Transient Response of a Short Length (L/D = 0.2) Open Ends Elastically Supported Squeeze Film Damper: Centered and Largely Off Centered Whirl Motions

Abstract

Commonly employed in air breathing (gas turbine) engines, squeeze film dampers (SFDs) reduce the amplitude of rotor vibration while traversing system critical speeds or in transient events such as during a maneuver load, a hard landing, a blade loss, or an engine startup/shutdown sequence that could instantaneously shift a damper journal eccentricity (es) to near its clearance (c). Experiments investigate the dynamic force performance of an open ends, shortlength (L/D = 0.2) SFD test rig with radial clearance c = 267 خ¼m and undergoing centered (es/c = 0) to largely offcentered (es/c → 1) whirl orbit motions induced by both a large static load plus a dynamic load. Four rods, symmetrically arranged to resemble a squirrel cage, elastically support the SFD test rig. A hydraulic load system displaces the test damper structure into static eccentricity (es/c). One of two types of dynamic load with amplitude FX = FY excite the SFD: a singlefrequency, stepping from low frequency to high frequency discretely; or a sinesweep frequency growing linearly with time at 6 Hz/s, 33 Hz/s, 40 Hz/s, or 55 Hz/s. For motions departing from es/c = 0.0, 0.95, and 0.99, the dynamic load uses a sinesweep frequency varying from 5 Hz to 245 Hz and evolving rapidly at ∼33 Hz/s. Measurements of SFD displacements characterize the behavior of the SFD rig during its transient response which crosses two system natural frequencies. For motions departing from a largely offcentered condition (es → c), the dynamic load forces the damper to whirl with highly elliptical orbits, in particular while crossing a resonance (damped natural frequency). Moreover, the dynamic motions departing from es ∼ c are smaller in amplitude than those arising from a centered condition (es/c = 0). The larger damping produced by a very small squeeze film thickness explains the difference in response amplitude. At a largely offcentered condition (es/c = 0.99) and a low excitation frequency (f < 40 Hz), intermittent contact between the damper journal and its housing occurs as evidenced by a large magnitude recorded dynamic pressure (on the order of MPa). For whirl motions around various static eccentricity positions, es/c = 0.0–0.75, the dynamic load covers a frequency range from 10 Hz to 100 Hz using either a singlefrequency excitation or a sinesweep frequency excitation with a slow growth rate ∼6.5 Hz/s to induce a quasisteadystate response. The experimental procedure builds complex stiffness in the frequency domain for identification of SFD stiffness, damping, and added mass force coefficients, (K, C, M)SFD. For motions centered around small to large static eccentricities, es/c = 0–0.75, the identified (K, C, M)SFD coefficients from sinesweep frequency dynamic loads coincide with those extracted from singlefrequency dynamic load tests over the same frequency range. Shortlength SFD theory predictions for damping coefficients agree with the experimental results. Predicted added mass or inertia coefficients, like the model, fall short of the target experimental magnitudes. The test results give practitioners the credence to certify the ability of a SFD to control rotor response amplitude during typical transient events.