arXiv subjects

J. Rabault

Publications and source records attributed to J. Rabault.

10 recordsLinked to original sources

arXiv2025

SvalMIZ-25 Svalbard Marginal Ice Zone Campaign 2025 -- Cruise Report

The coupling of weather, sea-ice, ocean, and wave forecasting systems has been a long-standing research focus to improve Arctic forecasting systems and their realism and is also a priority of international initiatives such as the WMO research project PCAPS. The goal of the Svalbard Marginal Ice Zone 2025 Campaign (SvalMIZ-25) was to observe and better understand the complex interplay between atmosphere, waves, and sea-ice in the winter Marginal Ice Zone (MIZ) in order to advance predictive skill of coupled Arctic forecasting systems. The main objective has been to set up a network of observations with a spatial distribution that allows for a representative comparison between in situ observations and gridded model data. The observed variables include air and surface temperature, sea-ice drift, and wave energy spectra. With the support of the Norwegian Coast Guard, we participated in the research cruise with KV Svalbard from 22.April - 11.May 2025. In total 21 buoys were deployed in the Marginal Ice Zone north of the Svalbard Archipelago.

physics.ao-ph↗

arXiv2025

Discovering Flow Separation Control Strategies in 3D Wings via Deep Reinforcement Learning

In this work, deep reinforcement learning (DRL) is applied to active flow control (AFC) over a threedimensional SD7003 wing at a Reynolds number of Re = 60,000 and angle of attack of AoA = 14 degrees. In the uncontrolled baseline case, the flow exhibits massive separation and a fully turbulent wake. Using a GPU-accelerated CFD solver and multi-agent training, DRL discovers control strategies that enhance lift (79%), reduce drag (65%), and improve aerodynamic efficiency (408%). Flow visualizations confirm reattachment of the separated shear layer, demonstrating the potential of DRL for complex and turbulent flows.

cs.CE↗

arXiv2025

Deep Reinforcement Learning for Active Flow Control around a Three-Dimensional Flow-Separated Wing at Re = 1,000

This study explores the use of deep reinforcement learning (DRL) for active flow control (AFC) to reduce flow separation on wings at high angles of attack. Concretely, here the DRL agent controls the flow over the three-dimensional NACA0012 wing section at the Reynolds number Re = 1,000 and angle of attack AoA = 20 degrees, autonomously identifying optimal control actions through real-time flow data and a reward function focused on improving aerodynamic performance. The framework integrates the GPU-accelerated computational fluid dynamics (CFD) solver SOD2D with the TF-Agents DRL library via a Redis in-memory database, enabling rapid training. This work builds on previous DRL flow-control studies, demonstrating DRL potential to address complex aerodynamic challenges and push the boundaries of traditional AFC methods.

cs.CE↗

arXiv2024

Flow control of three-dimensional cylinders transitioning to turbulence via multi-agent reinforcement learning

Designing active-flow-control (AFC) strategies for three-dimensional (3D) bluff bodies is a challenging task with critical industrial implications. In this study we explore the potential of discovering novel control strategies for drag reduction using deep reinforcement learning. We introduce a high-dimensional AFC setup on a 3D cylinder, considering Reynolds numbers ($Re_D$) from $100$ to $400$, which is a range including the transition to 3D wake instabilities. The setup involves multiple zero-net-mass-flux jets positioned on the top and bottom surfaces, aligned into two slots. The method relies on coupling the computational-fluid-dynamics solver with a multi-agent reinforcement-learning (MARL) framework based on the proximal-policy-optimization algorithm. MARL offers several advantages: it exploits local invariance, adaptable control across geometries, facilitates transfer learning and cross-application of agents, and results in a significant training speedup. \rev{For instance, our results demonstrate $16\%$ drag reduction for $Re_D=400$, outperforming classical periodic control, which yields up to $6\%$ reduction.} A proper-orthogonal-decomposition (POD) analysis at $Re_D=400$ reveals that the DRL control results in a stable wake structure with longer recirculation bubble. To the authors' knowledge, the present MARL-based framework represents the first time where training is conducted in 3D cylinders. This breakthrough paves the way for conducting AFC on progressively more complex turbulent-flow configurations.

physics.flu-dyn↗

arXiv2024

Active flow control for drag reduction through multi-agent reinforcement learning on a turbulent cylinder at $Re_D=3900$

This study presents novel drag reduction active-flow-control (AFC) strategies} for a three-dimensional cylinder immersed in a flow at a Reynolds number based on freestream velocity and cylinder diameter of $Re_D=3900$. The cylinder in this subcritical flow regime has been extensively studied in the literature and is considered a classic case of turbulent flow arising from a bluff body. The strategies presented are explored through the use of deep reinforcement learning. The cylinder is equipped with 10 independent zero-net-mass-flux jet pairs, distributed on the top and bottom surfaces, which define the AFC setup. The method is based on the coupling between a computational-fluid-dynamics solver and a multi-agent reinforcement-learning (MARL) framework using the proximal-policy-optimization algorithm. This work introduces a multi-stage training approach to expand the exploration space and enhance drag reduction stabilization. By accelerating training through the exploitation of local invariants with MARL, a drag reduction of approximately 9% is achieved. The cooperative closed-loop strategy developed by the agents is sophisticated, as it utilizes a wide bandwidth of mass-flow-rate frequencies, which classical control methods are unable to match. Notably, the mass cost efficiency is demonstrated to be two orders of magnitude lower than that of classical control methods reported in the literature. These developments represent a significant advancement in active flow control in turbulent regimes, critical for industrial applications.

physics.flu-dyn↗

arXiv2024

Buoy measurements of strong waves in ice amplitude modulation: a signature of complex physics governing waves in ice attenuation

The Marginal Ice Zone (MIZ) forms a critical transition region between the ocean and sea ice cover as it protects the close ice further in from the effect of the steepest and most energetic open ocean waves. As waves propagate through the MIZ, they get exponentially attenuated. Unfortunately, the associated attenuation coefficient is difficult to accurately estimate and model, and there are still large uncertainties around which attenuation mechanisms dominate depending on the conditions. This makes it challenging to predict waves in ice attenuation, as well as sea ice breakup and dynamics. Here, we report in-situ observations of strongly modulated waves-in-ice amplitude, with a modulation period of around 12 hours. We show that simple explanations, such as changes in the incoming open water waves, or the effect of tides and currents and bathymetry, cannot explain for the observed modulation. Therefore, the significant wave height modulation observed in the ice most likely comes from a modulation of the waves-in-ice attenuation coefficient. To explain this, we conjecture that one or several waves-in-ice attenuation mechanisms are periodically modulated and switched on and off in the area of interest. We gather evidence that sea ice convergence and divergence is likely the factor driving this change in the waves in ice attenuation mechanisms and attenuation coefficient, for example by modulating the intensity of floe-floe interaction mechanisms.

physics.ao-ph↗

arXiv2023

Drag-reduction strategies in wall-bounded turbulent flows using deep reinforcement learning

In this work we compare different drag-reduction strategies that compute their actuation based on the fluctuations at a given wall-normal location in turbulent open channel flow. In order to perform this study, we implement and describe in detail the reinforcement-learning interface to a computationally-efficient, parallelized, high-fidelity solver for fluid-flow simulations. We consider opposition control (Choi, Moin, and Kim, Journal of Fluid Mechanics 262, 1994) and the policies learnt using deep reinforcement learning (DRL) based on the state of the flow at two inner-scaled locations ($y^+ = 10$ and $y^+ = 15$). By using deep deterministic policy gradient (DDPG) algorithm, we are able to discover control strategies that outperform existing control methods. This represents a first step in the exploration of the capability of DRL algorithm to discover effective drag-reduction policies using information from different locations in the flow.

physics.flu-dyn↗

arXiv2023

Estimating sea ice properties from wave observations in sea ice

The Marginal Ice Zone is a highly dynamic region where the atmosphere, ocean, waves and sea ice meet. Waves play a fundamental role in this coupled system, but progress in our understanding of wave-ice interactions is currently hindered by the lack of observations of sea ice properties in-situ. In this study we aim to estimate the ice thickness and effective elastic modulus of sea ice passively using observations of waves in ice. Specifically, we use three low-cost geophones deployed in triangular formation with sides of about 200 m on fast ice. The focus here is on three major wave events recorded, each consisting of initial high frequency vibrations with a frequency at around 10 Hz, followed by low frequency dispersive waves within a frequency range of 0.08-0.28 Hz. Based on the phase speed of the initial high frequency vibrations, we estimate the purely elastic effective modulus to be 4-4.5 GPa. By comparing the arrival times of the dispersive low frequency wave events to the dispersion relationship of waves in a thin elastic ice sheet we obtain estimates of the effective elastic modulus in the range of 0.4-0.7 GPa. This is close to the measured effective elastic modulus through cantilever experiments of 0.5 GPa but considerably smaller than the default value of 5.5 GPa currently in use in contemporary wave models. We could not, however, obtain explicit estimates of the ice thickness and effective modulus individually as their impacts on the shape of the dispersion relationship are similar in this frequency range. Distinction is only possible for dispersive wave events larger than about 1 Hz, such as for vibrations generated by (thermal) cracking, which are frequently observed in our dataset. With our results, we show that low-cost geophones can be used to estimate sea ice properties in fast ice and substantiate that this to be possible on very large ice floes as well.

physics.geo-ph↗

arXiv2023

Deep reinforcement learning for turbulent drag reduction in channel flows

We introduce a reinforcement learning (RL) environment to design and benchmark control strategies aimed at reducing drag in turbulent fluid flows enclosed in a channel. The environment provides a framework for computationally-efficient, parallelized, high-fidelity fluid simulations, ready to interface with established RL agent programming interfaces. This allows for both testing existing deep reinforcement learning (DRL) algorithms against a challenging task, and advancing our knowledge of a complex, turbulent physical system that has been a major topic of research for over two centuries, and remains, even today, the subject of many unanswered questions. The control is applied in the form of blowing and suction at the wall, while the observable state is configurable, allowing to choose different variables such as velocity and pressure, in different locations of the domain. Given the complex nonlinear nature of turbulent flows, the control strategies proposed so far in the literature are physically grounded, but too simple. DRL, by contrast, enables leveraging the high-dimensional data that can be sampled from flow simulations to design advanced control strategies. In an effort to establish a benchmark for testing data-driven control strategies, we compare opposition control, a state-of-the-art turbulence-control strategy from the literature, and a commonly-used DRL algorithm, deep deterministic policy gradient. Our results show that DRL leads to 43% and 46% drag reduction in a minimal and a larger channel (at a friction Reynolds number of 180), respectively, outperforming the classical opposition control by around 20 percentage points.

physics.flu-dyn↗

arXiv2020

Experimental evidence for a universal threshold characterizing wave-induced sea ice break-up

Waves can drastically transform a sea ice cover by inducing break-up over vast distances in the course of a few hours. However, relatively few detailed studies have described this phenomenon in a quantitative manner, and the process of sea ice break-up by waves needs to be further parameterized and verified before it can be reliably included in forecasting models. In the present work, we discuss sea ice break-up parameterization and demonstrate the existence of an observational threshold separating breaking and non-breaking cases. This threshold is based on information from two recent field campaigns, supplemented with existing observations of sea ice break-up. The data used cover a wide range of scales, from laboratory-grown sea ice to polar field observations. Remarkably, we show that both field and laboratory observations tend to converge to a single quantitative threshold at which the wave-induced sea ice break-up takes place, which opens a promising avenue for robust parametrization in operational forecasting models.

physics.ao-ph↗