Project Details
Projekt
Inverse design of sustainable wrought aluminum alloys with significant cast scrap content
Applicants
Professor Dr. Gerhard Dehm; Professor Dr.-Ing. Markus Kästner; Professor Dr.-Ing. Dierk Raabe
Subject Area
Mechanics
Mechanical Properties of Metallic Materials and their Microstructural Origins
Mechanical Properties of Metallic Materials and their Microstructural Origins
Term
since 2025
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 562094916
Aluminum (Al) alloys produced entirely through scrap remelting can reduce greenhouse gas emissions by over 90% compared to primary synthesis. However, a significant portion of the recycled Al comes from post-consumer scrap. This leads to the enrichment of impurities, particularly iron (Fe) and silicon (Si), in the recycled alloys. A major source of these impurities is end-of-life vehicles where cast Al alloys, with Si as primary alloying element, account for up to 65% of the Al content in internal combustion engine vehicles. With the growing demand for electric vehicles, the usage of cast Al in automobiles is expected to decrease. This shift will lead to a decline in the demand for secondary cast alloys, and to a surplus of components rich in Si and Fe available for recycling. At the same time, to make electric vehicles lighter, the demand for Al wrought alloys is expected to increase. For this and various other applications it is crucial to develop wrought Al alloys capable of accommodating these impurities. The objective of this project is to design novel sustainable wrought Al alloys with maximum scrap content while ensuring moderate formability. To achieve this goal, we develop a data-driven inverse design approach that combines knowledge-guided experimental investigations with high-fidelity modeling and simulation to set up, analyze and invert Composition-Process-Structure-Property (CPSP) linkages that will be represented in terms of surrogate models using microstructural descriptors. The experimental part starts with a screening to identify key parameters that influence the mechanical properties of interest. The identified parameters are then translated into microstructural goals, which inform compositional and processing requirements. Experimental microstructure data serves as input for the generation of numerical models that capture relevant features. This also requires knowledge of the detrimental second phase inclusions, which will be addressed by site specific small-scale experimental testing. In line with the overall objectives of DaMic, the microstructure forms an integrating factor of the project as descriptor-based microstructure characterization and reconstruction allow to integrate data from experiments and simulations. Efficient spectral solvers are utilized in combination with phase-field models of fracture to simulate the mechanical behavior and predict effective properties to augment the experimental CPSP linkages. In this context, an automated inverse design approach based on surrogate models, which represent the CPSP linkages in this project, is employed for an efficient exploration and to select optimal microstructural morphologies that satisfy the formability requirements and provide a high tolerance to impurities. We expect that Active Learning will make navigating the combinatorial space of alloy composition and processing parameters more efficient. This way, a direct link from property to composition is in reach.
DFG Programme
Priority Programmes
