears to query only a single nearest facet ('Facet←NearestNeighbors(BBox,FacetsBary)') and increments Elem.nFacets by 1, whereas the text says the number of intersecting facets is stored. If the threshold is truly >1, elements crossed by exactly one geometry facet would never be refined, creating gaps in the refined band; if the pseudocode is literal, many overlapping facets are missed. Please clarify the actual implementation and correct the pseudocode, since this is the core of the refinement algorithm.
  3. [§5 and Eq. (44)] The penalization parameter η is central to the volume-penalization source term (Eq. 44) and is listed as a free parameter, but its numerical value is not reported for any test case (cylinder, NACA0012, MDA30P30N, or optimization). Without η (and the refinement box dimensions), the simulations cannot be reproduced. Please provide the η values and box-region sizes used in each case.
  4. [Abstract and §5.3] The abstract claims the meshes enable simulations 'in an accurate and efficient manner,' but no efficiency data are reported: no wall-clock times, CPU-hour counts, mesh-generation times, or comparison with the cost of a full body-fitted re-mesh. The fine MDA30P30N IBM mesh has 4,585,591 elements versus 736,472 for the body-fitted mesh, yet the paper does not discuss whether the IBM+refinement workflow is actually cheaper. Please add quantitative efficiency metrics or temper the efficiency claim.
  5. [§5.1–§5.3 and Conclusions] All validation cases are extruded 2D: the computational domain is a disk extruded one element in the y-direction with symmetry boundary conditions on the front and back faces. The tool's stated capability for genuine three-dimensional hybrid meshes (tetrahedra, pyramids, etc.) is therefore not tested by any three-dimensional case. The conclusions generalize to 3D industrial geometries and mention extensions to 'a billion elements,' but the present results do not substantiate that. Please add at least one truly 3D demonstration or explicitly restrict the scope to 2D-extruded configurations.
minor comments (5)
  1. [§5.1] Typo: 'vecinity' should be 'vicinity'.
  2. [§2.2] The turbulence model is repeatedly called 'Spallart–Allmaras'; the correct spelling is 'Spalart–Allmaras'.
  3. [Figures 8, 10, 12] The legend labels contain garbled characters (e.g., '∝O⌋⊣↕2-BFM'); these should be rendered as 'FV-II-BFM', 'DG-III-IBM', etc.
  4. [§5.4] The optimization result is presented as 'consistent with a similar optimization workflow with only body-fitted meshes' in the Conclusions, but no comparison data or reference to the companion paper is shown. Please provide evidence or qualify the statement.
  5. [Algorithm 6] The loop upper bound 'L' is not defined. It should be the current maximum refinement level; please define it explicitly.

Circularity Check

0 steps flagged

No circular reasoning: the refinement tool is validated against independent body-fitted and experimental data; self-citations are component reuse, not load-bearing derivations.

full rationale

The paper's central claim is that the preprocessing tool refines hybrid unstructured meshes around immersed geometries and that CODA with volume penalization on these meshes reproduces body-fitted and experimental force coefficients. This is not circular: the refinement algorithms (Algorithms 1-7) are defined in terms of element splitting and geometry overlap, not in terms of the output quantities (Cp, Cl, Cd). The validation compares against independent references, including Canuto & Taira [10], Yu & Pantano [67], Crumpton et al. [15], Jawahar & Kamath [26], and Murayama et al. [35], as well as full body-fitted CODA runs. No parameter is fitted to the target coefficients; refinement levels, mesh sizes, and the SA constants are stated independently. The authors' previous volume-penalization implementation [37] is invoked as a component of the workflow, but the present study does not derive its conclusions from that citation; the reported agreement is checkable from the provided tables and figures. The concern that CODA's support for the deliberately created hanging nodes and nonconforming faces is not documented is a real correctness/verifiability risk, not a circularity, because it concerns an external capability assumption rather than an equivalence between output and input.

Axiom & Free-Parameter Ledger

4 free parameters · 3 axioms · 0 invented entities

The central claim depends on: (1) CODA's undocumented ability to accept hanging-node/nonconforming meshes; (2) the robustness of kd-tree overlap detection for arbitrary facet sizes; and (3) watertight STL geometry. There are no invented physical entities. User-selected refinement levels and facet sizes are inputs, not fit parameters; the validation quantities are compared with external data.

free parameters (4)
  • penalization parameter eta = not reported
    Chosen small for volume penalization; not fitted to the target data, but its value is not stated, so results cannot be re-run exactly.
  • maximum refinement level per case = 5, 8, 4, 14, 7 depending on case
    User-selected per test; determines smallest cell size near the immersed body and thus the resolution of the validation. Not fitted to match outputs, but a manually chosen input.
  • geometry facet characteristic length = 5e-3, 2.5e-3, 2.0e-3 depending on case
    STL facet size chosen per case; the paper recommends it be the same order as the refined element size for reliable overlap detection (Algorithm 2).
  • neighbor search radius factor in Algorithm 2 = 10 times element size
    Manual choice controlling how many geometry facets are tested against each element; no sensitivity study is provided.
axioms (3)
  • domain assumption CODA solver accepts meshes with hanging nodes / nonconforming faces produced by the refinement.
    Section 4.1 states refinement creates 'nonconforming faces between elements'; Section 5 runs CODA on these meshes, but the paper never describes any conforming step or CODA's support for nonconforming interfaces. If CODA does not support them, the exported meshes would not run as claimed.
  • domain assumption Geometry overlap detection via kd-tree nearest neighbors within 10 times element size is sufficient for all immersed configurations.
    Algorithm 2 defines the overlap search by a sphere of radius ten times the element size and advises facet size matching element size; no sensitivity study is given for configurations where facets are much smaller or larger.
  • domain assumption The ray-casting mask function correctly classifies inside/outside for the STL surfaces.
    Section 3.4 uses even/odd ray intersections; this presumes watertight, consistently oriented geometry, which is not checked or discussed.

pith-pipeline@v1.3.0-daily-deepseek · 27807 in / 13798 out tokens · 110596 ms · 2026-08-01T05:15:20.654130+00:00 · methodology

0 comments
read the original abstract

In this work, we describe a new preprocessing tool for mesh adaptation on hybrid unstructured meshes with a target application on immersed boundary methods. The tool has as input an unstructured, hybrid, and conforming mesh generated by an external mesh generation software, and the main goal is to refine this mesh around immersed geometries in such a way that the CFD solver using the immersed boundary method can simulate flow problems in an accurate and efficient manner. The input background mesh can be made of different types of elements, like tetrahedra, hexahedra, prisms, and pyramids, which, unlike Cartesian meshes, permit for a more flexible mesh. Hybrid unstructured meshes enable one to use the immersed boundary technology in a new class of flow problems where the full geometry is decomposed into a fixed geometry part and a changing geometry part. A body-fitted mesh is generated for the fixed geometry while for the changing one is used the immersed boundary method. We simulate several flow problems to test the new meshes, including subsonic flow past a cylinder and subsonic flow past an NACA0012 airfoil, both using finite volume and discontinuous Galerkin methods and solving the Navier--Stokes equations. As an industrial example of our mesh generation, we consider the simulation of a multi-element airfoil: in this case, a mesh generation software generates an unstructured conforming background mesh for the slat and main airfoil, while the flap is placed as immersed geometry in this body-fitted mesh. As accurate and efficient results are sought, this mesh is refined around the flap and then the subsonic flow at high-lift flow conditions is simulated with a finite volume method coupled with an immersed boundary method and using the Reynolds--averaged Navier--Stokes equations. The reported numerical simulations are in good agreement with experimental data.

Figures

Figures reproduced from arXiv: 2607.27580 by Esteban Ferrer, Eusebio Valero, Jonatan N\'u\~nez-de la Rosa.

Figure 1
Figure 1. Figure 1: Schematic diagram of a hybrid background mesh for the immersed boundary method. A mask function is used to discriminate between those points and cells that are inside or outside the immersed body. and 0 < 𝜂 ≪ 1 is the penalization parameter. If turbulent flow is simulated with the Reynolds-averaged Navier–Stokes equations and the Spalart–Allmaras model is used, the eddy viscosity, ̃𝜈, is set to zero inside… view at source ↗
Figure 2
Figure 2. Figure 2: Hybrid full body-fitted meshes around complex geometries: unstructured grid of a NASA CRM-HL aircraft (left), and unstructured grid of a multi-element airfoil (right). 1 2 3 4 1 2 3 4 5 6 7 8 9 10 Tetrahedron 1 2 3 4 5 1 2 3 4 5 6 7 9 8 10 11 12 13 14 Pyramid 1 2 3 4 5 6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Prism 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 9 10 12 11 13 14 15 16 17 18 20 19 21 22 23 25 24 26 2… view at source ↗
Figure 3
Figure 3. Figure 3: Splitting of different types of elements implemented in the preprocessing tool. A tetrahedron is split into 8 tetrahedra. A pyramid is split into 6 pyramids and 4 tetrahedra. A prism is split into 8 prisms. A hexahedron is split into 8 hexahedra. computational fluid dynamics community. Next, we present the node numbering convention for the children resulting from the element splitting. In the figure hangin… view at source ↗
Figure 4
Figure 4. Figure 4: Hanging nodes numbering convention used by the preprocessing tool for the TRI4TRI (left) and a QUAD4QUAD (right) interfaces. These interfaces arise between an element with refinement level 𝑙 and a neighboring element that has been refined up to level 𝑙 + 1. level 𝑙 with its direct neighbors that have been refined to the level 𝑙 + 1. We use a convention regarding the nodes numbering to describe the TRI4TRI … view at source ↗
Figure 5
Figure 5. Figure 5: Flow diagram sketching the necessary steps to refine a hybrid mesh around an immersed geometry. level ratio 2:1 between neighboring elements is preserved (algorithm 6), and finally, the refinement of the flagged elements (algorithm 7). The algorithms implemented in the preprocessing tool work mainly on the following data structures: arrays, linked lists, hash tables, and 𝑘𝑑-trees. An array of logical value… view at source ↗
Figure 6
Figure 6. Figure 6: Initial unstructured background mesh created with GMSH [20], and used in the immersed boundary computations. The full computational domain is a disk of radius 𝑅 = 500 (left), with an inner square subdomain with dimensions [−20,+20] × [0, 1] × [−20,+20]. This inner subdomain is made of either hybrid shape elements (center), or only hexahedral elements (right). 10: case HEXA8 11: call SplitElement_HEXA8(Elem… view at source ↗
Figure 7
Figure 7. Figure 7: Flow around a cylinder: computational meshes (top) and contour plots of the velocity magnitude (bottom). The simulation on the body-fitted mesh was performed with a second-order finite volume method (left), the immersed boundary method simulation on the hybrid unstructured mesh with refinement level 𝑙 = 5 was computed with a second-order finite volume method (center), and the immersed boundary method compu… view at source ↗
Figure 8
Figure 8. Figure 8: Flow around a cylinder: comparison of pressure coefficient at Re = 40, 𝑀 = 0.2 and 𝛼 = 0◦ . Values were obtained with a second-order finite volume method on a full body-fitted mesh (FV-2-BFM), a second-order finite volume scheme + immersed boundary method on a hybrid mesh (FV-2-IBM) and a third-order discontinuous Galerkin method + immersed boundary method on a hexahedral mesh (DG-3-IBM) [PITH_FULL_IMA… view at source ↗
Figure 9
Figure 9. Figure 9: Flow around an NACA0012 airfoil: computational meshes (top) and contour plots of the velocity magnitude (bottom). The simulation on the body-fitted mesh was performed with a second-order finite volume method (left), the immersed boundary method simulation on the hybrid unstructured mesh with refinement level 𝑙 = 4 was computed with a second-order finite volume method (center), and the immersed boundary met… view at source ↗
Figure 10
Figure 10. Figure 10: Flow around an NACA0012 airfoil: comparison of pressure coefficient at Re = 5000, 𝑀 = 0.5 and 𝛼 = 0◦ . Values were obtained with a second-order finite volume method on a full body-fitted mesh (FV-2-BFM), a second-order finite volume scheme + immersed boundary method on a hybrid mesh (FV-2-IBM) and a second-order discontinuous Galerkin method + immersed boundary method on a hexahedral mesh (DG-2-IBM) [… view at source ↗
Figure 11
Figure 11. Figure 11: Flow around an MDA30P30N multi-element airfoil: computational meshes (top) and contour plots of the velocity magnitude (bottom). The simulation on the full body-fitted mesh for the wing-slat-flap system was performed with a second-order finite volume method (left), and a second-order finite volume method coupled with the immersed boundary volume penalization method was used in the simulations on the hybri… view at source ↗
Figure 12
Figure 12. Figure 12: Flow around an MDA30P30N multi-element airfoil: comparison of pressure coefficient at Re = 9.0 × 106 , 𝑀 = 0.2 and 𝛼 = 5.5 ◦ . Values were obtained with a second-order finite volume method on a full body-fitted mesh (FV-2-BFM), a second-order finite volume scheme + immersed boundary method on a coarse hybrid mesh and a fine hybrid mesh (FV-2-IBM). Experimental data from Murayama et al. [35] are also sho… view at source ↗
Figure 13
Figure 13. Figure 13: Wing/flap 3D configuration (left) and its 2D cross-section at 𝑦 = 5 m (right). characterized by the variables: gap, overlap, and deflection angle. The gap and overlap variables are measured with respect to the trailing edge coordinates of the wing and the leading edge of the flap. The deflection angle is the angle between the chord of the wing and the chord of the flap, as seen in figure 13 (right). The o… view at source ↗
Figure 14
Figure 14. Figure 14: Flowchart of the optimization methodology used in this work. mesh of the domain containing the wing is created, and this mesh is further refined in the region where the flap is located according to the position given by the optimizer. The CODA CFD solver simulates the flow past the wing/flap system for the current position of the flap, with appropriate initial and boundary conditions according to the high… view at source ↗
Figure 15
Figure 15. Figure 15: Optimization of flap position: Convergence history of the optimization process, showing the angle of attack and flap deflection design variables, and the target variable lift coefficient. 6. Conclusions We have developed a new preprocessing tool that enables mesh adaptation on hybrid unstructured meshes for immersed boundary method simulations, providing flexibility for complex geometries in computational… view at source ↗
Figure 16
Figure 16. Figure 16: Optimization of flap position: contour plots of the velocity magnitude for the baseline configuration of the airfoil-flap system (left) and the optimized airfoil-flap configuration (right). Several numerical computations were performed by solving the Navier–Stokes and Reynolds-Averaged Navier– Stokes equations: the subsonic flow past a cylinder, the subsonic flow past an NACA0012 airfoil, and the subsonic… view at source ↗

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

67 extracted references · 41 canonical work pages

  1. [1]

    Immersed boundary method implemented for the simulationof an external flow on unstructured meshes

    Abalakin, I., Zhdanova, N., Kozubskaya, T., 2016. Immersed boundary method implemented for the simulationof an external flow on unstructured meshes. Mathematical Models and Computer Simulations 8, 219–230. doi:10.1134/S2070048216030029

    work page doi:10.1134/s2070048216030029 2016

  2. [2]

    Navier–Stokes computations and experimental comparisons for multielement airfoil configurations

    Anderson, W., Bonhaus, D., McGhee, R., Walker, B., 1995. Navier–Stokes computations and experimental comparisons for multielement airfoil configurations. Journal of Aircraft 32, 1246–1253. doi:10.2514/3.46871

    work page doi:10.2514/3.46871 1995

  3. [3]

    On the flexibility of agglomeration based physical space discontinuous Galerkin discretizations

    Bassi, F., Botti, L., Colombo, A., Di Pietro, D., Tesini, P., 2012. On the flexibility of agglomeration based physical space discontinuous Galerkin discretizations. Journal of Computational Physics 231, 45–65. doi:10.1016/j.jcp.2011.08.018

    work page doi:10.1016/j.jcp.2011.08.018 2012

  4. [4]

    A high-order accurate discontinuous finite element method for the numerical solution of the compressible Navier– Stokes equations

    Bassi, F., Rebay, S., 1997. A high-order accurate discontinuous finite element method for the numerical solution of the compressible Navier– Stokes equations. Journal of Computational Physics 131, 267–279. doi:10.1006/jcph.1996.5572

    arXiv 1997

  5. [5]

    AerodynamicandFSIanalysisofwindturbineswiththe ALE-VMS and ST-VMS methods

    Bazilevs,Y.,Takizawa,K.,Tezduyar,T.,Hsu,M.C.,Kostov,N.,McIntyre,S.,2014. AerodynamicandFSIanalysisofwindturbineswiththe ALE-VMS and ST-VMS methods. Archives of Computational Methods in Engineering 21, 359–398. doi:10.1007/s11831-014-9119-7

    work page doi:10.1007/s11831-014-9119-7 2014

  6. [6]

    A4th-orderaccuratefinitevolumemethodforidealclassicalandspecialrelativisticMHD based on pointwise reconstructions

    Berta,V.,Mignone,A.,Bugli,M.,Mattia,G.,2024. A4th-orderaccuratefinitevolumemethodforidealclassicalandspecialrelativisticMHD based on pointwise reconstructions. Journal of Computational Physics 499, 112701. doi:10.1016/j.jcp.2023.112701

    arXiv 2024

  7. [7]

    Computational Fluid Dynamics: Principles and Applications

    Blazek, J., 2015. Computational Fluid Dynamics: Principles and Applications. 3 ed., Elsevier Science, Amsterdam

    2015

  8. [8]

    PHOENIX: MHD spectral code for rotating laboratory and gravitating astrophysical plasmas

    Blokland, J.W., van der Holst, B., Keppens, R., Goedbloed, H., 2007. PHOENIX: MHD spectral code for rotating laboratory and gravitating astrophysical plasmas. Journal of Computational Physics 226, 509–533

    2007

  9. [9]

    Acombinedghost-point-forcing/direct-forcingimmersedboundarymethod(IBM) for compressible flow simulations

    Boukharfane,R.,Ribeiro,F.,Bouali,Z.,Mura,A.,2018. Acombinedghost-point-forcing/direct-forcingimmersedboundarymethod(IBM) for compressible flow simulations. Computers & Fluids 162, 91–112. doi:10.1016/j.compfluid.2017.11.018

    work page doi:10.1016/j.compfluid.2017.11.018 2018

  10. [10]

    Two-dimensionalcompressibleviscousflowaroundacircularcylinder

    Canuto,D.,Taira,K.,2015. Two-dimensionalcompressibleviscousflowaroundacircularcylinder. JournalofFluidMechanics785,349–371. doi:10.1017/jfm.2015.635

    work page doi:10.1017/jfm.2015.635 2015

  11. [11]

    Turbulent wall model for immersed boundary methods

    Capizzano, F., 2011. Turbulent wall model for immersed boundary methods. AIAA Journal 49, 2367–2381. doi:10.2514/1.J050466

    work page doi:10.2514/1.j050466 2011

  12. [12]

    Coupling a wall diffusion model with an immersed boundary technique

    Capizzano, F., 2016. Coupling a wall diffusion model with an immersed boundary technique. AIAA Journal 54, 728–734. doi:10.2514/1. J054197

    doi:10.2514/1 2016

  13. [13]

    Automatic generation of locally refined Cartesian meshes: Data management and algorithms

    Capizzano, F., 2018. Automatic generation of locally refined Cartesian meshes: Data management and algorithms. International Journal of Numerical Methods in Engineering 113, 789–813. doi:10.1002/nme.5636

    work page doi:10.1002/nme.5636 2018

  14. [14]

    Journal of Computational Physics 519, 113441

    Constant,B.,Péron,S.,Beaugendre,H.,Benoit,C.,2024.AnimprovedimmersedboundarymethodforturbulentflowsimulationsonCartesian grids: extension of a global geometric approach for thin boundary layers and strong flow incidence. Journal of Computational Physics 519, 113441. doi:10.1016/j.jcp.2024.113441

    arXiv 2024

  15. [15]

    doi:10.1006/jcph.1993.1194

    Crumpton,P.,Mackenzie,J.,Morton,K.,1993.CellvertexalgorithmsforthecompressibleNavier–Stokesequations.JournalofComputational Physics 109, 1–15. doi:10.1006/jcph.1993.1194

    arXiv 1993

  16. [16]

    A sharp-interface immersed boundary method for moving objects in compressible viscous flows

    De Vanna, F., Picano, F., Benini, E., 2020. A sharp-interface immersed boundary method for moving objects in compressible viscous flows. Computers & Fluids 201, 104415. doi:10.1016/j.compfluid.2019.104415

    arXiv 2020

  17. [17]

    Numerical simulation of fluid-structure interaction with the volume penalization method

    Engels, T., Kolomenskiy, D., Schneider, K., Sesterhenn, J., 2015. Numerical simulation of fluid-structure interaction with the volume penalization method. Journal of Computational Physics 281, 96–115. doi:10.1016/j.jcp.2014.10.005

    work page doi:10.1016/j.jcp.2014.10.005 2015

  18. [18]

    CombinedImmersed-BoundaryFinite-DifferenceMethodsforThree-Dimensional Complex Flow Simulations

    Fadlun,E.,Verzicco,R.,Orlandi,P.,MohdYusof,J.,2000. CombinedImmersed-BoundaryFinite-DifferenceMethodsforThree-Dimensional Complex Flow Simulations. Journal of Computational Physics 161, 35–60. doi:10.1006/jcph.2000.6484

    arXiv 2000

  19. [19]

    Gallard, F., Vanaret, C., Guenot, D., Gachelin, V., Lafage, R., Pauwels, B., Barjhoux, P.J., Gazaix, A., 8–12 January 2018. GEMS: A Python library for automation of multidisciplinary design optimization process generation, in: Proceedings of the 2018 AIAA/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference, Kissimmee, Florida, USA. doi:...

    work page doi:10.2514/6.2018-0657 2018

  20. [20]

    Gmsh: A 3-D finite element mesh generator with built-in pre- and post-processing facilities

    Geuzaine, C., Remacle, J.F., 2009. Gmsh: A 3-D finite element mesh generator with built-in pre- and post-processing facilities. International Journal for Numerical Methods in Engineering 79, 1309–1331. doi:10.1002/nme.2579

    doi:10.1002/nme.2579 2009

  21. [21]

    Asharpinterfaceimmersedboundarymethodforcompressibleviscousflows

    Ghias,R.,Mittal,R.,Dong,H.,2007. Asharpinterfaceimmersedboundarymethodforcompressibleviscousflows. JournalofComputational Physics 225, 528–553. doi:10.1016/j.jcp.2006.12.007

    work page doi:10.1016/j.jcp.2006.12.007 2007

  22. [22]

    Modeling a no-slip flow boundary with an external force field

    Goldstein, D., Handler, R., Sirovich, L., 1993. Modeling a no-slip flow boundary with an external force field. Journal of Computational Physics 105, 354–366. doi:10.1006/jcph.1993.1081

    arXiv 1993

  23. [23]

    Immersed methods for fluid-structure interaction

    Griffith, B., Patankar, N., 2020. Immersed methods for fluid-structure interaction. Annual Review of Fluid Mechanics 52, 421–448. doi:10.1146/annurev-fluid-010719-060228

    doi:10.1146/annurev-fluid-010719-060228 2020

  24. [24]

    AghostcellmethodforturbulentcompressibleviscousflowsonadaptiveCartesiangrids

    Hu,O.,Zhao,N.,Liu,J.,2013. AghostcellmethodforturbulentcompressibleviscousflowsonadaptiveCartesiangrids. ProcediaEngineering 67, 241–249. doi:10.1016/j.proeng.2013.12.023

    work page doi:10.1016/j.proeng.2013.12.023 2013

  25. [25]

    Immersed boundary technique for turbulent flow simulations

    Iaccarino, G., Verzicco, R., 2003. Immersed boundary technique for turbulent flow simulations. Applied Mechanics Reviews 56, 331–347. doi:10.1115/1.1563627

    work page doi:10.1115/1.1563627 2003

  26. [26]

    A high-resolution procedure for Euler and Navier-–Stokes computations on unstructured grids

    Jawahar, P., Kamath, H., 2000. A high-resolution procedure for Euler and Navier-–Stokes computations on unstructured grids. Journal of Computational Physics 164, 165–203. doi:10.1006/jcph.2000.6596

    arXiv 2000

  27. [27]

    Comparative results from a CFD challenge over a 2D three-element high-lift airfoil

    Klausmeyer, S., Lin, J., 1997. Comparative results from a CFD challenge over a 2D three-element high-lift airfoil. NASA Technical Memorandum NASA-TM-112858. NASA Langley Research Center. Hampton, Virginia, USA

    1997

  28. [28]

    DLR project Digital-X: towards virtual aircraft design and flight testing based on high-fidelity methods

    Kroll,N.,AbuZurayk,M.,Dimitrov,D.,Franz,T.,Führer,T.,Gerhold,T.,Görtz,S.,Heinrich,R.,Ilic,C.,Jepsen,J.,Jägersküpper,J.,Kruse, Page 29 of 31 Mesh adaptation for immersed boundary methods M., Krumbein, A., Langer, S., Liu, D., Liepelt, R., Reimer, L., Ritter, M., Schwöppe, A., Scherer, J., Spiering, F., Thormann, R., Togiti, V., Vollmer, D., Wendisch, J.H.,...

    2016

  29. [29]

    ABrinkmanpenalizationmethodforcompressibleflowsincomplexgeometries

    Liu,Q.,Vasilyev,O.,2007. ABrinkmanpenalizationmethodforcompressibleflowsincomplexgeometries. JournalofComputationalPhysics 227, 946–966. doi:10.1016/j.jcp.2007.07.037

    doi:10.1016/j.jcp.2007.07.037 2007

  30. [30]

    A novel algorithm for fluid dynamics simulation with cell-adaptive Cartesian grids and immersed boundary method

    Luo, C., Bi, L., Meng, S., Gao, H., Yuan, X., Tang, Z., 2025. A novel algorithm for fluid dynamics simulation with cell-adaptive Cartesian grids and immersed boundary method. Physics of Fluids 37, 046122. doi:10.1063/5.0262707

    work page doi:10.1063/5.0262707 2025

  31. [31]

    A Hermite WENO reconstruction-based discontinuous Galerkin method for the Euler equations on tetrahedral grids

    Luo, H., Xia, Y., Li, S., Nourgaliev, R., Cai, C., 2012. A Hermite WENO reconstruction-based discontinuous Galerkin method for the Euler equations on tetrahedral grids. Journal of Computational Physics 231, 5489–5503. doi:10.1016/j.jcp.2012.05.011

    work page doi:10.1016/j.jcp.2012.05.011 2012

  32. [32]

    Numerical methods for fluid–structure interaction models of aortic valves

    Marom, G., 2015. Numerical methods for fluid–structure interaction models of aortic valves. Archives of Computational Methods in Engineering 22, 595–620. doi:10.1007/s11831-014-9133-9

    doi:10.1007/s11831-014-9133-9 2015

  33. [33]

    EnhancedstrategyforadaptiveCartesiangridgenerationwitharbitrarilycomplex 3D geometry

    Meng,S.,Zhou,D.,Yuan,X.,Li,X.,Chen,H.,Bi,L.,2022. EnhancedstrategyforadaptiveCartesiangridgenerationwitharbitrarilycomplex 3D geometry. Advances in Engineering Software 174, 103304. doi:10.1016/j.advengsoft.2022.103304

    arXiv 2022

  34. [34]

    Immersed Boundary Methods

    Mittal, R., Iaccarino, G., 2005. Immersed Boundary Methods. Annual Review of Fluid Mechanics 37, 239–261. doi:10.1146/annurev. fluid.37.061903.175743

    arXiv 2005

  35. [35]

    Murayama, M., Nakakita, K., Yamamoto, K., Ura, H., Ito, Y., Choudhari, M., 16–20 June 2014. Experimental study on slat noise from 30P30N three-element high-lift airfoil at JAXA Hard-Wall Lowspeed Wind Tunnel, in: Proceedings of the 20th AIAA/CEAS Aeroacoustics Conference, Atlanta, Georgia, USA. doi:10.2514/6.2014-2080

    work page doi:10.2514/6.2014-2080 2014

  36. [36]

    Mesh adaptation on hybrid unstructured meshes for immersed boundary methods with applications to industrial aerodynamics

    Núñez-de la Rosa, J., Ferrer, E., Valero, E., 2026a. Mesh adaptation on hybrid unstructured meshes for immersed boundary methods with applications to industrial aerodynamics. Engineering Proceedings 133, 62. doi:10.3390/engproc2026133001

    work page doi:10.3390/engproc2026133001

  37. [37]

    Implementation of immersed boundaries via volume penalization in the industrial aeronautical computational fluid dynamics solver CODA

    Núñez-de la Rosa, J., Huergo, D., Lodares, D., Shrestha, S., Guerra, J., Florenciano, J., Ferrer, E., Valero, E., 2025. Implementation of immersed boundaries via volume penalization in the industrial aeronautical computational fluid dynamics solver CODA. Engineering with Computers doi:10.1007/s00366-025-02119-x

    work page doi:10.1007/s00366-025-02119-x 2025

  38. [38]

    Optimization of the flap position of a high-lift multi-element airfoil using a body-fitted mesh along with immersed boundary methods

    Núñez-de la Rosa, J., Mateo, A., Ferrer, E., Valero, E., 2026b. Optimization of the flap position of a high-lift multi-element airfoil using a body-fitted mesh along with immersed boundary methods. Engineering Proceedings 133, 61. doi:10.3390/engproc2026133002

    work page doi:10.3390/engproc2026133002

  39. [39]

    XTROEM-FV: a new code for computational astrophysics based on very high order finite-volume methods–i.Magnetohydrodynamics

    Núñez-de la Rosa, J., Munz, C.D., 2016a. XTROEM-FV: a new code for computational astrophysics based on very high order finite-volume methods–i.Magnetohydrodynamics. MonthlyNoticesoftheRoyalAstronomicalSociety455,3458–3479. doi:10.1093/mnras/stv2531

    work page doi:10.1093/mnras/stv2531

  40. [40]

    XTROEM-FV: a new code for computational astrophysics based on very high order finite- volume methods – ii

    Núñez-de la Rosa, J., Munz, C.D., 2016b. XTROEM-FV: a new code for computational astrophysics based on very high order finite- volume methods – ii. Relativistic hydro- and magnetohydrodynamics. Monthly Notices of the Royal Astronomical Society 460, 535–559. doi:10.1093/mnras/stw999

    work page doi:10.1093/mnras/stw999

  41. [41]

    Hybrid DG/FV schemes for magnetohydrodynamics and relativistic hydrodynamics

    Núñez-de la Rosa, J., Munz, C.D., 2018. Hybrid DG/FV schemes for magnetohydrodynamics and relativistic hydrodynamics. Computer Physics Communications 222, 113–135. doi:10.1016/j.cpc.2017.09.026

    work page doi:10.1016/j.cpc.2017.09.026 2018

  42. [42]

    Cut-elementbasedimmersedboundarymethodformovinggeometriesincompressible liquid flows with cavitation

    Örley,F.,Pasquariello,V.,Hickel,S.,Adams,N.,2015. Cut-elementbasedimmersedboundarymethodformovinggeometriesincompressible liquid flows with cavitation. Journal of Computational Physics 283, 1–22. doi:10.1016/j.jcp.2014.11.028

    work page doi:10.1016/j.jcp.2014.11.028 2015

  43. [43]

    Cartesianmeshadaptation:immersedboundarymethodbasedonhigh-order discontinuous Galerkin method

    Ouyang,W.,Huang,J.,Wang,T.,An,W.,Liu,X.,Lyu,H.,2024. Cartesianmeshadaptation:immersedboundarymethodbasedonhigh-order discontinuous Galerkin method. Physics of Fluids 36, 096139. doi:10.1063/5.0223305

    work page doi:10.1063/5.0223305 2024

  44. [44]

    Engineering with Computers 37, 2419–2437

    Péron,S.,Benoit,C.,Renaud,T.,Mary,I.,2021.AnimmersedboundarymethodonCartesianadaptivegridsforthesimulationofcompressible flows around arbitrary geometries. Engineering with Computers 37, 2419–2437. doi:10.1007/s00366-020-00950-y

    work page doi:10.1007/s00366-020-00950-y 2021

  45. [45]

    Flow patterns around heart valves: a numerical method

    Peskin, C., 1972. Flow patterns around heart valves: a numerical method. Journal of Computational Physics 10, 252–271. doi:10.1016/ 0021-9991(72)90065-4

    1972

  46. [46]

    High-order discontinuous Galerkin method with immersed boundary treatment for compressible flows on parallel adaptive Cartesian grids

    Qi, X., Wang, Z., Zhu, J., Tian, L., Zhao, N., 2024. High-order discontinuous Galerkin method with immersed boundary treatment for compressible flows on parallel adaptive Cartesian grids. Physics of Fluids 36, 116139. doi:10.1063/5.0238605

    doi:10.1063/5.0238605 2024

  47. [47]

    A boundary condition-enforced immersed boundary method for compressible viscous flows

    Qiu, Y., Shu, C., Wu, J., Sun, Y., Yang, L., Guo, T., 2016. A boundary condition-enforced immersed boundary method for compressible viscous flows. Computers & Fluids 136, 104–113. doi:10.1016/j.compfluid.2016.06.004

    work page doi:10.1016/j.compfluid.2016.06.004 2016

  48. [48]

    Progress in high-lift aerodynamic calculations

    Rogers, S., 1994. Progress in high-lift aerodynamic calculations. Journal of Aircraft 31, 1244–1251. doi:10.2514/3.46642

    work page doi:10.2514/3.46642 1994

  49. [49]

    A comparison of turbulence models in computing multi-element airfoil flows, in: Proceedings of the 32nd AIAA Aerospace Sciences Meeting and Exhibit, Reno, Nevada, USA

    Rogers, S., Menter, F., Durbin, P., Mansour, N., 10–13 January 1994. A comparison of turbulence models in computing multi-element airfoil flows, in: Proceedings of the 32nd AIAA Aerospace Sciences Meeting and Exhibit, Reno, Nevada, USA. doi:10.2514/6.1994-291

    work page doi:10.2514/6.1994-291 1994

  50. [50]

    Predictionofhigh-liftflowsusingturbulentclosuremodels

    Rumsey,C.,Gatski,T.,Ying,S.,Bertelrud,A.,1998. Predictionofhigh-liftflowsusingturbulentclosuremodels. AIAAJournal36,765–774. doi:10.2514/2.435

    work page doi:10.2514/2.435 1998

  51. [51]

    Immersed boundary methods for numerical simulation of confined fluid and plasma turbulence in complex geometries: a review

    Schneider, K., 2015. Immersed boundary methods for numerical simulation of confined fluid and plasma turbulence in complex geometries: a review. Journal of Plasma Physics 81, 435810601. doi:10.1017/S0022377815000598

    work page doi:10.1017/s0022377815000598 2015

  52. [52]

    Analysis of the Common Research Model using structured and unstructured meshes

    Sclafani, A., Vassberg, J., Winkler, C., Dorgan, A., Mani, M., Olsen, M., Coder, J., 2014. Analysis of the Common Research Model using structured and unstructured meshes. Journal of Aircraft 51, 1223–1243. doi:10.2514/1.C032411

    work page doi:10.2514/1.c032411 2014

  53. [53]

    Development of an immersed boundary-phase field-lattice Boltzmann method for Neumann boundary condition to study contact line dynamics

    Shao, J.Y., Shu, C., Chew, Y.T., 2013. Development of an immersed boundary-phase field-lattice Boltzmann method for Neumann boundary condition to study contact line dynamics. Journal of Computational Physics 234, 8–32. doi:10.1016/j.jcp.2012.08.040

    work page doi:10.1016/j.jcp.2012.08.040 2013

  54. [54]

    Immersed boundary methods for simulating fluid-structure interaction

    Sotiropoulos, F., Yang, X., 2014. Immersed boundary methods for simulating fluid-structure interaction. Progress in Aerospace Sciences 65, 1–21. doi:10.1016/j.paerosci.2013.09.003

    work page doi:10.1016/j.paerosci.2013.09.003 2014

  55. [55]

    HighReynoldsnumber,multielementairfoilflowfieldmeasurements

    Spaid,F.,2000. HighReynoldsnumber,multielementairfoilflowfieldmeasurements. JournalofAircraft37,499–507. doi:10.2514/2.2626

    work page doi:10.2514/2.2626 2000

  56. [56]

    NASA Langley Research Center

    Swanson,R.,Langer,S.,2016.ComparisonofNACA0012laminarflowsolutions:structuredandunstructuredgridmethods.NASATechnical Memorandum NASA-TM-2016-219003. NASA Langley Research Center. Hampton, Virginia, USA

    2016

  57. [57]

    Near-wall modification of Spalart–Allmaras turbulence model for immersed boundary method

    Tamaki, Y., Harada, M., Imamura, T., 2017. Near-wall modification of Spalart–Allmaras turbulence model for immersed boundary method. AIAA Journal 55, 3027–3039. doi:10.2514/1.J055824

    work page doi:10.2514/1.j055824 2017

  58. [58]

    A ghost-cell immersed boundary method for flow in complex geometry

    Tseng, Y.H., Ferziger, J., 2003. A ghost-cell immersed boundary method for flow in complex geometry. Journal of Computational Physics Page 30 of 31 Mesh adaptation for immersed boundary methods 192, 593–623. doi:10.1016/j.jcp.2003.07.024

    work page doi:10.1016/j.jcp.2003.07.024 2003

  59. [59]

    ASharpInterfaceCartesianGridMethodforSimulatingFlowswithComplex Moving Boundaries

    Udaykumar,H.,Mittal,R.,Rampunggoon,P.,Khanna,A.,2001. ASharpInterfaceCartesianGridMethodforSimulatingFlowswithComplex Moving Boundaries. Journal of Computational Physics 174, 345–380. doi:10.1006/jcph.2001.6916

    arXiv 2001

  60. [60]

    High-lift testing at high Reynolds numbers, in: Proceedings of the 17th AIAA Aerospace Ground Testing Conference, Nashville, Tennessee, USA

    Valarezo, W., 6–8 July 1992. High-lift testing at high Reynolds numbers, in: Proceedings of the 17th AIAA Aerospace Ground Testing Conference, Nashville, Tennessee, USA. doi:10.2514/6.1992-3986

    work page doi:10.2514/6.1992-3986 1992

  61. [61]

    Multi-element airfoil optimization for maximum lift at high Reynolds numbers, in: Proceedings of the 9th AIAA Applied Aerodynamics Conference, Baltimore, Maryland, USA

    Valarezo, W., Dominik, C., McGhee, R., Goodman, W., Paschal, K., 23–25 September 1991. Multi-element airfoil optimization for maximum lift at high Reynolds numbers, in: Proceedings of the 9th AIAA Applied Aerodynamics Conference, Baltimore, Maryland, USA. doi:10.2514/6.1991-3332

    work page doi:10.2514/6.1991-3332 1991

  62. [62]

    Immersedboundarymethods:historicalperspectiveandfutureoutlook

    Verzicco,R.,2023. Immersedboundarymethods:historicalperspectiveandfutureoutlook. AnnualReviewofFluidMechanics55,129–155. doi:10.1146/annurev-fluid-120720-022129

    doi:10.1146/annurev-fluid-120720-022129 2023

  63. [63]

    AircraftsimulationsusingthenewCFDsoftwarefrom ONERA, DLR, and Airbus

    Volpiani,P.,Chapelier,J.B.,Schwöppe,A.,Jägersküpper,J.,Champagneux,S.,2024. AircraftsimulationsusingthenewCFDsoftwarefrom ONERA, DLR, and Airbus. Journal of Aircraft 61, 857–869. doi:10.2514/1.C037506

    work page doi:10.2514/1.c037506 2024

  64. [64]

    An immersed boundary-lattice Boltzmann flux solver and its applications to fluid-structure interaction problems

    Wang, Y., Shu, C., Teo, C.J., Wu, J., 2015. An immersed boundary-lattice Boltzmann flux solver and its applications to fluid-structure interaction problems. Journal of Fluids and Structures 54, 440–465. doi:10.1016/j.jfluidstructs.2014.12.003

    work page doi:10.1016/j.jfluidstructs.2014.12.003 2015

  65. [65]

    Adaptive mesh refinement for simulating fluid-structure interaction using a sharp interface immersed boundary method

    Wang, Z., Sun, X., 2020. Adaptive mesh refinement for simulating fluid-structure interaction using a sharp interface immersed boundary method. International Journal of Numerical Methods in Engineering 92, 1890–1913. doi:10.1002/fld.4853

    work page doi:10.1002/fld.4853 2020

  66. [66]

    Self-adaptive turbulence eddy simulation with a high-order finite differencing method for high Reynolds number complex flows

    Wu, W., Min, Y., Han, X., Ma, Y., Yan, Z., Deng, X., 2023. Self-adaptive turbulence eddy simulation with a high-order finite differencing method for high Reynolds number complex flows. Aerospace Science and Technology 141, 108562. doi:10.1016/j.ast.2023.108562

    arXiv 2023

  67. [67]

    Animmersed boundarymethod withimplicitbody forcefor compressibleviscous flow

    Yu,H., Pantano,C., 2022. Animmersed boundarymethod withimplicitbody forcefor compressibleviscous flow. Journalof Computational Physics 459, 111125. doi:10.1016/j.jcp.2022.111125. Page 31 of 31

    arXiv 2022