Citation :

Amol A. Dhakane, R. A. Kapgate, "Machine Learning-Based Modeling of Process-Structure-Property Relationships in IN-36 Alloys: A Random Forest and L18 DOE Approach," International Journal of Engineering Trends and Technology (IJETT), vol. 74, no. 5, pp. 250-262, 2026. Crossref, https://doi.org/10.14445/22315381/IJETT-V74I5P117

Abstract

The study combines a controlled experiment with guided machine learning to create explicable process structure property relationships of low-expansion Fe–36Ni-based alloys. Three different alloys (IN-36, IN-36Cr, IN-36Ti) on an L18 orthogonal array and cold-work levels (0, 30, and 60 percent) were used. They were then aged at 400 or 450 o C temperature for a period of 2, 7, or 24 hours and then stabilized at a temperature of 750 °C for 1 hour. The coefficient of thermal expansion (CTE, 20 -100 °C) and Vickers hardness, yield strength, ultimate tensile strength, elongation to failure, and stabilized grain size are the values that we measured under each of the conditions (Triplicate Averages). These data were used to train a multi-output Random Forest regressor, and a shallow neural network was used as a baseline. To ensure that it could work well with small data, we ran 6-fold cross-validation. The surrogate precisely measures the alterations within hardness, yield strength, ultimate tensile strength, elongation, and the size of grains; the coefficients of determination range 0.80-0.92, and the mean absolute error is low. Nevertheless, it does not perform so well with CTE since it possesses a small dynamic range. Feature-importance analysis indicates that cold work is the primary predictor for all mechanical and microstructural outputs, alloy type exerts a secondary yet significant influence, while aging temperature and time have lesser impacts within the examined range. Response surfaces created from the trained model show clear trade-offs between strength, ductility, and grain size. They also help find process windows that combine low CTE with high strength and acceptable elongation. In general, the work shows that a compact L18 design and an interpretable machine learning surrogate can help researchers find and improve processing strategies for Ni–Fe superalloys that don't expand much.

Keywords

Property linkages, Machine learning, Random Forest surrogate, Invar, Fe–36Ni alloys.

References

[1] Abu Anand, Szu-Jia Liu, and Chandra Veer Singh, “Recent Advances in Computational Design of Structural Multi-Principal Element Alloys,” iScience, vol. 26, no. 10, pp. 1-15, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[2] Ravindranadh Bobbili, B. Ramakrishna, and Vemuri Madhu, “Development of Machine Learning based Models for Design of High Entropy Alloys,” Materials Technology, vol. 37, no. 13, pp. 2580-2587, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[3] Jaka Fajar Fatriansyah et al., “Physical Property Prediction of High-Temperature Nickel and Iron-Nickel Superalloys using Direct and Inverse Composition Machine Learning Models,” Metals, vol. 15, no. 5, pp. 1-13, 2025.
[CrossRef] [Google Scholar] [Publisher Link]

[4] Cristiano Fragassa et al., “Predicting the Tensile Behaviour of Cast Alloys by a Pattern Recognition Analysis on Experimental Data,” Metals, vol. 9, no. 5, pp. 1-21, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[5] Kevin Kaufmann, and Kenneth S. Vecchio, “Searching for High Entropy Alloys: A Machine Learning Approach,” Acta Materialia, vol. 198, pp. 178-222, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[6] Saphal Kumar et al., “Machine Learning Enabled Processing Map Generation for High-Entropy Alloy,” Scripta Materialia, vol. 234, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[7] Juila Ling et al., “Machine Learning for Alloy Composition and Process Optimization,” Proceedings of the ASME Turbo Expo 2018: Turbomachinery Technical Conference and Exposition. Volume 6: Ceramics; Controls, Diagnostics, and Instrumentation; Education; Manufacturing Materials and Metallurgy, Oslo, Norway, pp. 11-15, 2018.
[CrossRef] [Google Scholar] [Publisher Link]

[8] Pei Liu et al., “Machine Learning Assisted Design of Γ′-Strengthened Co-Base Superalloys with Multi-Performance Optimization,” npj Computational Materials, vol. 6, no. 1, pp. 1-9, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[9] Sen Liu et al., “Physics-Informed Machine Learning for Composition-Process-Property Design: Shape Memory Alloy Demonstration,” Applied Materials Today, vol. 22, pp. 1-37, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[10] Srimanta Mitra et al., “A Machine Learning Approach to Predict the Structural and Magnetic Properties of Heusler Alloy Families,” Computational Materials Science, vol. 216, pp. 1-36, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[11] Dane Morgan, and Ryan Jacobs, “Opportunities and Challenges for Machine Learning in Materials Science,” Annual Review of Materials Research, vol. 50, pp. 71-103, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[12] Jian Peng et al., “Coupling Physics in Machine Learning to Predict Properties of High-Temperature Alloys,” npj Computational Materials, vol. 6, no. 1, pp. 1-7, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[13] Atiqur Rahman et al., “Physics-based Machine Learning Framework for Predicting Structure-Property Relationships in DED-Fabricated Low-Alloy Steels,” Metals, vol. 15, no. 9, pp.  1-32, 2025.
[CrossRef] [Google Scholar] [Publisher Link]

[14] Atul Raj, Joy Prakash Misra, and Dinesh Khanduja, “Modeling of Wire Electro-Spark Machining of Inconel 690 Superalloy using Support Vector Machine and Random Forest Regression Approaches,” Journal of Advanced Manufacturing Systems, vol. 21, no. 3, pp. 557-571, 2022.
[CrossRef] [Google Scholar] [Publisher Link]

[15] Rampi Ramprasad et al., “Machine Learning in Materials Informatics: Recent Applications and Prospects,” npj Computational Materials, vol. 3, no. 1, pp. 1-13, 2017.
[CrossRef] [Google Scholar] [Publisher Link]

[16] Vivek Revi et al., “Machine Learning Elastic Constants of Multi-Component Alloys,” Computational Materials Science, vol. 198, pp. 1-20, 2021.
[CrossRef] [Google Scholar] [Publisher Link]

[17] Sucheta Swetlana, Nikhil Khatavkar, and Abhishek Kumar Singh, “Development of Vickers Hardness Prediction Models Via Microstructural Analysis and Machine Learning,” Journal of Materials Science, vol. 55, no. 33, pp. 15845-15856, 2020.
[CrossRef] [Google Scholar] [Publisher Link]

[18] George Stephen Thoppil, Jian-Feng Nie, and Alankar Alankar, “Bayesian Approach for Inferrable Machine Learning Models of Process-Structure-Property Linkages in Complex Concentrated Alloys,” Journal of Alloys and Compounds, vol. 967, 2023.
[CrossRef] [Google Scholar] [Publisher Link]

[19] Cheng Wen et al., “Machine Learning Assisted Design of High Entropy Alloys with Desired Property,” Acta Materialia, vol. 170, pp. 109-117, 2019.
[CrossRef] [Google Scholar] [Publisher Link]

[20] Jie Xiong, San-Qiang Shi, and Tong Yi Zhang, “A Machine-Learning Approach to Predicting and Understanding the Properties of Amorphous Metallic Alloys,” Materials and Design, vol. 187, pp. 1-13, 2020.
        [CrossRef] [Google Scholar] [Publisher Link]