13–17 Sept 2021 Virtual Conference
Virtual
Europe/Vienna timezone
Thank you very much for your participation!

Numerical prediction of layer composition for multi-material Directed Energy Deposition

Not scheduled
3m
Virtual

Virtual

Poster C1. Additive manufacturing processes and modelling (incl. C2 & D10) C1_Poster Session

Speaker

Mr Vincent Jacquier (Université Paris-Saclay, CEA, Service d'Etudes Analytiques et de Réactivité des Surfaces)

Description

The Directed Energy Deposition (DED), a metal additive manufacturing process, opens up new possibilities for the manufacturing of multi-material structures by controlling the feedstock composition fused by the laser and mixed in the melt pool. Defect-free fabrications and control of the layers' composition however require the complex selection and fine-tuning of a large process parameter set. Indeed it is often necessary to modify the parameters depending on the melt pool’s target composition, while also taking into account the mixing with the previous layer.
Here we present a 3D finite element model of the melt pool aimed at fast prediction of the dimensions and composition field of the layer. The influence of track overlap on the melt pool geometry and composition is taken into account by updating the upstream boundary conditions with the composition and free-surface shape of the calculated molten track. The eddy viscosity originating from turbulence is approximated by a simple formula validated by a more complete v²-f RANS model. Microstructural and elemental analysis of a bi-metallic interface created by LPD is used to assess the validity of the model through SEM and EDS characterizations.
Simulation results illustrate that the fluid flow induced by surface tension gradients plays a critical role in the final shape and composition of the tracks and thus cannot be ignored. Comparison between simulation and experiments is satisfactory. A sensitivity study highlights the key parameters and material properties which affect the melt pool and must be set or known accurately enough to allow predictive simulations and parametric optimization, the latter being made possible by the short computation time.

Keywords: characterization, modelling, simulation, dissimilar materials, Additive Manufacturing, DED

Speaker Country France

Author

Mr Vincent Jacquier (Université Paris-Saclay, CEA, Service d'Etudes Analytiques et de Réactivité des Surfaces)

Co-authors

Dr Frédéric SCHUSTER (Cross-Cutting Program on Materials and Processes Skills, CEA, Université Paris-Saclay, 91191 Gif-sur-Yvette, France) Dr Hicham MASKROT (Université Paris-Saclay, CEA, Service d'Etudes Analytiques et de Réactivité des Surfaces, 91191, Gif-sur-Yvette, France) Prof. Julien ZOLLINGER (IJL, Université de Lorraine, CNRS, 54000 Nancy, France) Prof. Morgan DAL (PIMM, Arts et Metiers Institute of Technology, CNRS, CNAM, HESAM University, 75013 Paris, France) Dr Philippe ZELLER (Université Paris-Saclay, CEA, Service de la Corrosion et du Comportement des Matériaux dans leur Environnement, 91191, Gif-sur-Yvette, France) Dr Wilfried PACQUENTIN (Université Paris-Saclay, CEA, Service d'Etudes Analytiques et de Réactivité des Surfaces, 91191, Gif-sur-Yvette, France)

Presentation materials