The reviewed record of science sign in
Pith

arxiv: 2605.26388 · v3 · pith:KF7TLHNM · submitted 2026-05-25 · physics.comp-ph · math-ph· math.MP

MARUT: An Exascale-Ready, GPU-Accelerated High-Order CFD Framework with AMR for High-Speed Flows and Finite-Rate Chemistry

Trishit Mondal , Ameya D. Jagtap This is my paper

Reviewed by Pith T0 review T1 audit T2 compute T3 formal T4 kernel 2026-06-29 18:57 UTCgrok-4.3pith:KF7TLHNMrecord.jsonopen to challenge →

classification physics.comp-ph math-phmath.MP
keywords computational fluid dynamicsGPU accelerationdiscontinuous Galerkinadaptive mesh refinementfinite-rate chemistryhigh-speed flowsparallel scalingcompressible flows
0
0 comments X

The pith

MARUT achieves near-linear strong scaling on multiple GPUs while matching reference solutions for inviscid, viscous, and reactive compressible flows.

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

The paper introduces MARUT, a multi-GPU computational fluid dynamics framework built to simulate compressible flows from subsonic to hypersonic speeds, including those with finite-rate chemistry. It combines high-order spectral discontinuous Galerkin methods, strong-stability-preserving Runge-Kutta time integration, and dynamic adaptive mesh refinement inside an MPI-based infrastructure that runs natively on NVIDIA GPUs. The central effort is showing that this setup delivers accurate results at scale while concentrating work only where the physics is complex. A sympathetic reader would care because current supercomputers require tools that stay efficient and stable when modeling shocks, vortices, and reacting interfaces without excessive manual tuning. Validation across canonical benchmarks confirms close agreement with established references and near-linear parallel efficiency.

Core claim

MARUT is a scalable multi-GPU CFD framework that employs high-order spectral discontinuous Galerkin discretizations with strong-stability-preserving Runge-Kutta time integration and dynamic AMR to perform high-fidelity simulations of compressible flows spanning subsonic to hypersonic regimes, including multi-species nonequilibrium chemistry. The framework uses distributed-memory MPI parallelism and native GPU implementation to achieve near-linear strong scaling. Numerical predictions on a broad suite of inviscid, viscous, and reactive benchmark problems show close agreement with established reference solutions.

What carries the argument

High-order spectral discontinuous Galerkin discretizations combined with dynamic adaptive mesh refinement on a distributed-memory MPI and native NVIDIA GPU infrastructure.

If this is right

  • Concentrates computational resources only in regions of localized physical complexity through AMR, thereby reducing overall cost while preserving fidelity.
  • Supports low-dissipation, high-resolution capture of shocks, vortical structures, and reactive interfaces across speed regimes.
  • Handles multi-species nonequilibrium chemistry in high-speed flows as part of the same scalable infrastructure.
  • Maintains strong parallel efficiency through GPU-resident computations and scalable MPI communication.
  • Reflects a shift toward modular, adaptive simulation tools compatible with emerging heterogeneous architectures.

Where Pith is reading between the lines

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

  • The modular design may support coupling with reduced-order or data-driven models to further accelerate specific flow regimes.
  • The same AMR and high-order approach could be tested on other multiscale problems involving sharp interfaces, such as multiphase or plasma flows.
  • If scaling persists to full exascale node counts, routine three-dimensional simulations of complex reacting flows would become feasible on standard allocations.

Load-bearing premise

The high-order spectral discontinuous Galerkin discretizations combined with AMR will maintain accuracy and stability for the full range of subsonic to hypersonic regimes with finite-rate chemistry without additional tuning or case-specific adjustments.

What would settle it

A hypersonic reactive flow test case where MARUT produces results that deviate substantially from established reference solutions or shows numerical instability without case-specific changes.

Figures

Figures reproduced from arXiv: 2605.26388 by Ameya D. Jagtap, Trishit Mondal.

Figure 1
Figure 1. Figure 1: High-level architecture of the MARUT solver. The spatial-discretisation components configure the GPU￾resident RHS kernels, which are advanced in time by the SSP-RK integrator under the control of runtime callbacks. MARUT provides a robust AMR framework based on hierarchical tree-structured meshes, enabling dynamically adaptive high-order DG simulations with excellent conservation and parallel scalability p… view at source ↗
Figure 2
Figure 2. Figure 2: Mach 3 supersonic Euler flow past a cylinder in a channel: density field (left column) and AMR mesh (right [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Subsonic TGV (Case 3, Ms = 0.1, Re = 1600) at t = 10, ≈ 120,000 active elements: AMR mesh (left), density slice (middle), and iso-surface of Q = 0.01 coloured by velocity magnitude (right) [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Subsonic TGV: kinetic energy Ek(t) (left) and dissipation rate ε(t) = −dEk/dt (right) for the three grids of Sec. 4.2 (Case 1: 123 , Case 2: 243 , Case 3: 83 with AMR up to ℓmax = 3), against pseudo-spectral DNS reference data from van Rees et al. [26]. 14 [PITH_FULL_IMAGE:figures/full_fig_p014_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Supersonic TGV (Ms = 1.25, Re = 1600) at t = 11, approximately 140,000 active elements: same panel layout as [PITH_FULL_IMAGE:figures/full_fig_p015_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Supersonic TGV: Ek(t) (left), solenoidal dissipation εs(t) (middle), and dilatational dissipation εd(t) (right) for the same three grids as in [PITH_FULL_IMAGE:figures/full_fig_p016_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: (Left) Three-dimensional computational domain bounding the ONERA M6 wing, with extents expressed [PITH_FULL_IMAGE:figures/full_fig_p017_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Surface pressure coefficient Cp distributions on the ONERA M6 wing at M∞ = 0.84 for the seven spanwise stations (z/b) ∈ {0.20, 0.44, 0.65, 0.80, 0.90, 0.95, 0.99}. Solid lines: MARUT solution at t = 9.0 convective time units on the static 217,088-element P = 3 mesh; black markers: Schmitt and Charpin wind-tunnel data [29] (downward triangles: lower/pressure surface, upward triangles: upper/suction surface)… view at source ↗
Figure 9
Figure 9. Figure 9: Iso-surfaces of the Q-criterion coloured by velocity magnitude for the transonic ONERA M6 wing at M∞ = 0.84, showing the tip-vortex and the trailing-edge shear layer. 18 [PITH_FULL_IMAGE:figures/full_fig_p018_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Non-equilibrium reactive blast wave at tend = 2 × 10−4 s on a 64 × 64 grid with P = 7 spectral DG and SSPRK54. Top row, left to right: atomic nitrogen ρN, molecular nitrogen ρN2 , nitric oxide ρNO. Bottom row: atomic oxygen ρO, molecular oxygen ρO2 , and translational–rotational temperature T (in K). Color scales are individual to each panel. The cylindrical shock has propagated from the initial interface… view at source ↗
Figure 11
Figure 11. Figure 11: 2D Euler cylinder benchmark: wall-time per SSPRK54 step vs. number of grid points, comparing CPU [PITH_FULL_IMAGE:figures/full_fig_p020_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: 3D Navier–Stokes TGV benchmark: wall-time per SSPRK54 step vs. number of grid points, for a 64-thread [PITH_FULL_IMAGE:figures/full_fig_p021_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: MARUT multi-GPU representation. The mesh is partitioned across ranks, and each GPU advances its own partition through an identical pipeline; ghost-cell exchanges overlap with the interior compute on each rank. 5.2 Strong and Weak Scaling The parallel efficiency of MARUT is assessed through both strong and weak scaling experiments on up to four NVIDIA L40S GPUs interconnected via InfiniBand, using CPU-stag… view at source ↗
Figure 14
Figure 14. Figure 14: Strong scaling on up to four L40S GPUs. (Left) Wall time per SSPRK54 step versus GPU count for three [PITH_FULL_IMAGE:figures/full_fig_p023_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Weak scaling on up to four L40S GPUs. (Left) Weak scaling efficiency versus GPU count for three [PITH_FULL_IMAGE:figures/full_fig_p024_15.png] view at source ↗
read the original abstract

We present MARUT, a scalable multi-GPU computational fluid dynamics (CFD) framework designed for high-fidelity simulations of compressible flows spanning subsonic to hypersonic regimes, including chemically reacting nonequilibrium flows with finite-rate chemistry and adaptive mesh refinement (AMR). The framework addresses a central challenge in contemporary scientific computing: the development of numerically accurate and computationally scalable algorithms capable of resolving strongly nonlinear, multiscale flow physics on emerging heterogeneous supercomputing architectures. Built around a distributed-memory MPI-parallel infrastructure and implemented natively on NVIDIA GPUs, MARUT combines high-order spectral discontinuous Galerkin discretisations with strong-stability-preserving Runge--Kutta time integration to achieve low-dissipation and high-resolution representation of shocks, vortical structures and reactive interfaces. Dynamic AMR further enables efficient concentration of computational resources in localized regions of physical complexity, thereby substantially reducing computational cost while preserving solution fidelity. MARUT is designed to maintain strong parallel efficiency through GPU-resident computations and scalable MPI communication strategies, achieving near-linear strong scaling across multiple GPUs. The solver is validated against a broad suite of canonical benchmark problems involving inviscid, viscous, and reactive compressible flows, including subsonic, transonic, supersonic, and hypersonic configurations with multi-species nonequilibrium chemistry. The numerical predictions show close agreement with established reference solutions. Beyond its immediate performance characteristics, the framework reflects the broader transition of computational science towards modular, adaptive and AI-compatible simulation ecosystems.

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 / 0 minor

Summary. The manuscript presents MARUT, a multi-GPU CFD framework using high-order spectral discontinuous Galerkin discretizations, strong-stability-preserving Runge-Kutta time integration, and dynamic AMR for compressible flows spanning subsonic to hypersonic regimes with finite-rate chemistry. It claims near-linear strong scaling across multiple GPUs via GPU-resident computations and scalable MPI, plus close agreement with reference solutions across inviscid, viscous, and reactive benchmark problems.

Significance. If the scaling and validation claims hold, the work would be significant for exascale CFD by demonstrating a modular, GPU-native implementation of high-order DG+AMR methods that efficiently handles multiscale reactive flows on heterogeneous architectures, supporting the shift toward adaptive, high-fidelity simulation ecosystems.

major comments (2)
  1. [Abstract] Abstract: the central claim of 'close agreement with established reference solutions' for the full suite of subsonic-to-hypersonic cases with finite-rate chemistry is load-bearing for the validation narrative, yet the provided text supplies no quantitative metrics (e.g., L2 errors, convergence rates, or specific benchmark tables) to support it.
  2. [Abstract] Abstract: the assertion that high-order DG+AMR 'maintain[s] accuracy and stability' across all regimes without case-specific adjustments is presented as a strength but rests on an untested assumption in the absence of any stability analysis, limiter details, or chemistry-specific discretization choices.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their detailed review and constructive comments on the MARUT manuscript. We address each major comment below with specific responses and indicate planned revisions where appropriate.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim of 'close agreement with established reference solutions' for the full suite of subsonic-to-hypersonic cases with finite-rate chemistry is load-bearing for the validation narrative, yet the provided text supplies no quantitative metrics (e.g., L2 errors, convergence rates, or specific benchmark tables) to support it.

    Authors: We agree that the abstract would be strengthened by including quantitative support for the validation claims. The full manuscript contains detailed L2 error norms, convergence rates, and benchmark tables in the results sections for all cases. We will revise the abstract to add a concise summary of representative quantitative metrics (e.g., maximum L2 errors for key test problems) while retaining the overall length constraints. revision: yes

  2. Referee: [Abstract] Abstract: the assertion that high-order DG+AMR 'maintain[s] accuracy and stability' across all regimes without case-specific adjustments is presented as a strength but rests on an untested assumption in the absence of any stability analysis, limiter details, or chemistry-specific discretization choices.

    Authors: The manuscript demonstrates accuracy and stability through comprehensive benchmark results spanning the regimes without requiring case-specific adjustments, as shown in the validation sections. However, we acknowledge the absence of a dedicated von Neumann or nonlinear stability analysis. We will revise the abstract to qualify the statement as being supported by the presented numerical evidence rather than as a general untested claim, and ensure the methods section explicitly references limiter and discretization details for chemistry. revision: partial

Circularity Check

0 steps flagged

No significant circularity; implementation and validation claims are externally benchmarked

full rationale

The paper presents a software framework (MARUT) for CFD simulations, focusing on implementation details (DG discretization, AMR, GPU/MPI scaling) and empirical validation against independent reference solutions across flow regimes. No derivation chain exists that reduces predictions or uniqueness claims to fitted inputs, self-citations, or ansatzes by construction. Validation statements reference external benchmarks rather than internal fits, and scaling results are performance measurements, not mathematical derivations. This matches the default expectation of non-circularity for implementation papers.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no specific parameters or axioms detailed.

pith-pipeline@v0.9.1-grok · 5806 in / 1000 out tokens · 24990 ms · 2026-06-29T18:57:11.466995+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

38 extracted references · 2 canonical work pages

  1. [1]

    Direct numerical simulation: a tool in turbulence research.Annual review of fluid mechanics, 30(1):539–578, 1998

    Parviz Moin and Krishnan Mahesh. Direct numerical simulation: a tool in turbulence research.Annual review of fluid mechanics, 30(1):539–578, 1998

    1998

  2. [2]

    Progress in shock wave/boundary layer interactions.Progress in Aerospace Sciences, 72:80–99, 2015

    Datta V Gaitonde. Progress in shock wave/boundary layer interactions.Progress in Aerospace Sciences, 72:80–99, 2015

    2015

  3. [3]

    Cfd vision 2030 study: a path to revolutionary computational aerosciences

    Jeffrey P Slotnick, Abdollah Khodadoust, Juan Alonso, David Darmofal, William Gropp, Elizabeth Lurie, and Dimitri J Mavriplis. Cfd vision 2030 study: a path to revolutionary computational aerosciences. Technical report, 2014

    2030

  4. [4]

    On the utility of gpu accelerated high-order methods for unsteady flow simulations: A comparison with industry-standard tools.Journal of Computational Physics, 334:497–521, 2017

    Brian C Vermeire, Freddie D Witherden, and Peter E Vincent. On the utility of gpu accelerated high-order methods for unsteady flow simulations: A comparison with industry-standard tools.Journal of Computational Physics, 334:497–521, 2017

    2017

  5. [5]

    Tvb runge-kutta local projection discontinuous galerkin finite element method for conservation laws

    Bernardo Cockburn and Chi-Wang Shu. Tvb runge-kutta local projection discontinuous galerkin finite element method for conservation laws. ii. general framework.Mathematics of computation, 52(186):411–435, 1989

    1989

  6. [6]

    Hesthaven and Tim Warburton.Nodal Discontinuous Galerkin Methods: Algorithms, Analysis, and Applications, volume 54 ofTexts in Applied Mathematics

    Jan S. Hesthaven and Tim Warburton.Nodal Discontinuous Galerkin Methods: Algorithms, Analysis, and Applications, volume 54 ofTexts in Applied Mathematics. Springer, New York, 2008

    2008

  7. [7]

    Nodal discontinuous galerkin methods on graphics processors.Journal of Computational Physics, 228(21):7863–7882, 2009

    Andreas Kl¨ ockner, Tim Warburton, Jeff Bridge, and Jan S Hesthaven. Nodal discontinuous galerkin methods on graphics processors.Journal of Computational Physics, 228(21):7863–7882, 2009

    2009

  8. [8]

    Local adaptive mesh refinement for shock hydrodynamics.Journal of computational Physics, 82(1):64–84, 1989

    Marsha J Berger and Phillip Colella. Local adaptive mesh refinement for shock hydrodynamics.Journal of computational Physics, 82(1):64–84, 1989

    1989

  9. [9]

    Trixi.jl: Adaptive high-order numerical simulations of hyperbolic PDEs in Julia

    Michael Schlottke-Lakemper, Gregor J Gassner, Hendrik Ranocha, Andrew R Winters, Jesse Chan, and Andr´ es Rueda-Ram´ ırez. Trixi.jl: Adaptive high-order numerical simulations of hyperbolic PDEs in Julia. https: //github.com/trixi-framework/Trixi.jl, 2025

    2025

  10. [10]

    Aiaa, 1989

    John David Anderson.Hypersonic and high temperature gas dynamics. Aiaa, 1989. 33

    1989

  11. [11]

    Nonequilibrium hypersonic aerothermodynamics.Nonequilibrium hypersonic aerothermodynamics by Park, page 29860, 1989

    Chul Park. Nonequilibrium hypersonic aerothermodynamics.Nonequilibrium hypersonic aerothermodynamics by Park, page 29860, 1989

    1989

  12. [12]

    Computation of weakly ionized hypersonic flows in thermo- chemical nonequilibrium.Journal of Thermophysics and heat transfer, 5(3):266–273, 1991

    Graham V Candler and Robert W MacCormack. Computation of weakly ionized hypersonic flows in thermo- chemical nonequilibrium.Journal of Thermophysics and heat transfer, 5(3):266–273, 1991

    1991

  13. [13]

    Wilcox, and Omar Ghattas

    Carsten Burstedde, Lucas C. Wilcox, and Omar Ghattas. p4est: Scalable algorithms for parallel adaptive mesh refinement on forests of octrees.SIAM Journal on Scientific Computing, 33(3):1103–1133, 2011

    2011

  14. [14]

    An adaptive finite element scheme for transient problems in CFD.Computer Methods in Applied Mechanics and Engineering, 61(3):323–338, 1987

    Rainald L¨ ohner. An adaptive finite element scheme for transient problems in CFD.Computer Methods in Applied Mechanics and Engineering, 61(3):323–338, 1987

    1987

  15. [15]

    Rueda-Ram´ ırez, Florian J

    Sebastian Hennemann, Andr´ es M. Rueda-Ram´ ırez, Florian J. Hindenlang, and Gregor J. Gassner. A provably entropy stable subcell shock capturing approach for high order split form DG for the compressible Euler equations. Journal of Computational Physics, 426:109935, 2021

    2021

  16. [16]

    Kopriva.Implementing Spectral Methods for Partial Differential Equations: Algorithms for Scientists and Engineers

    David A. Kopriva.Implementing Spectral Methods for Partial Differential Equations: Algorithms for Scientists and Engineers. Scientific Computation. Springer, Dordrecht, 2009

    2009

  17. [17]

    Fisher and Mark H

    Travis C. Fisher and Mark H. Carpenter. High-order entropy stable finite difference schemes for nonlinear conservation laws: Finite domains.Journal of Computational Physics, 252:518–557, 2013

    2013

  18. [18]

    Gassner, Andrew R

    Gregor J. Gassner, Andrew R. Winters, and David A. Kopriva. Split form nodal discontinuous Galerkin schemes with summation-by-parts property for the compressible Euler equations.Journal of Computational Physics, 327:39–66, 2016

    2016

  19. [19]

    Comparison of some entropy conservative numerical fluxes for the Euler equations.Journal of Scientific Computing, 76(1):216–242, 2018

    Hendrik Ranocha. Comparison of some entropy conservative numerical fluxes for the Euler equations.Journal of Scientific Computing, 76(1):216–242, 2018

    2018

  20. [20]

    Michael G. Edwards. The dominant wave-capturing flux: A finite-volume scheme without decomposition for systems of hyperbolic conservation laws.Journal of Computational Physics, 218(1):275–294, 2006

    2006

  21. [21]

    Lax, and Bram van Leer

    Amiram Harten, Peter D. Lax, and Bram van Leer. On upstream differencing and Godunov-type schemes for hyperbolic conservation laws.SIAM Review, 25(1):35–62, 1983

    1983

  22. [22]

    David A. Kopriva. A conservative staggered-grid Chebyshev multidomain method for compressible flows. II. A semi-structured method.Journal of Computational Physics, 128(2):475–488, 1996

    1996

  23. [23]

    Bassi and S

    F. Bassi and S. Rebay. A high-order accurate discontinuous finite element method for the numerical solution of the compressible Navier–Stokes equations.Journal of Computational Physics, 131(2):267–279, 1997

    1997

  24. [24]

    The viscosity of gases and molecular force.Philosophical Magazine, Series 5, 36(223):507–531, 1893

    William Sutherland. The viscosity of gases and molecular force.Philosophical Magazine, Series 5, 36(223):507–531, 1893

  25. [25]

    G. M. Morton. A computer oriented geodetic data base and a new technique in file sequencing. Technical report, IBM Ltd., Ottawa, Ontario, Canada, 1966

    1966

  26. [26]

    van Rees, Anthony Leonard, D

    Wim M. van Rees, Anthony Leonard, D. I. Pullin, and Petros Koumoutsakos. A comparison of vortex and pseudo-spectral methods for the simulation of periodic vortical flows at high Reynolds numbers.Journal of Computational Physics, 230(8):2794–2805, 2011

    2011

  27. [27]

    H3PC: Hypersonic, high-order, high-performance code with adaptive mesh refinement and real chemistry.arXiv preprint arXiv:2511.17551, 2025

    Ahmad Peyvan, Khemraj Shukla, and George Em Karniadakis. H3PC: Hypersonic, high-order, high-performance code with adaptive mesh refinement and real chemistry.arXiv preprint arXiv:2511.17551, 2025

    work page arXiv 2025

  28. [28]

    Lusher, William Van Noordt, Christoph Wenzel, Tobias Gibis, Pascal Mossier, Andrea Beck, Guido Lodato, Christoph Brehm, Matteo Ruggeri, Carlo Scalo, and Neil Sandham

    Jean-Baptiste Chapelier, David J. Lusher, William Van Noordt, Christoph Wenzel, Tobias Gibis, Pascal Mossier, Andrea Beck, Guido Lodato, Christoph Brehm, Matteo Ruggeri, Carlo Scalo, and Neil Sandham. Comparison of high-order numerical methodologies for the simulation of the supersonic Taylor–Green vortex flow.Physics of Fluids, 36(5):055146, 2024

    2024

  29. [29]

    Pressure distributions on the onera m6-wing at transonic mach numbers, experimental data base for computer program assessment.AGARD AR-138, 1979

    V Schmitt. Pressure distributions on the onera m6-wing at transonic mach numbers, experimental data base for computer program assessment.AGARD AR-138, 1979

    1979

  30. [30]

    Flux-split algorithms for flows with non-equilibrium chemistry and vibrational relaxation.Journal of Computational Physics, 88(1):131–168, 1990

    B Grossman and P Cinnella. Flux-split algorithms for flows with non-equilibrium chemistry and vibrational relaxation.Journal of Computational Physics, 88(1):131–168, 1990. 34

    1990

  31. [31]

    Evaluating modern gpu interconnect: Pcie, nvlink, nv-sli, nvswitch and gpudirect.IEEE Transactions on Parallel and Distributed Systems, 31(1):94–110, 2019

    Ang Li, Shuaiwen Leon Song, Jieyang Chen, Jiajia Li, Xu Liu, Nathan R Tallent, and Kevin J Barker. Evaluating modern gpu interconnect: Pcie, nvlink, nv-sli, nvswitch and gpudirect.IEEE Transactions on Parallel and Distributed Systems, 31(1):94–110, 2019

    2019

  32. [32]

    Adaptive mesh refinement for hyperbolic partial differential equations

    Marsha J Berger and Joseph Oliger. Adaptive mesh refinement for hyperbolic partial differential equations. Journal of computational Physics, 53(3):484–512, 1984

    1984

  33. [33]

    Amrex: Block-structured adaptive mesh refinement for multiphysics applications.The International Journal of High Performance Computing Applications, 35(6):508–526, 2021

    Weiqun Zhang, Andrew Myers, Kevin Gott, Ann Almgren, and John Bell. Amrex: Block-structured adaptive mesh refinement for multiphysics applications.The International Journal of High Performance Computing Applications, 35(6):508–526, 2021

    2021

  34. [34]

    Shocks under control: Taming transonic compressible flow over an rae2822 airfoil with deep reinforcement learning.arXiv preprint arXiv:2511.07564, 2025

    Trishit Mondal, Ricardo Vinuesa, and Ameya D Jagtap. Shocks under control: Taming transonic compressible flow over an rae2822 airfoil with deep reinforcement learning.arXiv preprint arXiv:2511.07564, 2025

    work page arXiv 2025

  35. [35]

    Trishit Mondal and Ameya D. Jagtap. Hypersonic flow control: Generalized deep reinforcement learning for hypersonic intake unstart control under uncertainty.Physics of Fluids, 38(4):046112, 2026

    2026

  36. [36]

    A review of reaction rates and thermodynamic and transport properties for an 11-species air model for chemical and thermal nonequilibrium calculations to 30000 k

    Roop N Gupta, Jerrold M Yos, Richard A Thompson, and Kam-Pui Lee. A review of reaction rates and thermodynamic and transport properties for an 11-species air model for chemical and thermal nonequilibrium calculations to 30000 k. 1990

    1990

  37. [37]

    Spiteri and Steven J

    Raymond J. Spiteri and Steven J. Ruuth. A new class of optimal high-order strong-stability-preserving time discretization methods.SIAM Journal on Numerical Analysis, 40(2):469–491, 2002

    2002

  38. [38]

    Efficient implementation of essentially non-oscillatory shock-capturing schemes

    Chi-Wang Shu and Stanley Osher. Efficient implementation of essentially non-oscillatory shock-capturing schemes. Journal of Computational Physics, 77(2):439–471, 1988. 35

    1988