HAL-MINES ParisTech
Not a member yet
27348 research outputs found
Sort by
Non-Locally Controllable but Trackable Magnetic Head Flagellated Swimmer
No full textUnlike macroscopic swimmers, microswimmers operate in a low-Reynolds-number regime dominated by viscous forces. This paper investigates the controllability of a magnetic microswimmer composed of a spherical magnetic head and an elastic, non-magnetic flagellum. The swimmer evolves in a Stokes flow and is modeled using the resistive force theory. We prove that, under planar motion, the system is not small-time locally controllable and numerically identify regions that remain inaccessible. Nevertheless, simulations show that trajectory tracking can still be achieved via Bayesian optimization, though it requires large-amplitude transverse deformations
Metallurgically-driven thermomechanical analysis of multiple side-to-side laser melting on a 316L substrate
No full textInternational audienceIn additive manufacturing, the solidification grain structure has a significant influence on the properties of as-built material. In this context, the solidifi-cation grain structure and internal stress evolution during laser scanning of polycrystalline 316L stainless steel are simulated. A strongly coupled crystal viscoplasticity model is developed and integrated with a cellular automaton finite element (CAFE) approach to accurately capture grain structure and stress evolution, where the CAFE model is validated based on a literature experiment. The crystal viscoplasticity model is calibrated using stress-strain curves of annealed 316L from experiments considering small thermo-elasto-viscoplastic (TEVP) deformations. The resolution algorithm dynamically couples heat transfer, melting and solidification simulations while concurrently computing stress and strain evolution within the grain structure. Four scanning strategies are simulated using the coupled CAFE-crystal viscoplas-ticity approach, capturing stress evolution during grain growth. This enables the simultaneous thermo-viscoplastic modeling in the mushy zone and TEVP modeling in the solid, providing insights into stress evolution and grain orientation over a large domain. The melting-solidification process involves variations in compression and tension, leading to stress concentration within neighboring grains with significant orientation differences, extending along elongated grains. A framework for multiscale process-structure-mechanical investigation is established based on microscale stress evolution in additive manufacturing
ARCHITECTURE STUDIES AND OPTIMISATION OF A TWO-PHASE EXPANSION TURBINE
No full textInternational audienceThis work aims to compare the performance of several two-phase impulse turbine architectures. Radial and axial architectures will be modeled for refrigeration application. A 1D velocity triangle model is used to carry out a pre-design of the wheel. This model will be identical for all architectures. Then, using a Computational Fluid Dynamics (CFD) code, various architectures are modeled including for example axial and radial ones. Besides, the effect of the rotational speed are also analyzed.This preliminary work aims to explore the loss mechanisms of the studied architectures. It has been for example observed for radial architectures that a secondary fluid flow accumulates at the periphery of the wheel producing substantial losses. The study results in a pre-selection of the architecture with the best potential, which will then be optimised, particularly with regard to the blades.</div
Circuits quantiques supraconducteurs fortement entraînés et transfert d’état vers une mémoire à spins
No full textScaling up quantum processors to run practical quantum algorithms remains a major challenge due to hardware constraintsand control limitations. This dissertation explores two complementary directions to address the scalability bottlenecksin superconducting quantum architectures: the detrimental impact of strong drives required for fast and high-fidelityoperations, and the integration of quantum memories into hybrid architectures to reduce the hardware overhead.In the first part, we investigate how strong drives can generate detrimental effects in the circuit. In this regime, transmonscan be structurally unstable. Due to chaotic effects, the computational manifold is no longer well separated from theremainder of the spectrum, which correlates with enhanced offset-charge sensitivity and destructive effects in readout. Weshow here that these detrimental effects can further propagate to other degrees of freedom, for example to neighbouringqubits in a multi-qubit system. Specifically, a coherently driven transmon can act as a source of incoherent noise toanother circuit element coupled to it. By using a full quantum model and a semiclassical analysis, we perform the noisespectroscopy of the driven transmon coupled to a spectator two-level system (TLS), and we show that, in a certainlimit, the interaction with the driven transmon can be modelled as a stochastic diffusive process driving the TLS, therebyrevealing a fundamental source of drive-induced crosstalk in multi-qubit setups.In the second part, we turn to a modular approach for scalable quantum computation, which combines the processingunit with a storage module. We focus on spin-based quantum memories, where itinerant electromagnetic fields arestored in large ensembles of effective two-level systems embedded in a cavity. Using a mean-field framework, we modelthe ensemble as an effective spin communication channel and develop a cascaded quantum model to describe bothabsorption and emission processes. We derive optimal time-dependent modulations of the cavity linewidth that maximizestorage and retrieval efficiency for finite-duration wavepackets. Our analysis yields an upper bound on efficiency, whichcan be met in the narrow bandwidth regime. It also shows the existence of a critical bandwidth above which the efficiencyseverely decreases.Passage à l’échelle des processeurs quantiques pour exécuter des algorithmes quantiques pratiques demeure undéfi majeur, en raison des contraintes matérielles et des limitations de contrôle. Cette thèse explore deux axescomplémentaires pour surmonter les verrous de scalabilité dans les architectures quantiques supraconductrices : l’impactnéfaste des excitations fortes nécessaires aux opérations rapides et de haute-fidélité, et l’intégration de mémoires quantiquesdans des architectures hybrides afin de réduire la complexité matérielle.Dans la première partie, nous étudions comment les excitations fortes peuvent engendrer des effets délétères dans lecircuit. Dans ce régime, les transmons peuvent devenir structurellement instables. En raison d’effets chaotiques, le sous-espacecomputationnel n’est plus bien séparé du reste du spectre, ce qui se traduit par une sensibilité accrue au bruit decharge et par des effets destructeurs lors de la lecture. Nous montrons que ces effets indésirables peuvent se propagerà d’autres degrés de liberté, par exemple à des qubits voisins dans un système multi-qubits. En particulier, un transmonexcité de manière cohérente peut agir comme une source de bruit incohérent pour un autre élément du circuit auquel ilest couplé. En combinant un modèle quantique complet et une analyse semi-classique, nous réalisons la spectroscopiedu bruit d’un transmon excité couplé à un système spectateur à deux niveaux (TLS), et montrons que, dans une certainelimite, cette interaction peut être modélisée comme un processus diffusif stochastique agissant sur le TLS. Ce résultatmet en évidence une source fondamentale de diaphonie induite par les excitations dans les dispositifs multi-qubits.Dans la seconde partie, nous adoptons une approche modulaire du calcul quantique évolutif, qui associe une unité detraitement à un module de stockage. Nous nous intéressons aux mémoires quantiques à base de spins, dans lesquellesles champs électromagnétiques itinérants sont stockés dans de grands ensembles de systèmes effectifs à deux niveaux,intégrés dans une cavité. En utilisant un cadre de champ moyen, nous modélisons l’ensemble comme un canal de communicationquantique effectif, et développons un modèle quantique en cascade pour décrire les processus d’absorptionet d’émission. Nous dérivons des modulations optimales, dépendantes du temps, de la largeur de raie de la cavité afinde maximiser l’efficacité du stockage et de la récupération pour des paquets d’ondes de durée finie. Notre analyse fournitune borne supérieure sur l’efficacité, atteignable dans le régime de faible bande passante, et met en évidence l’existenced’une bande critique au-delà de laquelle l’efficacité décroît fortement
Machine learning based thermodynamic modelling of acid gas absorption in aqueous MDEA and aqueous Piperazine
No full textAcid gas chemical absorption with aqueous amine solvents is an important industrial technology for gas processing and CO2 capture. Vapor-Liquid Equilibrium (VLE) reflects the efficiency of solvents and is essential to model the thermodynamics of the absorption and solvent regeneration process. In this study, several machine learning (ML) approaches were used to develop VLE models for acid gas absorption in aqueous methyldiethanolamine (MDEA) and piperazine (Pz), namely CO2-MDEA-H2O, H2S-MDEA-H2O, CO2-H2S-MDEA-H2O and CO2-Pz-H2O systems. New experimental data are presented for the CO2-MDEA-H2O and H2S-MDEA-H2O ternary systems, and they are used to compare the accuracy of the ML models to an earlier reported activity coefficient (e-NRTL)-based thermodynamic model. For the quaternary system CO2-H2S-MDEA-H2O, the ML models and the physical model are compared using experimental data from literature, because the physical model of the quaternary system is only trained on the ternary experimental systems. The results indicate that the optimal ML approach is not systematically more accurate than the physics-based model
From Anatomical To Biomechanical Personalization: Comparing Patient-Specific Implants And Miniplates For Le Fort I Osteotomy Using Finite Element Modelling
No full textInternational audienc
Antioxidant depletion in polyethylene used in chlorinated seawater: from accelerated ageing to modelling
No full textInternational audienceThis study focuses on the loss of stabilizers used in polyethylene (PE) formulations exposed to chlorinated seawater. A PE model containing antioxidants is aged in three different environments: deoxygenated water, oxygenated water and chlorinated seawater with oxygen. To elucidate slow-occurring phenomena, accelerated ageing experiments were performed across a range of chlorine concentrations and temperatures. Additionally, ageing tests are performed on samples of different thicknesses to distinguish between surface processes, reaction-based mechanisms and diffusion-involved mechanisms. In all cases, the loss of stabilizers is monitored by measuring the oxidation induction time. The results show that the physical extraction of stabilizers (leaching) can be described using existing theories from the literature, with the extraction constants identified based on tests in deoxygenated water at 60, 80 and 90°C. It appears that the presence of oxygen in water leads to a faster decrease in stabilizers over time, explained by the reaction of these molecules with free radicals formed during the oxidation process. Kinetic rates associated with this process were determined at 40, 60, 80 and 90°C. Finally, for the first time, the kinetics of stabilizer loss in chlorinated seawater are identified for two pro-oxidant concentrations at 21°C. Based on theoretical considerations from the literature, a kinetic model of stabilizer loss is proposed to describe the results obtained in this study
