HydroGym: A Multi-Solver Reinforcement Learning Platform for Fluid Dynamics

Zenodo (CERN European Organization for Nuclear Research)

Abstract

HydroGym is a multi-solver reinforcement learning platform for data-driven modeling and control of fluid dynamics. This release (v1.0.1) extends HydroGym from a single-solver backend to a native multi-solver reinforcement learning framework for flow control, with environments spanning six CFD solver backends: Firedrake (finite element), MAIA LBM (lattice Boltzmann), MAIA structured FV (finite volume), NEK5000 (spectral element), JAX, and JAX-Fluids (differentiable). It provides 61 pre-configured, Gymnasium-compatible environments ranging from canonical 2D benchmarks to 3D turbulent flows at Reynolds numbers up to 400,000, compatible with standard RL libraries (Stable-Baselines3, RLlib, etc.). This release introduces a data manager abstraction that handles environment data natively and uniformly across solver backends. Accompanying environment checkpoints are hosted in the companion HuggingFace dataset repository (https://huggingface.co/datasets/dynamicslab/HydroGym-environments) and are auto-accessed at runtime. Documentation is available at https://dynamicslab.github.io/hydrogym. This archive corresponds to the software accompanying the associated Nature publication.

Authors 21

1. Christian Lagemann

   University of Washington

   Affiliation as printed

   University of Washington

2. Sajeda Mokbel

   University of Washington

   Affiliation as printed

   University of Washington

3. Miro Gondrum Aachen

   RWTH Aachen University

   Affiliation as printed

   RWTH Aachen University

4. Mario Rüttgers

   Inha University

   Affiliation as printed

   Inha University

5. Yuning Wang

   University of Michigan

   Affiliation as printed

   University of Michigan

6. Pol Suárez

   KTH Royal Institute of Technology

   Affiliation as printed

   KTH Royal Institute of Technology

7. Ludger Paehler

   Technical University of Munich

   Affiliation as printed

   Technical University of Munich

8. Deniz A. Bezgin

   Technical University of Munich

   Affiliation as printed

   Technical University of Munich

9. Aaron B. Buhendwa

   Technical University of Munich

   Affiliation as printed

   Technical University of Munich

10. Jared Callaham

    University of Washington

    Affiliation as printed

    University of Washington

11. Samuel Ahnert

    University of Washington

    Affiliation as printed

    University of Washington

12. Nicholas Zolman

    University of Washington

    Affiliation as printed

    University of Washington

13. Xingchen Shao Aachen

    RWTH Aachen University

    Affiliation as printed

    RWTH Aachen University

14. Jean-Christophe Loiseau

    HESAM Université

    Affiliation as printed

    HESAM Université

15. Nikolaus A. Adams

    Technical University of Munich

    Affiliation as printed

    Technical University of Munich

16. Matthias H. Meinke Aachen

    RWTH Aachen University

    Affiliation as printed

    RWTH Aachen University

17. Wolfgang Schröder Aachen

    RWTH Aachen University

    Affiliation as printed

    RWTH Aachen University

18. Kai Lagemann

    MediaTek (China) · MediaTek (Taiwan)

    Affiliation as printed

    MediaTek Research

19. Esther Lagemann

    University of Washington

    Affiliation as printed

    University of Washington

20. Ricardo Vinuesa

    University of Michigan

    Affiliation as printed

    University of Michigan

21. Steven L. Brunton

    University of Washington

    Affiliation as printed

    University of Washington