Gas foil bearings (GFBs) find widespread usage in oil-free turbo expanders, APUs and micro gas turbines for distributed power due to their low drag friction and ability to tolerate high level vibrations, including transient rubs and shaft misalignment, static and dynamic. The static load capacity and dynamic forced performance of GFBs depends largely on the material properties of the support elastic structure, i.e. a smooth foil on top of bump strips. Conventional models include only the bumps as an equivalent stiffness uniformly distributed around the bearing circumference. More complex models couple directly the elastic deformations of the top foil to the bump underlying structure as well as to the hydrodynamics of the gas film. This paper details two FE models for the top foil supported on bump strips, one considers a 2D shell anisotropic structure and the other a 1D beam-like structure. The Cholesky decomposition of the stiffness matrix representing the top foil and bump strips is performed off-line prior to computations coupling it to the gas film analysis governed by Reynolds equation. The procedure greatly enhances the computational efficiency of the numerical scheme. Predictions of journal attitude angle and minimum film thickness for increasing static loads and two journal speeds are obtained for a GFB tested decades ago. 2D FE model predictions overestimate the minimum film thickness at the bearing centerline, while underestimating it at the bearing edges. Predictions from the 1D FE model compare best to the limited tests data; reproducing closely the experimental circumferential wavy-like minimum film thickness profile. The 1D top foil model is recommended due to its low computational cost. Predicted stiffness and damping coefficients versus excitation frequency show that the two FE top foil structural models result in slightly lower direct stiffness and damping coefficients than those from the simple elastic foundation model.

High performance oil-free turbomachinery implements gas foil bearings (FBs) to improve mechanical efficiency in compact units. FB design, however, is still largely empirical due to their mechanical complexity. The paper provides test results for the structural parameters in a bump-type foil bearing. The stiffness and damping (Coulomb or viscous type) coefficients characterize the bearing compliant structure. The test bearing, 38.1 mm in diameter and length, consists of a thin top foil supported on bump-foil strips. A prior investigation identified the stiffness due to static loads. Presently, the test FB is mounted on a non-rotating stiff shaft and a shaker exerts single frequency loads on the bearing. The dynamic tests are conducted at shaft surface temperatures from 25 °C to 75°C. Time and frequency domain methods are implemented to determine the FB parameters from the recorded periodic load and bearing motions. Both methods deliver identical parameters. The dry friction coefficient ranges from 0.05 to 0.20, increasing as the amplitude of load increases. The recorded motions evidence a resonance at the system natural frequency, i.e. null damping. The test derived equivalent viscous damping is inversely proportional to the motion amplitude and excitation frequency. The characteristic stick-slip of dry friction is dominant at small amplitude dynamic loads leading to a hardening effect (stiffening) of the FB structure. The operating temperature produces shaft growth generating a bearing preload. However, the temperature does not affect significantly the identified FB parameters, albeit the experimental range was too small considering the bearings intended use in industry.

Gas foil bearings (GFBs) satisfy the requirements for oil-free turbomachinery, i.e. simple construction and ensuring low drag friction and reliable high speed operation. However, GFBs have a limited load capacity and minimal damping, as well as frequency and amplitude dependent stiffness and damping characteristics. This paper provides experimental results of the rotordynamic performance of a small rotor supported on two bump-type GFBs of length and diameter equal to 38.10 mm. Coast down rotor responses from 25 krpm to rest are recorded for various imbalance conditions and increasing air feed pressures. The peak amplitudes of rotor synchronous motion at the system critical speed are not proportional to the imbalance introduced. Furthermore, for the largest imbalance, the test system shows subsynchronous motions from 20.5 krpm to 15 krpm with a whirl frequency at ∼ 50% of shaft speed. Rotor imbalance exacerbates the severity of subsynchronous motions, thus denoting a forced nonlinearity in the GFBs. The rotor dynamic analysis with calculated GFB force coefficients predicts a critical speed at 8.5 krpm, as in the experiments; and importantly enough, unstable operation in the same speed range as the test results for the largest imbalance. Predicted imbalance responses do not agree with the rotor measurements while crossing the critical speed, except for the lowest imbalance case. Gas pressurization through the bearings’ side ameliorates rotor subsynchronous motions and reduces the peak amplitudes at the critical speed. Post-test inspection reveal wear spots on the top foils and rotor surface.

Reliable gas bearings will enable the rapid deployment of high speed oil-free micro-turbomachinery. This paper presents analysis and experiments of the dynamic performance of a small rotor supported on Rayleigh step gas bearings. Comprehensive tests demonstrate that Rayleigh step hybrid gas bearings exhibit adequate stiffness and damping capability in a narrow range of shaft speeds, up to ∼ 20 krpm. Rotor coast down responses were performed with two test bearing sets with nominal radial clearance of 25 μm and 38 μm. A near-frictionless carbon (NFC) coating was applied on the rotor to reduce friction at liftoff and touchdown. However, the rotor could not lift easily and severe rubbing occurred at shaft speeds below ∼ 4,000 rpm. The tests show that the supply pressure raises the rotor critical speed and decreases the system damping ratio, while only affecting slightly the rotor-bearing system onset speed of instability. Whirl frequencies are nearly fixed at the system natural frequency (∼ 120 Hz) with subsynchronous amplitude motions of very large magnitude that prevented rotor operation above ∼ 20 krpm. The geometry of the Rayleigh steps distributed on the rotor surface generates a time varying pressure field, resulting in a sizable 4X super synchronous component of bearing transmitted load. Predictions show the synchronous stiffness and damping coefficients decrease with shaft speed. Predicted threshold speeds of instability are much lower than measured values due to the analytical model limitations assuming a grooved stator. The predicted synchronous responses to imbalance correlate well with the measurements. The Rayleigh step gas bearings are the most unreliable rigid bearing configuration tested to date.

The multiple-shoe brush seal, a variation of a standard brush seal, accommodates arcuate pads at the bristles free ends. This novel design allows reverse shaft rotation operation, and reduces and even eliminates bristle wear, since the pads lift off due to the generation of a hydrodynamic film during rotor spinning. This type of seal, able to work at both cold and high temperatures, not only restricts secondary leakage but also acts as an effective vibration damper. The dynamic operation of the shoed-brush seals, along with the validation of reliable predictive tools, relies on the appropriate estimation of the seal structural stiffness and energy dissipation features. Single frequency external load tests conducted on a controlled motion test rig and without shaft rotation allow the identification of the structural stiffness and equivalent damping of a 20-pad brush seal, 153 mm in diameter. The seal energy dissipation mechanism, represented by a structural loss factor and a dry friction coefficient, characterizes the energy dissipated by the bristles and the dry friction interaction of the brush seal bristles rubbing against each other. The physical model used reproduces well the measured system motions, even for frequencies well above the identification range.