Kinetic Study of the Crystallization Behavior of Mold Flux Based on Computer Vision Technique
摘要
A novel method on kinetic study of mold flux crystallization based on computer vision technique was proposed, after the high-temperature single hot thermocouple technology (SHTT) test. The results suggested that the crystalline fraction of crystal in molten flux can be efficiency and accuracy obtained through the image smoothing, grayscale, and binary processing, as well as Canny edge detection via the open-source computer vision technique. Moreover, the SHTT tests show that the crystallization ability of flux A is higher than that of flux B, as it takes 92s for flux A and 123s for flux B to get completely crystallized at the isothermal process. Kinetics analysis indicates that the isothermal crystallization mechanism for flux A is constant nucleation rate, three-dimensional growth, controlled by diffusion, while it is constant nucleation rate, two-dimensional growth, controlled by diffusion for flux B.