|M.Sc Student||Ferdinskoif Alexander|
|Subject||Laboratory Platform for Simulation, Validation and|
Design of High-Speed Rotors
|Department||Department of Mechanical Engineering||Supervisor||Professor Izhak Bucher|
|Full Thesis text|
In the conceptual design phase of a rotating machine, rotor parameters are determined according to the machine's operational requirements. Some of the rotor's requirements are critical for the desired operation conditions of the machine, such as absence of critical speeds in the rotation speed range and sustainable vibration levels due to imbalance. To withstand these requirements, Finite Element analyses are carried out at the preliminary phase of the design and design modifications are subsequently implemented.
Often, simulations and experimentally-obtained results disagreements are identified in an advanced stage of the development process. Sometimes, it does not affect the intended operation conditions, but occasionally, it may lead to critical faults, machine failure or safety hazards. At this point, design modifications are very costly and time consuming since the entire system is in its late development stages.
This research introduces a comprehensive simulation, validation and design-assisting laboratory testing environment. It aims to pinpoint the causes for gaps between the finite element model and experimental system measurements and provide a platform for practical validation of design modifications. As a part of the research, a laboratory simulator was designed and manufactured, allowing generating a wide range of dynamic loads while the system rotates at high rotation speed, to examine the difference between the anticipated and observed dynamical parameters. The test-system developed within this research allows controlling the supports' stiffness, the rotation speed and acceleration profile. It enables to vary the dynamic behavior of the rotor, isolate the system modes of vibration and identify the parameters causing the deviation from the finite element model. The controlled stiffness range allows reordering the appearance of rigid and flexible vibration modes as rotation speed is increased. As part of the experimental system, voice-coil actuators were added to enable open excitation and closed-loop damping of critical speeds. Testing can be carried out using several types of excitations, ranging from swept and constant sinusoidal forces, damping enhancement and parametric stiffness modulations. The effects of such excitation are investigated vs. the numerical model. Specific signal processing and model identification techniques are tailored to the experimental system, allowing it to isolate the relevant, speed and condition dependent, dynamical components. The research examines the causes for modeling imperfections exploiting the experimental facility, signal and force controlling system that is developed for this research.