ASME TURBO EXPO 2010 Glasgow Scotland UK June

ASME TURBO EXPO 2010, Glasgow, Scotland, UK (June 2010) Identification of Rotordynamic Force Coefficients of a Metal Mesh Foil Bearing using Impact Load Excitations Luis San Andrés Thomas Abraham Chirathadam Mast-Childs Professor Fellow ASME Research Assistant Texas A&M University ASME GT 2010 -22440 ASME J. Eng Gas Turbines & Power (in print) Supported by TAMU Turbomachinery Research Consortium

Metal Mesh Foil Bearing (MMFB) MMFB COMPONENTS: Bearing Cartridge, Metal mesh ring and Top Foil Hydrodynamic air film develops between rotating shaft and top foil Potential applications: ACMs, micro gas turbines, turbo expanders, turbo compressors, turbo blowers, automotive turbochargers, APU n Large damping (material hysteresis) offered by metal mesh n Tolerant to misalignment, and applicable to a wide temperature range n Suitable tribological coatings needed to Top reduce friction at start-up & shutdown Foil Metal mesh ring Cartridge

Metal Mesh Foil Bearings (+/-) n No lubrication (oil-free). NO High or Low temperature limits. n Resilient structure with lots of material damping. n Simple construction ( in comparison with other foil bearings) n Cheap! n Metal mesh tends to sag or creep over time n Damping NOT viscous. Modeling difficulties n Unknown rotordynamic force coefficients

Past work in Metal Mesh Dampers METAL MESH DAMPERS provide large amounts of damping. Inexpensive. Oil-free Zarzour and Vance (2000) J. Eng. Gas Turb. & Power, Vol. 122 Advantages of Metal Mesh Dampers over SFDs Capable of operating at low and high temperatures No changes in performance if soaked in oil Al-Khateeb and Vance (2001) GT-2001 -0247 Test metal mesh donut and squirrel cage( in parallel) Metal Mesh damping not affected by modifying squirrel cage stiffness Choudhry and Vance (2005) Proc. GT 2005 Develop design equations, empirically based, to predict structural stiffness and viscous damping coefficient

Past work in MMFBs San Andrés et al. (2010) J. Eng. Gas Turb. & Power, Vol. 132(3) Assembled the first prototype MMFB (L=D=28 mm). Load vs Deflection with hysteresis shows large structural damping (g~ 0. 7). Frequency dependent stiffness agree with predictions. San Andrés et al. (2009) ASME GT 2009 -59920 Demonstrated operation to 45 krpm with early rotor lift off. Educated undergraduate students. San Andrés et al. (2009) Proc. AHS 65 th Annual Forum, May, 2009 Start and shut down to measure torque and lift-off speed. Low friction factor ~ 0. 01 at high speed 60 krpm.

MMFB assembly Simple construction and assembly procedure BEARING CARTRIDGE METAL MESH RING TOP FOIL

MMFB dimensions and specs Dimensions and Specifications Bearing Cartridge outer diameter, DBo(mm) 58. 15 Bearing Cartridge inner diameter, DBi(mm) 42. 10 Bearing Axial length, L (mm) 28. 00 Metal mesh donut outer diameter, DMMo (mm) 42. 10 Metal mesh donut inner diameter, DMMi(mm) 28. 30 Metal mesh density, ρMM (%) 20 Top foil thickness, Ttf (mm) 0. 127 Metal wire diameter, DW (mm) 0. 30 Young’s modulus of Copper, E (GPa), at 21 ºC Poisson’s ratio of Copper, υ 114 Bearing mass (Cartridge Foil+sensors), M (kg) PICTURE + Mesh 0. 33 + 0. 380 Mesh thickness= 7 mm

Static load vs. MMFB deflection San Andres, L. , Chirathadam, T. A. , and Kim, T. H. , 2010, ASME J. Eng. Gas Turbines Power, 132 (3), p. 032503. Start 3 Cycles: loading & unloading Nonlinear F(X) Large hysteresis loop : Mechanical energy dissipation Displacement: [-0. 12, 0. 12] mm Load: [-120, 150 ]N MMFB Structural Characteristic (wire density ~

MMFB structural stiffness K=d. F/dx San Andres, L. , Chirathadam, T. A. , and Kim, T. H. , 2010, ASME J. Eng. Gas Turbines Power, 132 (3) During Load reversal : jump in structural stiffness Lower stiffness values for small displacement amplitudes Max. Stiffness ~ 2. 5 MN/m MMFB Structure Characterization (wire density ~ 20%)

MMFB structural stiffness vs. freq. San Andres, L. , Chirathadam, T. A. , and Kim, T. H. , 2010, ASME J. Eng. Gas Turbines Power, 132 (3) Motion amplitude increases 12. 7 um At low frequencies (25100 Hz), stiffness decreases At higher frequencies, stiffness gradually increases 25. 4 um 38. 1 um Bearing stiffness is frequency and motion amplitude dependent Al-Khateeb & Vance model: reduction of stiffness with force magnitude (amplitude dependent)

MMFB eq. damping vs. frequency Amplitude increases San Andres, L. , Chirathadam, T. A. , and Kim, T. H. , 2010, ASME J. Eng. Gas Turbines Power, 132 (3) 12. 7 μm 25. 4 μm 38. 1 μm MMFB equiv. viscous damping decreases as the excitation frequency increases and as motion amplitude increases Predicted equivalent viscous damping coefficients in good agreement with measurements

MMFB rotordynamic test rig MMFB TC cross-sectional view Max. operating speed: 75 krpm Turbocharger driven rotor Regulated air supply: 9. 30 bar Journal: length 55 mm, 28 mm diameter , weight=0. 22 kg Journal press fitted on Shaft Stub Twin ball bearing turbocharger Model T 25

Journal speed and torque vs time Rotor starts Valve Constant speed Valve ~ 65 krpm open close Rotor stops Applied Load: 17. 8 N WD= 3. 6 N Iift off speed Manual speed up to 65 krpm, steady state operation, and deceleration to rest Startup torque ~ 110 Nmm Shutdown torque ~ 80 Nmm 3 N-mm Once airborne, drag torque is ~ 3 % of Startup ‘breakaway’ torque San Andres, L. , Kim, T. H. , Chirathadam, T. A. , and Ryu, K. , 2009, Proc. AHS 65 th Annual forum, Grapevine, TX, May 27 -29. Lift off speed at lowest torque : airborne operation Top shaft speed = 65 krpm

Bearing power loss vs rotor speed Mixed lubrication Increasing static load (Ws) to 35. 6 N (8 lb) Hydrodynamic lubrication Dead weight (WD= 3. 6 N) 35. 6 N (8 lb) 26. 7 N (6 lb) 17. 8 N (4 lb) 8. 9 N (2 lb) Rotor accelerates Power loss decreases to a minimum during mixed lubrication regime and then increases with increasing rotor speed

Friction coefficient vs rotor speed f = (Torque/Radius)/(Static load) f ~ 0. 01 Friction coefficient ( f ) decreases with increasing static load 8. 9 N (2 lb) 17. 8 N (4 lb) 26. 7 N (6 lb) Rotor accelerates f ~ 0. 3 -0. 4 Dry sliding Airborne (hydrodynamic) 35. 6 N (8 lb) San Andres, L. , Kim, T. H. , Chirathadam, T. A. , and Ryu, K. , 2009, Proc. AHS 65 th Annual forum, Grapevine, TX, May 27 -29.

Identification of stiffness and damping coeff. Top foil fixed end Force gauge Flexible string TC IMPACT HAMMER MMFB Journal (28 mm) Eddy current sensor Accelerometer Positioning table (FRONT VIEW) (SIDE VIEW) 5 cm Impact loads

Identification model Equations of motion: Assembly mass, M = 0. 38 kg Deliver impacts along Y direction only KYY, CYY Y Impact force, f. Y KXY, CXY Bearing Cartridg e Ω KXX, CXX Journal X KYX, CYX Record displacements (relative to rotor) and bearing acceleration

Identification model: freq. domain Lightly loaded bearing (3. 5 N). Assume: SHAFT STATIONARY ( 1 -DOF) SHAFT ROTATING (2 -DOF) Multiple tests (10) : frequency averages

Impact force and displacements Frequency domain averages of 10 impacts Y Time domain Frequency domain Shaft not rotating Time domain Frequency domain Rapid decay of MMFB displacement indicates large material damping

Impact force and displacements Y direction X direction Time domain Frequency domain Y direction X direction Shaft speed = 50 krpm Synchronous response, 1 X Appreciable cross directional motions (X)

Bearing acceleration & relative disp. Acceleration from displacement relative to shaft, |ω2 Y| Bearing acceleration Acceleration from displacement relative to shaft Y direction Measured SHAFT STATIONARY Acceleration derived from bearing displacement relative to shaft, |-ω2 Y|, is not equal to bearing acceleration since the TC shaft stub is rather flexible. Important to measure both: displacements and accelerations (X, Y) X direction SHAFT speed = 50 krpm (833 Hz)

Test MMFB stiffnesses (K, k) Structure K ( No rotation) Direct K Applied load = 3. 5 N. Weight=3. 5 N Shaft speed=50 krpm Direct stiffness K at 50 krpm < structure stiffness ( ~ 25 % reduction at 200 hz) Cross k Direct and cross coupled stiffnesses (airborne) and bearing structural stiffness increase with frequency

Test MMFB damping (C, c) Applied load = 3. 5 N. Weight=3. 5 N Shaft speed=50 krpm Structure C Direct C Cross c MMFB direct viscous damping C are similar, with and without shaft rotation. Metal mesh provides ++ damping Equivalent viscous damping decays rapidly with increasing frequency

Test MMFB loss factor (g) γ=Cω/K Applied load = 3. 5 N. Weight=3. 5 N Shaft speed=50 krpm No rotation Loss factor is large g ~ 0. 5, with little variation in frequency Large loss factor magnitude = large energy dissipation mechanism

Waterfalls of start up Not all measurements showed acceptable rotordynamic performance. At certain speeds, rotor-bearing system shows large subharmonic motions. Is this behavior a typical rotordynamic instability or a forced nonlinearity ?

Bearing displacements relative to shaft H load =3. 5 N Subharmonic whirl motions of large amplitudes locked at system natural frequency ½ frequeny whirl for lightly loaded bearing

Bearing displacements relative to shaft H load = 18 N Large sub harmonic motions locked at natural frequency ½ frequeny whirl absent with larger applied loads

Conclusions Metal mesh foil bearing assembled using cheap, commercially available materials. ØWhile airborne, bearing power loss increases with rotor speed (little friction). Min power loss found. ØMMFB direct stiffness (airborne) slightly < structural stiffness. Cross-stiffness small. MMFB viscous damping nearly independent of shaft speed though decreasing fast with frequency. LOSS factor is large (~0. 50) Ø Start up shows ½ frequency whirl=natural frequency for low static loads & speeds < 50 krpm Ø MMFBs are promising inexpensive bearings for oil-free turbomachinery

Acknowledgments Thanks support of • Turbomachinery Research Consortium • Honeywell Turbo. Charging Systems Learn more http: //rotorlab. tamu. edu Questions ?