on the failure of a gas foil bearing: high temperature ......failure (bearing seizure) while...
TRANSCRIPT
-
11
ASME GT2013-94244
Luis San AndrésMast-Childs Professor, Fellow ASME
Texas A&M University
This material is based upon work supported by the TAMU Turbomachinery Research Consortium (TRC) and NASA GRC.
2013 ASME Turbo Expo Conference, June 3-7 , 2013, San Antonio, TX
On The Failure Of A Gas Foil Bearing: High Temperature
Operation Without Cooling FlowKeun Ryu
Assistant Professor
Hanyang University
-
2
To develop a detailed, physics-based computational model of gas-lubricated foil journal bearings including thermal effects to predict bearing performance.
GFB research at TAMU
Since 2003, supported by NSF, NASA, Capstone and TRC
Bearing cartridge
Rotor
Ω
Foil spot weld
Top foil
Bump strip layers
In support of oil-free systems reduce overall system weight, complexity (low footprint) increase system efficiency due to low drag power losses extend maintenance intervals.
Objective
-
3
Gas foil bearings (+/-)• Increase reliability & good load capacity (< 20 psi)• Dispense with oil lubrication• Reduced weight & number of parts• High and low temperatures • Tolerate high vibration and shock load.
• Load capacity: less than rolling or oil lubricated bearings
• Endurance: wear during start up & shut down (lift off speed)
• Thermal management for high temperature applications (gas turbines, turbochargers)
• Predictive models lack validation for GFB operation at HIGH TEMPERATURE
-
4
Year Topic2008-13 Metal Mesh Foil Bearings: construction, verification of lift off performance and load
capacity, identification of structural stiffness and damping coefficients, identification of rotordynamic force coefficients
2008-10 Performance at high temperatures, temperature and rotordynamic measurements. Extend nonlinear rotordynamic analysis
2007-09 Thermoelastohydrodynamic model for prediction of GFB static and dynamic forced performance at high temperatures
2005-07 Integration of Finite Element structure model for prediction of GFB static and dynamic forced performance Effect of feed pressure and preload (shims) on stability of FBS. Measurements of rotordynamic response.
2005-07 Rotordynamic measurements: instability vs. forced nonlinearity?
2005-06 Model for ultimate load capacity, Isothermal model for prediction of GFB static and dynamic forced performance
2004-09 Measurement of static load capacity, Identification of structural stiffness and damping coefficients. Ambient and high temperatures
TAMU research on foil bearings
-
5
Objectives & tasks
Temperature and rotordynamic measurements of a heated rotor supported on gas foil bearings without cooling flow
• Measure temperature of bearings and rotor and the motions of rotor for increasing shaft temperatures• Identify post-test condition of test rotor and bearings after sudden failure• Estimate bearing clearance and top foil temperature to until bearing seizure occurs• Compare the experimental results to predictions from an in-house computational program
GT2013-94244
-
6
Thermal effects in GFB SystemsGas bearings (when airborne) are nearly friction
free, with small drag power and temperature rise.
However, the material structure of a foil bearing tends to heat up quickly since the lubricant, a gas, has low density and low thermal conductivity.
Rises in temperature change material properties (solids and gas) and the bearing clearance…..
Applications demand large cooling flows to control thermal growth of components and to remove efficiently mechanical energy from the rotor mainly.
-
7
Cooling effectiveness
Heater set temperature = 100ºC
0
0.2
0.4
0.6
0.8
1
1.2
0 100 200 300 400 500Cooling flow rate [L/min]
No rotor spinning10 krpm20 krpm30 krpm
[ºC/L
/min
]Te
mpe
ratu
re ri
se C
oolin
g flo
w ra
te
T 1~4 -T amb
Cooling flow rate increases
TrFE
Te
TrDE
EnclosureTh
Bearing housing
T1~T4 T6~T9
Free end bearing
0
0.2
0.4
0.6
0.8
1
1.2
0 100 200 300 400 500Cooling flow rate [L/min]
No rotor spinning10 krpm20 krpm30 krpm
[ºC/L
/min
]Te
mpe
ratu
re ri
se C
oolin
g flo
w ra
te
Tr FE -T amb
Rotor free end
GT2012-68074Best Paper Award
Cooling flow rate increases
-Even a little flow can cool the bearings & rotor. Too large cooling flows stop being effective. -The cooling effect of the supplied flows is most distinct when the rotor is hottest and at the highest rotor speed.
-
8
No cooling flow?
The current experimental results show poor operating procedure and ignorance
on the predicted system behavior.
The following details the test procedure, temperature measurements, and
discusses the possible causes leading to the failure.
Inadequate thermal management caused a sudden system failure (bearing seizure) while operating at a rotor speed of
37 krpm with the shaft OD temperature at ~250 °C.
-
9
The test rig
-
10
Hot rotor-FB test rig
Drive motor (max. 50 krpm) Instrumentation for high temperature.
Significant temperature gradient along rotor axisHeat source warms (unevenly) rotor and its bearings
Drive motor
Infrared thermometersCooling air supply for coupling
Infrared tachometer
Flexible couplingGFB support housing
Cartridge heater
Fiberoptic displacement sensors
Hollow rotor
Heater reference temperature at 600°C
Reference thermocouple to control heater temperature
Test GFBs
-
11
Hot rotor-FB test rig Dimensions
44.4
5
62.23
19.69
153.
04
144.78
15.9
0
17.9
0
254
50
180
200.66
Hollow rotorCartridge Heater (1.6 kW)
FB support housing
Free endGFB
Drive endGFB
Enclosure
5
Front view
Te
Th
Hollow rotor (AISI 4140): 1.31 kgLength: 200.7 mm, OD 36.46 mm and ID 17.90 mm. NO coating on rotor surface
-
12
Hot rotor-FB test rig Instrumentation
Foil bearings
Cartridge heater Drive motor
Heater temperature
controller
~208 V AC
Temperature digital panel meter
Thermocouples (K-type)
Displacement sensors
FFT Analyzer
Oscilloscopes
Signal conditionerTachometer
Data acquisition board
PC
Infrared thermometer
Infrared thermometer
~460 V AC
Motor controller
Thermocouples: 1 x heater, 1 x Bearing housing enclosure
2 x 4 FB outboard, 2 x Bearing housing2x Drive motor, 1 x ambient
+ infrared thermometers 2 x rotor surface (Total = 17)
-
13
The test bearings
-
14
Test foil bearings
Top foil
Cartridge sheet
Bumps
FB nominal dimensions
Parameter [Dimension] Drive end Free endCartridge inner diameter [mm] 37.98 37.92Cartridge outer diameter [mm] 44.64 44.58Axial bearing length [mm] 25.40 25.40Number of bumps 24× 3 24× 3
Bearing radial (assembly) clearance [mm] 0.076 0.042
Generation II FBThree (axial) bump strip layers, each with 24 bumps.Patented solid lubricant coating (up to 800°F) on top foil surface.
Other data proprietary
-
15
Oblique view
Thermocouples in test FB
Four (4) thermocouples placed in machined axial slots.
uncoated (Gen II)
TrFET1~T4
Th GFBs
Free End (FE)
Drive End(DE)
TrDET6~T9
Thermocouple fixed at mid-span bearingFree end
T1
T2
T3
T4
Ω
Drive end
ΩT8
T7
T6
T9
-
16
Test Procedure
-
17
Test procedure
Rotor spins at 37 krpm
Th is set at 200°C, 400°C, 600°C(~60 minute intervals)
050
100150200250300350400450500550600650
0 20 40 60 80 100 120 140 160 180 200
Tem
pera
ture
[ºC
]
Time [min]
Heater (Th)
Heater(Th)
Heater off Ths=200ºC Ths=400ºC Ths=600ºC
Failure at ~190 minute of elapsed
test time
RBS failure
Heater reference temperature at 600°C
Reference thermocouple to control heater temperature
-
18
Temperature measurements
TrFE
Te
TrDE
EnclosureTh
Bearing housing
T1~T4 T6~T9
-
19
Rotor OD temperature
050
100150200250300350400450500550600650
0 20 40 60 80 100 120 140 160 180 200
Tem
pera
ture
[ºC
]
Time [min]
FE rotor (TrFE)DE rotor (TrDE)Heater (Th)
Heater off Ths=200ºC Ths=400ºC Ths=600ºC
Significant axial thermal gradient from the rotor free end towards its drive end
TrFE >> TrDE
Max. 251ºC
Free end rotor OD(Tr FE)
Drive end rotor OD (Tr DE)
Heater(Th)
RBS failure
-
20
Bearing OD temperature
0
20
40
60
80
100
120
0 20 40 60 80 100 120 140 160 180 200
Tem
pera
ture
[ºC
]
Time [min]
T1
T2
T3
T4
Heater off Ths=200ºC Ths=400ºC Ths=600ºC
FB sleeve OD temperatures are different depending on their circumferential location; a result related to the heater warming the shaft and bearings unevenly, even without rotor spinning.
Max. 114ºCT4
T2T1
T3 RBS failure
-
21
Steady state temperatures
TrFE
TrFE
TrFE
TrFE
T4
T4
T4
T4
Td
Td
Td
Td
T7T7
T7T7
TrDETrDE
TrDE
TrDE
0
50
100
150
200
250
300
1 2 3 4Heater set temperature, Th [°C]
Tem
pera
ture
[°C
]
Th=200°C Th=400°C Th=600°CHeater off
TrFE TrDET4 T9Te
TrFE
TrDET4 Te T9
TrFE
TrDET4Te
T9
TrFE
TrDETe
T9
* * *
* Steady state temperature (~60 minute heating)
T4
Transient temperature (~10 minute heating)
TrFE TrDET4 T9Td
Te
Large thermal
gradient in the rotor
81 ºC
127 ºC
Te and TrDE rise above the temperatures in the bearing sleeves (T4, T9) as the heater temperature increases
None of the temperature measurements gives any evidence of an impending failure!
-
22
Rotor motion measurements
FV
GFBs
DV
FH DH
g
-
23
Rotor motion components at its free end show no distinctive differences as the rotor temperature increases.
Waterfall of rotor response37 krpm
Free end (Vertical)
FV
GFBs
DV
FH DH
g
Time increases
Th=200ºC (56 min)
Th=400ºC (63 min.)
Th=600ºC (10 min.)
Heater off (58 min.)
1X 0.5X 0.25X
FE Test bearing failure occurs
FE rotor OD temp: ~250ºC
No cooling flow into bearings
-
24
Synchronous rotor response
Rotor temperature does not affect the size and orientation of the FE rotor orbits.
0
50
100
150
200
250
300
350
0 20 40 60 80 100 120 140 160 180 200Time [min]
Tem
pera
ture
[ºC
]
Test bearing failure occurs
Free end rotor OD(Tr FE)
Heater offTh=200ºC Th=400ºC
Th=600ºC
FV
GFBs
DV
FH DH
g
37 krpm No cooling flow into bearings
FE bearing temperatures difference is largest
along horizontal plane Produce uneven
thermal growth leading to bearing ellipcity!
-
25
Post-test conditions
-
26
Rotor OD surface conditions
HT on FE rotor OD renders a much darker color!
Rotor surface where the FE bearing is held shows considerable wear and noteworthy scratches.
Drive end bearing location Free end bearing location
Before operation
AFTER incident
NO protective coatings on Rotor surface
Smears of top foil protective
coating
-
27
Drive end foil bearing after incident
Minor wear marks on the inboard top foil
Wear marks
Wear marks
Bearing still functional!Outboard
Inboard
Outboard
Inboard
-
28
Free end foil bearing after incident
Protective coating evaporated, and distinct sections of the top foil melted, in particular at the location where the top foil contacts the bump foil crests
Galled coating
Ripple traces
Melted top foil
Melted top foil
Beyond repair!Melting temperature of the foil material is 1,430ºC!
Outboard
Inboard
Outboard
Inboard
-
29
Close up view of FE bearingMost metal melted right at the line contacts with the crest of the bump foils Bearing seizure!
Bump
Flexible top foil
Pressure Pressure
“Sag”
Large amount of energy generated, producing an increasing temperature, first melted the protective coating (max. 400°C) and later the top foil.
-
30
Thermal expansion of test rig components
-
31
Predicted thermal expansion
Thermal expansion of the rotor OD is more pronounced since it is hotter than the bearing sleeve
FE Rotor (AISI4140) THERMAL EXPANSION
20 50 80 110 140 170 200 230 2600
20
40
60
80
100
Rotor temperature [C]
Ther
mal
exp
ansi
on [u
m]
Heateroff
Th=200º C
Th=400º C
Th=600º C
Rotor temperature [ºC]
20 40 60 80 100 1200
2
4
6
8
10
Bearing temperature [C]
Ther
mal
exp
ansi
on [u
m]
Heateroff
Th=200º C
Th=400º C
Th=600º C
Bearing cartridge temperature [ºC]
FE Bearing sleeve (Inconel 718) THERMAL CONTRACTION
-
32
Estimated bearing clearance
Free end FB
-20
-10
0
10
20
30
40
50
20 120 220 320 420 520 620Bea
ring
radi
al c
lear
ance
[µm
]
Heater temperature (Th) [ºC]
35 59 85 111 138 164
28 42 56 70 84 98
FE rotor (underneath bearing) temperature [ºC]
FE bearing (mean) temperature [ºC]
Th=400ºC Th=600ºCTh=200ºC
c=42 µm at cold condition(21°C of bearing & rotor temperature)
FE bearing eventually loses its clearance, followed by sudden bearing seizure and system failure!
C=Cr-ST-SC-BTCr: radial bearing clearance (ambient)ST : rotor radial thermal expansionSC : rotor centrifugal growthBT : bearing sleeve thermal expansion
FE bearing clearance is nil or disappears (turning negative)!
-
33
Damage on inboard side of bearing?
Trapped air in the enclosure became a thermal sink causing FE bearing inboard temperature to be higher than that on outboard side of the bearing!
Melted top foil
Outboard
Inboard
TrFE TrDET4 T9Td
Te
Inboard side of the bearing sleeve shrunk more than outboard side or at the mid-span.
The enclosure was closed, not even permitting the natural
convection of hot air into the environment.
-
34
Temperature predictions versus test data
-
35
TAMU GFB TEHD model
San Andrés and Kim (2008)
Gas filmReynolds eqn. for hydrodynamic pressure generation Energy transport eqn. for mean flow temperatureVarious surface heat convection modelsMixing of temperature at leading edge of top foil
Top foil & underspring
Thermo-elastic deformation eqns.Finite Elements and discrete parameter for bump strips.Thermal energy conduction paths to side cooling flow and bearing housing.
Bearing clearance
Material properties (gas & foils) = f (Temperature)Shaft thermal and centrifugal growthBearing thermal growth
Software licensed by TAMU
-
36
Recap: test rotor and FB
Outer flow stream Top foil
Bearing housing “Bump” layerzx
PCo , TCo
Bearing housing
PaThin film flow
z=0 z=LT∞
Ω RSo
Heatsource
Th
Coolingstream
Hollow rotor
Hot air (out)
ambient
View of rotor , heater cartridge + side cooling stream
Natural convection on
exposed surfaces of
bearing OD and shaft ID
-
37
predictions & testsBearing temperature
20
40
60
80
100
120
20 40 60 80 100 120 140
Bea
ring
tem
pera
ture
[ºC
]
Rotor temperature [ºC]
FE bearing-TEST
DE bearing-TEST
FE bearing-PREDICT
DE bearing-PREDICT
Predictions (DE bearing) agree well with test data!!
X
Y
Ω
45º
Θ
Load
Discrepancies are due to large temperature
gradient along heater axial length FE -TEST
DE -TEST
FE - PREDICT
DE - PREDICT
Thermal mixing coefficient λ=0.65 37 krpm
-
38
Estimation of the temperature raise in the top foil
-
39Dry friction contact generated large amounts of energy that could not be convected quickly to the bearing sleeve or removed by the gas film since there was no forced cooling flow. Hence, the temperature increased very rapidly.
Estimation of top foil temperature)( Ω=℘ SOfriction RWµ
)( Ω=℘ SOfriction RWµ
Generated Mechanical power due to rubbing
( ) ( )Foilp Foil frictionFoild T
C M Ah T Tdt ∞ ∞
+ − =℘
Top foil temp. increase
Foil specific heat and mass
Contact area
Heat convection coefficient (air)
130 W (with 6 N and 37 krpm)
( ) 1
t tfrictionFoil t o
air
PT T e T e
Ahτ τ− −
∞ = + + −
τ= ( )P FoilC M
Ah∞
Solution is valid for TFoil < 1,430 oC (Tmelt)
-
40
Rapid growth of the foil temperature to reach its melting point in ~14 s.
Predicted top foil temperature
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 180
300
600
900
1200
1500
1800
time (s )
Top
foil
tem
pera
ture
rise
(C)
Tmelt
*
Top foil coating temperature limit
Rubbing begins
-
41
Unusual operating condition, w/o cooling flow and too large increment in rotor temperature (up to 250°C) led to the incident which destroyed one of the test bearings. Bearing clearance decreases notably as rotor temperature increases until seizure occurs. Upon contact between the rotor and top foil, dry-friction quickly generated vast amounts of energy that melted the protective coating and metal top foil. In spite of the loss of one foil bearing, the other system components (rotor, 2nd bearing and instrumentation) remain functional, showing little damage.
Conclusions ASME GT2013-94244
Predictive tool validated & benchmarked to reliable test data base.
-
42
Even small quantities of air could be effective to promote the evacuation of hot air. The qualitative assessment for GFB system thermal management requires considerable experience! A cautious predictive design and conservative operation of the GFB system, along with an adequate thermal control strategy, could have prevented the bearing seizure and the rotor bearing system failure
Recommendations ASME GT2013-94244
Both (a) engineering thermal management with adequate cooing scheme and (b) adequate assessment of thermal effects prior to actual operation are mandatory for high temperature operation.
-
43
AcknowledgmentsThanks to the support of• NASA GRC (2007-09)• TRC (2004-10)
Learn more http://rotorlab.tamu.edu
Questions (?)
On The Failure Of A Gas Foil Bearing: High Temperature Operation Without Cooling FlowSlide Number 2Gas foil bearings (+/-)Slide Number 4Objectives & tasksThermal effects in GFB SystemsSlide Number 7No cooling flow?The test rigHot rotor-FB test rigHot rotor-FB test rigHot rotor-FB test rigSlide Number 13Test foil bearingsSlide Number 15Test ProcedureSlide Number 17Temperature measurementsSlide Number 19Slide Number 20Slide Number 21Rotor motion measurementsSlide Number 23Synchronous rotor responsePost-test conditionsRotor OD surface conditionsDrive end foil bearing after incidentFree end foil bearing after incidentClose up view of FE bearingThermal expansion of test rig componentsPredicted thermal expansionSlide Number 32Damage on inboard side of bearing? Temperature predictions versus test dataTAMU GFB TEHD modelSlide Number 36Slide Number 37Estimation of the temperature raise in the top foilSlide Number 39Slide Number 40Slide Number 41Slide Number 42Slide Number 43