1
Study of CFD Variation on TransportConfigurations from the Second Drag
Prediction Workshop
Christopher L. Rumsey
S. Melissa Rivers
Joseph H. Morrison
NASA Langley Research Center
2
Methods
• CFL3D – finite volume upwind Roe– using 1-to-1 ICEM grids (thin-layer in all 3 directions)
and overset grids (thin-layer in normal direction)
– 3 turbulence models (SA, SST, EASM-ko)
– EASM-ko is nonlinear explicit algebraic stress ink-omega formulation
• OVERFLOW – finite difference upwind Roe– using overset grids (thin-layer in normal direction)
– 2 turbulence models (SA, SST)
3
Case 1 – CL=0.5, fully turbulent
1.241.23SSTOVERFLOW,overset
1.221.211.201.191.181.17SAOVERFLOW,overset
1.161.151.141.131.121.11SACFL3D,overset
1.101.9EASMCFL3D, 1to1
1.81.7SSTCFL3D, 1to1
1.61.51.41.31.21.1SACFL3D, 1to1
wbnp(f)wbnp(m)wbnp(c)wb(f)wb(m)wb(c)Turbmodel
Code / Grid
(shaded = not done yet)
4
Case 2 – drag polar, transition specified
2.502.492.482.472.462.452.44wbnp(m)
SAOV, O
2.432.422.412.402.392.382.37wbnp(m)
SACF, O
2.362.352.342.332.322.312.30wbnp(m)
SACF, 1
2.292.282.272.262.252.242.23wb(m)SAOV, O
2.222.212.202.192.182.172.16wb(m)SACF, O
2.152.142.132.12wb(m)EASMCF, 1
2.112.102.92.8wb(m)SSTCF, 1
2.72.62.52.42.32.22.1wb(m)SACF, 1
1.510-1-1.5-2-3CaseTurbCode /Grid
5
Case 3 – CL=0.5, transition specified
3.63.5SAOV, O
3.43.3SACF, O
3.23.1SACF, 1
wbnp(m)wb(m)TurbCode / Grid
6
Grid issues
ICEM 1-to-1 grids: - less than factor 2 between successive grids - no uniform refinement, e.g., similar surface point distribution between successive grids
7
Grid issues, cont’d
8
Grid issues, cont’d
ICEM 1-to-1 WBNP grids: - non-smooth - non-orthogonal in places - sudden grid spacing changes
9
Grid effect, CL=0.5fully turbulent
N-2/3
CD
0 2E-05 4E-05 6E-05 8E-050.027
0.028
0.029
0.03
0.031
0.032
0.033
0.034
0.035
0.036
1-to-1 grids, CFL3Doverset grids, CFL3Doverset grids, OVERFLOW
WB 1-to-1 & overset grids, CL=0.5
finer
coarser
N-2/3
CD
0 2E-05 4E-05 6E-05 8E-050.027
0.028
0.029
0.03
0.031
0.032
0.033
0.034
0.035
0.036
1-to-1 grids, CFL3Doverset grids, CFL3Doverset grids, OVERFLOW
WBNP 1-to-1 & overset grids, CL=0.5
finer
coarser
10
Lift curve on medium gridstransition specified
alpha
CL
-5 -4 -3 -2 -1 0 1 2-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
CFL3D, 1-to-1 SACFL3D, overset SAOVERFLOW, overset SAexp
alpha
CL
-5 -4 -3 -2 -1 0 1 2-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
CFL3D SA, 1-to-1CFL3D SA, oversetOVERFLOW SA, oversetexp
WB WBNP
11
Drag polar on medium gridstransition specified
CD
CL
0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
CFL3D, 1-to-1 SACFL3D, overset SAOVERFLOW, overset SAexp
CD
CL
0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
CFL3D, 1-to-1 SACFL3D, overset SAOVERFLOW, overset SAexp
WB WBNP
12
CM on medium gridstransition specified
CM
CL
-0.2 -0.15 -0.1 -0.05 0-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
CFL3D SA, 1-to-1CFL3D SA, oversetOVERFLOW SA, oversetexp
CM
CL
-0.2 -0.15 -0.1 -0.05 0-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
CFL3D SA, 1-to-1CFL3D SA, oversetOVERFLOW SA, oversetexp
WB WBNP
13
Turbulence model effect on dragWB, CFL3D, medium 1-to-1 grid, transition specified
CD
CL
0.015 0.02 0.025 0.03 0.035 0.04 0.045 0.05-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
SASSTEASM-koexp
14 cts
7 cts
10 cts
14
Turbulence model effect on Cptypically significant only at root station
x/c-C
p0 0.2 0.4 0.6 0.8 1-1.2
-0.8
-0.4
0
0.4
0.8
1.2
1.6
CF, 1m, CL.5, SA(ft)CF, 1m, CL.5, SST(ft)CF, 1m, CL.5, EASM-ko(ft)exp, WBNP 183
2y/B=0.847
x/c
-Cp
0 0.2 0.4 0.6 0.8 1-1.2
-0.8
-0.4
0
0.4
0.8
1.2
1.62y/B=0.15
15
Overset grid density effect on Cplargest effects seen at root and tip stationsbiggest change between med & fine grids
x/c
-Cp
0 0.2 0.4 0.6 0.8 1-1.2
-0.8
-0.4
0
0.4
0.8
1.2
1.62y/B=0.15
x/c-C
p0 0.2 0.4 0.6 0.8 1-1.2
-0.8
-0.4
0
0.4
0.8
1.2
1.6
OV, Of, CL.5, SA(ft)OV, Om, CL.5, SA(ft)OV, Oc, CL.5, SA(ft)exp, WBNP 183
2y/B=0.847
16
Grid type effect on Cpdramatic difference at station inboard of nacelle
x/c
-Cp
0 0.2 0.4 0.6 0.8 1-1.2
-0.8
-0.4
0
0.4
0.8
1.2
1.62y/B=0.331
x/c-C
p0 0.2 0.4 0.6 0.8 1-1.2
-0.8
-0.4
0
0.4
0.8
1.2
1.6
CF, 1m, CL.5, SA(ft)CF, Om, CL.5, SA(ft)OV, Om, CL.5, SA(ft)exp, WBNP 183
2y/B=0.847
17
Grid type effect on nacelle Cpdifference at inboard location
X/C
-Cp
0 0.25 0.5 0.75 1-1.6
-1.2
-0.8
-0.4
0
0.4
0.8F = 60 (outboard nacelle)
X/C
-Cp
0 0.25 0.5 0.75 1-1.6
-1.2
-0.8
-0.4
0
0.4
0.8F = 300 (inboard nacelle)
X/C
-Cp
0 0.25 0.5 0.75 1-1.6
-1.2
-0.8
-0.4
0
0.4
0.8
CF, 1m, CL.5, SA(ft)CF, Om, CL.5, SA(ft)OV, Om, CL.5, SA(ft)exp, WBNP 183
F = 180 (lower nacelle)
18
Lower-wing pressure contoursoverset 1-to-1
19
Lower wing streamlinesoverset 1-to-1
20
Upper wing streamlinesoverset 1-to-1
21
T.E. wing-root bubble size
BLBUB (y-loc width)
FS
BU
B(x
-lo
cl e
ng
th)
80 85 90 95 100 105 110
140
150
160
170
180
190
200
210
open=wbfilled=wbnp
square=SA CFL3Dtriangle=SST CFL3Ddelta=EASM CFL3Ddiamond=SA OVERFLOW
small=1-to-1large=overset
Wing root separation bubble width and length
more separated
less separated
22
Summary of various effects on dragat CL=0.5
dr a
gco
un
t s( 1
ct=
0.0
00
1i n
CD)
0
2
4
6
8
10
12
14
16
18
20
Grid type Code Turb TransGrid type Code Turb Trans
WB WBNP
23
Summary• Grid has a big effect on physics (inboard of nacelle),
although this effect may be masked if only look atintegrated quantities
• Provided “b2b-icem” 1-to-1 grids not suitable for grid study(too little variation); WBNP grid of poor quality
• SA turbulence model tends to give the most wing rootseparation, EASM the least
• SST turbulence model drag generally lowest; by 7-15 cts
• Individual effects due to grid type, code, turb model, andtransition at CL=0.5: 5-10 cts for WB, 10-15 cts for WBNP
24
Typical CPU timingsfor the DPW-II cases
• Typically 2000-5000 cycles needed per case• CFL3D
– 60e-6 sec/gridpt/cycle/node PC cluster (1-to-1)– 30e-6 sec/gridpt/cycle/node PC cluster with Myrinet (1-to-1)– 60-110e-6 sec/gridpt/cycle/node Alpha cluster with Myrinet
(1-to-1 faster than overset for multi-processors)– 230e-6 sec/gridpt/cycle/node Origin 2000 cluster (1-to-1)– 17e-6 sec/gridpt/cycle Cray SV1 single processor
• OVERFLOW– 30e-6 sec/gridpt/cycle/node PC cluster– 75e-6 sec/gridpt/cycle/node Origin 3000 cluster (mlp version)