Source abstract
Abstract
Magnesium (Mg)-based alloys exhibit excellent mechanical and electrical properties for fuel cell bipolar plates. Traditional Mg-based alloy design relies on theoretical and experimental methods with long design cycles based on trial-and-error that focuses on one material. We analyzed the elastic modulus and electrical conductivity of 248 Mg-based alloys by combining density functional theory (DFT) and machine learning (ML). The decision tree classification (DTC) and gradient boosting regression (GBR) algorithms were chosen to predict the mechanical stability and electrical conductivity of the Mg-based alloys (the accuracy and R 2 values of DTC and GBR were 0.94 and 0.83, respectively), and the atomic volume, density, and boiling point were identified as the three most important factors affecting alloy properties. ML successfully predicted the mechanical stability and scalar conductivity of HfMgPt 2 ( T = 350 K, 0.666 × 10 21 ). We constructed a new Mg-based quaternary alloy, Mg 3 ZrTi 4 Ir 8, based on the predicted Mg-based ternary alloy with a high scalar electrical conductivity and verified its good mechanical properties. This work provides a theoretical basis for the design of Mg-based alloys for fuel cell bipolar plates, accelerates the screening of Mg-based alloys, and reveals the reliability of the “composition–property–composition” model established by DFT and ML.
Source: openalex_inverted_index · Quality: baseline
Affiliations
School of Mechanical Engineering, Shaanxi University of Technology, Hanzhong, Shaanxi 723001, China
Shaanxi Key Laboratory of Industrial Automation, Shaanxi University of Technology, Hanzhong, Shaanxi 723001, China
School of Physics and Materials Science, Nanchang University, Nanchang 330031, China
School of Physical Science and Technology, Southwest University, Chongqing 400715, China
Funding
Shaanxi University of Science and Technology · Scientific Research Plan Projects of Shaanxi Education Department · Natural Science Basic Research Program of Shaanxi Province · National Natural Science Foundation of China