Search

International audience; An experimental study has been carried out to describe the evolution of the failure load relative to the maximum misalignment angle of curved carbon–epoxy laminates with out-of-plane wrinkle defects for a large range of misalignments. L-angle specimens with different levels of fibre wrinkling were manufactured and tested in four-point bending tests. Digital Image Correlation (DIC) and infrared thermography were used to monitor the tests. They allowed the stress concentrations induced by the wrinkles to be identified, together with their influences on the failure load of the sample. In most cases, increases in wrinkling levels led to decreases in failure loads.

International audience; Lead-halide perovskite solar cells (PSCs) based on unusual semiconductors made with ions, have shown impressive improvement in photovoltaic performance in few years exceeding nowadays 25 % power conversion efficiency. However, PSCs suffer from a lack of stability and show significant hysteresis in current-voltage curves, which are impeding commercialization. We confirmed the importance of halide ion migration in the hysteresis effect which has direct consequences on device efficiency. Using impedance spectroscopy, in addition to the geometrical capacitance found at high frequency for fresh samples without bias, we observed a second capacitance at low frequency after ageing or under bias. This second capacitance is interpreted as a charge accumulation layer at interfaces, which can be promoted by the presence of grain boundaries. Through glow-discharge optical emission spectroscopy elemental depth profiles, we found that under dark conditions, iodide ions diffuse through the electron transport layer versus ageing time. These ions interact chemically with the front-end electrode after four weeks and form silver iodide.

International audience; To decrease development costs of aeronautical composite structures, industrials and academic researchers are turning to “virtual testing” methods. To help achieve this objective, a new methodology has been developed at the Institut Clément Ader. It is based on a complex loading test rig for technological specimens (more representative than the usual level of coupons) and the Discrete Ply Model (DPM), a model allowing the main damage occurring in composite structures to be simulated. This research focuses on the combined loading of CFRP plates after impact. In a previous work, plates were experimentally impacted and loaded after the impact in the VERTEX test rig and a methodology to transfer boundary conditions to the numerical model was validated using a simple model without damage. In the work presented here, the DPM was used to simulate both the impact and the complex loading after impact tests. Numerical results and experimental results showed satisfactory correlation. The research undertaken demonstrates that the impact damage has very little influence when the tests are performed at larger scales than coupons. It is therefore paramount to use a pyramidal approach to design and test composite structures, in order to avoid oversizing.

International audience; Composite failure phenomena remain complex and lead to over-size structures in many industries, involving long and costly experimental campaigns. Numerical approaches are a good alternative to decrease sizing costs. The Discrete Ply Model developed in Institut Clément Ader over these last ten years shows good results in simulation of impact and compression after impact of composite laminates, ply drop-offs, open hole tensile tests… But the present approach leads to certain limitations for complex stacking sequences and induces unwanted characteristic lengths between consecutive plies. A user element for delamination calculation is developed that allows to define an interface corresponding to the overlapping zone between the two volume elements of upper and lower plies, whereas this overlapping zone is not directly defined by nodes of the mesh (volume elements have non-coincident nodes). The present paper details the new interface element implementation, its validation from DCB and ENF test simulations, and then its improvement is discussed.

International audience; The development and certification of aeronautical composite structures is still largely based on the pyramid of tests. This approach is extremely costly in terms of number of tests, especially at the level of coupons. Moreover, these tests are highly conservative, under uniaxial loading, and do not represent the actual behavior at structure scale. To overcome these drawbacks, a new methodology has been developed at the Institut Clément Ader, which uses a complex loading test rig for technological specimens. This research focuses on the combined loading after impact of CFRP plates and highlights a specific behavior quite different from the usual CAI (Compression After Impact) response at the scale of coupons. In particular, compression, shear and combined shear/compression loadings were applied to large Carbon Fiber Reinforced Plastics (CFRP) laminated plates and the interaction of the impact damage with the post-buckling behavior has been investigated.

International audience; Highly curved laminated parts are used at the junction between two different perpendicular panels on aircraft primary structures. Usually, these laminates fail by delamination due to a bending moment, which appears when the part is loaded. The bending moment tries to flatten the part and out-of-plane tensile stresses are generated in the curvature. This failure is traditionally called unfolding failure. The modelling strategy called the 'Discrete Ply Model' (DPM) is used to simulate four-point bending tests on L-angle specimens. Experimental results of four point bending tests carried out at ONERA are used to validate the approach. Four different stacking sequences with the same thickness are taken into consideration in this study. In a second part, a sensitivity analysis on frictional coefficient, intralaminar matrix cracking, transverse tensile and shearing strength, and critical energy release rate in modes I and II is performed numerically and provides an original explanation of the failure scenario and the most influential parameter.

International audience; Composites are known to be vulnerable to out-of-plane loading such as impact. Investigating laminate's residual properties in function of damage detection is the main purpose of impact damage tolerance design in aeronautics. As a good alternative to experimental campaigns, numerical approaches would lead to saving of time. The model developed in Institut Clément Ader over the last years enables to represent the behavior of composite laminates subjected to low velocity/low energy impact - including permanent indentation - and Compression After Impact. Damages such as permanent indentation, fiber failures, matrix cracks and delamination are taken into consideration at each step thanks to a discrete ply modelling. The work presented here deals with the use of this model to make a composite laminate design optimization according to impact damage tolerance design. A method to improve optimization by reducing computation time is also proposed, based on a “best candidates” selection.

International audience; Impact damage tolerance of hybrid carbon and glass fibers woven-ply reinforced PolyEther Ether Ketone (PEEK) thermoplastic (TP) laminates obtained by consolidation process is investigated. Service temperature being one of the most important parameters to screen TP or thermosetting matrix for aeronautical purposes, impact testing at room temperature (RT) and near the glass transition temperature (Tg) has been conducted. From the results, it turns out that temperature has little influence on the impact behavior in terms of maximum force developed or maximum deflection, though it reduces the dissipated energy especially at lower impact energy. However, temperature has a strong effect on the internal and external damages caused to the plate, as it increases the permanent indentation and it limits the projected delaminated area. As for the influence of temperature on the compressive residual strength of the laminates, it also appears that the classical experimental set-up for CAI tests is not completely appropriate to draw a clear cut conclusion. At last the obtained results show that the considered TP-based laminates are characterized by a very good impact behavior and a high-degree of damage tolerance.

International audience

International audience; Three-dimensional (3D) real-time object detection and tracking is an important task in the case of autonomous vehicles and road and railway smart mobility in order to allow them to analyze their environment for navigation and obstacle avoidance purposes. In this paper, we try to improve the efficiency of 3D monocular object detection by using dataset combination, knowledge distillation, and creating a lightweight model. Firstly, we combine real and synthetic datasets to increase the diversity and richness of the training data. Then, we use knowledge distillation to transfer the knowledge from a large, pre-trained model to a smaller, lightweight model. Finally, we created a lightweight model by selecting the combinations of width, depth and resolution in order to reach a target complexity and computation time. Our experiments show that using each method improves either the accuracy or the efficiency of our model with no significant drawbacks. Using all these approaches is especially useful for resource-constrained environments, such as self-driving cars and railway systems.; La détection et le suivi d'objets tridimensionnels (3D) temps-réel est une tâche importante dans le cas des véhicules autonomes et de la mobilité intelligente routière et ferroviaire, afin de leur permettre d'analyser leur environnement à des fins de navigation et d'évitement d'obstacles. Dans cet article, nous essayons d'améliorer l'efficacité de la détection d'objets 3D monoculaire en utilisant la combinaison de jeux de données, la distillation de connaissances et la création d'un modèle léger. Tout d'abord, nous combinons des ensembles de données réelles et synthétiques pour augmenter la diversité et la richesse des données d'entraînement. Ensuite, nous utilisons la distillation des connaissances pour transférer les connaissances d'un grand modèle pré-entraîné vers un modèle plus petit et léger. Enfin, nous avons créé un modèle léger en sélectionnant les combinaisons de largeur, de profondeur et de résolution afin d'atteindre une complexité et un temps de calcul cibles. Nos expériences montrent que l'utilisation de chaque méthode améliore soit la précision, soit l'efficacité de notre modèle, sans inconvénient majeur. L'utilisation de toutes ces approches est particulièrement utile pour les environnements à ressources limitées, tels que les voitures à conduite autonome et les systèmes ferroviaires.

International audience; Oxidative desulfurization (ODS) was applied to three marine fuels with viscosities between 380 and 700 cSt (centistokes) and sulfur contents between 0.79 and 3.27 wt.%, using a MoO3/Al2O3 catalyst. H2O2 was chosen as oxidizing agent with a low oxidant to sulfur ratio of 3, for economic and environmental reasons. Global sulfur removal was obtained after extraction of sulfones by dimethylformamide. Oxidation efficiency was evaluated by monitoring the conversion of alkyl-benzothiophenes and dibenzothiophene molecule by gas chromatography (GC) with a sulfur specific detector. Applying ultrasounds prior to oxidation was found to significantly enhance conversions and sulfur removal, the effect being more pronounced when initial efficiency was low, that is for more refractory molecules and heavier fuels. Ultrasound assisted ODS (UAOD) was effective on raw fuels with initial viscosity below 700 cSt. Indeed, dilution decreased the viscosity of the fuels, allowing reaction to be effective even on initially very heavy fuels. The presence of asphaltenes dissolved or precipitated in the feed did not affect the conversion of Cx-BTs and DBT significantly, with only slight variations of global sulfur removal rates. Global desulfurization rates between 60 and 80 % were obtained, clearly showing the efficiency of UAOD applied to complex matrices such as marine fuels.

International audience; Des sollicitations multiaxiales sur structures stratifiées composites impactées sont réalisées à l'aide d'un banc d'essais développé à l'Institut Clément Ader (Toulouse). Une modélisation ElémentsFinis ,le « Discrete Ply Model », est utilisée pour simuler le comportement à rupture de ces plaques composites. Les calculs explicites prennent en compte à la fois la sollicitation d'impact et le chargement multiaxial post-impact. Un suivi par caméras optiques et infrarouge est mis en place afin de valider les simulations numériques.

International audience; This study evaluated the potential of using poly(lactic acid)/poly(ε-caprolactone) (PLA/PCL) blends for fused filament fabrication (FFF) and assembly with pure PLA for biomedical applications. PLA/PCL binary blends were meltblended in a twin-screw extruder at different ratios (20/80 to 80/20) and then formed into filaments with a calibrated diameter for FFF. The microstructure, surface properties, and rheological and mechanical behaviors of the blends were assessed. The blends were immiscible but showed signs of adhesion between the phases. It was determined that the fibrillar morphology of inclusions for PLA/PCL ratios higher than 30/70 proved to be driven by the manufacturing process. The tensile mechanical behaviors of printed and injected samples were similar, and their Young's modulus was simulated using Halpin-Tsai and Mori Tanaka models based on the sample microstructure. The ductility of the blends was strongly driven by the behavior of its majority phase. Finally, specific samples were designed to characterize the tensile strength between PLA and its blends by entangling layers of both materials. The strength of the assembly was found to be dependent on the phase that was continuous and was governed by the strength and the viscosity of the blend.

International audience; The management of French freshwater bodies (lakes, reservoirs, ponds, etc.) needs to face the challenges imposed by the effects of climate change and by legal requirements to develop standards for water temperature and other physicochemical indicators. While 1D hydrodynamical models could be helpful in improving the knowledge of the thermal and hydrodynamic behaviour of French water bodies, the scarcity of data for calibration and validation, in addition to the unavailability of some forcing data, makes their application difficult for most French water bodies. In this article we explore an alternative statistical approach that takes advantage of the available data in order to inform future modelling applications. We used official monitoring data and satellite measurements to study the thermal characteristics of more than 400 French water bodies (depth: 1–310 m, surface area: 0.1–577 km2, volume: 0.1–89,000 hm3). The objective was to identify the importance of size on their thermal behaviour. For this we analysed the annual temperature cycle, the summer temperature profiles and within-lake spatial variability. Together with size, geography and transparency determine the thermal behaviour of water bodies through complex interactions. As a result of the analysis we propose a classification of French water bodies as a function of depth and surface area that reflects the summer vertical temperature gradient and the processes influencing the deepening of the thermocline.

International audience; The development and certification of aeronautical composite structures is still largely based on the pyramid of tests. This approach requires a large number of tests going from the coupon to the full-size aircraft. To reduce the associated costs, predictive modelling methodologies are developed in several research centers [1]-[4]. Out of plane stresses emanating from impacts are well known to be critical for the residual compressive strength of laminated composites. To design composites structures, engineers use a damage tolerance approach following standards of low velocity impact and compression after impact tests [1], [2]. These standardized tests do not represent real structural loadings because of their conservative uniaxial boundary conditions that do not allow post-buckling failures. In this work, medium-scaled specimens are impacted at medium velocities with a gas launcher then multi-axially loaded in order to analyze interactions between impact damage and postbuckling.

International audience; Numerical and experimental research programs have been carried out to investigate the effect of scaling on the tensile strength of notched composites. This paper presents a computational study of scaled open-hole tensile tests using the Discrete Ply Model (DPM) method. This finite element model is discrete, and only a small number of parameters are required from experimental characterization tests. Experimental and numerical strength values are compared here, and reveal that DPM simulations tend to slightly overestimate strength values, with an average discrepancy of 9.7%. However, DPM Results show that such modeling simulates both the reduction in strength when specimen size is increased for sublaminate level scaled specimens, where failure is fiber dominated, and the increase in strength when specimen size is increased for ply level scaled specimens, where failure is delamination dominated. In all cases, increasing the total thickness of the specimen leads to a decrease in strength and this effect is dominant over the effect of increasing hole diameter. As well as the variation in strength, three distinct failure mechanisms are observed: fiber failure with extensive matrix damage (pull-out failure), fiber failure with little or no matrix damage (brittle failure) and delamination failure. Comparisons with experiments demonstrate that tensile strengths, damage propagation scenarios and failure patterns are predicted with acceptable accuracy.

International audience; Within the challenging race for alternative energy sources, hybrid metal halide perovskite solar cells (PSCs) have undergone unprecedented progress with efficiencies reaching now 25.5% [1-3]. These remarkable performances result from the exceptional optoelectrical properties of hybrid perovskite materials. Coupled with their potential for low fabrication cost, perovskite solar cell technology is very promising. However, since hybrid halide perovskites have a highly ionic character, they can decompose under external stresses such as moisture, solvents and heating cycles [4-6]. Reducing environmental stresses imposed by moisture or oxygen for example, in order to improve the long-term stability of perovskite solar cells, is critical to the deployment of this technology.In 2017, Grancini et al [7] published a structure proven to be stable for more than 10,000 hrs, measured uned controlled standard conditions, by engineering an ultra-stable 2D/3D (HOOC(CH2)4NH3)2PbI4/CH3NH3PbI3 perovskite junction. This structure is based on a fully printable architecture made of three mesoporous layers in which the perovskite is embed.In this communication, we investigate the effects of changing the perovskite composition in this architecture by adding and/or replacing iodine with bromine into its chemical composition to tune device, increasing the band gap of the material [8], opening the gate to potential application in water splitting (VOC > 1.23 V at pH = 0 [9]). After characterization of their photovoltaic properties, the cells, reaching VOC ~1.3 V are studied by SEM coupled to EDX, XRD, Raman, absorption and photoluminescence spectroscopies.1. Green, M. A., Ho-Baillie, A. & Snaith, H. J. The emergence of perovskite solar cells. Nature Photonics 8, 506–514 (2014).2. Green, M. A. & Ho-Baillie, A. Perovskite Solar Cells: The Birth of a New Era in Photovoltaics. ACS Energy Lett. 2, 822–830 (2017).3. Best Research Cell Efficiencies. NREL https://www.nrel.gov/pv/cell-efficiency.html (2021).4. Berhe, T. A. et al. Organometal halide perovskite solar cells: degradation and stability. Energy Environ. Sci. 9, 323–356 (2016).5. Leijtens, T. et al. Stability of Metal Halide Perovskite Solar Cells. Adv. Energy Mater. 5, 1500963 (2015).6. Rong, Y., Liu, L., Mei, A., Li, X. & Han, H. Beyond Efficiency: the Challenge of Stability in Mesoscopic Perovskite Solar Cells. Adv. Energy Mater. 5, 1501066 (2015).7. Grancini, G. et al. One-Year stable perovskite solar cells by 2D/3D interface engineering. Nature Communications 8, 15684 (2017).8. Yu, W. et al. Diversity of band gap and photoluminescence properties of lead halide perovskite: A halogen-dependent spectroscopic study. Chemical Physics Letters 699, 93–98 (2018).9. Walter, M. G. et al. Solar Water Splitting Cells. Chem. Rev. 110, 6446–6473 (2010).

International audience

International audience; In this article we investigate the potential of using visuo-haptic illusions in Virtual Reality environment to learn motor skills in a real environment. We report on an empirical study where 20 participants perform a multi-object pick-and-place task. The results show that although users do not perform the same motion trajectories in the virtual and real environments, skills acquired in VR augmented with visuo-haptic illusions can be successfully reused in a real environment: There is a high amount of skill transfer (78.5%), similar to the one obtained in an optimal real training environment (82.4%); Finally, participants did not notice the illusion and were enthusiastic about the VR environment. Our findings invite designers and researchers to consider visuo-haptic illusions to help operators to learn motor skills in a cost-effective environment.

International audience

International audience

COnsensus DEgenerate Hybrid Oligonucleotide Primers (CODEHOP) were developed for the detection of the dsz B desulfinase gene (2′-hydroxybiphenyl-2-sulfinate desulfinase; EC 3.13.1.3) by polymerase chain reaction (PCR), which allow to reveal larger diversity than traditional primers. The new developed primers were used as molecular monitoring tool to drive a procedure for the isolation of desulfurizing microorganisms. The primers revealed a large dsz B gene diversity in environmental samples, particularly in diesel-contaminated soil that served as inoculum for enrichment cultures . The isolation procedure using the dibenzothiophene sulfone (DBTO 2) as sole sulfur source reduced drastically the dszB gene diversity. A dszB gene closely related to that carried by Gordonia species was selected. The desulfurization activity was confirmed by the production of desulfurized 2-hydroxybiphenyl (2-HBP). Metagenomic 16S rRNA gene sequencing showed that the Gordonia genus was represented at low abundance in the initial bacterial community. Such observation highlighted that the culture medium and conditions represent the bottleneck for isolating novel desulfurizing microorganisms. The new developed primers constitute useful tool for the development of appropriate cultural-dependent procedures, including medium and culture conditions, to access novel desulfurizing microorganisms useful for the petroleum industry.

International audience; Point cloud acquisition techniques are an essential tool for the digitization of industrial plants, yet the bulk of a designer's work remains manual. A first step to automatize drawing generation is to extract the semantics of the point cloud. Towards this goal, we investigate the use of deep learning to semantically segment oil and gas industrial scenes. We focus on domain characteristics such as high variation of object size, increased concavity and lack of annotated data, which hampers the use of conventional approaches. To address these issues, we advocate the use of synthetic data, adaptive downsampling and context sharing.

International audience

Based on a multidisciplinary and configurable modeling approach, this work deals with the optimal choice and the design of a hybrid propulsion with the associated power management strategy to replace a conventional propulsion in a low tonnage river ferry operating on short cycles, with the aim of reducing its environmental impact and the costs over its entire lifetime. The considered ferry is used for the transport of people and vehicles crossing the Seine river, with an installed propulsive power of 330 KW. The operating cycle of the ferry and the energy consumption of its classical propulsion have been determined experimentally and then used as references in simulations for validation and comparison purposes. Two hybrid structures involving the use of batteries and supercapacitors were proposed and compared. It is shown that the hybridization leads to a substantial reduction in CO2 emissions. The supercapacitor- and battery-based hybrid structures lead respectively to 18% and 29.7% CO2 reduction compared to classical propulsion, representing, respectively, about 382 and 626 tons of CO2 reduction over 20 years of operation. Despite the fact that the use of batteries leads to a more significant reduction in CO2 emissions, the solution using supercapacitors is chosen following a technical-economic study over 20 years of operation.

International audience; The development and certification of aeronautical composite structures is still largely based on the pyramid of tests. This approach requires a large number of tests going from the coupon to the full-size aircraft. To reduce the associated costs, predictive modelling methodologies are developed in several research centers [1]-[4]. Out of plane stresses emanating from impacts are well known to be critical for the residual compressive strength of laminated composites. To design composites structures, engineers use a damage tolerance approach following standards of low velocity impact and compression after impact tests [1], [2]. These standardized tests do not represent real structural loadings because of their conservative uniaxial boundary conditions that do not allow post-buckling failures. In this work, medium-scaled specimens are impacted at medium velocities with a gas launcher then multi-axially loaded in order to analyze interactions between impact damage and postbuckling.

conférence en visio; International audience

International audience

International audience; Compressed air energy storage systems (CAES) have demonstrated the potential for the energy storage of power plants. One of the key factors to improve the efficiency of CAES is the efficient thermal management to achieve near isothermal air compression/expansion processes. This paper presents a review on the Liquid Piston (LP) technology for CAES as a timely documentary on this topic with rapidly growing interests. Various aspects are discussed including the state-of-the-art on LP projects all over the world and the trend of development, the coupled fluid flow and heat transfer during the compression/expansion operations, and different actions proposed and implemented to enhance the heat transfer inside the piston column.It has been found that LP is a promising concept for isothermal CAES. However, the complex and transient fluid flow and heat transfer behaviors inside the LP are difficult to characterize and master. To enhance the heat transfer and increase the efficiency of the compression/expansion processes many approaches have been tested including liquid spray, wire mesh, porous media, optimal trajectory, hollow spheres and optimal geometry of the piston column. Numerous empirical correlations of Nusselt number have also been proposed to estimate the heat transfer in different types of LP, as reviewed and summarized in this paper.

International audience; A successful implementation of Model Predictive Control (MPC) requires appropriately tuned parameters. In This paper an Artificial Neural Network (ANN) based approach is presented. To build the data learning base of the ANN, we adopted the Particle Swarm Optimisation (PSO) method, and we used an Online Sequential Extreme-Learning-Machine (OS-ELM) algorithm to learn the ANN. The objective of this work is to to ensure robustness of the closed-loop system against neglecting dynamics, disturbances and sensor noises, and also show that good tuning of MPC parameters allows to reach closed-loop stability without using robustification approaches in addition to MPC. The effectiveness of our approach has been emphasized by comparing the obtained performances to Generalized Predictive Control (GPC) approach without robustification approach, and also to a robustified (GPC) using Youla parametrisation in the presence of disturbances.

International audience; A successful implementation of Model Predictive Control (MPC) requires appropriately tuned parameters. This paper presents a novel tuning approach based on Artificial Neural Network (ANN). To build the data learning base of the ANN, we adopted the Particle Swarm Optimisation (PSO) method, and we used the reliable algorithm, Online Sequential Extreme-Learning-Machine (OS-ELM) to learn the ANN. The objective of this work is to show that good tuning of MPC parameters makes it possible to reach closed-loop stability and ensure robustness against disturbances and sensor noises, without using robustification approaches in addition to MPC. The effectiveness of our approach is brought to light by comparing the obtained performances to other MPC tuning approaches without disturbances, and also to a robustified Generalized Predictive Control (GPC) using Youla parametrisation in the presence of disturbances.

Transferring as fast as possible the functioning of our brain to artificial intelligence is an ambitious goal that would help advance the state of the art in AI and robotics. It is in this perspective that we propose to start from hypotheses derived from an empirical study in a human-robot interaction and to verify if they are validated in the same way for children as for a basic reinforcement learning algorithm. Thus, we check whether receiving help from an expert when solving a simple close-ended task (the Towers of Hano\"i) allows to accelerate or not the learning of this task, depending on whether the intervention is canonical or requested by the player. Our experiences have allowed us to conclude that, whether requested or not, a Q-learning algorithm benefits in the same way from expert help as children do., Comment: Accepted at Brain-PIL Workshop - ICRA2020

Controlling phosphorus is fundamental to limit the risk of eutrophication of continental aquatic ecosystems. Integrated modelling of its concentration in the aquatic continuum requires specific tools for water bodies. However, although simple static empirical models and complex biogeochemical models are numerous, there are few relatively simple and flexible models able to simulate seasonal variations in phosphorus concentrations in water bodies and particularly in reservoirs. In this study, the two-layer dynamic model, LakeMab, simulating phosphorus variations in water bodies, was enhanced to consider some tributary characteristics and reservoir specificities. It was then applied to the case of a reservoir in western France, Lake Bultiere. Without any calibration, the modified model reproduced reasonably well seasonal variations in phosphorus concentration in the lake. A sensitivity analysis showed the importance of improvements related to reservoir functioning (outlet depth, water level fluctuations) and the smaller importance of those related to tributaries (variable concentrations, depth of riverine inputs). The model can easily be applied to diverse lentic systems and could be coupled to stream models, thereby making it a useful tool for managing water quality in lakes and reservoirs.

International audience; Metamodeling allows to capture domain knowledge through the definition of its structure (concepts and relations between them) and its constraints (logical expressions) often written in OCL. The OCL constraints added to a metamodel are of two types: 1) domainrelated constraints: they differ from one domain to another and are expressed based on the knowledge of experts; and 2) those that are added to the majority of metamodels to precise some inaccurate structures that may cause problems when instantiating models. We call these structures Metamodel Inaccurate Structures (MIS). In this paper, we performed an empirical study in order to point out the metamodel inaccurate structures. As a first step, a study is conducted on a set of OCL constraints taken from the UML metamodel to investigate the relation between structure and constraints. Then, to confirm our findings, we realised a quantitative analysis in order to count the occurrences of constraints that complete inaccurate structures. We believe that our results can help designers in the quest of creating or refactoring metamodels and specifying constraints that precisely capture domain knowledge to ensure consistency of the derived artifacts.

Using a color code on human-machine interfaces could sensibly reduce the informational density. In this paper, we focused on the supervision interfaces used in engine room of Merchant Navy vessels. There is no regulation about color use on these interfaces, despite the color code playing an important role in showing information. The ISO Standard 14726-2008 regulates the color code printed on physical pipes and is used to represent the fluids flowing through it. The purpose of this study is to determine if we could use this standardized color code on the interfaces and create a connection between the physical pipes and their graphical representation. Two experimentations are presented, in which we compared the effects of two color codes on the performance at searching tasks carried out by novices and experts. One code is extracted from ISO Standard 14726-2008 (Normative Pipe Identification Color Code), and the other is created by taking into account general guidelines and common uses (Ergonomic Recommendations Color Code). The results showed that the NPICC, and its bicolor version in particular, increased the mental workload and the searching time required by novices. Experts had more difficulties to recognize the fluids represented by two colors and they used the two-colored fluid system less frequently than novices to retrieve an information.

International audience; Through the artificial manipulation of hydrological residence time, hydropower operations on alpine lakes should not only alter their hydro-biogeochemical characteristics but also modulate the degree of sensitivity of the lake to atmospheric forcing. These hypotheses were addressed by means of in situ observations that were combined with a three-dimensional lake model of a high-altitude pump-storage reservoir, i.e., Lake Corne, during the ice-free period. The fitted model (Pumping scenario) was used to simulate the hydrodynamics and oxygen concentrations of the lake under natural conditions (without pump-storage). Thereafter, the lake response to direct or catchment-mediated meteorological conditions (e.g., changes in inflows during storms) was investigated with a sensitivity analysis of the forcing parameters under both the pumping and natural hydrological scenarios. The pumping operation resulted in significant changes in water temperature, heat content and water mass stability during summer, as compared to a natural scenario. The lake hydrodynamics during the ice-free season were highly responsive to changes in the summer weather conditions, and more sensitive to realistic changes in cloudiness and water transparency than changes in air temperatures for both the pumped and natural scenarios. The spatial and temporal evolution of dissolved oxygen was comparatively less responsive to summer weather conditions, in both scenarios. The pump-storage operation from Lake Corne had a comparatively smaller effect on the lake functioning than the natural meteorological variability in summer. However, the pump-storage operation decreased the sensitivity of the lake hydrodynamics to changes in water transparency and limited water mass stability during summer.

Many high-performance models suffer from a lack of interpretability. There has been an increasing influx of work on explainable artificial intelligence (XAI) in order to disentangle what is meant and expected by XAI. Nevertheless, there is no general consensus on how to produce and judge explanations. In this paper, we discuss why techniques integrating connectionist and symbolic paradigms are the most efficient solutions to produce explanations for non-technical users and we propose a reasoning model, based on definitions by Doran et al. [2017] (arXiv:1710.00794) to explain a neural network's decision. We use this explanation in order to correct bias in the network's decision rationale. We accompany this model with an example of its potential use, based on the image captioning method in Burns et al. [2018] (arXiv:1803.09797)., Comment: Accepted at IJCAI19 Neural-Symbolic Learning and Reasoning Workshop (https://sites.google.com/view/nesy2019/home)

International audience; Composites are known to be vulnerable to out-of-plane loading such as impact. To design composite structures without performing long and costly experimental campaigns several numerical approaches have been developed. The model developed in Institut Clément Ader over the last year enables representation of main failures appearing in a composite laminate subjected to low velocity/low energy impact. The work presented here deals with the validation of this model to account for damages in composite laminates under medium velocity impact.; Les matériaux composites sont connus pour être vulnérables aux sollicitations hors plan comme les impacts. Pour s'abstenir de réaliser des campagnes expérimentales couteuses, de nombreux outils de simulations numériques sont développés. Le modèle développé à l'Institut Clément Ader depuis une dizaine d'années permet de représenter les principaux endommagements présents dans un stratifié composite lors d'un impact à faible vitesse/faible énergie. Le travail présenté ici a pour but de valider le bon fonctionnement du modèle pour des impacts à moyenne vitesse.

National audience; Thanks to their low mass and high performances, composites draw an increasing interest in many fields, such as aeronautics. Nevertheless, composites structures are particularly vulnerable to out-of-plane solicitations. Research [1-3] have shown differences between woven and unidirectional composites. Woven composites have better damage tolerances thanks to textile architecture and mechanisms of effort diffusion [4]. Moreover, they have better properties on out-of-plan solicitations and a greater dissipative power than unidirectional composites. Woven composites have shown a particular interest for aeronautical manufacturers, yet their breaking mechanisms remain complex.This study is focused on failure mechanisms of woven composite. We used the numerical model which has developed at the Institute Clement Ader over the last years called Discrete Ply Model (DPM), with the aim of simulating tensile tests on notched and unnotched specimens of woven composites. The objective is to study translaminar crack growth propagations and improve the understanding of failure mechanism. We take into consideration the critical energy release rate of the fiber failure in mode I and II, the shearing plasticity and the effect of weaving point.; Grâce à leurs faibles masses et leurs hautes propriétés mécaniques, les composites sont des matériaux de plus en plus utilisés dans de nombreux domaines, notamment dans l’aéronautique. Néanmoins, les structures composites sont particulièrement vulnérables vis-à-vis des sollicitations hors plan. Les recherches [1-3] ont montré des différences significatives entre composites unidirectionnels et tissés. Grâce à leurs architectures et aux mécanismes de diffusion des efforts, les tissés possèdent de meilleures propriétés en sollicitation hors plan que les unidirectionnels [4]. Ils présentent donc un intérêt non négligeable pour les constructeurs aéronautiques ; cependant leurs mécanismes de rupture restent complexes. Dans ce travail nous nous proposons d’étudier les mécanismes de ruptures des composites tissés. Pour cela nous avons utilisé le modèle numérique développé à l’Institut Clément Ader depuis une dizaine d’années, le Discrete Ply Model (DPM), afin de simuler des essais de traction sur éprouvettes tissées avec et sans pré-entaille. Le but est d’étudier la propagation de fissures translaminaires et d’améliorer la compréhension des mécanismes d’endommagements. Nous nous sommes notamment intéressés à la prise en compte du taux de restitution d’énergie critique en rupture de fibres en mode I et II, à la plasticité en cisaillement ainsi qu’à l’effet des points de tissages.

International audience; Thanks to their low mass and high performances, composites drawan increasing interest in many fields, such as aeronautics. Nevertheless, composites structures are particularly vulnerable to out-of-plane solicitations.Research [1-3] have shown differences between woven and unidirectional composites. Woven composites have better damage tolerances thanks to textile architecture and mechanisms of effort diffusion [4]. Moreover, they have better properties on out-of-plan solicitations and a greater dissipative power than unidirectional composites.Woven composites have shown a particular interest for aeronautical manufacturers, yet their breaking mechanisms remain complex.This study is focused on failure mechanisms of woven composite. We used the numerical model which has developed at the Institute Clement Ader over the last years called Discrete Ply Model (DPM), with the aim of simulating tensile tests on notched and unnotched specimens of woven composites. The objective is to study translaminar crack growth propagations and improve the understanding of failure mechanism. We take into consideration the critical energy release rate of the fiber failure in mode I and II, the shearing plasticity and the effect of weaving point.; Grâce à leurs faibles masses et leurs hautes propriétés mécaniques, les composites sont des matériaux de plus en plus utilisés dans de nombreux domaines,notamment dans l’aéronautique. Néanmoins, les structures composites sont particulièrement vulnérables vis-à-vis des sollicitations hors plan.Les recherches [1-3] ont montré des différences significatives entre composites unidirectionnels et tissés. Grâce à leurs architectures et aux mécanismes de diffusion des efforts,les tissés possèdent de meilleures propriétés en sollicitation hors plan que les unidirectionnels[4]. Ils présentent donc un intérêt non négligeable pour les constructeurs aéronautiques;cependant leurs mécanismes de rupture restent complexes.Dans ce travail nous nous proposons d’étudier les mécanismes de ruptures des composites tissés.Pour cela nous avons utilisé le modèle numérique développé à l’Institut Clément Ader depuis une dizaine d’années, le Discrete Ply Model (DPM), afin de simuler des essais de traction sur éprouvettes tissées avec et sans pré-entaille. Le but est d’étudier la propagation de fissures translaminaires et d’améliorer la compréhension des mécanismes d’endommagements. Nous nous sommes notamment intéressés à la prise en compte du taux de restitution d’énergie critique en rupture de fibres en mode I et II, à la plasticité en cisaillement ainsi qu’à l’effet des points de tissages.

National audience

International audience

A fluid-structure interaction’s effects on the dynamics of a hydrofoil immersed in a fluid flow of non-homogeneous density is presented and analyzed. A linearized model is applied to solve the fluid-structure coupled problem. Fluid density variations along the hydrofoil upper surface, based on the sinusoidal cavity oscillations, are used. It is shown that for the steady cavity case, the value of cavity length Lp does not affect the amplitude of the hydrofoil displacements. However, the natural frequency of the structure increases according to Lp. In the unsteady cavity case, the variations of the added mass and added damping (induced by the fluid density rate of change) generate frequency and amplitude modulations in the hydrofoil dynamics. To analyse this phenomena, the empirical mode decomposition, a well established data-driven method to handle such modulations, is used.

International audience; Non-contact detection and estimation of vital signs such as respiratory and cardiac frequencies is a powerful tool for surveillance applications. In particular, the continuous wave bio-radar has been widely investigated to determine the physiological parameters in a non-contact manner. Since the RF-reflected signal from the human body is corrupted by noise and random body movements, traditional Fourier analysis fails to detect the heart and breathing frequencies. In this effort, cyclostationary analysis has been used to improve the radar performance for non-invasive measurement of respiratory rate and heart rate. However, the preliminary works focus only on one frequency and do not include the impact of attenuation and random movement of the body in the analysis. Hence in this paper, we evaluate the impact of distance and noise on the cyclic features of the reflected signal. Furthermore, we explore the assessment of second order cyclostationary signal processing performance by developing the cyclic mean, the conjugate cyclic autocorrelation and the cyclic cumulant. In addition, the analysis is carried out using a reduced number of samples to reduce the response time. Implementation of the cyclostationary technique using a bi-static radar configuration at 2.5 GHz is shown as an example to demonstrate the proposed approach.

In the last years, Artificial Intelligence (AI) has achieved a notable momentum that may deliver the best of expectations over many application sectors across the field. For this to occur, the entire community stands in front of the barrier of explainability, an inherent problem of AI techniques brought by sub-symbolism (e.g. ensembles or Deep Neural Networks) that were not present in the last hype of AI. Paradigms underlying this problem fall within the so-called eXplainable AI (XAI) field, which is acknowledged as a crucial feature for the practical deployment of AI models. This overview examines the existing literature in the field of XAI, including a prospect toward what is yet to be reached. We summarize previous efforts to define explainability in Machine Learning, establishing a novel definition that covers prior conceptual propositions with a major focus on the audience for which explainability is sought. We then propose and discuss about a taxonomy of recent contributions related to the explainability of different Machine Learning models, including those aimed at Deep Learning methods for which a second taxonomy is built. This literature analysis serves as the background for a series of challenges faced by XAI, such as the crossroads between data fusion and explainability. Our prospects lead toward the concept of Responsible Artificial Intelligence, namely, a methodology for the large-scale implementation of AI methods in real organizations with fairness, model explainability and accountability at its core. Our ultimate goal is to provide newcomers to XAI with a reference material in order to stimulate future research advances, but also to encourage experts and professionals from other disciplines to embrace the benefits of AI in their activity sectors, without any prior bias for its lack of interpretability., 67 pages, 13 figures, accepted for its publication in Information Fusion

Aeronautics search high performance materials for structural weight reduction and impact energy absorption. Composite offers perspectives related to mass and stiffness ratio. However, the dispersion of their mechanical properties due to environmental conditions or impact behaviour hinders their development. Currently, honeycomb solutions and complex energy absorption mechanisms are used. The sizing of structures is controlled for conventional loads, but not for severe cases such avian impact. In the context of the development of an innovative reinforced aeronautical structure, the objective of this study is to identify and characterize a new concept of absorbent composite material at the coupon scale. In order to improve reliability and optimize its absorption characteristics during high-energy impact, works are realized to develop in parallel an experimental study methodology and a finite element model. The purpose of the latter is to have a predictive tool validated by correlation with experimental to ensure virtual testing. Lack of knowledge of the nonlinear behaviour of the composite material at high deformation, for different speeds, as well as the mechanisms of damage and fracture are locks to its development. Absence of a dedicated experiment matrix presents a difficulty for its characterization and sizing. The approach begins with the analysis and prioritization of the existing to determine the definition criteria of the new concept. After development of an experimental characterization study, the qualification and justification of the new concept is validated by correlation with a dedicated numerical simulation methodology. The results of the study highlight the analysis and development of an interesting concept.

International audience; Le travail présenté vise à étudier la possibilité de récupérer de l'énergie lors du freinage d'un navire de faible tonnage équipé d'une chaîne de propulsion entièrement électrique ou hybride. Les modèles des différents éléments de la chaîne de propulsion sont décrits puis intégrés pour une simulation dynamique sous Matlab/Simulink. Ce modèle comprend un modèle d'hélice à quatre quadrants qui permet de prendre en compte la récupération d'énergie. Une étude de cas est considérée sur la base des spécifications d'une navette fluviale entièrement électrique. Plusieurs stratégies de contrôle du moteur de propulsion pendant le freinage sont étudiées. Les résultats mettent en évidence les avantages de l'application d'une stratégie MPPT (Maximum Power Point Tracking) qui conduit à une augmentation considérable de l'énergie récupérée.

International audience; L'article présenté s'inscrit dans le cadre du développement d'un système de stockage offshore sous forme d'air comprimé. Dans l'objectif de maximiser les performances énergé- tiques de la chaîne de conversion, une approche déterministe de dimensionnement de la machine électrique (Machine Synchrone à Aimants Permanents, MSAP) est appliquée. Elle vise à optimiser à la fois la géométrie de la machine et la stratégie de commande qui maximise l'efficacité énergétique sur cycle de l'ensemble convertisseur machine. Les résultats de l'optimisation, avec prise en compte du transitoire thermique, appliquée à un système de stockage en mer sont présentés dans ce papier. Enfin, l'influence du convertisseur sur le dimensionnement de la machine et la commande sont étudiées.

International audience; The six-port measurement technique has played a major role in microwave metrology, particularly within the community of national standards laboratories. The six-port reflectometer provides a simple method of implementing one-port and dual-six-port vector network analyzer. This paper reviews the development of six-port technology for metrology applications.

International audience; Extreme weather events may be just as important as gradual trends for the long‐term trajectories of ecosystems. For alpine lakes, which are exposed to both exacerbated atmospheric warming and intense episodic weather events, future conditions might not be appropriately forecast by only climate change trends, i.e., warming, if extreme events have the potential to deflect their thermal and metabolic states from their seasonal ranges. We used high‐frequency monitoring data over three open‐water seasons with a one‐dimensional hydrodynamic model of the high‐altitude Lake Muzelle (France) to show that rainstorms or windstorms, notwithstanding their intensity, did not trigger long‐lasting consequences to the lake characteristics when light penetration into the lake was not modified. In contrast, storms associated with high turbidity input from the watershed (“turbid storms”) strongly modified the lacustrine hydrodynamics and metabolism for the rest of the open‐water season through reduced light penetration. The long‐lasting effects of turbid storms were related to the inputs and in‐lake persistence of very‐light glacial suspensoids from the watershed. The occurrence of the observed turbid storms was not related to the wind or rain intensities during the events. Instead, the turbid storms occurred after dry and atypically warm spells, i.e., meteorological conditions expected to be more frequent in this alpine region in the upcoming decades. Consequently, storm events, notwithstanding their intensity, are expected to strongly imprint the future ecological status of alpine lakes under climate warming.