ASME TURBO EXPO 2010 Glasgow Scotland UK Dynamic

ASME TURBO EXPO 2010, Glasgow, Scotland, UK Dynamic Response of a Rotor. Hybrid Gas Bearing System due to Base Induced Periodic Motions Luis San Andrés Mast-Childs Professor Fellow ASME Keun Ryu Research Assistant Yaying Niu Research Assistant TURBOMACHINERY LABORATORY TEXAS A&M UNIVERSITY ASME paper GT 2010 -22277 Supported by TAMU Turbomachinery Research Consortium

Microturbomachinery (< 250 k. W) Turbo Compressor 100 krpm, 10 k. W Advantages • Compact and fewer parts • Portable • High energy density • Lower emissions • Low operation/maintenance costs http: //www. hsturbo. de/en/produkte/turboverdichter. html Micro Turbo 500 krpm, 0. 1~0. 5 k. W http: //www. hsturbo. de/en/produkte/micro-turbo. html Oil-free turbocharger 120 krpm, 110 k. W http: //www. miti. cc/new_products. html

Gas bearings for microturbomachinery Metal Mesh Foil Bearing Advantages • Little friction and power losses • Simple configuration • High rotor speeds (DN value>4 M) • Operate at extreme temperatures Issues GT 2009 -59315 Gas Foil Bearing AIAA-2004 -5720 -984 • Small damping • Low static load capacity • Prone to instability Flexure Pivot Bearing GT 2004 -53621

Gas Bearing Research at TAMU 2001/2 - Three Lobe Bearings Stability depends on feed pressure. Stable to 80 krpm with 5 bar pressure 2003/4 - Rayleigh Step Bearings Worst performance to date with grooved bearings 2002 -09 - Flexure Pivot Tilting Pad Bearings Stable to 93 krpm w/o feed pressure. Operation to 100 krpm w/o problems. Easy to install and align. 2004 -10: Bump-type Foil Bearings Industry standard. Reliable but costly. Models anchored to test data. 2008 -10: Metal Mesh Foil Bearings Cheap technology. Still infant. Users needed

Objective & tasks Evaluate the reliability of rotor-air bearing systems to withstanding periodic base or foundation excitations • Set up an electromagnetic shaker under the base of test rig to deliver periodic load excitations • Measure the rig acceleration and rotordynamic responses due to shaker induced excitations • Model the rotor-air bearing system subject to base motions and compare predictions to test results

Gas Bearing Test Rig Rotor/motor Load cell Sensors Bearing Positioning Bolt Thrust pin Air supply 190 mm, 29 mm diam Rotor: 826 grams Bearings: L= 30 mm, D=29 mm LOP

Rotor and hybrid gas bearings Rotor • 0. 826 kg, 190 mm in length • Location of sensors and bearings noted Flexure Pivot Hybrid Bearings: Improved stability, no pivot wear Clearance ~42 mm, preload ~40%. Web rotational stiffness = 62 Nm/rad. Test rig tilted by 10°.

Previous work (GT 2009 -59199) Intermittent base shock load excitations Drop induced shocks ~30 g. Full recovery within ~ 0. 1 sec. Ps=2. 36 bar (ab) Rotor motion amplitudes increase with excitation of system natural frequency. NOT a rotordynamic instability!

Gas bearing test rig Base excitation Shaker & rod push base of test rig Front and side views (not to scale)

Hybrid gas bearing test rig Rod pushes base plate! (no rigid connection)

Waterfalls in coast down No base excitation Ps = 2. 36 bar Subsynchronous whirl > 30 krpm, fixed at system natural frequency = 193 Hz

Rotor speed coast down tests (35 krpm) No base excitation 1 X response Pressure 2. 36 bar 3. 72 bar 5. 08 bar Natural Freq 192 Hz 217 Hz 250 Hz Feed pressure increases natural frequency and lowers damping ratio

Acceleration (g) Natural frequency whole test rig (5 Hz) Soft mounts (coils) produce low natural frequency

Acceleration (g) Delivered excitations (6 Hz) Rotor speed: 34 krpm (567 Hz) Due to electric motor zoom Shaker transfers impacts to rig base! Super harmonic frequencies excited

Waterfalls in coast down Shaker frequency: 12 Hz Ps = 2. 36 bar (ab)

Rotor speed coast down Shaker frequency: 12 Hz Ps = 2. 36 bar (ab) Subsynchronous frequencies: 1) 24 Hz (2 x 12 Hz) 2) Natural frequency 193 Hz Synchronous motion dominates! Excitation of system natural frequency does NOT mean instability!

Effect of feed pressure Ps: 2. 36, 3. 72 & 5. 08 bar Shaker frequency: 12 Hz Rotor speed: 34 krpm (567 Hz) 12 Hz, 24 Hz, 36 hz, etc NOT due to base motion! Pressure increases 243 Hz Offset by 0. 01 mm 215 Hz 193 Hz Rotor motion amplitude at system natural frequency decreases as feed pressure increases

Effect of rotor speed 26, 30 & 34 krpm Shaker frequency: 12 Hz Feed pressure: 2. 36 bar (ab) 12 Hz, 24 Hz, 36 hz, etc Speed increases 193 Hz 180 Hz Rotor motion amplitude at system natural frequency increases as rotor speed increases

Effect of base frequency 0, 5, 6, 9, 12 Hz Rotor speed: 34 krpm (567 Hz) Feed pressure: 2. 36 bar (ab) 193 Hz Frequency increases NOT due to base motion! Rotor motion amplitude at natural frequency increases as excitation frequency increases

Rigid rotor model Rotor 1 st elastic mode: 1, 917 Hz (115 krpm) Equations of motion (linear system) U, Ub: rotor and base (abs) motions, Z=U-Ub M, G: rotor inertia and gyroscopic matrices W: rotor weight Fimb: imbalance “force” vector K, C: bearing stiffness and damping from gas bearing model (San Andres, 2006) Rework equations in terms of measured variables: System response = superposition of single frequency responses

Rigid rotor model Rotor speed Predicted natural frequencies 26 krpm 30 krpm 34 krpm Conical 191 Hz 200 Hz 208 Hz Cylindrical 184 Hz 192 Hz 200 Hz Measured from 1 X response tests Cylindrical 180 Hz 182 Hz 193 Hz Good agreement shows predicted bearing force coefficients are accurate For predictions: input RECORDED BASE accelerations (vertical)

Predictions vs. measurements Shaker input frequency: 12 Hz Feed pressure: 2. 36 bar (ab) Rotor speed: 34 krpm (567 Hz) Excitation freqs. Nat freq. 1 X Above natural frequency, RBS is isolated! Predictions in good agreement! Test rotor-bearing system shows good isolation.

Conclusions Base Excitations on Gas-Rotor Bearing Syst • Rotor response contains 1 X, excitation frequency (5 -12 Hz) and its super harmonics and system natural frequency. • Rotor motion amplitudes at natural frequency are smaller than synchronous amplitudes. • Excited rotor motion amplitude at system natural frequency increases as gas bearing feed pressure (5. 08~2. 36 bar) decreases, as rotor speed (26~34 krpm) increases, and as the shaker input frequency (5~12 Hz) increases. • Predicted rotor motion responses obtained from rigid rotor model show good correlation with test data. Demonstrated isolation of rotor-air bearing system to withstand base excitations at low freqs.

Acknowledgments Thanks support of • TAMU Turbomachinery Research Consortium • Bearings+ Co. (Houston) Learn more http: //rotorlab. tamu. edu Questions ?