Force coefficients for a large clearance open ends squeeze film damper with a central feed groove: Experiments and predictions

Highlights

• Comprehensive measurements of dynamic pressures in groove and film lands of a squeeze film damper.

• Identifies force coefficients from dynamic loads and bearing displacements for a large range of whirl frequencies.

• Correlates identified SFD parameters against classical model predictions and points out to the formulas limited value.

• Validates predictions from a flow model that incorporates flow interactions between a groove and film lands.

Abstract

The paper describes a large load squeeze film damper (SFD) test rig, details measurements of dynamic loads inducing circular orbits conducted on a large clearance (c=0.250 mm) open ends centrally grooved SFD, and presents the identified experimental SFD force coefficients for operation at three static eccentricities. The rig has a bearing cartridge supported atop four elastic rods and a stationary journal, 0.127 mm in diameter. The damper consists of two parallel film lands, 12.7 mm in length, separated by a central groove, 6.35 mm 9.5 mm in depth. In the journal, three equally spaced holes, 120° apart, supply a light lubricant into the central groove and squeeze film lands. The experimental SFD force coefficients are compared to test results obtained earlier for a damper with the same film land lengths but with a smaller clearance (c=0.140 mm) and against predictions obtained from an advanced physical model that accounts for the flow field in the central groove and the interaction with the adjacent film lands. Dynamic pressures in the film lands and in the central groove are (not) surprisingly of the same order of magnitude. The central groove affects the dynamic forced response of the test damper to generate large direct damping coefficients, ~3.5 times those derived from classical lubrication formulas. Experimental added mass coefficients are ~7.4 times the predictive classical values. Predictions from an advanced model correlate well with the test data when using a shallow groove depth. The measurements and analysis advance knowledge on the dynamic forced performance of SFDs, point out to the limited value of simplistic predictive formulas, and validate the accuracy of a modern predictive tool.

Introduction

Efficient multiple-stage turbomachinery implements tighter clearances in secondary flow passages, uses advanced materials; and implementing flexible rotors, demands of an increased tolerance to disk imbalances [1]. Operation at high shaft speeds induces severe dynamic loading with large amplitude journal motions at the bearing supports. In aircraft engines, squeeze film dampers (SFDs) are a proven mean to add viscous damping that aids to reduce rotor dynamic displacements and also serve to isolate the rotor(s) from the engine frame [2].

SFDs have been designed for industrial usage for well over half a century. Hence, the literature on SFDS is not scant. For example Refs. [1], [2] and the review papers [3], [4] detail the governing equations for generation of the hydrodynamic pressure field and fluid film reaction forces and the effect of SFDs on the rotordynamics of simple systems, rigid and flexible. These references also review relevant experimental data and establish there is modest agreement with predictions from classical lubrication theory, hence concluding the forced performance of SFDs is complicated, i.e., hard to predict, due to a number of effects (usually) not considered in the analysis and design processes. Some of these effects include fluid inertia—temporal and advective, in particular for configurations with circumferential oil supply grooves [5], [6], and the persistence of air ingestion and entrapment in dampers with open ends and not supplied copiously with enough flow to avoid the formation of bubbly mixtures [2]. More recently, Gehannin and Arghir [7] Delgado and San Andrés [8] extend the review of prior art to the present day, advance bulk-flow models to address realistically various issues of SFD performance, and provide extensive validations against measured squeeze film pressures and force coefficients.

In the author's opinion, disconnect between academic research and engineering practice persists due to the widespread belief that a squeeze film damper is just a plain journal bearing that does not spin. To date the vast amount of literature on the nonlinear dynamics of rotors supported on (short length journal) SFDs is suspicious due to the lack of supporting experimental evidence. That is, the implementation of SFDs in actual compressor and gas turbine applications shows them to be linear elements, as opposed to highly nonlinear elements that can induce jump phenomenon and multiple valued responses [2], [9], [10].

This paper presents measurements of true rotordynamic force coefficients¹ for an open ends SFD comprised of two parallel film lands separated by a deep central groove. The radial clearance of the test element is c=0.250 mm. An aircraft main engine application dictates the damper dimensions, lubricant properties and operating conditions. The contents of this paper complement prior work reported in the theses of Seshagiri [11] and Mahecha [12], both works summed in the archival literature as Refs. [13], [14]. The earlier work was conducted with the same mechanical element but with a smaller film radial clearance, c_s=0.140 mm. Details on the test rig, measurement procedure, and experimental results follow shortly.

As in a much earlier work [15], [16], the present experimental data shows the circumferential groove is not impervious to journal motions, hence generating significant dynamic film pressures that contribute greatly to the overall forced response of the test element. To this date it is customary to regard a deep and wide groove as an ideal source (or sink) of fluid flow, not affected by the journal kinetics, and hence decoupled to the dynamics of the adjacent squeeze film lands. The poor practice ignores a vast amount of experimental evidence, Refs. [11], [12], [13], [14], [15], [16] to name a few. Most importantly, as the reader will soon realize, predictions of the force coefficients for the test SFD using classical oversimplified formulas is futile. An advanced physical model offers a good correlation with the experimental force coefficients, albeit it uses an effective groove depth that remains as an empirical parameter.