HSR PROPULSION SYSTEM STUDIES: A STATUS REPORT ON THE DOWN-SELECT PROCESS
W.C. Strack
NASA Lewis Research Center
Cleveland, Ohio
HSR Airport Noise Challenge
-%/* 7 /
First Wish:
• Simple turbojet engines (V jet ~ 3200 ft/s)
• Acceptable 20 + dB mixer-ejector nozzles
1992 Status:
Unsuppressed
sideline noise
-123 EPNdB
Stage m Goal
102.5 EPNdB
Stage IV
Status with Generation I
mixer-ejector nozzles
and
without new operational procedures
When the HSR program began there was widespread belief that a simple and
familiar turbojet-like engine coupled to an advanced technology mixer-ejector
nozzle was the propulsion system of choice for achieving FAR 36-Stage 3 noise
requirements. Our ability to quickly demonstrate a practical 20 + dB suppression *
nozzle was confidently presumed by many. Our rate of progress towards that
objective, however, has been somewhat humbling. At the moment we are
reasonably confident of achieving about 1 5dB suppression with a mixer-ejector
nozzle designed for a high specific thrust turbojet-like cycle. Therefore, if we make
no further suppression progress and conservatively assume no new operational
procedures such as programmed lapse rate (PLR), then meeting the Stage III goal
requires a large amount of engine and/or wing oversizing which is economically
prohibitive. The scenario is further aggravated by the possibility of eventually
needing to comply with even more stringent regulations (Stage IV).
While this status may be somewhat disappointing to some, it must be remembered
that the HSR program plan involves two generations of mixer-ejector nozzles
beyond the current generation I nozzle designs. It is premature to conclude that
we cannot design a practical 20 + dB mixer-ejector nozzle. On the other hand, it is
prudent to consider alternative solutions to the noise problem. Thus, we are
investigating four other propulsion system concepts.
31-1
HSCT Noise Suppression Concepts
Rear End
—
Hybrid
Front End
Variable Cycle Engine
Flade
The highest specific thrust concept is the 1 -spool turbine bypass engine (a slightly
modified turbojet) combined with a very large mixer-ejector nozzle requiring
approximately 120 percent airflow augmentation during takeoff. The mixed flow
turbofan (MFTF) and variable cycle engine (VCE) concepts have intermediate jet
velocities because a low-spool driven fan absorbs much of the core energy.
Consequently, much less secondary air is required in the miker-ejector nozzle to
achieve low noise than the turbine bypass engine (TBE). The Flade engine is either a
VCE or a MFTF with a third flowpath surrounding the fan and scrolled to the lower half
of the engine. The fan driving this flowpath is modulated to absorb power during the
takeoff and this provides a fluid acoustic shield underneath the mixer nozzle (no
ejector). The TBE with an inlet flow valve (IFV) represents one member of the tandem
fan class of concepts wherein a compression system reconfiguration can occur.
During takeoff, auxiliary air is brought onboard and routed to the rear compressor
while the normal inlet airflow is processed only by the front compressor before
exhausting. In the cruise configuration, the auxiliary inlets are closed and the engine
becomes a turbojet with an extra pressure loss due to the IFV.
All of these candidate concepts achieve about 1 500 ft/s exhaust velocity during
takeoff by raising the total airflow to about 1100 Ib/sec. They differ in where the
airflow is introduced into the cycle and which technologies need to be developed to
achieve success.
31-2
Effect of Exhaust Velocity on Sideline Noise
Mach 0.322, 689 ft., 650 lb/s
Each of these five concepts can be characterized by its exhaust velocity and,
therefore, its suppression requirements compared to a conventional unsuppressed
nozzle. While a TBE presents a 20 + dB suppression problem to attain Stage III, the
TBE/IFV can be designed to achieve Stage III without an elaborate suppression
system, and the hybrid concepts fall somewhere in-between these extremes.
There are, of course, other discriminating attributes to be considered such as
weight, reliability, life, efficiency, thrust lapse, technology risk, tolerance to more
severe noise constraints, installation drag, and climb noise. What is needed is an
unbiased procedure to evaluate each of these concepts on a system basis that
accounts for all of these criteria simultaneously.
Overall HSCT Noise Issues
1. Which propulsion concept best achieves a
balanced compromise of performance, weight,
size, noise, complexity, and life ?
2. What price do we pay to achieve noise levels
below Stage III ?
issue 1 is important to resolve because the HSR program is resource-constrained to
pursue technologies specific to only two concepts at most. This is also a difficult
challenge to resolve with a high degree of confidence due to the large number of
independent variables, the complex interactions between propulsion and airframe,
multiplicity of merit criteria, and key technology and external uncertainties.
Since it is likely that noise regulations will be tightened sometime in the future it is
important to determine the economic penalty associated with such an eventuality.
Also, some propulsion concepts are able to accommodate severe noise constraints
better than other concepts. Thus, this information could be a key discriminator
during the concept selection process. It would be best if we generate a curve of
penalty (e.g., aDOC) versus AdB below Stage III rather than presume a definition of
Stage IV. Then, there could be more rational future rule-making.
31-4
A Propulsion System Down-Select is Needed to Focus the
Phase II Technology Program
Late 1993 Down-Select
Decision Gate
In addition to the mainline environmental technology feasibility effort in HSR Phase
I, a systems studies effort is also underway to address the issues listed on the
previous chart. The schedule calls for a down-select to both a primary and backup
concept by late 1 993. This will focus the technology effort in HSR Phase II. Note
that this down-select pertains to the NASA sponsored technology thrust only-i.e.,
it is not a production engine down-select. The intent of the '93 down-select is to
insure that the correct concept-specific technology is pursued in the earlier portion
of Phase II to enable a low-risk final down-select in late '95.
31-5
HSR Propulsion System Selection Process
Candidate
Concepts — ►
TBE
VCE
MFTF
Flade
TBE/IFV
Common — ►
Groundrules
The process by which the down-select information is acquired begins with the
establishment of a common set of groundrules for all participants to minimize the
risk of disparate results. GE/PW are to perform a preliminary concept screening
using takeoff gross weight as the prime evaluation criterion. This means that for
each candidate concept the cycle will be optimized and representative airplane and
mission models adopted. Propulsion-airframe installation (PAI) differences such as
interference drag will not be captured, however. The output of this first level
screening is passed to Boeing and Douglas for detailed comparative evaluations
that include PAI effects. Boeing and Douglas have adopted somewhat different
airframes and mission definitions (e.g., programmed lapse rate assumption) which
means that somewhat different results may ensue. The merit criteria will be direct
operating cost (DOC) and technical risk.
31-6
HSR Propulsion System Concept Selection Criterion
9 Prime discriminators: Direct operating cost (DOC)
Technical risk ( uncertainty band)
9 Risk to be incorporated into DOC when feasible
DOC
Concept 1 Concept 2 Concept 3 Concept 4 Concept 5
To avoid an unwieldy number of risks associated with the various technologies,
risks will be incorporated into DOC wherever possible. For example, if an
unconventional component such as a mixer-ejector nozzle requires an expensive
R&D program to reach acceptable risk, then the cost of this element is
incorporated into the DOC. The end result is an uncertainty band on DOC that
reflects the agglomerated risks associated with each concept. Conceivably, the
down-select process will produce results as depicted wherein the nominally lowest
DOC concept (arbitrarily drawn as concept 4) also has the highest risk. In this
case, the decision-maker will need to make a judgement concerning the balance
between DOC benefit and increased risk. In the example shown, concept 3 might
be preferred over concept 4 due to its lower technical risk.
31-7
Monte Carlo Simulation Risk Analysis
/ '
Direct
Operating
Cost Model
- —
/ s
Aircraft
System
Model
s >
In order to determine the DOC uncertainties for each concept, GE/PW and Douglas
are invoking a Monte Carlo simulation risk analysis. Component experts will
estimate the probability of attaining several values of the key component criteria
such as efficiency weight, acquisition cost, and maintenance cost. This will define
sets of probability curves for each propulsion component such as the mixer-ejector
nozzle, IFV, Flade fan, and mixed compression inlets. Random sampling of these
component probabilities done many times, together with a propulsion systems
model, will yield another (smaller) set of curves for each concept. These are
confidence curves for the complete propulsion system. Combining this information
with aircraft, mission, and economic models will lead to DOC confidence curves for
each of the five propulsion concepts. Finally, these DOC confidence curves can be
interrogated at three levels to yield the desired DOC uncertainty bands (e.g., 20
percent, 50 percent, and 80 percent).
Boeing prefers to do a more traditional risk analysis instead of a Monte Carlo
analysis. They will interrogate a group of experienced technology experts to
estimate risks associated with each candidate concept. This traditional approach
will provide a check on the Monte Carlo simulation.
31-8
Level 2 HSR Program Schedule - Propulsion System Studies
HSR Program Mtestone
Propulsion Design Studies
and Evaluations
Airframer Evaluations
NASA Design StucMes
and Evaluations
FY 91
FY92
FYM
FY94
Propulsion Down-select
m life Study Evaluations y
CE/PW
?|E3p
“/ II
II
Common Down-select Engine Engine
Ground rules Process Cost
\ \ \ \ \
THE MF7T IFV VCE Flade
\\\\ \
BAC&DAC
1 1
II
li
II
Fcllb*± A
Fallback
Strategies
URC
& Evaluations . Fallback A
Strategies
LaRC JA ftCfe
ewwoi/A
Having established common groundrules amongst GE/PW, Boeing, Douglas, and
NASA, the propulsion system design studies are well underway within the
propulsion community. Mach 2.4 data have been generated and delivered to the
airframers for the TBE* MFTF (bypass ratios of 0.4, .63, 1.13), and the TBE/IFV.
The VCE and Flade data are nearly complete. These data include performance,
weights, cost, and acoustic information. In the spring of 1993, Boeing and
Douglas will have completed a first pass comparison of all of the engine
candidates. At this point, all first-order technical issues and concept-specific
concerns will be identified. From then on, detailed analyses will be conducted to
insure that each concept is fairly judged. This entails exploring ways to mitigate
the weaknesses associated with each propulsion concept. The plan calls for
sufficient information to be acquired by the beginning of FY94 to enable a credible
down-select decision. NASA is also performing design studies and comparative
evaluations to strengthen the overall effort. Because the technical challenges are
complex and five organizations are involved in designing and evaluating numerous
engine and nozzle concepts, there is also an enormous information management
challenge to ensure effective use of available resources.
31-9
Status of NASA In-House Comparative Propulsion Studies
Mach 2.4, 100% supersonic 5000 n.mi. range, 292 passengers
Stage III sideline constraint
Assumes material goals and mixer-ejector nozzle goals are achieved
LeRC results as of September, 1992
Relative Takeoff
Gross Weight
Currently, there is not much comparative data to examine and what data does
exist is quite tentative and laced with caveats. Nevertheless, this chart displays
NASA's current state of understanding of three of the five concepts. Relative
takeoff gross weight is shown assuming the EPM materials goals and the HSR
mixer-ejector nozzle goals are achieved (e.g., the TBE nozzle delivers 18 + dB
suppression in an acceptable size), and that no significant PAI penalties exist.
Even with these assumptions the MFTF is superior to the TBE because its cruise
efficiency is significantly better. If the PAI differences do not hurt the MFTF, this
candidate appears to be very competitive. Because the VCE is essentially a MFTF
derivative, it too is expected to compete well--especially on missions with large
subsonic legs. The Flade is also anticipated to be quite competitive for similar
reasons.
31-10
Key Propulsion System Uncertainties
Technical
1. Adequate mixer - ejector nozzle aero / acoustic performance
2. Materials progress
External
3. More severe airport noise regulations
4. Operational procedure regulations (e.g., programmed lapse rate)
5. Climb noise
6. Mach number selection
By itself, the previous chart depicts an oversimplified situation. In reality, there are
a number of first-order uncertainties that need to be considered in the selection
process. The sensitivity of the comparison with respect to these uncertainties
needs to be determined to select wisely.
31-11
Impact of Noise Suppression Technology and Noise Constraint
Mach 2.4, all supersonic 5000 n.mi. range, 292 passengers
Cycle, nozzle, wing hading, thrust hading vary along each curve
Ignores aircraft instalhtion differences and possible climb noise constraint
NASA Lewis results as of November 1992
Relative Takeoff
Gross Weight
An example of the impact of the first three key uncertainties is displayed here in terms of
takeoff gross weight relative to a TBE powered airplane with a 2900°F T 41 . (The TBE would
yield lower TOGW if it were not constrained by turbine blade material limits and unavailability
of a suitable high-suppression nozzle.) Note that the TBE is quite sensitive to the degree of
success in achieving a quiet nozzle. If only 1 3dB of suppression is achieved, then the TBE's
TOGW penalty is about 1 4 percent. The TBE curve represents various amounts of wing and
engine oversizing to meet Stage III sideline noise. The MFTF (and the other concepts as well)
offer more degrees of freedom in the form of cycle changes (BPR) to mitigate the adverse
impact of a mixer-ejector nozzle technology shortfall. Hence the MFTF curve is less steeply
sloped, and it could accommodate even a 10dB suppressor nozzle without a show-stopping
penalty.
The NASA results shown here also indicate that a 400°F material temperature shortfall would
not be disastrous although 600°F would be. On the other hand, industry generated data show
at least twice the sensitivity displayed here. These differences will be resolved soon.
The impact of a Stage lll-5dB noise constraint may be determined by comparing results using
the lower abscissa scale with results using the upper scale. For example, the aTOGW for a
MFTF is about 7 percent for a 1 5dB suppressor nozzle.
Finally, it should be understood that this figure is just the beginning. Undoubtedly it will
change as more realism is added. Firm conclusions based on this alone are premature.
31-12
Variable Bypass Supercharged Core (VBSC)
Low-Row Mode
BPR -.60
Uuwr tan
Outer tan
High-Row Mode
BPR-1.13
VJ<1500 fps
IFV
Advantages
9 Core remains supercharged in high-flow mode
9 Less pressure drop, better cruise TSFC
9 Less auxiliary inlet air required
9 Lower engine weight
The previous chart contained several points for TBE/IFV and TJ/IFV engines. At
the moment there is some controversy concerning whether the plotted points are
too optimistic or not. Regardless of how that controversy is resolved, it is clear
that, while such high flow concepts are appealing because they obviate the need
for a high-risk mixer-ejector nozzle, they also suffer serious deficiencies. Namely, a
non-supercharged core and consequently low thrust in the takeoff mode, large and
heavy engines, and pressure drop through the IFV during cruise. These
deficiencies may be partially alleviated by the new concept illustrated here. It is a
turbofan/IFV with a flow splitter that keeps the core supercharged by the inner fan
flow at all times. It also features a core-driven aft fan stage that prevents bypass
ratio from rising at higher flight speeds (opposite of mission requirement).
31-13
VBSC
Takeoff and Climb to Cruise
Jet Velocity, Vj, fps
Mach
Another advantage of the variable bypass supercharged core (VBSC) concept is its
ability to efficiently stay in the high-flow mode throughout climb. This may prove
to be important to reduce climb noise. Shown here is the exhaust velocity V ( and
net thrust F n during a typical climb path. Note the modest Vj throughout-rising
from 1450 ft/s at Mach 0.3 to 2000 ft/s at Mach 0.9 at which point the mode
switch occurs. NASA has conceptualized this engine very recently and has
solicited industry feedback (pending) before adopting its inclusion into the down-
select process.
31-14
TOCW Sensitivity to Nozzle Performance and Weight
Mach 2.4 HSCT
Achieving high nozzle cruise efficiency is absolutely essential. A 1 percent C fg
shortfall can increase TOGW by over 4 percent. Clearly, we need confidence in
our predictive codes to substitute for lack of experimental data for many of the
unconventional nozzle concepts.
Nozzle weight is also a sensitive parameter. As the studies progress, the initially
large spread in weight estimates has significantly diminished.
Avoided on this figure is any mention of the takeoff nozzle performance which is
often cited as important also. (E.g., it is one of the two merit criteria in the oft-
used mixer-ejector nozzle technology goal charts.) This omission was deliberate
because some of the high-specific thrust engines are top-of-climb sized and
therefore do not suffer large penalties for takeoff C fg 's as low as 0.85 or so. There
is also some evidence that even takeoff sized engines could tolerate relatively poor
takeoff C fg 's if wing size is free to vary to compensate.
31-15
Thrust Augmentation Issue
Some cycles benefit from a
mini-augmentor during climb
TOGW
for
MFTF
0.4 BPR
But
Top of Climb
Nozzle Throat
Temp., °F
MFTF -BPR
One of the more recently discovered issues is whether to use a mini-thrust
augmentor or not in the high specific flow engines. For example, Boeing prefers to
use a mini-augmentor during the upper climb path to offset marginal thrust levels
that cause inefficient transonic system performance. A 3 percent TOGW reduction
is possible using a mini-augmentor rather than a dry engine for a MFTF with 0.4
bypass ratio. However, the use of augmentation also boosts the nozzle
temperature levels about 600°F from the 1 200-1 400°F level to the 1700-2000°F
level. The question is whether the TOGW payoff is worth the increased risk and
maintenance associated with the higher temperature experienced during the upper
climb. This issue is being investigated further.
31-16
Nacelle Placement Restraints
Another powerful influence on the down-select decision is propulsion-airframe
integration (PAI). For example, the nacelle shape, which is driven by the
propulsion geometry and changes significantly from one concept to another, and
placement can dramatically alter the interference wave drag. Hence, to compare
the alternative propulsion concepts we need to assure ourselves that PAI effects
are properly determined even if this requires more than the usual analysis depth to
understand. From early calculations it appears that some of the concepts do not
integrate easily with the airframe and some re-design effort is warranted to avoid
premature judgements.
31-17
Summary
1. Late 1993 propulsion system down-select requires
reliable M-E nozzle database
— Adequate progress but not established yet
2. Considerable concern exists about M-E nozzle risk
— Aero / acoustic performance — Weight / size — Life
3. Interest shifting toward low specific thrust cycle
solutions to noise challenge
4. Stage III-5 dB incurs approximately a 7% airplane
takeoff weight penalty for mixed-flow turbofans with
M-E nozzles
At the moment, the propulsion system down-select process is hindered by our lack
of an adequate experimental mixer-ejector nozzle database to enable high-
confidence aero/acoustic/weight modeling. Progress in establishing the needed
data base is progressing adequately but fitfully. Certainly there exists considerable
concern about M-E nozzles--enough to spawn a new wave of interest in the high-
specific flow alternatives. In the end, it is likely to come down to a matter of
which technology challenges do we prefer to pursue. The decision may depend
upon risks as much as on potential benefits.
31-18
