Thermohydrodynamic Analysis of Bump Type Gas Foil Bearings: A Model Anchored to Test Data

Transcription

1 June Update Thermohydrodynamic Analysis of Bump Type Gas Foil Bearings: A Model Anchored to Test Data Luis San Andrés Mast-Childs Professor Tae Ho Kim Post-Doc Research Associate Keun Ryu PhD Research Assistant This material is based upon work supported by NASA NNH6ZEA1N-SSRW2, Fundamental Aeronautics: Subsonic Rotary Wing Project and the Texas A&M Turbomachinery Research Consortium

2 Outline Statement of Work & Sources for Presentation Objectives and accomplished work in 27-8 Computational model. Validation with published data. Rotordynamic measurements at TAMU Objectives and accomplished work in 28-9 Description of test rig and foil bearings at TAMU Effect of temperature on bearing temperatures, coastdown speed and rotor motions Effect of cooling flow on shaft and bearing temperatures. Validation of computational model * The computational code Graphical User Interface. Further predictions GFB thermal management tests and preds. Added Value and Closure 2

3 Topic Statement of Work & Sources for Presentation Objectives and accomplished work in 27-8 Computational model. Validation with published data. Rotordynamic measurements at TAMU Objectives and accomplished work in 28-9 Description of test rig and foil bearings at TAMU Effect of temperature on bearing temperatures, coastdown speed and rotor motions Effect of cooling flow on bearing and shaft temperatures. Validation of computational model * The computational code Graphical User Interface. Further predictions GFB thermal management tests and preds. Closure & added value 3

4 Gas Foil Bearings (+/-) Increased reliability: large load capacity (< 1 psi) No lubricant supply system, i.e. reduce weight High and low temperature capability (up to 2,5 K) No scheduled maintenance Ability to sustain high vibration and shock load. Quiet operation Less load capacity than rolling or oil bearings Wear during start up & shut down No test data for rotordynamic force coefficients Thermal management issues Predictive models lack validation. Difficulties in modeling + dry-friction damping + effects of temperature on material properties and components expansion. Applications: ACMs, micro gas turbines, turbo expanders 4

5 SOW Main Objective Ω To develop a detailed, physicsbased computational model of gas-lubricated foil journal bearings including thermal effects to predict bearing performance. The result of this work shall include a fully tested and experimentally verified design tool for predicting gas foil journal bearing torque, load, gas film thickness, pressure, flow field, temperature distribution, thermal deformation, foil deflections, stiffness, damping, and any other important parameters. Agreement NASA NNH6ZEA1N-SSRW2 5

6 References Foil Bearings ASME GT ASME GT ASME GT NASA/TM ASME GT ASME GT J Eng Gas Turbines & Power 8th IFToMM Int. Conf. on Rotordynamics ASME GT AHS 29 paper ASME GT De Santiago, O., and San Andrés, L., 211, Parametric Study of Bump Foil Gas Bearings for Industrial Applications San Andrés, L.., and Ryu, K., 211, On the Nonlinear Dynamics of Rotor-Foil Bearing Systems: Effects of Shaft Acceleration, Mass Imbalance and Bearing Mechanical Energy Dissipation. Howard, S., and San Andrés, L., 211, A New Analysis Tool Assessment for Rotordynamic Modeling of Gas Foil Bearings, ASME J. Eng. Gas Turbines and Power, v 133 San Andrés, L., Ryu, K., and Kim, T-H, 211, Thermal Management and Rotordynamic Performance of a Hot Rotor-Gas Foil Bearings System. Part 2: Predictions versus Test Data, ASME J. Eng. Gas Turbines and Power, v 133 San Andrés, L., Ryu, K., and Kim, T-H, 211, Thermal Management and Rotordynamic Performance of a Hot Rotor-Gas Foil Bearings System. Part 1: Measurements, ASME J. Eng. Gas Turbines and Power, v 133 San Andrés, L., Ryu, K., and Kim, T.H., 211, Identification of Structural Stiffness and Energy Dissipation parameters in a 2nd Generation Foil Bearing; Effect of Shaft Temperature, ASME J. Eng. Gas Turbines Power, vol. 133 (March), pp San Andrés, L., Camero, J., Muller, S., Chirathadam, T., and Ryu, K., 21, Measurements of Drag Torque, Lift Off Speed, and Structural Parameters in a 1st Generation Floating Gas Foil Bearing, Seoul, S. Korea (Sept.) San Andrés, L., Kim, T.H., Ryu, K., Chirathadam, T. A., Hagen, K., Martinez, A., Rice, B., Niedbalski, N., Hung, W., and Johnson, M., 29, Gas Bearing Technology for Oil-Free Microturbomachinery Research Experience for Undergraduate (REU) Program at Texas A&M University Kim, T. H., and San Andrés, L., 21, Thermohydrodynamic Model Predictions and Performance Measurements of Bump-Type Foil Bearing for Oil-Free Turboshaft Engines in Rotorcraft Propulsion Systems, ASME J. of Tribology, v132 6 San Andrés, L., and Kim, T.H., 21, Thermohydrodynamic Analysis of Bump Type gas Foil Bearings: A Model Anchored to Test Data, ASME J. Eng. Gas Turbines and Power, v 132

7 References Foil Bearings IJTC ASME GT IJTC ASME GT J of Tribology Tribology International Kim, T.H., and San Andrés, L., 29, "Effects of a Mechanical Preload on the Dynamic Force Response of Gas Foil Bearings - Measurements and Model Predictions," Tribology Transactions, v52 Kim, T. H., and San Andrés, L., 29, Effect of Side End Pressurization on the Dynamic Performance of Gas Foil Bearings A Model Anchored to Test Data, ASME J. Eng. Gas Turbines and Power, v Best PAPER Rotordynamics IGTI San Andrés, L., and Kim, T.H., 29, Analysis of Gas Foil Bearings Integrating FE Top Foil Models, Tribology International, v42 Kim, T.H., Breedlove, A., and San Andrés, L., 29, Characterization of Foil Bearing Structure at Increasing Temperatures: Static Load and Dynamic Force Performance, ASME Journal of Tribology, v 131(3) San Andrés, L., and Kim, T.H., 28, Forced Nonlinear Response of Gas Foil Bearing Supported Rotors, Tribology International, 41(8), pp AIAA ASME GT ASME GT Kim, T-H, and L., San Andrés, 27, Analysis of Gas Foil Bearings with Piecewise Linear Elastic Supports. Tribology International, 4, pp San Andrés, L., and T.H. Kim, 27, Issues on Instability and Force Nonlinearity in Gas Foil Bearing Supported Rotors, 43 rd AIAA/ASME/SAE/ASEE Joint Propulsion Conference, Cincinnati, OH, July 9-11 Kim, T.H., and L. San Andrés, 28, Heavily Loaded Gas Foil Bearings: a Model Anchored to Test Data, ASME J. Eng. Gas Turbines and Power, v13 San Andrés, L., D. Rubio, and T.H. Kim, 27, Rotordynamic Performance of a Rotor Supported on Bump Type Foil Gas Bearings: Experiments and Predictions, ASME J. Eng. Gas Turbines and Power, v129 ASME GT ASME GT Rubio, D., and L. San Andrés, 27, Structural Stiffness, Dry-Friction Coefficient and Equivalent Viscous Damping in a Bump-Type Foil Gas Bearing, ASME J. Eng. Gas Turbines and Power, v Best PAPER Rotordynamics IGTI Structures and Dynamics Committee 7 San Andrés, L., and D. Rubio, 26, Bump-Type Foil Bearing Structural Stiffness: Experiments and Predictions, ASME J. Eng. Gas Turbines and Power, v128

8 References Metal mesh foil bearings ASME GT ASME GT ASME GT AHS Paper Other Journal of Tribology ASME DETC San Andrés, L., and Chirathadam, T., 211, Metal Mesh Foil Bearings: Effect of Excitation Frequency on Rotordynamic Force Coefficients San Andrés, L., and Chirathadam T.A., 21, Identification of Rotordynamic Force Coefficients of a Metal Mesh Foil Bearing Using Impact Load Excitations. San Andrés, L., Chirathadam, T. A., and Kim, T.H., 29, Measurements of Structural Stiffness and Damping Coefficients in a Metal Mesh Foil Bearing. San Andrés, L., Kim, T.H., Chirathadam, T.A., and Ryu, K., 29, Measurements of Drag Torque, Lift-Off Journal Speed and Temperature in a Metal Mesh Foil Bearing, American Helicopter Society 65th Annual Forum, Grapevine, Texas, May Kim, T.H., and L. San Andrés, 26, Limits for High Speed Operation of Gas Foil Bearings, ASME Journal of Tribology, 128, pp Gjika, K., C. Groves, L. San Andrés, and G. LaRue, 27, Nonlinear Dynamic Behavior of Turbocharger Rotor-Bearing Systems with Hydrodynamic Oil Film and Squeeze Film Damper in Series: Prediction and Experiment. 8

9 9

10 Topic Statement of Work & Sources for Presentation Objectives and accomplished work in 7-8 Computational model. Validation with published data. Rotordynamic measurements at TAMU Objectives and accomplished work in 28-9 Description of test rig and foil bearings at TAMU Effect of temperature on bearing temperatures, coastdown speed and rotor motions Effect of cooling flow on bearing and shaft temperatures. Validation of computational model * The computational code Graphical User Interface. Further predictions GFB thermal management tests and preds. Closure & added value 1

11 Research Objectives (27-8) THD model for prediction of GFB performance Perform physical analysis, derive governing equations, and implement numerical solution. Develop GUI for User ready use Compare GFB predictions to limited published test data (NASA mainly) Revamp existing test rig with cartridge heater, acquire new bearings, machine new rotor Perform structural tests on bearings and measure rotordynamic response for increasing shaft temperatures 11

12 Scheduled Timeline & completion Luis San Andres Tae Ho Kim MS student UG worker Task Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Computational analysis GFBS Development physical model for thermal transport in foil bearings Implementation thermal model (Finite Element Based) and coupling to existing STRUCTURAL MODEL Integration of thermal model with GFB FD computational code (gas film) Predictions of GFB performance for parametric studies Comparison of GFB predictions to measured performance from TAMU test rig Nonlinear analysis GFBS Development simple NONLINEAR physical model for foil bearings Prediction of performance and comparisons to available rotordynamic test data Test rig for identification of FB structure (High Temperature) Planning of modification, selection of instrumentation and cartridge heater, design of insulation cover Reception of parts and assembly of components, troubleshooting, connection to static loader and shaker Measurements of load & bearing deflection for increasing shaft temperatures (max 5 C), identification of FB structural parameters Rotordynamic-GFBs Test rig (High Temperature) Planning of modifications to existing, selection of instrumentation and cartridge heater, design of insulation cover and rotor Reception of parts and assembly of components, troubleshooting, connection to static loader and shaker Rotordynamic Measurements for increasing shaft temperatures (max 5 C), identification of GFB synchronous force coefficients 12

13 Accomplishments: Proposed & Actual Task Thermohydrodynamic Analysis of GFBS Physical model for thermal transport in foil bearings. Integration of thermal model with GFB FD computational code (gas film). Prediction of GFB performance: parametric study Validation of GFB predictions with measured temperatures from NASA & TAMU published research Nonlinear structural analysis of GFBS Development simple NONLINEAR physical model for foil bearings. Prediction of performance and comparisons to rotordynamic test data Test rig for identification of FB structure (High Temperature) Design & construction; selection & procurement of instrumentation and bearings; assembly, troubleshooting and operation at high temperature. Measurements of static load performance & comparison to predictions Rotordynamic-GFBs Test rig (High Temperature) Design & construction; selection & procurement of instrumentation and bearings; assembly, troubleshooting and operation at high temperature rotor-bearing test rig. Measurements of temperatures and rotordynamic performance with Foster-Miller GFBs completed (see Q7). Tests with MiTi bearings at higher temperatures in progress. Validation of computational model also in progress. Planned & Actual 1% comment Analysis completed. Code delivered on June 1, 29 See Q4 & Q7 reports Implementation in XLTRC2 for ready rotordynamic analyses See Q4, Q7 reports Completed Dec 21 1 st & 2 nd Years 13

14 Thermohydrodynamic model in a GFB X=RΘ Ω Y Top foil Bump strip layer Hollow shaft Inner flow stream Outer flow stream Side view of GFB with hollow shaft - Ideal gas with density, - Gas viscosity, ρ= P Rg T μ= α v T X Thin film flow Bearing housing External fluid medium - Gas Specific heat (c p ) and thermal conductivity (κ g ) at an effective temperature Reynolds equation in thin film 3 3 hp f f P f hp f f P f Ph f f + = Umz ( ) x 12μf gtf x z 12μf gtf z x gt R R R f 14

15 THD model Outer flow stream x z Bearing housing Bearing housing Bump layer Top foil GFB with cooling flows (inner and/or outer) P Co,T Co Thin film flow P a ΩR So P Ci, T Ci Inner flow stream Y Z z= z=l Bulk-flow temperature transport equation ( ρf hf U ftf ) ( ρf hf WfTf ) X c + + h T T h T T f i o x z ( ) ( ) p ff f F Sf S f Pf Pf 12μ f 2 1 ( ) 2 = U f hf + Wf hf + Wf + Um + U f Um x z hf 3 Hollow shaft T Convection of heat by fluid flow + diffusion to bounding surfaces = compression 15 work + dissipated energy

16 Heat flow paths in rotor - GFB system T Ci S C Q Heat carried by inner flow Q stream Si T Si T So So Q S f Q f F Q Heat carried by thin film flow Q Ci Heat conduction through shaft R Si R So R Fi R Fo R Bi R Bo T S T f T Fi T Fo T O F Q F O Q F Bi Q Heat carried by outer flow stream Hollow shaft Ω Q T Bi B O Bi Q Top foil Q Co T Bo B Q Bump layer Q B T Bearing housing Q=Rq External fluid Drag dissipation power (gas film) : Heat (+) dө : Heat (-) Heat flows & thermal resistances in a GFB & hollow shaft Heat conducted into bearing Cooling gas streams carry away heat 16

17 THD Model Validation Published data Generation I GFB with single top foil and bump strip layer Parameters Value / comment Parameters Value Bearing cartridge Gas properties at 21 C Bearing inner radius 25 mm Ref. [7] Gas Constant 287 J/(kg- K) Bearing length 41 mm Ref. [7] Viscosity 1-5 Pa-s Bearing cartridge thickness 5 mm Assumed Nominal radial clearance 2 μm Assumed Conductivity.257 W/m K Top foil and bump strip layer Density kg/m 3 Top foil thickness 127 μm Ref. [21] Specific heat 1,2 J/kg K Bump foil thickness 127 μm Ref. [21] Ambient pressure 1.14 x 1 5 Pa Bump half length Bump pitch Bump height Number of bumps x strips Bump foil Young s modulus Bump foil Poisson s ratio Bump foil stiffness mm 4.64 mm.58 mm 39 x 1 2 GPa GN/m 3 Assumed Assumed Assumed Assumed Gas viscosity & density & conductivity, foil Young s modulus, and clearance change with temperature. Radil and Zeszotek, 24 Dykas and Howard, 24 17

18 Peak film temperature Predictions & test data Peak temperature [ C] Test data (Mid-plane) Predictions (Mid-plane) Static load [N] Supply air (T Supply ), shaft (T S ), and bearing OD (T B ) temperatures at 21 C. Film temperature + higher than ambient, even for small load of 18 9 N. Good agreement preds with test data 5, rpm 4, rpm 3, rpm 2, rpm Predictions & test data Radil and Zeszotek, 24 T Supply =21 C Read more in ASME Paper GT

19 28 rotor-gfb test rig Max. temp. 13 C Infrared thermometer Electromagnet loader Strain gage load cell Tachometer Cooling air hose Drive motor (25 krpm). Cartridge heater max. temperature: 3F Eddy current sensor GFB housing Cartridge heater Test rotor Flexible coupling Driving motor Air flow meter (Max. 1 L/min at 14 psig) 19

20 28 hot rotor-gfb test rig Ambient temp. (T a ) ~ 22 C (71 F) Infrared Thermometer (Max. 54 ºC) Infrared Thermometer gun (Max. 5 ºC) Cartridge heater Eddy current sensor (Max. 177 ºC) Hollow test rotor Cooling air K-type thermocouples (Max. 48 ºC) Speed sensor Flexible coupling DC router motor (25 krpm) cm 1 1 Bearing sleeve temperature (at five locations along circumference) 2 Bearing outer surface temperature (Drive and bearing and free end bearing) 3 Rotor surface temperature (Drive end and free end) 4 Bearing support (housing) surface temperature (Drive end and free end) Numbers in circles show locations of temperature measurement. 2

21 THD Model Validation Bearings at TAMU Parameter [mm] Bearing cartridge Outer diameter Inner diameter Top foil and bump strip layer Top foil axial length Top foil thickness Bump foil thickness Number of Bumps Bump pitch Bump length Bump height Bump arc radius Bump arc angle [deg] Elastic Modulus 214 GPa, Poisson ratio=.29 Foster-Miller (2 nd gen.) axial KIST (1st gen.) axial MiTi (2 nd gen.) axial Foster-Miller FB with Teflon coating (Generation II) 21

22 Waterfall plots: Amplitude [microns] [μm, -Pk] Amplitude [μm, [microns] -Pk] X (a) Case 1 T c = 22 C Frequency [Hz] [Hz] 1X (b) Case 2 T c = 93 C Frequency [Hz] [Hz] coastdown responses Case 1-3 without cooling flow, T a ~ 22 C, and T hs = 22 C, 93 C, and 132 C 4krpm 15 krpm 25 krpm 4krpm 15 krpm 25 krpm Amplitude [μm, [microns] -Pk] Cartridge temperature (T hs ) increases 1X FE vertical plane 1X component dominant, i.e., no subsynchronous motion, during coastdown tests (c) Case 3 T c = 132 C 4krpm 15 krpm 25 krpm Frequency [Hz] [Hz]

23 Rotor speed up 1X response W/o cooling flow, T a ~ 22 C, and T hs = 22 C, 93 C & 132 C Amplitude [μm, -pk] Cartridge temperature (T hs ) increases (22 C) (93 C) Rotor speed [rpm] (132 C) DE vertical plane Speed - up Above critical speed ~ 14.5 krpm, amplitude drops. A 23 nonlinearity! As T hs increases, the peak amplitude decreases.

24 Rotor coastdown 1X response W/o cooling flow, T a ~ 22 C, and T hs = 22 C, 93 C, and 132 C Amplitude [μm, -pk] Cartridge temperature (T hs ) increases (22 C) (93 C) Rotor speed [rpm] (132 C) DE vertical plane Coastdown Read more in AHS Paper As T hs increases, critical speed raises by ~ 2 krpm and the peak amplitude decreases. Nonlinearity absent! 24

25 25

26 Topic Statement of Work & Sources for Presentation Objectives and accomplished work in 7-8 Computational model. Validation with published data. Rotordynamic measurements at TAMU Objectives and accomplished work in 28-9 Description of test rig and foil bearings at TAMU Effect of temperature on bearing temperatures, coastdown speed and rotor motions Effect of cooling flow on bearing and shaft temperatures. Validation of computational model * The computational code Graphical User Interface. Further predictions GFB thermal management tests and preds. Closure & added value 26

27 Research Objectives 28-9 Model validation with TAMU FB test data Complete test rig using cartridge heater for high temperature operation (up to 36C) Measure rotordynamic performance during speed coastdown from 3 krpm and for increasing shaft temperatures Quantify effect of side flow on cooling bearings (max. 15 LPM per bearing) Benchmark model predictions 27

28 THD Model Validation Bearings at TAMU Parameter [mm] Foster-Miller (2 nd gen.) KIST (1st gen.) MiTi (2 nd gen.) Bearing cartridge Outer diameter Inner diameter Top foil and bump strip layer Top foil axial length Top foil thickness Bump foil thickness Number of Bumps 25 5 axial 26 1 axial 24 3 axial Bump pitch Bump length Bump height Bump arc radius Bump arc angle [deg] Elastic Modulus 214 GPa, Poisson ratio=.29 28

29 29 hot rotor-gfb test rig Max. 36 C Instrumentation for high temperature. Insulation casing Gas flow meter (Max. 5 LPM). Drive motor (max. 65 krpm) Infrared thermometer Insulated safety cover ) Eddy current sensors Cartridge heater Tachometer Drive motor Hot heater inside rotor spinning 3 krpm Test hollow shaft (1.1 kg, 38.1mm OD, 21 mm length) Test GFBs Flexible coupling 29

30 Thermocouples in test rotor-gfb rig T1 Free end (FE) GFB 45º g Insulated safety cover g Drive end (DE) GFB 45º T6 Ω T2 Ω T3 T4 Th Tamb T11 T5 Hollow shaft T1 T12 T9 T8 Coupling cooling air T14 T16 T7 T15 Heater stand Cartridge heater Drive motor Cooling air T13 Foil bearings Overall 15 thermocouples for GFB cartridge outboard, Bearing support housing 3 surface, Drive motor, Test rig ambient, and Cartridge heater temperatures Two noncontact infrared thermometers for rotor surface temperature

31 Thermocouples in test GFB 5 5 thermocouple locations: 31 Θ = at 22, 94, 166, 238, 31 along bearing mid-plane Θ Oblique view Ω Spot weld Bump strip layer Top foil Bearing sleeve g Machined axial slot Bearing mid-plane X Y Foster-Miller FB uncoated (Generation II) five (5) thermocouples placed within machined axial slots. 31

32 Time to coast down rotor Effect of shaft temperature Speed 1 (krpm) Rotor speed [krpm] 1 1 Exponential decay Cartridge temperature (T hs ) increases No heating Ths=1ºC Ths=2ºC Ths=3ºC Ths=36ºC Coast down time [sec] Coastdown time (s) Baseline, Heater up to 36C. No forced cooling Long time to coastdown : very low viscous drag (no contact between rotor and bearings) Test Data Coastdown time lesser as rotor heats (reduced clearance) 32

33 Bearing outboard temperature predictions & test data Temperature rise [ C] Drive end FB Test data Predictions Max. Avg. Min. Error bar static load ~ 6.5N Rotor speed [krpm] rotor speed (krpm) No cooling 5 LPM Θ g Room temperature 21 C. Heater OFF w/o and w low cooling Rotor speed : 3 krpm Free End T1 45 Ω X Heat flux (Shaft bearing) Hot shaft (Isothermal) Y Spot weld Bump strip layer Top foil Bearing cartridge Drive End T6 Test data & predictions FB OD temperature rises with rotor speed and decreases with forced cooling stream ~ 5 LPM. Predictions agree with test data T11 T12

34 Rotor 1X response 1X motions as rotor heats Baseline. Heater to 36C. No forced cooling Tests Amplitude [μm, -pk] Drive End (H) DH Cartridge temperature (T hs ) increases No heating Ths=1ºC Ths=2ºC Ths=3ºC Ths=36ºC FH DH Free End T11=37º C Free End DH No heating Drive End T12=26º C Ths=2º C Drive End 1 System natural frequency FV g DV T11=92º C T12=7º C Free End Ths=36º C Drive End Speed [krpm] speed (krpm) Test Data T11=157º C T12=17º C As heater T increases to 36ºC, peak motion amplitude decreases 34 in speed range 7 krpm to 15 krpm

35 1X rotor response predictions & tests Baseline. Heater to 36C. No forced cooling 5 Drive End (H) Amplitude [μm, -pk] T hs =2ºC No heating T hs =36ºC Predictions Cartridge heater temperature increases Test data & predictions Rotor speed [krpm] Test data Rotor speed (krpm) As heater temperature raises, rotor amplitude decreases for speed 35 < 15 krpm & the critical speed increases from 14 krpm to 17 krpm

36 Bearing outboard temperature predictions & tests Bearing temperature rise [ºC] Bearing temperature rise [ºC] DE Bearing Temp Predictions Test data static load ~ 6.5N Shaft temperature rise [ºC] FE Bearing Temp Predictions Test data static load ~ 3.5N rise [ºC] Shaft temperature rise (C) Test data & predictions Free End Drive End T1 T6 T11 T12 Supply air (T Supply )~21 C. Heater up to 36C. No cooling flow Rotor speed : 3 krpm FB cartridge temperature increases linearly with shaft temperature

37 Cooling gas flow into GFBs Gas pressure Max. 1 psi heater AIR SUPPLY Cooling flow needed for thermal management: to remove heat from shear drag or to reduce thermal gradients in hot/cold engine sections 37

38 Cooling gas flow into GFBs Gas pressure Max. 1 psi heater AIR SUPPLY Heater warms unevenly rotor. Side cooling cools unevenly rotor and also heater 38

39 Heating of rotor Effect of rotor speed and side cooling Baseline imbalance, No side flow & 5 L/min : Temp. drop due to 5L/min cooling flow Free End T1 No heating Drive End T6 Temperature rise [ºC] krpm 2 krpm 3 krpm T11-Tamb T1-Tamb T6-Tamb T11 T12 Bearing cartridge and rotor temperatures increase steadily with time Rotor speed : 1, 2, 3 krpm Rotor speed makes rotor and bearings hotter T12-Tamb Time [sec] Test time (min) Test Data Heater OFF Cooling flow removes heat from shear dissipation in rotor, most 39 effective at high speed

40 Heater temperature Effect of cooling flow Heater temperature increases Heater up to 36C w/o & w cooling flows 4 Rotor speed : 3 krpm Heater temperature, Th [ºC] C 1C 2C 3C 36C No cooling & 5L/min 1L/min 15L/min Th Free End T1 T11 Heater power is limited Drive End T12 T Test Time time [min] (min) Test Data Cooling>1 LPM cools both rotor & HEATER! 4

41 Bearings OD temperatures Effect of cooling flow 7 Heater temperature increases Th Free End T1 Drive End T6 Temperature rise [ºC] C 1C 2C Cooling flow increases 3C T1-Tamb 36C T1-Tamb T6-Tamb T1-Tamb T6-Tamb T11 T12 High temp. (heater up to 36C). Cooling flow up to 15 L/min rotor speed : 3 krpm Time [min] T6-Tamb Test time (min) Test Data Cooling effective > 1 LPM and when heater 41 at highest temperature

42 Bearing cartridge temperature Effect of cooling flow Temperature rise [ºC] C 1C No heating Cooling flow increases T6-Tamb High temp. (heater up to 36C). Cooling flow to 15 LPM T1-Tamb rotor speed : 3 krpm Th Free End T1 T11 Drive End T12 T6 5 Cartridge temperature (T hs ) increases Cooling flow rate [L/min] Cooling flow (LPM) Turbulent flow > 1 LPM Test Data Bearing OD temperature decreases with cooling flow 42

43 Time to coast down rotor Effect of cooling flow 1 Cond. #7: No cooling, No heating Cond. #8: 5 L/min cooling, No heating Cond. #9: No cooling, Ths=36ºC Cond. #1: 5L/min cooling, Ths=36ºC Rotor speed [krpm] Rotor speed (krpm) 1 No cooling 5 L/min No heating 36C Coast down time [sec] Coastdown time (s) Test Data Coastdown time down by 2% (13 s) with cooling at 5 LPM 43

44 Bearing cartridge temperature Predictions & tests 6 DE Bearing temperature rise Supply air (T Supply )~21 C. Bearing temperature rise [ºC] Test data Predictions 5LPM No cooling 125LPM 1LPM 15 LPM static load ~ 6.5N Heater up to 36C. w/o & w cooling flow Rotor speed : 3 krpm Free End Drive End Th T1 T6 T11 T Shaft temperature rise [ºC] Shaft temperature rise (C) Test data & predictions As cooling flow rate increases, FB cartridge temperature decreases. Predictions agree with test data.

45 Post-test condition of rotor and GFBs Before operation FE Static load direction After extensive heating with rotor spinning DE Static load direction UNCOATED top foil! Before operation FE DE After extensive hearing with rotor spinning tests Wear marks on top foils are at side edges Rotor shows polishing marks at bearing locations. Deep wear 45 marks at outboard edges

46 Test data & predictions Amplitudes of rotor synchronous motion proportional to added imbalances. For operation with hot shaft, amplitude of rotor motion drops while crossing (rigid body mode) critical speed. As rotor and bearing temperatures increase, air becomes more viscous and bearing clearances decrease; hence coastdown time decreases. Thermal management with axial cooling streams is beneficial at high temperatures and with large flow rates ensuring turbulent flow conditions. Test foil bearings continue to survive high temperature & high vibration operation! 46

47 47

48 Topic Statement of Work & Sources for Presentation Objectives and accomplished work in 7-8 Computational model. Validation with published data. Rotordynamic measurements at TAMU Objectives and accomplished work in 28-9 Description of test rig and foil bearings at TAMU Effect of temperature on bearing temperatures, coastdown speed and rotor motions Effect of cooling flow on bearing and shaft temperatures. Validation of computational model * The computational code Graphical User Interface. Further predictions GFB thermal management tests and preds. Closure & added value 48

49 The computational program Windows XP OS and MS Excel 23 (minimum requirements) Fortran 99 Executables for FE underspring structure and gas film analyses. Prediction of forced static & dynamic- performance. Excel Graphical User Interface (US and SI physical units). Input & output (graphical) Compatible with XLTRC 2 and XLROTOR codes Code: XL_GFB_THD Delivered on June 29 49

50 Graphical User Interface Worksheet: Shaft & Bearing models (I) 5

51 Graphical User Interface Worksheet: Shaft & Bearing models (II) 51

52 Graphical User Interface Box 8 Box 9 Box 1 Worksheet: Top Foil and Bump Models 52

53 Graphical User Interface Worksheet: Foil Bearing (Operation and Results) 53

54 Static load parameters Drive End FB Journal eccentricity [μm] T hs =36ºC T hs =2ºC No heating Predictions Rotor speed [krpm] Th Cartridge heater temperature increases Rotor speed (krpm) Free End T1 T Journal attitude angle [deg] Minimum film thickness [μm] Drive End T hs =2ºC No heating Rotor speed [krpm] static load ~ 6.5 N No cooling flow T hs =36ºC Cartridge heater temperature increases As temperature increases, journal attitude angle and drag torque increase but journal eccentricity and minimum film thickness decrease due to 54 reduction in operating clearance T12 Predictions T6 Rotor speed (krpm) Drag torque [N-mm]

55 Bearing stiffnesses Predictions Drive End FB static load ~ 6.5 N No cooling flow Direct stiffness [MN/m] K XX K YY Predictions Cartridge heater temperature increases Rotor speed [krpm] Rotor speed (krpm) T hs =36ºC T hs =2ºC No heating Cross-coupled stiffness [MN/m] Cartridge heater temperature increases K YX K XY Rotor speed [krpm] No heating T hs =2ºC T hs =36ºC Rotor speed (krpm) As temperature increases, stiffnesses (K XX, K YY ) increase significantly, while difference (K XY -K YX ) increases slightly at low rotor speeds and decreases at high rotor speeds 55

56 Bearing damping Predictions Drive End FB Direct damping [kn-s/m] Cartridge heater temperature increases C XX C YY No heating T hs =36ºC T hs =2ºC Cross-coupled damping [kn-s/m] Cartridge heater temperature increases C YX C XY static load ~ 6.5 N No cooling flow No heating T hs =2ºC T hs =36ºC Rotor speed [krpm] Rotor speed (krpm) Rotor speed [krpm] Rotor speed (krpm) Predictions As temperature increases, damping (C XX, C YY ) increase. Cross 56 damping (C XY,C YX ) change little above 3 krpm.

57 Predictions on effect of cooling flow No cooling flow Peak temperature [ C] Predictions Laminar flow Outer cooling flow Inner and outer cooling flows Re D = 23 Cooling flow rate increases Turbulent flow Cooling flow rate [lit/min] 4, rpm 2, rpm Supply air (T Supply ), shaft (T S ), and bearing OD (T B ) temperatures at 21 C. Static load =89 N (18 ). T Supply =21 C Rotor speed : 2 & 4 krpm Peak temperature drops with strength of cooling stream. Sudden 57 drop at ~ 2 lit/min b/c of transition from laminar to turbulent flow

58 Predictions radial temperature Model DE FB Temperature rise [ C] T Ci T Si T So T f T Bo R Si R So R Fi R Fo R Bi R Bo Radial direction T Fi T Fo T Co T Bi Mean temperature Cooling stream increases No cooling Shaft temp. rise=79 C 5 L/min Shaft temp. rise=67 C 125 L/min Shaft temp. rise=39 C 15 L/min Shaft temp. rise=32 C T Natural convection on exposed surfaces of bearing OD and shaft ID w/o & w cooling flow Rotor speed : 3 krpm Hollow shaft Top foil Bump layer Bearing housing External fluid dө Model predictions With forced cooling, GFB operates 5 C cooler. Outer cooling stream is most effective in removing heat 58

59 59

60 Topic Statement of Work & Sources for Presentation Objectives and accomplished work in 7-8 Computational model. Validation with published data. Rotordynamic measurements at TAMU Objectives and accomplished work in 28-9 Description of test rig and foil bearings at TAMU Effect of temperature on bearing temperatures, coastdown speed and rotor motions Effect of cooling flow on bearing and shaft temperatures. Validation of computational model * The computational code Graphical User Interface. Further predictions Work with MiTi Bearings Closure & added value San Andrés, L., Ryu, K., and Kim, T.H., 211, Identification of Structural Stiffness and Energy Dissipation parameters in a 2nd Generation Foil Bearing; Effect of Shaft Temperature, ASME J. Eng. Gas Turbines Power, vol. 133 (March), pp

61 MiTi Korolon foil bearing Top foil Cartridge sheet Bumps Two generation II GFBs Three (axial) bump strip layers, each with 24 bumps. Korolon 8 coating (up to 8 F) on top foil surface. Bump arc radius [mm] Shaft diameter [mm] FB nominal dimensions Parameter [Dimension] Cartridge inner diameter [mm] Cartridge outer diameter [mm] Axial bearing length [mm] Number of bumps Bump pitch [mm] Bump length [mm] Bump foil thickness [mm] Bump height [mm] Top foil thickness [mm] Bearing Top foil inner diameter [mm] Nominal radial clearance [mm] Symbol D D O L N B s 2l o t h t T r B D T D J C Value

62 MiTi FB deflection versus static load Room temperature tests Lathe chuck Load cell Test 1 Test 2 Test Release Tests 1-3: Three cycles Rotor Eddy current sensor Bearing displacement [µm] e Push load Release Static load [N] Pull load F K X Nonlinear F(X) : Stiffness hardening FB fitted smoothly into 3.8 mm thick bearing shell 9 bearing orientation g Test bearing Live center Top foil spot weld Shaft OD mm: Highly preloaded FB Direction of static load Displacement sensor Large hysteresis loop : Mechanical energy dissipation due to dry-friction between top foil contacting bumps and bump strip layers contacting bearing cartridge sheet 62

63 MiTi FBstructural stiffness Room temperature tests Pull load Push load Poly. (Pull load) Poly. (Push load) y =.15x x x R 2 =.9718 Cubic polynomial curve fit over span of applied loads F=F +K 1 X+K 2 X 2 +K 3 X 3 Static load [N] e y =.127x x x R 2 = K=K 1 +2K 2 X+3K 3 X 2 25 Pull load -2 Bearing displacement [µm] Distinctive hardening effect as FB deflection increases Structrual stiffness [MN/m] Push load Bearing displacement [µm]

64 MiTi FB deflection versus static load Room temperature tests Test 1 Test 2 Test Release Tests 1-3: Three cycles Lathe chuck Load cell Eddy current sensor Bearing displacement [µm] -3 Push load -6 6 Pull load ROTOR OD mm FE Test bearing Rotor DE Test bearing -9 Release -12 Static load [N] 8 Live center FB sliding fit without play in 3.8 mm thick bearing shell! Identified from Cubic polynomial curve fit Static load [N] Pull load Push load 9 bearing orientation g Top foil spot weld Bearing displacement [µm] Direction of static load Displacement sensor 64

65 MiTi FB test setup for dynamic loads Thermocouples Cartridge heater 9 bearing orientation Bearing housing Eddy current sensor Load cell Shaker Test bearing Test shaft Index fixture Bearing housing Single frequency dynamic load in horizontal direction Shaft OD mm FB press fitted onto 15.5 mm thick bearing housing! FB Displacement controlled [µm] Frequency Range, Hz Shaft Temperature, C Bearing Mass M, kg 7.4, 11.1, 14.8, and (increment: 25 Hz) 23, 13, 183, and (load cell + attachment hardware) Uncoated rigid, nonrotating, hollow shaft 65 supported on FB.

66 Shaft heating using electric heater 134 mm Indexing fixture chuck T3 76 mm T4 T2 Bearing housing T1 Th MiTi B Ø25 mm Temperature [ C] Th T1 T2 T4 T3 Steady state temperature (heater 1 hr operation) Th T1 T2 T4 T3 Th Ø36.56 mm T1 T2 T4 Th=13 C 1 Th=183 C 2 Th=263 C 3 T3 Ø 25.4 mm Significant temperature gradient along shaft axis. Cartridge heater warms unevenly shaft 66 Heater set temperature (Th) [ C]

67 Parameter Identification (no shaft rotation) Equivalent Test System: 1DOF K stiffness, C eq viscous damping OR γ loss factor F ext M eq K eq M x+ K x+ C x = F eq () t x C eq F() t = F X F e i ω t O xt () = Xe i ω t O 2 Z = = ( K ω M) + iωceq Harmonic force & displacements Impedance Function E = πωc X dis Edis eq = πγk X 2 2 Viscous Dissipation or dry-friction Energy 67

68 Effect of temperature on dynamic stiffness 2 Re F O = ( K ω M) X T h = 23 C T h = 263 C Real part of impedance F/X [MN/m] FB motion amplitude increases 18.5 µm 7.4 µm 7.4 µm µm µm 18.5 µm Motion amplitude increases System natural frequency Frequency [Hz] Real part of impedance F/X [MN/m] µm 14.8 µm FB motion amplitude increases 7.4 µm µm 14.8 µm Frequency [Hz] Motion amplitude increases System natural frequency Real (F/X) decreases with FB motion amplitude & increases with shaft temperature. 68

69 Effect of temperature on structural stiffness K Structural stiffness [MN/m] FO 2 = Re + ω M X Motion amplitude increases Frequency [Hz] 7.4 µm 11.1 µm 14.8 µm 18.5 µm T h = 23 C, 9 bearing orientation Highly preloaded FB: K decreases as FB motion amplitude increases due to decrease in # of active bumps Structrual stiffness [MN/m] Dynamic structural K compared to static structural K Static pull load Static push load Dynamic load at 5 Hz Static load Dynamic load 2 Re F O ω 5 K = + M at Hz X Bearing displacement [µm] At larger FB deflections, static K is larger than dynamic K 69

70 Effect FB stiffness of temperature & viscous on damping stiffness & damping K Structural stiffness [MN/m] FO 2 = Re + ω M X ºC 13ºC 183ºC 263ºC Heater temperature increases Frequency [Hz] TEST FB cartridge OD is constrained within bearing housing. FB motion amplitude: 14.8 µm FB stiffness and viscous damping increase with shaft temperature and decrease with excitation frequency. FB radial clearance decreases as shaft temperature raises! Equivalent viscous damping [kns/m] C eq = F Im X ω O Heater temperature increases Frequency [Hz] 23ºC 13ºC 183ºC 263ºC 7

71 Effect Loss factor of temperature vs frequency on loss factor γ Structural (material) loss factor best represents energy dissipation in FB γ = C eq ω K Structural loss factor Heater temperature increases Frequency [Hz] 23ºC 13ºC 183ºC 263ºC FB motion amplitude: 14.8 µm The FB loss factor increases with excitation frequency and decreases slightly with shaft temperature. More damping expected in rotordynamic 71 measurements

72 72

73 Texas A&M University Mechanical Engineering Department Nov. 23, 21 EFFECT OF COOLING FLOW ON THE OPERATION OF A HOT ROTOR-GAS FOIL BEARING SYSTEM Ph. D. Final Exam Keun Ryu Chair of Advisory Committee: Dr. Luis San Andrés This material is based upon work supported by NASA GRC and the TAMU Turbomachinery Research Consortium (TRC)

74 Objective Quantify effect of cooling flow and shaft temperature on the rotordynamic performance of a GFB supported rotor. Investigate adequate thermal management strategies using forced cooling flow into the GFBs 74

75 Research tasks Revamp a GFB rotordynamic test rig for operation at high speed and extreme temperatures Measure temperature of bearings and rotor and the motions of rotor for increasing rotor speeds, shaft temperatures, and cooling flow rates Quantify effect of gas flow on cooling bearings (max. 5 L/min) Compare the experimental results (rotor responses and bearing temperatures) to predictions from an in-house computational program 75

76 TAMU Hot rotor-gfb test rig Instrumentation for high temperature Gas flow meter (Max. 5 LPM). Drive motor (max. 5 krpm) Infrared thermometers Hollow rotor Cooling air supply for coupling T h (Reference thermocouple to control heater temperature) Cartridge heater T5 FE GFB Displacement sensors Td (under cover) DE GFB T1 Infrared tachometer Flexible coupling Drive motor Tout Free end rotor temperature (Tr FE ) GFB support housing Free end rotor temperature (Tr DE ) 76

77 Cooling gas flow into GFBs Gas pressure Max. 1 psi heater AIR SUPPLY Cooling flow needed for thermal management: to remove heat from shear drag or to reduce thermal gradients from hot to cold engine sections 77

78 Cooling gas flow into GFBs Gas pressure Max. 1 psi heater AIR SUPPLY Heater warms unevenly test rotor. Side cooling flows cool unevenly the rotor. 78

79 Overview Thermal management Component-level tests Ruscitto et al (1978): Perform load capacity tests on 1st gen. GFB up to 45 kprm (1.7 MDN) and static load 111 N with 11 L/min cooling flow at 315C bearing temperature. DellaCorte (1998): No cooling flow. 3rd gen. GFB up to 7 krpm (2.4 MDN) at 7C. Bearing load capacity and torque decrease with temperature because of reduced bearing preload. Dykas (26): Investigates thermal management in foil thrust bearings. Cooling flow rates, to 45 L/min, increase bearing load capacity at high rotor speeds. Inadequate thermal management can give thermo-elastic distortions affecting load capacity of test FB Radil et al (27): Evaluate effectiveness of three cooling methods (axial cooling, direct and indirect shaft cooling) for thermal management in a hot GFB environment 79

80 Overview Thermal management System-level tests LaRue et al (26): Oil-free Turbocharger. Thermal management achieved by cooling the TC rotor and FBs. Lubell et al (26): Commercial oil-free micro-turbines. Cooling air flows axially through hollow rotor ID remarkably decrease rotor temperature. Heshmat et al (25): Demonstrates hot (65C) GFB operation in a turbojet engine to 6 krpm. Cooling flow rates to 57 L/min still give large axial thermal gradients (13ºC/cm) San Andrés et al (29): Forced cooling flow has limited effectiveness at low rotor temperatures. At high test temperatures, large cooling flows (turbulent) remove heat more efficiently. Gases have limited thermal capacity, hence (some) bearings demand large cooling flows to remove 8 heat from hot rotor sections.

81 Why thermal effects are important? Gas bearings (when airborne) are nearly friction free, hence the show small (drag) power loss and temperature raise. With hot rotors the lubricant in the bearings must also cool components. But gases have small thermal capacity and conductivity, and hence, get hot! Rises in temperature change material properties (solids and gas), and most importantly, change bearing clearance! 81

82 Lesson from previous demonstration 12/29: HT GFB test Temperature [ºC] Heater off Ths=2ºC Ths=4ºC Ths=6ºC FE rotor (TrFE) DE rotor (TrDE) Heater (Th) Heater (Th) Free end rotor OD (Tr FE) Drive end rotor OD (Tr DE) Time [min] Heater off Ths=2ºC Ths=4ºC Ths=6ºC 12 RBS failure NO COOLING FLOW!! Pressure Damaged foil bearing 1 T2 T1 T4 Temperature [ºC] Time [min] T3 T1 T2 T3 T4 RBS failure 82

83 Test Gas Foil Bearing 21: GFBs donated by KIST Bumps Top foil Test Gas Foil Bearing (Bump-Type) 1 st Generation. Diameter: mm Foil material: Inconel X-75 Reference: DellaCorte (2) Rule of Thumb UNCOATED TOP Foil! Hollow rotor (Inconel 718): 1.36 kg. Length: 2.66 mm. OD mm and ID 17.9 mm. HT Coating up to 4C

84 Foil Bearing Dimension KIST FB uncoated (Gen. I) Bearing cartridge Outer diameter Inner diameter Top foil and bump strip layer Top foil axial length Top foil thickness Bump foil thickness Number of Bumps Bump pitch, S o 4.4 Bump length, l B 2.5 Bump height, h B.5 Bump arc radius, r B 2.25 Bump arc angle [deg] Foil material: Inconel X-75 Parameter [mm] axial 67 Top foil Bump strip layer s O l B α r B Bearing sleeve 84 h B Ruler:.5 mm Each graduation

85 FB deflection versus static load Room temperature tests: Estimation of bearing clearance Load deflection tests 15 1 (Cyclic loading & unloading) 5 45 bearing orientation Bearing displacement [um] DE bearing FE bearing FB diametrical clearance ~ 2 µm Static load [N] F K X Nonlinear F(X) Hysteresis loop : Mechanical energy dissipation due to dry-friction between top foil contacting bumps and bump strip layers contacting bearing cartridge 85

86 Thermocouples in test GFB Thermocouple (Bearing mid-span) Oblique view 5 mm 3 mm 45º T1 T4 g T3 Th Free End (FE) GFBs Drive End (DE) T6 g T9 T8 45º T2 Free end (FE) bearing Tr FE T1~T4 T6~T9 Tr DE T7 Drive end (DE) bearing KIST FB uncoated (Generation I) Four (4) thermocouples placed within machined axial slots. 86

87 Thermocouples in bearing housing Duct Thermocouple (T d ) Rotor free end (FE) Rotor drive end (DE) Bearing housing T out Bearing support housing Thermocouple (T 5 ) Thermocouple (T 1 ) Rotor (free end) Heater Forced cooling stream temperature Gas foil bearing (Rotor free end) 1 in housing duct + 1 at outboard plane of free end bearing 87

88 Hot rotor-gfb test rig Cartridge heater 254 Free end GFB T d Drive end GFB Hollow rotor Duct Side view Hollow rotor T h Duct Cartridge heater GFB support housing Dimension [mm] Front view Dimensions 88

89 Hot rotor-gfb test rig FB Cooling air supply Valve 2 Pressure gauge 1 Thermocouples (K-type) T in Mass flow meter 28 V AC ~ Heater temperature controller Infrared thermometer Pressure gauge 2 Infrared thermometer FB Cooling flow path Pressure gauge 3 Cooling air supply to coupling Valve 3 Valve 1 Shop compressed air (up to 1 MPa) Cartridge heater Drive motor Motor controller Displacement sensors Foil bearings ~ 46 V AC Signal conditioner Oscilloscopes FFT Analyzer Instrumentation Data acquisition board PC Tachometer Temperature digital panel meter Thermocouples: 1 x heater, 3 x cooling air 2 x 4 FB outboard, 2 x Bearing housing 2x Drive motor, 1 x ambient + infrared thermometers 2 x rotor surface 89 (Total = 19)

90 Test Cases Test case # Heater set temperature [ºC] Rotor speed [ krpm] Set cooling flow rate (into two bearings) [L/min] Time [min] Off Overall 149 min 9

91 Rotor OD Temps. vs Time Temperature rise [ºC] Th No rotor spinning Time [min] The rotor is a source of HEAT! Free End (FE) Tr FE T1~T4 GFBs Drive End (DE) T6~T9 Tr DE Tr FE Tr DE Temperature rise [ºC] Thermal gradient Hot to cold FE rotor >DE rotor Cooling flow rate [L/min] Test cases #1 and #4 Heater set temperature = 15ºC Temperatures on shaft OD increase steadily with elapsed test time at each rotor speed and cooling flow rate condition 1 krpm Tr FE Tr DE Time [min] Cooling flow rate [L/min]

92 Duct & Outboard temperature rises vs time Temperature rise [ºC] Th T out T d T in Time [min] Tout Tr FE No rotor spinning Td Duct Bearing housing Tr DE Temperature rise [ºC] Cooling flow rate [L/min] Test cases #1 and #4 Heater set temperature = 15ºC T d Tin ~ Tamb Td > Tout > Tin T in Time [min] 1 krpm T out Cooling flow rate [L/min]

93 FE bearing temperature rise vs time Temperature rise [ºC] No rotor spinning Time [min] T3>T2>T4>T1 due to differences in rotor OD temp. along its circumference Rotor speed makes bearings slightly hotter (a few deg C) T 3 T T T 2 Temperature rise [ºC] Cooling flow rate [L/min] Test cases #1 and #4 Heater set temperature = 15ºC 1 krpm Free End Bearing Mid-plane T1 T4 T Time [min] T 1 T T T 2-85 T Cooling flow rate [L/min] 93

94 Bearing temperature rise vs duct temp. Bearing temperature rise [ºC] Th Duct air temperature rise [ºC] Tr FE T1~T4 Cooling flow decreases 23~28 L/min 35~42 L/min Td Duct Bearing housing No rotor spinning No cooling flow 5~65 L/min 15~17 L/min T6~T9 Tr DE Test cases #2 and #5 Heater set temperature = 1ºC Free end Bearing Temperatures on bearings ODs linearly increase with duct air temperature as the cooling flow rate into the bearings decreases Bearing temperature rise [ºC] ~8 L/min Cooling flow decreases 3 krpm 35~45 L/min 15~19 L/min 23~3 L/min Duct air temperature rise [ºC]

95 Rise in temperatures: Duct vs Rotor OD Td 3 No rotor spinning Test cases #2 and #5 Heater set temperature = 1ºC Duct air temperature rise [ºC] Cooling flow decreases No cooling flow 5~65 L/min 15~17 L/min 23~28 L/min Td decreases with cooling flow due to longer residence of air particles in the duct Td increases with rotor speed due to windage effect Th Td Duct 35~42 L/min Rotor OD temperature rise [ºC] Td Duct air temperature rise [ºC] ~8 L/min 3 krpm 35~45 L/min Cooling flow decreases 15~19 L/min 23~3 L/min Tr FE Bearing housing Tr DE Rotor OD temperature rise [ºC]

96 Bearing OD temperature rise vs. cooling flow Test cases #2 and #5 2 Heater set temperature = 1ºC Temperature rise [ºC] T 1~4 -T d No rotor spinning 1 krpm 2 krpm 3 krpm Th Tr FE T1~T4 Td Duct Bearing housing T6~T9 Tr DE Free end Bearing Temperature difference (T-Td) is invariant while increasing cooling flow rate! Cooling flow rate [L/min]? Bearing temperature is a small fraction of the heat source temperature 96 (heater and duct air).

97 Rotor OD temp. vs heater temp. (- duct???) No rotor spinning Temperature difference [ºC] L/min 25 L/min 15 L/min 5 L/min No cooling flow Tr FE -Td Tr DE -Td Test cases #1~#3 The cooling gas flow removes heat from the top foil back surface, thus cooling the rotor OD Temperature difference, (T h -T d ) [ºC] Th Td Duct Cartridge temperature (T hs ) increases Tr FE Bearing housing Tr97 DE

98 Rotor OD temp. vs heater temp. 3 krpm 8 Test cases #4 and #5 Temperature difference [ºC] Tr FE -Td Tr DE -Td ~35 L/min ~25 L/min ~15 L/min ~5 L/min Rotor OD temp (relative to duct temp Td) decrease with cooling flow Temperature difference, (T h -T d ) [ºC] Th Td Duct Cartridge temperature (T hs ) increases Tr FE Bearing housing Tr98 DE

99 Cooling Capability: Bearing OD temp. Test cases #2 and #5 [ºC/L/min] Temperature rise Cooling flow rate No rotor spinning 1 krpm 2 krpm 3 krpm ~ Constant Cooling flow rate increases Heater set temperature = 1ºC Free end Bearing Th The cooling capability of the forced axial flow on the bearing temperatures changes little with flow rate T1~T4 Td Duct T6~T9 Cooling flow rate [L/min] Tr FE Bearing housing 99 Tr DE

100 Cooling Capability: Rotor OD temp. [ºC/L/min] Temperature rise Cooling flow rate No rotor spinning 1 krpm 2 krpm 3 krpm Cooling flow rate [L/min] Cooling flow rate increases Test cases #2 and #5 Heater set temperature = 1ºC Th Free end rotor The cooling capability of the forced axial flow on the rotor temperatures appears to have an exponential decay character. The cooling effectiveness of the forced cooling stream is most distinct at the free end rotor OD. Td Duct Tr FE Bearing housing Tr 1 DE

101 Waterfalls of rotor motion Baseline, Ths=65C Drive End (Horizontal) FH DH GFBs g FV DV ~5 L/min ~35 L/min 1X 2X 3X 3 krpm ~5 L/min ~25 min 1X 2X 3X 2 krpm ~35 L/min ~5 L/min ~35 L/min 1X 2X 3X 1 krpm Cooling flow Test case #4 Free of sub synchronous whirl motions Cooling flow rate does not affect amplitude and frequency contents of rotordynamic displacement 11

102 Rotor motion measurements 12

103 Waterfalls of rotor motion Baseline, Ths=15C DH ~5 L/min 1X 2X 3X Drive End (Horizontal) FH DH GFBs FV DV g ~135 min 1 krpm ~35 L/min Cooling flow Test case #6 Rotor OD temperature does not affect rotor dynamic displacements! 13

104 Synchronous rotor response: effect of shaft temp. Test cases #7~#9 Ths=1ºC, 3 krpm, 35 L/min Free End Drive End Cartridge temperature (T hs ) increases Heater off, 3 krpm, 35 L/min Free End Drive End FH GFBs DH g TrFE=87º C TrDE=54º C TrFE=33º C TrDE=32º C FV DV Amplitude [μm, -pk] 6 DV Speed down (16.7 Hz/s) Heater off Ths=65ºC Ths=1ºC Amplitude [μm, -pk] DH Heater off Ths=65ºC Ths=1ºC Speed down (16.7 Hz/s) Speed [krpm] Speed [krpm] Flexible rotor mode at ~9 krpm (rap test) Critical speed (Rigid body mode) ~ 5 and 8 krpm Test cases #7~#9 14 No major differences in responses between cold and hot

105 GFB TEHD model By San Andrés and Kim (28) Gas film Top foil & underspring Bearing clearance Reynolds eqn. for hydrodynamic pressure generation Energy transport eqn. for mean flow temperature Various surface heat convection models Mixing of temperature at leading edge of top foil Thermo-elastic deformation eqns. Finite Elements and discrete parameter for bump strips. Thermal energy conduction paths to side cooling flow and bearing housing. Material properties (gas & foils) = f (Temperature) Shaft thermal and centrifugal growth Bearing thermal growth Excel GUI + executable licensed by TAMU 15

106 Recap: test rotor and FB Outer flow stream P Co,T Co Cooling stream x z Bearing housing Thin film flow Ω R So Bearing housing Bump layer Top foil P a Hot air (out) Heat source T h T ambient Hollow rotor z= z=l Natural convection on exposed surfaces of bearing OD and shaft ID Schematic view of rotor and heater cartridge + side cooling stream 16

107 Heat flow paths in rotor - GFB system T Ci S C Q Heat carried by inner flow stream Q Si T Si T So So Q S f Q f F Q Heat carried by thin film flow Q Ci Heat conduction through shaft R Si R So R Fi R Fo R Bi R Bo T S T f T Fi T Fo T O F Q F O Q F Bi Q Heat carried by outer flow stream Hollow shaft Ω Q T Bi B O Bi Q Top foil Q Co T Bo B Q Bump layer Q B T Bearing housing Q=Rq : Heat (+) External fluid Drag dissipation power (gas film) : Heat (-) dө Heat flows & thermal resistances in a GFB & hollow shaft Heat conducted into bearing Cooling gas streams carry away heat 17

108 Bearing temperature predictions & tests Free end FB Temperature [ºC] Test data static load ~ 5.94 N Predictions 3 krpm: Test 3krpm: Prediction 2 krpm: Test 2 krpm: Prediction 1 krpm: Test 1 krpm: Prediction Cooling flow rate per each bearing [L/min] Θ Top foil leading edge Top foil Top foil trailing edge 45º Hot shaft (isothermal) Load X Heat flux Shaft OD Bearing Shaft OD Cooling stream Ω Input: Measured ambient, rotor OD & inlet cooling flow temps. g Y Thin film flow Bump strip layer Bearing housing External fluid Predictions agree with test data!! 18

109 Bearing temperature predictions & tests Drive end FB Temperature [ºC] Test data Predictions static load ~ 7.39 N 3 krpm: Test 3krpm: Prediction 2 krpm: Test 2 krpm: Prediction 1 krpm: Test 1 krpm: Prediction Cooling flow rate per each bearing [L/min] Θ Top foil leading edge Top foil Top foil trailing edge 45º Hot shaft (isothermal) Load X Heat flux Shaft OD Bearing Shaft OD Cooling stream Ω g Y Thin film flow Bump strip layer Bearing housing External fluid Predictions follow test data: better at DEB since smaller 19 temperature gradient along heater axial length

110 Predictions: Temperature fields Test case #5 Ths=1C,3 krpm, Free end bearing Cooling flow 175 L/min Θ Top foil leading edge Top foil Top foil trailing edge 45º Load Heat flux Shaft OD Bearing Shaft OD Cooling stream Ω g Y External fluid Thin film flow Film temperature [ºC ] Axial coordinate Axial node number S1 S5 S9 S Shaft = 8C Cooling stream inlet = 33C Shear mechanical energy generated Circ coordinate [deg] Fresh gas enters Circumferential angle [deg] Hot shaft (isothermal) X Cooling flow 25 L/min Film temperature [ºC ] Axial node number S1 S5 S Bump strip layer Bearing housing Shear mechanical 4. energy 225 generated Circ coordinate [deg] Fresh gas enters Circumferential angle [deg] Shaft = 56C Cooling stream inlet = 41C 11

111 Bearing stiffnesses predictions Free End FB static load 5.94 N Kxx, 175 L/min Kxx, 125 L/min Kxy, 175 L/min Kxx, 75 L/min Kxx, 25 L/min Kxy, 75 L/min Kyy, 175 L/min Kyy, 125 L/min Kyx, 175 L/min Kyy, 75 L/min Kyy, 25 L/min Kyx, 75 L/min.6.6 Θ Top foil leading edge Top foil Top foil trailing edge 45º Hot shaft (isothermal) Load X Kxy, 125 L/min Kxy, 25 L/min Kyx, 125 L/min Kyx, 25 L/min Heat flux Shaft OD Bearing Shaft OD Cooling stream Ω g Y External fluid Thin film flow Bump strip layer Bearing housing Direct sitiffness [MN/m] K xx K yy Kxx > Kyy Cross-coupled sitiffness [MN/m] K xy K yx Rotor speed [krpm] Rotor speed [krpm] GFB rotorydnamic force coefficients do not change with 111 the strength of cooling flow rate

112 Bearing damping predictions Free End FB static load 5.94 N Θ Top foil leading edge Top foil Top foil trailing edge 45º Hot shaft (isothermal) Load X Heat flux Shaft OD Bearing Shaft OD Cooling stream Ω g Y External fluid Thin film flow Bump strip layer Bearing housing Cxx, 175 L/min Cxx, 75 L/min Cyy, 175 L/min Cyy, 75 L/min Cxx, 125 L/min Cxx, 25 L/min Cyy, 125 L/min Cyy, 25 L/min Cxy, 175 L/min Cxy, 75 L/min Cyx, 175 L/min Cyx, 75 L/min Cxy, 125 L/min Cxy, 25 L/min Cyx, 125 L/min Cyx, 25 L/min Direct damping [kn-s/m] C x C yy Cxx > Cyy Cross-coupled damping [kn-s/m] C yx C xy Rotor speed [krpm] Rotor speed [krpm] GFB rotorydnamic force coefficients do not change with 112 the strength of cooling flow rate.

113 FE Model of Test Rotor-Bearing System Shaft radius [ m] Shaft1 1 2 Coupling added mass and inertia 3 4 Flexible coupling Drive end Imbalance planes Foil bearings Foil bearing supports Free end Shaft Axial location [m] XLGFBTH predicts synchronous bearing force coefficients A linear rotordynamics software (XLTRC2 ) models test rotor GFBs 113 system and predicts the rotor synchronous responses

114 Damped Natural Frequency Map Test case #7, Heater off 3 Damped Eigenvalue Mode Shape Plot Natural frequency [krpm] st critical speed 2nd critical speed Conical mode Cylindrical mode Damped Eigenvalue Mode Shape Plot f= cpm d=.5915 zeta N=72 rpm f= cpm d=.468 zeta N=42 rpm Rotor speed [krpm] 114

115 1X rotor response predictions & tests Forced cooling 175 L/min per bearing Free End (H) Amplitude [μm, -pk] Prediction Test cases #7~#9 Test data Speed [krpm] The predicted rotor responses reasonably correlate with 115 the measurements. FH FV GFBs DV DH g

116 Conclusions - GFB temperatures linearly increase with the inlet cooling air temperature. - When the rotor spins, the bearing sleeve temperatures do not change with the cooling flow rate; albeit the rotor OD temperature increases with the strength of the cooling stream, -The cooling effect of the forced external flows is most distinct when the rotor is hottest and at the highest rotor speed. - Forced cooling flows do not affect the amplitude and frequency contents of the rotor motions. The test system (rigid-mode) critical speeds and modal damping ratio remain nearly invariant for increasing the rotor temperature and cooling flow strength. 116

117 Conclusions -A physics-based computational THD model predicts accurately measured FB OD temperatures for increasing shaft temperatures with cooling flow - Rotordynamic analysis integrating predicted FB force coefficients reproduces recorded rotor dynamic responses with increasing cooling flow rate and shaft temperature. Predictive tool validated & benchmarked to reliable test data base!!! 117

118 Major contribution The present work provides the most complete to date measurements of GFB temperatures and rotordynamic response thereby extending the GFB knowledge database. Comprehensive experiments and benchmarking of predictive tool serve to advance GFB applications for use into high temperature microturbomachinery. 118

119 Topic Statement of Work & Sources for Presentation Objectives and accomplished work in 7-8 Computational model. Validation with published data. Rotordynamic measurements at TAMU Objectives and accomplished work in 28-9 Description of test rig and foil bearings at TAMU Effect of temperature on bearing temperatures, coastdown speed and rotor motions Effect of cooling flow on bearing and shaft temperatures. Validation of computational model * The computational code Graphical User Interface. Further predictions Current work with MiTi Bearings Added value & closure 119

120 NSF-Research Undergraduate Experience in Microturbomachinery & Manufacturing To conduct hands-on training and research in mechanical, manufacturing, industrial, or materials engineering topics related to technological advances in microturbomachinery To develop microturbines to enhance defense, homeland security, transportation, and aerospace applications. (1 students /year) x 3 y NSF (6-9) $ 259 k Added value to NASA Project 12

121 29 REU MTM Program Project #1 REU student KIST FB (1st generation) Experimental Identification of Structural Stiffness and Damping in a 1st Generation Gas Foil Bearing for Oil-free MTM Shane Muller Mechanical Engineering Calvin College Oil inlet Project #2 REU student Measurement of Drag Torque, Power Loss, Friction Coefficient, Temperature, and Wear in a Foil Bearing & Coated Rotor System Jose Camero Mechanical Engineering U. of Texas, San Antonio KIST FB Valve Open Valve Close Rotor starts Constant Rotor speed starts ~5 krpm Rotor stops Rotor stops 121

122 Closure: objectives accomplished - To develop a physics-based computational model of GFB including thermal effects -To develop a fully tested and experimentally verified design tool for predicting GFB performance - To measure the rotordynamic performance of a HOT rotor supported on GFBs - To quantify the effect of feed gas flow on cooling GFBs Predictive tool validated & benchmarked to reliable test data base!!! 122

123 Acknowledgments NASA GRC: Drs. S. Howard & Dr. C. DellaCorte Turbomachinery Research Consortium NSF REUP Capstone Turbine, Inc. MiTi, Foster-Miller KIST: Korea Inst. Science & Technology Honeywell Turbocharging Technologies Learn more at: 123

124 Back up slides 124

125 THD Model Validation Bearings at TAMU Parameter [mm] Foster-Miller (2 nd gen.) KIST (1st gen.) MiTi (2 nd gen.) Bearing cartridge Outer diameter Inner diameter Top foil and bump strip layer Top foil axial length Top foil thickness Bump foil thickness Number of Bumps 25 5 axial 26 1 axial 24 3 axial Bump pitch Bump length Bump height Bump arc radius Bump arc angle [deg] Elastic Modulus 214 GPa, Poisson ratio=

126 Static load test setup Thermocouples Eddy current sensor Lathe chuck Cartridge heater Test shaft Test bearing Strain gauge type load cell Steady static load (or unload) proportional to linear movement of lathe tool holder 126

127 High temperature rotor NO COST! Inconel 718 shaft : photos taken by manufacturer (KIST) prior to machining of threaded holes at rotor ends and coating shaft at bearing locations. KIST proprietary solid lubricant (4 ºC) 127