Essais & Simulations n°140

DOSSIERSPÉCIAL JEC

DOSSIERSPÉCIAL JEC WORLDwise fashion, or sequentially activate angular swathes across thelayer. The elements are then allowed to naturally cool, and theresidual stresses are tracked in each element. Another layer ofelements is then activated in the model in the same way as thepreceding layer. The script simulates the complete process ofbuilding the part and tracks the residual stresses and deformationof each element.CALIBRATING THE SIMULATION MODELTo prove the quality of the simulation model, test structures wereproduced and the model calibrated to measured deformationand residual stresses. In the calibration process, the variationspace of the material parameter, the process parameter and thediscretization parameter is scanned by a design of experiment(DoE). From this, a metamodel of optimal prognosis (MOP) isgenerated by optiSLang. This metamodel shows how processvariability affects the results. The MOP is then used to calibratethe simulation model parameters to match the results of physicalmeasurements on the part. Important parameters used inthe calibration were the element size on the x, y and z axes, thelaser path (activating a complete layer, one rectangular elementat a time of various sizes, or an angular swatch across the layer),the time until melting of the next partial layer and the time untilplacement of the next powder layer.Measurement of the manufactured material revealed anisotropicdeformation and strength behavior, so engineers used Dynardo’smultiPlas, a custom anisotropic multisurface elastoplasticmaterial model in Ansys Mechanical, to match this anisotropicbehavior, and incorporated it into the additive manufacturingmodel. Comparing isotropic and anisotropic elastoplastic materialmodels, the team determined that the lower yield and ultimatestrength in the normal direction (between 80 percent and90 percent of the strength in the in-plane direction) has a veryimportant effect on the evolution of plastic strains. Employingthis anisotropic material model, the finite element model wascalibrated to predict the physical build to a high level of accuracy.Once the process parameter at the test structure was calibrated,the simulation workflow was ready to forecast deformation,stresses and cracks of the part to be qualified. ArianeGroup andDynardo engineers simulated the process of building a morecomplex part, an injector for a development prototype. Thefinite element model had 1,065,000 nodes and 620,000 quadraticvolume elements. It required 7 hours for thermal analysisand 32 hours for mechanical analysis on a personal computerwith 4 central processing units. The forecast using anisotropicmaterial models was an excellent match to the measurementsof the printed injector.OPTIMIZING THE PART GEOMETRY AND MANUFACTURINGPROCESSNext, engineers extended the workflow to investigate the effect ofpart geometry variation and key additive manufacturing processparameter variations on residual stress, plastic strain and distortionof the finished part. They created a fully automated workflowthat identifies the sensitivity of part quality to each designand process parameter incorporated into the DoE used to buildthe MOP. The workflow can optimize the part geometry and theadditive manufacturing process at the same time.The exceptionally high cost of a failure in the extremely competitiveaerospace industry makes it essential to perform a thoroughvalidation process before adopting new technologies. In thepast, this has meant a long trial-and-error process to validatenew manufacturing processes. Simulation can be combinedwith a much smaller volume of physical testing to provide fastqualification and insertion of new technologies without sacrificingmission safety. For example, this new workflow drasticallyreduces the time required to validate a new part, potentiallymaking it possible to optimize the part geometry and additivemanufacturing process with only two builds, one to validate thesimulation model and the second to validate the optimized partdesign and process. ArianeGroup engineers are planning to usethis process to reduce the time and cost required to validate partsfor the new Ariane 6 launch vehicle. ●Dieter Hummel, Thermomechanics Engineer ArianeGroupGmbH Ottobrunn, GermanyRoger Schlegel, Director of Consulting Dynardo GmbHWeimar, Germany60 IESSAIS & SIMULATIONS • N°140 • février - mars 2020

CyclesCodeFormationde Baseou SpécifiqueIntervenant et lieuDuréeen joursPrixAdhérentASTE HTDates proposéesMécanique vibratoireMesure et analyses des phénomènes vibratoires(Niveau 1)Mesure et analyses des phénomènes vibratoires(Niveau 2)31 mars-02 avrilMV13 1 570 €et 08-10 septembreBIUT du LimousinMV2 3 1 570 € 15-17 septembreApplication au domaine industriel MV3 B SOPEMEA (78) 3 1 570 €24-26 marset 13-15 octobreChocs mécaniques : mesures, spécifications, essaiset analyses de risquesMV4SChristian LALANNE, HenriGRZESKOWIAK et Yvon MORI (78)3 1 570 € 17-19 novembreTraitement des signauxPrincipes de base et caractérisation des signaux TS1 B IUT du Limousin 3 1 570 € 12-14 maiTraitement du signal avancé des signaux vibratoires TS2 SAnalyse modale et PilotagePierre-Augustin GRIVELET et BrunoCOLIN (78)3 1 570 € 15-17 septembrePilotage des générateurs de vibration : principes utiliséset applicationsAnalyse modale expérimentale et Initiation aux calculs de structureet essaisPV S SOPEMEA (78) 4 1 890 € 24-27 novembreAM S SOPEMEA ou AIRBUS D&S (31) 3 1 570 € 06-08 octobreClimatiquePrincipes de base et mesure des phénomènes thermiques CL1 B IUT du Limousin 3 1 570 € 17-19 novembreLes fondamentaux des essais climatiques CL2 B SOPEMEA (78) 2 1 350 €12-13 maiet 08-09 décembreElectromagnétismeSensibilisation à la compatibilité électromagnétique EL1 S IUT du Limousin 3 1 570 € 09-11 juinCompatibilité ÉlectroMagnétique (CEM) Exploitationdes normesEL2 S EMITECH (78) 2 1 170 € A définirPersonnalisation EnvironnementPrise en compte de l’environnement dans un programme industriel(norme NFX-50144-1)Prise en compte de l’environnement mécanique (norme NFX-50144-3)Principes de personnalisation de basePrise en compte de la norme NFX-50144 dans la conceptiondes systèmesPrise en compte de l’environnement mécanique (norme NFX-50144-3)Principes de personnalisation avancéesP1 S Henri GRZESKOWIAK (78) 2 1 170 € 15-16 septembreP2-1 S Bruno COLIN et Pascal LELAN (78) 3 1 570 € 06-08 octobreP3 S Bruno COLIN (78) 3 1 570 € 17-19 novembreP2-2 S Bruno COLIN et Pascal LELAN (78) 3 1570 24-26 novembrePrise en compte de l’environnement climatique(norme NFX-50144-4)P4SHenri GRZESKOWIAK et HenriTOLOSA (78)3 1 570 € 22-24 septembreMesureExtensomètrie : collage de jauge, analyse des résultatset de leur qualitéM1 S Raymond BUISSON (78) 3 1 570 € 01-03 décembreConcevoir, réaliser, exploiter une campagne de mesures M2 B Pascal LELAN (78) 2 1 170 € 08-09 décembreMesure tridimensionnelle M3 B IUT de LIMOGES 1 900 € 05 novembreFiabilité et EssaisConception et validation de la fiabilité - dimensionnementdes essais pour la validation de la conception des produitsE1 S Alaa CHATEAUNEF (78) 3 1 570 € A définirLes essais accélérés et aggravés E2 S Alaa CHATEAUNEF (78) 2 1 170 € A définirFatigue des matériaux métalliques :Essais, dimensionnement et calcul de durée de viesous chargement complexeE3 S Alexis BANVILLET 2 1 170 € 24-26 novembreGestion d’une Salle blanche : application dans un Centre d’Essais ME1 S AIRBUS D&S (31) 2 1 170 € A définirL’assurance qualité dans les laboratoires d’essais selon le référentielEN ISO/CEI 17025ME2 S EMITECH (78) 2 1 170 € A définirThermométrieThermométrie pour les essais vide thermique T1 S Alain BETTACCHIOLI (78) 1 900 € A définirSimulationLa simulation numérique et les essais : complémentarités -comparaisonsS1 B Philippe PASQUET (78) 2 1 170 € A définirFormations 2020