| Summary: | With increasing rotor flexibility and shaft speeds, turbomachinery undergoes large dynamic loads and displacements. Squeeze film dampers (SFDs) are a type of fluid film bearing used in rotating machinery to attenuate rotor vibration, provide mechanical isolation, and/or to tune the placement of system critical speeds. Industry has a keen interest in designing SFDs that are small, lightweight, and mechanically simple. To achieve this, one must have a full understanding of how various design features affect the SFD forced performance.
This thesis presents a comprehensive analysis, experimental and theoretical, of a short (L=25.4 mm) open ends SFD design incorporating three lubricant feed holes (without a circumferential feed groove). The damper radial clearance (c=127 μm), L/D ratio (0.2), and lubricant (ISO VG2) have similar dimensions and properties as in actual SFDs for aircraft engine applications. The work presents the identification of experimental force coefficients (K, C, M) from a 2-DOF system model for circular and elliptical orbit tests over the frequency range ω=10-250Hz. The whirl amplitudes range from r=0.05c-0.6c, while the static eccentricity ranges from eS=0-0.5c. Analysis of the measured film land pressures evidence that the deep end grooves
(provisions for installation of end seals) contribute to the generation of dynamic pressures in an almost purely inertial fashion. Film land dynamic pressures show both viscous and inertial effects. Experimental pressure traces show the occurrence of significant air ingestion for orbits with amplitudes r>0.4c, and lubricant vapor cavitation when pressures drop to the lubricant saturation pressure (PSAT~0 bar). Identified force coefficients show the damper configuration offers direct damping
coefficients that are more sensitive to increases in static eccentricity (eS) than to increases in amplitude of whirl (r). On the other hand, SFD inertia coefficients are more sensitive to increases in the amplitude of whirl than to increases in static eccentricity. For small amplitude motions, the added or virtual mass of the damper is as large as 27% of the bearing cartridge mass (MBC=15.15 kg). The identified force coefficients are shown to be insensitive to the orbit type (circular or elliptical) and the number of open feed holes (3, 2, or 1).
Comparisons of damping coefficients between a damper employing a circumferential feed groove1 and the current damper employing feed holes (no groove), show that both dampers offer similar damping coefficients, irrespective of the orbit amplitude or static eccentricity. On the other hand, the grooved damper shows much larger inertia force coefficients, at least ~60% more.
Predictions from a physics based model agree well with the experimental damping coefficients, however for large orbit motion, over predict inertia coefficients due to the model neglecting convective inertia effects.
Credence is given to the validity of the linearized force coefficients by comparing the actual dissipated energy to the estimated dissipated energy derived from the identified force coefficients. The percent difference is below 25% for all test conditions, and in fact is shown to be less than 5% for certain combinations of orbit amplitude (r), static eccentricity (eS), and whirl frequency (ω).
|
|---|
