IDENTIFICATION OF FORCE COEFFICIENTS IN A SQUEEZE FILM DAMPER WITH A MECHANICAL SEAL: LARGE CONTACT

1
IDENTIFICATION OF FORCE COEFFICIENTS IN A SQUEEZE
FILM DAMPER WITH A MECHANICAL SEAL LARGE CONTACT
FORCE
Paper C1-144 (Tuesday, Sept 9)
Luis San Andrés Mast-Childs Professor Texas AM
University
Adolfo Delgado Mechanical Engineer GE Global
Research Center
World Tribology Congress 2009
Kyoto, Japan
Supported by Turbomachinery Research Consortium
Honeywell Aerospace Corp.
2
SFD Operation Design Issues
In aircraft gas turbines and compressors, squeeze
film dampers aid to attenuate rotor vibrations
and to provide mechanical isolation. Too little
damping may not be enough to reduce
vibrations. Too much damping may lock damper
degrades system rotordynamic performance
In a SFD, the journal whirls but does not spin.
The lubricant film is squeezed due to rotor
motions, and fluid film (damping) forces are
generated as a function of the journal velocity.
3
SFD Operation Design Issues

• Damper performance depends on
• Geometry (L, D, c)
• Lubricant (density, viscosity)
• Supply pressure and through flow
• Sealing devices
• Operating speed (frequency)

• Flow regimes (laminar, superlaminar, turbulent)
• Type of lubricant cavitation
• gaseous or vapor
• air ingestion entrapment

4
End seals reduce through flow and aid to increase
damping coefficients Most sealing devices cannot
prevent air ingestion because of end gap
SFD end seal arrangements
5
Spring loaded contacting face seal (metal-metal).
Seal keeps lubricant for long periods of time
( No side leakage allowed) Forced performance
complicated by dry-friction at contact area
Flow in
Flow out
bearing
bearing
journal
film
Contact seal
Contact seal
SFD with mechanical seal
SFD mechanical seal
6
SFDs Relevant Past Work

• Della Pietra and Adilleta (2002) Comprehensive
  review of research conducted on SFDs over the
  last 40 years.
• Parameter identification in SFDs
• (1986) Roberts et al,
• (1990) Ellis et al. ,
• (1995) Miller et al.
• (1999) Diaz and San Andrés
• Sealed SFDs
• (1996) Levesley and Holmes
• (1999) De Santiago and San Andrés
• (2005) Kim and Lee
• (2006-2009) San Andrés and Delgado (SFD
  MECHANICAL SEAL)

2009, ASME GT2009-59175 (accepted for
Journal) 2009, ASME J. of Eng Gas Turbines
Power, 131 (Paper GT2008-50528) 2008, ASME J. of
Eng Gas Turbines Power, 130 (Paper
GT2007-27436) 2007, ASME J. of Eng Gas Turbines
Power, 129 (Paper GT 2006-90782) 2007, ASME J.
of Tribology, 129 (IJTC 2006-12041)
7
Schematic view of test rig
Sealed SFD TEST RIG
Bearing Assembly
SFD-mechanical seal test rig
8
SFD bearing design
L25.4 mm, D127 mm, c0.127 mm (5 mil)
Top view
9
IDENTIFICATION OF FORCE COEFFICIENTS IN A SQUEEZE
FILM DAMPER WITH A MECHANICAL SEAL
Clearance c 0.127 mm (5 mil) Diameter D 127 mm
(5 inch) Length L 25.4 mm (1 inch) ISO VG 2
oil
DRY FRICTION at contact surface
Depends on assembly force from wave spring
surfaces condition
Current tests- Larger contact force
Mechanical Seal
10
Objectives and Tasks

• SFD with mechanical end seal based on a
  commercial design
• Parameter identification structure, seal SFD
• Method allows simultaneous identification of seal
  friction force and squeeze film land damping and
  inertia force coefficients
• Evaluate seal effectiveness (any side leakage ?
  Prevents air ingestion and entrapment? )
• If closing force increases, quantify its effect
  in force coefficients of test system and squeeze
  film damper

11
Circular Centered Orbits
Lubricated SFD - ISO VG 2 Feed pressure 31
kPa Temperature (avg.) 24 0C Bearing orbit
amplitudes maintained (25 to 50 um) load
magnitude varies
Max. clearance 127 mm
130
65
25 mm
Flow restrictors f 2.8 mm
38 mm
Y Displacement mm
-65
-130
130
65
50 mm
-65
-130
X Displacement mm
Test bearing orbits at 60 Hz
12
Parameter Identification
ASME GT2008-50528
Equations of motion
SFD forces (No oil cavitation)
Mechanical Seal
13
Parameter Identification
ASME GT2008-50528
Non-linear input
Fx
x

Fd
u
Fy

y
Fd
v
Four-input/two-output system
Transfer functions
Frequency domain
14
Circular orbits
Parameter Identification
Dry friction force
3 independent tests at different orbit amplitudes
(same frequency)
Linear transfer function
SFD force coefficients
15
Lubricated SFD
Circular Centered orbit
3 independent tests (same frequency)
Seal Dry Friction Force (N)
ASME GT2007-24736
Excitation frequency (Hz)
1 San Andrés, L., and Delgado, A., 2007,
Squeeze Film Damper with a Mechanical End Seal
Experimental Force Coefficients, J. Eng. Gas
Turbine Power
Dry friction force 54 N
16
Circular Centered orbit
3 independent tests (same frequency)
Friction Force 96 N
Seal Dry Friction Force (N)
1
ASME GT2007-24736
Excitation frequency (Hz)
Larger contact force
1 San Andrés, L., and Delgado, A., 2007,
Squeeze Film Damper with a Mechanical End Seal
Experimental Force Coefficients, J. Eng. Gas
Turbine Power
Seal dry friction force
17
Linear transfer function
Real Part
Orbit amplitude 50 um
From impact tests Ksy 885 kN/m
R2 0.98
Re (Lyy) kN.s/m
Ksy w2Ms-yy
Classical theory predicts 2.1 kg (5 times
smaller)
SFD Added Mass 11.4 kg
Improved model
Excitation frequency (Hz)
San Andrés, L., and Delgado, A., 2009 ASME J.
Tribol. (under review)
SFD added mass (fluid inertia) coefficient
18
Linear transfer function
Imaginary Part
Orbit amplitude 50 um
wCSFDyy
Im (Lyy) kN.s/m
R20.99
CSFDyy 7.7 kNs/m (44 lbf-s/in)
Excitation frequency (Hz)
SFD damping coefficient (vs frequency)
19
Cs-yy
Dry friction force
97 N
ASME GT2007-24736
Damping coefficient kN.s/m
54 N
CSFDyy
Squeeze film damping independent of frequency,
increases with amplitude of orbit
Excitation frequency (Hz)
System squeeze film damping vs. frequency
20
CSFDyy
Similar squeeze film damping coefficients
independent of contact force
Improved predictions correlate well with data
EXPER. PRESENT
Damping coefficient kN.s/m
ASME GT2007-24736

Prediction
Orbit radius (mm)
San Andrés, L., and Delgado, A., 2009 ASME J.
Tribol. (under review)
Squeeze film damping vs. orbit radius
21
Conclusions

• Mechanical end seal effective in preventing air
  ingestion. Test system damping coefficients with
  larger contact force are two times larger than
  original. Squeeze film damping coefficients
  remain the same.
• The identification method distinguishes dry
  friction force from squeeze film forces.
• Classical theory predicts SFD added mass
  coefficients 5 times smaller than in experiments.
  Improved added mass coefficient predictions
  correlate best with experiments.
• Predicted damping coefficient agrees well with
  test derived SFD coefficients for small to
  moderate size orbit amplitudes.

22
Thanks to TAMU Turbomachinery Research Consortium
Honeywell Aerospace Corp.

• Questions ?

To learn more, visit
http//phn.tamu.edu/TRIBGroup