Huang Zihao, Shen Yirui, Zhou Mingxuan, Chen Manxi, Yang Heng, Li Shuang, Chen Hao-Sen
Beijing Key Laboratory of Lightweight Multi-Functional Composite Materials and Structures, Institute of Advanced Structure Technology, Beijing Institute of Technology, Beijing 100081, China.
School of Computer Science and Technology, Beijing Institute of Technology, Beijing, China.
Rev Sci Instrum. 2025 Apr 1;96(4). doi: 10.1063/5.0248652.
It is crucial to study the dynamic failure of materials because such failures occur widely in the fields of mechanics and earthquakes. However, capturing high-speed, high-resolution measurements of adiabatic shear bands temperature field during dynamic failure presents significant challenges. In this study, the spatial resolution of transient temperature field measurement was improved based on the self-developed high-speed infrared detector system combined with a novel multi-frame super-resolution method (3D-SwinIR). The super-resolution method employs 3D convolutional blocks and the Swin Transformer to effectively extract both shallow and deep features from multi-frame images, achieving high-quality super-resolution of high-speed infrared images. The effectiveness of the multi-frame super-resolution method was evaluated by a single structure measurement experiment, which indicated that spatial resolution has increased by 37%. An experiment was conducted on hat-shaped specimens using 3D-SwinIR for dynamic loading temperature measurement. The measurement errors for the peak and width of the temperature field were reduced by 51% and 33%, respectively. 3D-SwinIR provided a high spatial resolution measurement method for studying the dynamic thermal response of materials.
