Research On Surface Roughness Measurement Method Based On The Principle Of Spectral Confocal

Surface roughness is a significant technical indicator to characterize the microscopic unevenness of the workpiece surface,and it is especially major to measure it accurately and efficiently.Traditional measurement methods have shortcomings such as slow detection speed and easy scratching of the surface.In recent years,surface roughness measurement methods based on spectral confocal detection technology have developed rapidly.This non-contact detection technology has high efficiency and fast detection speed,which can avoid some problems in contact detection.Therefore,based on the principle of spectral confocal,this subject has studied a spectral confocal measurement method based on white LED light source and fiber coupling,which can realize surface roughness measurement.The major contents of this subject are as follows:Several common surface roughness measurement methods are analyzed.In order to achieve high-precision,non-contact surface roughness measurement,a spectral confocal measurement system is designed.Research on the surface roughness measurement method based on the principle of spectral confocal is carried out.Based on the discussion of the principle of the spectral confocal measurement method,the performance of the system is analyzed.The structure of the optical fiber-type spectral confocal measurement system is designed,and the design of the confocal optical system and the design of the spectral measurement system are completed by using the ZEMAX optical design software.In the design process of the confocal optical structure,the working wavelength of the dispersive objective lens is 400nm-700nm,the axial dispersion reaches 1.705mm,At the same time,the imaging situation of the dispersive objective lens is analyzed,and the Y-type fiber with a numerical aperture of 0.22 is selected.In the design process of the spectrum measurement system,the components of the system were analyzed,and the system structure based on the grating and photoelectric array detector CCD was designed.The resolution of the spectrum measurement system can reach 0.5nm.The spectral information processing algorithm is designed,which includes spectral information preprocessing,filtering and denoising algorithm,peak wavelength extraction algorithm and data fitting algorithm,which solves the influence of dark signal in the preprocessing process,and the influence of noise on peak wavelength extraction.The nonlinear fitting problem between the peak wavelength and the focal point distance.According to the characteristics of the spectral signal,median filtering is used to eliminate random noise.In order to ensure accurate peak finding,Gaussian algorithm is used to extract the peak wavelength.The designed global optimal segmented fitting method can eliminate the ill-conditioned problem of data fitting and effectively improve the accuracy of data fitting.Finally,a surface roughness detection system was built and debugged,and the comparative sample with a nominal roughness value of 6.3 ?m was tested.According to the related evaluation method of surface roughness,it is shown that the surface roughness parameter Ra value is 6.29?m and the Rz value is 29.52?m detected by this system.