Ziyuan Rao

Z. Li¹, R. Xie², H. Zhang², F. Körmann¹, J. Neugebauer¹, D. Raabe¹

¹Max-Planck-Institut für Eisenforschung, Düsseldorf, Germany;  ²TU Darmstadt

After the first discovery of the Fe₆₄Ni₃₆ Invar alloy in 1896 [1], new Invar alloys with improved and tailor-made properties have been developed for specific engineering applications [2]. However, Invar alloys are still not fully optimized in terms of several additional features such as strength, toughness, costs, mass density and corrosion resistance. In this context, the concept of HEAs/MEAs provides a great chance for the understanding of Invar effects and further designing of novel Invar alloys. In our work, we integrated machine learning with density-functional theory and experiments to investigate the Invar effect in HEAs/MEAs. An Invar alloy search map is constructed based on the present results and available literature data to visualize the relationships among saturation magnetization, Curie temperature and thermal expansion coefficient for a wide range of Invar alloys [3]. With an active learning loop, we identified 2 high-entropy Invar alloys with extremely low thermal expansion coefficients around 2×10^-6 K^-1 at 300 K. Our study thus opens a new pathway for the fast and automated discovery of high-entropy alloys with optimal thermal, magnetic and electrical properties [4].