AI-driven optimisation of metal alloys for space applications

Abstract This study investigates the application of artificial intelligence (AI) to accelerate the development of metal alloys for space applications, with a focus on aluminum, nickel, and titanium alloys. The research integrates data analysis, feature selection, and machine learning (ML) models to predict critical alloy properties for space environments, including Young’s modulus, yield strength, tensile strength, specific heat, and the coefficient of thermal expansion. The study optimises ML models, such as multi-layer perceptrons and ensemble techniques, demonstrating superior predictive accuracy compared to traditional benchmarks. Additionally, predictive models are employed to recommend novel alloy compositions, potentially enhancing specific properties crucial for aerospace applications. The proposed framework identifies significant predictors through correlation analysis, optimises models to achieve superior predictive accuracy, and recommends novel alloy compositions with the potential for enhanced performance in space applications. This study significantly contributes to materials science by integrating AI with traditional methods, offering a more efficient and targeted approach to alloy development, with the potential to enhance the design and durability of space vehicles and structures.