HAL Portal UTC Université de Technologie de Compiègne
Not a member yet
11652 research outputs found
Sort by
L’auto-efficacité des enseignantes et enseignants du supérieur à enseigner avec le numérique, de la période Covid à nos jours
No full textNational audienceWith the pandemic caused by Covid-19, the use of digital technologies has become imperative, resulting in highly diversified practices within the context of higher education. However, teaching remotely requires that instructors possess technical skills to effectively leverage digital tools to achieve their educational objectives. Drawing on Bandura’s social cognitive theory, this study employs a questionnaire survey of 290 instructors to validate a scale of self-efficacy in online teaching (SE) and to examine its relationships with attitudes, digital teaching practices, and the gender of respondents. After validating the unidimensional scale, the results reveal that a high SE is significantly correlated with a broader range of tools utilized and shows significant negative correlations with negative emotions and beliefs about digital technologies. The study also identifies a significant gender difference, with men exhibiting higher SE levels compared to women, despite the latter’s higher participation in digital training. In conclusion, SE emerges as a critical factor for instructors’ adaptation to the demands of digital technologies in higher education, highlighting the importance of training initiatives to enhance this aspect.Avec la pandémie causée par la Covid-19, l’usage du numérique s’est imposé et traduit par des pratiques très diversifiées dans le contexte de l’enseignement supérieur. Enseigner à distance suppose cependant que les personnes enseignantes (PE) détiennent des compétences techniques pour exploiter au mieux le numérique au service de leurs objectifs pédagogiques. Mobilisant la théorie sociocognitive de Bandura, cette étude utilise une enquête par questionnaire auprès de 290 PE pour valider une échelle d’auto efficacité à l’enseignement en ligne (AE) et examiner ses relations avec les attitudes, les pratiques d’enseignement à l’aide du numérique, ainsi que le sexe des répondants. Après validation de l’échelle unidimensionnelle, les résultats révèlent pour une AE élevée une corrélation positive et significative avec l’éventail d’outils mobilisés et des corrélations négatives (et significatives) avec les émotions et croyances négatives sur le numérique. L’étude identifie également une différence significative entre les sexes, les hommes ayant un niveau d’AE supérieur à celui des femmes, malgré une plus grande participation de ces dernières aux formations numériques. En conclusion, l’AE émerge comme un facteur déterminant pour l’adaptation des PE aux exigences du numérique en enseignement supérieur, soulignant l’importance des initiatives de formation pour renforcer ce dernier
Unraveling the Mechanisms of Hypertrophy-Induced Matrix Mineralization and Modifications in Articular Chondrocytes
No full textInternational audienceChondrocyte hypertrophic differentiation is a main event leading to articular cartilage degradation in osteoarthritis. It is associated with matrix remodeling and mineralization, the dynamics of which is not well characterized during chondrocyte hypertrophic differentiation in articular cartilage. Based on an in vitro model of progressive differentiation of immature murine articular chondrocytes (iMACs) into prehypertrophic (Prehyp) and hypertrophic (Hyp) chondrocytes, we performed kinetics of chondrocyte differentiation from Prehyp to Hyp to follow matrix mineralization and remodeling by immunofluorescence, biochemical, molecular, and physicochemical approaches, including atomic force microscopy, scanning electron microscopy associated with energy-dispersive X-ray spectroscopy (SEM–EDS), attenuated total reflection infrared analyses, and X-ray diffraction. Chondrocyte apoptosis was determined by TUNEL assay. The results show the formation of a mineral phase 7 days after Hyp induction, which spreads within the matrices to form poorly crystalline carbonate-substituted hydroxyapatite after 14 days, then the proportions of crystalline relative to amorphous content increases over time. Hyp differentiation also induced a matrix turnover that occurs over the first 7 days, characterized by a decrease in type II collagen and aggrecan and the concomitant appearance of type X collagen. This is accompanied by an increase in the enzymatic activity of MMP-13, the main collagenase in cartilage. The number of apoptotic chondrocytes slightly increased with Hyp differentiation and SEM–EDS analyses detected phosphorus-rich structures that could correspond to apoptotic bodies. Our findings highlight the mechanisms of matrix remodeling events leading to the mineralization of articular cartilage that may occur in osteoarthritis
Non-intrusive reduced order models for partitioned fluid–structure interactions
No full textInternational audienceThe main goal of this work is to develop a data-driven Reduced Order Model (ROM) strategy from high-fidelity simulation result data of a Full Order Model (FOM). The goal is to predict at lower computational cost the time evolution of solutions of Fluid–Structure Interaction (FSI) problems. For some FSI applications, the elastic solid FOM (often chosen as quasi-static) can take far more computational time than the fluid one. In this context, for the sake of performance one could only derive a ROM for the structure and try to achieve a partitioned FOM fluid solver coupled with a ROM solid one. In this paper, we present a data-driven partitioned ROM on two study cases: (i) a simplified 1D-1D FSI problem representing an axisymmetric elastic model of an arterial vessel, coupled with an incompressible fluid flow; (ii) an incompressible wake flow over a cylinder facing an elastic solid with two flaps. We evaluate the accuracy and performance of the proposed ROM-FOM strategy on these cases while investigating the effects of the model’s hyperparameters. We demonstrate a high prediction accuracy and significant speedup achievements using this strategy
Comparison between a delivery service of ready-to-cook ingredients and a meal prepared by a home helper for the elderly
No full textODD #12, #3, #10International audienc
Combined fungal and chemical pretreatment of lignocellulosic biomass for biogas production: Effect of pretreatment order and fungal strains
No full textInternational audienceThe recalcitrance of lignocellulosic poses a challenge for conversion to biogas and therefore, pretreatment is an essential step in addressing this. Combining fungal pretreatment with a biomimetic system, a mild Fenton reaction, was studied to improve the methane production from straw. The pretreatment conditions studied were able to reduce lignin content and cause an increase in the relative polysaccharides content. Interestingly, variations among the strains of the same species were observed in their ability to improve the biomethane yield. The order of pretreatment combination also played an important role as it affects the mechanism of action of subsequent pretreatment on the biomass, which indirectly affects the biogas yield
Amélioration des chaînes de traitement d’images satellites
No full textEncadrement - Pascal Mouquet - IRD (UMR Espace-Dev) Encadrement - Christophe Révillion - Université de La Réunion (UMR Espace-Dev) Encadrement - Didier Bouche - Université de La Réunion (DSI)Encadrement - Rodolophe Devillers - IRD (UMR Espace-Dev)Les stages en entreprise ou laboratoire font partie intégrante de la formation d’ingénieur de l’Université de Technologie de Compiègne (UTC).Dans ce cadre, j’ai réalisé un stage de 24 semaines à la station de Surveillance de l’Environnement Assistée par Satellite pour l’Ocean Indien (SEAS-OI). L’objectif principal était l’exploration de nouvelles fonctionnalités et le développement informatique en vue d’améliorer les chaînes de traitements de données spatiales existantes.Ce rapport est consacré au compte rendu de ce travail ; plus particulièrement aux missions qui m’ont été confiées, aux difficultés rencontrées et les solutions proposées ainsi qu’aux apprentissages et bilans acquis
Machine-Learning Enhanced Predictors for Accelerated Convergence of Partitioned Fluid-Structure Interaction Simulations
No full textStable partitioned techniques for simulating unsteady fluid-structure interaction (FSI) are known to be computationally expensive when high added-mass is involved. Multiple coupling strategies have been developed to accelerate these simulations, but often use predictors in the form of simple finite-difference extrapolations. In this work, we propose a non-intrusive data-driven predictor that couples reduced-order models of both the solid and fluid subproblems, providing an initial guess for the nonlinear problem of the next time step calculation. Each reduced order model is composed of a nonlinear encoder-regressor-decoder architecture and is equipped with an adaptive update strategy that adds robustness for extrapolation. In doing so, the proposed methodology leverages physics-based insights from high-fidelity solvers, thus establishing a physics-aware machine learning predictor. Using three strongly coupled FSI examples, this study demonstrates the improved convergence obtained with the new predictor and the overall computational speedup realized compared to classical approaches
Étude d'impact du prétraitement par D.I.C du Cannabis Sativa L.
No full textInternational audienceObjectivesInstant Controlled Pressure Drop (DIC) is an emerging agri-food technology that applies high-temperature, short-time treatments, altering material structures by expansion while preserving product quality. DIC shows promise for applications such as biomolecule drying, sterilization, and extraction, reducing energy costs and better preserving product quality compared to conventional processes. This research aims to determine the impact of DIC pretreatment on the extraction of essential oils and cannabidiol (CBD) from Cannabis Sativa L. buds. Objectives include decarboxylating the plant, improving extraction efficiency, preserving compounds, increasing storage stability, and exploring product and byproduct applications. This work investigates how to enhance process efficiency from harvest to product formulation by integrating DIC technology for hemp pretreatment, addressing scientific challenges, and offering higher-quality hemp products.MethodsDIC processing was performed using pilot equipment from ABCAR DIC Process (France). The DIC process involves increasing the pressure in the treatment chamber, maintaining it, and then depressurizing the chamber. Typical parameters include:• Pressure• Duration• Number of cyclesTreatment parameters were optimized based on the plant material and the target outcomes, aiming to decontaminate and decarboxylate the flowers, and improve the extraction of bioactive compounds. Parameters ranged from 1 to 35 seconds, 1 to 6 bars, and 1 to 8 cycles.CBD extraction kinetics were studied by dynamic maceration on samples of hemp buds:• Non-treated• Treated by DIC• Non-treated and steam distilled• Treated and steam distilledCBD yield was determined by gas phase chromatography analysis coupled with mass spectrometry (GC-MS).ResultsThe decarboxylation rate of CBDA to CBD was assessed by comparing concentrations of CBDA and CBD in the extracts. DIC treatment alone achieved a decarboxylation more than 80%. DIC treatment significantly enhanced the extraction kinetics of CBD. The conventional extraction method yielded 140.94 mg of CBD after 15 minutes, whereas with DIC pre-treatment, a similar amount (135.49 mg) was recovered in just 5 minutes. The aromatic profile of the essential oils, analyzed by GC-MS, revealed differences between oils extracted from DIC-treated buds and those obtained by traditional methods. Oils from DIC-treated buds demonstrated a fractionated extraction of volatile and non-volatile compounds. DIC treatment significantly improved drying efficiency, shortening the drying time to one-third of the traditional process. This results in reaching the target moisture content in just 640 minutes compared to the traditional 1920 minutes. The antimicrobial effectiveness of DIC was evaluated by comparing microbial colony counts (bacteria and fungi) on treated versus untreated hemp buds. DIC treatment reduced microbial counts by approximately 4 log (cfu/g of dried buds), highlighting DIC’s effectiveness for plant product decontamination and improved storage stability. Essential oils in the discharge waters of the DIC equipment formed stable emulsions, showing promise for bio-pesticide formulation. These emulsions, classified between nanoemulsions and macroemulsions, exhibited potential as fungicidal pesticides, insecticides, and nematicides, indicating additional applications for DIC-treated hemp flowers
De la structure à la dynamique d’un matériau fibreux
No full textFibrous materials, such as mineral wool, are commonly used in the acoustic insulation of various construction elements, like partitions and ceilings. The material investigated in this study is a network of random glass fibers partially bound by polymeric bonds. Despite its widespread use, the relationshipbetween its microstructure and acoustic behavior remains poorly understood. Previous studies have focused on macroscopic acoustic parameters, assuming a rigid fiber skeleton. However, this assumption is not accurate in modeling applications such as floating floors or partitions, where the displacement of the material structure must also be considered. Therefore, the main objective of this work is to study the dynamic behavior of mineral wool and to establish a link between its microstructure and macroscopic mechanical properties. Confocal microscope images and mesoscopic observations have revealed key properties of the material. Preliminary experiments confirmed that the material exhibits high anisotropy, even at small scales. The compressive stiffness across the thickness is relatively low, while fiber sheets were observed in the transverse direction. The polymer binder was found to be distributed randomly throughout the material. Macroscopic measurements of dynamic quantities, such as stiffness and loss factor, were performed on a set of model specimens designed for this purpose. The results demonstrated the influence of certain microscopic parameters on the dynamic behavior of the material. A simplified analytical model, based on existing research, was developed to explain the trends observed in the experimental results. This model identified the dominant mechanisms of energy dissipation. Simultaneously, a 3D finite element model was created to represent the material’s microstructure, using a geometry generation algorithm based on curved fibers. This model accounts for the dimensional variability of the material’s components and the polymer junctions, as well as their viscoelastic properties. Size effects are significant in this type of fibrous material model, making the extraction of homogeneous behavior very challenging. We have shown that the use of generalized boundary conditions helps limit these size effects and allows the numerical model to converge for a reasonable volume size. The many perspectives opened up by this work are also discussed, including potential applications for optimizing fibrous materials in acoustic insulation and further exploration of size effects in material modeling.Les matériaux fibreux, tels que les laines minérales, peuvent être utilisés pour l’isolation acoustique de différents éléments de construction (cloisons, plafonds...). Le matériau étudié est un réseau de fibres de verre aléatoires partiellement liées par des jonctions polymères dont la relation entre micro-structure et comportement acoustique est mal comprise. En effet, dans les travaux antérieurs, les paramètres acoustiques macroscopiques ont été appréhendés sous l’hypothèse d’un squelette fibreux rigide. Cette hypothèse s’avère inexacte pour modéliser certaines applications, telles que les planchers flottants ou les cloisons, lorsque le déplacement de la structure du matériau doit être pris en compte. L’objectif principal de cette thèse est donc d’étudier le comportement dynamique de la laine minérale, ainsi que d’établir un lien entre la micro-structure et les propriétés mécaniques macroscopiques. Des images obtenues au microscope confocal et des observations à l’échelle mésoscopique ont permis d’identifier certaines caractéristiques du matériau. Des expériences préliminaires ont confirmé qu’il est fortement anisotrope, même à petite échelle. La rigidité à la compression à travers l’épaisseur est relativement faible, alors que dans la direction transverse, des feuillets de fibres sont identifiés. Le liant polymère apparaît comme réparti très aléatoirement dans le matériau. Des mesures à l’échelle macroscopique des grandeurs dynamiques d’intérêt (rigidité, facteur de perte) en petites déformations ont été effectuées sur un lot d’échantillons modèles conçus à cette fin. Les résultats ont montré l’influence de certains paramètres microscopiques sur le comportement dynamique du matériau. Un modèle analytique simplifié a été développé sur la base de travaux existants. Ce modèle permet de retrouver les tendances observées par la mesure et d’identifier le phénomène dominant de la dissipation d’énergie. En parallèle des aspects expérimentaux, un modèle numérique éléments finis 3D représentatif de la micro-structure du matériau a été développé. Pour cela un algorithme de génération de géométrie basé sur des fibres courbes a été mis au point. Ce modèle prend en compte la variabilité des dimensions des composants du matériau et les jonctions crées par le polymère, ainsi que leur viscoélasticité. Les effets de taille sont très importants sur ce type de modèles de matériau fibreux et rendent l’extraction d’un comportement homogène très difficile. Nous avons montré que l’utilisation de conditions aux limites généralisées permet de limiter ces effets de taille et d’atteindre la convergence du modèle numérique pour un volume de taille raisonnable. Les nombreuses perspectives laissées ouvertes par ces travaux sont également discutées