Operational modal analysis based dynamic parameters identification in milling of thin-walled workpiece

The workpiece dynamics are an important factor in the planning of machining strategy. Usually, the structural dynamic parameters of workpiece are obtained by experimental modal analysis (EMA). However, the dynamics of thin-walled workpiece are varying due to material removal and tool-workpiece coupling. Operational modal analysis (OMA) provides a route to estimate the structural dynamic parameters during operation, but the input excitation of milling system is mainly periodic milling force, which violates the premise of OMA. In this paper, an output-only modal identification method for the dynamic parameters of thin-walled workpiece is proposed. The milling force is analyzed and the analysis results show the milling force contains white noise for OMA. However, strong harmonic components in milling force seriously interfere the identification of dynamic parameters. To remove the harmonic components in response signal, a revised least-squares (LS) method is proposed to fit harmonics with multiple fundamental frequencies. After the harmonic removal, OMA is conducted in frequency domain using the poly reference least squares complex frequency (p-LSCF) method based on positive power spectral density (PSD). To verify the feasibility of the proposed method, the simulation case is conducted by exciting a 3-degree of freedom structure using measured milling force signal, and the results show a good agreement with the theoretical values both in frequencies and damping ratios. Furthermore, a series of milling tests under different cutting parameters are conducted on a thin walled workpiece, where thin-film sensors are adopted to measure the structural response. The results show the method can be successfully applied to extract the workpiece dynamic parameters. Compared with results of EMA, the structural frequencies increase slightly, while the damping ratios show a large promotion due to the process damping.

Automated summary

This paper proposes an unsupervised modal identification method for the dynamic parameters of thin-walled workpiece, successfully extracting the structural dynamic parameters. The revised LS method effectively removes harmonic components in the milling force, enabling OMA using the p-LSCF method. Through simulation case and experimental validation, the results demonstrate the feasibility of the proposed method.