Les séminaires de l'UMET

To better apprehend the improved mechanical properties of highly structured multi-nanolayer films obtained by co-extrusion, as well as identifying the key parameters governing their structure-morphology, the present study focuses on model films made of low-density polyethylene (LDPE) and polystyrene (PS). Despite being two immiscible polymers with highly-mismatched rheological behavior, coextrusion by forced-layer-assembly allows to prepare well-controlled multilayer films (without the need of compatibilizers) with different levels of confinement through the fine tuning of their number of layers. More specifically, the layer confinement of a semi-crystalline polymer, here LDPE, by a glassy amorphous polymer, here PS, was investigated. In comparison to each of the two base materials, the obtained multilayer architecture brings out the best of the two worlds by combining the ductility of LDPE with the high strength of PS, greatly outperforming the corresponding 50/50 randomly structured blend. Interestingly, we have demonstrated that increasing the number of layers of PS/LDPE system (while remaining in the processing window of stable continuous layers), thus increasing the level of layer confinement, allows to stabilize the brittle damage mechanisms (i.e. crazes) of PS, enabling to reach large strains without macroscopic failure. In parallel to the multi-layer morphology characterization by SEM and TEM, the orientation and crystalline structure of the coextruded films were also characterized by WAXS and DSC. The geometrical confinement of LDPE nanolayer did not affect the thermal properties of LDPE crystalline phase, but it did affect its crystalline morphology, evolving from an isotropic-spherulitic shape to a more lamellae-oriented form.

Rock deformation and failure are governed by stress and strain heterogeneities that emerge at the grain scale and evolve across length scales, ultimately controlling macroscopic strength and rupture processes. Capturing these internal processes in situ, while a sample is under mechanical loading, remains a major challenge in both geoscience and materials science. High-energy synchrotron-based techniques such as three-dimensional X-ray diffraction (3DXRD) and its scanning variant (scanning-3DXRD) provide a unique route to address this problem by non-destructively probing bulk polycrystalline materials and reconstructing the crystallographic orientation and elastic strain of thousands of individual grains embedded within a three-dimensional volume.
This seminar will present recent methodological developments at the ESRF in 3DXRD and pencil-beam scanning-3DXRD, with a focus on their ability to resolve spatially heterogeneous stress fields and microstructural evolution during deformation. Applications to rock mechanics spans from operando experiments under quasi-static triaxial loading — revealing how stress builds up, localizes, and redistributes prior to failure — to post-mortem investigations of microstructures and heterogeneous residual strain fields in natural and experimentally deformed rocks. By delivering grain-resolved measurements of internal stress in deforming rocks at unprecedented resolution, these emerging approaches open new experimental avenues to explore the microphysical mechanisms governing deformation, rupture, and chemical reactions in rocks and other polycrystalline materials.

We propose a novel phase-field (PF) model to enhance the description of grain bound-
aries (GBs) and its effect on solute segregation behaviour under irradiation. Conven-
tional PF models typically treat GBs as perfect sinks for point defects (PDs) such as
vacancies and interstitials, often assuming the system’s chemical potentials as homo-
geneous. We present two different approaches to tackle this limitation: (i) the use of a
density model to take into account the reduction in atomic density within GBs and cal-
culate the induced GB thermodynamic and elastic properties and (ii) the representation
of low-angle symmetric tilt grain boundaries (STGBs) by a stack of dislocations.
Furthermore, we introduce a mixing term in the model to account for ballistic dam-
age induced by displacement cascades. This study demonstrates how our model featur-
ing the density function correctly predicts equilibrium segregation and its impact on
radiation-induced segregation (RIS) in Fe-Cr, Ni-Cr, Ni-Fe and Ni-Ti alloys. These
alloys are serious candidates for Gen IV nuclear reactors. For each of these systems,
the influence of a large number of parameters such as the nominal composition, tem-
perature and dose rate has been thoroughly investigated in order to quantify the con-
tribution of thermal equilibrium segregation compared to RIS. More particularly, our
methodology successfully reproduces the well-known “W-shape” segregation profiles
in both GB descriptions and provides insights into spinodal decomposition in Fe-Cr
alloys and ballistic mixing effects on GB segregation.
This advanced PF model offers a better understanding of GBs behaviour under irra-
diation, potentially contributing to improved material design for nuclear applications.

Le projet ANR Jeune Chercheur PhaMMAT (« PHAse-field Multiscale integrated simulation of Microstructure And Thermomechanical evolution of nuclear fuels »), débuté en novembre 2025 et d'une durée de quatre ans, a pour objectif le développement d'un outil de simulation multiéchelle dédié à l'évolution couplée de la microstructure et des champs thermomécaniques dans les matériaux inorganiques. Ce projet vise spécifiquement à permettre la modélisation des phénomènes de restructuration dans les combustibles nucléaires comme application de référence. Pour ce faire, le code de simulation INFERNO fondé sur la méthode de champ de phase est développé afin de capturer l'évolution complexe de la microstructure polycristalline de l'oxyde d'uranium à l'échelle mésoscopique, en intégrant les défauts, les bulles de gaz de fission et les dislocations sous l'influence des contraintes thermomécaniques. L'originalité de l'approche repose sur un couplage autocohérent où, à chaque pas de temps, le modèle de champ de phase interagit avec un code résolvant les équations de la mécanique pour actualiser les champs d'entrée. Le modèle physique s’appuie sur des données issues de simulations à plus basses échelles et s'applique à des structures polycristallines 3D réalistes, générées par le code MEROPE. Ces travaux permettront de mieux comprendre l'interaction entre le réseau de bulles de gaz de fission et l'apparition de sous-joints de grains à fort taux de combustion, des phénomènes cruciaux pour les propriétés du combustible sous irradiation, afin de fournir des modèles prédictifs aux codes de performance. Enfin, ce cadre numérique transverse offre des perspectives prometteuses pour d'autres matériaux de l'énergie, tels que les batteries à l'état solide ou les céramiques fonctionnelles, présentant des problématiques microstructurales similaires.

The ANR Jeune Chercheur PhaMMAT project (“PHAse-field Multiscale integrated simulation of Microstructure And Thermomechanical evolution of nuclear fuels”), which started in November 2025 and runs for four years, aims to develop a multiscale simulation tool dedicated to the coupled evolution of microstructure and thermomechanical fields in inorganic materials. The project specifically seeks to enable the modeling of restructuring phenomena in nuclear fuels as a reference application. To this end, the INFERNO simulation code, based on the phase-field method, is being developed to capture the complex evolution of the polycrystalline microstructure of uranium dioxide at the mesoscale, incorporating defects, fission gas bubbles, and dislocations under the influence of thermomechanical stresses. The originality of the approach lies in a self-consistent coupling in which, at each time step, the phase-field model interacts with a code solving the equations of mechanics in order to update the input fields. The physical model relies on data derived from lower-scale simulations and is applied to realistic three-dimensional polycrystalline structures generated by the MEROPE code. This work will improve the understanding of the interaction between the fission gas bubble network and the formation of high-burnup subgrain boundaries—phenomena that are crucial for fuel properties under irradiation—in order to provide predictive models for fuel performance codes. Finally, this transverse numerical framework offers promising perspectives for other energy-related materials, such as solid-state batteries or functional ceramics, which exhibit similar microstructural challenges.

Ce projet de recherche s’inscrit dans le développement d’une nouvelle génération d’élastomères
thermoplastiques (TPE) présentant des propriétés thermomécaniques modulables et évolutives,
répondant à la demande croissante en matériaux polymères à la fois durables et performants. Les TPE
se distinguent par leur capacité à combiner l’élasticité des caoutchoucs avec la transformabilité et la
recyclabilité des thermoplastiques. Traditionnellement conçus à partir de copolymères di- ou triblocs,
ils reposent sur un bloc souple, responsable de l’élasticité, et un ou deux blocs rigides, assurant la
stabilité thermique et la rigidité.
Notre approche innovante consiste en la conception de copolymères triblocs fonctionnalisés par des
groupements latéraux permettant des réticulations indépendantes, covalentes et réversibles. Le bloc
rigide est constitué de polyméthacrylate de méthyle (PMMA) fonctionnalisé par des unités furannes et
maléimides protégées, tandis que le bloc souple est formé de polycaprolactone (PCL) focntionnalisé par
des groupements anthracènes. Cette architecture offre la possibilité d’activer des réticulations
orthogonales : réticulation thermique via la réaction de Diels-Alder réversible pour le bloc rigide, et
réticulation photo-induite sous irradiations UV pour le bloc souple. Ce système à double stimulus facilite
non seulement l’ajustement local et sélectif des propriétés, mais permet également la conception de
matériaux à gradient de propriétés. Par ailleurs, les extrémités des blocs rigides et souples ont été
fonctionnalisées par des ligands 2,6-bis(1′-méthylbenzimidazol-2′-yl)pyridine (Mebip) en extrémités de
chaines, permettant leur auto-assemblage via la formation de complexes métal-ligand (MLC) (Figure1).
Cette approche de chimie métallo-supramoléculaire confère une flexibilité significative dans la
conception macromoléculaire, tout en introduisant un aspect thermosensible puisque les complexes
métal-ligands peuvent être assemblés ou désassemblés par chauffage, dotant ainsi les matériaux d’une
capacité de recyclabilité.

Membres du jury :

Pr. Benjamin Le Droumaguet, Université Paris-Est Créteil, ICMPE 
Pr. Blanca Martin-Vaca, Université de Toulouse, LHFA
Pr. Philippe Zinck, Université de Lille, UCCS
Dr. Vincent Lapinte, Université de Montpellier, ICGM
Dr. Fanny Coumes, Sorbonne Université, IPCM
Dr. David Fournier, Université de Lille, UMET
Dr. Grégory Stoclet, Université de Lille, UMET

Mars likely experienced a global magma ocean during its early evolution, strongly influencing the planet’s cooling history, crystallization processes, and mantle differentiation. Recent seismic observations from NASA’s InSight mission suggest the possible presence of a molten or partially molten silicate layer at the core–mantle boundary (CMB), raising the question of whether this basal melt layer could represent a long-lived remnant of the primordial magma ocean. Addressing this issue requires direct constraints on the physical properties of Fe-rich Martian melts—including viscosity, density, and seismic velocities—as well as on the melting relations and phase equilibria governing deep-mantle behaviour.

To investigate these processes, we conducted high-pressure and high-temperature experiments combining ex situ measurements at the Bayerisches Geoinstitut with in situ experiments at SPring-8 and DESY. Our results show that deep Martian melts are highly FeO-rich, exhibit low viscosities, and can significantly reduce seismic velocities even at very small melt fractions. Complementary melting experiments at ~19 GPa further constrain melting behaviour, melt chemistry, and phase relations expected near the Martian CMB.

Together, these measurements provide an experimental framework to evaluate the formation, stability, and seismic detectability of a basal melt layer, and to link the dynamics of Mars’ early magma ocean to the present-day structure of its interior.