Physical-information machine learning for strength and ductility prediction of metastable β titanium alloys
Source abstract
Abstract
Machine learning for predicting metastable β titanium alloy properties suffers from limited generalisation, unreliable extrapolation beyond data boundaries, and inadequate modelling of nonlinear effects of process parameters. This study proposes a physics-informed machine learning approach incorporating intrinsic physical attributes and phase transformation kinetics. Using 496 samples, it achieves R2 = 0.95 for ultimate tensile strength (UTS) and 0.90 for elongation (El), accurately predicting out-of-boundary alloys with similar heat treatment responses within 5.0% UTS and 2.5% El error. This approach reduces reliance on data completeness and complex algorithms, providing an accurate, generalisable path to UTS- and El-targeted titanium alloy design.
Source: openalex_inverted_index · Quality: baseline
Research context
OpenAlex topics: Machine Learning in Materials Science · Titanium Alloys Microstructure and Properties · Model Reduction and Neural Networks
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Shenyang University of Technology
Northwest Institute for Nonferrous Metal Research
University of Tennessee
Funding
National Natural Science Foundation of China · Major Science and Technology Projects in Liaoning Province
References (42)
Show source-provided references
1. 10.1186/s44147-023-00196-710.1186/s44147-023-00196-7 ↗
2. 10.1016/j.matdes.2024.11285010.1016/j.matdes.2024.112850 ↗
3. 10.1016/j.jacomc.2024.10001910.1016/j.jacomc.2024.100019 ↗
4. 10.1088/2053-1591/ada52010.1088/2053-1591/ada520 ↗
5. 10.3390/coatings1509107110.3390/coatings15091071 ↗
6. 10.3389/fmats.2024.145228810.3389/fmats.2024.1452288 ↗
7. 10.1080/21663831.2017.135021110.1080/21663831.2017.1350211 ↗
8. 10.1080/21663831.2023.227559910.1080/21663831.2023.2275599 ↗
9. 10.1016/j.scriptamat.2018.05.01010.1016/j.scriptamat.2018.05.010 ↗
10. 10.1016/j.actamat.2025.12150610.1016/j.actamat.2025.121506 ↗
11. 10.1002/advs.20251183410.1002/advs.202511834 ↗
12. 10.1016/j.scriptamat.2015.11.02010.1016/j.scriptamat.2015.11.020 ↗
13. 10.1016/j.jmrt.2024.01.04110.1016/j.jmrt.2024.01.041 ↗
14. 10.1016/j.jmrt.2024.12.24410.1016/j.jmrt.2024.12.244 ↗
15. 10.1109/ACCESS.2024.345051110.1109/ACCESS.2024.3450511 ↗
16. 10.1016/j.intermet.2025.10891410.1016/j.intermet.2025.108914 ↗
17. 10.1016/j.engfracmech.2024.11021410.1016/j.engfracmech.2024.110214 ↗
18. 10.1016/j.ijfatigue.2024.10820610.1016/j.ijfatigue.2024.108206 ↗
19. 10.1016/j.pnsc.2024.11.01010.1016/j.pnsc.2024.11.010 ↗
20. 10.1016/j.mtcomm.2024.11030910.1016/j.mtcomm.2024.110309 ↗
21. 10.1038/s42254-021-00314-510.1038/s42254-021-00314-5 ↗
22. 10.1016/j.scriptamat.2024.11635010.1016/j.scriptamat.2024.116350 ↗
23. 10.1016/j.engfracmech.2024.11067610.1016/j.engfracmech.2024.110676 ↗
24. 10.1039/C8ME00012C10.1039/C8ME00012C ↗
25. 10.1063/1.175038010.1063/1.1750380 ↗
26. 10.1016/0022-3697(92)90096-V10.1016/0022-3697(92)90096-V ↗
27. 10.1016/j.scriptamat.2018.10.04310.1016/j.scriptamat.2018.10.043 ↗
28. 10.1016/j.jallcom.2024.17790210.1016/j.jallcom.2024.177902 ↗
29. 10.1016/j.matdes.2017.02.04310.1016/j.matdes.2017.02.043 ↗
30. 10.1016/j.jallcom.2023.16950710.1016/j.jallcom.2023.169507 ↗
31. 10.1016/j.jallcom.2024.17580310.1016/j.jallcom.2024.175803 ↗
32. 10.1016/j.jallcom.2023.17002410.1016/j.jallcom.2023.170024 ↗
33. 10.1016/j.jmrt.2024.03.02910.1016/j.jmrt.2024.03.029 ↗
34. 10.2320/matertrans.45.157110.2320/matertrans.45.1571 ↗
35. 10.1016/j.msea.2022.14269610.1016/j.msea.2022.142696 ↗
36. 10.1557/JMR.1996.001210.1557/JMR.1996.0012 ↗
37. 10.1007/BF0267041510.1007/BF02670415 ↗
38. 10.1016/j.scriptamat.2006.04.02210.1016/j.scriptamat.2006.04.022 ↗
39. 10.1007/s11661-009-9798-010.1007/s11661-009-9798-0 ↗
40. 10.1080/21663831.2020.181509310.1080/21663831.2020.1815093 ↗
41. 10.1016/j.jmrt.2021.02.05510.1016/j.jmrt.2021.02.055 ↗
42. 10.3390/make304004810.3390/make3040048 ↗