Inverse Prediction of the CALPHAD-Modeled Physical Properties of Superalloys Using Explainable Artificial Intelligence and Artificial Neural Networks

Year 2025, Volume: 14 Issue: 1, 331 - 347, 26.03.2025

Yusuf Uzunoğlu , Yusuf Alaca

https://doi.org/10.17798/bitlisfen.1586564

Abstract

The CALPHAD methodology models the physical, mechanical, and thermodynamic properties of materials based on specified alloy compositions using phase equilibrium calculations and thermodynamic databases. With the CALPHAD approach, millions of material-property data can be obtained for each alloy composition over various temperature ranges. However, finding an alloy with the desired properties often requires lengthy trial-and-error processes that involve manually adjusting the composition. In this study, the goal is to inverse this approach using artificial intelligence to predict alloy compositions that yield the desired properties. Accordingly, in the JMatPro software based on the CALPHAD methodology, the physical properties (density, thermal conductivity, linear expansion, Young's modulus, bulk modulus, shear modulus, and Poisson's ratio) of 250 different Ni-Cr-Fe-based superalloys in the temperature range of 540–920 °C were modeled. A dataset with 5000 rows was created from the generated data, of which 80% was used to train the artificial intelligence model, while 20% was reserved for validation and testing. Through analyses using Explainable Artificial Intelligence (XAI) and Artificial Neural Networks (ANN), alloy compositions with the desired physical properties at a given temperature were predicted with a high accuracy rate of 98.03%. In conclusion, beyond obtaining material properties from alloy compositions through the CALPHAD approach, artificial intelligence techniques make it possible to accurately inverse predict alloy compositions that yield specified physical properties at a particular temperature.

Keywords

CALPHAD methodology, Explainable Artificial Intelligence, Artificial Neural Networks, Alloy Design, Superalloy.

Ethical Statement

The study complies with research and publication ethics.

Supporting Institution

This study did not receive any funding.

References

• N. Saunders and A. P. Miodownik, CALPHAD (Calculation of Phase Diagrams): A Comprehensive Guide. Pergamon Press, 1998.
• H. L. Lukas, S. G. Fries, and B. Sundman, Computational Thermodynamics: The Calphad Method. Cambridge University Press, 2007.
• Y. Du, S. L. Chen, and B. Huang, “An Overview of CALPHAD Applications in Superalloy Design,” J. Alloys Compd., vol. 456, no. 1–2, pp. 18–29, 2008.
• P. Raccuglia and others, “Machine-learning-assisted materials discovery using failed experiments,” Nature, vol. 533, no. 7601, pp. 73–77, 2016.
• T. Xie and J. C. Grossman, “Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties,” Phys. Rev. Lett., vol. 120, no. 14, p. 145301, 2018.
• K. T. Butler and others, “Machine learning for molecular and materials science,” Nature, vol. 559, no. 7715, pp. 547–555, 2018.
• L. Ward et al., “A general-purpose machine learning framework for predicting properties of inorganic materials,” npj Comput. Mater., vol. 2, no. 1, pp. 1–7, 2016.
• D. Jha and others, “ElemNet: Deep Learning the Chemistry of Materials from Only Elemental Composition,” Sci. Rep., vol. 8, no. 1, pp. 1–13, 2018.
• Y. Zhang and others, “Deep learning-based discovery of rare materials,” Nature, vol. 569, no. 7754, pp. 368–373, 2019.
• B. L. DeCost and others, “Scientific AI in materials science: A Pathway to Discovery and Insight,” MRS Bull., vol. 45, no. 8, pp. 614–623, 2020.
• X. Zhang, “Heat treatment effects on Inconel alloys.,” J. Mater. Eng., 2019.
• A. S. M. I. H. Committee and A. S. for M. H. T. Division, Heat treating, vol. 4. ASM international, 1991.
• P. Palumbo, G., & Ferro, “Comparison of CALPHAD-based tools for materials science applications.,” J. Phase Equilibria Diffus., no. 34(5), pp. 398–405, 2013.
• A. P. Zhao, J. C., Miodownik, “CALPHAD and its applications to materials science,” Prog. Mater. Sci., no. 51(3), pp. 241–312, 2006.
• C. Chen, J. Zhao, and L. Li, “Machine Learning for Alloy Design: An Overview of Recent Advances,” Mater. Today, vol. 48, pp. 15–27, 2021.
• K. Tran et al., “Active Learning Across Intermetallics to Guide Alloy Design,” Nat. Commun., vol. 9, p. 4074, 2018.
• A. Agrawal and A. Choudhary, “Deep Learning Applications and Challenges in Big Data Analytics for Manufacturing,” Manuf. Lett., vol. 15, pp. 12–19, 2019.
• L. Ward and C. Wolverton, “A Machine Learning Approach for Predicting and Understanding the Properties of Multiprincipal Element Alloys,” npj Comput. Mater., vol. 4, p. 36, 2018.
• J. Schmidt and others, “Recent Advances and Applications of Machine Learning in Solid-State Materials Science,” npj Comput. Mater., vol. 5, p. 83, 2019.
• A. Karpatne and others, “Regularization Techniques for Neural Networks: An Overview and Applicability in Materials Science,” J. Appl. Phys., vol. 128, no. 14, p. 141501, 2020.
• J. Ling, M. Hutchinson, and C. Wolverton, “Machine Learning for Multi-Property Predictions in Computational Materials,” Comput. Mater. Sci., vol. 138, pp. 140–152, 2017.
• Y. Zuo and others, “Accelerated Discovery of Metallic Glasses Using Machine Learning,” Mater. Horizons, vol. 6, no. 2, pp. 252–261, 2019.
• Y. Xu, J. Luo, and W. Chen, “Artificial Neural Networks in Alloy Design: Development and Challenges,” Adv. Eng. Mater., vol. 22, no. 4, p. 1901206, 2020.
• K. Kaufmann and K. S. Vecchio, “Artificial Intelligence in Materials Science: A New Frontier,” Mater. Sci. Eng. R Reports, vol. 144, p. 100595, 2020.