Computational alloy design
F. Tancret, I. Braems, E. Bertrand, L. Couturier
Our team, a pioneer in France in the domain, develops computer-aided alloy design methods, associating physical models (including computational thermodynamics using Calphad method and ab initio simulation) with artificial intelligence tools like data mining (machine learning) and multi-objective optimisation by genetic algorithms. These works, undertaken within academic or industrial collaborations, deal with many categories of materials (alloys based on Fe, Ni, Ti, Al, Mg; high entropy alloys HEA; complex concentrated alloys / compositionally complex alloys CCA), for a variety of applications (oil & gas / petrochemistry; energy / nuclear domain; aeronautics, automobile; micromechanics…).
Keywords: Alloy design, neural networks, first principles, Thermo-Calc, steels, stainles