Research On Complex Surface Slow Tool Servo Turning Path Generation And Surface Roughness Prediction Technology

Complex optical surface components play an increasingly important role in the fields of optics,medicine and national defense because of their excellent optical properties.Slow tool servo CNC lathe can drive the diamond tool relative to the workpiece end face in the cylindrical coordinate system through C,X and Z axis linkage to achieve the cutting motion of spiral trajectory to achieve the machining of complex surface.Slow tool servo turning technology is an international research hotspot,and this paper takes slow tool servo turning of complex surface as the research object,and studies tool path generation and simulation,surface roughness prediction technology and complex surface error evaluation technology.The specific content of the research is as follows:1.Complex surface slow tool servo turning tool path generation and simulationTool path generation is one of the key technologies to apply slow tool servo technology to process complex surfaces,and the advantages and disadvantages of the generation scheme directly affect the final surface accuracy.In this paper,based on the analysis of the principle and advantages and disadvantages of the equal angle and equal arc length tool contact discrete method,the normal tool compensation algorithm,and the Z-direction tool compensation algorithm,a tool shape compensation algorithm based on the chord length perpendicular is proposed;the PVT interpolation provided by IMAC is selected to realize the continuous tool motion trajectory;the mathematical essence of PVT interpolation Hermite interpolation is analyzed.Based on this,three-point method and area method are selected as PVT inlet parameter generation algorithms.Taking the torus surface and sinusoidal array surface as examples to compare the X-axis and Z-axis motion characteristics,the accuracy of the complex surface shape and the feasibility of the tool compensation algorithm.The results show that the tool trajectory with the chord length vertical compensation algorithm meets the requirements of the motion characteristics of all axes of the slow tool servo lathe and can be applied to turning processing.2.Study of surface roughness prediction technology for slow tool servo turningResponse surface experiments were conducted with tool nose radius,feed,depth of cut and discrete angle as cutting parameters and surface roughness value(R_a)as response value.A quadratic regression prediction model was established based on the experimental results,and the ANOVA concluded that the tool nose radius had a highly significant effect on the surface roughness.This is followed by the discrete angle and feed,and the depth of cut has no significant effect.The minimum value of R_a is used as the target to obtain the best combination of process parameters.The experimental results were used as sample data to establish BP neural networks with different numbers of hidden layer neurons,and the mind evolutionary algorithm(MEA)was selected to optimize the BP neural networks.The R_aprediction values of RSM,BP neural network and MEA-BP neural network under the optimal process parameters were obtained.Turning experiments were carried out under the optimal process parameters to obtain the real value of R_a,and the value of R_a was 0.052μm.The prediction relative error is obtained by comparing the real value with the predicted value to evaluate the prediction performance of each model.The results show that the MEA-BP neural network has the highest prediction accuracy,and the average relative error of prediction is8.1%.The quadratic regression model under the response surface method was the next best,but still within the error tolerance.The BP neural network had the worst prediction accuracy,with an average relative error of 22.55%in prediction.3.Research on turning experiment and error evaluation technology of complex surfaceThe CNC machine tool host control software was written in C#language to realize the independent machining of complex surface;the experiments on the surface roughness of toroidal surface and sinusoidal array surface under different process parameters were completed,and the experimental results showed that the surface roughness was influenced by the turning process parameters,and the correlation with the tool compensation algorithm was not significant under the same parameters.In order to obtain the shape error of complex curved surface under slow cutter servo turning,a complete error evaluation scheme was proposed.Firstly,to obtain the measured data of the surface,coordinate measuring machine(CMM)was used to obtain the discrete spherical core data with a circumferential detection path;secondly,Zernike’s first six polynomials were selected to reconstruct the discrete spherical core data in order to obtain the surface roughness.Finally,the ICP-GA algorithm was used to match the measured data with the theoretical data to obtain the complex surface shape error.The evaluation results of the torus surface shape error are as follows:the surface shape error PV value of the chord length vertical compensation algorithm is 0.008 mm,and the surface shape error PV value of the Z-direction compensation algorithm is 0.017 mm.The results show that the lower shape accuracy of the chord length vertical compensation algorithm is better than that of the Z-direction compensation algorithm.