Research on dual wavelength coaxial optical fiber sensor for detecting steel ball surface defects

Abstract The defects on a ground steel ball surface are very tiny and almost invisible; the existence of the defects will extremely influence the working stability of bearing system. To detect the surface quality on a steel ball effectively, dual wavelength coaxial optical fiber sensor technology is applied in this paper, and the defects are quantified as the characteristic parameters of displacement and roughness. The Matlab software was used to simulate the light intensity modulation characteristic ratio curve of displacement and roughness detection, the initial installation interval was determined as 0.8 mm between steel ball and the fiber sensor, and the wavelengths of light source were 1310 nm and 1550 nm. Experimental results showed that the coaxial optical fiber was suitable to detect the displacement, and dual wavelength to detect the roughness. The maximum errors of sensor displacement and roughness calibration experiment were 1.67% and 2.32% respectively in the measurement of steel ball with diameter 7.938 mm. The difference between the output curves was used to identify the defective balls from the good ones, which verifies the feasibility of the methods to detect the steel ball surface defects.