Surface Roughness Prediction Technology For Ultra-precision Turning Based On Tool Vibration

Surface roughness values are an important parameter in characterising the surface quality of a workpiece,and predicting and reducing surface roughness is vital in both production life and scientific practice.If roughness values can be accurately predicted in real time,they can to some extent be used as a reference for parameter changes in the machining process in order to improve the quality of the machined surface and reduce production costs.This paper analyses the factors affecting roughness based on cutting principles,integrating them into deterministic factors containing tool geometry parameters,cutting parameters and mechanical properties of the material,as well as random factors caused by relative vibrational displacement in the direction of the depth of cut between the tool and the workpiece.Based on the cutting characteristics,a vibration-free two-dimensional morphological model is developed that takes into account the geometrical copy of the tool,the minimum cutting thickness,the surface rebound after machining and the plastic flow under stress concentration.Based on the periodicity of cutting motion,a vibration-free three-dimensional morphology model based on the tool tip trajectory is established.On this basis,the mechanism of the relative vibration displacement in the cutting depth direction on the surface morphology is investigated,and a morphological model considering the influence of vibration is established.A tool shank vibration displacement measurement system with an air-bearing capacitive sensor LVDT as the core is built to measure the shank vibration displacement of two diamond tools with different radii of the tool tip.Based on the equivalent three-degree-of-freedom dynamics model of the tool,a dynamics simulation was carried out in the software to obtain the equivalent tool tip vibration displacement signal.By smoothing and filtering,the noise and the influence brought by the low and medium frequency bands are eliminated,and only the high frequency band signals related to roughness are retained.The wavelet packet analysis is carried out on the processed signal to obtain the frequency bands with a larger energy share,and the reconstructed time domain signal spectrum is analysed to obtain the main frequency information with the largest amplitude in each band.Based on the obtained deterministic and random factors,a morphological prediction simulation is performed in the model to obtain roughness prediction values.The prediction results are compared with the actual measurement results to calculate the error,and the results show that the established shape-prediction model predicts roughness with high accuracy,which verifies the accuracy of the established roughness prediction model.The reconstructed vibration equation of the full vibration frequency band is then input into the model as a random factor,and the 3D predicted shape is compared with the real 3D shape,and it is found that the simulation results have the same arrangement pattern as the measured shape image,which proves that the model meets the requirements for 3D shape prediction under full vibration and verifies the correctness of the established simulation shape model.