COMPUTATIONAL ANALYSIS OF ROTOR SYSTEMS WITH SMART MATERIAL FOIL AIR BEARINGS

Abstract

Research into foil air bearings (FABs, also known as air foil bearings or gas foil bearings) is a critical enabler of the rapidly expanding technology of oil-free turbomachinery, which is driven by well-documented environmental and technological benefits gained from eliminating oil or grease lubricated bearings. Such bearings are distinguished by their reliance on an air film that is pressurised by the hydrodynamic effect between the journal and one or more compliant foil pads that comprise a top foil and an underlying support structure (typically in the form of a bump foil). However, such physical complexity, compared to conventional fluid bearings, means that the prediction of the static and dynamic performance of FAB-rotor systems is far more challenging, requiring due consideration of the nonlinear multi-domain dynamical system that comprises the coupling between the rotor system, the air films and the foil structures. Nonetheless, altering the profile of the clearance that contains the air film is known to improve performance in terms of delaying the onset of instability speed (OIS), suppressing sub-synchronous vibration, and increasing load capacity. This has sparked an interest in smart (active) FABs where such modification can be applied in real time. The use of traditional actuating mechanisms (e.g. PZT stacks with amplifying levers) increases design complexity and can potentially introduce unwanted resonant effects from such mechanisms. An alternative way to create a smart FAB is to exploit the compliance of the foil structure by bonding piezoelectric material to its top foil, thus creating a piezoelectric foil air bearing (PFAB). Apart from the ability to effect changes to the clearance profile when the transducer is operated in the actuator mode, the PFAB can alternatively be operated in the generator mode, providing electrical output that can be used for sensing (condition monitoring) or vibration energy harvesting purposes. The only proof of concept conducted so far on journal PFABs has involved the simulation of the effect of actuation on the load capacity of the PFAB for a fixed journal position. What is needed, and addressed in this thesis, is a computational simulation of a rotor system fitted with journal PFABs that can be configured in either the actuator or generator mode. Rotordynamic analysis involving PFABs requires a realistic foil model that considers the detachment of the top foil from the bump foil. This is necessary for a realistic simulation of the top foil deflection, given that this is coupled with the electrical domain. Aside from piezoelectric actuation, such detachment also happens in any FAB wherever the air film pressure is below atmosphere. Apart from single -pad bearing, a three-pad bearing needs to be considered since having three controllable top foil segments may offer greater opportunity of influencing the clearance profile (and hence the performance). Foil detachment models have so far not been used to simulate multi-pad bearings, which are naturally more affected by the lack of realistic representation of the top foil segments deflection than single-pad bearings. Due to the need of an advanced foil model and consideration of multi-pad bearings, the simulations required are computationally intensive, especially when seeking an optimal configuration for the piezo patches. Hence, a reduced order modelling (ROM) approach for the nonlinear and linearised analysis of rotor-FAB systems is essential for such a task. A ROM that is capable of accommodating top foil detachment and/or multi-pad bearings needs to be developed. The overall aim of this thesis is therefore to present a computational simulation of a rotor system fitted with PFABs that can be configured in either the actuator or generator mode for improved rotordynamic performance (through suppression of nonlinear effects) and/or added value (through electrical output for sensing or energy scavenging purposes). The novel contribution

Date of Award	1 Aug 2023
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Philip Bonello (Main Supervisor) & David Hall (Supervisor)