Data- driven Correction of Gear Clamping Errors in Laser Interferometry

With the advantages of high efficiency, high precision and non-destructiveness, laser interferometry has been widely used in the measurement of surface topography errors of objects. Surface morphology quality is an important indicator of gear manufacturing accuracy, and its micron-level, or even submicron-level error will also directly affect the stress distribution on the surface contact surface of the parts and transmission performance, not only the overall performance of the equipment, but also related to the length of service life. During the measurement process, the quality of the interferometric image, as the key data information collected, has a direct impact on the measurement accuracy. Laser interference is an effective method to measure the geometrical error of helical gear teeth, and the quality of the interference image, as the raw data collected in the measurement, directly affects the final measurement accuracy. In the measurement, the precise clamping position of the measured gear directly determines the information integrity and processability of the acquired interference image, which is one of the prerequisites and key conditions for full-area, high-precision measurement of the tooth flank. The article proposes a data-driven gear clamping error correction method in order to obtain high-quality interferometric images. First, the data-driven dataset is constructed by collecting measured and simulated data; secondly, one-dimensional calibration points were calculated in the direction of shaft inclination to derive the trend of geometric distortion index with clamping position and predict the direction of clamping error; then, the masking ratio in the measured tooth interference image is used as a bridge to establish the mapping relationship between the measured and simulated interference images, and the optimal inclination approximation is used to calculate the adjustment angle; finally, the rationality of adjustment is verified according to the distortion law. Relevant experiments were designed, and the results showed that the adjusted interferometric image occlusion ratio was reduced by 4.87%, the calibration accuracy of the PSD scale was 1, and the mean PSD value was improved by 0.27. The quality of the interferometric image can be significantly improved after adjustment by this method, and the feasibility and correctness of this method are verified at the same time. The effective improvement of the stripe quality of the interferometric image provides technical support and means of realization for the improvement of the measurement accuracy of the interferometric image.