Simulation Of Grinding Surface Topography And Its Application In Roughness Measurement Based On Machine Vision

Grinding is an important processing technology of mechanical parts,and the surface roughness is an important index to evaluate the surface quality of the workpiece,which directly affects the performance and life of the workpie ce etc.Therefore,the research and measurement of the surface roughness is of great significance.Compared with the traditional grinding surface roughness measurement method,the grinding surface roughness measurement method based on machine v ision has the advantages of non-contact,large measuring range and high efficiency,and is beneficial to realize online detection.However,in the process of setting up the image index,a large number of grinding samples need to be prepared for experimental verification,and the manufacture of grinding samples has some problems such as time-consuming,laborious and poor controllability.At the same time,the construction of visual experiment is also time-consuming,high cost,easy to be disturbed by the external environment and so on.With the development of computer simulation technology,it is possible to simulate the surface topography of grinding.It provides a new way to solve the problem of the production of grinding samples.At the same time,the simulation sample can be applied to the computer simulation imaging system to simulate the process of grinding surface roughness measurement based on machine vision,which can not only reduce the actual cost of experiment and the disturbance of the external environment,but also improve the experimental efficiency and shorten the development cycle of the machine vision roughness detection system.In order to study the simulation of grinding surface topography and its application in roughness measurement based on machine vi sion,this paper conducts a simulation study from the grinding wheel topography and the grinding surface topography respectively.and solves the difficult problem of obtaining the grinding samples with different roughness,and compares the simulation topography with the measured topography from qualitative and quantitative perspectives,and the effectiveness of the simulation method is verified.Finally,the grinding surface simulation topography is applied to the surface roughness measurement based on machine vision.By comparing with the actual experiment,the validity of the simulation sample and the simulation imaging system is verified.The main research work of this paper is as follows:Firstly,the grinding process technology,classification and advan tages are introduced,and the definition,evaluation parameters and calculation methods of surface roughness are expounded and explained,which provide a theoretical basis for the next grinding surface topography simulation.At the same time,based on the rough surface imaging mechanism,the overall hardware scheme of the roughness detection system based on the machine vision is designed and selected,which provides the basis for the subsequent experimental verification.Secondly,in the grinding process,the grinding wheel is a machining tool and participates in the forming process of the grinding surface topography.Therefore,the surface topography of the grinding wheel is simulated.First,the surface topography data of the grinding wheel are obtained.By plotting the histogram of the frequency distribution and the data analysis based on the Kolmogorov-Smirnov(K-S)method,it is concluded that the surface height of the grinding wheel obeys the non-Gauss distribution.Then,the surface topography of the grinding wheel is generated by methods of filtering and Fourier transform.At last,the roughness parameters Ra,Rq,Rsk and Rku of the simulation topography and the actual wheel topography are calculated respectively,and the statistical characteristics ar e compared.The results show that the topography of the simulated wheel has a high consistency with the statistical characteristics of the topography of the measured wheel,which verifies the effectiveness of the simulation method and lays a foundation for the simulation of grinding surface topography.Then,the simulation of grinding surface topography is studied.First,the experimental samples with different roughness levels are obtained through the HR-618 S surface grinder.Then,a mathematical model of the grinding surface topography is established based on the multi-grain grinding trajectory.Considering the factors of multiple grinding and the grinding wheel wear,the mathematical model is modified,and the topography of the grinding surface is simulat ed.In the end,the one dimension,two-dimensional and three-dimensional morphologies of the simulation grinding surface topography are compared with the actual grinding surface topography,and the roughness parameters are calculated and compared.The experimental results show that the simulation grinding surface topography has a good similarity with the actual grinding surface topography,which proves the effectiveness and feasibility of the grinding surface topography simulation method.Finally,the application of grinding surface topography simulation in surface roughness measurement based on machine vision is studied.First,the mechanism of the surface roughness measurement based on machine vision is analyzed,and a new structural index SSCY is proposed by studying three indexes of EP,CD and SCY.Second,the point cloud data of the grinding surface topography is fitted and stretched to form a three-dimensional grinding sample block.The three-dimensional grinding sample block is imported into Tracepro optical simulation software and the simulation imaging environment is designed.The simulation pictures of the red and green light sources formed on the surface of the simulation grinding sample blocks with different roughness levels are obtained.Then,the actual pictures of the same red and green background light sources formed on the actual grinding surface are acquired by building the same experimental platform.At last,the values of different indexes of the simulation pictures and the actual pictures are separately calculated,and the relationship curves between the index and the surface roughness are plotted,and the performance of different indexes is evaluated using the Pearson linear correlation coefficient(PLCC).The experimental results show that the simulation imaging experiment results are consistent with the actual imaging experiments.The feasibility of the simulation imaging method is verified.At the same time,by comparing the PLCC of the picture indexes before and after preprocessing,the validity of the proposed preprocessing method is verified,which shows that the index SS CY has good feasibility.