On-machine inspection and compensation for thin-walled parts with sculptured surface considering cutting vibration and probe posture
Yanpeng Hao,Lida Zhu,Shaoqing Qin,Xiaoyu Pei,Tianming Yan,Qiuyu Qin,Hao Lu,Boling Yan
DOI: https://doi.org/10.1088/2631-7990/ad6de3
2024-08-13
International Journal of Extreme Manufacturing
Abstract:On-machine inspection has a significant impact on improving high-precision and efficient machining of sculptured surfaces. Due to the lack of machining information and the inability to adapt the parameters to the dynamic cutting conditions, theoretical modeling of profile inspection usually leads to insufficient adaptation, which causes inaccuracy problems. To address the above issues, a novel coupled model for profile inspection is proposed by combining the theoretical model and the data-driven model. The key process is to first realize local feature extraction based on the acquired vibration signals. The hybrid sampling model, which fuses geometric feature terms and time-frequency domain feature terms, is modeled by the lever principle. Then, the weights of each feature term are adaptively assigned by a multi-objective multi-verse optimizer. Finally, an inspection error compensation model based on the attention mechanism considering different probe postures is proposed to reduce the impact of pre-travel and radius errors on inspection accuracy. The anisotropy of the probe system error and its influence mechanism on the inspection accuracy are analyzed quantitatively and qualitatively. Compared with the previous models, the proposed hybrid profile inspection model can significantly improve the accuracy and efficiency of on-machine sampling. The proposed compensation model is able to correct the inspection error with better accuracy. Simulations and experiments show the feasibility and validity of the proposed method. The proposed model and corresponding new findings contribute to high-precision and efficient on-machine inspection, and help to understand the coupling mechanism of inspection errors.
engineering, manufacturing,materials science, multidisciplinary