材料科学
激光器
电工钢
硅
光学
激光功率缩放
激光烧蚀
微观结构
磁力显微镜
复合材料
磁化
光电子学
磁场
量子力学
物理
作者
Julien Dupuy,O. Maloberti,Nesser Manar,Yves Hernandez
摘要
We investigate the influence of different laser parameters on Grain-Oriented electrical silicon steels (GO SiFe). The advanced use of ultra-short laser pulses for scribing steel sheets with magnetic domains refinement and loss reduction purpose is proposed. The GO steel samples are commercial ones, commonly used in transformer cores. We compare laser impact on domains refinement obtained by thermal effects (with CW and long laser pulses), with thermal-free ablation produced by an ultrashort source. The laser scribing conditions (power or pulse energy, speed and pattern geometry, size and orientation) are optimized to have minimal deformation on material surface, while maximizing the influence and benefits on the magnetic properties. The steels used for these experiments have got one fixed composition (Silicon Iron alloy with 3% Silicon), the sheets used have got one thickness of 0.23 mm, very big grains size around 3 mm in average and one preferred lamination direction. The effects of the laser on steel samples are studied with different techniques, in order to show the impact on the surface topography, the metal microstructure and on the scalar magnetic properties. The confocal microscopy is used to probe the surface topography. The MFM (Magnetic Force Microscopy) is used to analyze in detail the formation of magnetic domains in the vicinity of laser patterns at the nanoscopic scale. The MOKE microscopy (Magneto-Optic Kerr Effect) is used to get images of magnetic domains at the microscopic scale. The SST (Single Sheet Tester) and the Epstein frame are used to measure the static and dynamic behavior and the power losses with two or one directional flux at the macroscopic scale. The motivation is to find the best laser treatment and patterns to improve the scalar magnetic properties and to reduce the energy consumption of electrical devices made with this steel.
科研通智能强力驱动
Strongly Powered by AbleSci AI