Work

Ignition-proof magnesium alloys with rare earth elements: A novel framework to predict combustion phases, surface morphologies, and hidden features using heating rates

Year: 2025
Type: article
Abstract: • ML framework predicted combustion phases and hidden surface behaviors. • Ignition-proof alloys with RE elements were successfully developed. • RE-rich alloys showed superior performance for aerospace and defense uses. • Image processing captured self-healing and crack-resistant oxide layers. • Heating rate simulations provided insights into ignition mechanisms. This study presents a proposed interdisciplinary framework for developing ignition-resistant magnesium alloys and analyzing their combustion behavior. It focuses on both commercial AZ31, AZ91, WE43 and formulated Mg-Gd-Y-Zn-Zr alloys with various rare earth elements (REEs) contents. The research integrates experimental methods, heating rate simulations, advanced image processing, and machine learning (ML) techniques to identify key mechanisms that enhance ignition resistance, particularly for aerospace and other industrial applications. A novel alloy composition, Mg-8Gd-6Y-0.6Zn-0.6Zr, demonstrated exceptional non-combustibility in air. The study is systematically to classifies the combustion process into distinct phases and surface morphologies by leveraging supervised and unsupervised learning models based on unseen heating rate features. Advanced image processing techniques reveal dynamic surface morphology changes, including thermal deformation, melting spots, gas bubble formation, and transformations during saturation and post-melting phases, while unsupervised ML models also validate these outstanding predictions of surface morphology features. Additionally, the research highlights the synergistic effects of REEs in forming dense, protective oxide layers, refining microstructures, and delaying ignition. This phase-based analysis provides the combustion behavior of magnesium alloys, which is crucial for evaluating their performance in industrial fire scenarios.
Language: English
Subfield: Biomaterials
Open Access status: gold
Estimated APC paid: $1,500
OpenAlex ID: W4407188244
FWCI: 1.11
Cites: 94
Cited by: 6
Related to: 10