A Method for Identification of Machine-tool Dynamics under Machining

Dynamic properties of machine-tool structures are likely to change under machining conditions. Thus, the dynamic parameters obtained by traditional experimental modal analysis in the static state may not characterize accurately the dynamics of the machine tool structure in operation. This paper proposes a new method of so-called AEMA (Active Excitation Modal Analysis) to identify the dynamic modal parameters of a machine tool structure during machining. A random cutting excitation technique realized by cutting a specially designed workpiece is proposed to provide strong and evenly distributed excitation within the frequency range of interest. The surface of the workpiece has a long narrow random zigzag width, which randomizes the resulting cutting forces. The LSCE (Least Square Complex Exponential) method is employed to estimate the modal parameters from just the measured responses. Then an algorithm based on two novel tools, the harmonic frequency fence and the spectrum abruptness ratio, is presented to eliminate the harmonic modes attributed to AC power and rotation frequency. The abruptness ratio is used to detect the basic frequency, and then the fence filters out the harmonic modes caused by peaks at integer multiples of the basic frequency through narrow frequency fence slots followed by a damping ratio limit. Finally, the proposed AEMA method is experimentally validated and shows satisfactory results.