Archive ouverte de Centrale Lyon
Not a member yet
32420 research outputs found
Sort by
Micro-Transfer-Printing integration of interband cascade lasers on Si
No full textInternational audienceThis study presents the technological development for the integration of an Interband Cascade Laser (ICL) source emitting at 4 µm onto a silicon platform, achieved through the Micro-Transfer Printing (µTP) technique
Cosmological implications of the Gaia Milky Way declining rotation curve
No full textInternational audienceAlthough the existence of dark matter has been widely acknowledged in the cosmology community, its nature remains unknown despite decades of research. This calls into questions its very existence. This never-ending search for dark matter has led us to consider alternatives. Since increasing the enclosed mass is the only way to explain the flat appearance of galaxies' rotation curves in a Newtonian framework, modified Newtonian dynamics (MOND) theory proposes a modification to Newton's dynamics when the acceleration of matter at given radius is around or below a threshold value, a 0 . Observed rotation curves, generally flat at large distances, are then usually well reproduced by MOND with an a 0 of ∼1.2 × 10 -10 m/s 2 . However, recent Gaia evidence indicates a decline in the Milky Way rotation curve, which a distinct behavior. Therefore, we examine whether MOND can accommodate the Gaia declining rotation curve of the Milky Way. We first use a standard model to describe the Milky Way's baryonic components.Secondly, we show that a Navarro, Frenk, and White model is able to fit the decline, assuming a scale radius (R s ) on the order of 4 kpc. In a third step, we show that the usual MOND paradigm is not able to reproduce the declining part in a standard baryonic model. Finally, we examine whether the MOND theory can accommodate the declining part of the rotation curve when the characteristics of the baryonic components are relaxed. To do so, we used a Markov chain Monte Carlo method to determine the characteristics of the stellar and HI disks, including their masses, that would result in the best possible fit of the decline. We find that the stellar disk should be massive, on the order of 10 11 M . The HI disk mass is capped at nearly 1.8 × 10 11 M but could also be negligible. Finally, a 0 is consistent with 0, with an upper limit of 0.53 × 10 -10 m/s 2 (95%), a value much lower than that usually advocated to explain standard flat rotation curves in MOND theory (∼1.2 × 10 -10 m/s 2 )
Correlation between surface morphology, superhydrophobicity, and icephobic behavior: a principal component analysis study
No full textInternational audienceThis study aimed to investigate the relationships between topographical parameters, wetting properties, and the icephobic behavior of textured surfaces. Principal component analysis (PCA) of the wetting data revealed a strong correlation between the slope of the patterns and the contact angle. However, because wetting states are not always reported with sufficient explanation, they were determined using the Extrand model. In some cases, discrepancies with experimental results were observed, likely due to metastable states caused by droplet pinning on micro-roughness. A separate PCA on icing delay time (IDT) data indicated a correlation between slope, contact angle, and IDT, without requiring prior knowledge of the wetting state. In contrast, PCA on ice adhesion data revealed no clear correlation, highlighting a lack of standardization in the field that hinders comparisons. To refine the results, another PCA was conducted using the Extrand-model-predicted wetting states, emphasizing the importance of accurately identifying the wetting regime to understand surface behavior. Given the challenges in correlating adhesion measurements, IDT emerges as a more relevant and accessible initial metric for evaluating icephobic properties. Finally, pattern slope appears to be a key topographical parameter. While other parameters showed no direct correlations, appropriate feature height and limited spacing are still critical to prevent complete droplet penetration
How the Structure and Wettability Properties of Morpho peleides Butterfly Wings Can Be a Source of Inspiration
No full textInternational audienceIn this study, the wettability of Morpho peleides (M. peleides) wings is studied. Using a goniometer, the contact angles of ~136° were measured. Although few studies have provided a general overview of the static wettability properties of M. peleides and the particularly anisotropic morphology of their wings, a detailed analysis of wettability properties is proposed. The results indicate anisotropic wettability, with a sliding angle of 7° when the wings were tilted away from the insect’s body and 29° when the wings were tilted toward the insect’s body. The Extrand model, coupled with a hierarchical approach, was also employed to describe the wettability behavior of M. peleides wings and its relationship with the dual roughness scale of the wings. Owing to these models, it has been demonstrated that the micrometric scale of the wing structure is primarily responsible for the static wettability properties of the M. peleides, while the nanometric scale influences the dynamic wettability of the wing. Moreover, compared to the topography, the wing’s chemistry has very little effect on the wettability properties of the M. peleides
A note on stable Kim-forking
No full textWe define weak stable Kim-forking, a notion that generalizes stable forking to the context of NSOP1 theories. We adapt some of the known results on stable forking to this context
Termination of Graph Rewriting using Weighted Type Graphs over Non-well-founded Semirings
No full textInternational audienceThe weighted type graph method is a technique for proving termination of double-pushout (DPO)graph rewriting systems. While existing approaches focus on weighted type graphs over well-founded semirings, we introduce an extension to non-well-founded semirings. To demonstrate the ef-fectiveness of our approach, we implemented both our method and a prior approach for edge-labeledgraph rewriting within a unified tool, enabling direct comparison
Entraînement informé par solveur pour l'intégration de contraintes logiques dans l'extraction de relations d'événements
No full textNational audienceL'extraction de relations d'événements (ERE) est une tâche cruciale dans le traitement du langage naturel, impliquant l'identification et la classification des relations sémantiques entre les événements décrits dans des documents textuels. Malgré les avancées récentes grâce aux approches d'extraction conjointe, les modèles actuels rencontrent encore des défis importants, notamment une précision insuffisante dans l'extraction des relations sous-représentées mais essentielles (telles que la causalité) et d'importantes incohérences logiques parmi les relations prédites. Pour remédier à ces limitations, nous proposons un framework pour l'ERE, conçu explicitement pour améliorer la performance d'extraction et assurer la cohérence logique globale. Notre approche combine l'encodage de documents basé sur les transformateurs avec un solveur de contraintes logiques dédié qui corrige systématiquement les prédictions brutes pour garantir la cohérence dans toutes les relations d'événements extraites. Nous introduisons le concept de relations fondamentales, un sous-ensemble de relations essentielles pour préserver la cohérence logique, et nous utilisons une stratégie d'entraînement consciente du solveur afin de prioriser explicitement ces relations. Des expérimentations approfondies sur l'ensemble de données complet MAVEN-ERE démontrent que notre framework obtient pas une précision d'extraction supérieures par rapport aux méthodes d'extraction conjointe existantes.</div
Un outil conversationnel basé sur un graphe de connaissances, des LLM et un modèle BERT pour les programmes d'alternance en France
No full textNational audienceLe suivi efficace de l'acquisition des compétences dans les programmes de l'alternance, présente des défis importants pour la technologie éducative. Cet article présente un nouvel agent conversationnel intégré dans un livret de formation numérique qui relève ces défis grâce à une architecture multimodale. Notre système intègre (1) un graphe de connaissances spécifique à un domaine, lié à des référentiels de compétences, (2) des grands modèles de langage (LLM) et (3) un composant génératif basé sur BERT. Cette approche hybride permet à la fois une représentation structurée des trajectoires d'apprentissage et des capacités d'interaction en langage naturel, ce qui permet un suivi nuancé des progrès et des interventions personnalisées. L'évaluation empirique démontre que le système fournit un retour d'information contextuellement pertinent qui s'adapte aux modèles d'apprentissage individuels, ce qui permet une acquisition plus efficace des compétences
NUMERICAL ASSESSMENT OF A MODE DETECTION ARRAY FOR AZIMUTHAL DECOMPOSITION
No full textInternational audienceThe efficiency of a mode detection array for azimuthal mode decomposition in turbofan nacelle experiments is numerically investigated by analyzing acoustic solutions computed using the commercial finite-element code AC-TRANTM. First, irrotational inviscid mean flow fields are computed for sideline conditions, with and without an external flow at a Mach number of 0.2. Then, the upstream propagation of duct modes representing fan noise components is performed for three frequencies by solving the Möhring equation in the frequency domain using ACTRANTM. The acoustic solutions are interpolated to the microphone locations of the mode detection array. In all cases, the azimuthal structure of the interpolated data is obtained using Fourier transforms. It is compared to a reference one, determined by a duct mode decomposition of the full acoustic solution. The results provide guidelines for the analysis of ongoing experiments. They indicate that the detection antenna will be able to efficiently capture the azimuthal mode distributions, but not the radial structure. The procedure can be applied to design improved microphone arrays by defining judicious axial and radial positions for azimuthal and radial mode decompositions. Moreover, the approach can determine the optimum discretization for azimuthal mode decomposition
Sharp exponent of acceleration in general non-local equations with a weak Allee effect
No full textInternational audienceWe study an acceleration phenomenon arising in monostable integro-differential equations with a weak Alleeeffect. Previous works have shown its occurrence and have given correct upper bounds on the rate of expansionin some particular cases, but precise lower bounds were still missing. In this paper, we provide a sharp lowerbound for this acceleration rate, which is valid for a large class of dispersion operators. Our results manage tocover fractional Laplace operators and standard convolutions in a unified way, which is new in the literature. Animportant result of the paper is a general flattening estimate of independent interest: this phenomenon appearsregularly in acceleration situations, but getting quantitative estimates is in general an open question. With thisestimate at hand, we construct a subsolution that captures the expected behaviour of the accelerating solution(rates of expansion and flattening) and identifies several regimes that appear in the dynamics depending on theparameters of the problem