Heat Transfer Enhancement by Convective Boiling in Microchannels applied to the Cooling of Computing Units in Data Centers
The proposed PhD thesis aims to improve the understanding and modeling of convective boiling phenomena in microchannels for new low-environmental-impact refrigerants. The candidate will adopt a combined experimental and multi-scale modeling approach, including the design of a test bench simulating the behavior of a micro-evaporator, the implementation of CFD simulations (ANSYS Fluent, CATHARE) to describe two-phase flow regimes, and the evaluation of various eco-friendly alternative fluids. The expected outcomes include, for each of these new fluids, the characterization of confined boiling mechanisms, the development of a predictive heat transfer model, and the proposal of innovative cooling solutions.
The growing demand for high-performance computing, driven by artificial intelligence and cloud technologies, leads to a significant increase in power dissipation in electronic chips. Current single-phase cooling technologies are reaching their limits when dealing with heat fluxes exceeding 100 W/cm². Two-phase cooling, based on fluid boiling to remove heat, can achieve much higher heat transfer performance than single-phase systems while reducing overall energy consumption. The results of this research will contribute to the development of more efficient and sustainable cooling solutions for future data centers, helping to reduce the digital sector’s energy footprint and strengthen European technological sovereignty in advanced cooling technologies.
Development of Machine Learning algorithm to optimize the control of absorption machines
The Thermal and Solar Technologies Laboratory (L2TS) and the Energy Systems for Territories Laboratory (LSET), located at the CEA LITEN site in Le Bourget-de-Lac, are offering a cross-disciplinary PhD thesis combining thermodynamics and optimization using Artificial Intelligence.
Specifically, this doctoral research project involves developing a machine learning algorithm to optimize the control of absorption machines. These machines are thermodynamic cycles able to produce heat or cold from an intermediate heat input; thus, offering potential valorization of industrial waste heat or renewable energies, such as solar thermal. Heat exchange is made possible by the absorption and desorption reactions of a gaseous refrigerant in a fluid. Specifically, the NH3-H2O mixture will be used. The dynamic operation of these cycles is extremely complex because the operational variables, physical parameters, and hydrodynamic aspects are highly intertwined. Thus, the use of a neural network is particularly relevant for establishing an adaptive control strategy for these machines.
The thesis will have a theoretical aspect, involving the study and selection of the most suitable algorithm to address the problem, and an experimental aspect of validation on a prototype absorption machine. The project will also involve the design of a controller for implementation.
The thesis will take place in a CEA laboratory in Bourget du Lac.
Resilience of fusion plasmas in a metallic environment, from WEST to ITER
Magnetic confinement nuclear fusion is an attractive option for contributing to the future energy mix, and the ITER project will, in the coming decade, mark a new milestone in the scientific and technological development of this field by producing more fusion energy than the energy deposited to sustain it. However, in a fusion power plant, the wall of the combustion chamber will be subjected to strong thermal and neutron stresses and must also limit the trapping of hydrogen isotopes used in the nuclear reaction.
The material considered the best compromise is tungsten, a metal whose high melting point and lack of chemical affinity with hydrogen are its main advantages. However, its high atomic number makes it highly radiative in the plasma where the reactions occur, which is detrimental to energy confinement and overall performance. It is therefore crucial to understand—both on current machines and through simulations for ITER—the impact of the inevitable tungsten dust (observed in the WEST tokamak) on turbulent transport, magneto-hydrodynamic stability, and ultimately on achieving a viable scenario for nuclear fusion. These aspects will form the foundation of the PhD project, combining experimental analysis on WEST at CEA with validation through simulations that include all relevant aspects, and extrapolation to the ITER environment. This work will be carried out in collaboration with ITER, the UKAEA (United Kingdom) for the simulation code, the CNR-Milano team for the tungsten dust trajectory, and the CEA teams at the IRFM.
Development of a new numerical scheme, based on T-coercivity, for discretizing the Navier-Stokes equations.
In the TrioCFD code, the discretization of the Navier-Stokes equations leads to a three-step algorithm (see Chorin'67, Temam'68): velocity prediction, pressure solution, velocity correction. If an implicit time discretization scheme is to be used, the pressure solution step is particularly costly. Thus, most simulations are performed using an explicit time scheme, for which the time step depends on the mesh size, which can be very restrictive. We would like to develop an implicit time discretization scheme using a stabilized formulation of the Navier-Stokes problem based on explicit T-coercivity (see Ciarlet-Jamelot'25). It would then be possible to solve an implicit scheme directly without a correction step, which could significantly improve the performance of the calculations. This would also allow the use of the P1-P0 finite element pair, which is frugal in terms of degrees of freedom but unstable for a classical formulation.
Exploring the Strategic Benefits of 0V Storage for Na-ion Batteries
Recently deployed on a commercial scale, the Na-ion battery technology demonstrates excellent behaviour during medium or long-term storage at zero voltage. This characteristic offers numerous safety advantages during the transport, assembly and storage of cells and modules, as well as during emergency shutdowns in the event of external issues. But are there no consequences for battery performance?
This research project aims to study and better understand the electrochemical mechanisms at play when the potential difference across the terminals is maintained at 0 V.
Initially, advanced dynamic characterisation techniques will be used to analyse and compare the electrochemical, thermal and mechanical properties of battery materials. The results will enrich calendar and cycling ageing models at the cell scale, thereby improving their accuracy and reliability. Subsequently, tests will be conducted on mini-battery modules assembled in various electrical architectures to study cell behaviour during cycling and ageing, particularly in response to the application of negative voltage. Specific battery management system (BMS) solutions could then be proposed to address these issues.
The scientific approach will involve implementing advanced characterisation and instrumentation techniques, conducting ageing and safety tests to identify mechanisms, and developing ageing models. This approach will draw on the expertise and testing facilities of CEA-Liten at the Bourget du Lac site in Savoie.
Code Development and Numerical Simulation of Gas Entrainment in Sodium-Cooled Fast Reactors
In sodium-cooled fast reactors (SFRs), the circulation of liquid sodium is ensured by immersed centrifugal pumps. Under certain conditions, vortices can develop in recirculation zones, promoting the entrainment of inert gas bubbles (typically argon) located above the free surface. If these bubbles are drawn into the primary circuit, they can damage pump components and compromise the safety of the installation. This phenomenon remains difficult to predict, particularly during the design phase, as it depends on numerous physical, geometrical, and numerical parameters.
The objective of this PhD work is to contribute to a better understanding and modeling of gas entrainment in free-surface flows typical of SFRs, through Computational Fluid Dynamics (CFD) simulations using the open-source code TrioCFD, developed by the CEA. This code includes an interface-tracking module (Front Tracking) that is particularly well-suited for simulating two-phase phenomena involving a deformable free interface.
development of a NET (Negative Emission Technologie) process combining CO2 capture and hydrogenation into synthetic fuel
Until recently, CO2 capture technologies were developed separately from CO2 utilization technologies, even though coupling the CO2 desorption stage with the chemical transformation of CO2, which is generally exothermic, would yield significant energy savings.
The first coupled solutions have recently been proposed, but they are mainly at moderate temperatures (100-180°C) [1], or even recently close to 225°C [2].
The objective of this doctoral thesis is to study, both experimentally and theoretically, a coupled system in the 250-325°C temperature range that allows via Fischer-Tropsch-type catalytic hydrogenation the direct production of higher value-added products
Fluid-structure interaction in a network of slender solids in a confined environment
As part of its study of progressive deformations in fuel assemblies within PWR cores, the CEA has developed two simulation tools. The first, Phorcys [1], calculates the flow of coolant in and around slightly deformed assemblies using a network of parametric pressure drops, then deduces the fluid forces acting on the structures. The second, DACC [2], uses finite element simulation to analyze thermomechanical behavior under irradiation and the interaction between assemblies during power cycles. Finally, fluid-structure interaction is analyzed using numerical coupling of these two tools, within which uncertainties can be propagated and analyzed [3].
The nuclear revival program (SMR, 4th generation reactors, PN, etc.) is providing new technologies and new core and fuel assembly topologies that need to be analyzed in terms of the risks associated with quasi-static deformations of core assemblies. With a view to both capitalizing on and extending the possibilities of simulation, the aim is to enable these two tools to handle the flow and deformation of slender structures in a more generic way in order to cover a wide range of nuclear technologies efficiently and quickly.
To do this, it will be necessary to identify, classify, and then model in a reduced but predictive manner the main flow structures that may occur within a fluid volume cluttered with slender structures with a large exchange surface area. The complete hydraulic model of the core will thus be created by concatenating elementary models that comply with strict interfacing conditions. A method for analyzing the overall flow obtained will then enable the quantification of the force field contributing to the deformations. A similar logic of classification and scaling would also be implemented with regard to the evaluation of reversible and irreversible deformations of a slender structure subjected to external stresses and severe irradiation. One difficulty is that the fine topology of a fuel assembly can exhibit nonlinearities at small scales that propagate in part to the macroscopic scale. Ultimately, a robust, cost-effective partitioned coupling will have to be implemented between the coolant flow and these individual structures, which deform and interact in a constrained environment.
The modeling framework thus constructed will make it possible to study the progressive deformations of assemblies and the associated risks for a wide range of nuclear reactor technologies.
Modeling of a non-equilibrium dispersed phase and its fragmentation
In the context of the sustainable use of nuclear energy to produce carbon-free electricity, fourth-generation reactors, also known as "fast neutron" reactors, are necessary to close the fuel cycle.
This thesis falls within the framework of safety studies associated with such sodium-cooled reactors, and more particularly the hypothetical situation of a molten core relocating by gravity towards the core catcher at the bottom of the reactor vessel. A jet of corium (mixture of molten fuel and structural elements of the core) then interacts violently with the coolant, inducing, among other things, the fragmentation of the corium jet into droplets coupled with film boiling of the coolant. Characteristics of the resulting dispersed phase of corium and its fragmentation are crucial for studying the risk of runaway and steam explosion.
The aim of this thesis is to model a dispersed phase and its fragmentation in a surrounding fluid, using an approach that is both efficient and able to account to the scale variations and thermal imbalances between the droplets and the carrier phase. The method considered to meet these objectives is the method of moments, which derives from a kinetic model. It requires adequate closure and numerical schemes that satisfy non-standard constraints, while offering, in return, a crucial cost/accuracy compromise in the context studied. The advancements will be a priori implemented in the CFD software SCONE, built on the CEA's open-source TRUST platform.
The main work location will be based at the LMAG (Laboratory of Severe Accidents Modeling) at the IRESNE Institute of CEA Cadarache. Part of the work will also be carried out at the EM2C Laboratory (Molecular and Macroscopic Energetics, Combustion) – CNRS/CentraleSupélec in Paris.
The future PhD will work in a scientific dynamic environment and will acquire skills enabling to aspire to academic and industrial R&D positions.
Keywords : Dispersed Phase, Fragmentation, Kinetic, Method of Moments, Multiphase, Numerical methods, Severe Accidents.