Introduction. Transform of the vibration signal, while the power spectrum has been estimated through the Blackman—Tukey method. However, because experimental vibration data are characterized by a combination of harmonic and stochastic processes, both techniques have proven inadequate for analyzing mixed signals.Problem Statement. The evaluation of rotor balance in gas turbine engines has been commonly based on themagnitude of the regular component of vibrations at the rotor speed. The amplitude spectrum of a gas turbine engine has traditionally been determined using the Fourier mponent of vibrations at the rotor’s rotational frequency. Yet, the vibration signal contains numerous additional components, including stochastic ones, which should be accounted for to obtain reliable estimates. Modeling the vibration signal as a periodically non-stationary random process (PNRP) has allowed separating the deterministic and the stochastic components. Purpose. The study has aimed to conduct a comparative analysis of periodic non-stationarity in vibration signals of gas turbine engines with balanced and unbalanced rotors.Materials and Methods. Vertical vibration components of balanced and unbalanced gas turbine engines in the low-frequency range (<2 kHz) have been analyzed. The vibration signal has been modeled as a periodically non-stationary random process, and an LS-functional has been applied to identify the fundamental frequencies of both deterministic and stochastic components.Results. Correlation and spectral functions of vibration signals have been estimated, and the fundamental frequencies of their regular and stochastic components have been determined. It has been shown that the vibration spectra of engines with both balanced and unbalanced rotors are mixed. The deterministic component has exhibited a polyharmonic spectrum, and the harmonic amplitudes have been quantified. Furthermore, engine vibrations have been shown to exhibit second-order periodic nonstationarity.Conclusions. The proposed vibration indicators have provided a basis for improving engine balancing during adjustment and repair procedures. The study has also highlighted the potential of these indicators for early detection of rotor defects. Future research will focus on analyzing the polyharmonic structure of vibration spectra and developing a refi ned methodology for defectdiagnostics at incipient stages.
The results of the processing of the wind turbine gearbox vibration signals are presented. The least square estimators of the periods of the deterministic part of the vibration signal and time variations power of stochastic part are obtained. The amplitude spectra of the deterministic and stochastic oscillations for different degrees of the gearbox damage are analyzed. The covariance structure of stochastic part of vibration was analyzed. The efficient indicator for practical evaluation of gear teeth damage, formed on the basis of amplitude spectra was proposed.
The estimators of deterministic part parameters for biperiodically nonstationary vibration signal, obtained by the least squares method (LSM), are analyzed. The formulae for LSM-estimator variances and bias, which describe their dependencies on realization length and signal covariance components, are obtained. It is shown that LSM allows to avoid leakage effects and to obtain unbiased estimators of biperiodic nonstationary vibration signal deterministic part for arbitrary realization length and harmonic frequencies.
