圆度误差的快速傅立叶变换分析

The Analysis of Roundness Error with FFT

回转零件正截面轮廓的圆度误差可以视为周期函数。对正截面轮廓等角度间隔采样获得的数据点,可进行离散傅立叶变换(DFT)。通过对数据点序列的快速傅立叶变换(FFT)结果加以分析,可以发现隐藏在圆度误差信号中的各种频率成分。文章首先简要介绍了圆度误差的数学模型,然后分析了各种频率成分对圆度误差的影响。最后提出了圆度误差计算的新方法。

The roundness error of the profile of the rotating parts can be viewed as a periodic function. Therefore, Discrete Fourier Transform can be applied to the data points we obtained by dividing the trace of the profile at equally spaced angles. By studying the data series of the trace with FFT, we can find various frequency components hidden in the signal of the roundness error. The article first gives a brief introduction to the mathematical model of roundness error. Then it focuses on how the various frequency components contribute to the roundness error. Finally, a new approach of evaluating roundness error is provided. Since the data points contain all the information of the roundness error, we can obtain the series of the roundness errors by eliminating the unwanted components in the frequency domain with FFT and changing back to the time domain with IFFT.