Boundary Layer Ingestion
Benefit Quantification and Analysis Framework
Alejandra Uranga
Gabilan Assistant Professor
Aerospace & Mechanical Engineering
6th International Workshop on Aviation and Climate ChangeUniversity of Toronto Institute for Aerospace Studies
May 16–18, 2018
Outline
1 Introduction
2 D8 Aircraft: Conceptual Studies
3 Experimental Measurement of BLI Benefit
4 Analysis Framework: “Data Mining”
5 Other BLI-Related Topics
A. Uranga (USC) 1 / 37
Boundary Layer Ingestion (BLI)
wake, or “draft”
WastedKinetic Energy
Zero NetMomentum
combined wake and jet
propulsor jet
+
+
+
+
+
+
+
-
-
-
I BLI reduces wasted KE in combined jet+wake (mixing losses)
I Long known to have large potential, never realized for aircraft
I Ambiguous decomposition into drag and thrust(airframe) (propulsion system)
) use of power balance instead of force accounting
A. Uranga (USC) 2 / 37
Summary
I Closer integration of propulsion system and airframe provides newopportunities to reduce fuel burn and emissions of commercial aircraft
I Boundary layer ingestion (BLI)
I Novel configurations
I System optimization (airframe, engine, operations)
I Flow power and dissipation in power balance frameworkprovide useful metrics for integrated configurations
I D8 wind tunnel testsI Quantification of aerodynamic BLI benefitI Proof-of-concept for use of BLI in transport aircraft
I Aerodynamic framework developed to analyze aircraft with BLI
A. Uranga (USC) 3 / 37
Major Collaborators
MIT
Mark DrelaEdward Greitzer
UTRC
Scott OchsGreg Tillman
Pratt & Whitney
Wes Lord (retired)
University of Michigan
Joaquim Martins
NASA
Scott Anders (LaRC)Greg Gatlin (LaRC)Judith Hannon (LaRC)James Heidmann (GRC)Natari Madavan (ARC)Shishir Pandya (ARC)Sally Viken (LaRC)
A. Uranga (USC) 4 / 37
Outline
1 Introduction
2 D8 Aircraft: Conceptual Studies
3 Experimental Measurement of BLI Benefit
4 Analysis Framework: “Data Mining”
5 Other BLI-Related Topics
A. Uranga (USC) 5 / 37
Phase 1: D8 Aircraft Concept2008-2010
MIT N+3 D8.2
A. Uranga (USC) 6 / 37
I B737-800/A320 class
I 180 PAX, 3,000 nm range
I Double-bubble lifting fuselagewith pi-tail
I Two aft, flush-mounted enginesingest ⇠ 40% of fuselage BL
I Cruise Mach 0.72
�37% fuel with current tech(configuration)
�66% fuel with advanced tech(2025-2035)
No “magic bullet”
E. Greitzer et al. 2010, NASA CR 2010-216794A. Uranga et al. 2014, AIAA 2014-0906 NASA-MIT Cooperative Agreement NNX08AW63A
A. Uranga (USC) 7 / 37
Photos NASA/George Homich
System Impact of BLI
BLI benefitsI
Aerodynamic (direct) benefitsI Reduced jet and wake dissipationI Reduced nacelle wetted area
ISystem-level (secondary) benefits
I Reduced engine weightI Reduced nacelle weightI Reduced vertical tail sizeI Compounding from reduced overall weight
“Morphing” sequence: B737-800 7! D8I Features of D8 introduced one at a time
I Sequence of conceptual designs, optimized at each step (TASOPT)
A. Uranga (USC) 8 / 37
E. Greitzer et al. 2010, NASA CR 2010-216794
M. Drela 2011, AIAA 2011-3970
Morphing Sequence: B737-800 7! D8.2 7! D8.6
0
0.2
0.4
0.6
0.8
188 %
81 % 82 %
67 % 66 %
100 %op
timiz
ed73
7-80
0M
=0.
8,C
FM56
engi
ne
0
slow
toM
=0.
72
1
D8
fuse
lage
,pit
ail
2
rear
podd
eden
gine
s
3
inte
grat
eden
gine
s,B
LI
4
optim
ize
engi
neB
PR
,FP
R
5
2010
engi
nes
8
FuelBurn
63 %
D8.2
6 7 9 10
2035
engi
nes
2035
mat
eria
ls
win
gbo
t.N
LF
smar
tstru
ct
48 %
38 % 35 % 34 %
D8.6- 18 %
- 23 %
- 21 %
A. Uranga (USC) 9 / 37
Phase 3: Trade-O↵s Summary2015-2017
Metric: Payload-Range Fuel Consumption =Fuel Energy Consumed
Payload Weight⇥Range
ID8 configuration benefit (20± 3)%relative to tube-and-wing at same cruise speed and technology
IN+3 technology benefit (45± 2)% relative to 1990s tech
I Tech advances benefit tube-and-wing more:D8 has lower structural/total weight and higher payload/total weight
ISlowing down from Mach 0.78 to 0.72 (5± 1.5)%
I Tube-and-wing benefits more from lower speed
A. Uranga (USC) 10 / 37
NASA-MIT Cooperative Agreement NNX15AM91A
Outline
1 Introduction
2 D8 Aircraft: Conceptual Studies
3 Experimental Measurement of BLI Benefit
4 Analysis Framework: “Data Mining”
5 Other BLI-Related Topics
A. Uranga (USC) 11 / 37
Phase 2: Airframe-Engine Integration2010-2015
Quantification of D8 BLI benefit (experimental/computational)
I Direct back-to-back comparison of BLI vs non-BLI
I Wind tunnel tests of 1:11 scale (4m span) powered models
A. Uranga (USC) 12 / 37
Non-BLI(Podded)
BLI(Integrated)
NASA-MIT Cooperative Agreement NNX11AB35A
BLI Benefit
BLI benefit (aerodynamic)
Savings in power required for given net stream-wise force
with BLI engines relative to non-BLI engines
Power metric
Mechanical flow power transmitted to the flow by the propulsors
PK =
I(po � po1)V · n̂ dS (incompressible)
BLI benefit ⌘ PKnon-BLI � PKBLIPK
non-BLI
����at given FX
A. Uranga (USC) 13 / 37
Obtaining PK
Method 1: Integration of the flow on propulsor stream-tube
PK =
Z
exit
(po � po1)V · n̂ dS �Z
inlet
(po � po1)V · n̂ dS
exitinlet inlet exit
Method 2: Conversion of electrical power provided to the propulsor motors
PK|{z}mechanical
flow power
= ⌘f|{z}fan
e�ciency
PK/PS
⇥ ⌘m|{z}motor
e�ciency
PS/PE
⇥ PE|{z}electrical
power
A. Uranga (USC) 14 / 37
Phase 2: Demonstrated Aerodynamic BLI Benefit2010-2015
0 0.02 0.04 0.06 0.08 0.1
-0.04
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
0.04
CPK
CX
BLI
non-BLI
unpowered
net drag
net thrust
cruise
Direct Rec = 0.68⇥106 (FHP)Direct Rec = 0.57⇥106 (FHP)Direct Rec = 0.57⇥106 (rake)Indirect Rec = 0.68⇥106Indirect Rec = 0.57⇥106
0.04 0.045 0.05 0.055-4
-3
-2
-1
0
1
2
3
4⇥10�3
CPK
CX
BLI
non-BLI
net drag
net thrust
cruise
Indirect Rec = 0.57MIndirect Rec = 0.68MDirect Rec = 0.57M (rake)Direct Rec = 0.57M (FHP)Direct Rec = 0.68M (FHP)
8.6% Benefitat Cruise
A. Uranga (USC) 15 / 37
NASA-MIT Cooperative Agreement NNX11AB35A
Outline
1 Introduction
2 D8 Aircraft: Conceptual Studies
3 Experimental Measurement of BLI Benefit
4 Analysis Framework: “Data Mining”
5 Other BLI-Related Topics
A. Uranga (USC) 16 / 37
BLI Benefit Sources
1 Lower propulsor jet dissipation and higher propulsive e�ciency:more useful power put into the flow
2 Lower surface dissipation (smaller nacelle size and surface velocities)
3 Lower wake dissipation (partial elimination of viscous wake)
4 Lower weight due to smaller nacelles and smaller engines,which in turn enables smaller wings, and thus even less weight
1 + 2 + 3 = aerodynamic benefit: less flight power requiredfor a given airframe operating at the same lift coe�cient
4 = system-level benefit after aircraft re-optimizations
A. Uranga (USC) 17 / 37
Power Balance Method
Consider mechanical energy sources and sinks:
[ Net Force ] = [ Dissipation ] � [ Power In ]
FX|{z}“drag – thrust”
V1 = ( �surf + �wake + �vortex + �jet ) � ( PK|{z}mechanical
flow power
+ ⇢⇢PV|{z}p dVpower
)
PK
�wake
�wake �vortex
�vortex�surf
�surf
PK
�jet
�jet
Non-BLI Configuration
BLI Configuration
A. Uranga (USC) 18 / 37
M. Drela 2009, AIAA Journal 47(7)
Airframe Dissipation (1/2)
I Conventional drag decomposition:
D
0V1 = �
0surf
+ �0wake| {z }
D0p V1
(profile)
+ �0vortex| {z }
D0i V1
(induced)
I Surface dissipation: �0surf
= (1�fwake
)D 0pV1
�surf
= (1�fwake
)D 0p V1 � ��surf
wheref
wake
⌘�0wake
�0surf
+ �0wake
��surf
⌘ �0surf
� �surf
> 0
A. Uranga (USC) 19 / 37
��surf
Airframe Dissipation (2/2)
I Wake dissipation:
�0wake
= fwake
(�0surf
+ �0wake
) = fwake
D
0p V1
�wake
= (1�fBLI
) �0wake
= (1�fBLI
) fwake
D
0p V1
where fBLI
⌘ boundary layer ingestion fraction= fraction of total airframe viscous kinetic energy defect
ingested by propulsors
I Vortex dissipation: �vortex = �0vortex
= D 0i V1
assuming comparison is made with same airframe at fixed CL
A. Uranga (USC) 20 / 37
(1� fBLI
)
Jet Dissipation
I Jet dissipation (with or without BLI):
�jet
=
ZZ
exit
1
2
(V�V1)2 dṁ
=1
2(V
jet
�V1)2 ṁ
for ṁ = Nprop
⇢jet
V
jet
A
jet
(total propulsor mass flow)
and assuming uniform velocity Vjet
across the jet
( jet pt non-uniformity ⌧ propulsor pt rise )
A. Uranga (USC) 21 / 37
Mechanical Flow Power
PK ⌘ �ZZ
prop
⇥p1�p + 1
2
⇢�V
2
1�V 2�⇤
V · n̂ dS
=1
2
�V
2
jet
� V 21�ṁ
| {z }(exit)
� (�fBLI
�0surf
)| {z }
(inlet)
=1
2
�V
2
jet
�V 21�ṁ + (1�f
wake
) fBLI
D
0p V1| {z }
extra power
A. Uranga (USC) 22 / 37
(1�fwake
) fBLI
D
0p V1| {z }
extra power
Parametric Expressions
Parametric expression for power required and stream-wise forcein terms of reference non-BLI configuration and propulsor operation
CPK = Nprop
"✓V
jet
V1
◆2
� 1#⇢jet
⇢1
V
jet
V1
A
jet
A
noz
A
noz
S
ref
+ (1�fwake
) fBLI
C
0Dp
CX = C0D � fBLIC 0Dp ��C�
surf
� 2Nprop
✓V
jet
V1� 1
◆⇢jet
⇢1
V
jet
V1
A
jet
A
noz
A
noz
S
ref
I Valid with and without BLI, depending on value of fBLI
I Jet properties (⇢jet
,Ajet
,Vjet
) with and without BLI may di↵er
I Quantify BLI benefit relative to known non-BLI configuration
A. Uranga (USC) 23 / 37
Major Design Parameters for BLI Aircraft
(i) Ingested dissipation fBLI
C
0Dp
(ii) How “well-designed” the BLI engine installation is ) ��surf
(iii) Propulsor operating points for each of the configurations) respective propulsor jet velocities or mass flows
A. Uranga (USC) 24 / 37
Data Mining: Application to D8 Wind Tunnel Tests
Use expressions for CPK and CX to fit experimental data (CL = C0L = 0.64)
0 0.02 0.04 0.06 0.08 0.1-0.04
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
0.04
CPK
CX
BLI
non-BLI
unpowered
8.2%BLI Benefit
(CI = ± 0.3%)
net drag
net thrust
cruise
Data Rec = 0.57⇥106
Power Balance Curve FitData Rec = 0.68⇥106
A. Uranga (USC) 25 / 37
C
0D = 0.0370
�C�
surf
= 0.0012
f
wake
= 0.080
A
jet
A
noz
= 0.955
CX0
= 0.0357
CDairframe
= 0.0338
Propulsive E�ciency ⌘p ⌘PK � �jet
PK
0.74 0.75 0.76 0.77 0.78 0.79 0.80.88
0.9
0.92
0.94
0.96
0.98
1
1.02
⌘p
PKP 0K
non-BLI
4.0%lower airframe
dissipation
8.2%benefit
at equalA
nozzle
BLI
propulsive efficiency gain
A. Uranga (USC) 26 / 37
63100
30100
Jet5.1%
Surface2.5%
Wake0.6%
7 100
Variable Nozzle Area: Di↵erent Propulsor Designs
0.74 0.75 0.76 0.77 0.78 0.79 0.80.88
0.9
0.92
0.94
0.96
0.98
1
1.02
⌘p
PKP 0K
non-BLI
BLI
Plug BPlug A
Plug C
8.5% 8.2% 7.6%
A. Uranga (USC) 27 / 37
Bases for Comparison
0.4 0.6 0.8 1 1.20.88
0.9
0.92
0.94
0.96
0.98
1
1.02
PKP 0K
non-BLIBLI
Anozzle /A0
nozzle
equalVjet
equal⌘p
equalA
nozzle
equal�jet
equalPK
equalṁ
A. Uranga (USC) 28 / 37
Bases for Comparison
0.4 0.6 0.8 1 1.20.88
0.9
0.92
0.94
0.96
0.98
1
1.02
PKP 0K
non-BLIBLI
Anozzle /A0
nozzle
0.3
0.4 equalVjet
equal⌘p
equalA
nozzle
0.1
equal�jet
equalPK
equalṁ
0.2
fBLI = 0
A. Uranga (USC) 29 / 37
Bases for Comparison
0.4 0.6 0.8 1 1.20.88
0.9
0.92
0.94
0.96
0.98
1
1.02
PKP 0K
non-BLIBLI
Anozzle /A0
nozzle
0.3
0.4 equalVjet
equal⌘p
equalA
nozzle
0.1
equal�jet
equalPK
equalṁ
0.2
fBLI = 0
A. Uranga (USC) 30 / 37
Outline
1 Introduction
2 D8 Aircraft: Conceptual Studies
3 Experimental Measurement of BLI Benefit
4 Analysis Framework: “Data Mining”
5 Other BLI-Related Topics
A. Uranga (USC) 31 / 37
High-E�ciency, High-OPR, Small Cores (N+3 Phase 2, P&W)
Pratt & Whitney – Lord et al., AIAA 2015-0071 : reverse core engine arch.
SAE INTERNATIONAL
“Engine Architecture for High Efficiency at Small Core Size” Lord et al., AIAA 2015-0071 - Pratt & Whitney
42 A. Uranga (USC) 32 / 37
D8 Transonic Design (N+3 Phase 3, U.Michigan)
Transonic wing and engine integration MDO for Mach = 0.72, 0.78 :aero-structural optimization with loosely coupled engine model
A. Uranga (USC) 33 / 37
Airframe-Propulsion Integration Challenges
Integration lines strongly dependent on optimization’s objective function
I Hard to identify MDO objective: need fully coupled engine model
I High sensitivity of fan face condition to di↵user shape
A. Uranga (USC) 34 / 37
Airframe-Propulsion Integration Challenges
A. Uranga (USC) 35 / 37
ReferencesDrela, M., “Power Balance in Aerodynamic Flows”, AIAA Journal, Vol. 47, No. 7,
2009, pp. 1761–1771. doi:10.2514/1.42409 [power balance method]
Uranga, A., Drela, M., Greitzer, E., Titchener, N., Lieu, M., Siu, N., Huang, A., Gatlin,G., and Hannon, J., “Preliminary Experimental Assessment of the Boundary LayerIngestion Benefit for the D8 Aircraft”, AIAA-2014-0906, 52nd AIAA AerospaceSciences Meeting, SciTech 2014, National Harbor, Maryland, 13–17 Jan. 2014.doi:10.2514/6.2014-0906 [preliminary wind tunnel test results, morphing chart]
Uranga, A., Drela, M., Greitzer, E. M., Hall, D. K., Titchener, N. A., Lieu, M. K., Siu,N. M., Casses, C., Huang, A. C., Gatlin, G. M., and Hannon, J. A., “Boundary LayerIngestion Benefit of the D8 Transport Aircraft”, AIAA Journal, Vol. 55, No. 11, pp.3693–3708, 2017. doi:10.2514/1.J055755 [wind tunnel tests]
Uranga, A., Drela, M., Hall, D. K., and Greitzer, E. M., “Analysis of the AerodynamicBenefit from Boundary Layer Ingestion for Transport Aircraft”, AIAA Journal, inpress [analysis, “data mining”]
Hall, D. K., Huang, A. C., Uranga, A., Greitzer, E. M., Drela, M., and Sato, S.,“Boundary Layer Ingestion Propulsion Benefit for Transport Aircraft”, Journal ofPropulsion and Power, Vol. 33, No. 5, pp. 1118–1129, 2017. doi:10.2514/1.B36321[analysis]
A. Uranga (USC) 36 / 37
Contact
Alejandra Uranga, Ph.D.
Gabilan Assistant Professor
Aerospace & Mechanical Engineering Dept.University of Southern California
Email: [email protected]: (213) 821 - 0846Web: uranga.usc.edu
A. Uranga (USC) 37 / 37