Accelerated Design of High γ′ Solvus Temperature and Yield Strength Cobalt-Based Superalloy Based on Machine Learning and Phase Diagram

Abstract

This study combines machine learning and a phase diagram to accelerate the design of a cobalt-based superalloy with a composition of Co-30Ni-10Al-6Ta (at%). The results show that Co-30Ni-10Al-6Ta alloy exhibits high γ′ solvus temperature (1,215 °C) and high yield strength (1,220 Mpa at 25 °C), which is comparable with commercial nickel-based polycrystalline superalloy M-Mar-247. Moreover, the wide processing window and excellent γ′ phase stability make it lucrative for further applications at high temperatures. Meanwhile, the alloy design method also provides a new idea for efficiently realizing the preparation of high-performance alloys.