Adolfo Delgado

Identification and Prediction of Force Coefficients in a Five-Pad and Four-Pad Tilting Pad Bearing for Load-on-Pad and Load-Between-Pad Configurations

This paper presents the identification of the rotordynamic force coefficients for direct lubrication five-pad and four-pad tilting pad bearings. The bearing is 110 mm in diameter with a L/D of 0.4 pad axial length (44 mm). The experiments include load-on-pad and load-between-pad configurations, with 0.5 and 0.6 pivot offsets, for rotor speeds ranging from 7500 rpm to 15,000 rpm. The bearing force coefficients are identified from multiple frequency excitations (20–300 Hz) exerted on the bearing housing by a pair of hydraulic shakers and are presented as a function of the excitation frequency and rotor speed for a 300 kPa unit load. The experimental results also include temperatures at the trailing edge of three pads. The direct force coefficients, identified from curve-fits of the complex dynamic stiffness, are frequency independent if considering an added mass term much smaller than the test device modal mass. The force coefficients from the four-pad bearing load-between-pad configuration show similar coefficients in the loaded and orthogonal directions. On the other hand, as expected, the five-pad bearing load-on-pad shows larger coefficients (∼25%) in the loaded direction. The maximum pad temperature recorded for the 0.5 pivot offset configurations is up to 20°C higher than those associated to the 0.6 offset configuration. Results from a predictive code are within 50% of the experimental results for the direct stiffness coefficients and within 30% for the direct damping coefficients.

Identification of Force Coefficients in a Squeeze Film Damper With a Mechanical End Seal—Centered Circular Orbit Tests

The damping capability of squeeze film dampers (SFDs) relies on adequate end sealing to prevent air ingestion and entrapment. The paper presents the parameter identification, procedure and damping coefficients, of a test SFD featuring a mechanical seal that effectively eliminates lubricant side leakage. The test damper reproduces an aircraft application intended to contain the lubricant in the film lands for extended periods of time. The test damper journal is 2.54 cm in length and 12.7 cm in diameter, with a nominal clearance of 0.127 mm. The SFD feed end is flooded with oil, while the discharge end contains a recirculation groove and four orifice ports. In a companion paper (ASME GT2006-90782), single frequency - unidirectional load excitation tests were conducted, without and with lubricant in the squeeze film lands, to determine the seal dry-friction force and viscous damping force coefficients. Presently, tests with single frequency excitation loads rendering circular centered orbits excitations are conducted to identify the SFD force coefficients. The identified parameters include the overall system damping and the individual contributions from the squeeze film, dry friction and structural damping. The identified system damping coefficients are frequency and motion amplitude dependent due to the dry friction interaction at the mechanical seal interface. Identified squeeze film force coefficients, damping and added mass, are in good agreement with predictions based on the full film, short length damper model.Copyright

Rotordynamic Force Coefficients for Three Types of Annular Gas Seals With Inlet Preswirl and High Differential Pressure Ratio

The following paper presents and compares rotordynamic force coefficients for three types of non-contact annular gas seals, which include a labyrinth (LABY), honeycomb (HC), and a fully partitioned damper seal (FPDS). These three annular seals represent the typical seal types used in process gas centrifugal compressors at the balance piston location or center seal location to limit internal leakage and ensure a robust rotordynamic design. Tests were conducted on 170.6 mm (6.716 in) diameter seals for rotor speeds up to 15 kprm, inlet air pressure of 6.9 bar (100 psi), ambient back pressure, and with inlet gas preswirl. The three seals were designed to have the same nominal clearance and similar axial lengths. Testing was conducted on a controlled motion test rig possessing non-synchronous excitation capability up to 250 Hz. Three different test methods were employed to give confidence in the rotordynamic coefficients, which include static force deflection tests, mechanical impedance tests, and dynamic cavity pressure tests. Results from experiments compare force coefficients for all seal configurations while paying special attention to the crossover frequencies of the effective damping term. All seals possessed negative effective damping at lower excitation frequencies with inlet preswirl, where the straight-through FPDS possessed the lowest cross over frequency of 40 Hz at 15 krpm. The testing also revealed that the preswirl parameter had significantly more influence on effective damping levels and crossover frequencies when compared to rotor speed.

Analysis of Performance and Rotordynamic Force Coefficients of Brush Seals With Reverse Rotation Ability

Multiple-shoed brush seals represent an alternative to resolve poor reliability resulting from bristle tip wear while also allowing for reverse rotation operation. The novel configuration incorporates pads contacting the shaft, and which under rotor spinning; lift off due to the generation of hydrodynamic pressures. The ensuing gas film prevents intermittent contact; thus lowering the operating temperature and thermal distortions, and even eliminating bristles’ wear. A computational analysis for the equilibrium and dynamic forced response of a brush seal with reverse rotation capability is presented. Small amplitude rotor motions about an equilibrium position lead to a nonlinear partial differential equation for the static pressure field, and a set of first order linear partial differential equations to determine the rotordynamic force coefficients, stiffness and damping, as function of the excitation frequency and other operating conditions. Predictions for the stiffness and damping coefficients of a 20 shoe-brush seal configuration operating over a range of rotor speeds are detailed. The parametric study varies the nominal gas film thickness, the supply to discharge pressure ratio, and the bristle bed structural loss (damping) coefficient. The results show that the film clearance and supply to discharge pressure ratio do not affect the shoed-brush seal force coefficients. On the other hand, the direct stiffness drops rapidly as the operating speed increases. The shoed-brush seal offers whirl frequency ratios much lower than 0.50 due to the (structural) damping arising from friction among the brush seal bristles.Copyright

A Model for Improved Prediction of Force Coefficients in Grooved Squeeze Film Dampers and Oil Seal Rings

Squeeze film damper (SFD) designs typically implement supply grooves to ensure adequate lubricant flow into the film lands. Oil seal rings, of land film clearance c, also incorporate short and shallow grooves (length ≤ 30c, depth ≤ 15c) to reduce cross-coupled stiffnesses, thus promoting dynamic stability without a penalty in increased leakage. However, extensive experimental results in the archival literature demonstrate that grooves do not reduce the force coefficients as much as theory predicts. A common assumption is that deep grooves do not influence a damper or oil seal ring forced response. However, unexpected large added mass coefficients, not adequately predicted, appear to be common in many tested SFD and oil seal configurations. In the case of oil seals, experiments demonstrate that circumferential grooves do reduce cross-coupled stiffnesses but to a lesser extent than predictions would otherwise indicate. A linear fluid inertia bulk-flow model for analysis of the forced response of SFDs and oil seal configurations with multiple grooves is advanced. A perturbation analysis for small amplitude journal motions about a centered position yields zeroth and first-order flow equations at each flow region (lands and grooves). At a groove region, a groove effective depth d η , differing from its actual physical value, is derived from qualitative observations of the laminar flow pattern through annular cavities. The boundary conditions at the inlet and exit planes depend on the actual seal or SFD configuration. Integration of the resulting first-order pressure fields on the journal surface yields the force coefficients (stiffness, damping, and inertia). Current model predictions are in excellent agreement with published test force coefficients for a grooved SFD and a grooved oil seal. The results confirm that large added mass coefficients arise from the flow interactions between the feed/discharge grooves and film lands in the test elements. Furthermore, the predictions, benchmarking experimental data, corroborate that short length inner-land grooves in an oil seal do not isolate the pressure fields of adjacent film lands and hence contribute greatly to the forced response of the mechanical element.

Measurements of Leakage and Power Loss in a Hybrid Brush Seal

Simplicity, low cost, and easy replacement make labyrinth seals the primary seal type in gas turbines. However, excessive leakage and potential for rotordynamic instability are well known issues. Brush seals effectively control leakage in air breathing engines, albeit only applied for relatively low pressure differentials. Hybrid brush seals (HBSs) are an alternative to resolve poor reliability resulting from bristle tip wear while also allowing for reverse rotation operation. The novel configuration incorporates pads contacting the shaft, which under rotor spinning lifts off due to the generation of a hydrodynamic pressure. The ensuing gas film prevents intermittent contact, thus lowering the operating temperature and thermal distortions and even eliminating bristle wear. The hybrid brush seal improves sealing, is more durable and reliable than conventional brush seals, and allows reverse shaft rotation without seal damage. This paper presents measurements of power loss and leakage in a HBS for increasing pressure differentials over a range of rotor speeds. The test HBS, Haynes-25 bristle pack (∼850 bristles/cm) and 45 deg lay angle, is 166.4 mm in diameter and integrates 20-arcuate pads connected with thin electrical-discharge machined webs (EDM-webs) to the seal casing. The webs are designed with low radial stiffness to allow for rotor excursions and high axial stiffness to avoid pad pitching motions resulting from high pressure differentials across the seal. Measured drag power at low rotor speeds (<11 m/s at 1300 rpm) decreases as the pressure differential across the seal increases. At a fixed rotor speed, a significant drop in drag torque (and drag power) ensues as the supply pressure increases, thus demonstrating that a gas film separates the rotor from the seal pads. Additionally, the operating temperature measured at the rotor/seal interface remains approximately constant (∼24°C) during tests with shaft rotation (power loss and drag torque measurements) under pressurized conditions, indicating that the rotor and seal pads are not in contact. Flow rate measurements at room temperature (25°C) show an improved sealing ability with a leakage reduction of about 36% when compared with a first generation shoed-brush seal. The HBS calculated effective clearance (∼50 pm) is approximately 70% smaller than the radial clearance (-180 μm) of an ideal noncontacting seal with similar rotor diameter. Improved brush seal technology will increase the efficiency of gas turbines while also aiding to improve the engine stability and to reduce vibrations.

Identification of Force Coefficients in a Squeeze Film Damper With a Mechanical End Seal—Part I: Unidirectional Load Tests

Squeeze film dampers (SFDs) with low levels of external pressurization and poor end sealing are prone to air entrapment, thus not generating enough damping capability. Single frequency, unidirectional load tests were conducted on a SFD test rig replicating a commercial jet-engine configuration. The damper journal is 2.54cm in length and 12.7cm in diameter, with nominal clearance of 0.127mm. The SFD feed end is flooded with oil, while the discharge end contains a recirculation groove and four orifice ports, and a mechanical seal ring in contact with the damper journal. A wave spring pushes the ring ensuring tight sealing to prevent gas ingestion. The mechanical seal also serves to contain the lubricant within the squeeze film land for extended periods of time and; while in operation, to prevent contamination of the ball bearing cartridge. The measurements conducted without and with lubricant in the squeeze film lands, along with a frequency domain identification procedure, render the mechanical seal dry-friction force and viscous damping force coefficients as functions of frequency and motion amplitude. The end seal arrangement is quite effective in eliminating side leakage and preventing air entrainment into the film lands. Importantly enough, the dry friction force, arising from the contact forces in relative motion, increases significantly the test element equivalent viscous damping coefficients. The identified system damping coefficients are thus frequency and motion amplitude dependent, albeit decreasing rapidly as the motion parameters increase. Identified squeeze film force coefficients, damping and added mass, agree well with predictions based on the full film, short length damper model.

Identification of Force Coefficients in a Squeeze Film Damper With a Mechanical Seal: Large Contact Force

Squeeze film dampers (SFDs) aid to reduce excessive vibration levels due to rotor imbalance and to raise stability thresholds in rotor-bearing systems. SFDs commonly include end seals to increase their damping capability with a lesser lubricant flow. Seals also aid to reduce the occurrence of air ingestion/entrapment that severely reduces the damper forced performance. However, most conventional end seals do not completely eliminate lubricant side leakage, which limits their effectiveness to prevent air ingestion. A novel end seal arrangement incorporates a spring loaded, contacting mechanical seal that effectively prevents lubricant side leakage and air ingestion. The mechanically sealed damper is intended for use in power engines for unmanned aircraft vehicles. The test damper journal is 2.54 cm in length and 12.7 cm in diameter, with a radial clearance of 0.127 mm. Prior literature reports dynamic load tests on the seal-SFD and measurements of orbital motions to characterize the mechanical parameters of both the mechanical seal and squeeze film damper section. The test data to date include damper operation for a single contact load (90 N) closing the mechanical seal. Presently, measurements of damper dynamic load performance are conducted with a larger contact force (260 N). A nonlinear parameter identification method in the frequency domain determines simultaneously the squeeze film damping and inertia coefficients and the seal dry-friction force. The test results show that the system equivalent viscous damping coefficients are twice as large as those obtained earlier with the smaller contact force. On the other hand, as expected, the squeeze film damper coefficients are nearly identical for both test configurations. Predicted squeeze film damping coefficients, from an improved model that includes the flow in the damper feed and discharge grooves, correlate well with the test data for small and moderate orbit radii. The experimental fluid added mass coefficients are in par with the actual mass of the bearing housing and accurately predicted.

Rotordynamic Force Coefficients of a Hybrid Brush Seal: Measurements and Predictions

Brush seals effectively control leakage in air breathing engines, albeit only applied for relatively low-pressure differentials. Hybrid brush seals (HBS) are an alternative to resolve poor reliability resulting from bristle tip wear while also allowing for reverse shaft rotation operations. A HBS incorporates pads contacting the shaft on assembly; and which under rotor spinning, lift off due to the generation of a hydrodynamic pressure. The ensuing gas film prevents intermittent contact, reducing wear, and thermal distortions. This paper presents rotordynamic measurements conducted on a test rig for evaluation of HBS technology. Single frequency shaker loads are exerted on a test rotor holding a hybrid brush seal, and measurements of rotor displacements follow for operating conditions with increasing gas supply pressures and two rotor speeds. A frequency domain identification method delivers the test system stiffness and damping coefficients. The HBS stiffness coefficients are not affected by rotor speed though the seal viscous damping shows a strong frequency dependency. The identified HBS direct stiffness decreases ∼15% as the supply/discharge pressure increases P r = 1.7-2.4. The HBS cross-coupled stiffnesses are insignificant, at least one order of magnitude smaller than the direct stiffnesses. A structural loss factor (γ) and dry-friction coefficient (μ) represent the energy dissipated in a HBS by the bristle-to-bristle and bristle-to-pad interactions. Predictions of HBS stiffness and damping coefficients correlate well with the test derived parameters. Both model predictions and test results show the dramatic reduction in the seal equivalent viscous damping coefficients as the excitation whirl frequency increases.

A Novel Bulk-Flow Model for Improved Predictions of Force Coefficients in Grooved Oil Seals Operating Eccentrically

Oil seals in centrifugal compressors reduce leakage of the process gas into the support bearings and ambient. Under certain operating conditions of speed and pressure, oil seals lock, becoming a source of hydrodynamic instability due to excessively large cross coupled stiffness coefficients. It is a common practice to machine circumferential grooves, breaking the seal land, to isolate shear flow induced film pressures in contiguous lands, and hence reducing the seal cross coupled stiffnesses. Published tests results for oil seal rings shows that an inner land groove, shallow or deep, does not actually reduce the cross-stiffnesses as much as conventional models predict. In addition, the tested grooved oil seals evidenced large added mass coefficients; while predictive models, based on classical lubrication theory, neglect fluid inertia effects. This paper introduces a bulk-flow model for groove oil seals operating eccentrically and its solution via the finite element method. The analysis relies on an effective groove depth, different from the physical depth, which delimits the upper boundary for the squeeze film flow. Predictions of rotordynamic force coefficients are compared to published experimental force coefficients for a smooth land seal and a seal with a single inner groove with depth equaling 15 times the land clearance. The test data represent operation at 10 krpm and 70 bar supply pressure, and four journal eccentricity ratios (e/c = 0, 0.3, 0.5, 0.7). Predictions from the current model agree with the test data for operation at the lowest eccentricities (e/c = 0.3); discrepancies increasing at larger journal eccentricities. The new flow model is a significant improvement towards the accurate estimation of grooved seal cross-coupled stiffnesses and added mass coefficients; the later previously ignored or largely under predicted.Copyright