Aerospace Technic and Technology

The work analyzes vibration signals obtained by simulating a turbine of a complex rotor system, for example, an aviation gas turbine engine, under conditions of stationary and non-stationary excitations. Four modes of vibration excitation are considered: stationary poly-harmonic excitation with the frequency of rotor rotation and super-harmonic components; stationary poly-harmonic excitation with the frequency of rotor rotation and sub-harmonic components; non-stationary vibration excitation with a linear increase in the rotor speed with super-harmonic and sub-harmonic components of the instantaneous rotor speed. In the course of the turbine model, vibration signals are generated, which are further analyzed without taking into account and taking into account additive noise. For signal processing, fractal and time-scale (wavelet) analysis were used. The determination of the fractal structure of the simulated vibration signals is made based on R / S analysis, or the method of normalized scope, as a result of which the Hurst exponent is determined. The Hurst exponent is a number that is interpreted as the ratio of the “strength” of a trend to the signal noise level and is used in the study to interpret the received vibration signals. The results showed that the vibration signals obtained in all considered modes of vibration excitation without taking into account the additive noise, in terms of the Hurst exponent, are classified as anti-persistent trend-non-stable signals. Taking into account additive noise, the Hurst exponent increases, the vibration properties in stationary excitation modes approach persistence and the appearance of a trend, and in non-stationary vibration excitation signals approach to processes such as white noise. For the vibration signal obtained at stationary poly-harmonic excitation with super-harmonic components, a preliminary wavelet - decomposition was carried out into a set of approximations and details, followed by determination of the Hurst exponent for each element of decomposition. The results obtained showed an ambiguous change in the Hurst exponent for various decomposition elements. The obtained results can be used to improve the methodological and algorithmic support systems for functional diagnostics of complex rotor systems with the appearance and propagation of damage to their rotating elements.

gas turbine engine; vibration signals; signal processing; fractal analysis; Hurst exponent; R/S analysis; wavelet decomposition

Chyhyryn, V. S., Sukhoviy, S. I. Vibroakustyka i vibrodiahnostyka hazoturbinnykh dvyhuniv. Navch. posib [Vibroacoustics and vibration diagnostics of gas turbine engines. Teach manual.]. Kharkiv, National Aerospace University "Kharkiv Aviation Institute", 2012. 264 p.

Bouraou, N. O., Yatsko, L. L., Pavlovskyi, O. M., Sopilka, Y. V. Metody tsyfrovoyi obrobky syhnaliv dlya vibratsiy-noyi diahnostyky aviatsiynykh dvyhuniv. Monohrafiya [Methods of digital signal processing for vibration diagnostics of aviation engines. Monograph]. Kiev, National Aviation University, 2012. 152 p.

Pavlovskyi, O., Bouraou, N., Iatsko, L. Multilevel vibration control system of aviation gas-turbine engines. Vibrations in Physical Systems, 2012, no. 25, pp. 323-328.

Bouraou, N. I., Ignatovych, C. P., Pazdrii, O. Y. Vykorystannya fraktalnoho analizu chastotno-chasovykh spektriv vibroakustychnykh syhnaliv dlya diahnostyky hazoturbinnykh dvyhuniv [Using fractal analysis of frequency-time spectra of vibroacoustic signals for diagnostics of gas turbine engines], Visnik NTUU KPI. Ser. Radioteh. radioaparatobuduv, 2018, no. 74, pp. 73-83.

Feder, J. Fractals. New York, Plenum Press, 1988. 254 p.

Kononyuk, A. E. Diskretno-nepreryvnaya matematika. (Poverkhnosti). V 12-i kn. Kn. 6. ch. 2 [Discrete-continuous mathematics (Surfaces). In the 12th book, Book 6 Part 2]. Kiev, Osvita Ukrainy, 2016. 618 p.

Zakharov, V. S. Dinamicheskiye i fraktal'nyye kharakteristiki vremennykh ryadov vydeleniya seysmicheskoy energii [Dynamic and fractal characteristics of time series of seismic energy release]. Nelineynyy mir, 2010, vol. 8, no. 4, pp. 234-242.

Aksenov, V. Y., Dmitriyev, V. N. Algoritmy fraktalnogo analiza vremennykh ryadov v sistemakh monitoringa sensornykh setey [Algorithms of fractal analysis of time series in monitoring systems of sensor networks]. Vestnik AGTU. Ser.:Upravleniye, vychislitel'naya tekhnika i informatika, 2012, no. 1, pp. 91-96.

Kiselev, B. V. Ob interpretatsii statisticheskogo R-S analiza. Pokazatel' Kharsta [The interpretation of statistical R-S analysis. Hurst exponent]. Voprosy geofiziki, 2007, no. 40, pp. 121-130.

Chumak, O. V. Entropii i fraktaly v analize dannykh [Entropies and fractals in data analysis]. Moscow, Izhevsk, NITS "Regulyarnaya i khaoticheskaya dinamika" Institute of Computer Science, 2011. 164 p.

Sanchez Granero, M. A., Trinidad Segovia, J. E., Garcia Perez, J. Some comments on Hurst exponent and the long memory processes on capital markets. Physica A, 2008, no. 387, рр. 5543-5551.