BACHELOR’S DEGREE:

AEROSPACE VEHICLES ENGINEERING

AUTHOR:
JUAN HITA GARCÍA

TITLE:
TECHNICAL STUDY OF A MODEL PLANE FOR THE AIR CARGO
CHALLENGE

DIRECTOR:
DANIEL GARCÍA ALMIÑANA

COURSE:
2014-2015 Q2

CONTENT OF THIS DOCUMENT:

ATTACHMENTS
 SUMMARY OF CONTENTS

LIST OF FIGURES ............................................................................................ 3

1. XFLR5 TUTORIAL ........................................................................................ 4

2. CATIA WING MODELLING ........................................................................... 8

2.1. Generating airfoil spline .......................................................................... 8

2.2. Generating surfaces................................................................................ 9

2.3. Generating ribs........................................................................................ 9

3. SCORE SCRIPT ......................................................................................... 11

4. ANALYTICAL STRUCTURAL SPREADSHEET .......................................... 15

Juan Hita García Attachments 2

 LIST OF FIGURES

Fig. 1 E205 airfoil imported to XFLR5 ............................................................... 4

Fig. 2 Multi-threaded analysis window............................................................... 5
Fig. 3 E205 airfoil 2D polars at different Reynolds ............................................ 5
Fig. 4 Wing geometry window ........................................................................... 6
Fig. 5 3D polars for fixed lift analysis ................................................................. 7
Fig. 6 Airfoil spline in Catia ................................................................................ 8
Fig. 7 Wing surface in Catia .............................................................................. 9
Fig. 8 Ribs generated through surface intersection ......................................... 10
Fig. 9 Wing product in Catia ............................................................................ 10
Fig. 10 Geometry spreadsheet ........................................................................ 17
Fig. 11 Materials and structural geometry ....................................................... 18
Fig. 12 External loads imported from XFLR5 .................................................. 19
Fig. 13 Stress calculation ................................................................................ 20

Juan Hita García Attachments 3

 1. XFLR5 TUTORIAL

In order to start analyzing a wing, the latest version of the software shall be
downloaded from the following web:

http://sourceforge.net/projects/xflr5/files/

Once the software is launched, it is necessary to import the airfoil data. On

the following page there are multiple airfoils classified according to their family
and purpose. Moreover, coordinates file can be downloaded in a .txt or .dat file
which we will be imported to XFLR.

http://m-selig.ae.illinois.edu/ads/coord_database.html

The next step is to start 2D analysis of the airfoil. Click FILE – DIRECT FOIL
DESIGN. Next, import airfoil data FILE – OPEN (select .txt or .dat). It is important
to check that the shown geometry is a closed curve. Otherwise, the results will not
converge. In that case, try downloading the .dat file from another site or just adjust
manually the points.

Fig. 1 E205 airfoil imported to XFLR5

Next, click FILE – XFOIL DIRECT ANALYSIS. This tool allows to obtain
polars for different Reynolds and angles of attack. Click ANALYSIS – MULTI-
THREADED BATCH ANALYSIS. This option provides faster calculation because
does not shown graphic iterations.

Juan Hita García Attachments 4

 Fig. 2 Multi-threaded analysis window

It must be pointed out that the range of Reynolds number chosen is critical
later. As we have mentioned earlier, interpolations for 3D analysis are based on
this 2D data. This means that every point of the span corresponds to different chord
and freestream speed, hence having its own Reynolds. For instance, tip chord is
typically lower than root chord, which means Reynolds will be lower at the tip. If
the previous 2D analysis has not included this Reynolds range, interpolation will
not be able to be done presenting errors on the solver window. The same problem
can occur if the alpha range is not large enough, due to 3D induced angle.

Fig. 3 E205 airfoil 2D polars at different Reynolds

Juan Hita García Attachments 5

 As it can be seen in the graphs, low Reynolds curves are not as accurate
as the higher Reynolds one, since they are more difficult to converge, while higher
Reynolds present continuous behavior. It is important to avoid this low Reynolds
zones where there is no realistic data and efficiency is lower. Due to this fact,
attention must be paid when choosing the tip chord or other tapered sections.

Once 2D data has been obtained, proceed with the 3D module. FILE –
WING AND PLANE DESIGN. Click PLANE – DEFINE NEW PLANE. First of all,
wing geometry must be defined. Fin and elevator can be included, but for a first
analysis and specially to choose which airfoil and geometry fits better our mission
requirements, these are not necessary.

Typical wing parameters include a planar wing at the central part (without
tapering) to facilitate manufacturing, no torsion and dihedral is an option if we are
looking for stability. On the following figure, there is an example of geometry. It is
important to choose a representative number of panels and its distribution. There
should be a higher number of panels in these sections where pressure distribution
has a higher gradient, which are the leading edge and the tip chord (cosine
distribution at X-panels and –sin distribution at Y-panels).

Fig. 4 Wing geometry window

The next step is to obtain 3D data. Click ANALYSIS – DEFINE AN

ANALYSIS. Select Fixed lift options in order to calculate wing polars at different
operating conditions and a constant load factor (for instance, n=1 is lift=weight,
which allows a level flight). In the same line to 2D analysis, a large range of alpha

Juan Hita García Attachments 6

 will be selected as the software itself will discard cases were analysis cannot be
converged. To understand how the program works, when the maximum alpha
converged is exceeded, we are asking the software to obtain a possible operating
point that cannot be achieved because maximum CL has been reached. Therefore,
under a certain speed the aircraft cannot meet the lift required to keep the level
flight. As it has been in the report (6.1.2), the equation follows approximately the
following expression, hence the different points belong to each combination of
speed and alpha.

1
𝐿 = 𝜌𝑉∞ 2 𝑆(𝐶𝐿 𝑜 + 2𝜋𝛼) (Eq. 1)
2

Fig. 5 3D polars for fixed lift analysis

In the case of studying aerodynamics during take-off, fixed-angle option

would be chosen instead of fixed-lift. This option provides polars at a constant
angle of attack. These points can be used later for a performance study.

Juan Hita García Attachments 7

 2. CATIA WING MODELLING

2.1. Generating airfoil spline

So as to import the geometry from the .txt or .dat file used in the
aerodynamic analysis to our CAD file, in this case Catia is used since it is the most
common software used in the industry, therefore it is an added value for the team
members learning to handle the different tools provided.

First step is to open the following excel file C:\Program files(x86)\Dassault

Systemes\B19\intel_a\code\command\GSD_PointSplineLoftFromExcel.xls. The
following table () shows how to import .dat points corresponding to X Y Z
components. Obviously, these coordinates are related to chord= 1mm. To import
an airfoil whose chord has a different value it is possible to scale the profile inside
the CAD software, but working with 1mm spline leads to instability. In order to
avoid this problem, it is highly recommended to scale the airfoil in Excel before
importing to Catia.

Using a macro included in this Excel, points from the .dat file are
introduced in the Part Body as a Geometrical set (notice that this operation is only
possible within the Part module). Click VIEW – MACRO – FEUIL1.MAIN and
execute. A window will appear asking what kind of geometry it is intended to be
created (points or spline), type 2 and continue.

Fig. 6 Airfoil spline in Catia

Juan Hita García Attachments 8

 2.2. Generating surfaces

Next step is to import of scaled airfoil sections (for instance, root chord
could be 2000mm and tip chord 1500mm). After including this geometrical sets
start the Wireframe and Surface Design module, using Multi-sections surface
command different profiles will be joined.

Fig. 7 Wing surface in Catia

2.3. Generating ribs

Once all the surfaces have been generated with the surface module, open
Part design module and define planes where ribs will be placed. So as to generate
them, intersect in a sketch every surface with this plane and pad the resulting
profile. Note that flaps and ailerons must be separated bodies.

Juan Hita García Attachments 9

 Fig. 8 Ribs generated through surface intersection

After completing all the operations, this is the final wing product.

Fig. 9 Wing product in Catia

Juan Hita García Attachments 10

 3. SCORE SCRIPT

%Global variables

rho=1.225; %[kg/m3]
g=9.81; %[m/s2]
Umax=26; %[m/s] Max speed considered in the analysis
Umin=12; %[m/s] Min speed considered in the analysis
%N=input('Enter the number of divisions to determine the analysis
resolution: ');
N=60;
MaxSpeed=0;

% THRUST DATA APC 13x6.5 8500rpm

% Thrust and Speed values obtained from graph for APC 13x6.5 found at:
% http://mypages.iit.edu/~vural/RC%20Airplane%20Design.pdf
TGraph_T=[16.87 11.97 10 7.16 4.22 0.88]; % [N]
TGraph_U=[10.3 14.9 17 19.7 22.4 25.5]; % [m/s]
% We fit (approximate) a first order polynomial to the thrust and
% convert it into an analytic function of velocity:
Tfit=polyfit(TGraph_U,TGraph_T,1);
T_analytic=@(U) (Tfit(1)*U+Tfit(2));

%CONSTANT LOAD FACTOR TURN

%CALCULATION OF MAXIMUM SPEED AT TURN START

% We read a csv file containing Drag vs Speed points at n=1 from an

XFLR5, type 2 simulation "Fixed Lift" imposing L=W (Load factor n=1)
[DGraph_D,DGraph_U]=importDrag(DGraph);
% Like with Thrust, we do another polynomial fit for the drag
coefficient:
Dfit=polyfit(DGraph_U,DGraph_D,6);
D_analytic=@(U)
(Dfit(1)*U.^6+Dfit(2)*U.^5+Dfit(3)*U.^4+Dfit(4)*U.^3+Dfit(5)*U.^2+Dfit
(6)*U+Dfit(7));

% We calculate a speed vector contained between Umin and Umax with N

divisions

for i=1:N
% This is the y-coordinate where our speed 'curve' cuts the y-axis
Uo=Umin-((Umax-Umin)/(N-1));
% Our speed vector U(i) in function of the #iteration and
resolution
U(i)=i*(Umax-Umin)/(N-1)+Uo;
end;

% The following algorithm calculates the maximum speed for the

aircraft in horizontal forward uniform flight condition by imposing
T=D
for i=Umin:0.01:Umax
if(((T_analytic(i)-D_analytic(i))<0.1)&&((T_analytic(i)-
D_analytic(i))>(-0.1)))

Juan Hita García Attachments 11

 MaxSpeed=i;
i=Umax;
end
end
MaxSpeed; %Show MaxSpeed in m/s

% The following plots are used to check the previous calculations

% plot(DGraph_U,DGraph_D,'ro') % Graph data for Drag-speed
% hold on
% plot(U,D_analytic(U),'b') % Analytic exp. for Drag-speed
% plot(TGraph_U,TGraph_T,'ro') % Graph data for Thrust-speed
% plot(U,T_analytic(U),'b') % Analytic exp. for Thrust-speed

%CALCULATION OF TURN TIME IMPOSING A N>1 LOAD FACTOR

% New Drag vs Speed calculation for a n=2,5 load factor

% To do this in XFLR5, do perform a 'fixed lift' analysis and multiply
the airplane's mass by n=2
[DGraph2_D,DGraph2_U]=importDrag(DGraph2);
Dfit=polyfit(DGraph2_U,DGraph2_D,6);
D_analytic2=@(U)
(Dfit(1)*U.^6+Dfit(2)*U.^5+Dfit(3)*U.^4+Dfit(4)*U.^3+Dfit(5)*U.^2+Dfit
(6)*U+Dfit(7));

% Turn conditions
roll=acos(1/n); % [rad]
roll_deg=(360/(2*pi))*acos(1/n); % [º]
% To calculate alpha we make an approximation in fact alpha is a
function
% of speed but since it only affects the thrust*cos(alpha) calculation
and it's a low angles approximation there's no need to waste time
importing
% the full data... just plot alpha vs speed on XFLR5 and check what
would be the worst scenario alpha when leaving the turn and use that
for the calculations.

%Initial conditions:
TURN_TIME=0; % Turn time counter initialized at 0 [s]
V0=MaxSpeed; % Initial turn speed [m/s]
R0=V0/(n*g*sin(roll)); % Initial turn radius [m]
deltaT=R0/V0*(pi/(1500)); % Initial time step [s]

% The following algorithm calculates the turn data in an incremental

fashion
for i=1:1500
% Drag has 1,4 factor to account for fuselage, tail and landing
gear
Vnew=V0+((T_analytic(V0)*cos(alpha)-
1.4*D_analytic2(V0))/m)*deltaT; % New tangential speed.
Rnew=V0.^2/(n*g*sin(roll)); % New turn radius

deltaT=((R0+Rnew)/(V0+Vnew))*(pi/(1500)); % Time Step

TURN_TIME=TURN_TIME+deltaT;
R0=Rnew;
V0=Vnew;
end

Juan Hita García Attachments 12

Vfinal_TURN=V0; % [m/s]
Rfinal_TURN=R0; % [m]
DeltaV=V0-MaxSpeed; % [m/s]
TURN_TIME; % [s]

% The following plots are used to check the difference between drag vs
speed at N=1 and N=2,5
plot(DGraph_U,DGraph_D,'ro') % Graph data for Drag-speed
hold on
plot(U,D_analytic(U),'b') % Analytic exp. for Drag-speed
plot(DGraph2_U,DGraph2_D,'ro') % Graph data for Drag-speed
plot(U,D_analytic2(U),'b') % Analytic exp. for Drag-speed
xlabel('U_i_n_f [m/s]');
ylabel('D [N]');
title('Drag vs Speed at n=1, n=2');
hold off

%NON CONSTANT STRAIGHT HORIZONTAL FLIGHT

%Initial conditions:
FWD_TIME=0; % Forward flight time counter initialized at 0 [s]
V0=V0; % Initial speed [m/s]
X0=0; % Initial position [m]
deltaT=2/(10*V0); % Initial Time Step [s]

% The following algorithm calculates the straight segment data in an

incremental fashion
while 1
if X0>100
break;
end

Vnew=V0+((T_analytic(V0)-D_analytic(V0))/m)*deltaT; % New
tangential speed.
Xnew=X0+((V0+Vnew)/2)*deltaT; % New x-position

deltaT=2/(10*(V0+Vnew)); % new Time Step

FWD_TIME=FWD_TIME+deltaT;
X0=Xnew;
V0=Vnew;
end

Vfinal_FWD=V0; % Final velocity at 100m segment end [m/s]

Xfinal_TURN=X0; % Final position [m]
FWD_TIME; % Time used to cover the 100m [s]

%FLIGHT SCORE CALCULATOR

TIME=0;
legs=0;
while TIME<120
TIME=TIME+FWD_TIME;
if TIME<120
legs=legs+1;
TIME=TIME+TURN_TIME;
end

Juan Hita García Attachments 13

end

TIME;
legs;
FlightScore=(PL*2)*(legs+3);

end

Juan Hita García Attachments 14

4. ANALYTICAL STRUCTURAL SPREADSHEET

This spreadsheet is divided in several pages. First of all, the user must
introduce the airfoil data and wing parameters (root chord, tip chord). On the
second page, torsion box parameters are introduced (x positions, thickness…). A
visual sketch of the structure is automatically calculated.

Afterwards, external loads are imported from XFLR5 and the spreadsheet
calculates wing stresses.

Regarding shear calculations, first of all, shear center is determined

through the imaginary cut in the closed profile.

Using forces equilibrium in vertical axis:

𝑄𝑧 = 𝑞′1 ℎ1 + 𝑞2 ℎ2 (Eq. 2)

From shear flow equation:

𝑄𝑧
𝑞′1 = 0 − 𝑧 𝐴 (Eq. 3)
𝐼𝑧 1𝑑𝑜𝑤𝑛 𝑟𝑒𝑖𝑛𝑓

Since there are no forces in x axis, 𝑞′3 = 0. Next step is to set the condition
of no rotation when Qz is applied in the shear center:

𝑞𝑑𝑠 (Eq. 4)
∮ =0
𝐺𝑡

ℎ1 ℎ2 1 1 ℎ2 ℎ1
𝑞𝑜 ( + + ( + ) 𝑑) + 𝑞′2 − 𝑞′1 = 0 (Eq. 5)
𝑡1 𝑡2 𝑡3 𝑡4 𝑡2 𝑡1

Juan Hita García Attachments 15

 Now, applying momentum equilibrium in x=x1, it is possible to find the
shear center location, 𝑥𝑠𝑐 .

𝑄𝑧 𝑥𝑠𝑐 = 𝑞𝑜 ℎ2 ℎ3 + 𝑞2 ℎ2 ℎ3 (Eq. 6)

Finally, shear stress due to torsion is calculated and it is possible to obtain

the total distribution of shear stresses along the walls.

1
𝑇𝐴 𝑄𝑧 (𝑥𝑐𝑝 − 𝑥𝑠𝑐 ) + 𝐶𝑚 2 𝜌𝑣 2 𝑆𝑐 (Eq. 7)
𝑞𝑡𝑜𝑟𝑠𝑖𝑜𝑛 = =
2𝐴𝑖𝑛𝑡 2𝐴𝑖𝑛𝑡

𝑞𝑡𝑜𝑡𝑎𝑙 = 𝑞𝑠ℎ𝑒𝑎𝑟 + 𝑞𝑡𝑜𝑟𝑠𝑖𝑜𝑛 (Eq. 8)

Juan Hita García Attachments 16

 Fig. 10 Geometry spreadsheet

Juan Hita García Attachments 17

 Fig. 11 Materials and structural geometry

Juan Hita García Attachments 18

 Fig. 12 External loads imported from XFLR5

Juan Hita García Attachments 19

 Fig. 13 Stress calculation

Juan Hita García Attachments 20