29th Turbomachinery Consortium Meeting

1
29th Turbomachinery Consortium Meeting
Dynamic Response of a Rotor-Air Bearing System
due to Base Induced Periodic Motions
Luis San Andrés Mast-Childs Professor Principal
Investigator
Yaying Niu Research Assistant
TRC-BC-1-09
2009 TRC Project GAS BEARINGS FOR OIL-FREE
TURBOMACHINERY
Start date Oct 1st, 2008
2
Micro Turbomachinery (lt 0.5 MW)
Turbo Compressor 100 krpm, 10 kW
Advantages

• Compact and fewer parts
• Portable and easily sized
• High energy density
• Lower pollutant emissions
• Low operation cost

http//www.hsturbo.de/en/produkte/turboverdichter.
html
Micro Turbo 500 krpm, 0.10.5 kW
Oil-free turbocharger 120 krpm, 110 kW
http//www.hsturbo.de/en/produkte/micro-turbo.html
http//www.miti.cc/new_products.html
3
Gas Bearings for MTM
Advantages
Metal Mesh Foil Bearing

• Small friction and power losses
• Less heat generation
• Simple configuration
• High rotating speed (DN valuegt4M)
• Operate at extreme temperatures

Issues

• Small damping
• Low load carrying capacity
• Prone to instability

GT 2009-59315
Gas Foil Bearing
Flexure Pivot Bearing
AIAA-2004-5720-984
GT 2004-53621
4
Gas Bearings for MTM
GT 2009-59199
2008 Intermittent base shock load excitations
Drop induced shocks 30 g. Full recovery within
0.1 sec.
Ps2.36 bar (ab)
Rotor motion amplitude increases largely.
Subsynchronous amplitudes larger than
synchronous. Excitation of system natural
frequency. NOT a rotordynamic instability!
5
2008-2009 Objectives
Evaluate the reliability of rotor-air bearing
systems to withstanding base load 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 the predictions to test
  results

6
2009 Gas bearing test rig
The rod merely pushes on the base plate!
7
Test rotor and gas bearings
Flexure Pivot Hybrid Bearings Improved
stability, no pivot wear.
Rotor

• 0.825 kg in weight
• 190 mm in length
• Location of sensors and bearings noted

Gas bearing
Clearances 42 mm, Preload 40. Web rotational
stiffness 62 Nm/rad. Test rig tilts by 10. NOT
Load-on-Pad (LOP) !
8
Shaker delivered accelerations
Rotor speed 34 krpm (567 Hz)
Due to electric motor
Shaker transfers impacts to the rig base! Super
harmonic frequencies are excited.
9
Rotor speed coast down tests
No base excitation
Ps 2.36 bar (ab)
Rotor coasts down from 35 krpm
No added imbalance
Slow roll compensated Synchronous response
Subsynchronous whirling starts beyond 30
krpm, fixed at system natural frequency 193 Hz
Pressure 2.36bar 3.72bar 5.08bar Natural
Freq 192Hz 217Hz 250Hz
10
Rotor speed coast down tests
Ps 2.36 bar (ab)
Shaker input frequency 12Hz

• Subsynchronous response
• 24 Hz (Harmonic of 12 Hz)
• Natural frequency 193 Hz

Synchronous Dominant! Excitation of system
natural frequency does NOT mean instability!
11
Fixed speed, increasing pressures
Shaker input frequency 12Hz Rotor speed 34 krpm
(567 Hz)
12Hz, 24Hz, 36hz, etc
NOT due to base motion!
Pressure increases
243Hz
215Hz
Offset by 0.01 mm
193Hz
Rotor response amplitude at the system
natural frequency decreases, as the feed pressure
increases.
12
Fixed pressure, increasing speeds
Shaker input frequency 12Hz Feed pressure 2.36
bar (ab)
12Hz, 24Hz, 36hz, etc
Speed increases
193Hz
180Hz
180Hz
Rotor response amplitude at the system
natural frequency increases, as the rotor speed
increases.
13
Fixed speed and pressure, increasing input
frequency
Rotor speed 34 krpm (567Hz) Feed pressure 2.36
bar (ab)
193Hz
Frequency increases
NOT due to base motion!
Same excitation magnitude for 6, 9, and 12 Hz
Rotor response amplitude at the system
natural frequency increases, as the input
frequency increases.
14
XLTRC2 prediction
FE rotor model
Shaker input frequency 12Hz Feed pressure 2.36
bar (ab)
Conical
System Natural Freq
XLTRC2
Cylindrical
Measured
Input acceleration in XLTRC2, simulate actual
acceleration.
XLTiltPadHGB
Prediction uses synchronous speed bearing force
coefficients. In actuality, gas bearing force
coefficients are frequency dependent!
15
XLTRC2 prediction
Shaker input frequency 12Hz Feed pressure 2.36
bar (ab) Rotor speed 34 krpm (567 Hz)
Input acceleration only on VERTICAL direction
Measured N.F. component
Predicted natural frequency component
Prediction frequency step 1.25 Hz.
XLTRC2 predicts absolute rotor motions! Measured
rotor response is relative to bearing housing.
16
Rigid rotor model prediction
Rotor 1st bending mode 1,917 Hz (115 krpm)
Equations of Motion
Shaker input frequency 12Hz Feed pressure 2.36
bar (ab)
Steady-State!
Absolute rotor response
Input acceleration in rigid rotor model,
VERTICAL direction only!
Relative rotor response
System Natural Frequency
Rigid rotor model
XLTRC2
Measured
System response equals to the superposition of
unique single frequency responses.
17
Rigid rotor model prediction
Shaker input frequency 12Hz Feed pressure 2.36
bar (ab) Rotor speed 34 krpm (567 Hz)
Measured N.F. component
Above the natural frequency, the system is
isolated!
Predicted natural frequency component
Relative rotor motion
Rigid rotor model predicts relative rotor
motions! The test rotor-bearing system shows good
isolation.
18
Conclusions

• The recorded rotor response contains the main
  input frequency (5-12 Hz) and its super
  harmonics, and the rotor-bearing system natural
  frequency.
• The motion amplitudes at the natural frequency
  are smaller than the components synchronous with
  rotor speed.
• The rotor motion amplitude at the system natural
  frequency increases as the gas bearing feed
  pressure (5.082.36bar) decreases, as the rotor
  speed (2634krpm) increases, and as the shaker
  input frequency (512 Hz) increases.
• Predicted rotor motion responses obtained from
  XLTRC2 and an analytical rigid rotor model show
  good correlation with the measurements. The
  system shows reliable isolation.

Reliability of rotor-air bearing system to
withstanding base load excitations demonstrated