Portail HAL ONERA
Not a member yet
12842 research outputs found
Sort by
Predicting Aerodynamic Loads on Flexible Wings through Low-Dimensional Balanced Projection
No full textInternational audienceAccurate and efficient prediction is vital for various tasks in aerospace engineering, including real-time data assimilation in aero-elastic phenomena, hybrid numerical-experimental setups, and shape optimization processes. These applications often demand minimal inference times, highlighting the necessity for simple yet effective reduced-order models. In this study, we focus on predicting the steady pressure field around a wing in inviscid, compressible flow, using as inputs the angle of attack and wing surface mesh. To this aim, a design of experiment (DoE) is generated by deforming the nominal wing surface using a basis of shape functions, to then compute the flow solution around the wing via the Euler equations. The resulting dataset of shapes and pressure distributions is used to train data-driven models that capture the relationship between these two fields at the wing surface. We initially employ classical Proper Orthogonal Decomposition (POD) to separately compress displacement fields and pressure distributions. By learning the mapping between their latent space variables using a fully-connected neural network, we establish a relation that allows for efficient reconstruction of the pressure field. Next, we explore Balanced Proper Orthogonal Decomposition (BPOD), which identifies correlated directions between displacement and pressure fields. This approach simultaneously provides a basis for each field, offering a more integrated approach to capturing the essential input/output behaviour. This method can be seen as an extension that leverages the interdependencies between the fields, potentially enhancing the accuracy and efficiency of the reconstruction process. Finally, we compare these linear compression methods with a Convolutional Neural Network (CNN)-based network to evaluate the reconstructed pressure field accuracy and the training/inference times for each method. This analysis aims to identify the most effective approach for real-time aerospace engineering applications that balance computational efficiency with precision in capturing complex aerodynamic interactions
Contrôle en boucle ouverte des écoulements en cavité basé sur des méthodes variationnelles d'ensemble
No full textInternational audienceFlow control in fluid mechanics is an important concept in ensuring that fluid systems operate efficiently and reliably. It involves guiding the flow to achieve specific goals, such as reducing energy losses or improving stability. Often, flow control is viewed as an optimization problem aimed at achieving objectivessuch as preventing the growth of flow disturbances or maintaining desired flow behavior. In this study, we aim to address the optimization of an open-loop control strategy using ensemble-based variational techniques (EnVar), offering an alternative to the adjoint-based optimization method. In fact, while adjoint methods require intensive gradient computations, such as Jacobian calculations, the EnVar method, which originated in the data assimilation community, is non-intrusive and therefore easier to formulate and implement. The search for an optimal control strategy takes place within a subspace defined by an ensemble of realizations, corresponding to various simulations with different control parameters. The iterative procedure aims to determine the optimum through a linear combination of controls in the ensemble of realizations.Our goal is to apply this methodology to perform the control of a two-dimensional cavity flow, which exhibits multiple instabilities at high Reynolds numbers. To this end, we first introduce a parameterized (in frequency and amplitude) localized forcing term in the momentum equations, which mimics blowing/suction close to the cavity walls. Subsequently, a forcing term that varies in both time and space is also considered, greatly increasing the dimension of the control vector. The considered control vectors are optimized in order to minimize a cost function that is defined as the norm of the velocity fluctuations relative to the base flow, i.e. a fixed-point solution of the Navier-Stokes equations, considering the full (i.e. nonlinear) unsteady Navier-Stokes equations as equality constraints.The control vectors that result from the optimization framework successfully lead to a significant reduction in the kinetic energy of the fluctuations. In addition, we compare the EnVar optimization method with the adjoint method to highlight the advantages of the newly implemented approach, and we explore different aspects of the optimization process – such as the generation of the ensemble members – with a future focus on applying the EnVar methodology to identify open-loop control strategies for complex and unstable applications, such as rotor-stator cavities
Estimating differential pistons for the Extremely Large Telescope using focal plane imaging and a residual network
No full textInternational audienceAs the Extremely Large Telescope (ELT) approaches operational status, optimising its imaging performance is critical. A differential piston, arising from either the adaptive optics (AO) control loop, thermomechanical effects, or other sources, significantly degrades the image quality and is detrimental to the telescope’s overall performance. Aims. In a numerical simulation set-up, we propose a method for estimating the differential piston between the petals of the ELT’s M4 mirror using images from a 2 × 2 Shack-Hartmann wavefront sensor (SH-WFS), commonly used in the ELT’s tomographic AO mode. We aim to identify the limitations of this approach by evaluating its sensitivity to various observing conditions and sources of noise. Methods. Using a deep learning model based on a ResNet architecture, we trained a neural network (NN) on simulated datasets to estimate the differential piston. We assessed the robustness of the method under various conditions, including variations in Strehl ratio, polychromaticity, and detector noise. The performance was quantified using the root mean square error (RMSE) of the estimated differential piston aberration. Results. This method demonstrates the ability to extract differential piston information from 2 × 2 SH-WFS images. Temporal averaging of frames makes the differential piston signal emerge from the turbulence-induced speckle field and leads to a significant improvement in the RMSE calculation. As expected, better seeing conditions result in improved accuracy. Polychromaticity only degrades the performance by less than 5%, compared to the monochromatic case. In a realistic scenario, detector noise is not a limiting factor, as the primary limitation rather arises from the need for sufficient speckle averaging. The network was also shown to be applicable to input images other than the 2 × 2 SH-WFS data
Characterization of debris disks observed with SPHERE
No full textInternational audienceAims. This study aims to characterize debris disk targets observed with SPHERE across multiple programs, with the goal of identifying systematic trends in disk morphology, dust mass, and grain properties as a function of stellar parameters. By combining scattered-light imaging with photometric and parametric modeling, we seek to improve our understanding of the composition and evolution of circumstellar material in young debris systems and to place debris disks in the broader context of planetary system architectures. Methods . We analyzed a sample of 161 young main-sequence stars using archival SPHERE observations at optical and near-infrared (IR) wavelengths. Disk geometries were derived from ellipse fitting and model grids, while dust mass and properties were constrained by modified blackbody (MBB) and size distribution (SD) modeling of spectral energy distributions (SEDs). We also carried out dynamical modeling to assess whether the observed disk structures can be explained by the presence of unseen planets. Results . We resolve 51 debris disks, including four new detections where disks are resolved for the first time: HD 36968, BD-20 951, and the inner belts of HR 8799 and HD 36546. In addition, we find a second transiting giant planet in the HD 114082 system, with a radius of 1.29 ± 0.05 R Jup and an orbital distance of ~1 au, providing an important new benchmark for planet–disk interaction studies. Beyond these new detections, we identify nine multi-belt systems, with outer-to-inner belt radius ratios of 1.5–2, and find close agreement between scattered-light and millimeter continuum belt radii with a mean ratio R belt (near-IR)/ R belt (mm) of 1.05 ± 0.04. Belt radii scale weakly with stellar luminosity ( R belt ∝ L ⋆ 0.11±0.05 ), but show steeper dependencies when separated by CO and CO 2 freeze-out regimes, and also increase with age as R belt ∝ t age 0.37±0.11 . Uniform image modeling yields vertical disk aspect ratios of 0.02–0.06, consistent with collisionally stirred belts, while gas-rich systems show unusually small values. Inner density slopes steepen with stellar luminosity, indicating more efficient dust removal around luminous stars. Disk fractional luminosities follow collisional decay trends, declining as t age −1.18±0.14 for A-type and t age −0.81±0.12 for F-type stars. SD modeling yields minimum grain sizes consistently above the blowout limit, typically >0.8 μm, with a mean SD index of q = 3.6, assuming astrosilicate composition. The inferred dust masses span 10 −5 −1 M ⊕ from MBB modeling (and 0.01–1 M ⊕ from SD modeling for detected disks). These masses scale as R belt n with n > 2 in belt radius and super-linearly with stellar mass, consistent with trends seen in protoplanetary disks (PPDs). Our detailed analysis of disk scattered-light non-detections indicates that they are mainly caused by low dust masses, unfavorable viewing geometries, or suboptimal observing conditions. SD modeling combined with Mie theory further shows that bulk albedos are consistently above 0.5 with little variation, making albedo differences an unlikely explanation. To explore this further, we introduced a new parametric approach based on scattered-light and polarized-light images, which provides independent estimates of dust albedo and maximum polarization fraction. We find a correlation between measured disk polarized flux and IR excess, with a slope shallower than that of optical total-intensity fluxes measured with HST/STIS. The offset of ~1 dex between total-intensity and polarized fluxes arises because polarized flux represents only a fraction of the total scattered light which depends on both grain properties and disk inclination. Finally, a comparison of planetary architectures shows that most benchmark systems resemble the Solar System, with multiple planets located inside wide Kuiper-belt analogues. Dynamical modeling further indicates that many observed gaps and inner edges can be explained by unseen planets below current detection thresholds, typically with Neptune- to sub-Jovian masses, underscoring the likely ubiquity of such planets in shaping debris disk morphologies
Mappes difféomorphiques pour la réduction de modèle appliqué en aérodynamique
No full textParametric fluid dynamics simulations present significant computational challenges in aerodynamics, where high-fidelity computational fluid dynamics (CFD) models involve millions of degrees of freedom. Reduced-order models (ROMs) address this computational burden by constructing low-dimensional approximations of the full-order system. However, standard projection-based approaches exhibit poor performance for advectiondominated flows. A proposed solution is to rely on coordinate transformation mappings to enhance ROMs compression. This thesis addresses the issue of developing a general framework for diffeomorphic mappings to align aerodynamic structures for application in coordinate transformation-based ROMs. To answer this problem, we first establish the mathematical foundations of the method. For this, the mapping of interest is defined as the minimizer of an objective function. This constitutes the registration problem. The objective function combines a data misfit term based on point-set alignment of coherent structures with a regularization term derived from differential operators. The mapping is defined from a parametrization of a velocity field. This guarantees diffeomorphic transformations in bounded CFD domains under tangent boundary conditions. The minimization problem is solved using a finiteelement discretization, gradient-based optimization, and an Expectation-Maximization algorithm for automatic coherent-structure labeling. The methodology is validated on three representative test cases of increasing complexity: (i) coalescing Gaussian mixtures, to illustrate the alignment of merging structures; (ii) transonic Euler flow around a NACA0012 airfoil, demonstrating multi-structure registration; and (iii) viscous RANS flow around the ONERA M6 wing, showcasing the alignment of complex 3D lambda shocks. Across these cases, the proposed approach enhances ROM accuracy and demonstrates robustness to aerodynamic problems.Les simulations paramétriques de dynamique des fluides présentent des défis computationnels significatifs en aérodynamique. Les modèles de dynamique des fluides computationnelle (CFD) de haute fidélité impliquent des millions de degrés de liberté. Les modèles d’ordre réduit (ROM) résolvent cette charge computationnelle en construisant des approximations de faible dimension du système d’ordre complet. Cependant, les approches standard basées sur la projection linéaire des équations présentent de mauvaises performances pour les écoulements dominés par la convection. Une solution proposée consiste à s’appuyer sur des changements de coordonnées pour améliorer la compression des ROMs. Cette thèse aborde la question du calcul des transformations difféomorphes dans le cadre général afin d’aligner les structures aérodynamiques pour l’application de ROMs plus robustes. Pour répondre à ce problème, nous établissons d’abord les fondements mathématiques de la méthode. Pour cela, la transformation d’intérêt est définie comme le minimiseur d’une fonction objective. Ceci constitue le problème de recalage (ou d’alignement). La fonction objective combine un terme d’inadéquation des données, basé sur l’alignement d’ensembles de points de structures cohérentes, avec un terme de régularisation dérivé d’un opérateur différentiel. La transformation est définie à partir d’une paramétrisation d’un champ de vitesse. Ceci garantit des transformations difféomorphes dans les domaines CFD bornés grâce à des conditions aux limites tangentielles. Le problème de minimisation est résolu en utilisant une discrétisation par éléments finis, une optimisation basée sur le gradient et un algorithme d’Espérance-Maximisation pour l’étiquetage automatique des structures cohérentes. La méthodologie est validée sur trois cas de test représentatifs de complexité croissante : (i) des mélanges gaussiens coalescents, pour illustrer l’alignement de structures fusionnantes ; (ii) un écoulement Euler transsonique autour d’un profil NACA0012, démontrant le recalage multi-structures ; et (iii) un écoulement RANS visqueux autour de l’aile ONERA M6, présentant l’alignement de chocs en lambda dans des géométries 3D. À travers ces cas, l’approche proposée améliore la précision des ROM et démontre une robustesse aux problèmes aérodynamiques
Effets du nombre de Reynolds et de la séparation sur le scénario de transition au-dessus d'un objet cône-cylindre-jupe en hypersonique
No full textInternational audienceThis experimental and numerical study examines transition to turbulence for a Cone-Cylinder-Flare geometry at Mach 7 and across a broad Reynolds number range. The focus is set on both attached boundary layers and separated shock-boundary layer interactions. The campaign is conducted in the R2Ch facility. Unsteady wall pressure fluctuations and high-speed schlieren images are analysed using data-driven techniques and compared with base flow computations and global linear stability analysis. The results distinguish two transition regimes. At high Reynolds numbers, transition is dominated by the second Mack mode and its non-linear interactions on the cone. High-frequency wall pressure measurements and schlieren imaging permit the capture of both fundamental waves and their non-linear harmonics. Non-linear interaction regions are resolved with unprecedented detail, clarifying the boundary-layer state before rapid breakdown at reattachment. At lower Reynolds numbers, the transition scenario is more intricate, marked by the coexistence of low- and high-frequency modes. A complex coupling between separated flow and convective instabilities is revealed, with trapped acoustic waves inside the recirculation region measured experimentally for the first time. Their linear origin is demonstrated through global stability analysis, and a simple acoustic duct model is provided to predict their frequencies. These waves offer a new interpretation of low-frequency pressure signatures and suggest a mechanism for energy transfer from high to low frequencies, ultimately driving transition on the flare. The findings advance understanding of hypersonic boundary-layer transition and its dependence on Reynolds number and flow separation.Une étude expérimentale et numérique de la transition vers la turbulence dans la couche limite et à travers l'interaction entre la couche limite et le choc séparé est réalisée pour une géométrie cône-cylindre-jupe en régime hypersonique froid à un nombre de Mach de 7 et pour une large gamme de nombres de Reynolds. La campagne expérimentale est menée dans l'installation R2Ch et permet de collecter des fluctuations de pression instationnaires a la paroi et des images strioscopiques à grande vitesse pour tous les régimes d'écoulement. Les données collectées sont ensuite post-traitées à l'aide d'une analyse basée sur les données et comparées aux calculs d'écoulement de base et à l'analyse de stabilité linéaire globale afin de mieux comprendre les mécanismes en jeu dans la transition vers la turbulence observée dans les expériences.Pour les cas à nombre de Reynolds élevé, la transition s'avère être entraînée par le deuxième mode de Mack et ses non-linéarités sur le cône grâce. À des nombres de Reynolds plus faibles, le scénario devient plus complexe, avec la coexistence des premier et deuxième modes, un couplage complexe entre la séparation causée par l'interaction entre la couche limite de choc et la transition dominée par les modes convectifs, avec une forte variation de la longueur de la région séparée en fonction du nombre de Reynolds. Enfin, pour la première fois, les ondes acoustiques piégées à l'intérieur de la région de recirculation sont mesurées et leur origine linéaire est démontrée
Precise orbit prediction for GEO and MEO non-cooperative objects: experimental validation using a passive ground-based optical station
No full textInternational audienceEnhanced tracking of resident space objects (RSOs) plays a key role in improving the safety and sustainability of space operations. However, the accuracy of observations constrains the orbit determination (OD) of RSOs. Passive ground-based optical stations, which serve as a convenient source of sensing devices for medium-and high-Earth-orbits (MHOs), provide angular measurements with typical accuracies ranging from 150 to several hundred milliarcseconds (mas). Following successive nights of observations, an RSO's orbit can be determined with an accuracy of a few hundred meters. The CICLOPE station, which was recently developed at ONERA, provides data with 50 mas angular accuracy. This corresponds to an accuracy of 5-10 m at MHO distance, which is sufficient to detect deviations from well-known gravitational dynamics over a few hours, such as those due to solar radiation pressure (SRP).This paper demonstrates what we call precise orbit prediction (POP), which achieves decametric accuracy over two days for non-cooperative MHO RSOs using angle-only measurements. In practice, we first determine which contributors are needed to model forces with sufficient accuracy and then build the propagator accordingly. Based on the constructed force model, we develop a precise orbit determination (POD) algorithm using a weighted least squares method and a Jacobian derived from variational equations. After validating the algorithm, we evaluate the POP performance using observations derived from precise ephemerides of reference satellites. We add a characteristic 50 mas random noise to simulate experimental angle measurements overnight with a low acquisition frequency. Initially, using only standard gravitational models, we obtain hectometric errors on propagated orbits, which underscores the importance of NGP residuals. Adding next-night observations to estimate SRP parameters accurately yields an accuracy of 14 m over the POD time span, increasing to 30 m over the 48-hour POP period. Finally, we demonstrate the decametric accuracy of our POP solution using CICLOPE data.</p
Microwave field imaging inside an atomic cell by fluorescence thermography
No full textInternational audienceRydberg atom-based electromagnetic field sensors are currently the focus of intense research efforts, promising high sensitivity, tunability, and compactness with SI-traceable measurements. However, the presence of a cell around the atoms alters the microwave field distribution, particularly due to reflection by materials like quartz. To characterize these sensors, measuring how the presence of the cell affects the microwave field distribution is therefore essential. Electro-optical probes offer high spatial resolution but require time-consuming scanning. Infrared electromagnetic thermography provides direct field mapping but is limited by the opacity of glass to infrared radiation. In contrast, fluorescence thermography allows for electric field mapping through optically transparent media. In this paper, we report the application of this technique to map the microwave electric field inside a quartz cell intended for atomic cooling and trapping. This provides a direct observation of the standing waves inside the cell, in good agreement with numerical simulations. This technique could be a valuable tool for the design and optimization of atomic cells for applications requiring good control of the microwave field distribution inside the cell, such as atomic clocks or electric field sensors based on hot or cold Rydberg atoms
Simulation of Lifting Bodies with Vortex Particles Through Active Enstrophy Control
No full textInternational audienceThe Vortex Particle Method (VPM) represents incompressible, unsteady, viscous flows by using independent particles carrying small concentrations of vorticity. This method is suitable for simulating vortex reconnection in the wake as well as flow mixing. It is capable of simulating complex flow fields at a moderate computational cost compared to CFD methods. The VPM can be used alongside a lifting line to estimate the loads and the wake shed by thin lifting bodies. The coupling procedure and the shedding of vortex structures are detailed. The coupled method is validated on a representative range of aeronautical applications of fixed and rotating lifting bodies of increasing complexity. It is well established that turbulent regimes in the wake can lead to numerical instabilities. A new method, based on the control of the enstrophy, is proposed to better represent turbulent flow regimes, thus improving the numerical stability of the VPM. The Diffusion Velocity Method is used to transport and diffuse vortices and represent wake decay. The proposed method is satisfactory in preserving the coherence of the vortex structures and is able to simulate turbulent cases such as rotor-on-rotor interactions or hovering helicopter rotors. The coupled method is able to accurately represent the loads on the lifting bodies and the evolution of their wake