Model and Experimental Verification of the Dynamic Forced Performance of a Tightly Sealed Squeeze Film Damper Supplied With a Bubbly Mixture

Abstract Practice and experiments with squeeze film dampers (SFDs) sealed with piston rings (PRs) show the lubricant exits through the PR slit, i.e. the gap made by the PR abutted ends when installed, forced as a jet during the portion of a rotor whirl cycle generating a positive squeeze film pressure. In the other portion of a whirl cycle, a sub ambient dynamic pressure ingests air into the film that mixes with the lubricant to produce a bubbly mixture. To reduce persistent air ingestion, commercial air breathing engines utilizing PRSFDs demand of a sufficiently large lubricant supply pressure (Ps), and hence a larger flow rate that is proportional to the journal squeeze velocity (vs = amplitude r × frequency of motion ω). The stringent requirement clearly limits the applicability and usefulness of SFDs. This paper presents a computational physics model for a sealed ends SFD operating with a mixture and delivers predictions benchmarked against profuse laboratory test data. The model implements a Reynolds equation adapted for a homogeneous bubbly mixture, includes temporal fluid inertia effects, and uses physics based inlet and outlet lubricant conditions through feed holes and PR slit, respectively. In the experiments for model validation, a SFD damper, 127 mm in diameter D, film land length L = 25.4 mm (L/D = 0.2), and radial clearance c = 0.371 mm, is supplied with an air in ISO VG2 oil bubbly mixture of known GVF, zero (pure oil) to 50% in steps of 10%. The mixture supply pressure varies from Ps = 2.06 bar-g (30 psig) to 6.20 bar-g (90 psig). Located in grooves at the top and bottom of the journal, a piston ring (PR) and an O-ring (OR) seal the film land. The OR does not allow any oil leakage or air ingestion; hence the supplied mixture discharges thru the PR slit into a vessel submerged within a large volume of lubricant. Dynamic load tests with a single frequency ω, varying from 10 Hz to 60 Hz, produce circular centered orbits with amplitude r = 0.2c. The measurements record the exerted forces and journal motions and an analysis delivers force coefficients, damping and inertia, representative of the exerted frequency range. The model predicts the pressure field and evolution of the gas volume fraction (GVF) within the film land and, in a simulated process replicating the experimental procedure, delivers representative force coefficients. For all Ps conditions, both predictions and tests show the SFD added mass coefficients significantly decrease as the inlet GVF (βs) increases. The experimentally derived damping coefficients do not show a significant change, except for tests with the largest concentration of air (βs = 0.5). The predicted damping differs by 10% with the test derived coefficient which does not readily decrease as the inlet GVF (βs) increases. The added mass coefficients, test and predicted, decrease with βs, both being impervious to the magnitude of supply pressure. The test PRSFD shows a quadrature stiffness due to the sliding friction between the PR being pushed against the journal. An increase in supply pressure exacerbates this unique stiffness that may impair the action of the squeeze film to dissipate mechanical energy. The comprehensive test results, first of their kind, demonstrate that accurate modeling of SFDs operating with air ingestion remains difficult as the flow process and the paths of its major components (air and liquid) are rather complex.