Portail HAL ONERA
Not a member yet
12842 research outputs found
Sort by
Observation of sodium D-line quantum beat in Laser Induced Fluorescence using stepwise fs excitation
No full textInternational audienceWe report a direct observation of quantum beating between 3P1/2 and 3P3/2 levels of atomic sodium using ultrashort laser pulses in a stepwise excitation scheme. Two broadband femtosecond laser pulses are used to excite the 6S level from the 3S ground level in sodium vapor. The first one is simultaneously resonant with the D1 and D2 lines of the 3S→3P transition, and the second one with the 3P→6S transition. The delay between the two pulses is scanned in order to probe the dephasing between the two excited intermediate levels. Quantum beating is observed by monitoring the fluorescence signals at 330 nm and 515 nm , namely transition 4P→3S and 6S→3P, respectively. Signals are compared to a semiclassical model that reproduces the oscillations with a very good agreement
Expériences en soufflerie sur de longs arcs en écoulement transversal : pied d’arc anodiques et influence du champ de vitesse
No full textInternational audienceThis paper presents an experimental investigation into the dynamics of long electrical arcs attached to an airfoil immersed in crossflow. The arcs are D.C., with around 3 A current and up to 1 m in length. Diagnostics include simultaneous high-speed video imaging of the arc column and root, particle image velocimetry (PIV) measurements of the flow velocity field, and electrical measurements of arc voltage and current. The study focuses on an anodic airfoil polarity, with variations in wind speed (1–4 m s −1 ), angle of attack (0 ∘ –20 ∘ ), and different cathode-electrode configurations (ball and rail). Experimental results reveal that, for smaller angles of attack and higher wind speeds, the arc root adopts a ‘jumping’ behavior along the airfoil, while separated flow at higher angles of attack tends to stagnate the root near the separation point. Despite significant arc elongation and tortuosity induced by the flow, the instantaneous value of the arc’s internal electric field aligns well with values for free-burning arcs reported in the literature. PIV and optical-flow analyses further indicate that the arc column is advected nearly at the local flow speed, highlighting the strong coupling between arc behavior and flow environment
Anisotropic h-adaptation on unstructured grids for modal Discontinuous Galerkin schemes with application to aircraft configurations
No full textInternational audienc
Assessment of uncertainty quantification strategies for the robust optimization of a propeller
No full textInternational audienceThe CFD community is increasingly aware of the impact of uncertainties on the results of flow simulation and CFD based optimization. In practice, uncertainties coming from the geometrical shape (from manufacturing tolerance, wear and tear, etc.) or aerodynamic conditions are evaluated using non-intrusive stochastic methods. These methods are subject to the curse of dimensionality, as the number of simulations needed increase exponentially with the number of uncertain parameters. Thus, the selection of the most efficient technique is crucial for the practicality of the method in an industrial context. In this work, we propose to assess efficient Uncertainty Quantification (UQ) methods on a 2D test case, a transonic airfoil representative of an open rotor design. The airfoil geometry is affected by twelve uncertain parameters, while two aerodynamic conditions are also uncertain. A precise surrogate model has been built by Safran-Tech on this test case, using the Kriging technique on a design of experiment of size 600. Thanks to this surrogate, an assessment of several UQ methods based on generalized Polynomial Chaos (gPC) expansion has been made. More precisely, the polynomial chaos expansions in this work are built with colocation technics, namely Least Square Approximation (LSA), Least Square Approximation with Gradient Enhancement (LSAGE), Least Angle Regression (LARS) and Basis Pursuit Denoise (BPDN). A combination of compressed sensing (BPDN) enhanced with gradient information is also evaluated. These methods will be employed by using the open-source UQ toolboxes OpenTurns and equadratures. Methods used by three partners of the NEXTAIR project (Safran-Tech, ONERA and UNICA) have been evaluated and compared on this common test case.Once the uncertainty quantification step is mastered, one has access to the statistics moment of the quantity of interest (drag for instance). Optimization taking as objective and constraints these moments (robust optimization), can then be carried out. This work will present a first robust optimization test case. A focus will be made on how to compute the gradient of the statistics moments, necessary for gradient based optimization
Étude expérimentale du tenseur de sous-mailles basée sur les fonctions de structure de second ordre dans la turbulence de grille
No full textInternational audienceMeasurements are carried out in a turbulent decaying flow behind a grid of the same dimensions as the wind tunnel cross section (150 mm X 300 mm) with a squared mesh size of M=40mm. A Dantec 55P11 hot-wire is used with a thickness of 5 micrometers and a length of 1.25mm. The time resolution frequency is 166kHz with an anti-aliasing filter set at 100kHz so that the smallest resolved time scales are much smaller than Kolmogorov turnover time scale (5%).Different mean flow speeds are tested from 13 m/s to 20 m/s corresponding to Taylor Reynolds number values between 190 and 360. The measurements are conducted at different streamwise positions between x/M=7.4 and x/M=19.9. A Taylor hypothesis is used to evaluate the second order structure functions scalings with dissipation and mean turbulent kinetic energy. The mean structure functions, normalized by dissipation and Kolmogorov scale, collapse reasonably well. The same results are observed for the mean subgrid-stress at different filter widths and these results are interpreted in light of the exact connection between structure functions and the subgrid-stress introduced by Germano (2007). A rather good collapse is observed, once normalized by u^3/L, at each location, where u is the velocity standard deviation and L the integral scale. The mean subgrid-stress contribution is observed to be well predicted by the Smagorinsky model with the same coefficient as the one measured in Meneveau (1994) but the Smagorinsky model local fluctuations are completely different to the real subgrid-stress contribution. Different exact decompositions of the subgrid stress based on the exact Germano (2007) equation are analysed with the experimental data to identify a possible decomposition of the subgrid-stress into a dissipative and non-dissipative part
Graphene as infrared and electron transparent electrode applied to the design of narrow band gap nanocrystal-based photodiode
No full textInternational audienceColloidal nanocrystals (NCs) are a promising platform for infrared optoelectronics. Efforts focus on designing NCs that absorb in the short-and mid-wave infrared and integrating them into diode stacks. A major challenge is coupling these sensors to read-out integrated circuits (ROICs) for infrared imaging, which requires infrared-transparent top electrodes. Conventional materials like tin-doped indium oxide lose transparency at longer wavelengths, limiting their effectiveness. Metallic grids have emerged as an alternative but struggle to maintain a uniform potential, as shown by nanobeam X-ray photoemission microscopy. To address this, graphene is explored as a transparent electrode. A novel diode stack is proposed to maintain a backside mirror, accommodate HgTe NCs' chemical constraints, and incorporate electrodes that efficiently extract both electrons and holes. Unlike conventional designs limited to near-zero bias, this stack operates optimally under CMOS ROIC conditions. Additionally, its transparent electrode allows photoelectron emission from within the diode, enabling in-situ electric field analysis. This capability could rationalize the optimization process of photodiode desig
How Does Disability Affect Incomes? An Empirical Study on Older European Workers
No full textInternational audienceThis paper studies the impact of the onset of disability on personal income. Using the Survey of Health, Ageing and Retirement in Europe, we compare income trajectories of individuals who experience disability with those who remain healthy over the same period. We hypothesize that the onset of disability reduces overall personal income, as the loss in wages is not fully compensated by disability-related benefits. To identify the causal impact, we combine a difference-in-differences approach with kernel propensity score matching, controlling for both observable and time-invariantunobservable individual characteristics. Our results confirm this hypothesis, show-ing a substantial decrease in personal income driven primarily by reduced wages.We further investigate heterogeneous effects by gender and the generosity of social welfare systems, showing that more generous welfare states mitigate income losses due to disability
Emetteurs thermiques nanostructurés haute température
No full textMetasurfaces, based on arrays of nanoantennas, enable the control of the optical response (reflectivity, absorption, transmission) of a surface within a subwavelength thickness. As such, an initially reflective surface can become fully absorbing at certain wavelengths and, according to Kirchhoff's law, act as a selective thermal emitter. The control of infrared emission opens the way to numerous applications, ranging from optical sources to stealth devices and thermophotovoltaic (TPV) energy conversion. One of the major challenges for these applications lies in the design of nanostructured emitters capable of operating at high temperatures (>1000~K) while allowing precise engineering of their spectral response. Ultra-refractory materials, such as zirconium carbide or tungsten, make it possible to reach this temperature range, although they exhibit higher optical losses than noble metals. Mastering their processing, along with an in-depth characterization of their optical, thermal, and mechanical properties, is therefore essential. The work presented in this manuscript is organized around four main objectives: the deposition and structuring of thin films of ultra-refractory materials, the characterization of their temperature-dependent properties, the simulation of metasurfaces based on these measured properties in order to control emissivity, and finally the fabrication and experimental validation of nanophotonic structures. To address these points, the manuscript presents an introduction to the thermal quantities required to understand the underlying phenomena, the deposition of thin films in a cleanroom environment, and the characterization of their properties as a function of temperature. Based on these results, coupled Fabry-Perot-type structures were designed and fabricated with the aim of controlling emissivity at high temperatures. Finally, building upon this understanding of emissivity structuring, the manuscript focuses on a more detailed study of two specific applications: infrared stealth and solar thermophotovoltaics.Les métasurfaces, basées sur l'agencement de nanoantennes, permettent de contrôler la réponse optique (réflectivité, absorption, transmission) d'une surface avec une épaisseur sub-longueur d'onde. Ainsi, une surface initialement réfléchissante peut devenir totalement absorbante à certaines longueurs d'onde et, d'après la loi de Kirchhoff, un émetteur thermique sélectif. Le contrôle de l'émission infrarouge ouvre la voie à de nombreuses applications, allant des sources optiques aux dispositifs de furtivité et à la conversion thermophotovoltaïque (TPV). L'un des défis majeurs pour ces applications consiste à concevoir des émetteurs nanostructurés capables de fonctionner à haute température (>1000 K) avec une ingénierie de leur réponse spectrale. Les matériaux ultraréfractaires, tels que le carbure de zirconium ou le tungstène, permettent d'atteindre cette gamme de température, bien qu'ils présentent des pertes optiques plus élevées que les métaux nobles. Leur maîtrise technologique, ainsi qu'une caractérisation approfondie de leurs propriétés optiques, thermiques et mécaniques, est nécessaire. Le travail présenté dans ce manuscrit s'articule ainsi autour de quatre enjeux principaux : l'élaboration et la structuration de couches minces de matériaux ultraréfractaires, la caractérisation de leurs propriétés en fonction de la température, la simulation de métasurfaces à partir de ces propriétés mesurées afin de contrôler l'émissivité, et enfin la fabrication suivie de la validation expérimentale des structures nanophotoniques. Pour y répondre, nous présentons au fil du manuscrit une introduction aux grandeurs thermiques nécessaires à la compréhension des phénomènes, le dépôt des couches en salle blanche et la caractérisation de leurs propriétés ainsi que leur évolution en fonction de la température. Sur cette base, des structures de type Fabry-Perot couplés ont été étudiées et réalisées dans l'optique de contrôler l'émissivité à haute température. Enfin, à partir de cette compréhension de la structuration de l'émissivité, le manuscrit se concentre sur l'étude plus approfondie de deux applications spécifiques : la furtivité infrarouge et le thermophotovoltaïque solaire
Comparison of Optical and Altimetry floe/lead classification using co-located Sentinel-3 OLCI/SRAL data
No full textInternational audienceSubmarines and satellite records from 1958 and 2018 reveal a significant decline of sea ice thickness [5]. In the context of polar and global changes, the estimation of sea ice thickness becomes more and more crucial to accuratelyquantify the energy exchange between ocean and atmosphere in climate models.Altimeter sensors, such as SIRAL/CryoSat-2 and SRAL/Sentinel-3, allow to derive sea ice thickness from the measures of ice elevation and sea elevation [7]. The precision of the estimation relies on the classification of the altimeter’s measure into valid floe or lead surfaces [1] [2]. This classification is based on the analysis of the echo’s waveform [8] [3], which depends on sea ice roughness, snow cover and other features included in the footprint of the measure [6]. However, the variability of the surfaces results in classification ambiguities. To improve the precision of sea ice thickness products, we propose a method to refine this classification by sharpening the selection of valid floe/lead measures used in the estimation.The Sentinel-3 mission represents a unique opportunity to study coincident sea ice surfaces and thickness data from the co-located SRAL altimeter and OLCI multispectral imager at high spatial and temporal resolutions. Using this synchronized data, our goal is to produce a dataset that combines OLCI classified images with SRAL’s waveforms. This dataset can be used to derive new robust criteria for the SRAL’s floe/lead classification. In that purpose, we developed a classification method for OLCI’s images. To classify OLCI’s images, a hand-made dataset is built, sampling 108 labeled, cloud-free subset images evenly distributed from March to May 2022 over Svalbard. The samples correspond to 100x100 pixels subset of the OLCI’s red band that are centered on the localisation of a SRAL’s measure. Given the variability of the intensity of both the floe and the lead classes, we use a homogeneous region containing the SRAL measure to make the classification decision. This segment is outputed by the Segment Anything Model (SAM) [4] algorithm using the SRAL measure as a prompt in a zero-shot learning framework. First results of the SAM segments lead us to extend the floe class to brash ice and pack ice classes as the scale and the shape of these three types of objects significantly differ. Thus, the OLCI’s classification we developed is composed of four labels: lead, floes (well-shaped ice floe), brash ice and pack ice. These labeled segments will allow to define shaped (SAM predicted IoU, continuity, etc.) and pixel intensity (mean, variance, etc.) criteria to discriminate each class. A same dataset is built from March to May 2023 in order to validate the method.Extended to the Arctic, this method could be used to define better criteria to select valid floe and lead waveforms and increase the number of sea-ice classes, such as separating floes from brash ice. It could even serve to create an extensive dataset to train a deep-learning waveform classifier. Beyond the altimetry scope, the method of sea ice surfaces classification could be used to study sea ice at the floe level, especially during spring periods when altimetry sea ice thickness estimation is limited
Étude des effets non linéaires dans les fibres amplificatrices à large diamètre de mode dopées ytterbium pour la réalisation d'un laser UV adapté à la mesure de vent par un lidar aéroporté
No full textONERA is developing a lidar demons-trator based on direct detection and Rayleighbackscattering. The receiver for this lidar, develo-ped by T. Boulant, consists of a Newtonian te-lescope accompanied by a Quadri-MachZehnder(QMZ). Studying the reception system yielded op-timal laser characteristics in the UV range. Conse-quently, we need to generate 20 W at a 40 kHz re-petition rate in the UV, with pulses having a spec-tral width close to hundreds of megahertz. My the-sis work aims to develop this UV laser system. Wechose to design a Master Oscillator Power FiberAmplifier (MOPFA) type laser system. Fibers havethe advantage of being inherently robust to vibra-tions compared to free-space laser systems. Howe-ver, they are highly sensitive to nonlinear effects,notably, in our case, Stimulated Brillouin Scatte-ring (SBS). To overcome this nonlinear effect, afrequency comb obtained through phase modula-tion is used in the system. If its characteristics arewell-chosen, it can drastically increase the Brillouinthreshold power without impacting the measure-ment performed by the QMZ. Furthermore, topush back nonlinear effects even further, nume-rous large-core fibers were tested and compared.Nevertheless, two detrimental phenomena relatedto using the frequency comb in fibers were obser-ved : Brillouin self-injection and four-wave mixing.Both phenomena were studied as they can signifi-cantly impact the final system. Finally, a free-spaceamplification system was added at the output ofthe fiber amplifiers to increase the peak signal po-wer, thus facilitating UV generation in nonlinearcrystals.L'ONERA développe un démonstrateurlidar basé sur la détection directe et la rétrodiffu-sion de Rayleigh. Le récepteur de ce lidar a été dé-veloppé par T.Boulant et il s'agit d'un télescope deNewton accompagné d'un Quadri-Mach Zehnder(QMZ). L'étude du système de réception a permisd'obtenir les caractéristiques lasers optimales dansl'UV. Il faut donc générer 20 W à une cadencede 40 kHz dans l'UV avec des impulsions possé-dant une largeur spectrale proche de la centainede mégahertz. Mon travail de thèse vise à dévelop-per ce système laser UV. Nous avons fait le choixde concevoir un système laser type Master Oscil-lator Power Fiber Amplifier (MOPFA). Les fibresont l'avantage d'être intrinsèquement robustes auxvibrations en comparaison des systèmes lasers enespace libre. Cependant, elles sont très sensiblesaux effets non linéaires notamment, dans notrecas, la diffusion Brillouin stimulée. Pour outrepas-ser cet effet non linéaire, un peigne de fréquenceobtenu par modulation de phase est utilisé dans lesystème. Ce dernier permet, si ses caractéristiquessont bien choisies, d'augmenter drastiquement lapuissance seuil Brillouin tout en n'impactant la me-sure effectuée par le QMZ. De plus, afin de repous-ser encore les effets non linéaires, de nombreusesfibres à large cœur ont été testées et comparées.Cependant, deux phénomènes délétères liés à l'uti-lisation du peigne de fréquence dans les fibres ontété observés : l'auto-injection du Brillouin et le mé-lange à quatre ondes. Ces deux phénomènes ontfait l'objet d'une étude car ils peuvent avoir unimpact très important sur le système final. Fina-lement, un système d'amplification en espace librea été ajouté en sortie des amplificateurs fibrés afind'augmenter la puissance crête signal et donc fa-ciliter la génération de l'UV dans des cristaux nonlinéaires