DIAMS : Design by artificial intelligence and high-throughput data of advanced alloys and innovative metallurgical concepts for structural applications

Responsible: Alexis DESCHAMPS,
INP Grenoble

Keywords: Metallurgy, alloy design, high-throughput screening, high-throughput characterization, optimization by artificial intelligence

The targeted DIAMS project aims to set up infrastructures allowing the accelerated design of metal alloys for structural applications, to develop a methodology allowing the implementation of these platforms for the entire French metallurgy community, and to demonstrate the possibilities through three demonstrator projects carried out in partnership between the project stakeholders.

The accelerated synthesis of metal alloys will be implemented through high-throughput alloy screening platforms. This screening can be carried out using powder/wire mixture additive manufacturing platforms made available and developed as part of the project at the CEA, IJL and SIMAP, as well as by the manufacture of gradient alloys. composition. A high-throughput characterization platform, allowing in-situ structural and microstructural study during thermo-mechanical treatments representative of industrial production processes, will be deployed on a set of high-energy lines of the European ESRF synchrotron. This tool will allow the creation of unique metallurgical databases on the alloys developed in the project. This platform will be complemented on the other synchrotron lines dedicated to the DIADEM program, and in particular for the DIAMS project with the resources of the SOLEIL MARS line. In parallel with these alloy manufacturing and data acquisition, artificial intelligence, coupled with physical modeling, will be implemented at all levels of the process, from proposing alloy compositions, to providing assistance for high-throughput characterization and modeling of microstructural evolutions and optimization of alloys.

The three demonstrator projects of the targeted DIAMS project will each involve collaboration between several project stakeholders, each implementing their own platforms or skills in the approach:

  • The first project will implement an advanced methodology for the design of high entropy alloys and concentrated complex alloys, for applications in harsh environments (irradiation/corrosion). This will follow a process of proposing alloys using calculations and AI, followed by screening and high-throughput characterization of the microstructure, then high-throughput characterization of the behavior in a harsh environment (in partnership with the targeted A-DREAM project).
  • The second project will implement the manufacturing of composition gradient alloys to evaluate the effects of alloys on the design of 3rd generation high strength steels (medium manganese), and their high throughput characterization coupled with AI modeling thanks to the thermo-mechanical platform on a high-energy synchrotron beamline.
  • The third project will implement the design of light aluminum-based alloys specific for additive manufacturing, using the project’s powder mixing platforms. The alloys screened by these synthesis methods can also be characterized at high throughput and the data thus obtained can be included in an AI approach to alloy design.