The process parameters to be measured during field tests (WP6) will be defined and the process monitoring equipment will be acquired and set up. Process data will be collected for the development of ML expert system (WP4).
AHSS grades for all tests will be supplied by the steelmakers. Different lab testing methodologies will be performed to determine mechanical properties and cut edge formability behaviour of AHSS. The mechanical characterization will include the measurement of fracture toughness of the AHSS sheets, aimed at gaining knowledge to understand the crack propagation in stretch-flangeability tests. The most relevant variables affecting stretch flangeability results will be identified to accurately reproduce in laboratory tests the edge-cracking resistance of AHSS. The results will be used as input data and for model calibration in WP5, and for the development of the ML expert system in WP4.
In this WP the fundamental part of the digital twin will be developed. To understand material behaviour an essential part is to develop phenomenological and micromechanical models by XFEM (ABAQUS/LS-DYNA) and SPG (LS-DYNA) of cutting process damage will be used together with a constitutive- and damage models developed within the project. These numerical approaches will allow predicting mechanical edge properties, burrs and cracks morphologies in the edges and surrounding areas.
Following data driven approaches and Industry 4.0 strategies, data fusion techniques and AI and Machine Learning algorithms will be investigated for exploiting the available data (material characterization and process sensors) to establish the correlations that will lead to a data driven edge-cracking prediction. The software prototype will be tested in WP6.
Tests will be performed to validate the edge-cracking prediction methods (experimental and modelling) developed in WP2 and WP3. It will permit to discern between different material properties, stretch-flangeability methodologies and models, the most adequate to predict and prevent edge-cracking. The results of these semi-industrial tests will be also used to select the AHSS grades for the industrial tests in WP6.
The following industrial tests are planned: (i) non-instrumented industrial tests with seat industrial parts to evaluate and validate the reliability of lab tests (WP2) and modelling (WP3) in real conditions; (ii) instrumented industrial tests to introduce, in addition to materials properties, process parameters in the prediction of edge-cracking using Industry 4.0.
According to project results, lab tests methodologies, damage models and machine learning solutions for edge-cracking prediction will be proposed.
Coordination, risk management and dissemination tasks.