General information

ICON project
Started: 01/06/2022, Duration: 36 months
Status: Running
Program: STREAM
Project contact: Tom Craeghs (Materialise)


Industry: MatchID, Siemens Industry Software, Vibrant (Germany, USA)

Knowledge institutions: UGent - MMS and KU Leuven

Earlier research within SIM STREAM and M3 projects mainly focused on a better quality and performance of AM parts under a range of mechanical loading conditions. From these project results, it became clear that the structural performance of AM parts over their lifetime is also strongly influenced by small material defects/variations already induced during the printing process (e.g. geometrical deviation, lack of fusion, porosity, crack, residual stress).

Current nondestructive testing (NDT) methods to detect such defects mainly rely on ultrasound and/or X-ray tomography. However, for typical AM parts, these techniques become impractical due to their long experimental time, limited accuracy, poor defect detectability and/or difficulty in inspecting complex shaped parts. In the framework of the SIM M3 DETECT-ION project (ended October 2021), encouraging results were obtained by means of a novel and sophisticated vibrational inspection approach, which is coupled to learning algorithms.

The RESONAM project partners want to extend the developed vibrational methodology, and to couple it to both vibro-acoustic and camera-based inspection techniques. It is believed that this will allow to accurately detect defects and non-acceptable print variations in complex-shaped medium-series AM components, within a fast inspection time (order of few seconds). The methods will also provide insight on the current print variability of the material (e.g. small deviations due to build location, build number …), and how the print parameters and print environment can be tuned to further improve the nominal print quality of the AM material. This can also help to further improve the quality of the material that is printed and understand the influence of printing parameters on final material quality. To relax the required number of AM parts for accurate training of the learning algorithm, a virtual database will be generated which partially replaces the experiments.

This project is part of:

STRuctural Engineering materials (metals, plastics, composites) processed by means of additive manufacturing (AM) result in light weight parts with a high freedom of design, complex geometries and customer specific features. The AM-process performance is determined by the printing material properties, the printing parameters and the supporting monitoring and simulating software tools.

General information

Started: 01/12/2013
Theme: Durable & Sustainable Structural Materials
Program manager: Tom Craeghs (Materialise)