Publications

Due to its capacity to evolve in a large solution space, the Simulated Annealing (SA) algorithm has shown very promising results for the Reverse Engineering of editable CAD geometries including parametric 2D sketches, 3D CAD parts and assem blies. However, parameter setting is a key factor for its performance, but it is also awkward work. This paper addresses the way a SA-based Reverse Engineering technique can be enhanced by identifying its optimal default setting parameters for the ftting of CAD geometries to point clouds of digitized parts. The method integrates a sensitivity analysis to characterize the impact of the variations in the parameters of a CAD model on the evolution of the deviation between the CAD model itself and the point cloud to be ftted. The principles underpinning the adopted ftting algorithm are briefy recalled. A framework that uses design of experiments (DOEs) is introduced to identify and save in a database the best setting parameter values for given CAD models. This database is then exploited when considering the ftting of a new CAD model. Using similar ity assessment, it is then possible to reuse the best setting parameter values of the most similar CAD model found in the database. The applied sensitivity analysis is described together with the comparison of the resulting sensitivity evolution curves with the changes in the CAD model parameters imposed by the SA algorithm. Possible improvements suggested by the analysis are implemented to enhance the efciency of SA-based ftting. The overall approach is illustrated on the ftting of single mechanical parts but it can be directly extended to the ftting of parts’ assemblies. It is particularly interesting in the context of the Industry 4.0 to update and maintain the coherence of the digital twins with respect to the evolution of the associated physical products and systems.

Depuis 2020, la crise sanitaire a profondément bouleversé l’enseignement supérieur, initiant une transition accélérée vers des pratiques pédagogiques numériques. Cette mutation, d’abord contrainte, s’est inscrite dans une réflexion durable sur la flexibilité, l’inclusivité et l’innovation pédagogique. Le projet du Learning Center AtriuM s’inscrit au sein de l’Ecole nationale d’Arts et Métiers au travers du programme Evolutive Learning Factories (ELF) et de la Direction de l’Information Scientifique et de la Science Ouverte (DISSO). Son développement notamment sur le campus Arts et Métiers ENSAM de Metz, illustre cette transformation en proposant un espace d’apprentissage connecté et évolutif. En nous appuyant sur un état de l’art international des pratiques dans les Learning center universitaires, sur des enquêtes approfondies pour saisir les demandes des futurs usagers, l’AtriuM Metz repose sur : - Un espace modulable intégrant technologies numériques et émergentes, ateliers de fabrication et environnements adaptés à divers besoins pédagogiques. - Un guichet unique pour (enseignants, chercheurs, personnels et extérieurs) centralisant des services comme le prêt de ressources et objets connectés, des formations ciblées et la mise à disposition des espaces équipés favorisant le travail collaboratif ou individuel. Ce lieu incarne une réponse concrète aux défis posés par les technologies émergentes, notamment la réalité virtuelle (VR) et l’intelligence artificielle générative (IA). Ces outils permettent d’immerger les étudiants dans des scénarios d’apprentissage innovants tout en posant des questions éthiques et pédagogiques, comme la préservation de l’interaction humaine et de la réflexion critique. Au-delà de l’accompagnement et de la formation, l’AtriuM joue un rôle clé dans la diffusion scientifique, organisant des conférences et ateliers pour rapprocher la recherche de la société civile. En sensibilisant aux enjeux de durabilité et aux compétences technologiques, il contribue à une pédagogie résiliente et inclusive. Cet article sur le projet AtriuM témoigne de la capacité de l’enseignement supérieur à évoluer face aux défis contemporains, marquant une nouvelle ère où technologies, espaces et pédagogie s’allient pour anticiper les besoins du XXIᵉ siècle.

Reliable detection of barely visible impact damage is critical to ensure the structural integrity of composite components in service, particularly in safety-critical applications such as pressure vessels and transportation systems. This study presents a solution for detecting such damage in woven glass fiber-reinforced thermoplastic composites using terahertz (THz) time-of-flight tomography and convolutional neural networks. THz provides non-contact, non-ionizing, high-axial-resolution imaging of subsurface and back-surface damage, addressing key limitations of surface-based inspection methods. While THz imaging alone may not always permit conclusive damage identification, we bridge this gap by training neural network classifiers on depth-resolved THz B-scan images using ground truth from co-located X-ray micro-computed tomography. Among several pretrained architectures tested via transfer learning, DenseNet-121 exhibits the highest accuracy. The model remains robust even when trained on truncated B-scans excluding surface indentation features, confirming its ability to detect structural anomalies located internally or on the back surface. This is particularly relevant for applications where back-side access is not feasible. Experimental validation is performed on impacted glass-fiber-reinforced thermoplastic coupons prepared in accordance with ASTM D7136, with damage severity quantified through force–displacement data and micro-tomographic analysis. Labeling for supervised learning conforms to acceptance criteria from industrial standards for composite pressure vessels (ASME BPVC Section X, CGA C-6.2), ensuring regulatory alignment and enabling deployment in quality control workflows. The proposed method minimizes the need for expert interpretation or secondary validation and offers direct applicability to in-service inspection and manufacturing quality control.

The work-hardening curve of sheet metals under large plastic strains can be extracted from the Plane Strain Compression Test (PSCT) using an analytical method that relies on several simplifying assumptions and correction factors (friction, boundary conditions, lateral spreading, tool geometry, yield criterion, anisotropy). This study rigorously assesses each of these correction factors using finite element simulations. Synthetic materials with predefined hardening laws are used to enable direct comparison between the reference curves and those extracted from simulated PSCTs. Dedicated simulation setups were developed to isolate the effect of each factor through progressive 2D and 3D configurations. The results show that the analytical method is generally valid when appropriate corrections are applied, with improved accuracy observed when using rounded tools with small radii under low-friction conditions. Recommendations for the selection of correction factors are provided to enhance the reliability of flow curves obtained through this method.

Elasticipy is a Python library designed to streamline computation and manipulation of elasticity tensors for materials and crystalline materials, taking their specific symmetries into account. It provides tools to manipulate, visualise, and analyse tensors –such as stress, strain and stiffness tensors– simplifying workflows for materials scientists and engineers.

A micromechanics-informed neural network framework is developed for homogenization of periodic unidirectional thermoconductive composites with cylindrically orthotropic fibers. The framework hard-imposes the steady-state governing heat conduction equations within the network architecture, enabling accurate capture of singular heat flux fields at the fiber center that are challenging for conventional approaches. In contrast, continuity and periodicity conditions are enforced via boundary collocation points in the loss function. Validation against finite element simulations across a wide range of fiber volume fractions shows that accurate and converged temperature distributions can be achieved after 9000 training epochs using 8-16 harmonic terms. Additional higher-order harmonics are difficult to train reliably and may degrade predictions. While strong agreement is observed in the matrix heat flux distributions, noticeable discrepancies persist in the fiber phase due to varying ability to capture the singular heat flux fields. Furthermore, uniform collocation points converge faster than random points during solution refinement. Finally, transfer learning is employed to accelerate training for new configurations, allowing the network to achieve comparable accuracy after only 2000 training epochs, which is substantially fewer than the 9,000 epochs required when training from scratch.

In this short review, the multiscale modeling of dissipative composites undergoing fully coupled thermomechanical processes is outlined through the models presented in a collection of recent works. The aim is to demonstrate the challenges and limitations of: (1) the multiscale approaches (full-field or mean-field techniques), (2) the computational approaches dealing with complex material systems, (3) the alternative methodologies dedicated to the analysis of composite structures, such as those founded on the data-driven modeling and the model order reduction techniques.

In this study, a ZrO2 thin film was deposited onto a Ti6Al4V substrate using the Oblique Angle Deposition (OAD) technique. The influence of the substrate/Zr target an-gle (15°, 30°, 45°, and 60°) was investigated, with a fixed azimuthal orientation (Phi) of 180°. The primary objective of this work is to develop and characterize novel biocompatible coat-ings for hip prosthesis implants with a complex 3D spherical geometry. The OAD method enables thin film deposition on such geometries and enhances understanding of how the par-ticle incidence angle affects the surface morphology and microstructure of zirconium oxide (ZrO2) thin films. This study combines an experimental approach DC magnetron sputtering with a multi-scale numerical approach using Monte Carlo codes (SRIM, SIMTRA, and NASCAM). The structure, texture, and growth of the ZrO2 coatings were analyzed via X-ray diffraction (XRD), while microstructure and surface morphology were examined using scan-ning electron microscopy (SEM). Hardness and Young’s modulus were determined through nanoindentation testing. Results indicate that increasing the oblique angle leads to a decrease in hardness. Experimental and numerical findings complement each other, offering deeper insight into the deposition phenomena. SIMTRA simulations closely replicate experimental observations: a higher number of incident particles results in increased coating thickness. Additionally, the film thickness decreases with increasing substrate inclination angle. The microstructure of ZrO₂ thin films is strongly influenced by substrate orientation, and coated substrates demonstrate superior performance compared to their uncoated counterparts.

Introduction Adolescent Idiopathic Scoliosis (AIS) is classically evaluated through static X-rays and health-related quality of life questionnaires that do not reflect the functional limitations of patients during daily life activities, such as walking. The aim was to investigate kinematic strategies in non-operated AIS with different types of curvature during walking using 3D gait analysis. Methods 13 AIS with Lenke 5 (major Cobb: 23 ± 8°), 30 AIS with Lenke 1 (major Cobb: 40 ± 14°) in addition to 24 controls underwent biplanar X-rays followed by 3D gait analysis. The kinematic parameters of the head, trunk, spinal segments, pelvis and lower limbs were compared between groups. Results AIS Lenke 5 had a lumbar segment bending while walking (T12L3-L3L5: 5 ± 7° vs. -3 ± 7° in controls) to the concave side of the scoliosis. They walked with an increased pelvic frontal mobility (12 ± 3° vs. 9 ± 3°) and internal rotation of the right foot (-2 ± 6° vs. -11 ± 8°; all p < 0.05). AIS Lenke 1 increased their thoracic & lumbar segment bending to the concave and to the opposite side respectively (T6T9-T9T12: -4 ± 9° vs. 1 ± 4°; T12L3-L3L5: 8 ± 12° vs. -2 ± 7°). However, they tended to reduce their lumbo-pelvic mobility (7 ± 5° vs. 12 ± 5°; all p < 0.05). Conclusion In response to their inherent lumbar stiffness and bending, AIS Lenke 5 patients tended to increase their pelvic frontal mobility and to develop a homolateral internal foot rotation, ensuring a dynamic alignment during gait. AIS Lenke 1, by producing opposite bending movement at the thoracic and lumbar segments, tended to reduce their lumbo-pelvic mobility and ensure coronal dynamic alignment.

This paper investigates the influence of substrate grain size on the behavior of a multilayer CrN/CrAlN coating, with the bilayer thickness varying across the cross-section in the range of 200–1000 nm. The substrate tools were made of WC-Co sintered carbide with three different grain sizes. The coatings were subjected to mechanical and tribological tests to assess their performance, including nanohardness, scratch resistance, and tribological testing. The coating’s roughness was measured using a 2D profilometer. Additionally, the chemical composition and surface morphology were analyzed using Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray Spectroscopy (EDX). The durability tests were performed on an industrial CNC machine tool on the particleboard. The results revealed that tools with ultra-fine nano-grain (S) and micro-grain (T) WC-Co substrates exhibited a significant increase in tool durability by 28% and 44%, respectively. Significant differences in the microgeometry of the substrate U, especially in relation to the tool based on substrate S, explain the lack of improvement in its durability despite the use of a multilayer coating.