Abstract
Motivated by the overall progress towards "smart rotating machinery" for extending life time, safety and availability of rotating machines, the main target of this project is the development of a general strategy that acquires extended information via sensors, processes this information for fault diagnosis and is able to react to the occurrence or the progression of faults by setting up an appropriate control action. This strategy of Fault Tolerant Control is developed for rotor systems, supported by hydrodynamic bearings, being those machines strategic equipments in conventional and off-shore energy power plants. Therefore, the potential industrial applications intended for the results are mainly power generation rotating machinery. Given the complexity of the mechanical systems considered, uncertainties are unavoidable and must be properly considered in various phases, including evaluation of uncertainty propagation, parameter identification from uncertain models and/or noise-contaminated dynamic responses, reliability evaluation, and robust control in the presence of uncertainties. In this context, the expert skills in rotor-bearing-structure systems modeling (particularly those supported by hydrodynamic bearings), fault diagnosis and identification, uncertainty propagation and reliability analysis and active control of such systems will be combined. The faults to be considered in the model are based on linear and nonlinear modeling of critical components of rotor-bearing-structure system, considering unbalance effects, shaft misalignment, bearing cavitation, bearing wear and structural dynamic effects. On the basis of theoretical fault models, methods of fault detection and isolation (FDI) are to be set up for providing accurate information on occurring faults in the rotor system. Based on this information, a stabilizing and vibration reducing control is applied. Robust controllers will be applied accounting for eventual nonlinearities and variations of the faults. The influence of uncertainties on the performance of FDI methods and the levels of reliability of the controlled system will be evaluated. The developed strategies and the fault models will be validated in rotor test rigs with different characteristics, mainly for hydrodynamic supported rotors with an attached electromagnetic actuator, enabling realization of the considered strategies for faults in distinct characteristics. This work, therefore, aims at the progress of knowledge in this field, in order to foster the industrial applicability of the methods investigated. (AU)
Scientific publications
(36)
(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)
ALVES, DIOGO STUANI;
DANIEL, GREGORY BREGION;
DE CASTRO, HELIO FIORI;
MACHADO, TIAGO HENRIQUE;
CAVALCA, KATIA LUCCHESI;
GECGEL, OZHAN;
DIAS, JOAO PAULO;
EKWARO-OSIRE, STEPHEN.
Uncertainty quantification in deep convolutional neural network diagnostics of journal bearings with ovalization fault.
MECHANISM AND MACHINE THEORY,
v. 149,
JUL 2020.
Web of Science Citations: 0.
BALTHAZAR, JOSE M.;
TUSSET, ANGELO M.;
BRASIL, REYOLANDO M. L. R. F.;
FELIX, JORGE L. P.;
ROCHA, RODRIGO T.;
JANZEN, FREDERIC C.;
NABARRETE, AIRTON;
OLIVEIRA, CLIVALDO.
An overview on the appearance of the Sommerfeld effect and saturation phenomenon in non-ideal vibrating systems (NIS) in macro and MEMS scales.
NONLINEAR DYNAMICS,
v. 93,
n. 1, SI,
p. 19-40,
JUL 2018.
Web of Science Citations: 6.