Experimental Force Coefficients for Two Sealed Ends Squeeze Film Dampers (Piston Rings and O-rings): An Assessment of Their Similarities and Differences
- Additional Document Info
- View All
© 2019 by ASME. Squeeze film dampers (SFDs) in aircraft engines effectively aid to reduce rotor motion amplitudes, in particular when traversing a critical speed, and help to alleviate rotor whirl instabilities. The current work is a long-term endeavor focused on quantifying the dynamic force performance of practical SFDs, exploring novel design damper configurations, and producing physically sound predictive SFD models validated by experimental data. Piston rings (PRs) and O-rings (ORs), commonly used as end seals in SFDs for commercial and military gas turbine engines, respectively, amplify viscous damping in a short physical length and while operating with a modicum of lubricant flow. This paper presents experimental force coefficients (damping and inertia) for two identical geometry SFDs with end seals, one configuration hosts PRs, and the other one ORs. The test rig comprises a stationary journal and bearing cartridge (BC) hosting the SFD and supported on four elastic rods to emulate a squirrel cage. The damper film land length, diameter, and clearance are L25.4 mm, D 5L, and c0.373mm (D/c340), respectively. A supply feeds ISO VG 2 oil to the film land at its middle plane through either one hole or three holes, 2.5mm in diameter, 120 deg apart. In the PRSFD, the lubricant exits through the slit opening at the ring butted ends. The ORs suppress oil leakage; hence, lubricant evacuates through a 1mm hole at 1=4 L near one journal end. The ORs when installed add significant stiffness and damping to the test structure. The ORSFD produces 20% more damping than the PRSFD, whereas both sealed ends SFDs show similar size added mass. For oil supplied at 0.69 bar(g) through a single orifice produces larger damping, 60-80% more than when the damper operates with three oil feedholes. A computational model reproducing the test conditions delivers force coefficients in agreement with the test data. Archival literature calls for measurement of a single pressure signal to estimate SFD reaction forces. For circular centered orbits (CCOs), the dynamic pressure field, in the absence of any geometrical asymmetry or feed/discharge oil condition, rotates around the bearing with a speed equal to the whirl frequency. The paper presents force coefficients estimated from (a) measurements of the applied forces and ensuing displacements, and (b) the dynamic pressure recorded at a fixed angular location and integrated over the journal surface. The first method delivers a damping coefficient that is large even with lubricant supplied at a low oil supply pressure whereas the inertia coefficient increases steadily with feed pressure. Predictions show good agreement with the test results from measured forces and displacements, in particular the added mass. On the other hand, identified damping and inertia coefficients from dynamic pressures show a marked difference from one pressure sensor to another, and vastly disagreeing with test results from the first method or predictions. The rationale for the discrepancy relies on local distortions in the dynamic pressure fields that show zones of oil vapor cavitation at a near zero absolute pressure and/or with air ingestion producing high frequency spikes from bubble collapsing; both phenomena depend on the magnitude of the oil supply pressure. An increase in lubricant supply pressure suppresses both oil vapor cavitation and air ingestion, which produces an increase of both damping and inertia force coefficients. No prior art compares the performance of a PRSFD vis-a-vis that of an ORSFD. Supplying lubricant with a large enough pressure (flow rate) is crucial to avoid the pervasiveness of air ingestion. Last, the discussion on force coefficients obtained from two distinct methods questions the use of an oversimplifying assumption; the dynamic pressure field is not invariant in a rotating coordinate frame.
author list (cited authors)
San Andres, L., Koo, B., & Jeung, S.