TY - GEN

T1 - comparison of different oma techniques and their application to an axial compressor test rig

AU - Amer, Mona

AU - Wallaschek, Joerg

AU - Seume, Joerg R.

AU - Ventura, Carlos E.

PY - 2022

Y1 - 2022

N2 - Identifying the modal properties of mechanical engineering systems while these are under operating conditions is still associated with major challenges, especially when said operation is induced by rotation of components. In this case, ambient vibration excitation, which is typically assumed for the application of output only techniques, is superposed by a deterministic periodic excitation component. This paper studies different techniques to perform an operational modal analysis in this particular situation. The methods are applied to an axial compressor test rig at the Institute of Turbomachinery and Fluid-Dynamics. Among them, three approaches, the Hilbert-Huang transform, the Bayesian OMA, and the stochastic subspace identification, are investigated. For the stochastic subspace identification, two different pole picking techniques to automatize the selection of stable poles from the stability diagram are studied. Furthermore, various approaches to deal with the harmonic excitation, which can lead to falsely identified modal parameters in the analysis process, are presented and applied to the data sets. Measurements at different rotational speeds of the rotor were collected and are used for the analysis. The results are summarized and the relative strengths and weaknesses of the different methods are discussed. The results obtained in this study are a first step towards a digital model to monitor the vibration behavior of rotating machinery using OMA techniques.

AB - Identifying the modal properties of mechanical engineering systems while these are under operating conditions is still associated with major challenges, especially when said operation is induced by rotation of components. In this case, ambient vibration excitation, which is typically assumed for the application of output only techniques, is superposed by a deterministic periodic excitation component. This paper studies different techniques to perform an operational modal analysis in this particular situation. The methods are applied to an axial compressor test rig at the Institute of Turbomachinery and Fluid-Dynamics. Among them, three approaches, the Hilbert-Huang transform, the Bayesian OMA, and the stochastic subspace identification, are investigated. For the stochastic subspace identification, two different pole picking techniques to automatize the selection of stable poles from the stability diagram are studied. Furthermore, various approaches to deal with the harmonic excitation, which can lead to falsely identified modal parameters in the analysis process, are presented and applied to the data sets. Measurements at different rotational speeds of the rotor were collected and are used for the analysis. The results are summarized and the relative strengths and weaknesses of the different methods are discussed. The results obtained in this study are a first step towards a digital model to monitor the vibration behavior of rotating machinery using OMA techniques.

KW - axial compressor

KW - harmonic detection

KW - Hilbert-Huang transform

KW - pole picking

KW - stochastic subspace identification

UR - http://www.scopus.com/inward/record.url?scp=85147843712&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85147843712

T3 - 9th IOMAC International Operational Modal Analysis Conference, Proceedings

SP - 77

EP - 90

BT - 9th IOMAC International Operational Modal Analysis Conference, Proceedings

A2 - Ventura, Carlos E.

A2 - Motamedi, Mehrtash

A2 - Mendler, Alexander

A2 - Aenlle-Lopez, Manuel

T2 - 9th International Operational Modal Analysis Conference, IOMAC 2022

Y2 - 3 July 2022 through 6 July 2022

ER -