{"success":true,"data":{"author":"Zachary Cooper-Baldock, Paulo E. Santos, Russell S. A. Brinkworth, Karl Sammut","bucketAccessType":"public","bucketName":"ZacharyCB99/WAKESET","bucketPrefix":"","category":"Physics","cloudProvider":"huggingface","createdAt":"2026-11-18T00:06:05.956Z","description":"Large CFD dataset of turbulent wake flow around a generalized extra-large uncrewed underwater vehicle (XLUUV) hull, for machine learning of wake dynamics. It holds 1,091 Reynolds-averaged Navier-Stokes (RANS) simulations parameterized by forward speed and turning (yaw) angle, augmented to 4,364 instances. Volumes/ holds fields resampled onto a regular 128 x 128 x 128 Cartesian grid (CUBE_128) as ML-ready tensors; Planes/ holds vertical and horizontal planar slices (VERTPLN, HORZPLN). Files follow Forward_<speed mm/s>_ms_Angle_<deg>_<type>; variables are velocity components u, v, w and magnitude, total, absolute and dynamic pressure, vorticity magnitude and turbulence intensity. Examples/Python/ has loader scripts. The repository is about 480 GB. Released by Cooper-Baldock, Santos, Brinkworth and Sammut under CC BY 4.0 as stated in the card; cite arXiv:2602.01379 and doi:10.57967/hf/7698.","displayTitle":"WAKESET: 1,091 High-Reynolds-Number RANS Simulations of Turbulent Wakes behind an XLUUV","doi":"10.57967/hf/7698","huggingFaceUrl":"https://huggingface.co/datasets/ZacharyCB99/WAKESET","id":"d4b498bd-4802-5d17-83b5-57600c20c693","license":"CC-BY-4.0","licenseEvidenceUrl":"https://huggingface.co/datasets/ZacharyCB99/WAKESET","objectCountEstimate":3283,"probeStatus":"ok","referenceUrl":"https://arxiv.org/abs/2602.01379","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"ZacharyCB99/WAKESET\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"ZacharyCB99/WAKESET\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=ZacharyCB99, license:cc-by-4.0, size_categories:n<1K, format:text, downloads=2238"],"format":"mixed","generatedAt":"2026-11-18T00:10:02.414Z","layout":{"example":"Planes/Vertical/Forward_0100_ms_Angle_00_VERTPLN_ALL_processed.npz","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"Planes/Vertical/Forward_0100_ms_Angle_00_VERTPLN_ALL_processed.npz","truncatedAt":512}],"sampleObjects":["Examples/Python/WAKESET_pytorch.py","Examples/Python/load_planes.py","Examples/Python/load_visualizations.py","Examples/Python/load_volumes.py","Examples/Python/requirements.txt"],"sidecars":[{"bytes":"# WAKESET: A Large-Scale, High-Reynolds Number Flow Dataset\r\n\r\n**WAKESET** is a comprehensive Computational Fluid Dynamics (CFD) dataset designed for Machine Learning applications in fluid mechanics. It captures the complex hydrodynamic interactions of an Extra Large Uncrewed Underwater Vehicle (XLUUV).\r\n\r\nThe dataset comprises **1,091 high-fidelity RANS simulations** (augmented to 4,364 instances via the provided `WAKESET_pytorch.py`), covering Reynolds numbers up to $1.09 \\times 10^8$ and turning angles up to 60 degrees.\r\n\r\n## Directory Structure\r\n\r\n```text\r\nWAKESET/\r\n|-- Volumes/           # 3D interpolated grids (128x128x128)\r\n|   |-- Forward_0100_ms_Angle_00_CUBE_128/\r\n|   |-- Forward_0100_ms_Angle_05_CUBE_128/\r\n|   |-- ...\r\n|-- Planes/            # 2D slices (Vertical and Horizontal)\r\n|   |-- Vertical/\r\n|   |   |-- Forward_0100_ms_Angle_00_VERTPLN_ALL/\r\n|   |   |-- Forward_0100_ms_Angle_05_VERTPLN_ALL/\r\n|   |   |-- ...\r\n|   |-- Horizontal/\r\n|   |   |-- Forward_0100_ms_Angle_00_HORZPLN_ALL/\r\n|   |   |-- Forward_0100_ms_Angle_05_HORZPLN_ALL/\r\n|   |   |-- ...\r\n|-- Examples/          # Python Toolkit\r\n|   |-- Python/\r\n|       |-- requirements.txt\r\n|       |-- WAKESET_pytorch.py     # ML Dataloader with on-the-fly augmentation\r\n|       |-- load_planes.py         # Utilities for 2D data\r\n|       |-- load_volumes.py        # Utilities for 3D data\r\n|       |-- load_visualizations.py # Plotting tools\r\n```\r\n\r\n# Quick Start\r\n\r\n## 1. Installation\r\nThe dataset includes a Python toolkit to streamline loading, parsing, and augmentation.\r\n\r\n```bash\r\ncd Examples/Python\r\npip install -r requirements.txt\r\n```\r\n\r\n## 2. Loading 3D Volumes (CFD Data)\r\nThe `load_volumes.py` script handles the parsing of sparse CFD exports and reshapes them into structured grids.\r\n\r\n```python\r\nimport sys\r\nsys.path.append(\"Examples/Python\")\r\nfrom load_volumes import load_volume\r\n\r\n# Load a specific volume\r\nvol_data = load_volume(\r\n    velocity=1.0, \r\n    angle=0, \r\n    variable=\"velocity_magnitude\", \r\n    data_dir=\"../../Volumes\"\r\n)\r\n\r\nprint(f\"Loaded Volume Shape: {vol_data.values.shape}\") # (128, 128, 128)\r\n```\r\n\r\n## 3. Machine Learning (PyTorch)\r\nUse the `WAKESET_pytorch.py` wrapper to plug the dataset directly into an ML pipeline. This loader handles on-the-fly augmentation (rotation and flipping) to expand the effective dataset size to 4,364 instances without using extra disk space.\r\n\r\n```python\r\nfrom torch.utils.data import DataLoader\r\nfrom WAKESET_pytorch import WakesetVolumeDataset\r\n\r\n# Initialize Dataset\r\ndataset = WakesetVolumeDataset(\r\n    root_dir=\"../../\", \r\n    subset='train', \r\n    augment=True  # Enables physics-informed rotation/flipping\r\n)\r\n\r\nloader = DataLoader(dataset, batch_size=4, shuffle=True)\r\n\r\n# Training Loop\r\nfor flow_field, kinematics in loader:\r\n    # flow_field: [Batch, 1, 128, 128, 128]\r\n    # kinematics: [Batch, 2] (Speed, Angle)\r\n    print(flow_field.shape, kinematics.shape)\r\n    break\r\n```\r\n\r\n## 4. Visualization\r\nVisualise the data using `load_visualizations.py`.\r\n\r\n```python\r\nfrom load_visualizations import visualize_volume_slices\r\n\r\n# Visualize the center slices of the previously loaded volume\r\nvisualize_volume_slices(vol_data, variable=\"velocity_magnitude\")\r\n```\r\n\r\n# Citation\r\nIf you use WAKESET in your research, please cite:\r\n\r\nCooper-Baldock, Z., Santos, P. E., Brinkworth, R. S. A., & Sammut, K. (2026). WAKESET: A Large-Scale, High-Reynolds Number Flow Dataset for Machine Learning of Turbulent Wake Dynamics.\r\n","encoding":"utf8","key":"Examples/Python/README.md","name":"README.md","size":3474,"truncated":false},{"bytes":"Creative Commons Attribution 4.0 International Public License\n\nBy exercising the Licensed Rights (defined below), You accept and agree\nto be bound by the terms and conditions of this Creative Commons\nAttribution 4.0 International Public License (\"Public License\"). To the\nextent this Public License may be interpreted as a contract, You are\ngranted the Licensed Rights in consideration of Your acceptance of these\nterms and conditions, and the Licensor grants You such rights in\nconsideration of benefits the Licensor receives from making the Licensed\nMaterial available under these terms and conditions.\n\nSection 1 – Definitions.\n\na. Adapted Material means material subject to Copyright and Similar\nRights that is derived from or based upon the Licensed Material and in\nwhich the Licensed Material is translated, altered, arranged,\ntransformed, or otherwise modified in a manner requiring permission\nunder the Copyright and Similar Rights held by the Licensor. For purposes\nof this Public License, where the Licensed Material is a musical work,\nperformance, or sound recording, Adapted Material is always produced\nwhere the Licensed Material is synched in timed relation with a moving\nimage.\n\nb. Adapter's License means the license You apply to Your Copyright and\nSimilar Rights in Your contributions to Adapted Material in accordance\nwith the terms and conditions of this Public License.\n\nc. Copyright and Similar Rights means copyright and/or similar rights\nclosely related to copyright including, without limitation, performance,\nbroadcast, sound recording, and Sui Generis Database Rights, without\nregard to how the rights are labeled or categorized. For purposes of\nthis Public License, the rights specified in Section 2(b)(1)-(2) are not\nCopyright and Similar Rights.\n\nd. Effective Technological Measures means those measures that, in the\nabsence of proper authority, may not be circumvented under laws\nfulfilling obligations under Article 11 of the WIPO Copyright Treaty\nadopted on December 20, 1996, and/or similar international agreements.\n\ne. Exceptions and Limitations means fair use, fair dealing, and/or any\nother exception or limitation to Copyright and Similar Rights that\napplies to Your use of the Licensed Material.\n\nf. Licensed Material means the artistic or literary work, database, or\nother material to which the Licensor applied this Public License.\n\ng. Licensed Rights means the rights granted to You subject to the terms\nand conditions of this Public License, which are limited to all Copyright\nand Similar Rights that apply to Your use of the Licensed Material and\nthat the Licensor has authority to license.\n\nh. Licensor means the individual(s) or entity(ies) granting rights under\nthis Public License.\n\ni. Share means to provide material to the public by any means or process\nthat requires permission under the Licensed Rights, such as reproduction,\npublic display, public performance, distribution, dissemination,\ncommunication, or importation, and to make material available to the\npublic including in ways that members of the public may access the\nmaterial from a place and at a time individually chosen by them.\n\nj. Sui Generis Database Rights means rights other than copyright resulting\nfrom Directive 96/9/EC of the European Parliament and of the Council of\n11 March 1996 on the legal protection of databases, as amended and/or\nsucceeded, as well as other essentially equivalent rights anywhere in the\nworld.\n\nk. You means the individual or entity exercising the Licensed Rights\nunder this Public License. Your has a corresponding meaning.\n\nSection 2 – Scope.\n\na. License grant.\n\n1. Subject to the terms and conditions of this Public License, the\nLicensor hereby grants You a worldwide, royalty-free, non-sublicensable,\nnon-exclusive, irrevocable license to exercise the Licensed Rights in the\nLicensed Material to:\n\nA. reproduce and Share the Licensed Material, in whole or in part; and\n\nB. produce, reproduce, and Share Adapted Material.\n\n2. Exceptions and Limitations. For the avoidance of doubt, where\nExceptions and Limitations apply to Your use, this P","encoding":"utf8","key":"LICENSE","name":"LICENSE","size":13928,"truncated":true},{"bytes":"---\npretty_name: WAKESET\nlicense: cc-by-4.0\nlanguage:\n  - en\nsize_categories:\n  - 1K<n<10K\ntags:\n  - computational-fluid-dynamics\n  - turbulence\n  - rans\n  - 3d\n  - underwater-vehicles\ntask_categories:\n  - other\n  - text-to-3d\n  - image-to-3d\n---\n\n# WAKESET: A Large-Scale, High-Reynolds Number Flow Dataset for Machine Learning of Turbulent Wake Dynamics\n\n## What this is\nWAKESET is a large-scale CFD dataset of turbulent wake flow fields around a generalized XLUUV geometry, spanning forward speed and turning angle parameterizations. It comprises 1,091 high-fidelity Reynolds-Averaged Navier-Stokes (RANS) simulations, augmented to 4,364 instances.\n\n## Contents\n- **Volumes**: `CUBE_128` (regular 128×128×128 Cartesian grid; ML-friendly volumetric tensors)\n- **Planes**: `VERTPLN` and `HORZPLN` (planar exports / slices on the CFD mesh or interpolated grids, depending on file type)\n- **Examples**: `Examples/Python/` (loader scripts and minimal usage examples)\n\n## File naming convention\n`Forward_[VVVV]_ms_Angle_[AA]_[TYPE]_[SUFFIX]`\n\n- `VVVV`: forward speed in mm/s (e.g., `5000` = 5.00 m/s)\n- `AA`: yaw angle in degrees (e.g., `20`)\n- `TYPE`: `CUBE_128`, `VERTPLN`, `HORZPLN` (and/or your specific suffix variants)\n\n## Variables (units)\n- Velocity components: `u, v, w` [m/s]\n- Velocity magnitude: `v_mag` [m/s]\n- Pressure: `p_total`, `p_abs`, `p_dyn` [Pa]\n- Vorticity magnitude: `omega_mag` [1/s]\n- Turbulence intensity: `I` [%]\n\n## License\n- **Dataset (CFD outputs / derived fields):** released under **CC BY 4.0** (`license: cc-by-4.0`).\n  - You may share and adapt, including commercially, **with attribution**.\n  - See `LICENSE` for full terms.\n\n## Attribution / Citation\nIf you use WAKESET, please:\n1. **Cite** the associated paper\n2. Include attribution: “WAKESET dataset by Zachary Cooper-Baldock et al., CC BY 4.0.”\n\n### BibTeX\n```bibtex\n@dataset{wakeset,\n  title   = {WAKESET: A Large-Scale, High-Reynolds Number Flow Dataset for Machine Learning of Turbulent Wake Dynamics},\n  author  = {Cooper-Baldock, Zachary and Santos, Paulo E. and Brinkworth, Russell S. A. and Sammut, Karl},\n  year    = {2026},\n  publisher = {Hugging Face},\n  note    = {CC BY 4.0},\n  url     = {https://huggingface.co/datasets/ZacharyCB99/WAKESET}\n  doi     = {10.57967/hf/7698}\n}\n\n@article{wakeset_paper,\n  title         = {WAKESET: A Large-Scale, High-Reynolds Number Flow Dataset for Machine Learning of Turbulent Wake Dynamics},\n  author        = {Cooper-Baldock, Zachary and Santos, Paulo E. and Brinkworth, Russell S. A. and Sammut, Karl},\n  year          = {2026},\n  eprint        = {arXiv:2602.01379},\n  archivePrefix = {arXiv},\n  primaryClass  = {physics.flu-dyn} \n  doi           = {10.57967/HF/7698}\n}\n","encoding":"utf8","key":"README.md","name":"README.md","size":2712,"truncated":false}],"sourceObjectKey":"Planes/Vertical/Forward_0100_ms_Angle_00_VERTPLN_ALL_processed.npz","starterCell":"from huggingface_hub import hf_hub_download\npath = hf_hub_download(repo_id=\"ZacharyCB99/WAKESET\", filename=\"Planes/Vertical/Forward_0100_ms_Angle_00_VERTPLN_ALL_processed.npz\", repo_type=\"dataset\")\nprint(path)","starterSamplePath":"Examples/Python/requirements.txt","totalObjects":3283},"summary":"Large CFD dataset of turbulent wake flow around a generalized extra-large uncrewed underwater vehicle (XLUUV) hull, for machine learning of wake dynamics.","title":"WAKESET: 1,091 High-Reynolds-Number RANS Simulations of Turbulent Wakes behind an XLUUV","updatedAt":"2026-11-18T03:56:48.457Z","visibility":"public","websiteUrl":"https://huggingface.co/datasets/ZacharyCB99/WAKESET"}}