Table of Contents

FULL LOAD, FULL SPEED TEST OF

TURBOEXPANDER-COMPRESSOR WITH ACTIVE MAGNETIC BEARINGS

by
Frank Davis
Engineering Consultant
Doha, Qatar
Reza Agahi
Consultant
Irvine, California
and
Randy Wu
Senior Engineer
GE Oil & Gas
Rancho Dominguez, California

Frank Davis is an Engineering Consultant with ExxonMobil, Randy Chih-Chien Wu is a Senior Engineer with GE Oil & Gas
assigned to Ras Laffan LNG Company Ltd., in Doha, Qatar. He has Operations LLC-North America, in Rancho Dominguez, California.
more than 30 years of experience in the field of rotating machinery. He began his career with Rotoflow Corporation (now part of GE Oil
Mr. Davis specializes in project engineering for machinery & Gas) as Thermodynamic Design and Test Engineer for expander
applications, bid reviews, and machinery acceptance testing. After and compressor products. In his 19 years of working experience, he
20 years machinery engineering experience with Exxon and Mobil has been involved in various disciplines in designing and developing
Corporations, he was a member of the Project Task Force managing of expander and compressor products. His current responsibilities
the engineering and construction of four LNG plants in the State of are focusing on risk review and NPI activities as well as test and
Qatar over the past 10 years. He has developed specifications, commission operation of expander and compressor product.
completed bid reviews, and followed manufacturing, testing, and Mr. Wu received a B.S. degree (Physics and Mechanical
participated in the startup of more than 40 machinery trains including Engineering, 1976) from Chung Yuan University and an M.S. degree
65 MW gas turbine driven refrigeration compressors and gas (Mechanical Engineering, 1983) from the University of Nebraska at
expander/compressor units. Lincoln. He is a member of ASME.
Mr. Davis received a B.S. degree (Mechanical Engineering, 1966)
from New Jersey Institute of Technology and is a registered
Professional Engineer in the State of New Jersey. INTRODUCTION
The application of the turboexpander in natural gas processing
Reza Agahi is a Consultant, in Irvine, and the petrochemical industry had its beginning at a small gas
California. He recently retired after 33 years plant in Southwest Texas where Dr. Judson S. Swearingen installed
with GE Rotoflow where he last served as the first natural gas turboexpander (Swearingen, 1999).
Director of Marketing, Sales, and Turboexpander technology has developed considerably in the
Commercial Operations. Dr. Agahi has last 40 years. For example:
taught in universities in Southern California • Advances in fluid dynamics theory and computational fluid
and has authored more than 40 articles and dynamics have made it possible to design a turboexpander with
papers in system engineering and turboma- high isentropic efficiency and performance predictability;
chinery applications. He is the inventor and
coinventor of several GE Rotoflow Patents. • Progress in rotordynamics evaluation and modern finite element
Dr. Agahi received B.S. and M.S. degrees (Mechanical analysis capabilities have resulted in more reliable turbomachinery.
Engineering, 1968) from Tehran University, and an M.S. degree • Increase in demand and economies of scale have resulted in
(1974) and Ph. D. degree (Operations Research and Systems natural gas processing and petrochemical plants becoming larger
Engineering, 1977) from the University of Southern California. and larger (Figure 1) (Agahi, 2003).
81
 Table of Contents
82 PROCEEDINGS OF THE THIRTY-FIFTH TURBOMACHINERY SYMPOSIUM • 2006

compressor application should incorporate features to address

those deficiencies and that the turboexpander compressor should
be tested as thoroughly as possible prior to shipment.
Turboexpander compressors now employ magnetic radial and thrust
bearings as a solution to the major problem of oil contamination.
Thrust calculations and thrust balancing designs have improved and
the load can be measured with the active magnetic bearing system.
Inlet guide vane controls were updated by the addition of an electric
actuator but it was not certain that the control would be as accurate
as desired. The active magnetic bearing system, however, lacks
redundancy and presents the new concern of a component failure
resulting in a loss of levitation under loaded conditions. In order to
demonstrate the ruggedness of the design, an FLFST was planned in
series with the residue gas compressor for the same plant.
Figure 1. Turboexpander Flow Development Since 1960.
The turboexpander compressor package for this project used
Three large size turboexpander compressor (TEC) units with active 240 mm (9.45 inch) active magnetic bearings. Table 1 shows the
magnetic bearings (AMB) were installed in a gas plant in the early turboexpander compressor wheel characteristics. This package was
1990s (Agahi, et al., 1994). Because of some difficulties with the to be installed in a closed loop with other LNG train equipment such
operation of the inlet guide vanes (IGV) at the commissioning stage as a gas turbine (GT) driven residue gas compressor. The project GT
caused by hydrate formation, oil and gas companies and engineering- provides the power required to drive the residue gas compressor,
procurement companies (EPCs) were somewhat skeptical about AMB which in turn delivers the required flow and pressure to the
technology and its reliability. Furthermore, with more offshore turboexpander compressor to bring the latter up to the full speed,
platforms using turboexpander compressor units, ease of commissioning full load condition. Table 2 shows the design guaranteed conditions
and reliability were among the major concerns of end users. and Table 3 depicts the simulated conditions for the FLFST.
In order to address these concerns, the end users of turboexpander
compressor units began to demand testing of the mechanical and Table 1. Turboexpander Compressor Wheel Characteristics.
control designs of custom designed equipment, and full load, full
speed testing (FLFST) of turboexpander compressor units. The first
tests were carried out in the late 1990s (Bergmann, et al., 1996).
These FLFSTs used a hydrocarbon gas mixture that was intended to
simulate the actual process gas as closely as possible.
Full load, full speed testing of a turboexpander compressor
with hydrocarbon gas proved to be somewhat difficult and too
expensive. Furthermore, obtaining permits to conduct such tests is a
challenge for most test facilities. To circumvent some of these issues,
a mixture of nitrogen and helium was used in some turboexpander
compressor FLFSTs in early 2000. In this approach, a mixture of
nitrogen and helium formulated to simulate the molecular weight of
the actual process gas is used in a closed test loop.
Table 2. Turboexpander Compressor Design/Guaranteed
BACKGROUND Operating Conditions.
A major Middle East gas production company was contemplating the
development of six large liquefied natural gas (LNG) plants. The
designs were to be similar, all based on the use of turboexpander
compressor units for maximum production of LNG. The turboexpander
compressor power was estimated to be in the range of 9.0 to 10.0 MW,
assuming that some of the plants would have units in parallel operation
to limit the size of the expanders to modules already demonstrated.
The company had experience with similar plants and had identified
a large turboexpander compressor as a major contributor to plant
outages or production limitations. A thorough review of the problems
experienced in the existing plants was conducted with the responsible
operating and rotating equipment engineers. The objective of the
review was to identify specific design features of the turboexpander
compressor that required upgrading from previous designs.
Several specific features of the existing turboexpander compressor
were identified as responsible for the great majority of equipment
outages. In order of importance, these were:
• Dilution of lubricating oil by process gas and subsequent loss of
oil viscosity resulting in excessive vibration and bearing wear.
• Inadequate design of the thrust balancing provisions and poor
operation of the thrust balance mechanism resulting in thrust
bearing overload.
• Clamping of inlet guide vanes due to lack of controls resulting
in erratic unit performance.
The review group concluded that a new, large turboexpander
 Table of Contents
FULL LOAD, FULL SPEED TEST OF 83
TURBOEXPANDER-COMPRESSOR WITH ACTIVE MAGNETIC BEARINGS
Table 3. Turboexpander Compressor FLFS Operating Conditions. • Partial load operation
• Design/normal speed operation
• Normal shutdown
• Emergency shutdown (ESD) due to process upset
• Two coastdown auxiliary bearing landing tests
• Verification of critical functions such as ESD and process
shutdown system trips
THE CLOSED LOOP TEST SETUP
Figure 3 shows a schematic of the test setup and Figure 4 shows
the FLFST piping around the turboexpander compressor. The
turboexpander compressor unit was integrated into the residue gas
compressor (RGC) test loop. The project residue gas compressor
was driven by its dedicated GT and delivered test fluid to the
expander at 64 barg (928.2 psig) and 50ºC (122ºF). The expander
extracted energy from the gas stream by expanding to 20 barg
(290.1 psig) and cooling down to 221ºC (25.8ºF). The cold gas
from the expander discharge was then fed to the recompressor,
which used the expander power and boosted the test fluid to a
discharge pressure of 31 barg (449.6 psig) before returning it to the
RGC to repeat the cycle.

FULL LOAD FULL SPEED TEST OBJECTIVES

The main objectives of the FLFS test that were agreed to by the
customer and expander manufacturer are as follows:
• Verification of the mechanical integrity of the turboexpander
compressor with active magnetic bearings
• Verification of the control functions related to the inlet guide
vanes of the turboexpander compressor unit, automatic thrust
balancing, and the antisurge valve while operating at FLFS
• Identification and correction of any faults or defects of the
turboexpander compressor system and repeat of the FLFS test to
verify that the issues were indeed rectified
• The FLFST sequence is shown in Figure 2 Figure 3. Closed Loop Process and Instruments Diagram for Full
Load/Full Speed Test.

Figure 2. Full Load/Full Speed Test Sequence.

The following conditions were to be monitored and recorded:
• Startup
• Near trip speed operation
Full load operation, turboexpander compressor operates at
maximum continuous speed (MCS) Figure 4. Test Loop Around Turboexpander Compressor.
 Table of Contents
84 PROCEEDINGS OF THE THIRTY-FIFTH TURBOMACHINERY SYMPOSIUM • 2006

EQUIPMENT The TEC speed was ramped up at 10 percent increments until the
speed closely approached the trip speed of 12,400 rpm and remained
The major components and supporting equipment in this test
at that speed for 15 minutes. Then the turboexpander compressor
loop are listed below:
speed was reduced to a maximum continuous speed (MCS) of 11,813
• Residue gas compressor, centrifugal compressor driven by a GT rpm for two hours. The test loop equilibrium state was achieved by
slowly adjusting the recycle valves of the recompressor and residue
• Residue gas compressor aftercooler gas compressor. The turboexpander compressor speed was further
• Residue gas compressor auxiliary support system reduced to the normal speed of 11,250 rpm and operated at this speed
for another two hours. At the end of this test run, the turboexpander
• Test loop piping compressor speed was gradually increased to the shutdown speed
• Flow measuring devices and the turboexpander compressor tripped on high speed. The
• Pressure measuring devices turboexpander compressor rotor coasted down under normal
conditions, i.e., the active magnetic bearing system was in levitating
• Temperature measuring devices mode during coastdown. Table 4 shows a sample of the FLFST
• Pressure transducers, cabling, data acquisition/analysis equipment parameters that were monitored.
• Startup seal gas supply, instrumentation air, etc. Table 4. A Sample of Test Parameters Monitored/Recorded.
• Gas analyzer and reporting system
• Vibration recording and analyzing equipment compatible with
active magnetic bearing system
• Residue gas compressor inline inlet strainer (60 mesh)
• Noise meter
The auxiliary equipment and components were as follows:
• Turboexpander compressor package with control system,
including active magnetic bearing signal interlock with test facility
and startup seal gas supply
• Expander bypass, Joule Thompson (JT) valve
• Expander inlet quick shutoff valve The turboexpander compressor was restarted and the speed was
• Compressor surge control system and recycle valve increased to 10,550 rpm for about 15 minutes in order for the
system to reach thermal equilibrium. A special method was applied
• Check valve downstream of the recompressor to bypass the active magnetic bearing controller and completely
• Expander inline inlet strainer (60 mesh) disable the AMB amplifiers in order to activate delevitation.
• Compressor inline inlet strainer (20 mesh) INLET GUIDE VANE SENSITIVITY
The majority of the operating parameters such as flow rates, AND FLOW CONTROLLABILITY
pressures, temperatures, vibration, etc., were monitored and logged The turboexpander compressor flow is linearly proportional to
by the automatic data acquisition system. the opening of its inlet guide vanes except in small opening and
Considering the practical aspects of the FLFST, some design full open positions. By controlling the sensitivity of the inlet guide
parameters could not be simulated. Figure 5 shows process vanes to within 1 percent, i.e., deviation between process signal to
parameters that were different during the FLFST compared to the inlet guide vanes and feedback signal to actuator system, it could
normal site conditions. be demonstrated that the expander flow controllability is within 1
percent of the total flow. As the trended data in Figure 6 show, the
differences between the inlet guide vane input signals and the
corresponding feedback signals were mostly less than1.0 percent
from ramp up to the FLFST condition. It is interesting to note that
the injection of additional fluid to increase the test loop pressure
did not influence the inlet guide vane sensitivity or flow
controllability. The inlet guide vane sensitivity remained within
1.0 percent even when the expander inlet pressure was increased
to 64 barg (928.2 psig).

Figure 5. Comparison Between the Design and Test Parameters.

FLFST Operation
Before startup, the test loop was pressurized to 24 barg (348.1
psig). The test header pressure reached 52 barg (754.2 psig) after the
residue gas compressor developed a stable pressure. Flow was
admitted to the turboexpander compressor by opening the inlet guide
vanes and closing the JT valve simultaneously. The recompressor
antisurge valve was at the full open position at the startup. The
antisurge valve began closing to load up the recompressor when the
turboexpander compressor speed reached approximately 5000 rpm. Figure 6. IGV Sensitivity Analysis.
 Table of Contents
FULL LOAD, FULL SPEED TEST OF 85
TURBOEXPANDER-COMPRESSOR WITH ACTIVE MAGNETIC BEARINGS

AMB AUXILIARY BEARING LANDING TEST Table 5. Comparison of Radial and Axial Air Gap Before and After
Landing Tests.
Auxiliary bearings support the turboexpander compressor rotor
when it is not levitated. Another function of the auxiliary bearings
is to catch the rotor upon loss of the magnetic field resulting in
delevitation as the machine coasts down from full load and full
speed (FLFS) to a full stop. To demonstrate the functionality of the
auxiliary bearings and show their ability to support the rotor upon
delevitation at FLFS, landing tests were performed during tur-
boexpander compressor shop tests. To implement this test, the
turboexpander compressor speed was increased to 10,550 rpm, and
then the expander was delevitated by intervening with the active
magnetic bearing control system. The delevitation signal shut off
the expander inlet quick shutoff valve and opened the JT valve. The
active magnetic bearing controllers were bypassed by introducing
jumpers in the control cabinet. As a result, all radial and axial
The tear down and inspection showed that there were light
amplifiers were disabled, and the rotor landed on the auxiliary
touches on the compressor impeller blade tips. The rest of the
bearings and coasted down to a complete stop. For both landings,
rotor and its corresponding stator parts such as the expander
it took approximately 2.4 seconds from the rotor landing until the
wheel, shaft seals, sensor rings, thrust disk, and magnetic
quick shutoff valve shut off; it took 4.6 seconds for the turboex-
bearings were found to have no touch marks and were in
pander compressor to coast down to a complete stop; the rotor
excellent condition.
landed in the auxiliary bearings for a total of 7.0 seconds; and rotor
There were light marks on the ball bearing inner rings and
whirling stopped within 4.0 to 4.2 seconds. Figures 7 and 8 provide
landing sleeves in both radial and axial surfaces but the ball
detailed records of these tests.
bearings could roll freely.
ACTIVE MAGNETIC BEARING ROTOR VIBRATION
Before spinning the residue gas compressor for the FLFST,
fine tuning of the active magnetic bearings, clearance, and
tuning checks were carried out to ensure that the air gaps were
consistent with the design values, transfer functions were up to
date, and all the required securities were set correctly. The
bearing system was equipped with antivibration rejection and
automatic balancing system logic. The antivibration rejection
activation deactivation limits were set at 3400 and 4600 rpm,
respectively. The rotor first critical speed was estimated to be
approximately 37 Hz. At higher speeds the automatic balancing
system took over the control function. These two systems
ensured that the active magnetic bearing rotor always rotates
around its inertia center. Figure 9 shows the turboexpander
compressor rotor vibration throughout the course of the FLFST
including both landing tests. The higher rotor vibration was
Figure 7. First Landing Test. observed during ramp up, at a speed range between 6,000 to
9,000 rpm. The highest vibration reached 50 mm and was
mainly in the subsynchronous spectrum, from approximately
37 percent to 44 percent of the synchronous frequency, and
occurred only during the startup period. The vibration at this
level was considered normal compared to the alarm setting of
90 mm. The tangential velocity component, i.e., exit from
the inlet guide vanes, could cause swirling around the
turboexpander wheel and resonate at subsynchronous
frequencies. At FLFST conditions, subsynchronous displacements
were almost nonexistent and overall vibration levels were about
15 µm. The axial vibration was about 13 mm.

Figure 8. Second Landing Test.

Before and after the landing test, both the radial and axial air
gaps between the rotor and auxiliary bearings were measured,
compared, and contrasted (Table 5). The data showed that there
were no changes in gap dimensions after two landing tests. Figure 9. Rotor Total Displacement/Vibration.
 Table of Contents
86 PROCEEDINGS OF THE THIRTY-FIFTH TURBOMACHINERY SYMPOSIUM • 2006

The active magnetic bearing current chart, Figure 10, shows the
current for each bearing during the FLFST. The relatively flat curves
indicate that the rotor was quite stable at the FLFST conditions.

Figure 12. Axial Bearing Current.

Figure 10. AMB Current. EXPANDER INLET QUICK

SHUTOFF VALVE TIMING
The bearing temperature chart, Figure 11, shows that the bearing
coil temperatures were normal. The highest temperature of 70ºC The project quick shutoff valve was used during the FLFST.
(158ºF) was observed when the turboexpander compressor was Three trips were initiated. One trip was triggered by an overspeed
operated near the trip speed. The rotor length expansion was measured trip at the end of the four-hour mechanical running test, and two
at 2150 mm. This measurement shows that the relative distance of the trips were performed during landing tests. It took about 2.4
rotor and housing was increasing during the test (Figure 11). seconds to complete the valve closing process. This duration
included the signal transmission time, the valve actuator response
time, and the stem traveling time. Based on the factory bench test
report, the valve closing time was 0.602 seconds. Therefore, the
signal transmission time in the test loop control system took
approximately 1.8 seconds.
BEARING HOUSING, SEAL GAS, AND
VENT GAS TEMPERATURE MONITORING
The turboexpander compressor seal gas control system was not in
operation during the FLFST. The seal gas pressure was controlled
manually with a bypass valve. The bearing housing vent gas
temperature alarm was initially set at 55ºC (131ºF). This setting was
too low for the FLFST conditions and had to be revised. The active
magnetic bearing high temperature alarm was set at 110ºC (230ºF)
and shut down was set at 130ºC (266ºF). Therefore, the vent gas
temperature setting was revised to 95ºC (203ºF).
Figure 11. Shaft Extension.
TEAR DOWN INSPECTION
ROTOR VIBRATION DURING
After completion of the four-hour mechanical running and two
TURBOEXPANDER TRIP
landing tests, the turboexpander compressor was disassembled for
One of the FLFST criteria was to observe the rotor behavior inspection. All parts were in good condition. The auxiliary ball
during shutdown and coastdown to stop. Upon a request for a bearings were replaced with a new set despite being in good
turboexpander compressor trip, the normal sequence of the events condition. The turboexpander variable inlet guide vane assembly
is to close the quick shutoff valve and open the antisurge valve. was also removed and inspected. There were scratch marks on the
Following this procedure, during turboexpander shutdown and interfacing surfaces between the guide vanes and nozzle clamping
coastdown while the rotor was levitated, there were no rings as well as on the nozzle cover. These parts were sent to the
significant changes in vibration levels, unbalance, temperatures, or metallurgical laboratory for determination of the root cause. The
bearing currents. conclusion was that debris carried by the gas stream of the test loop
was trapped in the nozzle grooves and was dragged into the
ACTIVE MAGNETIC THRUST BEARING AND interfacing surfaces causing the scratch marks when the vanes
AUTOMATIC THRUST BALANCING SYSTEM moved. All scratches were removed and the surfaces were restored
The turboexpander automatic thrust balancing system operated before the inlet guide vanes were reassembled.
flawlessly in conjunction with the active magnetic bearing thrust FLFST RESULTS
bearing control system. There were no indications of axial thrust
biases throughout the FLFST. The interlock logic between the The turboexpander compressor operation during the FLFST was
active magnetic bearing and turboexpander automatic thrust without any major problems or unexpected events. Uninterrupted
balancing system was set such that when the axial current reached FLFST continued for four hours at 12.5 MW and 11,250 rpm. The
14 A, the turboexpander system would take action to open or close rotordynamics performance was stable and all dynamic parameters
the automatic thrust balancing valve to relieve the thrust load. The were within the design limits and as predicted. The normal trips at
action automatically stopped when the bearing thrust current was FLFS and ESD trips with landing at FLFS into auxiliary bearings
reduced to 12 A. During the FLFST, the valve opening varied from were carried out successfully. Rotor coastdown and vibration
50 percent to 60 percent. Figure 12 shows the axial bearing current levels for all trips were smooth and at safe levels. The inlet guide
during the FLFST. vane test for ease of operation under full pressure, FLFS, ramping
 Table of Contents
FULL LOAD, FULL SPEED TEST OF 87
TURBOEXPANDER-COMPRESSOR WITH ACTIVE MAGNETIC BEARINGS
up, and closing down conditions did not show any faults, or produced the same performance that was predicted at the design
indications of blow by or clamping problems. The response of the stage and that would have resulted from extrapolation of the open
inlet guide vanes to process signals and controllability were loop air test. Therefore one may conclude that the FLFST does not
demonstrated to be consistent with other controls in the plant and provide any additional information through the gas dynamics
hence could control plant flow with the desired precision. The performance tests.
automatic thrust balancing systems of the bearing and
turboexpander maintained the axial position of the rotor in the PRESENT CONDITION
desired center position during the various tests and operating The turboexpander compressor package was installed,
conditions. Gas dynamics performance test results were as commissioned, and put into normal operation in early 2006. The unit
expected and consistent with the predicted values. Expander has been in normal operation since then.
isentropic and compressor polytropic efficiencies were better than
the guaranteed values. CONCLUSIONS
The end user and EPC for a large LNG facility in the Middle
LESSONS LEARNED
East had specific requirements for the design, manufacturing, and
The FLFST requires detailed planning and coordination with the FLFST of a turboexpander compressor package with an active
various facilities that are involved. The expander vendor had an magnetic bearing system. The turboexpander vendor worked
earlier FLFST at a facility outside their group and this one was closely and diligently with them and incorporated all the special
within their group. It turned out that more planning and attention requirements that were consistent with the turboexpander com-
to details were necessary in the latter FLFST because each team pressor design. The FLFST was carried out in a loop where the
tended to leave something out anticipating that the other team residue gas compressor for the same project supplied the required
would pick it up. boost for this package. The FLFST was conducted successfully and
Turboexpander compressor units are normally skid-mounted with relatively few problems. All the specified FLFST criteria were
packages ready to be installed on a foundation. This package fulfilled satisfactorily. The turboexpander compressor package is
should be complete with all auxiliary and support systems before it in normal operation at the present time.
is installed for the FLFST. There were delays, confusion and
manual (in lieu of automatic) operation because some components REFERENCES
were not shipped to the test site. Aghai, R. , 2003, “Turboexpander Technology Evaluation and
The turboexpander compressor shutdown loops should be Application in Natural Gas Processing,” Proceedings of
dedicated and have absolute minimum response time to guarantee Eighty-Second Gas Processing Association Annual
the safety and security of the turboexpander and its processes. Convention, San Antonio, Texas.
The noise level for this turboexpander compressor was
estimated at 85 dBA with a noise jacket installed on the casings. Agahi, R., Ershaghi, B., Leonard, M., Bosen, W. and Brunet, M., 1994,
The turboexpander, bearing housing, and compressor did not have “The First High-Power Natural Gas Turboexpander/Compressor
a noise jacket during the FLFST and no background noise with Magnetic Bearings and Dry Face Seals,” Proceedings of the
correction was applied. Noise levels were measured at 115 dBA, 23rd Turbomachinery Symposium, Revolve Technologies Inc.,
some 20 percent higher than calculated/expected. This test helped Calgary, Alberta, Canada.
to highlight the need to review and revise formulas and algorithms Bergman, D. and Nijhuis, H., 1996, “Full Power, Full Pressure,
used for noise level estimation. Closed Loop Test of a Natural Gas Turboexpander,”
The integrity and ruggedness of the auxiliary bearings were Proceedings of the Twenty-Fifth Turbomachinery Symposium,
tested and demonstrated by multiple landing tests. Inspection of Turbomachinery Laboratory, Texas A&M University, College
rotor parts after landing tests showed that there was no damage. Station, Texas, pp. 89-94.
This could be considered as justification to delete landing tests that
normally are requested during open loop air tests. “Dr. Judson S. Swearingen,” 1999, ORBIT, 20, (4), pp. 68-69.
Gas dynamics performance tests of the expander and compressor
 Table of Contents
88 PROCEEDINGS OF THE THIRTY-FIFTH TURBOMACHINERY SYMPOSIUM • 2006