Innovative techniques for evaluating critical steps and limiting factors for batteries formation
The battery manufacturing sector in Europe is currently experiencing strong growth. The electrical formation step that follows battery assembly and precedes delivery has received little academic attention, despite being crucial for battery performance (lifespan, internal resistance, defects, etc.). It is an essential time-consuming and costly step in the process (>30% of the cell manufacturing cost, and 25% of the equipment cost in a Gigafactory) that would greatly benefit from optimization.
In this thesis, we propose studying battery formation using innovative, complementary, operando non-intrusive techniques. The goal is to identify the limiting mechanisms of the electrolyte impregnation step (filling electrode pores) and of the initial charge. The candidate will implement experimental methods to monitor and analyze these mechanisms. He will also establish a methodology and protocols for studying these steps, combining electrochemical measurements with non-intrusive physical characterizations under operating conditions. The research will focus on optimizing formation time and quality control during this stage.
Differential phase contrast imaging based on quad-pixel image sensor
Biopharmaceutical production is booming and consists of using cells to produce molecules of interest. To achieve this, monitoring the culture and the state of the cells is essential. Quantitative phase imaging by holography is a label-free optical method that has already demonstrated its ability to measure the concentration and viability of cultured cells. However, implementing this technique in a bioreactor faces several challenges related to the high cell density. It is therefore necessary to develop new quantitative phase imaging methods, such as differential phase contrast imaging.
The objective of the PhD is to develop this technique using a specific image sensor for which a prototype has been designed at CEA-LETI. The PhD candidate will use this new sensor and develop the reconstruction and image-processing algorithms. They will also identify the limitations of the current prototype and define the specifications for a second prototype that will be developed at CEA-LETI. Finally, they will consider the design of an inline probe to be immersed in the bioreactor.
Designing a hybrid CPU-GPU estimator for neutron transport: Advancing eco-efficient Monte Carlo simulations
Digital twins incorporating Monte Carlo simulation models are currently being developed for the design, operation, and decommissioning of nuclear facilities. These twins are capable of predicting physical quantities such as particle fluxes, gamma/neutron heating, and dose equivalent rates. However, the Monte Carlo method presents a major drawback: high computational time to achieve acceptable variance levels.
To enhance simulation efficiency, the eTLE estimator has been developed and integrated into the TRIPOLI-4® Monte Carlo code. Compared to the conventional TLE (Track Length Estimator), eTLE offers lower theoretical variance, particularly in highly absorbing media, by contributing to the detector response even when particles do not physically reach it. Nevertheless, its computational cost remains significant, especially when evaluating multiple detectors.
Two recent PhD works have proposed variants to overcome this limitation. The Forced Detection eTLE- (Guadagni, EPJ Plus 2021) employs preferential sampling that directs pseudo-particles toward the detector at each collision. It is particularly effective for small detectors and configurations with moderate shielding, especially for fast neutrons. The Split Exponential TLE (Hutinet & Antonsanti, EPJ Web 2024) is based on an asynchronous GPU approach, offloading straight-line particle transport to the graphics processor. Through multiple sampling, it maximizes GPU utilization and enables more efficient exploration of phase space.
The proposed thesis aims to combine these two approaches into a hybrid estimator named seTLE-DF. This new estimator could be used either directly or to generate importance maps without relying on auxiliary deterministic calculations. Its implementation will require dedicated GPU developments, particularly to optimize the geometry library and memory management in complex geometries.
This research topic aligns with green computing objectives, aiming to reduce the carbon footprint of high-performance computing. It relies on a hybrid CPU-GPU strategy, avoiding full porting of the Monte Carlo code to GPU. Solutions such as half-precision formats will be considered, and an energy impact assessment will be conducted before and after implementation. The future PhD student will be welcomed with the IRESNE Institute (CEA Cadarache)and will acquire strong expertise in neutron transport simulation, facilitating integration into major research institutions or companies within the nuclear sector.
Design artificial intelligence tools for tracking Fission Product release out of nuclear fuel
The Laboratory for the Analysis of Radionuclide Migration (LAMIR), part of the Institute for Research on Nuclear Systems (IRESNE) at CEA Cadarache, has developed a set of advanced measurement methods to characterize the release of fission products from nuclear fuel during thermal transients. Among these innovative tools is an operando in situ imaging system that enables real-time observation of these phenomena. The large amount of data generated by these experiments requires dedicated digital processing techniques that account for both the specificities of nuclear instrumentation and the underlying physical mechanisms.
The goal of this PhD project is to develop an optimized data processing approach based on state-of-the-art Artificial Intelligence (AI) methods.
In the first phase, the focus will be on processing thermal sequence images to detect and analyze material movements, aiming to identify an optimal image-processing strategy defined by rigorous quantitative criteria.
In the second phase, the methodology will be extended to all experimental data collected during a thermal sequence. The long-term objective is to create a real-time diagnostic tool capable of supporting experiment monitoring and interpretation.
This PhD will be carried out within a collaborative framework between LAMIR, which has recognized expertise in nuclear fuel behavior analysis and imaging, and the Institut Fresnel in Marseille, known for its strong background in image analysis and artificial intelligence.
The candidate will benefit from a multidisciplinary and stimulating research environment, with opportunities to present and publish their work at national and international conferences and in peer-reviewed journals.
Development of a dosimeter based on the capture of xenon in a zeolite
Reactor dosimetry makes possible to characterize the neutron spectrum (neutron energy distribution) and to determine the neutron fluence received during irradiation for monitoring the embrittlement of materials. This technique relies on analyzing the radioactivity of irradiated dosimeters, made of pure metals or alloys of known compositions, some isotopes of which undergo activation or fission reactions.
There are numerous dosimeters sensitive to 2 MeV, a few between 1 MeV and 2 MeV, but Zr is the only one suitable for the energy range between 1 keV and 1 MeV. Moreover, few dosimeters respond with a threshold close to 1 MeV in moderate-flux R&D reactors. The only one practically usable, Rh, has a half-life < 1 h, and its measurement relies solely on highly self-absorbed X-rays, requiring very thin dosimeters and complicating measurements. There is therefore a real need to develop a dosimeter capable of responding between 1 keV and 1 MeV.
In this context, Xe not only exhibits an interesting reaction already identified between 1 keV and 1 MeV, but also has two reactions close to 1 MeV producing two nuclides with half-lives of about ten days, well suited to the irradiation cycles of the upcoming high-flux experimental reactor at CEA: the Jules Horowitz Reactor (JHR).
The main idea of this thesis topic would be to use adsorbent materials to trap a sufficient mass of Xe in a reduced volume. Some commercial zeolites can now trap up to 30% by weight of Xe when exposed to only 1 bar of Xe at room temperature.
The thesis will consist of producing a Xe dosimeter trapped on a zeolite at CNRS MADIREL (frequent trips to the Saint Jérôme campus in Marseille in the first year) as well as a simplified Xe-filled chamber manufactured in in the workshops of our laboratory. The common irradiation of a dosimeter and a chamber in a reactor such as CABRI in Cadarache will allow the evaluation of the self-absorption factors by the zeolite of the gamma lines emitted by the isotopes of interest, verification of their measurability with the MADERE platform of our laboratory, as well as assessment of the ageing of zeolites under strong neutron irradiation. The dosimeter will then be tested at higher neutron flux, for example in the TRIGA reactor at JSI (one-week trip to Slovenia to be expected), through the uninterrupted CEA-JSI collaboration since 2008, in order to qualify this dosimeter for JHR.
By acquiring expertise in the field of nuclear measurement, the future PhD graduate will be well prepared for professional integration into major French and international research organizations, or in nuclear companies.
Proximal primal-dual method for joint estimation of the object and of unknown acquisition parameters in Computed Tomography.
As part of the sustainable and safe use of nuclear energy in the transition to a carbon-free energy future, the Jules Horowitz research reactor, currently under construction at the CEA Cadarache site, is a key tool for studying the behaviour of materials under irradiation. A tomographic imaging system will be exploited in support of experimental measures to obtain real-time images of sample degradation. This imaging system has extraordinary characteristics due to its geometry and to the size of the objects to be characterized. As a result, some acquisition parameters, which are essential to obtain a sufficient image reconstruction quality, are not known with precision. This can lead to a significant degradation of the final image.
The objective of this PhD thesis is to propose methods for the joint estimation of the object under study and of the unknown acquisition parameters. These methods will be based on modern convex optimization tools. This thesis will also explore machine learning methods in order to automate and optimize the choice of hyperparameters for the problem.
The thesis will be carried out in collaboration between the Marseille Institute of Mathematics (I2M CNRS UMR 7373, Aix-Marseille University, Saint Charles campus) and the Nuclear Measurement Laboratory of the IRESNE institute of the French Alternative Energies and Atomic Energy Commission (CEA Cadarache, Saint Paul les Durance). The doctoral student will work in a stimulating research environment focused on strategic questions related to non-destructive testing. He or she will also have the opportunity to promote his or her research work in France and abroad.
What mechano-thermal coupling is necessary for fast transients? Evaluation of the contributions of thermodynamics to irreversible processes.
The Laboratory for the Analysis of Radioelement Migration (LAMIR) at the Institute for Research on Nuclear Systems (IRESNE) of the CEA Cadarache has developed a set of measurement methods to characterize the release of fission products from nuclear fuel during transient thermal transients. For these transients, it is important to simulate the mechanical stresses associated with temperature changes that could lead to fracturing of the tested fuel samples . This thesis focuses on modeling hypothetical and very rapid accidental power transients. Its objective is to implement a new model based on the thermodynamics of irreversible processes (TIP).
The first part of this thesis will aim to validate the thermomechanical coupling model in TIP, which was proposed in our laboratory (https://www.mdpi.com/2813-4648/3/4/33). This will be an essentially analytical approach to establish the orders of magnitude of the various mechanisms involved. The second part will apply this formalism to experimental results obtained during rapid heating experiments using laser beams.
One of the main challenges of numerical simulation with TIP is calculating the temperature and stress fields simultaneously, rather than sequentially as in current models. We will start with a 1D program (in Python or another language) that will be progressively refined. Comparing the results obtained with TIP and with current models will help us identify situations in which TIP-specific couplings must be taken into account to achieve accurate predictions.
The PhD candidate will benefit from the support of experts in thermodynamics, mechanics, and programming. The research will lead to scientific publications and conference presentations. Owing to the diversity of the fields involved, this thesis topic offers excellent career prospects in both industry and academic research.
Design of an integrated circuit for decoding motor brain activity for autonomous use of a brain-machine interface for motor substitution
This work is part of the development of brain-machine interfaces dedicated to restoring mobility for patients with severe chronic motor disabilities. The proposed technological solutions are based on decoding brain signals acquired at the motor cortex level in order to extract movement intentions. These intentions serve as commands for motor compensation systems. Our team is a pioneer in this field, having developed WIMAGINE, one of the first chronic wireless implants, as well as a decoder and effectors adapted to the needs of paraplegic or quadriplegic patients (Benabid et al, The Lancet Neurology, 2019 ; Lorach et al, Nature 2023).
The proposed research follows on from an initial thesis whose objective was to design an integrated circuit capable of replicating the performance of the brain signal decoder with extremely low energy consumption, using a fixed model. However, due to changes in the user's strategy or the natural evolution of their brain structures, the performance of the decoding model tends to deteriorate over time, requiring regular recalibration. Initial strategies to compensate for these phenomena have been identified. The candidate's objective will be to refine these strategies and propose an implementation in the form of a low-power digital circuit.
The thesis will be carried out in Grenoble, within a dynamic project team composed of recognized experts in the design and clinical validation of brain-machine interfaces. The team is particularly distinguished in the design of specific integrated circuits and the development of signal decoding algorithms. This framework will allow the doctoral student to evolve in a stimulating scientific environment and to promote their research work, both in France and abroad.
Acoustic and Ultrasound-based Predictive Maintenance Systems for Industrial Equipment
Power converters are essential in numerous applications such as industry, photovoltaic systems, electric vehicles, and data centers. Their conventional maintenance is often based on fixed schedules, leading to premature replacement of components and significant electronic waste.
This PhD project aims to develop a novel non-invasive and low-cost ultrasound-based monitoring approach to assess the state of health and remaining useful life (RUL) of power converters deployed across various industries.
The research will focus on identifying and characterizing ultrasonic signatures emitted by aging electronic components, and on developing physics-informed neural networks (PINNs) to model their degradation mechanisms. The project will combine experimental studies with advanced signal processing and AI techniques (compressed sensing), aiming to detect early signs of failure and enable predictive maintenance strategies executed locally (edge deployment).
The research will be carried out within a Marie Sklodowska-Curie Actions (MSCA) Doctoral Network, offering international training, interdisciplinary collaboration, and secondments at leading academic and industrial partners across Europe (Italy and Netherlands for this PhD offer).
On-line monitoring of bioproduction processes using 3D holographic imaging
The culture of adherent is a promising approach for various bioproduction applications, such as drug manufacturing and delivery, regenerative medicine, and tracking of cellular differentiation. However, the analysis of single cell morphology and behavior without affecting the substrate integrity remains a major challenge. Lens-free holographic imaging is emerging as a promising solution for real-time, non-invasive monitoring of cellular processes. This technique captures wide field of view images without requiring exogenous labeling or sample manipulation, thus preserving the integrity of the cellular environment.
This thesis proposes the development of a 3D lens-free imaging system to monitor adherents cells in near real-time. The microscope will be coupled with advanced algorithms for data reconstruction and analysis and tested on different cell models. The use of deep learning techniques will allow for real-time segmentation and analysis of single cells, facilitating the tracking of cellular dynamics. This innovative project paves the way to a non-invasive monitoring of 3D multicellular samples, with potential applications on organ-on-chip and more complex organoids systems.