AID4GREENEST · AI powereD characterization and modelling for GREEn STeel technology

The fourth industrial revolution and market demands for advanced steels are driving the research towards transformation of the manufacturing processes and to ever-more sustainable steel compositions. The conventional ‘trial and error’ approach traditionally used to develop metallurgical processes still prevails in the industrial steel plants. However, it is a time-consuming, labour-intensive process entailing high material waste and associated carbon emissions. Also, it can ultimately lead down to a repetitive path that consists of creating a process design, putting it into production, and detecting possible process design flaws too late, resulting in high component rejection rates. Ascertaining the inadvertent flaws in the manufacturing approach before its implementation on industrial lines could be the key to major cost savings. With the introduction of AI- and simulation-driven design, back-and-forth interaction between part and process designs can be significantly diminished.The main objective of AID4GREENEST is to develop six new AI - based rapid characterization methods and modelling tools. AID4GREENEST tools’ scope will cover the steel design (chemistry and microstructure), process design (processing parameters), product design (processing and heat treatments) and product performance (creep) stages. Proposed tools will be complemented with a roadmap designed to enable model-based innovation processes, from materials design to product development, while considering the industry needs: enhanced material quality, reduction of carbon emission and waste generation, and reduced supply risk of critical raw materials. In order to facilitate the knowledge transfer of the characterization and modelling data generated in this project and across the wider European characterization and modelling community, the project will also develop an open online platform, based on a standardized and interoperable data management system and following the EMMC, EMCC and EMMO approach.