|
|
Applications of Machine Learning on Aero-Engine Titanium Alloys |
MI Guangbao1, SUN Yuanzhi1, 2, WU Mingyu1, 2, LI Peijie2 |
1. Aviation Key Laboratory of Science and Technology on Advanced Titanium Alloys, AECC Beijing Institute of Aeronautical Materials, Beijing 100095, China;
2. National Center of Novel Materials for International Research, Tsinghua University, Beijing 100084, China) |
|
|
Abstract The continuous development of high-performance aero-engines has put forward higher requirements for the comprehensive performance of titanium alloys. The composition design of titanium alloys based on the mechanism of different alloying elements is an important means to achieve titanium alloy modification. For the increasingly complex titanium alloy system for aero-engines, the interaction mechanism and precise design of different elements are extremely difficult. Traditional alloy design methods based on Mo equivalent or density functional theory cannot meet future needs, while machine learning has become a feasible and efficient theoretical method. The basic principles and methods of titanium alloy machine learning is introduced in this review, and the latest research achievements on the element design and processing optimization of titanium alloy for aero-engines through machine learning are summarized. This review focuses on the comparison of the characteristics and advantages between different machine learning models in predicting mechanical properties and high-temperature oxidation performance. Finally, a prospect is proposed for the future research methods for designing titanium alloy components in aero-engines based on active learning frameworks. It is supposed that the combination of Mo/Al equivalent design with machine learning, and the simplification design of complex multi-element alloy materials, are important development directions in the future.
|
|
|
|
|
|
|
|