Lors de l'opération d’usinage à grande vitesse, la résistance à l'usure des outils de coupe est améliorée par l’utilisation des revêtements mono ou multicouches sur les faces actives de l’outil. Cependant, le chargement thermomécanique généré à l'interface outil-pièce affecte considérablement les zones de contact. Par cet effet, plusieurs modes d'usure tels que la fissuration, l’abrasion, l’adhésion et le délaminage du revêtement peuvent se manifester. L'étude du comportement des revêtements et de leurs différents modes de dégradation permet de mieux comprendre leur impact sur la durée de vie de l'outil et ainsi optimiser le procédé d'usinage. Dans ce travail de thèse, une approche numérique multi-étapes a été proposée pour prédire l'usure des outils de coupe revêtus. Cette approche est composée par trois principales étapes. La première consiste à effectuer une simulation éléments finis de l’usinage pour une courte durée (jusqu’à la stabilisation du chargement à l’interface outil/pièce). La deuxième étape consiste à récupérer ce chargement et de l’utiliser comme une entrée du modèle XFEM/Level-set. Ce dernier permet d’analyser le comportement des couches de revêtement sans recours à un maillage conforme aux interfaces. Par conséquence, la distorsion du maillage est évitée lorsque le profil d'outil usé est mis à jour, ainsi que le temps de calcul CPU est drastiquement réduit. La dernière étape de cette approche consiste à calculer le taux d’usure et ainsi prédire le déplacement des nœuds de l’outil de coupe affectés par l’usure. Les essais expérimentaux ont permis d’une part d’identifier les paramètres de contact outil/pièce, et d’autre part de valider l’approche proposée / In high speed machining, wear resistance of the cutting tools is improved by depositing single or multilayered coatings on their surface. However, the thermomechanical loading generated at the tool-workpiece interface greatly affects the contact zones. For this purpose, several wear modes such as cracking, abrasion, adhesion and delamination of the coating can be occurred. The study of the coatings behavior and their different degradation modes lead to better understanding of their impact on the tool life and machining process under optimal conditions. In this PhD thesis work, a multi-step numerical approach has been proposed to predict wear of the coated cutting tools. This approach involves three main steps. The first is to perform a finite element simulation of the orthogonal cutting for a short time (until the loading stabilization at the tool/workpiece interface). The second step is to recover this loading and use it as an input for the XFEM/Level-set model. The latter allow to take into account the coating layers presence without any need of mesh conforming to the interfaces. As a result, the mesh distortion is avoided when the worn tool profile is updated, as well as the CPU calculation time is drastically reduced. The final step of this approach is to convert the wear rate equation into a nodal displacement, thus representing the cutting tool wear. Based on the experimental tests, a procedure for identifying tool/workpiece contact parameters, and for calibrating the wear equation for each coating layer has been proposed. Experimental trials have been also used to validate the proposed approach

The need for miniaturized, and high performance microwave devices has focused significant attention onto new fabrication technologies that can simultaneously achieve high performance and low manufacturing complexity. Additive manufacturing (AM) has proven its capability in fabricating high performance, compact and light weight microwave circuits and antennas, as well as the ability to achieve designs that are complicated to fabricate using other manufacturing approaches. Direct print additive manufacturing (DPAM) is an emerging AM process that combines the fused deposition modeling (FDM) of thermoplastics with micro-dispensing of conductive and insulating pastes. DPAM has the potential to jointly combine high performance and low manufacturing complexity, along with the possibility of real-time tuning. This dissertation aims to leverage the powerful capabilities of DPAM to come-up with new designs and solutions that meet the requirements of rapidly evolving wireless systems and applications. Furthermore, the work in this dissertation provides new techniques and approaches to alleviate the drawbacks and limitations of DPAM fabrication technology. Firstly, the development of 3D packaged antenna, and antenna array are presented along with an analysis of the inherent roughness of 3D printed structures to provide a deeper understanding of the antenna RF performance. The single element presents a new volumetric approach to realizing a 3D half-wave dipole in a packaged format, where it provides the ability to keep a signal distribution network in close proximity to the ground plane, facilitating the implementation of ground connections (e.g. for an active device), mitigating potential surface wave losses, as well as achieving a modest (10.6%) length reduction. In addition, a new approach of implementing conformal antennas using DPAM is presented by printing thin and flexible substrate that can be adhered to 3D structures to facilitate the fabrication and reduce the surface roughness. The array design leverages direct digital manufacturing (DDM) technology to realize a shaped substrate structure that is used to control the array beamwidth. The non-planar substrate allows the element spacing to be changed without affecting the length of the feed network or the distance to the underlying ground plane. The second part describes the first high-Q capacitively-loaded cavity resonator and filter that is compatible with direct print additive manufacturing. The presented design is a compromise between quality factor, cost and manufacturing complexity and to the best of our knowledge is the highest Q-factor resonator demonstrated to date using DPAM compatible materials and processes. The final version of the single resonator achieves a measured unloaded quality factor of 200-325 over the frequency range from 2.0 to 6.5 GHz. The two pole filter is designed using a coupled-resonator approach to operate at 2.44 GHz with 1.9% fractional bandwidth. The presented design approach simplifies evanescent-mode filter fabrication, eliminating the need for micromachining and vias, and achieving a total weight of 1.97 g. The design is fabricated to provide a proof-of-principle for the high-Q resonator and filter that compromises between performance, cost, size, and complexity. A stacked version of the two-pole filter is presented to provide a novel design for multi-layer embedded applications. The fabrication is performed using an nScrypt Tabletop 3Dn printer. Acrylonitrile Butadiene Styrene (ABS) (relative permittivity of 2.7 and loss tangent of 0.008) is deposited using fused deposition modeling to form the antenna, array, resonator, and filter structures, and Dupont CB028 silver paste is used to form the conductive traces conductive regions (the paste is dried at 90 °C for 60 minutes, achieving a bulk DC conductivity of 1.5×106 S/m.). A 1064 nm pulsed picosecond Nd:YAG laser is used to laser machine the resonator and filter input and output feedlines.

In the introductory part of the work carried out the theoretical research of advanced methods of cutting. Chosen methods are laser, water jet and EDM wire cutting and trenching the wire. There is described the history, the principle of separation and classification. The practical part of these methods are discussed in detail with emphasis on their use in practice. The text is supplemented with illustrative examples of products. Finally, the work focuses on the comparison of the described methods. Overall, the work is conceived as a methodological text useful in teaching this topic.

O fresamento do aço inoxidável austenítico é um processo importante para a produção de peças em que se deseja alta resistência mecânica e à corrosão. No entanto, a usinagem desse material representa um desafio por suas características adversas ao corte. A alta taxa de encruamento e a alta dureza relativa fazem-no resistente ao corte, podendo gerar vibrações em diferentes faixas de frequência. Uma técnica importante ao processamento de sinais de vibração é a Transformada de Wavelet que permite analisar diferentes frequências do sinal através da subdivisão em aproximações e detalhamentos. Neste trabalho analisaram-se vibrações em alta e baixa frequência geradas no fresamento frontal do aço inoxidável austenítico AISI 316 a partir de sinais de força, coletados por meio de um dinamômetro piezelétrico e processados via Transformada de Wavelet Discreta. Também se fez a investigação dos perfis de rugosidade, dos parâmetros de rugosidade média (Ra) e média parcial (Rz) e das ondulações gerados na superfície fresada. Nos ensaios, foram utilizados insertos com três raios de ponta distintos, hastes da fresa com três diferentes comprimentos em balanço e foram variadas a rotação do eixo-árvore e a profundidade de corte axial em três níveis cada, totalizando 81 combinações de parâmetros. Constatou-se que a profundidade de corte representou a maior influência na vibração. Na usinagem com rotações abaixo do valor mínimo recomendado pelo fabricante (1600 rpm), houve dificuldades na formação e remoção do cavaco. A modificação do raio de ponta influenciou mais a vibração em pequenas profundidades de corte ou em zonas próximas às condições de instabilidade. O comprimento da haste mostrou comportamentos diferentes para a vibração, podendo estar relacionado com a mudança das frequências naturais do sistema e também pode definir entrada em regime instável. Verificou-se correlação do detalhamento (D1) da força resultante (vibração em altas frequências) com o parâmetro Ra para condições de vibrações intensas (maiores amplitudes), mas em regime estável. Assim, o parâmetro D1 pode ser utilizado na detecção de vibrações chatter no processo de fresamento frontal do aço inoxidável AISI 316. / The milling of austenitic stainless steel is important process for the production of part that require mechanical and corrosion resistance. However, the machining of this steel represents a challenge by its adverse features. The high hardening rate and the high relative hardness make it resistant to cutting and can generate vibrations in different frequency ranges. An important technique for the processing of vibration signals is the Wavelet Transform that allows the analysis of different signal frequencies through the subdivisions into approximations and details. In this work, high and low frequency vibration generated in end milling of AISI 316 stainless steel were analyzed from force signals collected through a piezoelectric dynamometer and processed via Wavelet Discrete Transform. Besides, the roughness profiles were investigated, as well as average (Ra) and partial mean (Rz) roughness parameters, and waviness generated on the milled surface. Three different insert nose radius, end mill tool lengths, depths of cut and spindle speeds were used in the experiments, totaling 81 combinations of parameters. It was found that depth of cut represented the greatest influence on vibration. In end milling with spindle speed below the minimum recommended by the tool manufacturer (1600 rpm) it occurred difficulties in the chip formation and removal. The modification of tool nose radius greater influenced the vibration at small depths of cut or in regions close to the stability limit. The end mill tool length showed different behaviors for the vibration, which may be related to the change of natural frequencies of the mechanical system and may also define an unstable state. The correlation of detail (D1) of the resulting force (vibration at high frequencies) with the parameter Ra for intense vibration conditions (larger amplitudes) was verified, but in stable state. Thus, D1 can be used for detecting chatter in end milling process of AISI 316 stainless steel.

The optimal choice of cutting conditions, as well cutting tool, has a significant effect on the quality of machined part and is an essential parameter to reduce the cost of production. This thesis describes CNC turning process of disc springs made of 51CrV4 material, especially using new cutting tools and the use of higher cutting speeds. A comparison of four manufacturers of cutting tools in terms of surface quality, tool path and the machined material volume will be introduced. It was experimentally verified that our proposed facility has achieved the best result.

This master thesis work concerns monitoring implementing methods on cutting operations for high cost part in the aeronautics and space fields. The final goal is to set tools to support industrial implementing but even more sustain acknowledgments and methods beyond the mission. This mission covers both a technical and a project management issue with related human factors. In this report, the following axis will be developed: • Monitoring on cutting operations stat of the Art. • The developed simulation function to secure monitoring. • Technical modifications at Numerical Command (NC) programming level and Programmable Logic Controller (PLC) level to obtain a simplified use of monitoring and reach the targeted production functioning. • The obtained results and expectations.

Les travaux présentés dans cette thèse sont réalisés dans le cadre du projet européen ACCENT et s’inscrivent dans la problématique de l’utilisation de la surveillance d’usinage dans la fabrication des pièces critiques de moteurs aéronautiques. L’objectif est de définir une technique de surveillance de l’usure de l’outil dans l’usinage de l’alésage central d’un disque haute pression réalisé en superalliage à base nickel. Trois grandes étapes peuvent être identifiées dans notre démarche : l’identification des modes de dégradation dans le perçage de l’Udimet® 720 LI, la définition de relations entre les différents types d’usure et les grandeurs physiques mesurées et finalement, la proposition de stratégies de surveillance basées sur les résultats précédents. Afin de répondre aux objectifs, un protocole expérimental a été défini. Les campagnes expérimentales nous ont permis de caractériser l’usure du foret et de définir une cartographie des modes de dégradation en fonction de conditions de coupe testées. Par ailleurs, l’impact des perturbations d’usinage (coupure de lubrification, modification du pourcentage d’émulsion du lubrifiant, défaut de position de l’outil par rapport à l’avant-trou, etc.) a été étudié. La deuxième partie consiste à identifier les relations entre les modes de dégradation et les signaux de surveillance enregistrés (efforts, couple, puissance, vibrations, etc.). Pour cela, nous avons privilégié deux pistes : une explication directe de l’évolution des signaux par des phénomènes physiques et une approche selon « la base de données », pour développer des modèles d’usures par des critères statistiques extraits sur les signaux. Cette analyse a aussi permis la détection de certaines perturbations d’usinage en utilisant les grandeurs physiques mesurées. Il s’agit notamment d’un défaut d’excentration de l’outil dans le porte-outil et d’un problème d’évacuation de copeaux. La troisième étape porte sur la sensibilité des signaux afin de détecter l’ensemble des phénomènes et la définition des stratégies de surveillance pouvant être mises en œuvre. Pour chaque relation définie, nous avons déterminé une probabilité de détection. Ensuite, nous avons utilisé les signaux en parallèle afin d’accroitre la robustesse de la détection. Les taux de probabilité ont été améliorés, en se basant sur l’hypothèse d’indépendance des signaux (sans calculer les biais existants). Enfin, l’association de l’approche phénoménologique à l’approche dite « statistique » a permis d’augmenter la robustesse de la surveillance. / The work presented in this thesis is a part of European project ACCENT and fits into the context of using process monitoring in the manufacture of critical parts for aircraft engines. The aim is to define a monitoring technique for the tool wear in the machining of the central bore of a high pressure disk made in nickel based superalloy. The major stages can be identified in our approach: identification of tool failure modes in drilling Udimet® 720 LI, defining the relationships between wear and measured physical quantities and finally, a proposal for monitoring strategies based on the previous results. In order to complete the objectives, an experimental protocol was defined. The experimental campaigns allowed us to characterize the drill wear and to define cartography of tool failure modes depending on cutting conditions tested. Moreover, the impact of machining disturbances (lubricant stop, change in percentage of lubricant emulsion, the position over the before hole, etc.) was studied. In the second part, the relations between the tool failure modes and monitoring signals recorded during the machining (forces, torque, power, vibrations, etc.) were identified. For this task, we focused on two directions: a straightforward explanation of signals evolution by physical phenomena and “black box” approach to develop wear models from a database obtained by statistical criteria extracts from the signals. This analysis also allowed the detection of certain machining disturbances using the physical quantities. These include an eccentricity of the tool in the tool holder and a chips evacuation problem. The third part concerns the signals sensitivity to detect the different phenomena and the definition of monitoring strategies which can be implemented. For each relation defined, a detection probability was determined. Further, we used signals in parallel to increase the robustness of detection. The probabilities were improved, with an assumption of independent signals (without calculating existing biases). Finally, the combination of the phenomenological approach to the so-called "statistical" has increased the robustness of the monitoring technique.

Dans le processus de fabrication en fraisage à grande vitesse ‘FGV’, l’étude de la réaction de la machine au cours de l’usinage est une tâche très délicate et importante. En effet, l’identification du comportement de la machine nécessite la modélisation de la loi de mouvement des axes et de la trajectoire réelle aux niveaux des discontinuités. Le nombre important de discontinuités engendrent une instabilité de la vitesse de déplacement des axes, ce qui implique une augmentation du temps d’usinage et un non-respect de la vitesse d’avance programmée, se traduisant par des problèmes de productivité et une sous-estimation du coût de l’usinage pour l’industriel. L’objectif de cette thèse est de développer un outil informatique qui permet de calculer la vitesse d’avance et de faire une estimation précise du temps de cycle pour n’importe quelle trajectoire générée par un logiciel de FAO. Pour ce faire, nous avons déterminé un modèle qui permet d’identifier le comportement cinématique des axes d’un centre d’usinage en FGV pour toute forme de trajectoire. À partir de la modélisation de la variation de la vitesse d’avance, nous avons déterminé le temps réel selon les trajectoires et l’erreur imposée par le bureau des méthodes. Enfin, nous utilisons ces résultats pour mettre en place une méthodologie pour l’aide au choix du diamètre de l’outil et de la stratégie d’usinage. Afin de valider les modèles et les méthodologies développés, une étude expérimentale a été réalisée sur des applications didactiques et industrielles. / In the context of high speed milling ‘‘HSM’’, the feed rate does not always reach the programmed value during the machining process which implies an increase of machining time and non-compliance with the programmed feed rate. This phenomenon leads to productivity issues and an underestimation of the cost of machining for the industry.The aim of this study is to develop a computerised tool in order to automate the determination process of the evolution of the feed rate for an imposed error and the estimation of cycle time and production cost. To begin with,a modeling approach in order to evaluate feed rate during any type of discontinuity between linear and circular contours in different combination by taking into account the specific machining tolerances.is presented. Then, the cycle time will be estimated with a maximum error of 7% between the actual and the prediction cycle time. The proposed method permits to develop a methodology to determine the optimal diameter of the tool and the optimal strategy. Finally, an industrial application was carried out in order to validate models and to determine the influence of feed rate evolution on the cycle time.

La mise en œuvre des matériaux composites, fait souvent appel à des procédés d’usinage conventionnel, comme l’opération de perçage utilisée lors de l’assemblage de structures par rivetage. Ces opérations peuvent générer dans la pièce usinée différents types d’endommagement: arrachement des fibres, rupture de la matrice, délaminage intralaminaire et interlaminaire, dégradation thermique de la matrice, ce qui peut provoquer une baisse des performances mécaniques de la structure. L’objectif de la thèse est de mettre en place un modèle numérique scientifiquement rigoureux pour analyser l’usinage des composites CFRP et prédire les mécanismes d’endommagement induits par l’outil coupant. Ce modèle basé sur une loi constitutive mésomécanique combine l’effet de la chute de rigidité dans le comportement du matériau, la plasticité, l’initiation et l’évolution de l’endommagement durant le processus d’usinage. Ensuite, deux modèles 2D et 3D adoptant un schéma explicite ont été implémentés via la sub-routine VUMAT dans Abaqus. Le délaminage interplis a été pris en considération à l’aide des éléments cohésifs disponibles dans le code ABAQUS/Explicit. Ce travail a permis de reproduire de manière réaliste les opérations de coupe orthogonale et de perçage des composites CFRP en termes de processus de formation du copeau, la prédiction des forces de coupe et celle de l’endommagement induit. Ces études ont montré que l’orientation des fibres et la profondeur de coupe sont les paramètres les plus influents en coupe orthogonale tandis que pour le perçage se sont les vitesses d’avance et la géométrie des outils / The machining of composite materials is often necessary for material removal operations by cutting tools such as drilling. These operations can generate a lot of damage in the machined workpiece (fiber fracture, matrix craking, intralaminar and interlaminar delamination and thermal degradation of the matrix), which can cause a decrease of mechanical performance of the structure. The PhD thesis objective is to set up a reliable accurate model to analyze the machining of CFRP composites and to predict the different damage modes induced by the cutting tool. This model is based on a mesomechanical constitutive law combining the stiffness degradation concept into the material behavior, the plasticity, the initiation and the evolution of the damage during the machining process. Two 2D and 3D models adopting an explicit scheme were implemented in Abaqus/Explicit analysis code through the user subroutine VUMAT. Furthermore, interlaminar delamination is taken into account using the cohesive elements available in the ABAQUS / Explicit code. This work allowed to realistic numerical simulation of orthogonal cutting and drilling operations of CFRP composites in terms of chip formation process, cutting forces prediction and induced damage. These studies have shown that the fiber orientation and the depth of cut were the most influential parameters in orthogonal cutting while for the drilling process, the feed rate and the tool geometry are the most important parameters

Le processus de rétro-conception (RC), dans la littérature, permet de retrouver un modèle CAO pauvrement paramétré, les modifications sont difficilement réalisables. C’est à partir de ce dernier et d’un logiciel FAO (Fabrication Assistée par Ordinateur) qu’une gamme de fabrication est générée. Cette thèse propose une méthodologie de RC de composants mécaniques dans un contexte de fabrication, nommée « Reverse Engineering For Manufacturing ». Une gamme de fabrication incluant les informations de type opérations d’usinage, posages, phases… est obtenue. Une fois cette gamme générée, elle sera stockée afin d’être réutilisée sur des cas similaires. L’intégration des contraintes métier dans le processus de RC fait penser aux concepts de Design For Manufacturing (DFM) et Knowledge Based Engineering (KBE). La réutilisation de stratégies d’usinage afin de supporter le contexte récurrent fait penser aux travaux sur le « Shape Matching ». En effet, des travaux sur les descripteurs topologiques permettent de reconnaitre, après numérisation, la nature d’une pièce et ainsi appliquer une stratégie d’usinage existante. Cette thèse propose donc un rapprochement entre deux domaines de recherches: la reconnaissance de formes (Shape Matching) et les méthodologies de gestion des données techniques (DFM et KBE). Cette thèse vise à proposer une nouvelle approche de RC dans un contexte d’usinage, et à développer un démonstrateur de RC qui permet de gérer les aspects récurrents de la RC en réutilisant des cas d’études connus / Reverse Engineering (RE) process, in the literature, allows to find a poorly parametrized CAD model, the changes are very difficult. It is from this CAD model and a CAM (Computer Aided Manufacturing) software that a CAPP (Computer Aided Process Planning) model is generated. This thesis proposes a RE methodology of mechanical components in a manufacturing context, called “Reverse Engineering For Manufacturing”. A CAPP model including information like machining operations, fixtures, steps… is obtained. Once this CAPP generated, it will be stored for reuse in similar cases.The integration of design intents in the RE process requires the use of Design For Manufacturing (DFM) and Knowledge Based Engineering (KBE) concepts. The reuse of machining strategies to support the recurrent context requires the use of Shape Matching works. Indeed, works on topological descriptors allow to recognize, after scanning, the nature of a part and thus apply an existing machining strategy. This thesis proposes to combine two research domains: Shape Matching and technical data management methodologies (DFM and KBE). This thesis aims to propose a new RE approach in a machine context, and to develop a RE viewer for managing recurrent aspects of RE by reusing known case studies