Modelling of gear stochastic vibration considering non-white noise errors

A model to simulate the non-Gaussian non-white-noise random errors of gears is developed with the influence of rotational speed considered. The gear transmission errors is first decomposed into harmonic and random components based on the spectral analysis. Then, a second-order Markov process with time-variant parameters is used to simulate the non-Gaussian non-white-noise random component of gears. The second-order Markov process is used as a shaping filter. The input of the filter is a white Gaussian noise process, and the output is the random erros of gears. Furthermore, a new dynamic and stochastic simulation model for the rotational bivration of gear transmission system is proposed. The influences of the random errors on vibration acceleration are examined by use of the developed simulation system.