The IRIDISCENTE project aims to develop a digital methodology to automatically and intelligently design, process and characterise new safe and sustainable metallic materials with improved properties, through the efficient flow of data along the material value chain, establishing the basis for the effective industrial use of Artificial Intelligence (AI) technologies. Data generated along the entire material value chain will be optimised through the use of Machine Learning (ML) techniques, making visible the applicability of AI technologies to: 1) the design and production phases of new iron-based alloys in powder form, starting from scrap generated throughout the current steel manufacturing process, 2) the subsequent processing phases of the new materials by using advanced manufacturing techniques, and 3) the characterisation phase of the manufactured materials by applying automatic processing techniques and data obtained through advanced characterisation techniques. All this will increase the industry’s capacity to design and produce new iron-based alloys, reducing dependence on third countries for the supply of strategic metals and favouring the development and optimisation of more flexible and sustainable industrial processes applied to the production of high added-value components.

The new digital methodology will focus on the steel value chain and will propose a systematic approach to optimise and increase the use of prime scrap with a higher content of alloying elements, facilitating the design of high performance alloys from gas atomised powders. The project will investigate 4 near-net-shape manufacturing technologies for the further processing of powders, including 2 direct additive manufacturing techniques: Direct Energy Deposition (DED), Powder Bed Melting (SLM) and two sintering-based techniques, 3D Extrusion Moulding (CEM) and Powder Injection Moulding (MIM). All these processes allow combining different material compositions in powder form to obtain components with improved properties in an efficient and sustainable way.

IRIDISCENTE will investigate new digital strategies to evaluate the processability of new materials by near-net-shape processes, investigating the development of reduced models elaborated by ML algorithms, based on data from multiphysics and multiscale simulation models of the processes, microstructure and properties of the materials, and experimental test data. This will allow defining and adjusting the optimal parameter window for each process, enabling the development of experiment design strategies and minimising the consumption of material resources for the evaluation and experimental validation of the processes.