articles+ search results

Accuracy, Computer numerical control (CNC), Identification, Monitoring, LPBF, and [SPI.MECA.GEME]Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanical engineering [physics.class-ph]

International audience; The quality of the parts produced in additive manufacturing by laser powder bed fusion (LPBF) depends on the understanding of all the physics involved in the process. In LPBF, the effective path of the laser spot is rarely studied and existing works are mainly focused on the choice of the scanning strategy and its parameters. However, the programmed scanning path during the CAM stage is adapted by the numerical control (NC) unit of the manufacturing machine in order to generate admissible setpoints for the actuators (galvanometers). This modification locally generates deviations on the scanning path and significant decrease of the laser spot velocity. In order to identify the NC unit behaviour, a test bench that replicates an industrial machine has been developed. This test bench allows the acquisition of the setpoints sent to the actuators and their real positions at a frequency of 100 kHz, as well as the energy deposited by the laser spot in the work plane. An analysis of those data shows that the processing done by the NC is based on a filtering method which in some cases can generate a deviation of the scanning path of more than 100 µm for a programmed speed of 2 m/s. Finally, in order to correctly estimate the amount of energy brought by the laser on the powder, a dedicated model has been developed. This model takes into account the dynamics of the actuators, the behaviour of the NC unit and the optical chain of the system. Experimental tests have been conducted to validate the simulations produced by the proposed model.

[SPI.MECA.GEME]Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanical engineering [physics.class-ph]

International audience; The additive manufacturing machines used in laser powder bed fusion processes are composed of an opto-mechanical chain whose purpose is to focus the laser in the work plane to locally fuse the material with the required energy. The opto-mechanical chain does not only position the laser spot, it also induces a modification of its shape and thus the distribution of energy brought to the material. As a result, the shape of the laser spot is not the same at every point on the machine, which can have an influence on the characteristics of the produced parts. This aspect related to opto-mechanical chain technology is very often ignored, particularly in thermal modelling. In this work, we study the impact of opto-mechanical chains composed of post-objective focusing systems on the energy distribution of the laser spot in the entire achievable space by the system. For this purpose, mathematical models are used to evaluate the position of the laser spot in the work plane as well as the orientation of the laser beam. An energy model of Gaussian laser beams is then used and applied to these geometric models. The coupling of these two models makes it possible to study the energy distribution of the laser spot at any point in the workspace of an additive manufacturing machine. The work is confirmed with experimental measurements made on a test bench representative of an industrial additive manufacturing machine.

[SPI.MECA.GEME]Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanical engineering [physics.class-ph] and ComputingMilieux_MISCELLANEOUS

International audience

numerical control, filtering, geometric model, additive manufacturing, opto-mechanical chain, fabrication additive, chaîne opto-mécanique, modèle géométrique, commande numérique, filtrage, and [SPI.MECA.GEME]Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanical engineering [physics.class-ph]

In metal additive manufacturing by laser powder bed fusion, the geometry and mechanical characteristics of the produced parts are generated during the manufacturing process. These two aspects are greatly influenced by the laser spot trajectories, and by the control of the energy provided to the powder locally. The numerical control system, whose purpose is to generate instructions to be sent to actuators, has therefore a significant impact on the quality of the parts produced.This work proposes to study the local impact of the operations carried out in the numerical control on both the trajectories executed and the energy provided to the material. In the literature, few studies have addressed these aspects in additive manufacturing. For this reason, an experimental platform is implemented and used to analyze and better understand the operations currently implemented in industrial numerical controls.First, a mathematical model representative of the machine geometry is established. This model converts the laser spot trajectories into instructions for actuators. The model developed is used to improve the calibration step of the machines. Once the system is calibrated, the instructions sent to the actuators are studied. The various processes carried out in the industrial numerical control are analysed, limitations are highlighted and several proposals for improvements are implemented. All these developments are then used to precisely control the energy supplied to the material in the case of certain trajectories adapted to the process. The scientific developments proposed in these works are all validated experimentally on an additive manufacturing machine or on the test bench developed. The work carried out makes it possible to envisage many perspectives concerning the improvement of the treatments carried out inside the numerical control in additive manufacturing.; En fabrication additive métallique par fusion laser sur lit de poudre, la géométrie et les caractéristiques mécaniques des pièces produites sont générées au cours de la fabrication.Ces deux aspects sont grandement influencés par les trajectoires du spot laser et par la maîtrise de l’énergie apportée à la poudre localement. La commande numérique dont le rôle est de générer les consignes à envoyer aux actionneurs a donc un impact conséquent sur la qualité des pièces produites.Ces travaux proposent d’étudier l’impact des traitements effectués dans la commande numérique sur les trajectoires réalisées et sur l’énergie apportée à la matière. Dans la littérature, peu de travaux traitent de ces aspects en fabrication additive. C’est pourquoi une plateforme expérimentale est mise en œuvre et utilisée afin d’analyser et de mieux comprendre les opérations actuellement implémentées dans les commandes numériques industrielles.Un modèle mathématique représentatif de la géométrie de la machine est d’abord établi. Ce modèle permet de convertir les trajectoires du spot laser en consigne pour les actionneurs. Le modèle développé est utilisé afin d’améliorer l’étape de calibration des machines. Une fois le système calibré, les consignes envoyées aux actionneurs sont étudiées. Les différents traitements effectués dans la commande numérique industrielle sont analysés, des limitations sont mises en évidence et plusieurs propositions d’améliorations sont implémentées. Tous ces développements sont ensuite utilisés afin de maîtriser finement l’énergie apportée à la matière dans le cas de certaines trajectoires adaptées au procédé. Les développements scientifiques proposés dans ces travaux sont tous validés expérimentalement sur une machine de fabrication additive ou sur le banc d’essai développé. Les travaux effectués permettent d’envisager de nombreuses perspectives concernant l’amélioration des traitements réalisés dans la commande numérique en fabrication additive.

fabrication additive, modèles géométriques, défauts d'assemblage, erreur volumétrique, and [SPI.MECA.GEME]Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanical engineering [physics.class-ph]

International audience; Les machines de fabrication additive utilisées dans les procédés SLM et SLA sont composées d'un système de balayage laser permettant de focaliser et diriger un faisceau laser pour fusionner la matière brute (poudre ou liquide). Une des problématiques majeures consiste à maitriser la position du spot laser dans le plan de travail. Celle-ci dépend essentiellement de la géométrie de la machine et de la position angulaire des galvanomètres. Dans la littérature, la plupart des modèles utilisent des hypothèses restrictives sur la géométrie du système de balayage. Ces hypothèses induisent des déviations sur la position réelle du spot laser comparée à la position désirée, ce qui affecte les performances de la machine. Cette communication présente deux modèles géométriques du système de balayage laser : un modèle nominal et un modèle avec prise en compte des défauts géométriques d'assemblage. Ces modèles géométriques, souvent utilisés en usinage et en robotique, sont ici mis en oeuvre pour créer une machine de fabrication additive virtuelle. En utilisant ces modèles, la position du spot laser peut être simulée en connaissant certains paramètres géométriques de la machine et l'orientation des galvanomètres. Le second modèle géométrique est utilisé pour extraire l'influence des défauts d'assemblage sur la position du spot laser. Il permet également d'accélérer le processus de construction des tables de correspondance tout en maîtrisant les erreurs d'interpolation effectuées. Le travail réalisé permet de mettre en évidence et quantifier l'impact de chaque défaut d'assemblage sur la précision de positionnement du spot laser.

Accuracy, Assembly, Defect, Geometric modelling, and [SPI.MECA.GEME]Engineering Sciences [physics]/Mechanics [physics.med-ph]/Mechanical engineering [physics.class-ph]

International audience; Additive manufacturing (AM) machines used in SLM or SLA are composed of galvanometric scanning system in order to focus and steer a laser beam to melt raw materials. A major problematic in these AM processes is to master the laser spot position which essentially depends on the machine geometry and the galvanometers angular positions. In literature, most of existing models make strong assumptions concerning the geometry of the laser scanning system. Moreover, the position of the laser spot is often obtained by interpolating a table of correspondence, experimentally determined, between the angular positions of the galvanometers and the cartesian coordinates in the working plane. All these approximations induce deviations of the laser spot compared to the desired position and affect machine performance. This article presents two kinematic models of the galvanometric laser scanning system in an AM machine: a nominal model of the system and a model with assembly defects consideration. These kinematic models, often used for machine tools and robots, are here applied to create a virtual AM machine. Thereby, the laser spot position can be simulated knowing the geometry of the machine, the possible assembly defects, and the orientation of the galvanometers. The second kinematic model is then used to extract the influence of assembly defects on the laser spot position. The work described in this paper allows us to highlight and quantify the theoretical impact of an assembly defect on the precision of the laser spot position in an AM machine.