Professional Documents
Culture Documents
W. K. Lord
Pratt & Whitney
East Hartford, CT
Fluids 2000
19-22 June 2000 / Denver, CO
AIAA 2000-2234
*
Development Engineer, Senior Member, AIAA.
1
American Institute of Aeronautics and Astronautics
RDT&E
11.2%
Operation
And
Support
20.6%
Procurement
Maintenance 44.4%
20.5%
Figure 1. F119-PW-100 engine (used in F-22 Advanced Tactical Fighter). 3-stage fan, POL 3.3%
6-stage compressor, single stage LPT and HPT, thrust vectoring nozzle. Military Fighter LCC
Crew
21.5% Depreciation
21.2%
Maintenance
20.5%
POL
33.6%
Reliability
Noise
Emissions
Low-observability (IR and radar cross-section)
Primary Propulsion airframe integration (PAI)
nozzle Trade factors can be established between different
metrics, these depend on the engine / airframe
combination, mission (e.g. short vs. long range), price
Fan nozzle of fuel, and so forth. Defining a single number as
typical is therefore not possible; for weight, for
Figure 4. PW 4168 (100" fan) showing the nacelle for a example, it can range from less than $500 of initial
separate flow configuration. engine cost per pound to as high as $3000.
2
American Institute of Aeronautics and Astronautics
technologies to be considered. Because the utilization performance with existing technology. A plot of TSFC
of a military engine is much lower than for a vs. year of entry into service for commercial engines
commercial engine, the fuel consumption is much less would show a similar trend; an initial rapid
critical than other factors, as shown in Figure 3. improvement in early development of gas turbine
Weight is an issue for both military and commercial engines, with a more recent tendency towards more
engines. Note that an extra pound of engine weight incremental improvements. Flow control is one set of
requires extra wing area for added lift, extra fuel, and possible technologies that may alter the game and
thus ultimately translates into much more than the enable us to provide significantly greater value to gas
original pound in extra total system weight. turbine engine customers.
8
Noise and emissions are increasingly becoming
important differentiators in the commercial engine 7.5
3
American Institute of Aeronautics and Astronautics
The FADEC has a number of additional control issue here than in the turbine because the compressor is
functions. In the compression section of the engine, producing a pressure rise. The highest performance is
both bleed flow extraction and variable-stagger stator typically reached close to the stall boundary, but
vanes are used to provide adequate stall margin at each adequate stall margin must be maintained at all
operating condition. Schedules for actuation of bleed conditions. Aeroelastic behavior such as flutter is also
valves and the unison rings that drive the variable stator a design consideration, particularly for the fan. The
vanes are developed based on both steady-state engine flutter margin requirement can set blade aspect ratio in
operating parameters and transient information such as todays wide-chord shroudless fans. High cycle fatigue
rate of change of RPM. In the turbine section, turbine (HCF) in the compressor, generally generated by
case cooling (TCC) is applied at cruise to obtain tighter upstream stator row wakes, can lead to premature blade
clearances for higher efficiency. These control actions failure if resonant stresses are not correctly predicted,
for compressor stability and TCC air are essentially resulting in undesirable maintenance and availability
open-loop controls. That is, in todays engines we do impact. Compressor stall is currently avoided through
not actually measure compressor operating line or scheduled bleeds and variable-stagger stator vanes,
turbine tip clearances and feed that information back to together with a stall margin that allows for effects of
the controller. inlet distortion and tip clearance deterioration. Military
engines also have variable geometry inlet guide vanes
In military fighter engines, control of the variable- (IGVs) in front of the fan to enable optimum behavior
geometry exhaust nozzle is an important function. at multiple points in the operating map.
Control of the convergent section and throat area must
be coordinated with not only the engine operating One of the first flow control applications to receive
condition but also the augmentor operation. In the case considerable attention in gas turbine applications is
of thrust-vectoring nozzles, control of the divergent active compressor stability control. If the compressor
flaps that set the thrust vector direction has to be operated closer to the stall boundary, then additional
integrated with the vehicle flight control system. work could be obtained per stage, and hence fewer
stages would be required. The current approach avoids
Future engine systems that may incorporate flow stall by scheduled bleed and stator vane actuation. This
control technology will likely require distributed schedule is based on engine tests, and must therefore
controls, as opposed to a single centralized FADEC. allow stall margin to account for engine to engine
Because much of the engine environment is at high variability, wear (primarily tip clearances), thermal
temperature, compact high-temperature electronics transients, and disturbances such as pressure or thermal
technology will be required to enable distributed distortion. This margin could be reduced if feedback
controls. With the development of compact high- were used; this requires some precursor to detect the
temperature sensors, controls, and actuators already imminent onset of stall. Furthermore, if one could use
underway14,15,16 it may be possible to envision micro- feedback to stabilize the unstable dynamics, one could
adaptive flow control for engine applications on the potentially operate past the stall boundary, resulting in
scale of a single airfoil or an individual fuel nozzle. In even greater work per stage. There are considerable
the following sections, some possibilities for references for this subject, see for example Reference
application of flow control technology in gas turbine 17 and the references therein. Actuation approaches
components are outlined. include bleed18 and air injection near the blade tips19 .
Note that in contrast to many of the other applications
FAN AND COMPRESSOR discussed, these flow control approaches treat the
The compressor adds energy to the flow, achieving dynamics of the overall compressor system, rather than
pressure ratios that can exceed 40:1 in current the dynamics on the scale of a shear layer.
commercial engines, with expectations of increasing
this further in the future so as to improve efficiency. Cost, weight, and length of the engine are driven to a
Military engines tend towards lower pressure ratio, due great extent by the number of stages of compression
to the optimization of thrust / weight, rather than TSFC. (see Figures 1 and 2). The number of stages is set by
Pressure rise per stage is typically less than 1.5, work per stage, which is limited by blade separation.
resulting in many stages of compression (as shown in Aspiration on the blade can be used to delay separation,
Figure 1, the F119 has 9 stages, while the PW4084 in with the added benefit of removing high entropy fluid 20 .
Figure 2 has 18, including the fan). Compressor Design of a single-stage fan at 3.5 pressure ratio with
efficiencies are on the order of 90 percent. The primary 6% bleed, 3% each from stator and rotor, has been
issues with the compressor and fan are twofold. First, described by Kerrebrock et al22 . A longer-term goal is
is the cost and weight associated with the number of to reach a 30:1 compression in 3 blade rows 21 . The
airfoils. Second is stability, or operability; more of an pressure ratios that are predicted to be achievable with
4
American Institute of Aeronautics and Astronautics
aspirated airfoils 22 are compared with those for current circulation control or jet flap) could result in weight
turbomachinery in Figure 6 and 7. Doubling the work savings, reduced cost, and a simpler subsystem. In and
per stage results in a dramatic reduction in the number of itself, this may not revolutionize the engine,
of stages, though not necessarily an exactly however, this is a clear example of an application that
proportional reduction in compressor weight due to the would use flow control if the technology was mature;
added complexity of blades and added plumbing. For there are many such opportunities within the engine.
fighter applications this technology looks very Because anti-ice air is required on IGVs to ensure
attractive due to its potential to improve thrust/weight operability, it may be possible to implement flow
of an engine. Clearly there are issues that need to be control for this application without additional
worked relative to how to use the resulting bleed air so plumbing.
that there is not too large a TSFC penalty, but the
potential to change the trend in Figure 5 is clear. Tip leakage flows resulting from loose clearances can
3.5
have a detrimental effect on both efficiency and stall
margin of a compressor. Maneuvers that load the rotor
can cause wear of the casing, as will thermal transients.
3
As the casing wears, the clearance gradually increases.
s
ign
2.5
d
ate
ta
Da
2 an
geF maneuvers, thus reducing the need for stall margin,
Sta
gle
Sin maintaining original efficiency, and reducing
1.5 maintenance costs. Control of the tip leakage flow
through flow control has been suggested, either through
1
a reduction of the effective tip gap (a virtual shape),
400 600 800 1000 1200
Tip Speed
1400 1600 1800 2000
or through enhanced mixing of the tip gap shear layer23 .
Thus, for example, an array of synthetic jets mounted in
Figure 6. Comparison of aspirated and conventional fan the casing and driven synchronized with the blade
designs; pressure rise vs. tip speed. Conventional fans passage frequency, could create a blockage that would
typically fall within the shaded region; the data for prevent the tip leakage flows that degrade efficiency
aspirated fans comes from Ref. 22. and stall margin, as shown in Figure 8. Similar
30
approaches could also be used to reduce seal leakage
MIT Aspirated Compressor goal: flows elsewhere in the engine 24 .
30 to 1 compression with 3 blade rows
25
20
Pressure Rise
15
rD
so
es
flow blocked by
pr
m
Co
10
synthetic jet output
5
0
0 5 10 15 Rotation
Number of Stages
5
American Institute of Aeronautics and Astronautics
highly loaded turbomachinery stages and more lightly and HPC). Because the low spool speed is limited by
damped integrally-bladed rotors (IBRs) mean that the tip speed of the fan (in 2-spool non-geared engines),
alternative solutions could be useful. The primary the LPC requires a high offset to keep tip speeds up and
source of HCF is typically the upstream stator wake on retain efficiency and reasonable work per stage. The
the rotor. This provides a harmonic excitation at N intermediate case must turn the flow radially inward,
times the rotation speed, where N is the number of and is therefore limited by separation, indeed there may
stators. If the speed of the engine lines up with a be some small regions of separated flow in parts of the
resonant mode of the rotor, significant stresses can operating envelope of military engines. The Directed
result, ultimately leading to blade fatigue. Modifying Synthetic Jet3 or DSJ, shown in Figure 9, is one
the wake of the upstream stator stage is one way of possible separation control approach that may be
solving this problem6 . Note that in addition to filling in appropriate for this application. The DSJ enables
the wake, it is possible to selectively increase the wake complete reattachment of the flow through boundary
in order to be orthogonal to the vibratory modes25 . layer energization; both the instroke and the outstroke
of the synthetic jet contribute to improving the
A similar problem from a technical perspective is the boundary layer profile. Applying separation control to
noise generated by the fan in a high bypass ratio engine. the intermediate case could enable more aggressive
One major source of fan noise is rotor/stator interaction turning, similar to the turbine transition duct shown in
noise, caused by wakes and turbulence from the fan Figure 12. This would allow either a shorter transition,
rotor impinging on the downstream fan exit stator. The or a higher offset, or some combination of these. This
periodic unsteadiness of the wake excitation produces could be traded either for weight and length reduction
unsteady lift and discrete tone noise at harmonics of of the IC, or to keep the LPC aft flowpath at higher
blade pass frequency. The turbulence interaction with radius for higher work and fewer stages.
the stator is a major source of fan broadband noise. Time-Averaged
One possible approach to reducing rotor/stator Mass and Momentum
interaction noise would be to apply trailing-edge Flow
6
American Institute of Aeronautics and Astronautics
The mixing zones of a gas turbine combustor are shown average exit temperature (see Figure 11) and/or
in Figure 10. The fuel/air ratio in the primary reducing the amount of turbine cooling air.
combustion zone of modern high temperature rise
Current
combustors is often designed to be rich, in order to
meet operability constraints, and hence the primary
air jets are needed to burn the remaining fuel. The Enhanced
Pattern Factor
transition from the rich primary zone to the lean Mixing
secondary combustion zone must pass through the
stoichiometric point, resulting in maximum flame Constraint set
temperatures, which is favorable for the production of by maximum
NO X. More rapid mixing here would significantly temperature
reduce emissions production. While NOx is produced Allowable Average
at the highest temperatures, a more important issue at Exit Temperature
low temperatures can be CO emissions. As the
Figure 11. Reduction in pattern factor enables higher
temperature increases, the CO reacts into CO2 . At low average exit temperature
power settings, therefore, the decreased temperature
results in less NOx being produced, and more CO. It Dilution air is used to reduce combustor exit gas
may therefore become beneficial to have less mixing, temperature, and the mixing of this air with the main
giving increased residence time at peak temperature so combustor flow helps determine the pattern factor.
as to react more CO; this impacts both emissions and Also note that this process must be accomplished in the
efficiency. Thus, in addition to technology to enhance shortest distance (in the flow direction) possible, so that
mixing, more generally, the combustor requires the length of the combustor can be minimized, thereby
controllable mixing. Mixing also affects operability of decreasing overall engine weight. As with emissions
the front-end of the combustor, since combustion production, technology to modify mixing behavior
cannot be sustained below some minimum equivalence could therefore enable significant improvements in
ratio if the fuel/air is uniformly mixed. Thus combustor design. Improved circumferential pattern
controllable mixing could also increase the achievable factor could also be enabled by control of the flow
ratio between maximum and minimum fuel/air. Flow through individual fuel nozzles. Closed loop feedback
control is being investigated to improve mixing of turbine temperature to low frequency fuel
behavior4 . For example, pulsed injection of air through modulation would reduce circumferential pattern factor,
the primary or dilution jets, rather than steady injection, for example, when inlet distortion has reduced the air
can increase jet penetration and alter mixing. Acoustic flow through a portion of the annulus.
excitation can also impact mixing behavior.
Considerable effort has also gone into the development
Current thrust performance of gas turbine aeroengines of control approaches for mitigating thermo-acoustic
is limited by the peak temperature exiting the instabilities at lean (and hence low NO X) conditions27 .
combustor into the turbine. Turbine cooling air The instability results from interaction between the
schemes must be designed to account for the peak gas combustion heat release, and the acoustic dynamics.
temperature as opposed to average temperature. In This unstable interaction can be controlled by pressure
extreme cases, unburned fuel can enter the turbine, with feedback; actuation could employ either modulation of
predictable results when contacting film cooling air the fuel with fast response valves (typically >100 Hz),
intended to protect the first vane row. The relationship or acoustic excitation. Military engines also frequently
between peak temperature and average exit temperature employ afterburners or augmentors, and this
is quantified by pattern factor: combustion process can also generate instabilities
T4MAX T3 (screech) leading to maintenance costs and design
PF = AVE constraints. Engines currently employ screech liners in
T4 T3 an effort to damp out these instabilities. These
where T4 is the gas temperature at the combustor exit instabilities could be controlled by modulating the fuel
and T3 is the gas temperature at the compressor exit. flow, or through shear flow actuation to directly
This pattern factor can be defined in both the influence the unstable feedback mechanism.
circumferential and radial directions, as well as overall.
High pattern factor corresponds to a non-uniform TURBINE
temperature exit distribution and can lead to reduced Undoubtedly the biggest challenge in turbine design is
turbine durability. If pattern factor can be reduced by associated with the temperature environment. As much
15%, it may be possible to increase engine thrust by as as 30% of the core flow is used for turbine cooling to
much as 20%, by operating the combustor at higher ensure durability. Maintaining reasonably constant
7
American Institute of Aeronautics and Astronautics
clearance is required for efficiency, and this is currently speed, yielding a 20% increase in work per stage, and
managed through scheduled thermal management of the thus would allow 20% fewer stages for a given turbine
case. High cycle fatigue, discussed previously in the work output requirement. Such flow control enabled
context of compression systems, is also an issue for benefits could lead to revolutionary changes in the
turbines, as is the need to increase stage loading in design of next generation turbines. The temperature
order to reduce overall engine weight and size. environment rules out some flow control approaches
such as suction, or many actuator technologies;
One of the challenges in working in the turbine area is however, steady blowing or high temperature actuated
the difficulty of sensing and actuating. For example, unsteady excitation are feasible.
the clearance control air flow is scheduled open loop. LPT
If sensors were available to monitor clearance, then entrance
LPT radius.
Increased
closed loop control could be used to maintain the
initially tight clearances as the engine wears, resulting
in as much as 2% efficiency improvements at the end of
the time on wing; longer maintenance intervals might
also be enabled. The cooling air flow currently has no
valve to allow even open loop control, but in principle, HPT exit
measurement of turbine inlet temperature (T4 ) and
temperature distribution would allow closed loop
control of at least part of the cooling air. This could Figure 12. A more aggressive turbine transition duct
therefore enable either reduced cooling flow geometry enabled by flow control would allow an increase
requirements, or higher fuel flow for the same turbine in low pressure turbine (LPT) radius.
life. A 30 error in T4 is estimated to result in a factor
0.0
of 2 difference in turbine life. Noting the substantial Expected
amount of core flow required for cooling the turbine, Performance
and the substantial maintenance cost associated with the
turbine, flow control technologies that improve the film
Change Of Efficiency
-2.0
In a commercial engine, the low pressure spool speed is
limited by fan tip speed. In order to obtain reasonable
low pressure turbine (LPT) work per stage, a high radial
offset is used; even so, there can be many LPT stages. -3.0
Aggressive transition ducts without flow separation 0.0 100,000 200,000 300,000 400,000
could reduce airfoil count and enable enhanced LPT Reynolds Number
performance, providing substantial reductions in weight
and overall propulsion system length. More aggressive Figure 13. Low pressure turbine efficiency. Performance
ducts permit increased radial offset between the HPT drops with Reynolds number.
and LPT, and a larger annulus area ratio across the duct, Low pressure turbine efficiency is a strong function of
as shown in Figure 12. The increased radius results in Reynolds number, as shown in Figure 13. Typical
more LPT wheel speed so that for the same optimum turbine efficiency drops by about 1 - 2 points between
loading parameter, tip speed squared over work, the sea level and altitude operation, and the problem can
number of stages can be reduced. Alternatively, a more become more severe for high altitude flight. The reason
aggressive transition duct could be traded for length. for this performance shift is due to laminar separation
The increased area ratio provides more airfoil on the turbine airfoil suction surface caused by the low
convergence (inlet flow area over exit flow area) in the Reynolds number encountered at altitude. More highly
front stages, which allows design of more effective loaded airfoils to increase stage loading and decrease
airfoil pressure distributions and improved turbine airfoil count were discussed earlier in the context of the
airfoil performance. The magnitude of these benefits, compressor, and this is certainly relevant for the turbine
which would be enabled by separation control, may be as well. Aspiration is clearly not desirable in the HPT,
roughly illustrated by imagining an aggressive where gas temperatures exceed the melting point of the
transition duct which provides a 10% increase in LPT airfoil material. Other means of separation control may
radius. This would result in a 10% increase in tip be possible 28 , although care must be taken not to
8
American Institute of Aeronautics and Astronautics
interfere with the film cooling. In the LPT, however, cross-winds and high-angle-of-attack operation. A
aspiration may be an option. common subsonic inlet design criterion provides for
separation-free inlet/diffuser performance up to 25
The compressor currently employs variable geometry angle of attack. A typical inlet lip is shown in Figure 4.
turning vanes to optimize the operating point, and avoid A thinner, more streamlined nacelle would reduce
stall. Variable geometry is challenging in the turbine, cruise drag. Thus, applying separation control during
and therefore using flow control is an attractive the take-off and climb portions of a transport mission
alternative approach to obtaining flow turning normally could enable a thin, cruise-optimized lip. Engine
associated with variable turning stators. The benefit to system weight could also be reduced, and reduced size
doing so is better accommodation of multiple turbine would improve propulsion/airframe integration.
operating points, widening of the airfoil loss bucket, Control of inlet distortion would also help reduce
and increasing the overall turbine operating range. A required stall margin of low-pressure-ratio turbofans.
turbine inlet guide vane would set the flow through the This could also enhance fan performance and
core, enabling a variable cycle and potentially replacing operability by providing improved take-off pressure
compressor stability bleeds. recovery, and reducing fan vibration.
NACELLE, INLET, and NOZZLE The development path for next-generation turbofan
The engine does not start and end with the engines leads in the direction of higher bypass ratio and
turbomachinery; the inlet and nozzle are essential parts lower fan pressure ratio. These features will provide
of the engine operation. The inlet, whether commercial substantial improvements to propulsive efficiency and
or military, must provide the fan with distortion-free air reductions in engine noise relative to current products.
with minimum pressure losses over a wide range of High bypass ratios require large engine diameters,
incidence angles and engine operating points. For resulting in potentially large contributions to overall
commercial engines, the inlet and the aft fan duct aircraft drag and increasingly adverse installation
provide noise reduction as well. For military engines, effects on the wing aerodynamics. These effects
the inlet may also be required to satisfy observability (among others) eventually offset the advantages of
constraints, while the military nozzle may be required higher bypass ratios. For modern high bypass ratio
not only to satisfy observability, but also to be variable (BPR~6) engines, the external nacelle drag can
in order to handle supersonic flight, afterburner contribute on the order of 3- 5% to the total aircraft
operation or to vector thrust. drag, and this fraction could be higher still for next-
generation turbofan engine nacelles (BPR > 10).
Buried propulsion system installations are common for Including internal nacelle drag increases the aircraft
military engines to reduce radar cross-section (RCS), drag fraction for current nacelles to as much as 8%.
and also for some configurations of the Blended Wing Strategies to reduce this can therefore have significant
Body (BWB) design29,30 . The S-duct inlet for such impact. While one might expect increased benefits
cases may be undesirably long to avoid significant from applying drag reduction strategies to the airframe
losses, or pressure distortions that could cause the and wing, the ready supply of high pressure air and the
engine to stall. Flow control such as the Directed self-contained nature of the nacelle subsystem makes
Synthetic Jet (DSJ) in Figure 9 could be used to enable the propulsion system an attractive place to first
a more aggressive inlet duct without the associated introduce these technologies.
performance penalties from separation, indeed vortex
generators are already in use to reduce inlet distortion Laminar flow control (LFC) has been demonstrated up
in some aircraft such as the F-18A/B. While inlet to the flight test level of technology development31,32 .
length is a significant propulsion/airframe integration Based upon current aircraft configurations and
challenge for manned vehicles, the inlet could become a commercial transport missions, net benefits of 1 2%
limiting factor for the entire vehicle size for some reduction of overall aircraft drag are predicted for
unmanned configurations. There are other approaches nacelle laminar flow control, which translate directly
one can use to avoid the strong radar return off of the into an equivalent reduction of specific fuel
spinning fan, however, these all involve some low- consumption. Natural laminar flow (NLF) systems
observables device in the front of the engine that will maintain extended regions of laminar flow due to
inevitably incur pressure loss. Flow control could be careful design and shaping of the forward nacelle
used to minimize the loss of such devices. contour. Hybrid systems employ both a natural laminar
flow region, as well as active suction within the adverse
The inlet lip thickness on todays subsonic turbofan pressure gradient region, to further delay transition.
engines is sized primarily for takeoff and low-speed Although progress has been made in the area of laminar
flight performance, where the inlet must handle large flow control, significant challenges remain. These
9
American Institute of Aeronautics and Astronautics
include the cost associated with the manufacture, nozzle flow areas between take-off and cruise would
installation, and maintenance of suction skins and the enable new, high-efficiency engine cycles. Fan surge
suction bleed source, and the sensitivity of laminar margin requirements presently limit the degree to which
boundary layers to small (order ~ 0.001 inch) surface low fan pressure ratio (and hence high propulsive
imperfections caused by insect or debris strikes, efficiency) engine cycles may be implemented. For a
manufacturing tolerances/defects, or natural joints next-generation, high-bypass ratio long range
produced in the nacelle construction. commercial turbofan (BPR > 10) it is estimated that up
to 20% fan nozzle area control would be required in
Methods of controlling the near-wall boundary layer order to enable fan operation and associated TSFC
flow beneath fully turbulent boundary layers provide engine cycle benefits on the order of 5% and higher.
alternatives to laminar flow control. Riblets have been For high bypass ratio engines, this represents a
studied for many years in much detail as a means to significant amount of fan flow needing to be controlled.
achieve skin friction drag reduction on propulsion Variable fan nozzle area could also provide aircraft trim
systems and aircraft wings. The interest level in these control to optimize engine performance during flight,
devices has been high due to their passive nature and and combined with trim vectoring of the fan nozzle
overall simplicity. Maximum riblet drag reduction thrust, could further optimize overall aircraft cruise
levels, based upon previous studies available in the drag during a transport mission. Finally, flow control
open literature, are in the 5 15% range33,34,35 . A study that enables the control of fan nozzle area can provide
by Lynch and Klinge11 points out some of the economic important acoustic benefits as well. Quiet engine
drawbacks associated with implementing riblets on cycles are associated with the resulting reduced fan
aircraft. Costs associated with the initial riblet pressure ratio (and hence reduced tip speed) and lower
installation by the manufacturer, combined with fan stream jet velocities that this capability provides.
maintenance/cleaning costs and the cost of re-installing
the riblet material approximately every 5 years, offset The nacelle on subsonic transports also typically has a
much if not all of the achievable riblet drag reduction thrust reverser, also in the fan duct region, which
benefit. diverts the fan stream radially outward and forward to
reverse thrust immediately after landing. A cascade-
Surface mass injection is another method for turbulent type thrust reverser consists of a movable cowl that
boundary layer drag reduction, both in terms of translates aft to expose a set of cascade vanes, and a
distributed normal mass injection as well as tangential blocker door that deploys simultaneously to force the
slot injection36,37 . Normal blowing with conventional fan air through the cascade rather than out the nozzle.
hole sizes suffers from the problem of large effective Target-type reversers deploy flow-impingement
roughness increases of the perforated surfaces. Normal hardware external to the nozzle. The engine flow hits
blowing through micro-perforate skins has been shown the target, and is directed outward and forwards. Pivot
recently to combine the effectiveness of normal mass door reversers perform in a similar manner, but deploy
injection for skin friction drag reduction with a porous flow-impingement hardware within the nacelle. These
surface possessing low effective roughness38 . This is systems work, but add substantial weight, particularly
achieved by using micro-sized holes of the proper when one considers what fraction of the total engine
dimension and shape. This concept has been life cycle that they operate. Thrust reversers can be
successfully demonstrated on a large-scale nacelle responsible for more than half the cost of the entire
model at flight Reynolds numbers to achieve skin nacelle subsystem. A blockerless system has been
friction reductions on the order of 50% over a portion proposed that uses fluidic actuation/flow blockage
of the nacelle 8 . Again, challenges remain regarding the instead of mechanical variable geometry.39 Note that
cost and weight of such a system, and engine cycle the efficiency of such a system is less critical for this
penalties associated with the required steady source of application because of when it is being used;
bleed air. Combining two or more flow control nonetheless, a substantial quantity of air must be
concepts that possess complementary features would diverted using air from the core stream, and hence this
help minimize such installation penalties. For example, remains a challenging application.
utilizing the flow removed from aspirated compressor
blades as a blowing source for micro-porous mass A final application to consider for the commercial
injection elsewhere in the engine would be an attractive engine is that of controlling jet noise. Engine noise has
alternative to scheduling bleed air for this function. decreased considerably over the last 30 years, however,
total aircraft traffic has increased, so that community
The aft end of the nacelle forms the nozzle for the fan noise remains an increasing problem. Primary engine
duct. In commercial and military transport noise sources include the fan, as discussed earlier, and
applications, flow control that allows for different fan jet noise. For current engines, both jet and fan noise are
10
American Institute of Aeronautics and Astronautics
roughly comparable sources of noise. The jet noise is mechanical complexity would therefore be extremely
created by turbulent mixing aft of the engine, and valuable 48,49,50,51 . Fluidic thrust vectoring is used in
driven primarily by the difference in jet velocity and rocket engines today. For gas-turbine engines, several
temperature between the engine exhaust and the fluidic mechanisms have been proposed for thrust
ambient air. For separate flow nacelles (where the fan vectoring, including shock-vector turning in the
and core streams are not mixed within the engine, e.g. divergent section, and throat-skewing, as shown in
as in Figure 4), there are two shear layers where noise Figure 14. Lockheed Martin has carried out
can be produced. Noise reductions to date have been computational and experimental studies of fluidic
obtained purely through reductions in jet velocity, nozzle throat area and thrust vector control49 , as shown
either by increasing bypass ratio (which also increases in Figure 15. This work demonstrated that effective
propulsive efficiency), or through enhanced mixing thrust vectoring could be accomplished through fluidic
using devices such as internal lobed mixers or mixer- skewing of the nozzle throat, a relatively low-loss,
ejectors40,41,42 . Internal lobed mixers and mixer-ejectors subsonic mechanism. In order to accomplish fluidic
have been successfully deployed on 727 and 737 thrust vectoring at realistic flight conditions, 1% of
aircraft during the 1990s to reduce jet noise to stage 3 injected flow for every 1.5 degrees of vector angle was
levels. Recent research, however, has created some required. This substantial bleed flow requirement
hope for reducing noise by modifying the dynamics of dictates the need for continued research in this area.
the mixing process itself. A discussion of the problem Note that while there are substantial aircraft system
and some potential flow control solutions can be found benefits, a useful fluidic system must not incur other
in Seiner43 . The relationship between mixing and the penalties such as excessive added base drag.
resulting noise is not yet well understood, and therefore Fluidic Throat Skewing Shock Vector Control
it isnt clear what control should do to the flow in order injector
injector
to improve noise44 . Several passive devices have
shown some promise, these include tabs45 , or an subsonic supersonic
shock
appropriately flexible wire trailing behind the core 46 , turning turning
11
American Institute of Aeronautics and Astronautics
Flow control could also be used for more rapid mixing changes in the overall engine design will be required to
of the exhaust jet, to enable more rapid reduction in the take advantage of this capability.
temperature 52 . For flight, this has obvious IR benefits.
The work in Reference 52 involves pulsed injection of In addition to the system issues, the gas turbine engine
compressor bleed air into the core exhaust to excite the provides a challenging environment within which to
jet flapping mode, resulting in much more rapid mixing. work. The primary issue is temperature. Gas-path
This technology is intended for deployment on the C-17 temperatures throughout the engine are shown in Figure
transport for use while the aircraft is on the ground. 16; this plot is already slightly dated in terms of
The improved mixing enables the ground crew to load temperatures reached in modern commercial engines,
the aircraft while the engine is running, for rapid turn- and military values are higher still. The case tends to
around, without having to resort to a heavy and costly be cooler, but can still reach 1000F. Actuation and
core thrust reverser system. Because this aircraft has a sensing must survive this harsh environment, must be
blown flap for high lift, temperature reduction can also extremely reliable, must not add excessive weight, and
reduce material costs required for protection from jet must not add excessive cost despite the relatively low
impingement. Finally, modification of the shape of the production volumes.
jet could increase the area of the flap assisted by engine
exhaust, thus improving the high lift capability of the
aircraft.
ISSUES
The potential for flow control to impact gas-turbine
engine design is clearly significant. However, for any
such technology to make it on to the product, system
trade studies are required. It is not sufficient to
improve a component, one must look at the overall
system and consider total value to customer, including
added cost and weight, additional complexity and
resulting maintenance or availability cost, and the risks
incurred by the engine program if the technology is not
thoroughly understood. Reliability is an absolute
requirement; for flight critical systems, failure rates of Figure 16. PW 4000 engine gas-path temperatures (from
10-9 per flight are required. Current commercial Ref. 53)
engines achieve in-flight shutdown rates better than
0.0026 in 1000 hours.
CONCLUDING REMARKS
For many of the flow control applications discussed, the Flow control has the potential to make a significant
maximum benefit is obtained if the propulsion system is impact on the turbofan engine. The ability to design the
designed around the technology. The most desired fluid dynamic behavior has always been the key
revolutionary benefits are typically achieved when a to designing gas turbine engines. Flow control has the
fundamental design constraint is removed. As a result, potential to be a breakthrough technology that enables a
many applications are much less attractive as a retrofit step change in our ability to design the fluid mechanics,
option. Thus for example, implementing the aspirated and break fundamental barriers that limit performance.
compressor requires thinking about where to use bleed Future engines that incorporate a range of these
air. Many more options are available with a clean sheet technologies may be able to yield dramatic
of paper than if one is considering a retrofit onto an improvements in key metrics and change the
existing engine. Such an analysis has been performed asymptotic nature apparent in plots such as Figure 5.
for this case; a design study considered the retrofit of a
single aspirated stage to replace the current 3-stage fan We have consciously chosen not to prioritize the
on the PW100-229. While many issues were identified, applications listed herein. There are three related
including estimates of weights and manufacturing costs, reasons for this. First is that some of the applications
the analysis indicated a net increase in life cycle cost, are more technically challenging than others, and the
based largely on TSFC penalty and added prioritization should be based on both impact and
manufacturing cost from hollow blades. This not technical risk. All of the applications have potential
surprising result simply emphasizes that for dramatic impact, and it may be that demonstrating technology in
changes in fundamental design constraints, significant a less risky but lower impact application becomes the
pathway by which the concepts get proven out and
12
American Institute of Aeronautics and Astronautics
ultimately implemented across a range of applications. turbine engine flow control applications based upon
Second is that the value depends on both the benefits, overall complexity and challenge.
and the costs (in terms of cost, weight, maintenance
impact, etc) of a particular solution, and thus cannot be ACKNOWLEDGEMENTS
determined solely on the application alone. Third is A broad paper such as this is ultimately the result of the
that the relative value of overcoming different problems comments and suggestions of many people. Almost
may differ for different aircraft and engine every paragraph corresponds to a conversation one of
combinations, or different missions. If a technology the authors has had with someone, and thus it is
makes the difference between being able to fit an impossible to acknowledge everyone who influenced
engine into a particular airframe or not, or meeting a this paper. However, special thanks are due to Jeff
regulatory constraint or not, then the value can become Cohen of UTRC (combustion mixing), Bob Neubert of
substantially larger than expected. Pratt & Whitney (aspirated compressor benefits
analysis), Mike Larkin of Pratt & Whitney (nacelle
The above comments notwithstanding, it is likely that aerodynamics), Dick Price and Joel Wagner of Pratt &
flow control implementation within the turbofan engine Whitney (turbine aerodynamics), Eric Gamble and Nate
will follow paths which are linked to the most proven Messersmith of Pratt & Whitney (fluidic nozzle
and reliable technology first. Thus, the more near-term technology) and Dan Miller of Lockheed Martin
applications are likely to involve extensions of elements (fluidic nozzle work in Figure 14 and 15), and Duane
currently in use (such as utilizing scheduled air for McCormick of UTRC (DSJ in Figure 9).
blowing in new separation control or cooling
applications). In addition, it is felt that low- REFERENCES
temperature, fixed engine components offer in general
fewer challenges relative to actuation authority, 1
robustness, integration, and flow control subsystem Prandtl, L. ber Flssigkeitsbewegung bei sehr
reliability. Examples would include the subsonic inlet kleiner Reibung, Proceedings, 3 rd International
at takeoff, the fan nozzle, fan case, and fan exit guide Mathematics Congress, Heidelberg, Germany, pp. 484-
vane. Within the mid-portion of the engine would 491, 1904.
2
come the next set of challenges as temperatures Wygnanski, I, Method and Apparatus for Delaying
increase within again fixed geometry flowpaths such as Separation of Flow from a Solid Surface, US Patent
turbine transition ducts, the intermediate case, and low 5,209,438, 1993.
3
pressure turbine guide vanes. Finally, the greatest McCormick, D.C. Boundary Layer Separation
challenges and likely least-near-term applications lie in Control with Directed Synthetic Jets, AIAA 00-0519,
the hot sections of the engine and on the rotating AIAA Aerospace Sciences Conference, Reno NV, 2000.
4
machinery airfoils themselves. High temperature, high Vermuelen, P.J., Grabinski, P. and Ramesh, V.,
Mach number, structural and rotational loads, and Mixing of an Acoustically Excited Air Jet With a
packaging are significant challenges in creating a Confined Hot Crossflow, ASME Journal of
reliable and functional flow control subsystem. Engineering for Gas Turbines and Power, Vol 114,
pp. 46-54, April 1992.
5
It is likely that non-flight critical applications will see Smith, B.L. and Glezer, A., Vectoring and Small-
implementation prior to those that are flight critical. Scale Motions Effected in Free Shear Flows Using
This includes such things as improvements in fuel burn Synthetic Jet Actuators, AIAA 97-0213, 35 th
or noise, rather than those that impact durability where Aerospace Sciences Conference, Reno, NV, 1997.
6
failure means the loss of the engine and airframe. The Rao, N.M., Feng, J., Burdisso, R.A., and Ng, W.F.,
engineer who has to put a product out the door is Active Flow Control to Reduce Fan Blade Vibration
necessarily conservative and can only consider proven and Noise, 5 th AIAA/CEAS Aeroacoustics Conference,
technology. The biggest barrier, therefore, may be the AIAA 99-1806, Seattle, WA, May 1999.
7
perception of a lack of reliability. Uninhabited air Brookfield, J.M., and Waitz, I.A., Trailing Edge
vehicles may also pose a more near-term technology Blowing for Reduction of Turbomachinery Fan Noise,
insertion path, because the flight safety issue is not 4 th AIAA/CEAS Aeroacoustics Conference, AIAA 98-
nearly so paramount. 2321, June 1998.
8
Tillman, T. G. and Hwang, D. P., Drag Reduction on
As mentioned above, specific application needs and a Large-Scale Nacelle Using a Micro-Blowing
goals will ultimately drive the final prioritization of Technique, AIAA 99-0130, AIAA Aerospace Sciences
flow control technology development. This discussion Conference, Reno, NV, 1999.
9
presents one view of a possible near-term class of gas Bewley, T.R., Moin, P., and Temam, R., DNS-based
predictive control of turbulence: an optimal benchmark
13
American Institute of Aeronautics and Astronautics
25
for feedback algorithms, under consideration for J. El-Aini, Y.M., Benedict, B. et al. Active rotor stage
Fluid Mechanics. vibration control, Patent # 6055805, May 2, 2000.
10 26
Gad-el-Hak, M., Introduction to Flow Control, in Walker, B.E., Hersh, A.S., Heidelberg, L.J., Sutliff,
Flow Control: Fundamentals and Practices, Springer- D.L, Spencer, M.E., Active Resonators for Control of
Verlag, 1998, pp. 1-108. Multiple Spinning Modes in an Axial Flow Fan Inlet,
11
Lynch, F. T., and Klinge, M. D., Some Practical 5 th AIAA/CEAS Aeroacoustics Conf., May 1999, Seattle
Aspects of Viscous Drag Reduction Concepts, SAE WA.
27
Paper 912129, Sept. 1991. Cohen, J.M., Rey, N.M., Jacobson, C.A. and
12
Aviation and the Global Atmosphere, A special report Anderson, T.J., Active Control of Combustion
of Working Groups I and III of the Intergovernmental Instability in a Liquid-Fueled Low-NO x Combustor,
Panel on Climate Change, ASME Journal of Engineering for Gas Turbines and
http://www.usgcrp.gov/usgcrp/IPCCINFO.html Power, Vol. 121, No. 2, pp. 281-284, April 1999.
13 28
Janes All the Worlds Aircraft. Bons, J.P., Sondergaard, R., Rivir, R.B., Turbine
14
McMichael, J.M., Progress and Prospects for Active Separation Control using Pulsed Vortex Generator
Flow control Using Microfabricated Electro- Jets, TurboExpo 2000, Munich, Germany. ASME
Mechanical Systems (MEMS), 34 th Aerospace paper 2000-GT-0262.
29
Sciences Conference, Reno NV, 1996. AIAA 96-0306. Anabtawi, A.J., Blackwelder, R.F., Lissaman, P.B.S.,
15
Grzybowski, R.R., Advances In Packaging and Liebeck, R.H., An Experimental Investigation of
Technologies For Electronics To 500C, Proceedings Boundary Layer Ingestion in a Diffusing S-Duct with
of Engineering Foundation 1998 High-Temperature and without Passive Flow Control, 37 th AIAA
Electronic Materials, Devices and Sensors Conference, Aerospace Sciences Conference, Reno NV, 1999.
San Diego, CA, Feb. 1998, pp. 207 - 215. AIAA paper 99-0739.
16 30
Grzybowski, R.R., Long Term Behavior of Passive Anabtawi, A.J., Blackwelder, R.F., Lissaman, P.B.S.,
Components For High Temperature Applications and Liebeck, R.H., An Experimental Study of Vortex
Proceedings of Engineering Foundation 1998 High- Generators in Boundary Layer Ingesting Diffusers with
Temperature Electronic Materials, Devices and Sensors a Centerline Offset, 35 th AIAA/ASME/SAE/ASEE Joint
Conference, San Diego, CA, Feb. 1998, pp. 172 - 179. Propulsion Conference, LA, CA, 1999. AIAA paper
17
de Jager, B., Rotating Stall and Surge Control: A 99-2110.
31
Survey, Proceedings, 34th IEEE Conference on Barry, B., Parke, S.J., Brown, N.W., Riedel, H., and
Decision and Control, New Orleans, LA, 1995, Sitzmann, M., The Flight Testing of Natural and
pp. 1857-1862. Hybrid Laminar Flow Nacelles, ASME Paper 94-GT-
18
Eveker, K.M., Gysling, D.L., Nett, C.N., and Sharma, 408, June 1994.
32
O.P., Integrated Control of Rotating Stall and Surge in Runyan, L.J, Bielak, G.W., Behbehani, R. Chen,
High-Speed Multi-Stage Compression Systems, ASME A.W., and Rozendaal, R.A., 757 NLF Glove Flight
J. Turbomachinery, 120(3), July 1998, pp. 451-457 Test and Results, Research in Natural Laminar Flow
19
D'Andrea, R., Behnken, R.L., Murray, R.M. Active and Laminar Flow Control, J.H. Hefner and F.E. Sabo,
Control of an Axial Flow Compressor via Pulsed Air compilers, NASA CP-2487, Part 3, pp. 795-818, 1987.
33
Injection, J. Turbomachinery, 119(4):742-752, 1997. Stockman, N.O., Latapy, M.O., Andrew, T.L., and
20
Kerrebrock, J.L., Reijnen, D.P., Ziminsky, W.S., Rogers, D.H., Scale Model Test of an Isolated
Smiig, L.M., Aspirated Compressors, ASME 97-GT- Turbofan Nacelle with Riblets, SAE Paper 912128,
525, Orlando, FL, 1997. Sept. 1991.
21 34
Freedman, J.H., Design of a Multi-Spool, High Speed Neiuwstadt, F.T.M., Wolthers, W., Leijdens, H.,
Counter-Rotating, Aspirated Compressor, MS Thesis, Krishna Prasad, K., and Schwarz-van Manen, A., The
Dept. of Aeronautics and Astronautics, MIT, reduction of skin friction by riblets under the influence
June 2000. of an adverse pressure gradient, Experiments in Fluids,
22
Kerrebrock, J.L., Drela, M., Merchant, A.A., Schuler, Vol. 15, pp. 17-26, 1993.
35
B.J., A Family of Designs for Aspirated Debisschop, J.R., and Nieuwstadt, F.T.M.,
Compressors, International Gas Turbine & Turbulent Boundary Layer in an Adverse Pressure
Aeroengine Congress, ASME 98-GT-196, 1998. Gradient: Effectiveness of Riblets, AIAA Journal,
23
Bae, J., Breuer, K., Tan, C., Active Control of Tip Vol. 34, No. 5, May 1996.
36
Clearance Flows in Axial Compressors, Fluids 2000, Hefner, J.N., and Bushnell, D.M., Viscous Drag
AIAA 2000-2233, Denver, CO, June 2000. Reduction via Surface Mass Injection, In Viscous
24
Kang, E., Breuer, K., Tan, C. Control of Leakage Drag Reduction in Boundary Layers. American
Flows using Periodic Excitation, Fluids 2000, AIAA Institute of Aeronautics and Astronoutics, Inc., 1989.
2000-2232, Denver, CO, June 2000.
14
American Institute of Aeronautics and Astronautics
37
Schlichting, H., Boundary Layer Theory. 7th Edition, 51
Anderson, C.J; Giuliano, V.J; Wing, D.J,
McGraw-Hill, New York, 1979. Investigation of hybrid fluidic/mechanical thrust
38
Hwang, D.P., A Proof of Concept Experiment for vectoring for fixed-exit exhaust nozzles, AIAA Paper
Reducing Skin Friction by using a Micro-Blowing 97-3148, 33 rd AIAA/ASME/SAE/ASEE Joint Propulsion
Technique, AIAA Paper 97-0546, Jan. 1997. Conference & Exhibit, Seattle, WA, July 6-9, 1997.
39 52
Marconi, F., Gilber, B., Tindell, R., Computational Kibens, V., Dorris III, J., Smith, D.M., Mossman,
Fluid Dynamics Support of the Development of a M.F., Active Flow Control Technology Transition:
Blockerless Engine Thrust Reverser Concept, AIAA The Boeing ACE Program, AIAA 99-3507.
53
97-3151, AIAA/ASME/SAE/ASEE Joint Propulsion The Aircraft Gas Turbine Engine and its Operation,
Conference, Seattle, WA, July 1997. United Technologies Corporation, 1988.
40
Lord, W.K., Jones, C.W., Stern, A.M., Head, V.L.,
and Krejsa, E.A., Mixer Ejector Nozzle for Jet Noise
Suppression. AIAA Paper 90-1909, July 1990.
41
Tillman, G.T., Paterson, R.W., and Presz, W.M., Jr.,
Supersonic Nozzle Mixer Ejector. AIAA Journal of
Propulsion and Power, Jan.-Feb. 1992.
42
Paterson, R.W., Turbofan Mixer Nozzle Flowfield-
A Benchmark Experimental Study, Journal of
Engineering for Gas Turbines and Power, Vol. 106,
July 1984, pp. 692-698.
43
Seiner, J.M., A New rational Approach to Jet Noise
Reduction, Theoretical and Computational Fluid
Dynamics, Vol 10, 1998, pp. 373-383.
44
Simonich, J.C., Narayanan, S., Barber, T.J.,
Nishimura, M., High Subsonic Jet Experiments Part I:
Aeroacoustic Characterization, Noise Reduction and
Dimensional Scaling Effects, 6 th AIAA/CEAS
Aeroacoustics Conference, Lahaina, HI, 2000. AIAA
Paper 2000-2022.
45
Zaman, K.B.M.Q., Effects of Delta Tabs on Mixing
and Axis Switching in Jets from Axisymmetric
Nozzles, 32 nd AIAA Aerospace Sciences Conf., Reno
NV, 1994. AIAA Paper 94-0186.
46
Anderson, B., Wygnanski, I., and Gutmark, E.,
Noise Reduction by Interaction of Flexible Filaments
with an Underexpanded Supersonic Jet, 37 th AIAA
Aerospace Sciences Conf., Reno NV, 1999. AIAA
paper 99-0080.
47
Corke, T., Glauser, M.N., Berkooz, G., Utilizing
Low Dimensional Dynamical Systems to Guide Control
Experiments, Applied Mechanics Review, 47, 6,
pp. 133-138, 1994.
48
Gridley, M.C. and Walker, S.H., Inlet and Nozzle
Technology for 21st Century Aircraft, ASME 96-GT-
244.
49
Miller, D.N., Yagle, P.J., Hamstra, J.W., Fluidic
Throat Skewing for Thrust Vectoring in Fixed
Geometry Nozzles, Paper AIAA 99-0365, 37 th AIAA
Aerospace Sciences Conference, Reno NV, 1999.
50
Giuliano, V.J; Wing, D.J, Static investigation of a
fixed-aperture nozzle employing fluidic injection for
multiaxis thrust vector control,, AIAA Paper 97-3149,
33 rd AIAA/ASME/SAE/ASEE Joint Propulsion
Conference & Exhibit, Seattle, WA, July 6-9, 1997.
15
American Institute of Aeronautics and Astronautics