far-field drag breakdown applied to the dlr-f6 configuration · 2003. 8. 27. · 4turbulence model...
TRANSCRIPT
-
2nd CFD Drag Prediction Workshop
21st of June 2003
Far-field drag breakdownapplied to the DLR-F6 configurationIn the framework of the 2nd CFD Drag Prediction Workshop
2nd CFD DPW
Presented by
Stéphane AmantAerodynamic methods engineer
-
21st of June 20032nd CFD Drag Prediction Workshop Page 2
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
Context of the topic
• To tackle the issue of aircraft performance4 Optimal shape design fi
need for reliable drag breakdown4 Cruise performance assessment fi
need for accurate overall drag level
• Prevailing role of CFD4 Generation of well adapted grids
4 Need for a breakthrough in drag assessment– information provided by near-field integration insufficient
– far-field abilities very attractive in terms of• improvement of the accuracy• physical drag breakdown• spatial distribution of drag sources
-
21st of June 20032nd CFD Drag Prediction Workshop Page 3
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
Tools used by Airbus for the 2nd DPW
• Structured multiblock solver : elsA4Developed by ONERA4Oriented-object structure (C++)4Turbulence model : k-w (Wilcox)4 Centred scheme4Dissipation : Jameson type scalar scheme4Implicit time integration : LU-SSOR method
• Far-field drag analysis tool : FFD414developed by ONERA / Airbus France
(van der Vooren / Destarac‘s theory)4industrialized by Airbus France4daily used by shape designers and aerodynamic data
engineers4capability for dealing with patched grids (soon AMR grids)
-
21st of June 20032nd CFD Drag Prediction Workshop Page 4
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
Comparison elsA / TAU of the drag polars(WB & WBPN configurations)
0,10
0,20
0,30
0,40
0,50
0,60
180,0 230,0 280,0 330,0 380,0 430,0
Cd (*10000.)
Cl Cd_total_WB (wind tunnel)
Cd_total_WB (elsA - ICEM grid)
Cd_total_WB (elsA - Airbus grid)
Cd_total_WB (TAU - medium)
Cd_total_WB (TAU - fine)
Cd_total_WBPN (wind tunnel)
Cd_total_WBPN (elsA - ICEM grid)
Cd_total_WBPN (TAU - medium)
Cd_total_WBPN (TAU - fine)
Cd_total_WBPN (elsA; -1,0°; fully turb.)
Drag polar (near-field integration)
10 drag counts
with forced transition
fully turbulent
-
21st of June 20032nd CFD Drag Prediction Workshop Page 5
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
Cp distributions(elsA, medium ICEM grid, WBPN, CL=0.50)
with fixed transitionfully turbulent
Wing
Nacelle
-
21st of June 20032nd CFD Drag Prediction Workshop Page 6
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
Physical and numerical drag breakdown
( ) iwvAfp DDDDD ++=+
( ) spD
spi
appiwvAfp DDDDDDD
i
++++=+43421
VWVV
• Exact near-field/far-field balance4no small disturbances assumption
• Numerical considerations4production of spurious drag
(connected to entropy variations in the flowfield)
4decay of axial vorticity in the wing-tip vortexfi transformation of induced drag into spurious drag
-
21st of June 20032nd CFD Drag Prediction Workshop Page 7
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
Far-field analysis
• Comments4lower overall far-field drag fi removal of the spurious drag
4quite high discrepancy on the viscous terms fi
turbulence model ? role of the grid size ?
4 good agreement on the wave drag magnitude
4uncertainty on the induced drag : correction of the vortex decay ?
4incremental drag better predicted than the near-field for elsA– TAU : DCD = 41,2.10-4
– elsA : DCD = 45,9.10-4
SolverGridCDfCDvpCDwCDiCDtotalTAU8 750 330 nodes125.363.31.389.6279.5WBelsA5 715 968 cells(ICEM)135.056.00.991.2283.5TAU12 321 068nodes149.177.24.889.6320.7WBPNelsA8 290 304 cells(ICEM)159.173.24.692.5329.4
-
21st of June 20032nd CFD Drag Prediction Workshop Page 8
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
Wave drag distribution
Excellent agreementon the local drag rise !
Valuable plotfor design improvement !
-
21st of June 20032nd CFD Drag Prediction Workshop Page 9
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
Viscous drag distribution
The magnitude of the root separationin terms of drag is identically predicted !
Excellent agreement on the local drag rise !
-
21st of June 20032nd CFD Drag Prediction Workshop Page 10
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
Drag breakdown from elsA computations(WB & WBPN configurations)
0,30
0,35
0,40
0,45
0,50
0,55
0,60
0,0 20,0 40,0 60,0 80,0 100,0 120,0 140,0 160,0 180,0
Cd (*10000.)
Cl
Cd_wave (WB)
Cd_wave (WBPN)
Cd_friction (WB)
Cd_friction (WBPN)
Cd_viscous_pressure (WB)
Cd_viscous_pressure (WBPN)
Cd_induced (WB)
Cd_induced (WBPN)
Drag polar breakdown (elsA computations)
Issue withthe enforced transition ?
Contribution of the lower side(interaction wing / pylon / nacelle)
Low impact of the engine installationon the induced drag
-
21st of June 20032nd CFD Drag Prediction Workshop Page 11
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
[email protected]° : an example of analysis
Wing root separation
Wing/Pylon separation
AoA = -1.0° AoA = 0.0°
Physical intuition confirmed
by the far-field analysis
fi very valuable information
provided by FFD41
-
21st of June 20032nd CFD Drag Prediction Workshop Page 12
© A
IRB
US
FR
AN
CE
S.A
.S. T
ous
droi
ts r
éser
vés.
Doc
umen
t con
fiden
tiel.
Conclusions
• Far-field drag analysis tools = an intelligent means of makingthe most of CFD4shape design improvement (including optimisation)
4identification of CFD issues (grids, solvers, etc.)
4New sensors for automatic refinement
• Physical drag breakdown is available for both Airbus solvers(elsA & TAU)4discrepancies remain to be addressed
4some obvious satisfactory trends
• The one-drag count accuracy : a utopia4solver able to capture the flow features
4well-built grid
4far-field drag assessment tool
the 1 d.c. variationis at hand