Url https://www.cimne.com/sgp/rtd/Project.aspx?id=765
LogoEntFinanc
Acronym ADaMANT
Project title Marco Computacional para la Fabricación Aditiva de Componentes de Aleaciones de Titanio
Reference DPI2017-85998-P
Principal investigator Luis Miguel CERVERA RUIZ - mcervera@cimne.upc.edu
Michele CHIUMENTI - michele@cimne.upc.edu
Start date 01/01/2018 End date 31/12/2020
Coordinator CIMNE
Consortium members
Program Fomento inv.cient.-téc.de Excelencia: Generación conocimiento Call Proyectos de I+D: Excelencia 2017
Subprogram Proyectos de I+D (Excelencia) Category Nacional
Funding body(ies) MCIU Grant 45.980,00 €
Abstract Additive Manufacturing (AM), the next industrial revolution, is a rapidly emerging manufacturing technology and has evolved as one of the most promising techniques for creating components of virtually any shape, based on digital models. Technological interest of AM has been mainly shown by industrial/bussiness machines, consumer products/electronics, automotive, aeronautical, and mechanical/dental sectors. Key benefits of AM are innovation, process and cost optimization, and enhanced mechanical properties. Critical issues related to AM are distortions, feedstock quality, residual stresses, porosity, cracking, delamination and swelling, substrate adherence and warping, and scan and deposition strategy. The goal of project ADaMANT is to develop highly-efficient numerical tools with reliable predictive capabilities for the enhanced-accuracy and high-fidelity simulation of multi-physics and multi-scale AM processes for Ti alloys. In order to reach this ultimate goal, four main research lines will be addressed: 1.Enhanced accuracy mixed FE formulation. Mixed FE formulations for nonlinear solid mechanics problems guarantee an enhancement over standard FE formulations in terms of pressure, stress and strain accuracy. In mixed formulations, the pressure and/or the stress and strain are approximated independently from the displacement field. In this way, more accurate pressure, stress and strain fields are computed, resulting in a more precise computation of the solids nonlinear behavior, particularly for low order FE. 2.High-accuracy multi-level hp-adative Finite Cell Method (FCM). The high-fidelity simulation of AM processes is an extremely highdemanding computational task which has to tackle complex multi-physics and multi-scale challenges in a time-evolving geometry. The use of body-fitted meshes is not a suitable option, while embedded domain methods (EDM) is a very attractive choice. To tackle those highdemanding challenges, enhanced-accuracy and high-fidelity models are mandatory. The Multi-Level hp-Adaptive FCM is a very attractive choice which combines the high-accuracy of high order FE with the flexibility of EDM, providing high-order convergence rates. The method allows dynamically refining or coarsening the embedded mesh size, keeping track of the moving heat source and the solidified domain, locally and globally, all along the process simulation. 3.Multi-scale Reduced Order Model (ROM). The high-fidelity simulation of AM processes would require huge meshes compatible with the layer thickness, and a huge number of time steps to follow the metal deposition based on the scanning sequence, being unaffordable for the actual computation capabilities. In this work, a multi-scale approach, based on ROM techniques, is proposed. The idea is to use an offline phase where the high-fidelity process is studied for a given Representative Volume Element (RVE) domain. 4.Microstructural model for Ti6Al4V Titanium alloy. The approach to microstructure modeling in AM has been generally based on the adaptation of models previously developed for casting and welding. The iterative metal depositions generate fast thermal cycles characterized by an unusual range of cooling and heating, leading to unique microstructural and mechanical properties different from the properties of cast or weld parts. Today, only a limited number of commercial alloys have been used and characterized in AM processes, being the Ti6Al4V alloys the most investigated.