Comparison Study on Simulation Effect of Improved Simulation Methods for Packaging Random Vibration Test

Abstract

Gaussian vibration signal generated by the current shaking table according to power spectral density (PSD) in international standards was widely used for packaging random vibration test in laboratory. However, the actual road vehicle vibration is non-Gaussian. The key limitation is that the simulated Gaussian vibration cannot reconstruct the shock events buried in the vibration, which may result in inaccurate reliability evaluation of cushion packaging. A new simulation technique called shock extraction method has been proposed and demonstrated in our earlier study. This article is a further study of our previous research. The shock extraction method was compared with other three representative methods including the single level PSD, three way split spectral and wavelet decomposition in terms of simulation effect. The results revealed that the signal simulated by the shock extraction method was closest to the original signal, which possessed the same vibration intensity, duration, approximate statistical characteristics and PSD plots with the original signal. The simulation effect of the shock extraction method was the best, and the worst was the single level PSD. The wavelet decomposition and three way split spectral were somewhere in the middle.