Portail HAL Ensta
Not a member yet
11080 research outputs found
Sort by
Water impact of V-shaped wedge at constant velocity: Influence of deadrise angle and stiffness panels
No full textInternational audienc
L'effet de la réalité augmentée sur la mémoire autobiographique involontaire
No full textInternational audienceWe know little about the impact of augmented reality (AR) on human cognition, particularly regarding involuntary autobiographical memory (IAM). IAMs are spontaneous recollections of personal events, ubiquitous in daily life but under-researched in both psychology and human-computer interaction. We first discuss the potential opportunities and risks of replacing conventional displays with AR to increase the likelihood of IAMs. We then report on a study investigating whether stimuli displayed on the same mobile device using video-see-through AR are more likely to resurface than those shown with a simple 3D viewer. We found that AR elicits approximately twice as many IAMs in controlled settings with immediate re-exposure to contextual cues, but no measurable effect was found in everyday settings with delayed re-exposure. Therefore, AR can enhance IAMs, but its effects may be modest and short-lived in most cases. Nevertheless, future studies could reveal stronger effects of AR in other settings. Supplementary material is available at https://osf.io/gscpv/.Nous savons peu de choses sur l'impact de la réalité augmentée (RA) sur la cognition humaine, en particulier en ce qui concerne la mémoire autobiographique involontaire (MAI). Les MAI sont des souvenirs spontanés d'événements personnels, omniprésents dans la vie quotidienne mais peu étudiés en psychologie et dans le domaine de l'interaction homme-machine. Nous examinons tout d'abord les possibilités et les risques potentiels liés au remplacement des affichages conventionnels par la RA afin d'augmenter la probabilité des MIA. Nous présentons ensuite une étude visant à déterminer si les stimuli affichés sur le même appareil mobile à l'aide de la RA vidéo sont plus susceptibles de refaire surface que ceux affichés à l'aide d'une simple visionneuse 3D. Nous avons constaté que la RA suscite environ deux fois plus de MIA dans des environnements contrôlés avec réexposition immédiate aux indices contextuels, mais aucun effet mesurable n'a été constaté dans des environnements quotidiens avec réexposition différée. Par conséquent, la RA peut renforcer les MAI, mais ses effets peuvent être modestes et de courte durée dans la plupart des cas. Néanmoins, des études futures pourraient révéler des effets plus importants de la RA dans d'autres contextes. Du matériel supplémentaire est disponible à https://osf.io/gscpv/
Reduced-order modeling for nonlinear vibrations of structures
No full textInternational audienceThis chapter is devoted to the presentation of model-order reduction techniques that are used in the field of structural vibrations. A special emphasis is placed on substructuring methods for localized nonlinearities and on nonlinear normal modes defined via invariant manifolds for distributed smooth nonlinearities, as key tools to perform efficient yet accurate dimensional reductions. Other reduction techniques such as proper orthogonal decomposition, implicit condensation, and modal derivatives are also briefly covered at the end of the survey. The contents of this chapter were written for the Handbook of Nonlinear Dynamics during the Summer of 2024
Wormhole formation in fluid-driven granular flow
No full textInternational audienceFluid-driven flow of granular material leads to complex behaviour and emergent instabilities in many natural and industrial settings. However, the effect of using fluid flow to vertically drive a dense bed of sedimenting grains is not well documented. Here we find contrasting behaviours in a submerged fluid-driven silo, including fingering patterns, porous flow, classical silo flow, and the formation of straight, semi-dilute wormhole-like channels. Once formed, these channels rapidly propagate towards the outlet and act as a bypass of the wider packing. The onset of this instability occurs when the gravity-driven grain flow at the free surface is insufficient to supply the fluid-assisted central region below the interface. Balancing empirical models of these flows predicts the height at which channels emerge as a function of grain size and flow rate. These findings provide a framework for predicting and controlling fluid-grain interactions in natural hazards, industrial processing, and geophysical flows
Navigation événementielle par cycles pour robots mobiles
No full textThis thesis proposes a new paradigm for autonomous underwater navigation in environments where traditional localization systems (GNSS, inertial, acoustic) are unavailable or unreliable. Inspired by biological behaviors, the approach relies on the use of stable cycles as a navigation principle, guiding the robot along periodic trajectories defined by a timed automaton. These trajectories are stabilized using a minimal set of sensors that trigger events in the automaton, enabling navigation without absolute localization. The theoretical analysis, based on control theory, set-based methods, and the notions of positively invariant sets and attraction basins, provides guarantees of both local and global stability. A cycle transition mechanism allows for extended navigation in complex environments. Validated through simulations and real-world experiments on a physical robot, this approach demonstrates robust and energy-efficient performance, paving the way for new navigation strategies for autonomous marine systems.Cette thèse propose un nouveau paradigme de navigation autonome sous-marine pour les environnements où les systèmes de localisation classiques (GNSS, inertiel, acoustique) sont indisponibles ou peu fiables. Inspirée des comportements biologiques, l'approche repose sur l'utilisation de cycles stables comme principe de navigation, guidant le robot le long de trajectoires périodiques définies par un automate temporisé. Ces trajectoires sont stabilisées à l'aide d'un ensemble minimal de capteurs déclanchant des évènements dans l'automate, permettant de naviguer sans localisation absolue. L'analyse théorique, fondée sur la théorie du contrôle, des outils ensemblistes et des notions d'ensemble positivement invariant et de bassin d'attraction, établit des garanties de stabilité locale et globale. Un mécanisme de transition entre cycles permet une navigation étendue dans des environnements complexes. Validée par simulations et essais sur un vrai robot, cette approche démontre des performances robustes et sobres en énergie, ouvrant la voie à de nouvelles stratégies de navigation pour les systèmes marins autonomes
Stable Cycles for Underwater Navigation
No full textInternational audienceNavigating GNSS-denied environments presents significant challenges. This is one of the challenges faced by underwater robotics. Underwater localization solutions are still expensive and difficult to integrate into small robots. This paper proposes to take advantage of stable cycles to enable navigation without getting lost. The robot is following a timed automaton that guides it along a predefined trajectory. By tuning the timings of the automaton based on measurements, the cycle is moved and stabilized in the environment. Several of these cycles may exist in areas to be explored. The goal of this paper is to provide strategies to navigate between these cycles, either using dead-reckoning for short-range navigation or using an isobath bounce strategy to reach the capture basin of the next cycle and to be sure to stabilize the trajectory of the robot on the next cycle
High-velocity laser-driven flyer impact on paraffin gel
No full textInternational audienceThe penetrating ballistic impact of thin 100 micrometers aluminum projectiles, accelerated at high velocities by laser-induced shock wave, on parafin gel is investigated. The laser-driven flyer experiments were conducted at BELENOS laser facility and allow the acceleration of projectile at high velocity ranging from 170 m s to 710 m s. The projectile is monitored during impact and penetration into gel targets using shadowgraphy with ultra-high speed camera. Its velocity is recorded by fast-imaging technics and correlated to Photonic Doppler Velocimeter (PDV) measurements. The ballistic impact phenomena such as the splash ejection on the front face of the gel target and the cavitation effect are analyzed. The strength resistance parameter in the Poncelet model of the gel is obtained from experimental data fit, which predicts the speed of a given fragment from its penetration depth in the target. The cavity dynamics highlights the influence of the strain rate on the mechanical behavior of paraffin gel target under penetrating ballistic impact
Évaluation combinée de la ténacité et de la performance de réparation dans les stratifiés unidirectionnels CF/PPS sous chargement statique et cyclique.
No full textInternational audienceÉvaluation combinée de la ténacité et de la performance de réparation dans les stratifiés unidirectionnels CF/PPS sous chargement statique et cyclique
Machine Learning Predictions for the Comparative Mechanical Analysis of Composite Laminates with Various Fibers
No full textInternational audienceThis article addresses the complex behavior of composite laminates under varied layer orientations during tensile tests, focusing on carbon fiber and epoxy matrix composites. Data characterizing the mechanical load behavior are obtained using twelve composite laminates with different layer orientations and the DIGIMAT-VA software (version 2023.3). First, these data were used to elaborate a complex comparative analysis of composite laminates from the perspective of materials science. Composite laminates belong to three classes: unidirectional, off-axis oriented, and symmetrically balanced laminates, each having a specific behavior. From the perspective of designing a new material, a prediction model that is faster than the finite element analysis is needed to apply this comparative analysis’s conclusions. As a novelty, this paper introduces several machine learning prediction models for composite laminates with 16 layers arranged in different orientations. The Regression Neural Network model performs best, effectively replacing expensive tensile test simulations and ensuring good statistics (RMSE = 34.385, R2 = 1, MAE = 19.829). The simulation time decreases from 34.5 s (in the case of finite element) to 0.6 s. The prediction model returns the stress–strain characteristic of the elastic zone given the new layer orientations. These models were implemented in the MATLAB system 2024, and their running proved good models’ generalization power and accuracy. Even specimens with randomly oriented layers were successfully tested
Experimental study and modelling of shock-induced compaction of autoclaved aerated concrete through MHz X-ray radioscopy
No full textInternational audienceExtreme scenarios related to sudden increase of stress states, such as mechanical impacts or collisions, can lead to severe physical damage on structures. However, through their compaction phase, porous materials absorb a part of mechanical energy and efficiently mitigate the shock wave induced damage. Autoclaved Aerated Concrete (AAC) is one type of cellular materials which exhibits such capability and also being non-flammable, which is of interest for most structures that must resist both impacts and fire. With good insulation properties and its low density, AAC is therefore a material of choice in protective design against shock loading. It is however difficult to study in real time, especially because of the cloud of dust produced during compaction. In the present study, two AAC of respective densities of 550 kg/m3 and 115 kg/m3 are considered. Plate impact tests were performed at the European Synchrotron Radiation Facility (ESRF) and the compaction process was observed in-situ by ultra-fast X-ray phase-contrast radioscopy for impact velocities ranging from 250 to 400 m/s. Through the records analysis, a compaction front is identified..The tracking of the compaction front and the initial velocity of the projectile provide a portion of the AAC compacted state. In line with the findings, an extended analysis extracting the mean pore size and the evolution of the densities and speed of sounds is conducted by the use of laser induced shock waves on samples pre-compacted. An analytical model is proposed to reproduce the compaction front dynamics, considering equivalent mass-spring system