MEC4447 Assignment 2

Yian Wang 32501765

Abstract
The NACA6409 airfoil and JAXA airfoil are used to predict the flow around airfoils.
The lift and drag coefficient are determined at different angle of attack for NACA6409
airfoil. The lift against angle of attack and drag polars are plotted. The lift coefficient
from simulation prediction matches with experimental data, but the drag coefficient
is greater than experiment data. The boundary layer resolution is sufficient as y+
value is close to 1. The mesh is refined for retracted JAXA airfoil and compound JAXA
airfoil by grid adaption. The take off weight of A320 from calculation is 74.12 tonnes,
which is close to actual take off weight of A320 (75 tonnes).

Introduction
The airfoil is used to generate lift for an aircraft. The flow behaviour of airfoil can be
predicted by CFD software as Fluent. The NACA6409 with 1m chord length and JAXA
airfoil with 0.7115m chord length is provided as airfoil profile data for simulation. The
area of airflow for NACA6409 airfoil is a semicircle with 25m radius from trailing edge
of airfoil and rectangle with 50m width and 11m length from the semicircle. The area
of airflow for JAXA airfoil is a semicircle with 10m radius from trailing edge of airfoil
and rectangle with 20m width and 20m length from the semicircle. The semicircle
and two adjacent sides are set as velocity inlet, the bottom side is set as pressure
outlet and the airfoil is set as wall. The magnitude of velocity is 1m/s at velocity inlet
for NACA6409 airfoil. The velocity at velocity inlet is 60 m/s for JAXA airfoil. The k-
omega Shear-stress transport (SST) turbulence model is used for both two cases with
convergence criteria of 1e-5.

Aim
 To perform CFD simulation on NACA6409 airfoil and JAXA airfoil.
 To determine the lift and drag coefficient of NACA6409 airfoil at different
angle of attack.
 To determine the take off weight with JAXA airfoil using refined mesh.

Method
Part 1 NACA6409 (Appendix 3)

The geometry is set up as introduction. The density of fluid is change to 1 kg/m 3. The
Reynold number of 60,000, 100,000 and 200,000 are used for simulation. The
viscosity is 1.667e-5 Ns/m2, 1e-5 Ns/m2 and 5e-6 Ns/m2 respectively from calculation
(Equation 1). The angle of attack from -2 up to 10 degrees with 2 degrees steps are
simulated for airfoil. The velocity is set by velocity component in x and y direction
separately. The reference density is set as 1 kg/m3. The model for each case is set in
Workbrench (Appendix 1). The edge sizing with 200 division and bios factor of 5
toward centre is used at airfoil. The edge sizing with 100 division and bios factor of
50 towards the x-axis is set from 0.5 m to x-axis. The edge sizing with 100 division
and bios factor of 20 towards the airfoil is used for 10m after the airfoil. The second-
order upwind momentum equation, first-order upwind turbulent kinetic energy and
first-order upwind specific dissipation rate are used for discretisation schemes. The
relaxation parameter is set to 0.25 for pressure and momentum, 0.5 for density and
body forces, 0.275 for turbulent kinetic energy and specific dissipation rate and 0.25
for turbulent viscosity. The number of iterations is approximately 450 for low angle of
attack and approximately 350 for high angle of attack. The lift and drag coefficient are
obtained from force report.

ℜ=ρUc / µ Equation 1

Part 2 JAXA airfoil

The geometry is set up as introduction. The standard air is used in simulation. The
reference velocity is set as 60 m/s. The model for each case is set in Workbrench
(Appendix 2). The mesh is refined by grid adaption for 3 times. The cells that greater
than 0.1 in the gradient of velocity magnitude is registered to refinement criteria.

Retracted JAXA airfoil (Appendix 4)

The edge sizing with 200 division and bios factor of 5 toward centre is used at airfoil
and the inflation layer of 10 is set from airfoil surface for retracted JAXA airfoil. The
second-order upwind momentum equation, first-order upwind turbulent kinetic
energy and first-order upwind specific dissipation rate are used for discretisation
schemes of retracted JAXA airfoil. The relaxation parameter is 0.25 for pressure and
momentum, 0.5 for density, body force and turbulent viscosity and 0.75 for turbulent
kinetic energy and 0.275 for specific dissipation energy. The number of iterations is
150 before refinement and 140 after refinement for retracted JAXA airfoil.

Compound JAXA airfoil (Appendix 5)

The edge sizing with 2mm element size is used at airfoil and the inflation layer of 7
and transition ratio of 0.35 is set from airfoil surface for compound JAXA airfoil. The
first-order upwind momentum equation, turbulent kinetic energy and specific
dissipation rate are used for discretisation schemes of compound JAXA airfoil. The
relaxation parameter is set to 0.25 for pressure and momentum, 0.5 for density and
body forces, 0.275 for turbulent kinetic energy and specific dissipation rate and 0.25
for turbulent viscosity. The number of iterations is 130 before refinement and 400
after refinement for compound JAXA airfoil. The sectional lift is obtained from force
report.

Discussion
Part 1 NACA6409

Figure 1: contour plot of static pressure at 6 degrees angle of attack

The contour plot of static pressure with 6 degrees angle of attack is labelled (figure
1). The maximum static pressure is occurred at the leading edge of airfoil. The upper
surface of airfoil has lower static pressure compared to lower surface of airfoil. The
minimum static pressure is occurred at the leading edge of upper surface and the
maximum static pressure is occurred at the leading edge of lower surface. The lift of
airfoil is caused by the pressure difference between the upper and lower surface of
airfoil.

Figure 2: lift coefficient against angle of attack at Re=60,000

Figure 3: lift coefficient against angle of attack at Re=100,000

Figure 4: lift coefficient against angle of attack at Re=200,000

The lift coefficient against angle of attack for different Reynold number is plotted
(figure 2, 3, 4). The lift coefficient has a positive linear trend with angle of attack. The
simulation predictions of lift coefficient have similar result compared to experiment
for Reynold number at 60,000, 100,000 and 200,000. The experimental result has
slightly higher lift coefficient at large angle of attack compare to simulation result.
Figure 5: drag polar at Re=60,000

Figure 6: drag polar at Re=100,000

Figure 7: drag polar at Re=200,000

The drag polars are plotted for different Reynold number (figure 5, 6, 7). The plots for
simulation result have c shape, which is similar to experimental plots. The simulated
drag coefficient of airfoil is much greater than the experimental drag coefficient at
100,000 and 200,000 Reynold number. The simulated drag coefficient of airfoil is
much smaller than the experimental drag coefficient at 60,000 Reynold number.
Figure 8: y+ along lower surface of airfoil

Figure 9: y+ along upper surface of airfoil

The y+ value of lower surface is approximately 1 (figure 8). The y+ value for upper
surface decreasing from 1.5 and reach 0 around 0.8 (figure 9). The y+ value then
increased from 0 to 0.5. The flow separation is occurred around 0.8 as the y+ value
reaches 0. The boundary layer resolution of airfoil is sufficient as the theoretical y+
value for k-omega SST model should be 1.

Part 2 JAXA airfoil

Figure 10: initial mesh of compound airfoil

Figure 11: final refined mesh of compound airfoil

Figure 12: initial mesh of retracted airfoil

Figure 13: final refined mesh of retracted airfoil

The quality of mesh is refined by grid adaption for both retracted JAXA airfoil (figure
12, 13) and compounded JAXA airfoil (figure 10, 11). The amount of mesh at rapid
velocity change is increased after mesh refinement. The area of mesh that located
close to airfoil is decreased and the accuracy of simulation is increased. The sectional
force is approach constant after 3 iteration of mesh refining. The predicted take-off
weight for A320 is 74.12 tonnes form calculation (Appendix 6). It is close to actual
take-off weight for A320 (75 tonnes).

Conclusion
The NACA6409 airfoil and JAXA airfoil are simulated by Ansys Fluent. The maximum
static pressure is occurred at the leading edge of bottom surface of airfoil and the
minimum static pressure is occurred at the leading edge of upper surface of airfoil
from static pressure plot (figure 1). The lift is generated by the difference of static
pressure between upper and bottom surface of airfoil. The lift and drag coefficient of
NACA6409 airfoil is determined. The lift coefficient has a positive linear trend with
angle of attack (figure 2, 3, 4). The plots of predicted results match with experiment
data. The drag polars of predicted result have similar c-shape plot with experiment
data (figure 5, 6, 7). However, the simulation prediction has greater drag coefficient
compared to experiment data. The y+ is plotted at 6 degree angle of attack (figure 8,
9). The boundary layer resolution is sufficient as y+ from simulation is close to
theoretic y+ value for k-omega SST model (1). The quality of mesh is refined by
adaptive for JAXA airfoil (figure 10, 11, 12, 13). The predicted take-off weight for
A320 is 74.12 tonnes, which is close to actual take-off weight for A320 (75 tonnes).

Appendix
Appendix 1: screen shot of the workbench for NACA airfoil

Appendix 2: screen shot of the workbench for JAXA airfoil

Appendix 3: NACA6409 airfoil

8

Appendix 4: Retracted JAXA airfoil

Appendix 5: compound JAXA airfoil

Appendix 6: Take off weight calculation

3505.1816∗0.5292
LJAXA = ∗4.6=1192.6 N
0.7115

( )
2
35.8
L A 320 =1192.6∗ =726379 N
4.6

L A 320 726379
m= = =74120 kg=74.12 tonnes
g 9.8