The technology to design element combinations and manufacturing processes for alloys used in personal air vehicles (PAVs) and high-speed trains using 'Explainable Artificial Intelligence (XAI)' technology, which explains why artificial intelligence (AI) recommended specific combinations and processes, is expected to be applied to the development of various alloy materials for future mobility.

UNIST (President Yong-Hoon Lee) announced on the 21st that a research team led by Professor Im-Doo Jeong of the Department of Mechanical Engineering has developed a new high-strength, lightweight aluminum alloy design technology using AI. This research was conducted jointly with Gyeongsang National University, the Korea Institute of Materials Science, and POSTECH.

The joint research team developed a deep learning AI model that quickly identifies combinations of additive elements and processing conditions that yield optimal strength and ductility. Using a recommendation algorithm, they also obtained processing conditions for alloys predicted to exhibit superior mechanical properties. The recommendation process takes less than five minutes, allowing designers to quickly obtain desired processing conditions without complex and lengthy experiments.

By manufacturing an actual 7,000 series aluminum alloy following the new chemical composition and process conditions recommended by AI, we were able to produce a high-strength alloy with a yield strength of over 710 MPa (megapascals) while maintaining 20% ductility. Commonly used commercial materials have yield strengths of around 590 MPa (megapascals) and ductility of around 8%.

A particular advantage of the developed technology is that it enables alloy design engineers to quantitatively determine the effects of chemical composition and process conditions on the mechanical properties of the alloy.

This is thanks to the application of explainable AI technology, which allows you to understand why the AI recommended a specific combination or process, thereby increasing the reliability of the AI model's results.

The research team analyzed the microstructure of the alloy actually manufactured based on AI recommendations and confirmed that the interpretation of the 'explainable algorithm' was well-matched with actual material engineering theory.

“This technology can be widely applied to the production of not only aluminum alloys but also other lightweight alloy materials, and we expect that it will be able to drastically reduce the material development period and cost,” said Park Seo-bin, the first author.

Professor Seong Hyo-kyung of Gyeongsang National University, a co-corresponding author, said, “This technology has increased its reliability and applicability in that it allows us to directly understand the key factors that enhance strength through explainable artificial intelligence,” and “It will be able to make a great contribution to the development of high-strength, ultra-light materials in the future.”

Professor Im-Doo Jeong, who led the research as the corresponding author, said, “We discovered a lightweight metal with high mechanical properties that was difficult to find through experimental methods alone through fusion research with explainable artificial intelligence. This is expected to be a key technology that can maximize stability while meeting the ever-increasing demand for lightweight vehicles in the carbon-neutral era of mobility production.”

The results of this study were published in the Journal of Alloys and Compounds, an international academic journal ranked within the top 7% of JCR in the metal field.The study was published in the Journal of Alloys and Compounds in January. This study was supported by the National Research Foundation of Korea, the Ministry of Trade, Industry and Energy, and the Korea Evaluation Institute of Industrial Technology (KEIT).