IIT Guwahati and UK Universities Develop AI-Driven Framework
To Design Sustainable Metal Alloys Without Critical Raw Materials
GUWAHATI :
Researchers from the Indian Institute of Technology (IIT) Guwahati, in collaboration with leading UK universities, have developed a Machine Learning (ML)-based framework to design high-performance, sustainable metal alloys without using Critical Raw Materials (CRMs). The breakthrough offers a promising solution to reduce dependence on scarce and expensive materials used in advanced engineering applications.
The international research collaboration includes scientists from London South Bank University, University of Manchester, and the University of Leeds. Their work demonstrates how artificial intelligence can accelerate the discovery of CRM-free Multi-Principal Element Alloys (MPEAs) while maintaining or even improving mechanical performance.
Addressing the Challenge of Critical Raw Materials
Modern high-performance alloys—especially High-Entropy Alloys (HEAs) used in aerospace, gas turbines, and nuclear power systems—often rely on CRMs such as tantalum, niobium, tungsten, and hafnium. These elements face high supply risk, environmental concerns, and geopolitical vulnerabilities.
Reducing dependence on CRMs is essential for sustainable manufacturing, supply-chain security, and long-term industrial resilience.
Machine Learning–Driven Alloy Design
To overcome this challenge, the IIT Guwahati–led team developed a machine learning–assisted alloy design framework that focuses exclusively on CRM-free compositions.
Key highlights of the approach include:
- Classification of CRMs into three risk levels based on supply risk, economic importance, and global availability
- Creation of a database of 3,608 alloy compositions, primarily unary and binary systems
- Evaluation of multiple ML models, with the Extra Trees Regressor delivering the most accurate predictions of Vickers hardness
- Integration of nature-inspired optimisation techniques to identify high-hardness, CRM-free alloys
Using this framework, the researchers identified a novel CRM-free alloy composition:
Ti₀.₀₁₁₁NiFe₀.₄Cu₀.₄
The model predicted that this alloy would outperform a well-known CRM-containing alloy with a hardness of approximately 480 HV.
Experimental Validation Confirms AI Predictions
The newly proposed Ti–Ni–Fe–Cu alloy was successfully developed at laboratory scale at IIT Kanpur. Experimental testing showed that the measured hardness closely matched the ML predictions, validating the effectiveness of the AI-based design framework.
The methodology can be extended to design alloys optimised for multiple properties, including:
- Strength and ductility
- Heat and wear resistance
- Corrosion resistance
- Thermal conductivity
Expert Insight
Speaking about the research, Prof. Shrikrishna N. Joshi, Professor, Department of Mechanical Engineering, IIT Guwahati, said:
“The developed CRM-free alloy is particularly suited for applications where high hardness is a primary requirement, while also avoiding the use of Critical Raw Materials. This makes it attractive for both performance-driven and sustainability-focused applications.”
He further added:
“This is the first validated computational framework for designing CRM-free multi-principal element alloys using a unary- and binary-based compositional database, without relying on microstructural or processing parameters. The framework is entirely composition-driven and highly transferable, making it suitable for other material systems with limited experimental data.”
Potential Industrial Applications
The newly developed alloys can be used in:
- Wear-resistant mechanical components
- Tooling and surface-contact parts
- Automotive and industrial machinery applications
Publication and Future Roadmap
The research findings have been published in Scientific Reports, a prestigious journal by the Nature Publishing Group. The paper is co-authored by Prof. Shrikrishna N. Joshi, Dr. Swati Singh (IIT Guwahati), Prof. Saurav Goel (London South Bank University), Dr. Mingwen Bai (University of Leeds), and Prof. Allan Matthews (University of Manchester).
As the next step, the research team plans to collaborate with industry partners and research laboratories to test the materials under real-world operating conditions and move toward commercial deployment and scale-up.
