Quantitative characterization of physical structure on carbon fiber surface based on image technique

In order to quantitatively characterize the surface physical structure of a bundle of carbon fibers, scanning electron microscope (SEM) was used to photograph the fiber section and edge differential detection method was used to automatically identify the fiber boundary. By finding the smallest circumscribed convex polygon of fiber section and the minimum circumscribed rectangle of grooves though MATLAB, the groove width, depth, contour length and area for 20 monofilaments were calculated. The algorithm was verified by establishing a groove model and used to characterize grooves of three different carbon fibers. The results show quantitative difference of average groove size of three carbon fibers. The distribution of groove size accords with the lognormal distribution, and the fitting degree is between 0.8 and 0.97. The groove shape index rectangularity R and surface roughness Ra of three carbon fibers were also calculated. It indicates that both groove shape and groove size contribute greatly to the fiber surface roughness.