Quantitative analysis of metal fiber morphology by level set image segmentation algorithms

Abstract

Abstract Due to their advantages of high porosity, large specific surface area, and abundant surface topography, porous fiber-sintered plates are widely used in microreactors. The fiber surface in the cutting process generates a rich micro-surface morphology which can effectively improve the catalyst load and promote reaction efficiency. There is currently a lack of quantitative analysis of the microchannel surface topography features. In this study, a new image segmentation technique based on level set algorithms was developed to study the microscopic features of the metal fibers processed under different processing parameters, allowing for the establishment of a matching relationship between the microscopic roughness of the fiber surface and the equivalent diameter of the fibers. The experimental results show that the surface topography characteristic coefficient (STCC) was approximately inversely proportional to the cutting speed (r × min−1), which was approximately proportional to the feed (mm × r−1). The cutting depth (mm) effect on the STCC value of the fibers was not obvious. At low speed, with a cutting speed n = 20 r × min−1, feed rate f = 0.08 mm × r−1, and cutting depth ap = 0.11 mm, the STCC of the metal fiber reached a maximum value of 115.927. At high speed, with a cutting speed of n = 400 r × min−1, feed rate f = 0.06 mm × r−1, and cutting depth ap = 0.17 mm, the metal fiber STCC reached a maximum of 128.605. The parameters corresponding to the maximum STCC value can be used as the optimal turning parameters of metal fiber material for making a porous fiber microchannel reactor.