Lecture Notes: Aerospace Structural Design

Copyright Notice

© 1997-2005 by Jafar Eskandari Jam, IAU

All rights reserved. Permissions are granted for
personal and educational uses only. Any other uses of
these lecture notes (such as for classroom lectures
outside the Islamic Azad University, trainings
elsewhere, and those of a commercial nature) are not
permitted, unless such uses have been granted in
writing by the author.

______________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
Lecture Notes: Aerospace Structural Design

Course Outline

Loads On Aircraft
The different types of loads exerted on an aircraft are briefly explained. These loads include both
aerodynamic and manoeuvre loads.

Function of Structural Components

An aircraft is made from many different components. Each component has a specific function but form an
integral whole.

Bending Of Beams With Non-Symmetrical Cross Section

Establishment of sign convention for an arbitrary beam section. The Effective Bending Moment and the
Effective Shear Forces are defined and derived. An example is given.

Stress, Strain and Displacement Relationship for Open and Closed Single
CellThin-Walled Beams
Concept of Shear Flow is introduced. The Shear Centre of an open beam section, twist and warping of a
closed beam section are explained through an example.

Torsion of Thin Walled Beam Sections

The application of a torque on a closed or open section beam produces warping. This section attempts to
illustrate the characteristics of such a distribution by using an open beam section.

Structural Idealization
Idealization of a wing section is performed to calculate the stresses on the stringers and shear on the
skin.This can be done using nodes to represent the stringers. Spread sheet examples are given.

Multicell Beams
The wing section can be divided into multiple cells with applied torques. Method of determining Rate of
Twist and Shear Flow distribution is shown in an example.

Tapered Beams
A real wing on an aircraft is often swept and tapered. This section shows the method of determining the
shear flow on an idealized tapered wing.

______________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
Lecture Notes: Aerospace Structural Design

Chapter 1

Loads on Aircraft

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
1
Lecture Notes: Aerospace Structural Design

INTRODUCTION

An aircraft is basically required to support two types of basic loads:

1) Ground Loads: Encountered by the aircraft during movement on the ground; ie:
taxying, landing, towing, etc

2) Air Loads: Loads exerted onto the structure during flight by the manoeuvres
carried out by the aircraft or by wind gusts (such as wind shear).

As well as these, other role specific loads may be generated by the aircraft, ie:

• High Altitude Flying : Pressurised cabin,

• Amphibious aircraft : Landing on water,
• Military Aircraft : High Speed Manoeuvres, Withstand considerable damage

These two loads classes may be further divided into:

A) Surface Loads : Act on the surface of the structure, such as aerodynamic

or hydrostatic loads, and

B) Body forces : Act over the volume of the structure and are generated by
gravitational and inertia effects.

AERODYNAMIC SURFACE LOADS

All flying aircraft flying under steady flight, manoeuvre or gust conditions experience
pressure distributions on the surface of the skin. The resultants of these pressures cause direct
loads such as: (bending, shear and torsion in all parts of the structure).

A conventional aircraft consists of a fuselage, a pair of wings, and a tail plane (Horizontal &
Vertical Tail section).

The fuselage carries crew, payload, passengers, cargo, weapons, or fuel. The wings provide
lift and the tailplane contributes to directional control.

As well you have ailerons, elevators and a rudder which enable the aircraft to be controlled,
and flaps provide extra lift during take-off and landing.

The force on an aerodynamic surface (Wing, vertical & horizontal tail) results from a
differential pressure distribution caused by incidence, camber or both.

For a typical wing, the chord wise pressure distribution looks like this:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
2
Lecture Notes: Aerospace Structural Design

Figure 1: Pressure distribution and resultant forces around airfoil

This wing with incoming wind direction as shown creates this pressure distribution. By
integrating the pressure in the vertical and horizontal components about the wind direction
the following terms can be obtained:

A) The vertical Lift Force component (L), perpendicular to the wind direction.

B) The horizontal Drag component (D), parallel to the wind direction.

Both these resultants act through the centre of pressure (CP) of the airfoil. But as this point
will move depending on the attitude of the airfoil to the incoming wind, the lift and drag
forces are moved to act about the aerodynamic centre (AC) (the quarter chord). This means
that an extra Moment (M) needs to be included to keep the system in equilibrium.

Looking at the wing in a span wise view, the lift distribution about the AC looks something
like Figure 2, where the wing Lift resultants forces act as shown.

Now that the pressures on the surfaces of an aircraft can be converted to loads, the overall
loads on an aircraft due to symmetrical manoeuvres need to be investigated.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
3
Lecture Notes: Aerospace Structural Design

Figure 2: Typical lift distribution for a wing fuselage combination

SYMMETRIC MANOEUVRE LOADS

Level Flight

Figure 3: Aircraft Loads in level flight

where:
LW : Lift acting at aerodynamic centre of wing

D : Aircraft Drag

LT : Horizontal Tail loading acting at AC of tail

T : Engine Thrust (parallel to direction of flight) (No vector thrusting)

W : Mass of aircraft acting at centre of gravity (= Mg)

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
4
Lecture Notes: Aerospace Structural Design

Accelerated Flight (Rapid Pull-out From Dive)

Increased downward load on horizontal tail, increases lift load causing upward acceleration
normal to the flight path.

This causes the load factor 'n' to be greater than 1. Creating inertia load on the structure to be
nMg, where:

n = 1 + V2/Rg

R = Radius of curvature of the flight path.

Steady Banked Turn

The aircraft flies in a horizontal turn without any sideslip and at a constant speed.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
5
Lecture Notes: Aerospace Structural Design

where:

n = secφ;

tanφ = V2/gR, for θ = 60o, n = 2

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
6
Lecture Notes: Aerospace Structural Design

Static Ground Condition

Figure 6: Aircraft Static ground loads

where:

RNose : Ground reaction at nose wheel

RMain : Main undercarriage ground loads

W : Mass of aircraft acting at centre of gravity (= Mg)

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
7
Lecture Notes: Aerospace Structural Design

Chapter 2

Function of Structural
Components

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
8
Lecture Notes: Aerospace Structural Design

FUNCTION OF STRUCTURAL COMPONENTS

The aircraft structure has a dual purpose:

a) It transmits and resists all loads applied to it, &

b) It acts as a cover to maintain the aerodynamic shape and protect its content.

The most efficient structure to achieve this is a thin shell structure supported by longitudinal
stiffeners and transverse frames, known as a semi-monocoque structure.

Figure 7: Typical semi-monocoque structural component. (a) Body, (b) Aerodynamic surface

The functions of the different structural members which go on to make an aircraft are as
follows:

SKIN

1) It transmits aerodynamic forces to the longitudinal and transverse members.

2) It develops shear stresses to react the applied torsional moments

3) Together with longitudinal members it resists axial loads and applied bending

SPAR WEB

Similar to (2) for the skin

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
9
Lecture Notes: Aerospace Structural Design

STRINGERS OR LONGERONS (with larger cross sectional area)

1) Resist bending and axial loads

2) Divide the skin into small panels

3) Act with skin to resist axial loads caused by pressurisation

SPAR CAPS

Similar to (1) and (2) of stringers.

FRAMES OR RINGS

These are the transverse members of the body, if they cover most of the cross section are
called bulkheads

1) Maintain cross sectional shape

2) Distribute concentrated loads

3) Establish column length to prevent buckling

4) Provide edge restraint for skin.

5) Act with skin to resist circumferential loads

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
10
Lecture Notes: Aerospace Structural Design

Chapter 3

Bending of Beams With Non

Symmetrical Cross Sections

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
11
Lecture Notes: Aerospace Structural Design

INTRODUCTION

The majority of aircraft structural components consist of beams with non-symmetrical cross
section acting in bending . For this reason an expression needs to be derived to allow for the
determination of the stresses induced by bending moments to such sections.

SIGN CONVENTION AND NOTATION

Look at the Oxyz system of axis, with an arbitrary beam parallel to the z-axis:

Figure 8 : Notation and sign convention for positive forces, moments and displacements

Where :

T = Torque

M = Bending Moment

S = Shear Force

w = Distributed load

P = Axial or direct load

u,v,w = Axial displacements

All of these externally applied loads are positive in the direction indicated in the figure.
Internal moment and forces applied to face A are in the same direction and sense as
externally applied loads. However on face B, the positive internal moments and forces are in
the opposite sense.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
12
Lecture Notes: Aerospace Structural Design

RESOLUTION OF BENDING MOMENTS

A bending moment M applied in any plane parallel to the z-axis can be resolved into Mx and
My components by normal vector rules.

By doing it in a visual way it will be easier to see:

Figure 9: Resolved bending moment about x and y axis

From Figure 9, the following relationships can be obtained:

Mx = M sinθ

My = M cosθ

and that these moments can have different sign depending on the value of θ. For example
if θ>π/2, Mx is positive and My is negative.

STRESS DISTRIBUTION DUE TO BENDING (ETB-NonSym)

Consider a beam of arbitrary non-symmetrical cross section, which supports bending

moments Mx and My, bending about some axis in the cross section. This is the plane of no
bending stress, called Neutral Axis (N.A.).

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
13
Lecture Notes: Aerospace Structural Design

Figure 10: Determination of Neutral Axis location

Let the axis origin coincide with the centroid G of the cross section, and that the neutral axis
is a distance p from G.

The direct stress σz on element dA at point (x,y) and distance ξ from the neutral axis is:

Look at the beam in a plane parallel to the neutral axis with two segments ij and kl which are
of equal length when the beam is undeflected:

Figure 11: Side view of undeflected beam with segments ij and kl marked

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
14
Lecture Notes: Aerospace Structural Design

Once the beam has been deflected this section will look like this:

Figure 12: Deflected beam

where:

R = the radius of curvature

dθ = angle between planes ik and jl

The strain in plane kl can be defined as:

with

and

giving

By substituting back into the stress equation it gives:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
15
Lecture Notes: Aerospace Structural Design

(3.2)

Now that a stress equation has been obtained, it is necessary to satisfy both rotational and
linear equilibrium at the ends of the beam. That is, the summation of the forces through the
depth of the beam is equal to '0', and the summation of the moments through the depth of the
beam is equal the applied moments Mx and My.

As the beam supports pure bending, the resultant load on the end section must be zero. Hence

Substituting equation 3.2 gives:

This equation defines the location of the centroid of the section; it follows that the neutral
axis must pass through the centroid.

Rather than use this equation to find the location of the centroid, it is much easier to locate
the centroid about the xy-axis by using equation 3.3.

In order to have moment equilibrium, it is necessary to re-draw Figure 10 but with the axis
passing through the centroid.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
16
Lecture Notes: Aerospace Structural Design

Figure 13: Beam section with Neutral Axis passing through centroid

In order to see this in more detail, Figure 14 shown a close up of the axis and the area dA.

Figure 14: Detail of area dA in beam cross-section.

If the inclination of the Neutral Axis (N.A.) is at an angle from the x-axis then:

ξ = x sin( α ) + y cos( α ) (3.4)

and substituting into 3.2 gives:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
17
Lecture Notes: Aerospace Structural Design

(3.5)

The moment resultants have the same sense as the applied moments, hence:

, (3.6)

and substituting equation 3.4 into equations 3.5 gives:

and

Both the sin(α) and cos(α) terms are not a function of dA, so they can be removed from the
integration. What remain are terms which only have to do with the characteristics of the cross
sectional shapes of the beam, and these are just the 2nd moments of area of the beam.

The second moments of area about the xy axes are:

which gives:

solving simultaneously gives,

Substituting into Equation (3.5) gives:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
18
Lecture Notes: Aerospace Structural Design

By defining the terms Effective Bending Moment, as:

and

Equation 3.8 can be re-written as follows:

Note that if the beam is symmetrical about the x -axis or y-axis then:

and the x & y axes are the principal axes.

POSITION OF NEUTRAL AXIS

The location of the Neutral Axis was defined by Equation (3.3).

For a beam with a symmetrical cross section, the centroid is the point defined by equation 3.3
and the Neutral Axis is parallel to the x and y axis. For a non-symmetrical beam cross section
however, the Neutral Axis passes at some angle α with respect to the x-axis. What needs to
be done define the angle of the neutral axis.

At the N.A. the normal bending stresses are equal to σz = 0, giving that:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
19
Lecture Notes: Aerospace Structural Design

where:

xN.A. and yN.A. are coordinates of points along the neutral axis, giving:

By taking the inverse tan of the angle 'α', the angle of the Neutral Axis with respect to the x-
axis can be found, given by equation 3.12.

Example 1: The beam shown is subjected to a bending moment of 150Nm about the x-axis.
Calculate the maximum direct stress due to bending stating where it acts.

Figure 15: Beam cross section with applied bending moment

a) Determine the location of the centroid about point A

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
20
Lecture Notes: Aerospace Structural Design

and

b) Shift axis to centroid

Figure 16: Beam cross section with axis shifted to centroid position

c) Determine other sectional properties

d) Calculate effective bending moments

Since Mx = 1500 and My = 0, substituting these values into Equations (3.9) and (3.10) gives:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
21
Lecture Notes: Aerospace Structural Design

e) Determine bending stress equation and cross section stresses

Substitute the values calculated above into Equation (3.11).

Substituting the coordinates of all the corner points from Figure 16 give that:

Point x (mm) y (mm) σz (MPa)

1 72 21.6 0.436
2 -48 21.6 5.1
3 -48 13.6 3.9
4 -16 13.6 2.66
5 -16 -66.4 -9.32
6 -8 -66.4 -9.63
7 -8 13.6 2.34
8 72 13.6 -0.76

You can use this method to determine the stresses due to bending in any type of beam with
anytype of cross section.

APPROXIMATION FOR THIN WALLED SECTIONS

Due to the thin nature of aircraft structures, the assumption can be made that stresses are
constant through the thickness 't' of the skin. By saying this, the square and higher powers of
't' can be neglected from the computation of sectional properties.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
22
Lecture Notes: Aerospace Structural Design

To see this, look at a channel section:

Figure 17: Thin-walled channel section.

1) Because the section is symmetrical about the x-axis, then Ixy = 0

2) The second moment of area Ixx is given by:

Expanding the RHS, gives:

by eliminating t2 powers and higher it becomes:

and similarly

This indicates that the sectional properties may be calculated as if the section was represented
by a thin line, as shown in Figure 18, disregarding any t2 or higher terms.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
23
Lecture Notes: Aerospace Structural Design

Figure 18: Approximation of channel section

For this discrete thin walled section, the sectional properties would be found as follows:

1) Determine location of centroid

2) Shift axis to centroid

Figure 19: Axis at centroid

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
24
Lecture Notes: Aerospace Structural Design

3) Determining moments of inertia

These results are exactly the same as for the section considering the skin material thickness
and then disregarding all t2 and higher terms.

Since not all skin sections will lie parallel to either the x or y axis, the local moments or area
for a section of skin at an angle θ with respect to the x-axis are given by the following
equations.

Figure 20: Thin skin section inclined at

an angle θ wrt the x-axis

GENERAL LOADING RELATIONSHIPS

Consider an element of length δz from a beam with an unsymmetrical cross section with all
types of loads applied in the y-z plane, Figure 21.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
25
Lecture Notes: Aerospace Structural Design

Figure 21: Equilibrium of generally loaded beam element in zy-

plane

Equilibrium of element in y direction gives:

dividing by δz and in the limit as δz 0, this equation simplifies to:

Taking moments about A gives:

dividing by δz and in the limit as δz=0, this equation simplifies to :

Combining these two equations gives:

Similarly, about the x-z plane:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
26
Lecture Notes: Aerospace Structural Design

Consider now an element of length δz from a beam with an unsymmetrical cross section with
only an applied torque about z-axis, Figure 22.

Figure 22: Torque equilibrium of beam section of length δz

Taking moments about the z-axis:

dividing by δz and in the limit as δz=0, this equation simplifies to:

By assuming that a parameter has the same relationship to as Sy has to Mx then by

differentiating equations (3.9) and (3.10) produces:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
27
Lecture Notes: Aerospace Structural Design

Also the parameters and are related to the load intensities wy and wx in the same
manner such that by differentiating Equations (3.16) and (3.17) gives:

The parameters , , and are called the EFFECTIVE SHEAR FORCES and
LOAD INTENSITIES.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
28
Lecture Notes: Aerospace Structural Design

Chapter 4

Stress, Strain and

Displacement Relationships
for Open and Closed Single
Cell Thin Walled Beams

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
29
Lecture Notes: Aerospace Structural Design

STRESS, STRAIN AND DISPLACEMENT RELATIONSHIPS FOR OPEN AND

CLOSED SINGLE CELL THIN WALLED BEAMS

Assumptions:

1) Axial constraint effects are negligible

2) Shear stresses normal to surface can be neglected

3) Direct and shear stresses on planes normal to surface are constant across thickness

4) Beams have uniform section, with skin thickness varying around the section but constant
along length of the beam

The parameter 's' in this analysis is the distance measured around the cross-section of the
tube/beam from some convenient origin.
For a loaded beam, look at an element of size δs x δz x t of its wall showing all shear and
direct stresses necessary to keep it in equilibrium.

Figure 23: Loaded beam structure

Figure 24: General stress system on element of close or open beam section

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
30
Lecture Notes: Aerospace Structural Design

These shear and direct stresses are produced by the bending moments, shear loads and
internal pressures. Although 't' can vary with 's' for each element of length δs, we can
assume that this length is small enough to make 't' constant over this length.

From Elasticity we have that:

Instead of using shear stress, the analysis will become easier if we introduce the term Shear
Flow q , which is the shear force per unit length rather than shear stress. It is represented by:

(4.1)

Shear flow is defined positive if it is in the same direction as increasing 's'.

What we are attempting to determine in this analysis are two things, firstly a relationship
between the shear and direct stresses, and secondly a relationship between shear strain and
the deformation of this element. So we may then use these relationships to determine a
shear flow equation and an angle of twist equation.

We now replace the shear stress values in the element with those of shear flow:

Figure 25: Element showing the direct stresses and shear flow

Using equilibrium about the z-axis and neglecting body forces gives:

Dividing by δz and in the limit as δz=0, it simplifies to:

(4.2)

and using equilibrium about the 's' direction Equation (4.3) is similarly derived.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
31
Lecture Notes: Aerospace Structural Design

(4.3)

We now need to look at strain relationships. Starting by defining the three components of
displacement at a point on the tube wall.

Figure 26: Axial, tangential and normal

components of displacement of a point in beam
wall

Where:

w = displacement in the z axis

vt = tangential displacement, positive with increasing 's'

vn = normal displacement, positive outwards

From Elasticity :

(4.4)

To define Shear strain look at how the element is distorted by shear:

Figure 27: Element distorted due to shear.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
32
Lecture Notes: Aerospace Structural Design

Shear strain is then defined as the addition of the two angles of rotation of the sides, such
that:

where in the limits as the element size goes to zero:

(4.5)

It is now necessary to define the term vt as a function of displacements u and v ( in x and y

axis) and angle of twist of the section θ. In order to do this it is necessary to assume that the
ribs are able to hold the cross section rigidly enough so that when it twists it holds its cross
sectional shape. However the ribs have no strength in a plane normal to them, allowing the
section to warp or deform in the z axis.

Define

ψ = Angle between the tangent to the surface of the beam's cross section and the x-axis.

Figure 28: Beam cross section rotated by angle θ, showing the rotation of
the normal to surface

Tangential displacement at any point N on the tube is :

Relating these displacements about a point R, which is the centre of twist gives:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
33
Lecture Notes: Aerospace Structural Design

Figure 29: Rotation of beam section about centre of twist

the displacement vt is

and

which when combined give:

These two equations which describe the tangential displacement of the tube. In order to
determine the second term of Equation (4.5) it is necessary to differentiate Equations (4.6)and
(4.7) with respect to z.

(4.8)

and

(4.9)

These two equations represent the same value, therefore the centre of rotation is:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
34
Lecture Notes: Aerospace Structural Design

These equations will be to use to determine the shear stress distribution in a thin walled open
or closed tube, as well as the displacement, warping and angle of twist of the section due to
these shear loads.

SHEAR FLOW IN BEAMS WITH OPEN SECTIONS

Look at a beam with an open section, with applied shear forces Sx and Sy about a point which
produces no twisting of the tube cross section (Shear Centre).

Figure 30: Open beam section loaded with two shear loads (Sy and Sx)

The relationship between shear flow and axial stress is given by Equation (4.2):

and the equation for direct stress is given by equation (3.11):

differentiating this equation gives:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
35
Lecture Notes: Aerospace Structural Design

substituting back into the first equation gives:

substituting for the effective shear forces gives:

Integrating this equation wrt 's' from one point to another along the beam surface gives:

But by starting at s = 0 where q = 0 for an open beam, then:

Example 2: Determine the shear flow distribution in the thin walled channel section loaded
by a single vertical force applied through the shear centre.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
36
Lecture Notes: Aerospace Structural Design

Figure 31: Channel section loaded vertically

through shear centre.

Since the applied vertical load passes through the shear centre, there is no torque applied to
the beam, so shear flow equation (4.13) applies. Since only Sy is applied, then:

Where :

, ,

Which gives that

So equation (4.13) becomes:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
37
Lecture Notes: Aerospace Structural Design

On the bottom flange 12, at s1 = 0, y = -h and at s1 = h, y = -h which by using the equation of

aline y = ms1 + b gives that: y = -h, where 0 < s1 < h so the shear flow between points 1 and 2
is:

giving:

Now at s1 = 0, q12 = 0, at s1 = h, q2 = 3Sy / 8h, and from equation (iii) it can be seen that we
have a linear increasing shear flow.
On the web 23, at s2 = 0 , y = -h and at s2 = 2h, y = h which by using the equation of a line
y = ms2 + b gives that y = - h + s2, where 0 < s2 < 2h , but at point 2 the shear flow is not 0so:

giving:

Which has the form of a parabola symmetrical about the x-axis, with the maximum value of
shear flow at s2 = h, of q23 = 9Sy / 16h, and at s2 = 2h, q3 = 3Sy / 8h.
On the top flange 34, at s3 = 0, y = h and at s3 = h, y = h which by using the equation of a
line y = ms3 + b gives that y = h, where 0 < s3 < h so the shear flow between points 3 and 4 is:

giving:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
38
Lecture Notes: Aerospace Structural Design

And at s3 = h , q4 = 0, which is a linearly decreasing shear flow.

The shear flow distribution looks like this:

Figure 32: Shear flow distribution on channel section

SHEAR CENTRE OF OPEN BEAM SECTION

The shear centre is that point in the cross section through which the shear loads produce no
twisting. It is also the centre of twist when torsional loads are applied. As a rule, if a cross
section has an axis of symmetry, then the shear centre must lie on that axis and in cruciform
or angle sections, the shear centre is located at the intersections. It is important to define the
position of the shear centre because although most wings are not loaded at this point, if we
know its location, we can represent the shear loads applied as combinations of shear loads
through the shear centre and a torque.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
39
Lecture Notes: Aerospace Structural Design

Figure 33: Shear centre locations for some typical open beam sections

To calculate the shear centre, determine the moment generated by the shear flow about an
appropriate point in the cross section. This moment is equal to the moment generated by the
applied shear force about this same point.

Example 3: For the open beam section of Example 2, determine the position of the shear
centre.

Figure 34: Channel section with load through shear

centre

Because shape is symmetrical the shear centre must lie on the x-axis, a distance scx from the
web.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
40
Lecture Notes: Aerospace Structural Design

The steps in determining the shear centre are as follows:

1) Determine the equations which describe the shear flow in the cross section, (done).

These were found to be:

2) Find an appropriate point in the cross section and take moments about it. In this case
point 3. This eliminates the moments caused by the shear flow in the web 23 and flange 34.
The moment equation is:

substituting the equation for q12 defined previously, gives:

which gives:

Note: In the case of unsymmetrical sections, the coordinates (scx, scy) of the shear centre
have to be found. This is best achieved by first applying a vertical shear force Sy, determining
scx, then applying a horizontal force Sx determining scy.

SHEAR FLOW OF CLOSED TUBES

This solution is similar to that for an open beam section but with 2 differences:

1) Shear loads may be applied through any point in cross section;

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
41
Lecture Notes: Aerospace Structural Design

2) At origin of 's', the value of shear flow qs,0 is unknown

Look at arbitrary closed beam, with applied shear forces Sx and Sy

Figure 35: Closed beam section with two shear loads (Sy and Sx)

If hoop stresses and body forces are absent, it is necessary to use equation 4.2:

and as for the analysis of open beam sections, when substituting for σz, we obtained:

However unlike for open beams, at s = 0, qs0 0, so when integrating we get:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
42
Lecture Notes: Aerospace Structural Design

The first two terms are identical to the equation of shear flow of open tube loaded through
shear centre. If we let this terms be ‘qb’ then:

To obtain qb we assume the beam section is cut at some point to produce an open tube, and
the shear flow distribution is then given by the following equation, which is equation (4.13).

where at s = 0 , qs = 0.
What is now required, is a way of determining the value of the shear stress at the point where
the beam was cut.
To do this, look at beam cross section loaded at some point:

Figure 36: Resolving moments due to applied loads

and shear flow about a point

Take moments about a convenient point inside the beam.

The moment equation will look like this

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
43
Lecture Notes: Aerospace Structural Design

but the shear flow term is given by equation 4.16, so substituting for q gives:

The term is the integral around the cross section.

By looking at the area enclosed by the elemental distance s and the point where moments
are taken, then:

Integrating this over the cross section as the element δs= 0, gives:

Which means that:

where :

A = Area enclosed by the mid-line of the beam section wall

Giving that:

But if we take moments about the points where the shear forces are applied, then this
equation becomes:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
44
Lecture Notes: Aerospace Structural Design

which can easily be used to determine the value of qs,0.

TWIST AND WARPING OF SHEAR LOADED CLOSED SECTIONS

If a shear load is not applied at the shear centre, the closed beam section will both twist and
have an out of plane axial displacement (warp).
From equation (4.1) shear flow is defined as:

and from elasticity:

with the term for shear strain given by equation (4.5)

Combining all three equations gives:

and substituting equation (4.8) gives:

Integrating this equation around the cross section wrt 's' gives:

Which is the rate of twist of the beam wrt z.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
45
Lecture Notes: Aerospace Structural Design

If however, equation (4.20) was integrated wrt 's' from some origin point on the surface of the
beam to any other point in the cross section, with the axis origin at the shear centre, it
gives:

where :

and

Figure 37: Beam section showing area swept by

generator

This equation gives the axial displacement or warping of the beam. If the cross section
wassingly or doubly symmetrical, at the axis of symmetry the warping would be zero. If the

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
46
Lecture Notes: Aerospace Structural Design

origin of 's' was taken at any of these points then at s = 0, w0 = 0 and the rest of the warping
would be easily found.
For unsymmetrical sections, the unknown warping displacement at s = 0 is given by:

However in order to obtain these values, you must first know the position of the shear
centre.

SHEAR CENTRE OF CLOSED BEAM SECTIONS

The position of the shear centre for a beam loaded as shown in Figure 38, can be found by
using equation (4.17).

where:

ηo = vertical distance from x-axis to shear centre from reference axis

εo = horizontal distance from y-axis to shear centre from reference axis

Figure 38: Determining location of the shear

centre for closed beam section

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
47
Lecture Notes: Aerospace Structural Design

In order to determine the shear centre we need to determine qs,0. From the definition for shear
centre, if a shear load is applied here, it produces no twist, so by using equation (4.21):

and substituting for qs,0 it gives:

which gives that:

and if Gt is constant:

With these equations the shear flow, shear centre, the rate of twist, and warping in a closed
beam section can now be determined.

Example 4: Determine shear flow, shear centre and warping in the following closed beam
section loaded with a shear force Sy at the s.c. G is constant.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
48
Lecture Notes: Aerospace Structural Design

Figure 39: Closed section of question 5

1) Determine sectional properties

Since section symmetrical Ixy = 0, and since only loaded vertically , only need Ixx, which is:

giving:

Therefore:

2) Determine Shear Flows

Between 1 and 2, y = s1 and at s1 = 0 , q1 = q0

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
49
Lecture Notes: Aerospace Structural Design

giving:

between 2 and 3, y = h/2, at s2 = 0, q2 = q0 - 9Sy/44h, giving:

between 3 and 4, y = h/2 - s3, at s3 = 0, q3 = q0 - 45Sy/44h, giving:

at s3 = h/2, q3 = q0 - 48Sy/44h

The shear flow distribution will look like this:

Figure 40: Shear flow distribution with constant q0

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
50
Lecture Notes: Aerospace Structural Design

3) Determine q0

Because 't' is not constant, equation (4.26) becomes:

where due to symmetry:

and

giving that:

Substituting this into (i), (ii) and (iii) gives:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
51
Lecture Notes: Aerospace Structural Design

Plotting the shear flow distribution, it would look like this:

Figure 41: Shear flow distribution of closed beam

4) Determine Shear Centre

To determine the shear centre you can use equation (4.18 mod.) or since qs,0 was found, use
equilibrium. If you use the latter it a lot easier.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
52
Lecture Notes: Aerospace Structural Design

Figure 42: Closed beam indicating applied shear force and the
direction of the shear flow.

From definition of applied moments and resultant moments due to internal stresses:

Taking moments about point 1

First term:

Second Term:

Substituting into (vii) gives:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
53
Lecture Notes: Aerospace Structural Design

5) Determine Section Warping

This is done using equation (4.22) but first the axis must be moved to the shear centre:

Figure 43: Closed beam with axis shifted to shear centre.

Since our origin is point 1, the at s = 0 , w0 = 0, so equation (4.22) becomes:

But because the load is applied through the shear centre, the right hand term goes to zero,
giving that

Between 1 and 2,

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
54
Lecture Notes: Aerospace Structural Design

At point 2,

and between 2 and 3,

at point 3,

and between 3 and 4

Because the section is symmetrical about the x-axis, the warping distribution is equal in
magnitude but opposite in direction. If this is plotted, it looks like this:

Figure 44: Diagram showing the warping distribution on the closed beam

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
55
Lecture Notes: Aerospace Structural Design

Chapter 5

Torsion of Thin Walled

Beam Sections

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
56
Lecture Notes: Aerospace Structural Design

INTRODUCTION

All aircraft structures, particularly wings and fuselages, as well as supporting direct shear
loads must support an applied torque. In a fuselage, a torque can be generated by the loads
applied to the vertical fin. In a wing a torque can be generated by the resultant pressure
distribution acting a distance away from the cross section's shear centre. This chapter deals
with determining the shear flow distribution due to an applied torque.

TORSION OF CLOSED BEAM SECTIONS

A beam with a closed section experiencing only a pure torque T and without any axial
constraints, does not develop direct stresses, ie σ z = 0.

So equations (4.2) and (4.3) become:

The only way to satisfy these equations would be if the shear flow 'q' was constant.

NOTE: Although 'q' is constant, the shear stress 'τ ' may not be if the wall thickness 't'
varied with 's'.

Figure 45: Closed beam with applied torque

To determine the relationship between applied torque and shear flow, apply equilibrium to
the end of the beam. In essence the applied Torque T must equal to the torque generated by
the shear flow.

Look at the end of beam, and a small section δs.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
57
Lecture Notes: Aerospace Structural Design

Figure 46: Equating applied torque

with moment generated by shear flow

The torque produced by the shear flow on element δs is pqsδ s. Integrating about the whole
section gives:

We have previously defined that:

Therefore:

Often referred to as the 'Bredt-Batho Formula'. Substituting this equation into (4.21) gives
the rate of twist due to the Torque 'T':

And substituting equation (5.1) into equation (4.22), the equation for the warping distribution
becomes:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
58
Lecture Notes: Aerospace Structural Design

where:

Example 5: Determine the rate of twist and warping distribution in the following
symmetrical, rectangular section when subjected by counterclockwise torque T.
Assume G is constant.

Figure 47: Torsion of rectangular beam section

Start with the rate of twist equation:

Now:

and let:

Giving that the rate of twist is:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
59
Lecture Notes: Aerospace Structural Design

Look at the warping equation, but starting from the symmetry line because here w0 = 0 then:

δ and A were calculated above, so to determine the warping we need to define δo,s and A0,s,
where:

From 0 to 1, 0 ≤ s1 ≤ b/2, so:

From 1 to 2, 0 ≤ s2 ≤ a, so:

and similarly, from 2 to 3, 0 ≤ s3 ≤ b, so:

and from 3 to 4, 0 ≤ s4 ≤ b, so:

All of which gives a linear equation for the warping on all four sides of the section. Meaning
that we need only look at the corners because here the warping is the greatest. At 1:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
60
Lecture Notes: Aerospace Structural Design

Because of symmetry you will find that w2 = -w1 = -w3 = w4.

Figure 48: Warping distribution of closed beam with applied torque

Note: a) if b/tb < a/ta the sign of w is reversed

b) if b/tb = a/ta the section has ZERO warping

TORSION OF OPEN BEAM SECTIONS

To do this analysis we need to consider torsion of a thin rectangular strip, which is bent to
form the open section.

Figure 49: Open beam section subject to torque

This analysis is detailed and complex, for that reason, only the final equations will be
given.
The equation for rate of twist is:

where the second polar moment of area J is given by:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
61
Lecture Notes: Aerospace Structural Design

The shear stress although constant over the section's length, varies linearly through the
thickness of the section and is given by:

with the maximum shear stress occurring at n = ± ½ t to give:

From this definition for shear stress we can now develop the equation for warping.
Note: Because shear stress is not constant through thickness, warping will be present a cross
thickness. This is much less than the warping of the centre line of the section and is ignored
in aircraft structural analysis.

The shear strain is given by equation (4.5):

Shifting the axis of the beam to coincide with the shear centre, makes equation (4.9) become:

Shear stress is given by:

Combining these 3 equations give:

At the centre line of the section, n = 0 and τ z,s = 0, giving:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
62
Lecture Notes: Aerospace Structural Design

Integrating w.r.t. 's' gives:

where:

and

AR is the area swept out by a generator, about the shear centre, from the point of zero
warping.

Note : AR is positive if movement of PR along the tangent to the surface in the direction
of 's' leads to a counterclockwise rotation of PR about the shear centre.

Example 6:

Determine the maximum shear stress τ max, the rate of twist and the warping for the open
beam channel section of Figure 50 when a torque of magnitude T = 10 Nm is applied at the
shear centre. Assume the material shear modulus to be and G = 25 GPa.

Figure 50: Open beam with applied torque of 10 kNmm.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
63
Lecture Notes: Aerospace Structural Design

a) Determining sectional property

b) Determining maximum shear stress

c) Determining rate of twist

d) Determining the sections warping

where:

Because the section is symmetrical, at 0, s1 = 0, w = 0. Therefore, between 0 and 1:

Figure 51: Diagram showing the

derivation of AR

so the warping equation is linearly varying between 0 and 1, and is:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
64
Lecture Notes: Aerospace Structural Design

at s1 = 25 mm, point 1 : w1 = -0.253 mm

Between 1 and 2

Figure 52: Diagram indicating the construction of AR02

So

at s2 = 25 mm, w2 = 0.5345 mm
As the section is symmetrical, warping on the lower section will be the reverse of this.
The warping distribution looks like this:

Figure 53: Warping distribution for channel section with 10 kNmm torque

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
65
Lecture Notes: Aerospace Structural Design

Note: If the section is unsymmetrical and you don't know where ws = 0, use the following
equation.

where:

A R,0 = Swept area by generator rotating about the shear centre from some convenient origin

and

where : section = Integral over the entire perimeter of the section

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
66
Lecture Notes: Aerospace Structural Design

Chapter 6

Sructural Iealisation

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
67
Lecture Notes: Aerospace Structural Design

INTRODUCTION

In the work done so far, the aircraft structural components analysed have been relatively
simple. However an aircraft wing (for example) can be made up of many cell compartments.
Each section of the wing would be covered by a thin skin, and the skins would be reinforced
by many stringers of Z, C or T section. The analysis of such a structure would be extremely
complex and time consuming. In order to simplify this, structural idealization should be
carried out.

STRUCTURAL IDEALISATION

1) The longitudinal stiffeners and spar flanges carry only axial stresses

2) The web, skin and spars webs carry only shear stresses

3) The axial stress is constant over the cross section of each longitudinal stiffener

4) The shearing stress is uniform through the thickness of the webs

5) Transverse frames and ribs are rigid within their own planes and have no rigidity
normal to their plane.

In idealising a structural component by following these five points the new simpler structure
looks something like this:

Figure 54: Original and Idealised wing sections

For which:

• The stiffeners are represented by circles called booms, which have a

concentrated massin the plane of the skin.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
68
Lecture Notes: Aerospace Structural Design

• The direct stresses are calculated at the centroid of these booms and are

assumed to have constant stress through their cross-section.

• Shear stresses are assumed uniform through the thickness of the skins and webs.
•
• The direct stress carrying capability of skin is represented as an addition to

existing booms or as additional separate booms.

Because the skin can carry some direct load, usually tensile (although some can be carried in
compression but this is determined by the buckling load of the skin), the Direct Stress
Carrying Thickness of the skin is defined as:

tD = Actual skin thickness t, if skin resists totally direct load

tD = Percentage of t, if partially resists applied load

tD = 0 if skin only able to resist shear load

We now need to look at how these idealisation techniques can be applied:

Look at a typical stiffened panel:

Figure 55: a) Actual panel,b) Idealised panel with same number of stiffeners and
booms,c) Idealised panel with reduced boom number,d) Further Idealised panel with all
stiffener and bending carrying areas lumped into two booms

To idealise a structure you can:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
69
Lecture Notes: Aerospace Structural Design

1) Replace the spar web by replacing its area to the major booms

2) Reduce the number of stiffeners or booms by combining the stiffener mass with that of
the skin into one boom

If you wish, this could be simplified even further by:

1) Increasing the cross-sectional area of just two of the booms with the direct carrying
capacity of the skin and stiffeners.

REMEMBER : When idealising a structure, the elastic characteristics of the idealised

structure must be the same as for the original structure.

IDEALISED SHEET TO SUPPORT TENSILE LOAD P

Figure 56: Actual sheet to be Idealised Figure 57: Idealised structure

What needs to be done is to idealise the structure of Figure 56 into that of Figure 57. In
orderto do this two things must be considered, force (stress) equilibrium and compatibility
(displacement).
In the real structure the direct stress is determined using:

and its elongation is given by :

In the idealised structure, however, the direct stress is given by:

(in the Booms)

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
70
Lecture Notes: Aerospace Structural Design

And its elongation is given by:

Since equations (6.2) and (6.4) should give the same elongation, and then we can define the
relationship:

If you look at the stress in the booms they are the same as those of the original structure.

IDEALISED SHEET TO SUPPORT BENDING MOMENT M

The bending moment is applied at the end of the plate as follows:

Figure 58: Actual sheet to be idealised with

applied bending moment M

The stress in this sheet is given by :

The second moment of area for the idealised booms is (using parallel axis theorem) is given
by:

and its stress is given by :

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
71
Lecture Notes: Aerospace Structural Design

Since equations 6.6 and 6.8 should give the same maximum stresses then:

As can be seen by this equation, a different idealised boom area equation is required for a
different loading condition.

IDEALISED SHEET TO SUPPORT DIRECT LOAD AND BENDING MOMENT

The stress distribution for a combined loading case with a direct load and a bending moment
looks like Figure 59.

Figure 59: Actual sheet supporting direct load and

bending moment

In the idealised structure, the maximum and minimum stresses need to be carried by the
booms as actual direct stresses, Figure 60.

Figure 60: Idealised sheet supporting

maximum stresses in booms

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
72
Lecture Notes: Aerospace Structural Design

Equating the direct loads along the z-axis gives that:

If the neutral axis in the idealised structure is in the same position as in the original sheet,
the bending moment carried by the idealised structure must be the same as that carried by the
real structure.
Equating moments about neutral plane gives the following:

multiplying by 2 gives:

Adding these equations and simplifying, gives:

By calculating the bending stresses, the structure can then be idealised.

WARNING : There is a problem with equations (6.3). It has to do with idealising sections

of skin very close to the neutral axis of the cross section, see Figure 61.

Figure 61: Skin segment very close to

Neutral Axis to be idealised into booms i & j

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
73
Lecture Notes: Aerospace Structural Design

Equation (6.3) can be generalised in the form of:

However if σ i << σ j or σi = 0, Bi becomes extremely large, giving an incorrect

approximation for the boom area.

To correct this problem, use the following correction. However be aware that this is just a
simplified correction method and that large errors can obtained if proper care is not taken
to observe how these corrections alter the solution.

IDEALISED SHEET TO SUPPORT DIRECT COMPRESSIVE LOAD

When supporting a compressive load, the metal sheet is usually critical in buckling. Because
of this, only a small fraction of its width 'b' is effective in compression. Experimental
research has led to the derivation of the following equations for the effective area of skin in
compression.

Figure 62: Sheet to be idealised in

Figure 63: Idealised sheet
compression

Area of boom:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
74
Lecture Notes: Aerospace Structural Design

where 'C' is the effective width of metal in compression, given by:

Note :
Although the idealisation given by equations (6.6) & (6.7) for a metal sheet in compression is
valid, it is sometimes better to idealise the structure using the equations derived for the other
three load cases. When the stresses on the metal skin are known we can then use critical
buckling stresses for thin metal sheets to determine if they will be effective in compression.

Example 7 :

Part of a wing section is in the form of a two-cell box shown. All the vertical spar webs are
connected to the skins through spar caps, all with cross-sectional areas of 300 mm2. Idealise
the section into direct stress carrying booms and hear stress only carrying skin when the wing
is subjected to an applied bending Moment Mx = -150 kNm. Position the booms at the
spar/skin junctions.

Figure 64: Wing section to be idealised,subject to a bending

moment Mx = -150 KNm

Since all skin panels experience both a tensile/compressive load together with a bending
moment, equation (6.4) is the ones to use when idealising this section.
Now, from symmetry: B1 = B6, B2 = B5 and B3 = B4, saving us some calculations. This would
not be the case if the wing wasn't symmetrical.
Lets start with determining the area of boom 1. This boom must have first of all the area of
spar cap 1, which is 300 mm2, together with the area contributions of the two adjacent skins.
Therefore:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
75
Lecture Notes: Aerospace Structural Design

Since the equation for stress for a beam with symmetrical cross section is :

Then by substituting this equation for the different stresses, their rations simplifies to just the
rations of their vertical distances.

Which gives:
So B1 = B6 = 891.67 mm2
For boom 2 we have to consider the contribution of the two spar caps, and the three adjacent
skins. Giving :

When substituting for the bending moment equation it becomes:

So B2 = B5 = 1791.67 mm2
And for boom 3 we only have one spar cap and two adjacent skins:

and when substituting for the bending stress equation:

giving that B3 = B4 = 1050 mm2

So the idealised wing now looks like this:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
76
Lecture Notes: Aerospace Structural Design

Figure 65: Final idealised wing with boom areas marked.

The above method explains how to calculate the boom areas if carried out by hand. The next
pages will carry out the same example but using table format, specifically using the program
Excel.

Figure 66: Excel sheet where the sectional properties and stresses of the real structure are
calculated.

67: Excel sheet where the boom areas are calculated based on the stresses calculated in the sheet of
Figure 66

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
77
Lecture Notes: Aerospace Structural Design

Figure 68: Excel sheet where the sectional properties and bending stresses of the idealised structure
are calculated

Example 8:
The singly symmetrical fuselage cross-section is subjected to a bending moment Mx = 100
kNm. If all direct stresses are carried by the booms, determine the average direct stress in
each boom.

Figure 69: Idealised fuselage cross section.

Because the section is symmetrical about the y-axis (Ix,y = 0), and My = 0, the effective
bending moment equations reduces to:

a) Determine location of centroid using equation (3.3)

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
78
 Lecture Notes: Aerospace Structural Design

b) Determine sectional properties and stresses in booms

Now that we have the position of the neutral axis, the best way to determine the stresses in
each boom is by using a table

Boom y (mm) B (mm2) ∆Ixx = By2 (mm4) σz (MPa)

1 661.5 640 2.8005e+08 35.67
2 601.5 600 2.1708e+08 32.43
3 421.5 600 1.0660e+08 22.73
4 229.5 600 3.1602e+07 12.37
5 26.5 620 4.3540e+05 1.43
6 -202.5 640 2.6244e+07 -10.92
7 -394.5 640 9.9603e+07 -21.27
8 -500.5 850 2.1293e+08 -26.99
9 -538.5 640 1.8559e+08 -29.04
Ixx = 1.8546e+09

Note: Because of symmetry, this table doesn't contain values for the booms on the left hand
side of the y-axis. However when calculating the second moment of area Ixx, the values of ∆Ixx
for booms 2 - 8 were multiplied by 2 when doing the summation.

SHEAR OF OPEN SECTION BEAMS

The equation for shear flow for an open beam section was found to be equation 4.12:

From its derivation, the value of 't' was the direct stress carrying thickness 'tD' of the skin.
Such hat it could either be equal to : tD = t, if the skin was fully effective in carrying direct
stress, or tD = 0, if skin was assumed to carry only shear stresses.
So:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
79
Lecture Notes: Aerospace Structural Design

So that if we are idealising the beam's cross section, based on our idealisation technique,
the value of 'tD' may or may not be equal to zero.
This equation, however does not consider the effect of the idealised booms to the overall
shear flow. To account for this, we need to consider the equilibrium of one such boom.

Figure 70: Shear loaded open beam with booms, and equilibrium of ith boom.

Summing the forces along the length of the elemental length of the boom gives:

this equation simplifies to :

but since the direct stress is given by equation (3.11), which is:

then substituting for it gives:

Giving the change in shear flow induced in the skin by the presence of a boom. If several
booms were present, their presence becomes an additive one, such that for a beam with 'n'
booms, equation (6.8) becomes:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
80
Lecture Notes: Aerospace Structural Design

But because in an idealised structure the skin only carries the shear stresses, then equation
(6.10) can be simplified to have only the summation terms. To give:

A simpler way of thinking of equation (6.11) is by looking at the shear flow in a skin segment
i due to the shear flow in skin segment i-1 and the change of shear flow due to the boom k
separating these two skin segments, Figure 71.

Figure 71: Sketch of skin segments i and

i-1 and boom k to show how the shear flow
is determined.
In general terms, the change in shear flow between any two skin elements (i and i-1) due to
the area of boom k of coordinates xk and yk about the structures' centroid is:

The shear flow in skin segment i is then a function of the shear flow in the previous skin
segment (i - 1) and the change in shear flow due to boom k ( ∆qk), equation (6.13).

Note:
Because of the idealisation process, the shear flow in any skin section between any
two booms always has a constant value.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
81
Lecture Notes: Aerospace Structural Design

Example 9:
Calculate the shear flow in the following channel section. The skin is only effective in shear
(tD=0), area of all booms 1 & 4 = 250 mm2 , areas of booms 2 & 3 = 350 mm2.

Figure 72: Open channel with 5 kN force applied through

shear centre.

Because the structure is symmetrical Ixy = 0 and as we are only applying a vertical load,
equation 6.11 becomes:

The second moment of area Ixx is:

1)

2)

3)

On the RHS of boom 1, qs = 0, so the change in shear flow due to boom 1 is:

1)

2)
3)

4)

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
82
Lecture Notes: Aerospace Structural Design

this shear flow is constant until boom 2, where:

1)

2)
3)
4)

which is constant until boom 3, where:

1)
2)

3)
4)

and you can see that after boom 4, q = 0.

Figure 73: Shear flow distribution

around channel of example 7.

And when this problem is done using a spread sheet program, the solution would look like
this:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
83
Lecture Notes: Aerospace Structural Design

Figure 74: Excel sheet indicating how this problem can be solved

SHEAR OF CLOSED SECTION BEAMS

The derivation of the equation to determine this shear flow is identical as the analysis for
open
beam sections, making equation (4.15) look like this:

Since the structure has been idealised, then equation (6.14) is simplified to:

where:
qbi is the open beam shear flow of skin segment i, and which can be calculated either
using equation (6.11) or (6.13).
qs,0 in the residual shear flow due to making the closed beam into an open beam
section. It can be calculated using equations (4.19, 4.26 or 4.27) depending if we
know the location of the applied load (4.19), or if we are trying to determine
the shear centre (4.26 or 4.27).

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
84
Lecture Notes: Aerospace Structural Design

Example 10:

Determine shear flow in the following wing structure with an applied vertical load in the
plane of booms 3 and 6. This structure has been idealised into direct stress carrying booms
and shear only stress skin (ie tD = 0). Boom areas are: B1 = B8 = 200 mm2, B2 = B7 = 300
mm2, B3 = B6 = 400 mm2, B4 = B5 = 150 mm2.

Figure 75: Closed wing section with applied shear force between booms 3 & 6

Because the section is symmetrical, Ixy = 0. Skin carries only shear load, tD = 0 so use
equation 6.7. Only a vertical load is applied, Sx=0. Reducing equation 6.7 to:

where:
Sy = 15 kN
Ixx = 2x(200x252 + 300x1252 + 400x1252 + 150x752) = 23.813x106 mm4

Substituting this into the above equation gives:

Where qb is the shear flow for an open section. Cutting the beam between sections 2 and 3
and doing the analysis in a counter clockwise sense we get:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
85
Lecture Notes: Aerospace Structural Design

It is now necessary to determine qs,0. This can be done by using equation 4.16, because we
know the position of the applied force. Or we could take moments anywhere about the line of
action of the applied force, both these statements mean the same thing.
Taking moments about point 3, gives:
0 = ( q4,5x150x150 + q5,6xx158.11x237.17 + q6,7x300x250 + q7,8x412.31x315.3 + q8,1x50x700
+ q1,2x412.31x72.76) + 2x165,000qs,0
giving that qs,0 = -8.08 N/mm
From equation (ii) above we have that qs = qb + qs,0 , so:

qs12 = 23.62 - 8.08 = 15.54 N/mm

qs23 = 0 - 8.08 = -8.08 N/mm

qs34 = -31.5 - 8.08 = - 39.58 N/mm

qs45 = -38.59 - 8.08 = - 46.67 N/mm

qs56 = -31.5 - 8.08 = - 39.58 N/mm

qs67 = 0 - 8.08 = - 8.08 N/mm

qs78 = 23.62 - 8.08 = 15.54 N/mm

qs81 = 26.67 - 8.08 = 18.59 N/mm

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
86
Lecture Notes: Aerospace Structural Design

Figure 76: Shear flow distribution of loaded closed wing section

When this problem is solved using a spread sheet program, the solution would look as
follows:

Figure 77: Spread sheet where the open beam shear flows are calculated

In order to calculate the value of qs0, moments were taken about the centroid for this spread
sheet example.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
87
Lecture Notes: Aerospace Structural Design

However, in order for the spread sheet to calculate the perpendicular distance between a point
in space and a segment of skin equation (6.16) was used, with reference to Figure 78.

Figure 78: Perpendicular distance d between a

point A and skin segment i-j

where:

Figure 79: Excel sheet showing the calculation of qs0 and the final shear flow distribution
Note: Column P gives the twice the area between the centroid and each skin segment, and
the value of cell P15 is equal to 2A, twice the enclosed area of the wing section

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
88
Lecture Notes: Aerospace Structural Design

Chapter 7

Multicell Beams

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
89
Lecture Notes: Aerospace Structural Design

MULTICELL BEAMS

BENDING

When determining the bending stresses, there is no difference if the section is multicell or not
because for bending only the boom areas are used. So in order to do this calculation just use
thee quations derived in Chapter 3.

TORSION

Look at a wing section made of 'N' cells carrying a torque 'T'.

Figure 80: Multicell wing structure with applied torque T

The torque T, generates individual torques in each cell, the sum of which must equal T, so:

But a torque in a closed cell (beam section) produces a constant shear flow, given by equation
(5.1), which is:

Substituting this gives:

However since we have 'N' cells this equation is statically indeterminate, so we need a
compatibility relationship.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
90
Lecture Notes: Aerospace Structural Design

Effect of Booms on Shear Flow Distribution

Look at an elemental section of wing of length z, where three skin sections are in contact with
a boom of area B, Figure 81.

Figure 81: Equilibrium of boom/skin/web junction

At this boom/skin/web junction, the shear flows must all be in equilibrium. The element δz of
the boom experiences forces only produced by shear flows because no direct stresses are
present. Therefore the booms don't affect the analysis in pure torsion.
Summing the forces along boom gives:

which simplifies to:

If no booms are present a similar relationship is obtained.

This equation means that the sum of the shear flows into a junction must equal the sum of the
shear flows out of the junction. This is similar to fluid flow in a multi pipe system. Which is
why we use the term shear flow.

Compatibility Equation for Each Cell

Due to the ribs in aircraft wings, the rate of twist is constant for all cells in the section. So
look at the rate of twist of cell i in the wing section of Figure 80 . Equation (4.21) then
becomes for cell i equal to:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
91
Lecture Notes: Aerospace Structural Design

In order to derive the rate of twist for cell i, it is necessary to look at the three cells: i - 1, i
and
i + 1, and determine the bound integral at cell i of equation (4.21).

Figure 82: Shear flow distribution in the ith cell of an 'N' cell wing beam
section

Let :

It is now necessary to determine the rate of twist for cell i giving:

Rearranging this gives:

and in general terms for the configuration of Figure 82, the rate of twist equation for cell i
becomes:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
92
Lecture Notes: Aerospace Structural Design

where:
∆(i - 1 , i ) =
length of wall common with the cell i and i - 1 cell divided by the
multiple of its thickness times its shear modulus
∆i = Sum of all ∆s for cell i
∆(i+1,i) = length of wall common with the i and i + 1 cell divided by the multiple
of its thickness times its shear modulus
With the rate of twist equation for all cells in the structure + the torque equation, you can
solve them simultaneously to determine the value of the shear flows in the structure.

Example 11: Calculate shear stress distribution in the walls of the 3 cell wing section
subjected to a positive torque of 12000 kNmm

Figure 83: Three cell wing section loaded by a pure torque

Where the cell areas are : AI = 258000 mm2, AII = 355000 mm2, AIII = 161000 mm2

Figure 84: Table where ∆ is calculated

Using equation 7.3, determine the rate of twist of all cells:

For cell I

For cell II

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
93
Lecture Notes: Aerospace Structural Design

For cell III

And from equation 7.1:

These four equations can now be solved simultaneously to give the shear flows. This is done
in the following way:

and

Solving iv, v & vi simultaneously gives:

qI = 8.6923 N/mm, qII = 8.4515 N/mm, qIII = 4.7021 N/mm

Figure 85: Shear flow distribution in three cell wing section with
applied torque

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
94
Lecture Notes: Aerospace Structural Design

Note: By dividing the above values of shear flow by the respective wall thickness, gives the
Shear Stresses in the walls in MPa.

SHEAR LOADS

To determine the shear flow distribution in a multicell beam we use the same analysis as for a
single close cell beam.

Figure 86: Multicell wing section with applied shear loads

Because a single closed beam is a statically indeterminate structure, we need to cut it in order
to determine the shear flow due to an open beam section, then determine the cut skin sections
shear flow, qs,0, add this to the open beam shear flow and determine the true shear flow
experienced by the section.
A multicell wing section can be made statically determinate by cutting a skin panel in each
cell. The best place to cut the cells is always at the centre of the top or bottom skin panels.
The reason for this is that when a cell is loaded vertically, the shear flows are zero at these
positions, so that when determining the cut section shear flow its value will be close to zero,
minimising numerical errors.
If we do this to our idealised wing section, then that panel's shear flow would be zero. If that
section was symmetrical, the bottom panel will also have a zero open cell shear flow. The
section will look like this:

Figure 87: Multicell wing structure loaded by shear loads, indicating the best location for
cutting the cell.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
95
Lecture Notes: Aerospace Structural Design

The shear flow equation for an idealised beam close section subjected to shear loads was
found in Chapter 6 and was given by equations (6.12), (6.13) and (6.15).

where qb is the cut section shear flow and qbi is the open beam shear flow.
By using equation (6.13) the open section shear flow distribution can be found in the
multicell wing section. However there remain N number of unknown values of shear flows at
each of the cuts, ie: qs,0,I, qs,0,II, qs,0,III, . . . . , qs,0,N, which are constant for each cell, shown in
Figure 88.

Figure 88: Redundant shear flow in the ith cell of N cell wing
structure loaded in shear

From equation (4.21), the rate of twist for ith cell:

Substituting equation (6.15) gives:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
96
Lecture Notes: Aerospace Structural Design

but because qs,0,i is a constant, then this term in the integral is similar to that of equation (7.3),
thus giving:

where as before:

where the bound integral term of equation (7.4) can be represented in summation of the open
beam shear flow multiplied by ∆ for all the skin sections surrounding cell i, equation (7.5).

where:
k = Skin segment around cell i and the compatibility equation is then that as for
section 7.2:

Using (7.4) and (7.6) for all cells, (N - 1) equations can be generated, requiring 1 more to
make N equations to solve for the unknown qs,0,i's. The extra equation is found by considering
the moment equilibrium of all cells.
Look at the ith cell, the moment Mq,i produced by the total shear flow about some point '0' is
as follows:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
97
Lecture Notes: Aerospace Structural Design

Figure 89: Moment equilibrium of ith cell

The moment about point 0 due to cell i is given by:

substituting equation (6.15) gives:

which simplifies to:

The sum of the moment of the individual cells is equal to the moment produced by the
applied loads so:

And if moments are taken about the point where the loads are applied then:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
98
Lecture Notes: Aerospace Structural Design

Example 12:
The following wing section carries a vertical load of 80 kN in the plane of web 34, determine
the shear flow distribution and the rate of twist. Boom areas, B1,6 = 2580 mm2, B2,5 = 3880
mm2, B3,4 = 3250 mm2.

Figure 90: Idealised wing section with applied vertical shear load

Where the cell areas are AI = 256000 mm2, AII = 560000 mm2, AIII = 413000 mm2

Figure 91: Table with calculated ∆

1) Determine sectional properties:

Because the section is symmetrical we only need Ixx = 810.99x106 mm4
2) Determine the open section shear flow by cutting the top skin panels, like this:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
99
Lecture Notes: Aerospace Structural Design

Figure 92: Three cell wing section cut to determine open beam shear
flows

Using equations (6.12) and (6.13) the open beam shear flows were calculated in the table of
Figure 93.

Figure 93: Excel table where the open beam shear flows were calculated

The shear flow distribution in the open beam section looks like this:

Figure 94: Idealised wing section showing open beam shear flows

3) Determine rate of twist for each cell using equation (7.4):

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
100
Lecture Notes: Aerospace Structural Design

For Cell I

For Cell II

For Cell III

Because all rates of twist are the same we can now equate them together:

4) Equate moments about the point of application of load gives:

which simplifies to:

Solving these 3 equations simultaneously gives:

Superimposing these shear flows with the open beam values gives:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
101
Lecture Notes: Aerospace Structural Design

Figure 95: Wing section with final shear flows

5) Determine rate of twist:

By substituting these values into any of the rate of twist equations we can find it to be :

SHEAR CENTRE

The position of the shear centre is found in an identical manner to that of section 4.5.
Arbitrary loads Sy and Sx are applied in turn through the shear centre s.c., the corresponding
shear flow distributions determined and moments taken about some convenient point.
The shear flow distributions are found as described in the previous section, except that the N
equations derived using equation (7.4) are enough to determine the unknown shear flows.

This is because the rate of twist of a beam loaded at the shear centre is zero, ( ).

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
102
Lecture Notes: Aerospace Structural Design

Chapter 8

Tapered Beams

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
103
Lecture Notes: Aerospace Structural Design

INTRODUCTION

Most aircraft structural components are tapered to improve structural efficiency. All the
analysis done so far assumes the beams have uniform cross sections. The effect of taper on
direct stresses produces by bending moments are minimal if taper is small and the sectional
properties are determined at the section analysed. However the effect of taper on shear
stresses is considerable.

SINGLE WEB BEAM

Consider an idealised section of beam of length δz in the x-y plane with 2 flanges and a web.
At the RHS, it is subjected to a positive Moment Mx and Shear force Sy. The horizontal
resultant forces carried by the booms due to the bending moment Mx are: Pz,1 and Pz,2, and are
defined by the equation:

Where
Bz = idealised boom area
The FBD of the beam section is:

Figure 96: Effect of taper on single web beam analysis

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
104
Lecture Notes: Aerospace Structural Design

From observation, the vertical components of the axial forces carried by the booms are:

Note : The values for δy could be either negative or positive.

And from trigonometry, we can get that:

Which can then be used to determine the effect of taper on the stresses in the booms.
Now, the internal shear force Sy is equal to the shear force carried by the web Sy,w + the shear
forces carried by the booms, Py,1 and Py,2. Thus:

Since we are interested in determining the shear force carried by the web, then by substituting
equation (8.1) into (8.4) gives:

For an idealised web, the shear flow is then given by :

OPEN AND CLOSED BEAM SECTIONS

The effect of taper on open and closed beam section is exactly the same, and is defined by the
analysis which follows.
Consider an idealised beam of length δz carrying positive shear loads Sy and Sx at the section
z. The beam is also subjected to bending moments Mx and My which produce a direct stress
σz in the booms.

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
105
Lecture Notes: Aerospace Structural Design

Note : The skin is only effective in shear.

The FBD for this boom is:

Figure 97: Effect of taper on the analysis of open or closed beam section

In boom i, the direct stress in z direction is: σz,i, and the component of axial force due to this
stress in the z direction is:

where Bi is area of ith boom

Look at the ith boom along the y-z and x-y axes gives:

Figure 98: Effect of taper on ith beam of wing section

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
106
Lecture Notes: Aerospace Structural Design

From these figures the relationships between Pyi, Pxi, & Pzi can be derived:

where:

Using these equations the true axial load in beam 'i' is:

or in terms of axial stress in z direction and taper is :

This equation give the effect of taper on the axial stress carried by the boom when this force
is divided by the boom's true cross sectional area.
The shear loads Sy and Sx are reacted by the boom loads Px,i and Py,i and the shear flows in the
skin and web panels, such that:

Substituting equation (8.7) into this and re-arranging it gives that:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
107
Lecture Notes: Aerospace Structural Design

The shear flow distribution in either open or closed beam sections are found as before, except
that instead of using Sy and Sx as the shear force terms you now use Sx,w and Sy,w. Which is
the fraction of the applied shear force carried by all skin and web panels.

Warning: When taking moments to determine shear centre or qs,0 you must also consider
the moments generated by all the Px,i and Py,i which are the fractions of the
applied shear forces carried by the booms, equation (8.12).

Example 13: An idealised cantilever beam is uniformly tapered along its length in the x and
y axis and carries a load of 100 kN at its free end. What are the boom forces
and the shear flow distribution in the walls at a section 2 m from the built in
end.
The boom areas are B1 = B3 = B4 = B6 = 900 mm2, B2 = B5 = 1200 mm2.

Figure 99: Tapered box beam

The view of the cross section 2 m from the end is:

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
108
Lecture Notes: Aerospace Structural Design

Figure 100: View of section 2 m from end

At 2 m from the end Sy = 100 kN, Sx = 0, My = 0 , Mx = -200kNm.

a) Sectional Properties

b) Determine Shear force carried by skin

Because of symmetry, Ixy = 0, so the direct stress equation becomes:

Equation for direct force is:

The best way to solve this problem is to create a table with all the necessary values.

From this table we can then obtain the resultant shear load carried by the skin, so using
equations (8.10) we get that: Syw = 100 - 33.33 = 66.67 kN, and Sxw = 0 - 0 = 0 kN

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
109
Lecture Notes: Aerospace Structural Design

c) Determine Open Beam Shear Flow by cutting between booms 1 and 2

Because of symmetry, equation (6.12) simplifies to

Substituting for Shear flow and sectional properties gives:

Applying this equation together with equation (6.13) in a counterclockwise sense, starting
with boom 2 gives:

The section now looks like this:

Figure 102: Diagram of open beam shear flows, shear forces carried by booms because of
taper and qs,0

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
110
Lecture Notes: Aerospace Structural Design

d) Determine qs,0 and closed beam shear flow

To do this take moments at any point, in this case moments were taken about the line of web
34.
100x103 x 200 = 10000x300 - 44.447x600x300 + 6666.7x600 + 5000x1200 - 10000x300
+ 33.335x600x1200 - 10000x300 + 5000x1200 + 6666.7x600 -
44.447x600x300 + 10000x300 + 2x1200x600xqs,0
giving that qs,0 = - 5.56 N/mm, and so, the shear flow distribution is

Figure 103: Final shear flow distribution in tapered

beam at 2 m from fixed end

___________________________________________________________________________
© 1997-2005 Jafar Eskandari Jam
111