Additive manufacturing opened completely new perspectives and possibilities for production processes. For industrial applications, powder-based laser additive manufacturing has proven especially relevant. However, it is not a matter of “simply printing” parts, because defects migtht occur during the process.
For critical components – for example in aerospace – it is impossible to use such parts with even the smallest defects. Hence, quality inspection is crucial. As of now, this is usually done with computed tomography (CT) – time-consuming and expensive. Also, parts can only be inspected with this method up to a certain size.
Machine Learning Enables Robust Process Monitoring
Thanks to supervised machine learning, new approaches become feasible: A robust quality assurance process would be possible in a reliable and cost-effective manner, because parts will be inspected already during prdouction instead of afterwards, as it’s done up to now.
To conduct research and development in this field, Testia GmbH is part of the EU-funded project: RobustAM – Robust and efficient processes for laser additive manufacturing (EFRE-LURAFO3001C)
In this project, an aerospace component is used as a specimen to demonstrate how the product life span can be increased by improving all process steps and monitor the whole process with all its interdependences via machine learning.
The collected in-situ data are analyzed by a neural network. The results then are compared with reference data from conventional CT inspection of the final components. This way it can be ensured that monitoring via machine learning is just as reliable as CT inspection.
Partners and Funding for RobustAM
RobustAM is a cooperation between Testia and WT-LW, AKON Robotics, AMSIS GmbH, BIAS GmbH, Materialise GmbH and the Leibniz-Institut für Werkstofforientierte Technologien (IWT) as well as Airbus Operations GmbH as an associated partner. The project receives funding from the European Regional Development Fund (ERDF) and is undertaken between April 2020 and June 2022.
Further information on the ERDF funding program in Bremen can be found on efre-bremen.de.
Goals of the RobustAM Research Project
The objective of the project is to reduce the variability of quality-relevant product parameters (e.g. porosity) in laser additive manufacturing of metallic components. This is based on software solutions for parts manufacturing and process simulation as well as in-situ detection and downstream minimalization of defects.
Using a Ti6Al4V component as an example, it shall be demonstrated that the service life of a part can be significantly improved compared to the state of the art through appropriate understanding of the interactions and further development of the individual process steps, as well as improved process monitoring.
Testia GmbH is involved in the CT, machine learning and process monitoring aspects of this project.