More Related Content Similar to akaydinD8asmeImece2016 Similar to akaydinD8asmeImece2016 (20) akaydinD8asmeImece20167. /27 7
How does BLI (boundary layer inges>on) work?
Uranga et al., Preliminary Experimental Assessment of the Boundary Layer Inges(on Benefit for the D8 Aircra@, AIAA-2014-0906
Drela, Development of the D8 Transport Configura(on, 2011, AIAA-2011-3970
Drela, Power Balance in Aerodynamic Flows, 2009, AIAA Journal 47(7)
Pk = Φjet+ Φwake+ Φfuse. + Φvortex + Ė = Φ
PK = p0∞−p0( )
!
V ⋅
!
n dS
propulsor
"∫∫ Φ = p0∞−p0( )
!
V ⋅
!
n dS
wake plane
− prop. reg.
"∫∫ + 1
2
V−V∞( )
2
ρ
!
V ⋅
!
n dS
wake plane
"∫∫
With BLI, the terms Φjet , Φwake and Ė become smaller.
Consider two configura>ons at cruise (zero-net force):
⇥⇤⌅
⇥⇤⌅⇧
⌃⌥ ⌦
⇥
⇥⇤⌅
⇤ ⌅⇥
⌃⌥ ⌦
⇧
⇥⇤⌅
↵ ⇥⇤ ⇤✏
⇣⌥ ⌃⌥ ⌦
◆ ⇥⇤ ⌅⇧⌃⇥ ⌥ ⇥ ⌦↵ ⇤⇥ ⇥⇥⇤⌅
⌃⌥ ⌦
⌥⌘⇧ ⌥⌅ ✓◆
⌫ ⌘⌥
⌫ ⌘⌥ ⇠⇥ ⇡⌥⇢
⇠⇥ ⇡⌥⇢⇣⌅⌥
⇣⌅⌥
⌧⌥⇡
⌧⌥⇡
⇥⇤⌅⇧⌃⌥ ⇥⇤ ⌦↵ ✏⇥⇤
⇧⌃⌥ ⇥⇤ ⌦↵ ✏⇥⇤
⌫⇠⇡ ⇢⇥⇤ ⌧ ⇤⌦ ⇡ ! "##
⌃⌥ ⌦↵ ✏⇣ ⌘✓⌘⌘ ◆ ↵ ⌫⇠⇡⇠⇢
Ė
Ė
Podded (non-BLI) configura>on
Integrated (BLI) configura>on
8. /27 8
Experimental and Computa>onal Tools
Pandya, External Aerodynamics Simula(ons for the MIT D8 “Double-Bubble” Aircra@ Design, 2012, ICCFD7-4304
EFD: NASA Langley 14x229 Wind Tunnel. 3D-printed skin and modular components
Overflow 2.2, NASA’s workhorse flow solver for structured curvilinear overset grids.
Chimera Grid Tools (CGT) for to generate these CFD grids.
CFD:
Nichols and Buning, Users Manual for OVERFLOW 2.2, heps://overflow.larc.nasa.gov
Chan, Gomez, Rogers and Buning, Best Prac(ces in Overset Grid Genera(on, 2002, AIAA-2002-3191
M∞≈0.1
13. /27 13
EFD and CFD complement each other.
⇥⇤⌅⇧⌃⇧⌃⌥ ⌦↵ ⌅ ✏⇣ ⌘ ✓ ◆ ⇤ ⌦ ⌫⇠
⇥⇤ ⌅⇧⌃⌥
⌦⇥↵
✏⇣⌘✓◆ ⌫✓⇣◆⇠⇡⇢ ⇡⇠⇡
⇥ ⇥⇤⇥ ⌅⇤⌅ ⇧ ⌃
⇡⇢⇢ ⇢✏⌧ ⇡◆⇠✓⇠✓
⇥ ⇥⇤⇥ ⌥⌅ ⇤⌅ ⇧ ⌃
⇡◆⇠✓⇠✓ ✏ ✓◆⇠⌫ ⇠
⇤ ⌦⌅ ↵
!⇧" ⌦#⇥⌥⌅" "⇥
"⇥"⌅⌥ ⌦⌅$ ⌃⌦!%
!⇧" ⌃⌦⌅"!⇥⇧
⌅⌦ ⌅
!⇧" ⌃⌦⌅"
⇠✏⇠⇡⇢ ⌫⇡&
⌫'( ⇡⇠⇡
⇥ ⇧ ⌃
⇠⇡⇠✓ ⌫⇡&
⌫'( ⇡⇠⇡
⇧ ⌃
⇠⇡⇠✓ )⌫✓⇠✏⇠⇡⇢ )⌫✓
*+ , + + ,
*+ ,
+
+ ,
⌅⌃
✏
*+ -
+
+ -
+ .
/ 0
/ ,
⇣⇤ ⌅⌃
⌘
⇧
✓
⇥⇤ ⌅⇧⌃ ⌥ ⌦↵ ↵ ✏ ⌃ ⇣↵⌘✓↵◆ ⌫⇠⇠
Pandya, External Aerodynamics Simula(ons for the MIT D8 “Double-Bubble” Aircra@ Design, 2012, ICCFD7-4304
Greitzer et al., Aircra@ and Technology Concepts for an N+3 Subsonic Transport – Phase 2. Year 3 Status Review, 2014
Local flow direc>ons on the rake survey
plane is determined by CFD
PK = p0∞−p0( )
!
V ⋅
!
n dS
propulsor
"∫∫
14. /27 14
EFD and CFD complement each other.
Pandya et al., Computa(onal Assessment of the Boundary Layer Inges(ng Nacelle Design of the D8 Aircra@, AIAA-2014-0907
⇤⌅⇧⌃⌥⇤⌅ ⌃⌦↵ ⌦⌃ ⇥⇧⌅ ⇤⌅ ✏ ⌦⇣
⇤ ⌅⇧⌃ ⌥ ⌦↵ ↵ ✏ ⌃ ⇣↵⌘✓↵◆
⇥
⇤⌅⇧⌃ ⌥ ⌥⌦↵ ✏⇣⌃ ⌥ ⌥⌦↵ ⌘⇤ ⌥ ✓↵⌅
◆ ⌃↵ ⌥ ⌅ ⌦ ⌅ ⌅ ⌅⌫⌃ ⇥
⇥⇥ ⇥⇥
⇤
⇥⌘⌥⇥
⇤⌅⇥⌘⌥⇥
⇥⇤⌅ ⇥⇤⇧ ⇥⇤⌃ ⇥ ⇥⇤⌃ ⇥⇤⇧ ⇥⇤⌅
⇥
⇥⇤⌃
⇥⇤⇧
⇥⇤⌅
⇥⇤⌥
⇤⌃
⇥⇤ ⌦ ⇥⇤⌦ ⇥⇤⌃⌦ ⇥
⇥
⇥⇤⌃
⇥⇤⇧
⇥⇤⌅
⇥⇤⌥
⇤⌃
↵
✏⇣⌘✓
⇥⇤⌥
⇥⇤⌅
⇥⇤⇧
⇥⇤⌃
⇥
⇥⇤⌅ ⇥⇤⇧ ⇥⇤⌃ ⇥ ⇥⇤⌃ ⇥⇤⇧ ⇥⇤⌅
⇤⌅⇥⌘⌥⇥
⇥⌘⌥⇥
⇤⌅⇥⌘⌥⇥
⇥⇤⌅ ⇥⇤⇧ ⇥⇤⌃ ⇥ ⇥⇤⌃ ⇥⇤⇧ ⇥⇤⌅
⇥
⇥⇤⌃
⇥⇤⇧
⇥⇤⌅
⇥⇤⌥
⇤⌃
⇥⇤ ⌦ ⇥⇤⌦ ⇥⇤⌃⌦ ⇥
⇥
⇥⇤⌃
⇥⇤⇧
⇥⇤⌅
⇥⇤⌥
⇤⌃
⇥⇤⌥
⇥⇤⌅
⇥⇤⇧
⇥⇤⌃
⇥
⇥⇤⌅ ⇥⇤⇧ ⇥⇤⌃ ⇥ ⇥⇤⌃ ⇥⇤⇧ ⇥⇤⌅
⇤⌅⇥⌘⌥⇥
⌥⌅ ⇢⌅⌫⌃
⇤
⇢⌥⌦⌃⌃ ⌫
⇤
⇠⇡ ⌅↵↵⌅⌫⌃ ⇢⌥ ⌫⇥⇤⌅⇧⌃⌥⇤⌅ ⌃⌦↵ ⌦⌃ ✏⇣⌘⇤ ✓ ⌦◆
⇥⇤ ⌅⇧⌃ ⌥ ⌦↵ ↵ ✏ ⌃ ⇣↵⌘✓↵◆
⇥
⇤⌅⇧⌃ ⌥ ⌥⌦↵ ✏⇣⌃ ⌥ ⌥⌦↵
◆ ⌃↵ ⌥ ⌅ ⌦ ⌅ ⌅ ⌅⌫⌃ ⇥
⇥⇥
⇤
⌅ ⇧⌃ ⌅ ⇧⌥ ⌅ ⇧⇤ ⇧⇤ ⇧⌥ ⇧⌃⌅ ⇧⌃ ⌅ ⇧⌥ ⌅ ⇧⇤ ⇧⇤ ⇧⌥ ⇧⌃
⌅ ⇧⌃
⌅ ⇧⌥
⌅ ⇧⇤
⇧⇤
⇧⌥
⇧⌃
⇥⌥⇥
⇤⌅⇥⌥⇥ ⇤⌅⇥⌥⇥
⌅ ⇧⌥
⇧⌥
⇧⇣
⇥⇧⇤
⇥⇧⌃
⌅ ⇧⌃ ⌅ ⇧⌥ ⌅ ⇧⇤ ⇧⇤ ⇧⌥
⌅ ⇧⌃
⌅ ⇧⌥
⌅ ⇧⇤
⇧⇤
⇧⌥
⇧⌃
⇥⌥⇥
⇤⌅⇥⌥⇥
⌅ ⇧⌥
⇧⌥
⇧⇣
⇥⇧⇤
⇥⇧⌃
⌅ ⇧⌥ ⌅ ⇧⇤ ⇧⇤ ⇧⌥ ⇧⌃
⇤⌅⇥⌥⇥
⇧⌃ ⌅ ⇧⌃
⇠⇡⌥⌅ ⇢⌅⌫⌃
⌃ ⇤
⌧ ⇢⌥⌦⌃⌃ ⌫
⌃ ⇤
⌧ ⇧⌫ ⌅⇧⇧⌃⌅
Greitzer et al., Aircra@ and Technology Concepts for an N+3 Subsonic Transport – Phase 2. Year 3 Status Review, 2014
EFD:
CFD:
Inlet Outlet
L R
Total pressure coefficients. L: Le9 propulsor, R: Right propulsor. Views from rear of the model.
L R
15. /27 15
CFD and EFD reassures confidence in results: ⇧⌃⌥ ⇥ ⇥⌦ ↵ ✏ ⇣⌘⌅ ⌃⌥⇤ ⇧ ✓ ◆
⇥⇤ ⌅⇧⌃⇧⌥ ⌃ ⇧⌦↵ ⇥ ✏
⌘ ✓◆◆ ⌦ ⌃ ⇧ ⌫⇧⌦ ↵⇧⌃ ⇠↵⇧ ⇡ ⇢ ⌧ ⌘ ✓◆◆ ⌦ ⌃
⇥ ⇤ ⇥ ⌅ ⇥ ⇧ ⇥ ⌃ ⇥⌥ ⇥⌥⇤
⇥
⇥ ⌅
⇥ ⌦
⇥ ⇤
⇥ ⌥
⇥ ⌥
⇥ ⇤
⇥ ⌦
⇥ ⌅
⇥
↵
↵
✏⇣⌘⌘✓⌘◆
⌫✓⇠⇡⇢⌫✓⌘◆
✏⇣⌘⌘✓⌘◆↵ ⌧
⌫✓⇠⇡⇢⌫✓⌘◆↵ ⌧
⌥ ⌦↵ ↵ ✏ ⌃ ⇣↵⌘✓↵◆ ⌃ ⌫⌫
ΔCPk= -9%
ΔCPk= -8%
cruise
EFD predicts an 8% ± 0.7%
reduc>on in power to sustain cruise.
CFD predicts a 9% reduc>on.
A very good agreement between experimental and computa>onal
predic>ons of BLI benefit:
Uranga et al., Preliminary Experimental Assessment of the Boundary Layer Inges(on Benefit for the D8 Aircra@, AIAA-2014-0906
Greitzer et al., Aircra@ and Technology Concepts for an N+3 Subsonic Transport – Phase 2. Final Review, Dec. 2014
Pandya et al., Computa(onal Assessment of the Boundary Layer Inges(ng Nacelle Design of the D8 Aircra@, AIAA-2014-0907
16. /27 16
CFD facilitates in-depth and alternate analysis methods
Uranga et al., Preliminary Experimental Assessment of the Boundary Layer Inges(on Benefit for the D8 Aircra@, AIAA-2014-0906
Drela, Development of the D8 Transport Configura(on, 2011, AIAA-2011-3970
Drela, Power Balance in Aerodynamic Flows, 2009, AIAA Journal 47(7)
Pk = Φjet+ Φwake+ Φfuse. + Φvortex + Ė = Φ
PK = p0∞−p0( )
!
V ⋅
!
n dS
propulsor
"∫∫ Φ = p0∞−p0( )
!
V ⋅
!
n dS
wake plane
− prop. reg.
"∫∫ + 1
2
V−V∞( )
2
ρ
!
V ⋅
!
n dS
wake plane
"∫∫
Φ can be computed in separate zones in the wake planes
to analyze Φ in terms of Φjet, Φwake, Φfuse … etc.
⇥⇤⌅
⇥⇤⌅⇧
⌃⌥ ⌦
⇥⇤⌅
⇤ ⌅⇥
⌃⌥ ⌦
⇥⇤⌅
↵ ⇥⇤ ⇤✏
⇣⌥ ⌃⌥ ⌦
⇥⇤⌅
⌃⌥ ⌦
⌥⌘⇧ ⌥⌅ ✓◆
⌫ ⌘⌥
⌫ ⌘⌥ ⇠⇥ ⇡⌥⇢
⇠⇥ ⇡⌥⇢⇣⌅⌥
⇣⌅⌥
⌧⌥⇡
⌧⌥⇡
⇥⇤⌅⇧⌃⌥ ⇥⇤ ⌦↵ ✏⇥⇤
⇧⌃⌥ ⇥⇤ ⌦↵ ✏⇥⇤
⌫⇠⇡ ⇢⇥⇤ ⌧ ⇤⌦ ⇡ ! "##
⌃⌥ ⌦↵ ✏⇣ ⌘✓⌘⌘ ◆ ↵ ⌫⇠⇡⇠⇢
Ė
Ė
Podded (non-BLI) configura>on
Integrated (BLI) configura>on
Understanding the losses:
20. /27 20
CFD helps plan and improve the experiments
⇥ ⇤ ⌅⇧⌃⌥ ⌦⌅↵⇥ ⇥ ✏⇣⇧⌃ ✏⌅⌘ ⇣✓ ⇣⇧⌃ ✓⇥ ◆ ⌦↵ ⌦ ⌥✓ ↵
⌫ ◆◆ ◆ ⌦↵ ⌦ ⌥✓ ↵ ⇠⌅ ↵⌅⇣✓ ◆ ⇥ ⌅⇡ ⇣◆ ⇣⇧⇥ ↵⌅⇢⇥⇣ ⇧ ✏⇣⇥ ✏⇣⇧⌃ ✏⌅⌘
⇥⇤ ⌅⇧⌃ ⌥ ⌦↵ ↵ ✏ ⌃ ⇣↵⌘✓↵◆
⌧
⇤
Greitzer et al., Aircra@ and Technology Concepts for an N+3 Subsonic Transport – Phase 2. Year 3 Status Review, 2014
Wing wake would par>ally
ingested by podded propulsor
at α=6° but not at α=2°.
A suitable dihedral angle θ for the
pods were found by simula>ng
mul>ple models (not by building
and tes>ng mul>ple models).
θ=0° θ=30° … ( and other θ angles simulated)
23. /27 23
Summary & Discussion
0 5 10 15
α
0.5
1
1.5
2
CL
WT
Overflow - Spalart-Allmaras
Overflow - SST
0 5 10 15
α
0.5
1
1.5
2
CL
WT
Overflow - Spalart-Allmaras
Overflow - SST
0 5 10 15
α
0.5
1
1.5
2
CL
WT
Overflow - Spalart-Allmaras
Overflow - SST
0 5 10 15
α
0.5
1
1.5
2
CL
WT
Overflow - Spalart-Allmaras
Overflow - SST
⇥⇤⌅⇧⌃⌥ ⇥ ⇥⌦ ↵ ✏ ⇣⌘⌅ ⌃⌥⇤ ⇧ ✓ ◆
⇥⇤ ⌅⇧⌃⇧⌥ ⌃ ⇧⌦↵ ⇥ ✏
⇣⌘ ✓◆◆ ⌦ ⌃ ⇧ ⌫⇧⌦ ↵⇧⌃ ⇠↵⇧ ⇡ ⇢ ⌧ ⌘ ✓◆◆ ⌦ ⌃
⇥ ⇤ ⇥ ⌅ ⇥ ⇧ ⇥ ⌃ ⇥⌥ ⇥⌥⇤
⇥
⇥ ⌅
⇥ ⌦
⇥ ⇤
⇥ ⌥
⇥ ⌥
⇥ ⇤
⇥ ⌦
⇥ ⌅
⇥
↵
↵
✏⇣⌘⌘✓⌘◆
⌫✓⇠⇡⇢⌫✓⌘◆
✏⇣⌘⌘✓⌘◆↵ ⌧
⌫✓⇠⇡⇢⌫✓⌘◆↵ ⌧
⇥⇤ ⌅⇧⌃ ⌥ ⌦↵ ↵ ✏ ⌃ ⇣↵⌘✓↵◆
ΔCPk= -9%
ΔCPk= -8%
cruise
• In D8 project, u>liza>on of EFD and CFD methods goes beyond cross-check & valida>on.
CL
24. /27 24
Summary & Discussion
⇥⇤⌅⇧⌃⇧⌃⌥ ⌦↵ ⌅ ✏⇣ ⌘ ✓ ◆ ⇤ ⌦ ⌫⇠
⇥⇤ ⌅⇧⌃⌥
⌦⇥↵
✏⇣⌘✓◆ ⌫✓⇣◆⇠⇡⇢ ⇡⇠⇡
⇥ ⇥⇤⇥ ⌅⇤⌅ ⇧ ⌃
⇡⇢⇢ ⇢✏⌧ ⇡◆⇠✓⇠✓
⇥ ⇥⇤⇥ ⌥⌅ ⇤⌅ ⇧ ⌃
⇡◆⇠✓⇠✓ ✏ ✓◆⇠⌫ ⇠
⇤ ⌦⌅ ↵
!⇧" ⌦#⇥⌥⌅" "⇥
"⇥"⌅⌥ ⌦⌅$ ⌃⌦!%
!⇧" ⌃⌦⌅"!⇥⇧
⌅⌦ ⌅
!⇧" ⌃⌦⌅"
⇠✏⇠⇡⇢ ⌫⇡&
⌫'( ⇡⇠⇡
⇥ ⇧ ⌃
⇠⇡⇠✓ ⌫⇡&
⌫'( ⇡⇠⇡
⇧ ⌃
⇠⇡⇠✓ )⌫✓⇠✏⇠⇡⇢ )⌫✓
*+ , + + ,
*+ ,
+
+ ,
⌅⌃
✏
*+ -
+
+ -
+ .
/ 0
/ ,
⇣⇤ ⌅⌃
⌘
⇧
✓
⇥⇤ ⌅⇧⌃ ⌥ ⌦↵ ↵ ✏ ⌃ ⇣↵⌘✓↵◆ ⌫⇠⇠
• In D8 project, u>liza>on of EFD and CFD methods goes beyond cross-check & valida>on.
• The methods also complement each other: CFD facilitates the duty of EFD and vice versa.
27. /27 27
Summary & Discussion
⇥⇤⌅⇧⌃⇧⌃⌥ ⌦↵ ⌅ ✏⇣ ⌘ ✓ ◆ ⇤ ⌦ ⌫⇠
⇥⇤ ⌅⇧⌃⌥
⌦⇥↵
✏⇣⌘✓◆ ⌫✓⇣◆⇠⇡⇢ ⇡⇠⇡
⇥ ⇥⇤⇥ ⌅⇤⌅ ⇧ ⌃
⇡⇢⇢ ⇢✏⌧ ⇡◆⇠✓⇠✓
⇥ ⇥⇤⇥ ⌥⌅ ⇤⌅ ⇧ ⌃
⇡◆⇠✓⇠✓ ✏ ✓◆⇠⌫ ⇠
⇤ ⌦⌅ ↵
!⇧" ⌦#⇥⌥⌅" "⇥
"⇥"⌅⌥ ⌦⌅$ ⌃⌦!%
!⇧" ⌃⌦⌅"!⇥⇧
⌅⌦ ⌅
!⇧" ⌃⌦⌅"
⇠✏⇠⇡⇢ ⌫⇡&
⌫'( ⇡⇠⇡
⇥ ⇧ ⌃
⇠⇡⇠✓ ⌫⇡&
⌫'( ⇡⇠⇡
⇧ ⌃
⇠⇡⇠✓ )⌫✓⇠✏⇠⇡⇢ )⌫✓
*+ , + + ,
*+ ,
+
+ ,
⌅⌃
✏
*+ -
+
+ -
+ .
/ 0
/ ,
⇣⇤ ⌅⌃
⌘
⇧
✓
⇥⇤ ⌅⇧⌃ ⌥ ⌦↵ ↵ ✏ ⌃ ⇣↵⌘✓↵◆ ⌫⇠⇠
• In D8 project, u>liza>on of EFD and CFD methods goes beyond cross-check & valida>on.
• The methods also complement each other: CFD facilitates the duty of EFD and vice versa.
• Ul>mately, this interplay between the CFD and EFD shortens the design cycle and reduces tes>ng costs.
↵ ⇥⌃ ⌃ ⌅✏ ⌅⇤ ⇣⌘⇥✓
⇥ ✏⇣⇧⌃ ✏⌅⌘ ⇣✓ ⇣⇧⌃ ✓⇥ ◆ ⌦↵ ⌦ ⌥✓ ↵
⇠⌅ ↵⌅⇣✓ ◆ ⇥ ⌅⇡ ⇣◆ ⇣⇧⇥ ↵⌅⇢⇥⇣ ⇧ ✏⇣⇥ ✏⇣⇧⌃ ✏⌅⌘
⌫⇠⇠
Acknowledgements: Thanks to MIT team (Edward Greitzer, Mark Drela, Alejandra Uranga, Arthur Huang and
David Hall) for technical discussions and many of the material presented here. Compu>ng resources for CFD
simula>ons were provided by NASA Advanced Supercompu>ng Division at NASA Ames Research Center.
This work is funded by NASA AATT (Advanced Air Transport Technologies) Project