CFD JAXA CFD CFD Navier-Stokes -1- -2-1
CFD CFD Navier-Stokes 1. 2. 2,500,000 CFD Copyright Boeing CFD Boeing B777 HP -3- -4-2
CFD European Transonic Wind tunnel (-163 ) JAXA 2mx2m CFD (Computational Fluid Dynamics) CFD CFD E. N. Tinoco AIAA 98-2512 -5- -6-3
: *=10 7 10 8, 0.8, 0.15 C 1/300,000C < L < 50C M =0.75, Re=6.5x10 6 0.02% C 2% C Re = )()() ( ) = ( ) () Navier-Stokes Stokes Navier-Stokes 5 ρ ρu Q = ρv, ρw E ρu 2 ρu + p τ xx F = ρvu τ xy, ρwu τ zx T ρhu ( τ u) x k x Q F G H + + + = 0 t x y z ρv ρuv τ xy 2 ρv + p τ yy G =, ρwv τ yz T ρhv ( τ u) y k y ρw ρuw τ xz H = ρvw τ yz 2 ρw + p τ zz T ρhw ( τ u) z k z -7- -8-4
Reynolds Navier-Stokes(RANS) - - - 2 6 7 Euler 5 1 2 Euler Prandtl-Glauret TRANAIR(1989) Boeing 90CFD TRANAIR F. T. Johnson et. al AIAA 2003-3439 -9- -10-5
CFD CFD CFD CFD - E. N. Tinoco AIAA 98-2512 F. T. Johnson et. al AIAA 2003-3439 CFD CFD CFD Navier-Stokes -11- -12-6
CFD CFD Navier-Stokes Navier-Stokes Stokes 15 1992 6 10 CFD ONERA M5 3 2 VP2600, VP400 43 523 5 100 15 CFD Capability Computing ONERA M5-13- -14-7
5 2001 (2001. 6) ONERA M5 650 98 SGI Onix3400 14CPU 12 11 - ONERA M5 NSIII NSIII AIAA Drag Prediction Workshop II (2003) DLR-F6 1% 0.5% Fujitsu PRIMEPOWER2500 100CPU 8001300 10 1 2 10 10% 5% 1 DLR-F6-15- -16-8
AIAA Drag Prediction Workshop 3 (2006) JAXA 1000 1 3003000 3-17- -18-9
0.5% 1 Capability Computing 1000 1000 C CD D 0.0295 0.0290 0.0285 0.0280 0.0275 0.0270 CFD CFD Navier-Stokes 0.0265 UPACS WB UPACS FX2B 0.0260 TAS WB TAS FX2B 0.0255 0.0E+00 1.0E-05 2.0E-05 3.0E-05 4.0E-05 5.0E-05 6.0E-05 N ^(-2/3) 1/N^(2/3) -19- -20-10
CFD CFD CFD 3 RANS RANS 5000 B747 Airbus A300-21- CFD,, 3 2-22- 11
CFD CFD European EUROLIFT 233 AIAA 2004-767 NASA Langley CFD Web JAXA CFD 100 Trapezoidal wing models JAXA CFD Code : TAS (Tohoku Univ. Aerodynamic Simulation Code) 6 Spalart-Allmaras 60m/s Re=2.1x10 6 FTFHLD JAXA NSIII (PRIMEPOWER HPC2500) Upper Surface Unstructured hybrid mesh Lower Surface -23- JAXA wind tunnel model B.L. transition locations (AoA=10deg) -24-12
(C L -α) CFD 3 2.5 C L 2 Exp. (Corrected) Comp. (Fully Turb.) 1.5 0 5 10 15 20 Angle of attack Exp. α=10.55 Comp.(w/oFTF) α=10 Exp. α=15.54 Comp.(w/oFTF) α=15-25- -26-13
CFD CFD Navier-Stokes 10,000-27- E. N. Tinoco AIAA 98-2512 -28-14
vs Kriging ANOVA (analysis of variance) SOM (self-organizing map) -29- -30-15
CFD AIAA Applied Aerodynamics Conf. 2005 MSC. NASTRAN / Navier-Stokes solver TAS) Block Fuel Aerodynamic Model (Unstructured mesh) Adaptive Range Multi-Objective Genetic Algorithm Structural Model (Wing box: Shell elements) Modified PARSEC : 9 design variables * 3 cross sections : 6 spanwise locations : 2 positions at kink and tip 35-31- -32-16
1 1 ARMOGA Individual # 1 Individual # 8 Static Aeroelastic Analysis Module Euler comp. @ =Const. No Subsonic Static Analysis Displacement Converge? Yes Subsonic @ 3conditions Euler eq. N-S eq. 11 3000 CFD Aerodynamic Model ite=ite+1 ite=1 Transonic Euler comp. @ =Const. No Static Analysis Displacement Converge? Yes Transonic @ 5conditions Structural Model if (ite=1) then Strength/Flutter Optimization endif Euler comp. @ =Const. Static Analysis Displacement Converge? Yes Off-design @ 3conditions N-S S Comps. - Evaluation of Three Objective Functions Block Fuel Max Takeoff Weight C D divergence (MDD) -Check of Constraints MLD Fuel Quantity Block Fuel Module Individual # i Off-design No -33-3 : Nondominated solution 4counts initial solution C D divergence Block Fuel [kg] 10kg 50kg Max Takeoff Weight [kg] -34-17
SOM Block Fuel Max Takeoff Weight C D Divergence N-S N-S1 N-S 100 CFD -35- -36-18
/NEDO Navier-Stokes CFD 10 RANSCFD CFD 100 RANS Large Eddy Simulation -37-19