| Conference | Green Aluminium for Transportation and Infrastructure, (INALCO 2023), Technical Session: Alloys-4, October 11-13, 2023, Québec City, Québec, Canada |
|---|
| Abstract | Modelling of process-microstructure-properties relationships in aluminium components with advanced microstructural characterization and machine learning Microstructural characterization, especially using optical microscopy, has been an important part of process development and process operation of aluminium components during the last decades. At the era of digitalisation and artificial intelligence, manufacturing processes are pushed to move toward a new paradigm where data intelligence and machine learning are fully integrated. This work presents a novel approach for optical microscopy and show how it can be integrated into a digital process data workflow and leveraged to gain critical insight on the process-microstructure-properties relationships with the help of machine learning. High pressure vacuum die casting (HPVDC) of Aural™-2 alloy and cold spray for additive manufacturing (CSAM) of AA6061 alloy, both in a research environment, are taken as example to illustrate the approach. It is demonstrated that with a proper database structure, advanced image analysis methods and a custom easy-to-use machine learning tool, it is possible to automate and speed-up a large part of the workflow and improve the overall value of the characterization on one hand by gaining insight on the process to microstructure relationships which can help to understand and improve the process and on the other hand, by developing a predictive model for the mechanical properties using the microstructure data as inputs for the model. |
|---|
