Postdoc at the University of Twente
4D-Flow or how to reconstruct vascular flow from ultrasound imaging with contrast agents. Supervision: Guillaume Lajoinie and Michel Versluis.
Théodore Michel
he/him
Postdoctoral Researcher, Physics of Fluids group, University of Twente
Enschede, the Netherlands
Hi! I'm a Postdoctoral Researcher in the Physics of Fluids group, University of Twente. My research lies at the intersection of fluid dynamics, machine learning and high-performance computing, currently working on 4D-Flow or how to reconstuct blood flow from ultrasound imaging with contrast agents.
I completed my PhD at Université PSL in 2026 with a thesis on building a digital twin of a solar power plant, combining high-fidelity CFD, adaptive meshing, deep reinforcement learning and graph neural network surrogates to predict and optimize how solar panel arrays should behave in strong winds. The PhD was in joint partnership with TSE, working on agrivoltaic systems.
I'm always happy to chat about research, collaborations or opportunities — feel free to reach out.
Currently working on
Graph-DRL for flow surrogate modeling and optimization
Graph neural networks and deep reinforcement learning for flow control and optimization.
ONECYL: benchmarking graph-based CFD surrogates
A benchmark of 450 high-fidelity simulations of unsteady flow past a cylinder, across laminar, transitional and high-Reynolds regimes, with a Graph Transformer baseline for long-horizon prediction.
News
- 09 / 2026Started a postdoct in the Physics of Fluids group at the University of Twente, supervised by Guillaume Lajoinie and Michel Versluis.
- 09 / 2026New preprint: ONE CYLinder: A Benchmark for Graph-Based Surrogate Modeling of Unsteady Bluff-Body Flows is on arXiv.
- 02 / 2026I defended my PhD thesis, Towards a Digital Twin of a Solar Power Plant, at Université PSL.
- 02 / 2025Presented our work on graph neural networks for CFD at DTE & AICOMAS 2025 in Paris.
- 12 / 2024Our paper on combining machine learning and CFD for solar panel tilt optimization in extreme winds was published in Physics of Fluids.
- 10 / 2024Our paper on fast characterization of large earthquakes from prompt elastogravity signals was published in Communications Earth & Environment.
- 07 / 2024Presented at the 16th World Congress on Computational Mechanics (WCCM 2024) in Vancouver.
CV at a glance
Full CVEducation
- 2022 - 2026 PhD in Computational Mathematics, HPC & Data Université PSL, at CEMEF Mines Paris Thesis: Towards a Digital Twin of a Solar Power Plant (CIFRE with TSE)
- 2018 - 2022 MSc & Executive Engineering in Applied Maths, AI & Big Data Mines Paris PSL
- 2016 - 2018 Preparatory classes (CPGE) in Maths & Physics Lycée Joffre, Montpellier
- 2018 Music Diploma (saxophone, music theory) Conservatoire à Rayonnement Régional de Montpellier
- 2016 International Baccalaureate Lycée Jules Guesde, Montpellier
Experience
- 2026 - present Postdoctoral Researcher Physics of Fluids, University of Twente Deep learning for ultrasound super-resolution with contrast agents, for ERC FALCON
- 2026 Postdoctoral Researcher CEMEF Mines Paris PSL GNN + deep RL surrogates for turbulent flows, for ERC Cure
- 2022 - 2025 Engineer in Research TSE, Bourgoin-Jallieu Digital twins of agrivoltaic power plants (CIFRE PhD)
- 2023 Visiting Researcher Farhat Research Group, Stanford University Subgrid modeling
- 2022 Climate Risk Analyst Intern AXA Climate, Paris
- 2021 Research Intern IRD Géoazur
- 2020 - 2021 Data Scientist Intern Solaïs, Sophia-Antipolis
- 2020 Research Intern CEMEF Mines Paris PSL