Portail HAL ONERA
Not a member yet
12842 research outputs found
Sort by
Metallurgy at the nanoscale: domain walls in nanoalloys
No full textInternational audienceIn binary alloys, domain walls play a central role not only on the phase transitions but also on their physical properties and were at the heart of the 70's metallurgy research. Whereas it can be predicted, with simple physics arguments, that such domain walls cannot exist at the nanometer scale due to the typical lengths of the statistical fluctuations of the order parameter, here we show, with both experimental and numerical approaches how orientational domain walls are formed in CuAu nanoparticles binary model systems. We demonstrate that the formation of domains in larger NPs is driven by elastic strain relaxation which is not needed in smaller NPs where surface effects dominate. Finally, we show how the multivariants NPs tend to form an isotropic material through a continuous model of elasticity
Performance Assessment of Counter-Drone Systems Using Bayesian Networks
No full textInternational audienceThis paper proposes a method to model and analyze the performance of a Counter-Drone System (CDS) using Bayesian Networks (BN). Quantitative performance indexes related to the sensor and tracking algorithm used in the CDS are proposed. They are used in a BN which also accounts for CDS functions related to alert, localization and engagement of neutralization means. A case study is proposed to illustrate how performance of the CDS can be evaluated under various scenario conditions, including different types of drones and influences of the environment. To illustrate how BN can also help for design considerations of a CDS, influence of the sensor location is also analyzed
An Equivariant von Mises-Gaussian distribution on SE(n) for Unscented Kalman Filtering
No full textInternational audienceThe characterization and propagation of uncertainties on the Special Euclidean Lie groups SE(2) and SE(3) are crucial in robotics and state estimation. Applications such as navigation and SLAM require accurate modeling of pose uncertainty involving both position and attitude. Lie groups offer a structured state space that preserves system properties and improves consistency in nonlinear estimation. Kalman filters on Lie groups improve robustness but rely on a Gaussian assumption, which fails for large uncertainties. To overcome these limitations, we used an alternative probability density function, based on a maximum entropy criterion, leading to a filter that we call vMG-UKF-LG. The resulting method is validated on experimental datasets against four benchmark filters and indicates improved accuracy when dealing with complex trajectories and overestimating the process noise
Un solveur FFT efficace pour la dynamique des dislocations discrètes dans les métaux CFC
No full textInternational audienceDiscrete Dislocation Dynamics (DDD) is an established mesoscopic numerical method for simulating dislocation motion to determine the plastic behaviour of metals. However, DDD’s reliance on analytical expressions for internal stress fields limits its application to infinite isotropic media. The Discrete-Continuous Model (DCM), which couples DDD with a finite element elastic solver, has been proposed to handle more complex boundary conditions by regularising plastic strain using the eigenstrain formalism. However, this model still relies on analytical solutions for short-range dislocation interactions, making it unsuitable for anisotropic media. In this work, we present an improved DCM framework that replaces the FE solver with a Fast Fourier transform (FFT) solver for improved computational efficiency and full numerical calculation of stress fields, eliminating the need for analytical corrections. The proposed FFT solver employs a discrete theory of Green’s operators and uses a sharp eigenstrain field to describe dislocations. The solver mesh is aligned with the face-centred cubic (fcc) lattice of the DDD, forming an octahedral cell to address symmetry artefacts around {111} slip planes. Our FFT-based approach successfully maintains numerical stability by representing fcc dislocations as sharp fields without generating oscillations. This coupling allows the study of plasticity in anisotropic materials and interactions between dislocations and diffuse inclusions, such as precipitates, without short-range stress corrections.La dynamique des dislocations discrètes (DDD) est une méthode numérique mésoscopique bien établie pour simuler le mouvement des dislocations et déterminer le comportement plastique des métaux. Toutefois, elle repose généralement sur les expressions analytiques des champs de contraintes obtenues dans des milieux isotropes infinis, ce qui en limite l’application. Le modèle discret-continu (DCM), qui couple la DDD avec un solveur élastique par éléments finis (EF) grâce au formalisme des déformations libres, avait été proposé pour traiter des conditions aux limites plus complexes. Cependant, ce modèle utilise également des solutions analytiques pour les interactions à courte portée, ce qui le rend inadapté aux milieux anisotropes. Dans ce travail, nous présentons une version améliorée du DCM en remplaçant le solveur EF par un solveur utilisant les transformées de Fourier rapides (FFT), plus efficace et supprimant la nécessité des corrections analytiques à courte portée. Pour cela, ce solveur s’appuie sur trois ingrédients : (i) un champ de déformation propre abrupt (discontinu à l’échelle du maillage) pour décrire les dislocations ; (ii) un maillage CFC conforme à celui du code DDD sur lequel un “stencil” octaédrique est utilisé pour préserver les symétries des plans de glissement ; (iii) enfin un opérateur de Green discret. Ce solveur parvient ainsi à gérer des dislocations décrites par des champs abrupts sans générer d’oscillations. Couplé à la DDD, il permet d’étudier la plasticité dans les matériaux anisotropes et les interactions entre dislocations et des inclusions diffuses, sans corriger les contraintes aux cœurs des dislocations comme ce qui est fait habituellement
Rapporto finale del progetto LexiKHuM (ANR 20 CE33 0012)
No full textThe ANR project LexiKHUM (Lexicon for the kinesthetic communication human-machine) - coordinated by ISIR (Sorbonne Université, L. Saint-Bauzel) in collaboration with FoReLLIS (Université de Poitiers, C.S. Bianchini), LISN (Université Paris-Saclay, O. Grynszpan) and DTIS (ONERA, B. Berberian) – has lasted 4 years, from 2021 to 2025. This project brought together researchers in linguistics, cognitive sciences, ergonomics and robotics to study the units of kinesthetic meaning in human-human and human-robot interactions. The originality of the work lies in the analysis of gestural interactions between humans in order to inform the development of automatic systems (robots, algorithms). This approach is structured around two main axes: gesture analysis (description, meaning, epistemic properties) and message reception (meaning transmitted as well as emotions and levels of confidence it generates), with the idea that the attitude and state of mind in which an individual attempts to communicate gesturally are generalizable. This project produced several results: a) identification of common properties between sign language movements, co-verbal gesturing and kinesthetic communication, particularly for messages involving negotiation or uncertainty transmission; b) clarification of the concept of agency in joint interactions; and c) demonstration of the potential of kinesthetic communication to convey information in contexts of negotiation, uncertainty and navigation.Le projet ANR LexiKHUM ((Lexique de communication kinesthésique humain-machine) - coordonné par ISIR (Sorbonne Université, L. Saint-Bauzel) en collaboration avec FoReLLIS (Université de Poitiers, C.S. Bianchini), LISN (Université Paris-Saclay, O. Grynszpan) et DTIS (ONERA, B. Berberian) - a duré 4 ans, de 2021 à 2025. Ce projet a rassemblé chercheurs en linguistique, sciences cognitives, ergonomie et robotique afin d’étudier les unités de sens kinesthésique dans les interactions humain-humain et humain-robot. L’originalité du travail repose sur l’analyse des interactions gestuelles entre humains afin d’éclairer le développement de systèmes automatiques (robots, algorithmes). Cette démarche est structurée autour de deux axes principaux: l’analyse du geste (description, signification, propriétés épistémiques) et la réception du message (sens transmis ainsi que les émotions et niveaux de confiance qu’il suscite), dans l’idée que l’attitude et l’état d’esprit dans lesquels un individu tente de communiquer gestuellement sont généralisables. Ce projet a produit plusieurs résultats: a) identification des propriétés communes entre les mouvements de la langue des signes, de la gestualité co-verbale et la communication kinesthésique, en particulier pour les messages impliquant une négociation ou la transmission d’incertitude; b) précision du concept d’agentivité dans les interactions conjointes; et c) démonstration du potentiel de la communication kinesthésique pour transmettre des informations dans des contextes de négociation, d’incertitude et de navigation.Il progetto ANR LexiKHUM (Lessico per la comunicazione cinestetica uomo-macchina) - coordinato dall'ISIR (Sorbonne Université, L. Saint-Bauzel) in collaborazione con FoReLLIS (Université de Poitiers, C.S. Bianchini), LISN (Université Paris-Saclay, O. Grynszpan) e DTIS (ONERA, B. Berberian) - è durato 4 anni, dal 2021 al 2025. Questo progetto ha riunito ricercatori in linguistica, scienze cognitive, ergonomia e robotica per studiare le unità di significato cinestetico nelle interazioni uomo-uomo e uomo-robot. L'originalità del lavoro risiede nell'analisi delle interazioni gestuali tra esseri umani al fine di informare lo sviluppo di sistemi automatici (robot, algoritmi). Questo approccio è strutturato attorno a due assi principali: analisi del gesto (descrizione, significato, proprietà epistemiche) e ricezione del messaggio (significato trasmesso nonché emozioni e livelli di fiducia generati), nell'idea che l'atteggiamento e lo stato d'animo con cui un individuo tenta di comunicare gestualmente siano generalizzabili. Questo progetto ha prodotto diversi risultati: a) identificazione di proprietà comuni tra movimenti della lingua dei segni, gesti co-verbali e comunicazione cinestetica, in particolare per messaggi che implicano negoziazione o trasmissione di incertezza; b) chiarimento del concetto di agency nelle interazioni congiunte; e c) dimostrazione del potenziale della comunicazione cinestetica nel trasmettere informazioni in contesti di negoziazione, di incertezza e di navigazione
VAE pour la détection d'anomalies en imagerie SAR
No full textInternational audienceIn this paper, we propose an unsupervised learning approach for anomaly detection in SAR imaging. The proposed model combines a preprocessing despeckling step, a -\textit{annealing} Variational Auto-Encoder (VAE) for unsupervised anomaly filtering, and an anomaly detector based on the change of the covariance matrix at the input and output of the -\textit{annealing} VAE network. Experiments have been carried out on X-band ONERA polarimetric SAR images to demonstrate the effectiveness of Beta-Annealing VAE compared with the methods proposed in the literature.Dans ce papier, nous proposons une approche non supervisée pour la détection d'anomalies en imagerie SAR. Le modèle proposé combine une étape de \textit{despeckling} en pré-traitement, un Auto-Encodeur Variationnel (VAE) de type -\textit{annealing} pour filtrer les anomalies de manière non supervisé et un détecteur d'anomalies basé sur le changement de la matrice de covariance en entrée et en sortie du réseau de -\textit{annealing} VAE. Des expériences ont été menées sur des images polarimétriques SAR de l'ONERA en bande X pour démontrer l'efficacité du Beta-Annealing VAE par rapport aux méthodes proposées dans la littérature
Thermométrie optique couplée à la mesure d'autres quantités (vitesse, pression)
No full textInternational audienceThe visualization and quantification of thermal mixing in turbulent fluid flows is key to the development of accurate turbulence modelling. These models are needed to predict the behaviour of flows in industrially relevant applications as well as geophysical flows (ocean, atmosphere, earth's mantle, etc.). This talk will describe thermometry techniques based on luminescent tracer particles which can be combined with particle-based velocimetry to image both temperature and velocity in laboratory flows. The emphasis will be mostly on inorganic luminescent crystals, typically referred to as thermographic phosphors. These particles exhibit a wide range of luminescence properties and can be chosen to match the application needs, e.g., for use in cryogenic flows, at physiological temperatures, or up to 1000 K. There are also various implementations of this measurement concept. We can exploit the temperature dependence of the luminescence emission spectrum or the decay time of the particles to measure the temperature. The velocity can be measured simultaneously using temporally separated images of the particles using light scattering as in traditional Particle Image Velocimetry (PIV), using luminescence light or even single images of phosphorescence streaks caused by the motion of the particles during their luminescence decay.Recent developments include high-resolution measurements in submillimeter boundary layers, a proof-of-concept study of 3D temperature and velocity measurements in gas flows, and 2D thermometry in water with sub-°C precision.La visualisation et la quantification du mélange thermique dans les écoulements fluides turbulents sont essentielles au développement d'une modélisation précise de la turbulence. Ces modèles sont nécessaires pour prédire le comportement des écoulements dans les applications industrielles ainsi que les écoulements géophysiques (océan, atmosphère, manteau terrestre, etc.). Cet exposé décrira les techniques de thermométrie basées sur des particules traceuses luminescentes qui peuvent être combinées avec la vélocimétrie à base de particules pour obtenir des images de la température et de la vitesse dans les écoulements de laboratoire. L'accent sera mis principalement sur les cristaux luminescents inorganiques, généralement appelés luminophores thermographiques. Ces particules présentent une large gamme de propriétés de luminescence et peuvent être choisies en fonction des besoins de l'application, par exemple pour une utilisation dans des flux cryogéniques, à des températures physiologiques ou jusqu'à 1000 K. Il existe également diverses applications de ce concept de mesure. On peut exploiter la dépendance à la température du spectre d'émission de la luminescence ou le temps de décroissance des particules pour mesurer la température. La vitesse peut être mesurée simultanément à l'aide d'images séparées dans le temps des particules en utilisant la diffusion de la lumière comme dans la PIV traditionnelle, en utilisant la lumière de luminescence ou même des images uniques de traînées de phosphorescence causées par le mouvement des particules pendant la décroissance de la luminescence. Les développements récents comprennent des mesures à haute résolution dans des couches limites submillimétriques, une étude de validation de concept de mesures de température et de vitesse en 3D dans des écoulements gazeux, et la thermométrie 2D dans l'eau avec une précision de l'ordre de quelques degrés Celsius
Intelligence Artificielle pour la Combinaison Cohérente de Lasers : de la simulation à la réalité
No full textInternational audienc
A framework for automated construction of modular space platforms integrating structure constraints and automated planning
No full textInternational audienceLarge-scale space structures have taken center stage as the future of space exploration, paving the way for ambitious endeavors like sprawling solar power stations, intricate telescopes, and giant space stations. This program lies not in manual assembly, but in empowering space robots to autonomously build these structures using plans provided by dedicated on-orbit assembly planning algorithms.This paper presents a novel framework for autonomously assembling these structures using crawling mobile robots, guided by dedicated on-orbit planning algorithms. Our approach integrates the co-design of robots and satellite to ensure stable dynamics, focusing on deployable truss beams equipped with standard interfaces that support payloads like solar panels or antennas. By leveraging an automated planning algorithm that optimizes assembly time based on spacecraft actuator capabilities, and verifying plan execution through Finite Element Model (FEM) tools, our framework ensures both stability and accuracy. This integrated approach enables effective co-design, streamlines deployment through optimized planning, and ultimately enhances reliability by minimizing risks associated with structural integrity during assembly, resulting in a more efficient and robust space structure assembly process
Introducing an initial contrail shape parameterization based on wake vortex dynamics simulation into pycontrails: a study of its influence on radiative forcing
No full textInternational audiencePredicting contrail formation and evolution is crucial for assessing aviation's impact on cloudiness and radiative balance. The CoCip method, implemented in pycontrails, simulates contrail formation using gaussian plume distributions. However, studies show that actual plumes rapidly deviate from this assumption shortly after emission. We investigate the optical impact of non-gaussian plume distributions at the end of the vortex phase and quantify errors incurred by assuming gaussian distributions under various atmospheric conditions, with the ultimate goal of enabling a more refined distribution modeling framework