Generation of sine on random vibrations for multi-axial fatigue tests

Abstract

Abstract This paper deals with the generation of multi-axial sine on random mixed vibration environments with specified references. Unlike pure vibration environments such as sine or random excitation, the difficulties of the superposition of sine on random vibration control lie in the separation of mixed vibration signals and simultaneous control of sine and random components. An improved correlation integral method is proposed to separate the mixed vibration signals. Considering the non-Gaussian random vibration characteristics, the control references of random component are comprised of multi-output response skewnesses, kurtoses, auto-power spectral densities, phase and coherence coefficients. The Hermite polynomial model transformation method is introduced to transform stationary Gaussian random signals to stationary non-Gaussian random signals with specified skewnesses and kurtoses. The control references of sine component consist of multi-output response amplitudes and phase differences. Closed loop control algorithms are presented to update the drive signals according to the deviations between responses and references. And finally, the results from a tri-axial sine on random mixed vibration control test show the feasibility and efficiency of the proposed methods.