Portail HAL edf
Not a member yet
11029 research outputs found
Sort by
Implementation of a liver health check in people with type 2 diabetes
No full textInternational audienceAs morbidity and mortality related to potentially preventable liver diseases are on the rise globally, early detection of liver fibrosis offers a window of opportunity to prevent disease progression. Early detection of non-alcoholic fatty liver disease allows for initiation and reinforcement of guidance on bodyweight management, risk stratification for advanced liver fibrosis, and treatment optimisation of diabetes and other metabolic complications. Identification of alcohol-related liver disease provides the opportunity to support patients with detoxification and abstinence programmes. In all patient groups, identification of cirrhosis ensures that patients are enrolled in surveillance programmes for hepatocellular carcinoma and portal hypertension. When considering early detection strategies, success can be achieved from applying ad-hoc screening for liver fibrosis in established frameworks of care. Patients with type 2 diabetes are an important group to consider case findings of advanced liver fibrosis and cirrhosis, as up to 19% have advanced fibrosis (which is ten times higher than the general population) and almost 70% have non-alcoholic fatty liver disease. Additionally, patients with type 2 diabetes with alcohol use disorders have the highest proportion of liver-related morbidity of people with type 2 diabetes generally. Patients with type 2 diabetes receive an annual diabetes review as part of their routine clinical care, in which the health of many organs are considered. Yet, liver health is seldom included in this review. This Viewpoint argues that augmenting the existing risk stratification strategy with an additional liver health check provides the opportunity to detect advanced liver fibrosis, thereby opening a window for early interventions to prevent end-stage liver disease and its complications, including hepatocellular carcinoma
Quantification of microscale factors for fatigue failure in NiTi shape memory alloys
No full textInternational audienceFatigue behavior is intrinsically linked to microstructural alterations induced by cyclic loading. However, the quantification of microstructural defects associated with fatigue damage of NiTi shape memory alloys (SMAs) is lacking, which hinders the development of a physically based fatigue criterion. To this end, a multi-scale experimental analysis was conducted on cyclically deformed NiTi SMAs, which evidenced a strong correlation between microstructural inhomogeneity and localized deformation behavior. The microstructural change associated with fatigue was quantified in terms of stored strain–energy, with the highest values observed in the regions where fatigue cracks initiate. Consequently, stored energy is deemed as an effective fatigue indicator, offering valuable insights for future work in the design and optimization of SMAs’ structures against fatigue
Whispering Gallery Modes and Frequency Combs: Two excursions in the world of photonic resonators
No full textInternational audienceIn photonics, resonators are versatile elements designed with the aim of guiding light. They encompass various orders of magnitude for theirsizes and the power of their laser sources. Accordingly some effects can be considered as negligible, leading to simplified mathematical models. We will focus on two such models: one related to Whispering Gallery Mode resonators and the other to ring/Fabry-Perot resonators. In the first case the retained model leads to a two-dimensional linear Helmholtz equation in the whole space with material laws having a jump along a bounded interface. The set of complex resonances is analyzed close to the real axis, corresponding to modes concentrated close to the interface. In the second case a one-dimensional model with cubic nonlinearity based on the LugiatoLefever equation is considered. Branches of stationary solutions issued from flat solutions are highlighted, providing frequency comb solutions
L'apprentissage profond inspiré de la physique est une méthode à noyau
No full textInternational audiencePhysics-informed machine learning combines the expressiveness of data-based approaches with the interpretability of physical models. In this context, we consider a general regression problem where the empirical risk is regularized by a partial differential equation that quantifies the physical inconsistency. We prove that for linear differential priors, the problem can be formulated as a kernel regression task. Taking advantage of kernel theory, we derive convergence rates for the minimizer of the regularized risk and show that it converges at least at the Sobolev minimax rate. However, faster rates can be achieved, depending on the physical error. This principle is illustrated with a one-dimensional example, supporting the claim that regularizing the empirical risk with physical information can be beneficial to the statistical performance of estimators.L'apprentissage automatique informé par la physique combine l'expressivité des approches basées sur les données avec l'interprétabilité des modèles physiques. Dans ce contexte, nous considérons un problème de régression où le risque empirique est régularisé par une équation aux dérivées partielles (EDP) qui quantifie l'incohérence physique. Nous prouvons que pour les EDP linéaires, le problème peut être formulé comme une de régression à noyau. En tirant parti de la théorie des noyaux, nous en déduisons des taux de convergence pour le minimiseur du risque régularisé et montrons qu'il converge au moins au taux minimax de Sobolev. Cependant, des taux plus rapides peuvent être atteints, en fonction de l'erreur physique. Ce principe est illustré par un exemple unidimensionnel, qui prouve que la régularisation du risque empirique à l'aide d'informations physiques peut améliorier la performance statistique des estimateurs
Manufacturing of 2507 super duplex stainless steel by laser powder-directed energy deposition: process optimization and microstructure analyses
No full textInternational audienceThis paper presents a thorough manufacturability study of 2507 super duplex stainless steel (SDSS) produced by the Laser Powder-Directed Energy Deposition (LP-DED) process. First, experimental observations on single-track surface morphology and geometrical characteristics are described. Then, dimensional and porosity analyses of thin walls and bulk samples are presented and discussed in order to identify the optimal process parameter combinations, from a process point of view, for further characterizations to determine the general properties of LP-DED-printed samples. Finally, the results obtained on wrought steel of the same grade are compared to those of the printed material in terms of microstructure, texture, and ferrite/austenite ratio. A set of meticulously chosen process parameters with regard to the process and the machine (nozzle size, maximum power deliverable) gives the optimal results for the as-built LP-DED material in terms of porosity and microstructure quality but with smaller grains (irrespective of the phase studied) and a higher proportion of ferrite (about 75%)
Comparison of Hyper-Reduction Methods Combined With POD: Model Order Reduction of a Squirrel Cage Induction Machine in Nonlinear Case
No full textInternational audienceIn an industrial context of electrical device design or expertise, it is common for engineer to use models based on the Finite Element (FE) method. In the case of nonlinear magneto-quasistatic problems, it can lead to prohibitive computational times. Then, Model Order Reduction (MOR) approaches based on the Proper Orthogonal Decomposition (POD) combined with a hyper-reduction method can be employed to reduce the computational time but often leads to numerical instabilities. In this article, we aim to compare two hyper-reduction methods that were developed to improve the stability of the nonlinear reduced model: the Gaussian Newton Augmented Tensors (GNAT) method and the Energy-Conserving mesh Sampling and Weighting (ECSW) method. The reduced models are evaluated in terms of accuracy on global and local quantities of interest and in terms of speedup, with application to a squirrel cage induction machine. Improvement of the ECSW method has been proposed such that it appears that this method is more competitive than the GNAT method and enables to obtain very good results both in terms of accuracy and time computation, which was reduced by a factor 40 compared to the FE model
Measurement of air temperature fields in radiative environments using optical fibers
No full textInternational audienc
Maximum principle for the mass fraction in a system with two mass balance equations
No full textInternational audienceThree Finite Volume schemes are proposed in this note to satisfy the maximum principle for the mass fraction y, solution of an unsteady balance equation, including a relative velocity between phases and a source term. The continuous maximum principle is examined first. Then, linear implicit discrete schemes are detailed in a multi-dimensional and unstructured framework
Mean-field neural networks-based algorithms for McKean-Vlasov control problems *
No full textThis paper is devoted to the numerical resolution of McKean-Vlasov control problems via the class of mean-field neural networks introduced in our companion paper [25] in order to learn the solution on the Wasserstein space. We propose several algorithms either based on dynamic programming with control learning by policy or value iteration, or backward SDE from stochastic maximum principle with global or local loss functions. Extensive numerical results on different examples are presented to illustrate the accuracy of each of our eight algorithms. We discuss and compare the pros and cons of all the tested methods
Modèle réduit de contact appliqué à l'intégration temporelle explicite d'un composant de réacteur nucléaire soumis à de large rotations et à des impacts frictionnels
No full textThis work addresses the need for precise dynamic simulations to predict the non-linear vibrations of the thermal sleeve in a reactor tank. Accurate prediction of wear requires simulations that provide highly localized impact data. While the modeling of flexible bodies undergoing rigid body motions is well-established, incorporating frictional impacts remains challenging. Traditional methods using impulse-based techniques do not meet the precision requirements, leading to the adoption of explicit time integration schemes with small time steps and the penalty method for shock treatment.The conducted work involved a detailed examination of time integration for Lagrange’s equation, beginning with large rotations and rigid body dynamics using Newmark-derived integrators. These were tested through Hamiltonian case studies. The simultaneous integration of elastic vibrations and large rotations is then studied using two different formalisms, both based on modal decomposition. The first approach, Linear Theory of Elastodynamics (LTE), treats elastic and rigid body motions independently, with coupling occurring only during shocks. The second formalism includes time-dependent coupling between reference motion and elastic vibrations, modeled using the original CDIN algorithm. A qualitative comparison of the two approaches shows the LTE to be significantly more efficient, and the corresponding energy drift acceptable for the thermal sleeve’s geometry.The next phase focuses on modeling the conical contact between the thermal sleeve and the adapter using the penalty method.A parametric study based on static finite element computations produced an analytical formulation for the contact forces. The final part of the investigation involved simulating the in-service behavior of the thermal sleeve, recalibrating the model to match experimental data and computing wear energy on the contact surfaces.Ce travail répond au besoin de simulations dynamiques précises pour prédire les vibrations non linéaires de la manchette thermique dans une cuve de réacteur. La prédiction précise de l'usure nécessite des simulations qui fournissent des données d'impact très localisées. Alors que la modélisation de corps flexibles soumis à des mouvements de corps flexibles soumis à des mouvements de corps rigides est bien établie, l'incorporation des impacts frictionnelles reste un défi.Les méthodes traditionnelles utilisant des techniques basées sur les impulsions ne fournissent pas de données suffisamment localisées sur les efforts de contact dans le cas de large tâches de contact. Ce travail a donc recours à des méthodes d'intégration temporelle explicite avec de petits pas de temps combinés à la méthode de pénalité pour le traitement des chocs.Le travail effectué a consisté en un examen détaillé de l'intégration temporelle pour l'équation d'Euler-Newton généralisée, en commençant par les grandes rotations et la dynamique des corps rigides en utilisant des intégrateurs dérivés de la famille de Newmark. Ces derniers ont été testés à l'aide d'études de cas hamiltoniens. L'intégration simultanée des vibrations élastiques et de grandes rotations est ensuite étudiée à l'aide de deux formalismes différents, tous deux basés sur la décomposition sur base modale.La première approche, la théorie linéaire de l'élastodynamique (LTE), traite les vibrations élastiques et les mouvements de corps rigides de manière indépendante, le couplage ne se produisant que pendant les chocs. Le second formalisme inclut un couplage dépendant du temps entre le mouvement de référence et les vibrations élastiques, modélisé à l'aide du nouvel algorithme CDIN. Une comparaison qualitative des deux approches montre que le LTE est nettement plus rapide. De plus, la dérive énergétique occasionnée par le découplage est jugée acceptable pour des poutres peu élancées comme la manchette thermique.La phase suivante se concentre sur la modélisation du contact conique entre la manchette thermique et l'adaptateur à l'aide de la méthode de pénalité.Une étude paramétrique basée sur des calculs statiques par éléments finis a permis de formuler analytiquement les forces de contact.La dernière partie de l'étude a consisté à simuler le comportement en service de la manchette thermique, à recalibrer le modèle pour qu'il corresponde aux données expérimentales et à calculer l'énergie d'usure sur les surfaces en contact