| About this Abstract |
| Meeting |
2024 TMS Annual Meeting & Exhibition
|
| Symposium
|
Accelerated Discovery and Insertion of Next Generation Structural Materials
|
| Presentation Title |
Machine Learning-CALPHAD Assisted Design of L12-strengthened Ni-Al-Co-Cr-Fe-Ti Complex Concentrated Superalloy for Multi-property Optimization |
| Author(s) |
Sudeepta Mukherjee, Surendra Kumar Makineni, B.S. Murty, Satyam Suwas |
| On-Site Speaker (Planned) |
Sudeepta Mukherjee |
| Abstract Scope |
The demand for high-performance superalloys that offer superior temperature capability and thermal stability is paramount in various high-temperature applications, like the hot sections of aircraft gas turbine engines. This study highlights the potential of statistical inference from machine learning (ML) in designing Ni-based CCSAs by leveraging the AutoSciKit Learn library and CALPHAD data from Thermocalc. We thus offer a promising pathway for the rapid and efficient design of low-cost advanced superalloys with unique targeted properties. We trained ML algorithms, on a database of experimental and theoretical data. Using these models, a new Ni-based CCSA with targeted properties was screened. Long term isothermal annealing studies upto 1000h at 900 ℃ on lab-scale samples revealed that the alloy displays exceptional high-temperature thermal stability (coarsening rate of 8.01 nm^(3) s^(-1) ), consistent with TC-Prisma predictions, and comparable to that of commercial Ni-based superalloys such as CMSX-4, demonstrating the effectiveness of our ML-based approach. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
High-Entropy Alloys, High-Temperature Materials, ICME |