Stress Manual Eads PDF

Technical Manual
MTS 004 Iss. C

External distribution authorised: YES X NO
Static stress manual,

1
metallic materials
5 4
2 4
1
2
5
Volume 1
Structural
Design Manuals
Purpose Methods for calculating static failure loads and stresses for
aircraft metallic structural details.
Scope All programmes, static justification of metallic structures.
EDP tool Not applicable.

supporting this Manual
Contents V1 - 1 Stiffened panels

V1 - 2 Buclking of plates and thin shells
V1 - 3 Stiffeners
V1 - 4 Thin web beams
V1 - 5 Stable web beams
V1 - 6 Bolted or rivetted junctions
V1 - 7 Lugs
V1 - 8 Hole reinforcements
V1 - 9 Stabilisers
Document Dept code: BTE/CC/CM Validation Name: JF. IMBERT

Manager
Name: J. HUET Function : Deputy Department
Group Leader
Dept code: BTE/CC/A
Date: 11/99
Signature
This document is the property of AEROSPATIALE MATRA AIRBUS; no part of it shall be reproduced or transmitted 3page 1
without authorization of AEROSPATIALE MATRA AIRBUS and its contents shall not be disclosed.
© AEROSPATIALE MATRA AIRBUS - 1999
Title - Annex
Reference documents C BE 019: Drawing up of the Structural Justification Dossier
Documents to be consulted See bibliography at the beginning of each chapter.
Abbreviations See Lexique Aerospatiale Airbus/ATR

See "General" paragraph of each chapter
Definitions List of words the definitions of which are integrated into the
Lexique Aerospatiale Airbus/ATR:
Highlights
Issue Date Pages modified Justification of the changes made

A 02/98 V1 - 1 à V1 - 3 New document.
V1 - 7 à V1 - 9
B 05/99 V1 - 7 Changes as per table page V1-7.i.
V1 - 4 New chapter.
C 11/99 V1 - 1 Changes as per table page V1-1.i.
Created paragraph V1-1-8.
V1 - 5 New chapter.
© AEROSPATIALE MATRA AIRBUS - 1999 MTS 004 Iss. C 3Ann. page

Static stress manual, metallic materials - Management information
NOT FOR DISTRIBUTION
List of approval
Dept. code Function Name / First name Signature

BTE/CC/CM Chef de Département CAZET G.
Key words Calcul
Bibliography Néant
Distribution list
Dept. code Function Name / First name (if necessary)

BQP/TE Archives Diderot SIBADE Alain
BQP/TE Bibliothèque BQP/TE SIBADE Alain
BTE/SM/MG Bibliothèque Technique BTE BOUTET Fernand
Distribution list managed in real time by BIO/D (Didocost application)
© AEROSPATIALE MATRA AIRBUS - 1999 MTS 004 Iss. C page IG1

MCS V1-1 • STIFFENED PANELS
CONTENTS
issue date change
V1-1 STIFFENED PANELS 2 8/1999 Creation V1-1 8

V1-1 1 General 1 1/1998 Creation
V1-1 2 Preparing basic data 1 1/1998 Creation
V1-1 3 Pocket folding stresses 2 1/1998 Creation
V1-1 4 Calculating stiffened panels single compression 2 1/1998 Creation
V1-1 5 Calculating flat stiffened panels single nominal shear 2 1/1998 Creation
V1-1 6 Calculating curved stiffened panels single nominal shear 2 1/1998 Creation
V1-1 7 Calculating stiffened panels compression and shear 2 1/1998 Creation
V1-1 8 Calculating stiffened panels compression and bending 1 8/1999 Creation
1 GENERAL
1.1- Definitions (V1-1y1/1)*
1.2- Behaviour, failure modes (V1-1y1/3)
1.3- Calculation methodology (V1-1y1/5)
1.4- Scope (V1-1y1/6)
2 PREPARING BASIC DATA
2.1- Material characteristics (V1-1y2/1)
2.2- Super-stiffener sections (V1-1y2/2)
2.3- Example (V1-1y2/5)
3 POCKET FOLDING STRESSES
3.1- Hypotheses (V1-1y3/1)
3.2- Boundary conditions (V1-1y3/2)
3.3- Design
3.4- Example (V1-1y3/3)
4 CALCULATING STIFFENED PANELS SINGLE COMPRESSION
4.1- Principles (V1-1y4/2)
4.2- Widths and load-carrying section (V1-1y4/3)
4.3- Effective buckling length, clamping factor (V1-1y4/4)
4.4- Allowable stresses at UL (V1-1y4/6)
4.4.1 Local buckling
4.4.2 Crippling, lateral buckling, allowable stress at zero slenderness ratio in stiffener
(V1-1y4/7)
4.4.3 Allowable stress at zero slenderness ratio in super-stiffener
4.5- Calculating margin at UL using Engesser formula (V1-1y4/8)
4.5.1 Principle
4.5.2 Modified Engesser formula
4.5.3 Limits, accuracy
4.5.4 Methodology
4.5.5 Example (V1-1y4/10)
*: Page number between brackets
Issue 2 Contents page •i

 AEROSPATIALE 1999
4.6- Calculating margin at UL using Johnson formula (V1-1y4/15)

4.6.1 Principle
4.6.2 Limits, accuracy
4.6.3 Methodology (V1-1y4/16)
4.6.4 Example (V1-1y4/18)
4.7- Simplified approach to the Johnson formula (V1-1y4/22)
4.7.1 Method
4.7.2 Example (V1-1y4/23)
5 CALCULATING FLAT STIFFENED PANELS SINGLE NOMINAL
SHEAR
5.1- Diagonal tension theories (V1-1y5/2)
5.2- Limits (V1-1y5/6)
5.3- Calculation principle for stiffened panels incomplete diagonal tension (V1-1y5/7)
5.4- Loading ratio, diagonal tension factor (V1-1y5/9)
5.5- Widths and load-carrying sections
5.6- Stresses in super-stiffener (V1-1y5/11)
5.6.1 General method
5.6.2 Simplified approaches (V1-1y5/12)
5.7- Stresses in skin (V1-1y5/14)
5.7.1 In pocket
5.7.2 At stiffener
5.8- Stresses in stiffener (V1-1y5/15)
5.9- Allowable stresses at UL
5.9.2 Forced crippling
5.9.3 Buckling of super-stiffener (V1-1y5/16)
5.9.4 Skin failure
5.10- Skin stiffness
5.11- General instability, design (V1-1y5/17)
5.12- Example (V1-1y5/18)
6 CALCULATING CURVED STIFFENED PANELS SINGLE NOMINAL
SHEAR
6.1- Diagonal tension theories (V1-1y6/2)
6.2- Limits (V1-1y6/5)
6.3- Calculation principle for stiffened panels incomplete diagonal tension (V1-1y6/6)
6.4- Loading ratio, diagonal tension factor (V1-1y6/7)
6.5- Widths and load-carrying sections (V1-1y6/8)
6.5.1 Stiffener orientation
6.5.2 Frame orientation (V1-1y6/10)
6.6- Stresses in super-members
6.7- Stresses in skin (V1-1y6/12)
6.7.1 In pocket
6.7.2 At stiffener
6.8- Stresses in stiffener (V1-1y6/13)
page •ii Contents Issue 2

6.9- Stresses in frames (V1-1y6/14)

6.10- Allowable stresses at UL
6.10.2 Forced crippling
6.10.3 Buckling of super-stiffener
6.10.4 Skin failure
6.11- Skin stiffness (V1-1y6/15)
6.12- General instability, local buckling of frames, design
6.13- Example (V1-1y6/17)
7 CALCULATING STIFFENED PANELS COMPRESSION AND SHEAR
7.1- Calculation principles (V1-1y7/2)
7.2- Limits (V1-1y7/4)
7.3- Loading ratio, diagonal tension factor
7.3.1 Loading ratio
7.3.2 Diagonal tension factor
7.4- Widths and load-carrying sections (V1-1y7/5)
7.4.1 Stiffener orientation
7.4.2 Frame orientation (V1-1y7/6)
7.5- Stresses in super-members
7.6- Stresses in super-stifffener (V1-1y7/8)
7.6.1 In skin
7.6.2 In stiffener
7.7- Stresses in frames (V1-1y7/9)
7.8- Allowable loads
7.8.1 At limit loads
7.8.2 At ultimate loads
7.9- Example (V1-1y7/11)
8 CALCULATING STIFFENED PANELS COMPRESSION AND
BENDING
8.1- Introduction (V1-1y8-1)
8.2- Allowable stresses at ultimate load (V1-1y8-2)
8.3- Pocket folding load compression / bending
8.3.1 Detection of minimum folding stress cp
8.3.2 Search for load transmitted Pt
8.3.3 Example (V1-1yy8-5)
8.4- Compression / bending stress (V1-1y8-8)
8.4.1 Position of extreme fibres
8.4.2 Compression / bending stresses at Max moment point
8.4.3 Compression / bending stresses at Min moment point (V1-1yy8-9)
8.4.4 Limits, accuracy (V1-1yy8-10)
8.4.5 Sign conventions
8.4.6 Example (V1-1yy8-11)
8.5- Amplified bending margin (V1-1y8-14)
Issue 2 Contents page •iii

Static stressing manual V1-1 • STIFFENED PANELS
SYMBOLS USED
A: Transverse Member pitch p: Fastener pitch

a: skin bay length q: Transverse distributed load
B: Stiffener pitch R: Loading rate
b: Skin bay width Rc: Single compression loading rate
C: Compression RF: Reserve factor
UL: Ultimate loads Rs: Single nominal shear loading rate
Cflf: Forced crippling coefficient r: Panel curvature radius
LL: Limit loads S: Cross section area
CS: Single shear T: Shear load
D: Fastener diameter TD: Diagonal tension
E: Young's modulus TDP: Pure diagonal tension
Es: Secant modulus TDI: Incomplete diagonal tension
Et: Tangent modulus t: Shear flow
e: skin bay thickness W: Cross section static moment
et: Skin pad thickness
es: Stiffener skin side flange thickness
G: Coulomb's modulus α: Angle of diagonal tension
GTDI: Effective shear modulus under TDi ∆: Flattening deflection
h: Stiffener height ε: Strain (expansion)
I: Section moment of inertia Φ: Normal flow (tension or compression)
IG: General instability γ: Shear strain
K: End fixity coefficient η: Plasticity factor
k: Diagonal tension factor λ: Slenderness ratio
L: Buckling effective length (column) v: Poisson factor
Lt: Load-carrying width σ: Normal stress
M: Bending moment σ0: Column buckling at zero slenderness ratio
MS: Static margin σ0.2: Conventional allowable compressive
N: Normal force yield stress
n: Stress-strain curve shape factor (Ramberg σcp: Skin bay buckling stress in compression
and Osgood) σcrit: Column buckling
P: Compression force (P>0) σenr: Column buckling at zero slenderness
PadmCE: Compression load-carrying capacity ratio stiffener alone
Pcrit: Column buckling load σfirp: Skin inter-rivet buckling stress
σflf: Forced crippling stress
σflr: Stiffener local buckling stress
Page V1-1/1 revision 1

© AEROSPATIALE 1998
σR: Failure tension stress

τ: Shear stress
τflf: Forced crippling shear loading stress
τmax: Maximum shear stress (in Tresca line of
thought)
τR: Failure shear stress

BIBLIOGRAPHY
1- Paul Kuhn, James P. Peterson, and L. Ross Levin: "A summary of diagonal tension", part 1
(methods and analysis), NACA TN 2661, Washington, May 1952.
2- Paul Kuhn, James P. Peterson, and L. Ross Levin: "A summary of diagonal tension", part 2
(experimental evidence), NACA TN 2662, Washington, May 1952.
3- Bruhn: "Analysis and design of flight vehicle structure" (in particular chapter C11).
4- Pablo Rodriguez: "Prédimensionnement d'une partie basse de fuselage en GLARE" (Pre-

sizing of a low part of the fuselage in GLARE), rapport de stage de fin d'étude (end of study
report), INSA Toulouse, 1995.
5- Koos Verolme: "The development of a design tool for fibre metallic laminate compression
panels", Ph. D. thesis, Delpht University, Nov. 1995.
6- F.R. Shanley: "Inelastic column theory", Journal of Aeronautical Sciences, Vol. 14, no. 5,
pp 261-267, May 1947.
7- F. Engesser: "On the buckling strength of straight columns", Zeitschift für Architectur
Ingenieurwesen, Vol. 35, no. 4, pp 455-462, 1889.
8- L. Euler: "Methodus inveniendi lineas curvas maximi minimive proprietate gaudentes",

Annexe 1 ("De curvis elasticis"), Bousquet, Lausanne et Genève, 1744.

V1-1 •1-GENERAL INFORMATION
A large proportion of aircraft structures is designed using a box beam type construction basis.
The box beams are made of stiffened panels.
As thin web beams are also considered as being part of this category, stiffened panels can be
considered as the main basic element of aircraft airframes.
V1-1 •1.1-DEFINITIONS
A stiffened panel is a flat or curved "skin" to which stiffeners are laid parallel to the normal
dominating stress direction. For example, with box beams, this direction corresponds to the
direction of internal forces generated by bending moments.
The stiffeners are longitudinal members attached to the skin over their entire length either
because they are "sewn" by means of bolts, rivets or tack welding and in this case they are
called "fastened stiffeners", or because the whole skin is machined from a thick plate - in this
case they are called "integrate stiffeners".
Thus, stiffened panels comprise transverse members, the function of which is to provide
supporting points, more or less evenly spaced, and more or less stiff for the stiffeners and
possibly the skin. For example, frames perform this function in fuselage panels.
Therefore, the skin is latticed in squares by members. The squares of panels defined in this
manner are called "bays".
For stressing purposes, the panels are broken down into "super-stiffeners" each consisting of a
stiffener and half skin bays.
General information - Page V1-1•1/1 revision 1

Pocket bottom
BAY
Pad Members
SUPER-STIFFENER
=
B
=
=
B
=
Stiffener

V1-1 •1.2-BEHAVIOUR, FAILURE MODES
The skin of stiffened panels is load-carrying unlike the fabric in wooden and fabric
constructions.
These panels are capable of carrying/transferring two types of loads:
- loads in the skin plane (Nx, Φy, Ty, My),
- loads normal to the skin plane, for example due to the pressure applied to it (Tz).
Φy
Ty G
Φy Nx
My
Tz
Stresses resulting from these forces are:
- Normal (σx) and tangential (τxz ; τzx) in the stiffener,
- Normal (σx), transverse (σy) and tangential (τxy ; τyx) in the skin.
The crux of the problem in the stressing of stiffened panels is due to the fact that skin
behaviour depends on the type and intensity of loading:
• Tension: the static limit of skin strength at UL is given by the characteristics of the
material.
• Shear: this is also the case even though the folds occur beyond the critical shear
stress of bays. These folds are oblique in relation to the edges. For this reason,
through misuse of language, the term "diagonal tension" is used.
• This property of dependence is further emphasised in the case of compression. This

is related to the instability of thin sheets.

Three types of instability are differentiated:
• SKIN INSTABILITY:
Thin sheets fold under relatively low compression and/or shear loads.
In compression, the skin is saturated when its buckling stress is reached. At greater values,
the load is picked up by the stiffeners and small portions of neighbouring skin which justifies
the idea of breaking down panels into super-stiffeners.
This saturation phenomenon does not exist under shear stressing. After folds occur, the skin
transfers the excess load partly in the form of shear stress and the remainder as diagonal
tension. The tangential flows at skin bay edges at constant loads are the same with or without
folds. On the other hand, the diagonal tension induces normal flow at these same edges
which causes overloads, especially in the stiffeners.
However, these phenomena only create a dependence of the overall behaviour of stiffened
panels on the intensity of the loads applied to them. Instability of the skin does not cause
panel failure and therefore is not a cause of limitation at UL.
• INSTABILITY OF SUPER-STIFFENERS:
If the transverse members spaced at a pitch A are sufficiently stiff, the super-stiffeners
behave as columns with a length A, generally simply supported.
A A
Therefore, the initial failure occurs in the most critical super-stiffener either by buckling of
the column or by local buckling. Our panels are designed in this manner.
• GENERAL INSTABILITY:
This may occur if the wave nodes of the buckle pattern are no longer at transverse members
due to insufficient stiffness of these members.
High-speed amplification of this type of buckle pattern entails the total ruin and practically
the explosion of the whole panel. This is to be prohibited.

V1-1 •1.3-CALCULATION METHODOLOGY
The information that this chapter of the manual makes possible to calculate is:
- The stresses in the studied structural element corresponding to the applied loads
(these values, amongst others, are needed as they can be compared with
measurements made during non-destructive testing).
- The load causing the occurrence of folds in skin bays. This should not occur too
early (in general, not before 80% of LL).
- The static failure load, i.e. the load allowable at UL.
- The safety margin at UL, which is determined by comparing the allowable load and
the applied load.
In simple cases (stable shear and tension, tension), the calculation of the stresses under a given
load does not cause any special problems. The panels are broken down into super-stiffeners
that are assimilated to beams. The failure load is the one at which the equivalent stress on Von
Mises reaches the allowable tension failure stress of the material (therefore, this criterion is
systematically verified).
If the skin folds before the failure load, then the stresses are no longer a simple linear function
of loads. In this case, the general procedure to use to calculate a super-stiffener is as follows:
- Select any stress value in the skin at the stiffener.
- Calculate the "load-carrying section" corresponding to it.
- The load corresponding to the selected skin stress is determined by summing the
stresses on the load-carrying section. Also, the properties of the load-carrying section
and the material properties are used to determine whether or not the failure load is
reached at this time.
By varying the initial parameters, a stress-load curve of the super-stiffener is obtained and we
can therefore associate a stress level to with a given load. Naturally, this curve is increasingly
monotonic.
Data varying with the skin stress is required to determine the allowable load at UL. This
problem is solved by plotting the evolution of the failure load (it is written in italics because
the true failure load is naturally a constant) depending on this same initial parameter. This
curve is decreasingly monotonic. The allowable load searched for is at the intersection of the
load-stress and failure load-stress curves.
Often, internal loads in the studied piece of structure are linear, with the general loads applied
into the entire structure (more or less). For this reason, the margins are calculated by
comparing the loads and not the stresses.

V1-1 •1.4-FIELD OF APPLICATION
The methods described in this chapter exclusively apply to the calculation of "typical areas":
- Stiffeners are parallel one with the other and perpendicular to the transverse
members.
- The studied stiffener does not edge an opening.
- No stiffener run-outs in the studied area.
- Each rectangular skin bay may be considered as having a constant thickness.
- Stiffeners may be considered as having constant cross sections between two

transverse members.
- The skin curvature is in the plane normal to the stiffeners.
The special cases indicated below are discussed in Chapter V1-4 (thin web beams):
- In plane bending of stiffeners edging an opening under the effect of the diagonal
tension.
- Other effects induced by the presence of an opening in the skin.
- Secondary bending at run-outs.
- Calculation of the fasteners binding the panels together.
- Sizing of the fasteners binding the stiffeners to the skin.

V1-1 •2-PREPARATION OF BASIC DATA
V1-1 •2.1-MATERIAL PROPERTIES
STABILITY CALCULATIONS:
The elastic-plastic behaviour of materials has to be taken into account in these calculations. To
this end, the Ramberg and Osgood model (refer to V2-2 "Material Behaviour") is used:
n
σ  σ 
ε= + 0,002  
E  σ 0, 2 
σ
Es =
ε
1 n 1− n
= +
Et Es E
Therefore, for each material used, the following is required:
• Young's compression modulus, E.
• Conventional allowable compression yield stress, σ0.2.
• R.&O. factor, n.
ALL CALCULATIONS:
Well designed stiffened panels (general buckling impossible) are damage tolerant structures.
Choose "B values" type statistical properties.
Preparation of basic data - Page V1-1•2/1 revision 1

V1-1 •2.2-SUPER-STIFFENER CROSS SECTIONS
Assumption related to the section of curved panels: the curvature of panels is sufficiently
small for them to be considered as flat panels.
CROSS SECTION (INTEGRAL STIFFENER):
Lt1 Lt2
d
e1 G e2
et
Initial cross section: Lti = Bi/2
B1 B2
PROPERTIES:
Area: S
Moment of inertia/Gy: I
Centre of gravity offset: d
Load-carrying width: Lt1, Lt2
Bay thicknesses: e1, e2

CROSS SECTION (FASTENED STIFFENER):
Lt1 Lt2
Y Gp Y
dp
d G
e1 e2
Gr dr
Initial cross section: Lti = bi/2
b1 b2
b1 and b2 are the distances between the end rows of fasteners
SKIN PROPERTIES: STIFFENER PROPERTIES:
Cross section area: Sp Sr

Static moment/YY: WYYp WYYr
Moment of inertia/YY: IYYp IYYr
Bay thickness: e1, e2
WYYp WYYr
Centre of gravity offset: dp = dr =
Sp Sr
Remark: WYYp is negative whereas WYYr is positive.

SKIN + FASTENED STIFFENER PROPERTIES:
The skin and the stiffener may be made from different materials. In this case, corrected cross
sections are used so as to be in the simple case of a beam made of a homogeneous fictitious
material (somewhat like an average material).
The secant modulus and the tangent modulus of the equivalent homogeneous material is:
Sp Sr
Es =   Esp +   Esr ; S = Sp + Sr
 S  S
Sp Sr
Et =   Etp +   Etr
 S  S
The properties of the corrected section are:
Esp Esr
S = Sp + Sr =   Sp +   Sr
 Es   Es 
Esp Esr
WYY =   WYYp +   WYYr
 Es   Es 
WYY
d=
S
Esp Esr
IYY =   IYYp +   IYYr
 Es   Es 
I = IYY - Sd2
The secant modulus may be replaced by the Young's modulus in the following formulas as
long as one remains in the linear elastic domain.
The average stress calculated by using the properties of the corrected cross section σ is
fictitious. The real stresses in the skin and the stiffener are:
Esp
σp =   σ
 Es 
Esr
σr =   σ
 Es 

V1-1 •2.3-EXAMPLE
The stress in the super-stiffener below, subject to a sufficiently low compression load for no
instability to occur (stability calculations are discussed in paragraph 3) has to be calculated.
The skin material properties (2024 PLT 351) are:

Ep = 70300MPa, σ0.2p = 270 MPa, np = 7,05
The stiffener material properties (7075 T 73510) are:

Er = 73800MPa, σ0.2r = 420 MPa, nr = 13,83
170
= =
3
2
35
= =
25
30 2
4
16
Stiffener: Skin:
Sr = 185 mm² Sxp0 = 375 mm²
IYYr = 62186 mm4 IYYp0 = 1485 mm4
dr = 13,9 mm dp0 = - 1,86 mm
WYYr = 2572 mm3 WYYp0 = - 698 mm3

Skin + stiffener section:

S = 185 + 375 = 560 mm²
HOMOGENEOUS FICTITIOUS MATERIAL:
Young's modulus (linear elastic material assumption):

185   375  70300 = 71456 MPa
E =   73800 +  
 560   560 
CORRECTED CROSS SECTION PROPERTIES:
Skin + stiffener corrected cross section area:

S = 560 mm²
73800  70300 
(Check that: 185 + 375 =   185 +   375 = 560)
 71456   71456 
Static moment in relation to YY:

70300   73800  2571,5 = 1969,6 mm3
WYY = -   697,5 +
 71456   71456 
Centre of gravity offset in relation to YY:

1969,6
d= = 3,52 mm
560
Moment of inertia in relation to YY:

70300  73800 
IYY =   1485 +   62186 = 65687 mm4
 71456   71456 
Moment of inertia at the centre of gravity:

I = 65687 - 560 x 3,52² = 58759 mm4
EXAMPLE OF STRESS CALCULATION:
If Nx is - 10000N, the "average stress" is:

10000
σ=- = - 17,9 MPa
560
The skin stress is:

70300 
σp = -   17,9 = - 17,6 MPa
 71456 
And the stiffener stress:

73800 
σr = -   17,9 = - 18,1 MPa
 71456 

V1-1 •3-SKIN BAY BUCKLING STRESSES
V1-1 •3.1-ASSUMPTIONS
Paragraph 2.2 explained that the properties of the cross section of a super-stiffener are
determined assuming that the skin is flat. This assumption is sound as long as the curvature
radius of our panels is long in comparison to the dimension of the skin bays (make sure that
the proportions comply with information in paragraph 6.2).
On the other hand, the curvature, no matter how slight it may be, has a non-negligible effect
on bay folding stresses. In particular, a curved bay is more stable than a flat bay, especially
under compression.
Each bay is assimilated to a constant thickness rectangular plate, whether a flat or curved. The
folding stresses σcp and τcp are buckling stresses (as indicated in paragraph 1.2 only shear and
compression are taken into account).
Generally, folding occurs at the same time on a set of contiguous bays. The appearance of the
skin is then wavy under compression:
BUMP HOLLOW B
HOLLOW BUMP B
Stiffeners are stopped from rotating in their longitudinal axis by means of cleats located at the
transverse members. However, members generally have an open section and their torsional
stiffness is low. This explains the conservative boundary conditions indicated in paragraph 3.2.
Refer to the next chapter (V1-2•BUCKLING OF THIN PLATES AND SHELLS) to

make these calculations.
Skin bay buckling stresses - Page V1-1•3/1 revision 2

V1-1 •3.2-BOUNDARY CONDITIONS
If the members are made of open sections, the edging conditions to be selected to
calculate pocket buckling stresses are: four simply hinged edges.
• BAY WIDTH, b (measured in transverse member direction):
b'
(distance between fasteners)
et
e
b"
Integrated stiffeners: b=B

Fastened stiffeners: b = b'
All types if (et ≥ 3e): b = b"
• BAY LENGTH, a (measured in stiffener direction):
Same principle with A, a', a".
V1-1 •3.3-DESIGN
Generally, requirements stipulate that bay buckling shall not occur before a certain percentage
of limit load. To check out this condition, the folding stresses are compared with the average
stresses at LL (linear static).

V1-1 •3.4-EXAMPLE
• EXAMPLE 1: Use the example in paragraph 2.3, assuming the bays are flat, under the
average stresses at LL:
σ = 45MPa ; τ = 50 MPa
Skin material: 2024PLT3

Ep = 70300 MPa ; ν = 0,33
Bay geometry to the right and to the left of the stiffener (zero curvature):
B = b' = b = 170 mm ; A = a' = a = 530 mm ; e = 2 mm
Boundary conditions:
Four simply hinged edges.
SINGLE COMPRESSION
 a ≥ 1 ⇒ kc = 4
b 
( )
Refer to example in V1-2.2.5:
σ
σcp0 = 36 MPa ; Rc = = 1,25
σcp0
SINGLE SHEAR
 b = 0,32 ⇒ ks = 5,74
a 
( )
Refer to example in V1-2.2.5:
τ
τcp0 = 52 MPa ; Rs = = 0,962
τcp0
COMPRESSION-SHEAR INTERACTION
 Rc + R c2 + 4 R s2 
R = σ = τ = = 1,77 ⇒ folding at 56% of LL
 σcp τcp 2 
 

• EXAMPLE 2: Use example 1 assuming that there are curved bays (r=2820mm), under
average stresses at LL:
σ = 45MPa ; τ = 50 MPa
SINGLE COMPRESSION
See the example in V1-2.3.5:
σ
σcp0 = 45 MPa ; Rc = =1
σcp0
SINGLE SHEAR
See the example in V1-2.3.5:
τ
τcp0 = 57 MPa ; Rs = = 0,877
τcp0
COMPRESSION-SHEAR INTERACTION
 Rc + R c2 + 4 R s2 
R = σ = τ = = 1,51 ⇒ folding at 66% of LL
 σcp τcp 2 
 

V1-1 •4-CALCULATION OF STIFFENED PANELS UNDER SINGLE

COMPRESSION
G
y P
z
A super-stiffener subject to a compression load P=-Nx, behaves like a column with a cross
section depending on P. At a given P, the section is assumed constant along the longitudinal
centreline of the stiffener.
The instability mode of super-stiffeners is described in paragraph 1.2 ("Behaviour, failure

modes"). The transverse members are of sufficient stiffness and therefore general buckling of
the panel cannot occur.
As seen in paragraph 1.3, the problem of the stressman is:
• To establish the relationship between the applied load P and stresses σp and σr.
• To calculate the failure load by column buckling, Pcrit and the allowable load at UL,
PadmCE (these two loads are not necessarily merged).
Stiffened panels under single compression - Page V1-1•4/1 revision 2

V1-1 •4.1-PRINCIPLES
• LOAD-CARRYING WIDTH PRINCIPLE:
As long as the skin stress σp is less than the smallest buckling stresses in compression of the
pocket bottoms involved, σp is uniform.
The pockets saturate after folding. The average stress in their centres is approximately
constant whereas the stress continues to increase at the stiffener at the location where the
skin is fastened to the stiffener.
The load-carrying width concept is used to simplify calculations. These widths are fictitious
widths of skin such that, assuming σp is uniform, the right distribution of loading in the
stiffener and in the skin is found.
σ
Lt1 Lt2
σp
σcp2
σcp1
Lt1.0 Lt2.0
The 0 index relates to the initial cross section, i.e. the section before folding. The load-
carrying widths of the initial section are given in paragraph 2.2 ("Super-stiffener sections").
The "load-carrying cross section" deduced from the load-carrying widths therefore depend on
P, σcp1, and σcp2. In fact, below it is explained that it is easier to deduce P from σp than the
opposite.
• FAILURE LOAD:
The cause of failure of the super-stiffener is either local buckling or column buckling.

In this text, local buckling designates highly localised phenomena, i.e. with a small buckle
pattern wave length, like local buckling of the inner flange or the web of the stiffener (if it is
of the open section type) or inter-rivet buckling of the skin. The phenomena occur abruptly
and cause just an abrupt drop in the strength characteristics of the column. Therefore, local
buckling stress constitutes peak values.
There are other stiffener buckling phenomena that could be called "semi-local". Warping of
open section stiffeners, or "crippling". It occurs more progressively and have a larger wave
length and entail permanent deformation of transverse sections. If this type of phenomenon
can occur before the yield stress of the material, it is combined with column buckling. The
result is the same as if σ0.2 was reduced. As σ0.2, the warping stress and the crippling stress
are parameters affecting column buckling and not simple peak values.
For as long as the slenderness ratio λ of a column is sufficient, the stress and the critical
column buckling load are given by the Euler formula:
∏2 E L I
σcrit = 2
;λ= ;ρ=
λ ρ S
∏ 2 EI
Pcrit =
L2
This formula is only valid in the linear elastic domain, being for σcrit less than the material
limit of proportionality. However, in aircraft construction, small slenderness ratios are used.
Buckling of super-stiffeners, when it can occur appears in the elastic-plastic domain. This
explains the two problem solving methods described in this chapter, based on:
- The Engesser formula.
- The Johnson formula.
V1-1 •4.2-EFFECTIVE WIDTH AND LOAD-CARRYING CROSS SECTIONS
The load-carrying widths to the right and to the left of the stiffener correspond to a given skin
stress σp, and are calculated using the Karman formula (see figures in paragraph 2.2):

bi
(σp ≤ σcpi ) ⇒ (Lti = Lti,0) ; Lti,0 =
2
 σcpi 
(σp ≥ σcpi ) ⇒  Lti = Lti.0 
 σp 
To calculate the characteristics of the load-carrying section, the simplest method is to subtract:
rather than reconstructing the entire cross section using load-carrying widths, remove from the
initial section (index 0) the two small skin rectangles that are not "load-carrying". Then the
following is obtained:
2
∆Si = (Lti,0 − Lti ) ei ; ∆S = ∑ ∆S
i=1
i
S = S0 - ∆S
INTEGRATED STIFFENER COLUMN:
1 2
ei 
d = d0 +
S
∑  d
i =1
0 −
2
 ∆Si
2
2  ei  e 2i 
∆I = ∑ 
i =1 


d 0 −
2
 +
12 
 ∆Si
I = I0 - ∆I - (d 0 − d ) S
2
INTEGRATED STIFFENER COLUMN:
2
ei 
WYYp = WYYp0 + ∑  e
i =1
t −
2
 ∆Si
2
 2
ei  e 2i 
IYYp = IYYp0 - ∑   et −  +  ∆Si
i =1 
 2 12 
Naturally, the properties of the stiffener do not change. To calculate the properties of the
super-stiffener, the corrected section design rules given in paragraph 2.2 are used.
V1-1 •4.3-EFFECTIVE LENGTH OF BUCKLING, END FIXITY

COEFFICIENT
The true length of a super-stiffener A is equal to the distance between the transverse members
delimiting it.

The effective length of buckling, L, is the theoretical length of the column of the same cross
section both ends simply supported, buckling under the same critical load. L therefore depends
on the supporting conditions at the ends of the super-stiffener:
L = KA
The main fixity end coefficients, K, are given below. For more details refer to Chapter V2-5
"COLUMNS".
In current areas, use K=1 (beam with two simple support ends).
MAIN END FIXITY COEFFICIENTS:
P P
K=0,5 K=1
P P
K=0,7 K=2

V1-1 •4.4-ALLOWABLE STRESSES AT ULTIMATE LOAD (UL)
V1-1 •4.4.1-Local buckling
SKIN: Inter-rivet buckling, σfirp.
STIFFENER: σflr is the smallest local buckling stress of the stiffener (web, inner flange,
inter-rivets, etc.)
• Inter-rivet buckling:
The pitch between the fasteners binding the stiffener to the skin, the thickness of the skin pad
under the stiffener and the thickness of the stiffener skin side flange must be sized so that
inter-rivet buckling cannot occur.
Skin and stiffener inter-rivet buckling stresses σfirp and σfirr are calculated considering that
the length of the strips of sheet are equal to the fastener pitch p and that the non-loaded edges
are free. The end fixity coefficient K depends on the type of link:
FLAT HEAD SCREW, TACK WELDS: K=0.54
COUNTERSUNK HEAD SCREW, RIVETS: K=0.66
Refer to Chapter V1-3 "STIFFENERS" to calculate the inter-rivet buckling stresses.
• Other local buckling stresses:
Refer to Chapter V1-3 ("STIFFENERS").

V1-1 •4.4.2-Crippling, warping, allowable stress at zero slenderness ratio in the

stiffener
Refer to Chapter V1-3 ("Stiffeners") to calculate the crippling σcr and the warping σdr stress.
The allowable stress associated with a zero slenderness ratio in the stiffener σenr, is the
smallest value amongst the conventional allowable compression yield stress of the
material, warping stress and crippling stress:
σend = min {σ 0.2 r ; σcr ; σdr}
V1-1 •4.4.3-Allowable stress at zero slenderness ratio in the super-stiffener
This is the stress above which the deformation of the transverse cross section is permanent.
This limit is considered to be reached when the skin stress in σ0.2 or when the stiffener stress is
σenr.
EXAMPLE:
Stiffener
σenr = σ0r Equivalent homogeneous material
σ0 Skin
σ0.2p
σ0p
ε0r ε0p
In this illustration, the limit is due to the stiffener even though the zero slenderness ratio
allowable stress in the stiffener is greater than the allowable compressive yield stress of the
skin.

V1-1 •4.5-CALCULATION OF THE ALLOWABLE LOAD AT UL USING

THE ENGESSER FORMULA
V1-1 •4.5.1-Principle
This is the Euler formula in which the Young's modulus, E, is replaced by the tangent modulus
Et:
∏ 2 Et ∏ 2 EtI
σcrit = ; P crit =
λ2 L2
However, when applied as such, these equations do not make it possible to take into account
possible warping or crippling of the stiffener.
V1-1 •4.5.2-Modified Engesser formula
In fact, this is a modification of the laws of behaviour of materials: Use the Ramberg and
Osgood model (refer to paragraph 2.1) replacing σ0.2 by σ0:
n
σ σ σ 1 n 1− n
ε= + 0.002   ; Es = ; = +
E  σ0  ε Et Es E
Furthermore, whatever the slenderness of the column:
max {σcrit} = σ0
V1-1 •4.5.3-Limits, accuracy
This method is used to calculate panels built of any homogeneous and isotropic metallic
material of which the increasing monotonic curves (σ, ε) can be modelled by means of R&O
formulas.
The accuracy of the failure load calculation is satisfactory as long as the possibilities of local
buckling and warping of stiffeners is correctly analysed. However, stability of a compressed
structural element may be affected by imperfections. For sizing, it is advisable to take into
account 15% minimum margins confirmed by partial tests. Furthermore, these margins may be
used to readjust σ0.
V1-1 •4.5.4-Methodology
Now the following data is available:

• Basic material data (refer to paragraph 2.1): (Ep ; σ 0.2 p ; np), (Er ; σ 0.2 r ; nr )
• Basic geometrical data (refer to paragraphs 2.2 and 4.3): (S0 ; I 0 ; L)
• Skin pocket folding (refer to paragraph 3): (σcp1 ; σcp 2 )
• Local buckling stresses (refer to paragraph 4.4.1): (σfirp ; σflr )

• Zero slenderness ratio allowable stresses
(refer to paragraph 4.4.3): (σ 0 p ; σ 0 r )
Also, a set of rules and equations can now be used to calculate:
• The load-carrying section area for a given skin stress (refer to § 4.2): (Sp, S)
• ε, σr and the tangent and secant moduli corresponding to σp, using the modified
material properties:
nr
σr σr σr 1 nr 1− nr
ε= + 0.002   ; Esr = ; = +
Er  σ0r  ε Etr Esr Er
np
σp  σp  σp 1 np 1− np
ε= + 0.002   ; Esp = ; = +
Ep  σ0p  ε Etp Esp Ep
• The secant modulus and the tangent modulus of the equivalent homogeneous
material corresponding to σp (refer to § 2.2):
Sp  Sr  E ; E =  Sp  E +  Sr  E
Es =   Esp +
 S  S  sr t  S  tp  S  tr
• The inertia of the load-carrying section (refer to § 2.2 and 4.2): I
By varying the parameter σp:
• The curve giving P as a function of σp, σr or σ or vice-versa (this equation is

biunivocal) is obtained:
P = Spσp + Srσr = Sσ
• Also, the curve giving the buckling load of the super-stiffener as a function of σp, σr
or σ and vice-versa (this equation is also biunivocal) is obtained:

∏ 2 EtI
Pcrit =
L2
Column buckling of the super-stiffener occurs when:
Pcrit
P = Pcrit ; and: σcrit = ≤ σ0
S
The load-carrying capacity of the super-stiffener is the allowable load at UL. If the
stiffener or skin local buckling value is reached before Pcrit then the load-carrying capacity is
the value of the corresponding P:
PadmCE = min {Pcrit, P( σfirp ), P( σflr )}
V1-1 •4.5.5-Example
Use the example in paragraph 2.3
The material properties of the skin (2024 PLT 351) are:

Ep = 70300MPa, σ0.2p = 270 MPa, np = 7.05
The material properties of the stiffener are (7075 T 73510) are:

Er = 73800MPa, σ0.2r = 420MPa, nr = 13,83
The load applied at UL is:

P = 3800daN
170
= =
3
2
35
= =
25
30 2
4
16

Stiffener: Skin:
Sr = 185 mm² Sxp0 = 375 mm²
IYYr = 62186 mm4 IYYp0 = 1485 mm4
dr = 13,9 mm dp0 = 1,86 mm
WYYr = 2572 mm3 WYYp0 = - 698 mm3
Pocket folding
Refer to example 1 in paragraph 3.3 (a1 = a2 = 530 mm ; b1 = b2 = 170 mm):

σcp1 = σcp2 = 36 MPa
Inter-rivet skin buckling
Linked by countersunk head bolts, pitch 25 mm: et = 3 mm ; K = 0,66
Details of the skin local buckling stress calculation, as well as that of the other local buckling
stresses, are given in Chapter V1-3 ("STIFFENERS").
σfirp = 289 MPa
Stiffener local buckling
inner flange: 504 MPa

web: 476 MPa σflr = 413 MPa
inter-rivet: 413 MPa
Allowable stress at zero slenderness ratio of the stiffener
Allowable compressive yield stress: σ0.2r = 420 MPa

Warping: σdr = 380 MPa
Crippling: σcr = 461 MPa
13.83
380 380 
σenr = 380 MPa ; ε0r = + 0.002   = 0.565%
73800  420 
Allowable stress at zero slenderness ratio of the super-stiffener
270
Skin allowable compressive yield stress: σ0.2p = 270 MPa ; ε0p = + 0.002 = 0.584%
70300
The limit is due to the stiffener:
 σ0 p  σ 0 p  
np
(ε 0 p > ε 0 r ) 
⇒  σ 0 p ε 0 = 0.565% = + 0.002 
 σ 0 .2 p  
and (σ 0 r = 380 MPa )
 Ep

This equation is not reversible. A numerical solution has to be found:
7. 05
0.565 267  267 
σ0p = 267 MPa checks: = + 0.002
100 70300  270 
Pocket folding load
(σp = σcp = 36 MPa ) → (Sp = Sp 0 = 375 mm 2 ; S = S0 = 560 mm 2 )
7. 05
36 36  36
and ε= + 0.002   ≈ = 0.512%
70300  270  70300
σr = 37.8 MPa
Stress which checks out:

13.83
37.8  37.8  37.8
ε= + 0.002   ≈ = 0.512%
73800  380  73800
The load corresponding to the stresses is:
Pcp = 36 x 375 + 37.8 x 185 = 20492 N, i.e. 81% LL

38000
(The load applied at LL is equal to: ≈ 25300 N).
1.5
Calculation of the load-carrying capacity and the stresses at UL
It can be seen that the local buckling stresses are greater than the allowable stresses at a
zero slenderness ratio both in the skin and in the stiffener. Therefore, the failure mode
of this super-stiffener is column buckling.
For these calculations, a spread sheet was used.
Stresses at UL: P = 38000 N → (σp = 84 MPa ; σr = 88 MPa)

Load carrying capacity: Pcrit = 90200 N → (σp = 225 MPa ; σr = 279 MPa)

Modelling of material curves:
epsilonp sigmap epsilonr sigmar Epsilon Sigmap Sigmar Sigma

0.00051 36 0.00051 37.8 0.051% 36.0 37.8 36.6
0.0012 84.13 0.0012 88.4 0.120% 84.1 88.4 85.9
0.00217 150 0.00217 160 0.217% 150.0 160.0 154.8
0.00268 180 0.00268 198.1 0.268% 180.0 198.1 189.0
0.00323 205 0.00323 237.9 0.323% 205.0 237.9 221.8
0.00336 210 0.00335 247.1 0.336% 210.0 247.2 229.1
0.00364 220 0.00364 267.5 0.364% 220.0 267.5 244.6
0.00381 225.38 0.00381 279.2 0.381% 225.4 279.2 253.4
0.00436 240 0.00436 311.9 0.436% 240.0 311.9 277.8
0.00481 250 0.00481 332.2 0.481% 250.0 332.2 293.6
0.00536 260 0.00536 349.3 0.536% 260.0 349.3 307.7
0.0058 267.05 0.0058 359.5 0.580% 267.1 359.5 316.7
0.006 270 0.006 363.4 0.600% 270.0 363.4 320.2
0.00609 365
Input data is the skin and stiffener stresses, columns 2 and 4. The stiffener stresses
corresponding to (ε=εp) in column 7 are obtained by linear interpolation in columns 3 and 4.
np nr
σp  σp  σr  σr  P
εp = + 0.002   ; εr = + 0.002 ;σ=
Ep  σ0p  Er  σ0r  S
Tangent and secant moduli:
epsilon Sp S P(N) Esp Esr Es Etp Etr Et

0.051% 375.0 560.0 20492 70300 73800 71456 70300 73800 71456
0.120% 257.4 442.4 38003 70266 73800 71744 70061 73800 71624
0.217% 201.6 386.6 59835 69187 73800 71395 63141 73797 68241
0.268% 187.1 372.1 70318 67050 73793 70403 52396 73753 63015
0.323% 177.5 362.5 80393 63536 73730 68739 40159 73307 57077
0.336% 175.8 360.8 82648 62591 73684 68279 37627 72992 55762
0.364% 172.5 357.5 87445 60436 73484 67188 32689 71630 52838
0.381% 170.9 355.9 90168 59131 73254 66472 30150 70116 50925
0.436% 166.7 351.7 97710 55082 71589 63765 23848 60606 43185
0.481% 164.0 349.0 102468 51932 69014 60987 20123 49568 35730
0.536% 161.5 346.5 106614 48536 65206 57436 16894 38255 28299
0.580% 159.8 344.8 109196 46032 61972 54584 14905 31464 23789
0.600% 159.2 344.2 110207 44966 60527 53331 14139 28986 22120
 σcp 
Columns 2 and 4: Sp = Sp0 - be 1 −  ; P = Spσp + Srσr
 σp 
−1
σ  n + 1 − n  ; column 10: E = Sp E + Sr + E
Columns 5 to 7: Es = ; columns 8 and 9: Et = t tp tr
ε  Es E  S S

Critical stresses and loads:
Sigma S WYYp d IYYp I P Pcrit Sigmap Sigmar

37 560 -698 3.52 1485 58759 20492 147524 36 38
86 442 -462 4.96 975 54059 38003 136043 84 88
155 387 -351 6.00 733 51088 59835 122492
189 372 -322 6.42 671 50478 70318 111764
222 362 -302 6.84 629 50333 80393 100941
229 361 -299 6.93 622 50341 82648 98630
245 358 -293 7.13 608 50383 87445 93536
253 356 -289 7.24 601 50411 90168 90200 225 279
278 352 -281 7.52 582 50435 97710 76526
Columns 3 to 6:
e 1 Esp Esr
(σ = 37) → S0 ; ∆S = S - S0 ; WYYp = WYYp0 +  et −  ∆S ; d =  WYYp + WYYr
 2  
S Es Es 
2
 e e2  Esp Esr
IYYp = IYYp0 -   et −  +  ∆S ; I = IYYp + IYYr - Sd2
 2 12  Es Es
∏ 2 EtI
Column 8: Pcrit =
L2
P and Pcrit as a function of σ:
P and Pcrit as a function of the average stress

Sigma
160000
140000
120000
100000 P
80000
Pcrit
60000
40000
20000
0
0 100 200 300 400
Sigma (MPa)

V1-1 •4.6-CALCULATION OF THE ALLOWABLE LOAD AT UL USING

THE JOHNSON FORMULA
V1-1 •4.6.1-Principle
The Johnson formula is used to estimate the critical buckling stress of a column with a low
slenderness ratio using a smoothing curve. This curve is a parabola expressing σcrit as a
function of the slenderness ratio between λ = 0 and the limit value λ'0. The point of tangency
with the Euler curve corresponds to this limit value.
σcrit is written as follows:
2E
λ'0 = ∏
σ0
 λ2 
Johnson: (0 ≤ λ ≤ λ'0) ⇒  σcrit = σ 0 − 2
σ 20 
 4∏ E 
 ∏ ²E 
Euler: (λ ≥ λ'0) ⇒  σcrit = 
 λ² 
σ0 is the allowable stress at zero slenderness ratio in the super-stiffener (refer to paragraph
4.4.3).
σcrit
σ0
Johnson
σ0/2
Euler
λ'0 λ
V1-1 •4.6.2-Limits, accuracy
This method is used for the stressing of panels made of aluminium alloy, typically used in the
aeronautical industry.

The accuracy of the failure load calculation depends on the form of the curves (σ, ε), the
quality of the local buckling analysis and the stiffener warping analysis. The result is
satisfactory (skin in 2024, stiffeners in 7xxx) to slightly conservative (skin and stiffeners in
7xxx). However, the stability of a compressed structural element may be affected by
imperfections. For sizing, it is advisable to take into account minimum 15% margins,
confirmed by partial tests. Moreover, these margins may be used to readjust σ0.
V1-1 •4.6.3-Methodology
Now the following data is available:
• Basic material data (refer to paragraph 2.1): (Ep ; σ 0.2 p ; np ), (E r ; σ 0.2 r ; nr )

• Basic geometrical data (refer to paragraphs 2.2 and 4.3): (S 0 ; I 0 ; L)
• Pocket folding (refer to paragraph 3): (σ cp1 ; σ cp 2 )
• Local buckling stresses (refer to paragraph 4.4.1): (σ firp ; σ flr )
• Zero slenderness ratio allowable stress
(refer to paragraph 4.4.3): (σ 0p ;σ0 r ;σ0 )
Also, a set of rules and equations can now be used to calculate:
• The load-carrying section area for a given skin stress (refer to paragraph 4.2): (Sp, S)
• ε, σr, and the tangent and secant moduli corresponding to σp, using the material
properties:
nr np
σ  σ  σp  σp 
ε = r + 0.002  r  = + 0.002  
Er  σ 0.2 r  Ep  σ 0 .2 p 
σr σp
Esr = ; Esp =
ε ε
• The secant modulus of the equivalent homogeneous material corresponding to σp

(refer to paragraph 2.2):
 Sp  S   Sp   Sr 
Es =   Esp +  r  Esr ; σ = Esε =   σp +   σr
 S  S  S S
• The inertia of the load-carrying section (refer to paragraphs 2.2 and 4.2): I

• Slenderness ratio:
L I
λ= ;ρ=
ρ S
By varying the parameter σp:
• The curve giving λ in function of σ or vice-versa (this equation is biunivocal) is

obtained.
• Also, the curve giving the buckling load of the super-stiffener as a function of λ and
vice-versa (this equation is also biunivocal) is obtained:
2E
λ'0 = ∏
σ0
 λ2 
Johnson: (0 ≤ λ ≤ λ'0) ⇒  σ crit = σ 0 − 2
σ 20 
 4∏ E 
 ∏2 E
Euler: (λ ≥ λ'0) ⇒  σ crit = 
 λ2 
Buckling of the super-stiffener occurs when:
σ = σcrit
Pcrit = Sσcrit
The load-carrying capacity of the super-stiffener is the allowable load at UL. If the
stiffener or skin local buckling value is reached before Pcrit then the load-carrying capacity is
the value of the corresponding P:
PadmUL = min {Pcrit, P(σfirp), P(σflr)}

Use the example in paragraph 4.4.5.
Allowable stress at zero slenderness ratio of the super-stiffener
σ0p = 267 MPa ; ε0 = 0.565%

σ0r = 380 MPa
Effective width and load-carrying section as per paragraph 4.2, for σp = σ0p:
36
Lt = 85 = 31.2 ; ∆S = 2 x (85 - 31.2) x 2 = 215
267
Sp = Sp0 - ∆S = 160 ; S = Sp + Sr = 345
Average allowable stress at zero slenderness ratio (refer to paragraph 4.4.3):
 Sp   Sr 
σ0 =   σ0p +   σ0r ≈ 328 MPa
 S  S
Calculation of the load-carrying capacity and stresses at UL:
A spread sheet was used for these calculations (tables and graphs on following pages).
Stresses at UL: P = 38000 N → (σp = 84 MPa ; σr = 88 MPa)

Load-carrying capacity: Pcrit = 89800 N → (σp = 225 MPa ; σr = 277 MPa)

Modelling of material curves:

0.00051 36 0.0005 37 0.051% 36.0 37.8 36.6
0.00114 80 0.00114 84 0.114% 80.0 84.0 81.7
0.00217 150 0.00215 159 0.217% 150.0 159.8 154.7
0.00268 180 0.00267 197 0.268% 180.0 197.4 188.7
0.00309 200 0.00308 227 0.309% 200.0 227.7 214.1
0.00333 210 0.00332 245 0.333% 210.0 245.5 228.2
0.0036 220 0.00359 265 0.360% 220.0 265.5 243.6
0.00376 225.154 0.00376 276.916 0.376% 225.2 276.9 252.1
0.00429 240 0.00428 313 0.429% 240.0 313.6 278.7
0.00472 250 0.00471 340 0.472% 250.0 340.2 297.8
0.00523 260 0.00521 364 0.523% 260.0 364.8 316.0
0.00565 267.05 0.00565 380 0.565% 267.1 380.0 327.6
0.00584 270 0.00583 385.5 0.584% 270.0 385.7 332.2
0.00585 386
Input data is the skin and stiffener stresses, columns 2 and 4. The stiffener stresses
corresponding to (ε=εp) in column 7 are obtained by linear interpolation in columns 3 and 4.
np nr
σp  σp  σ  σ  P
εp = + 0.002   ; εr = r + 0.002  r  ; σ =
Ep  σ 0 .2 p  Er  σ 0.2 r  S
Secant modulus:
epsilonp Sp S P (N) Esp Esr Es

0.051% 375.0 560.0 20492 70300 73800 71456
0.114% 263.1 448.1 36588 70277 73800 71731
0.217% 201.6 386.6 59799 69270 73800 71438
0.268% 187.1 372.1 70193 67286 73798 70524
0.309% 179.2 364.2 77978 64808 73790 69370
0.333% 175.8 360.8 82323 63115 73773 68580
0.360% 172.5 357.5 87081 61086 73727 67627
0.376% 171.0 356.0 89720 59905 73676 67062
0.429% 166.7 351.7 98025 56000 73180 65038
0.472% 164.0 349.0 103941 52981 72095 63113
0.523% 161.5 346.5 109485 49702 69739 60399
0.565% 159.8 344.8 112981 47268 67258 57993
0.584% 159.2 344.2 114319 46227 66031 56873
 σ cp  σ
Sp = Sp0 - be 1−  ; P = Spσp + Srσr ; Es =
 σ p  E

σ, ε) of equivalent homogeneous skin and stiffener materials:

Curves (σ
400
350
300
Sigmap
Sigmar
Stress (MPa)
250
Sigma
200
150
100
50
0
0 0.002 0.004 0.006
Epsilon
Critical load and stress:
Sigma S I Lambda sigcrit Sigmap Sigmar P Pcrit
37 560 58759 51.74 226 36 38 20492

86 442 54058 47.95 240 84 88 38002
155 387 51071 46.11 247 150 160
189 372 50436 45.52 249 180 197
221 362 50263 45.01 251 205 236
228 361 50267 44.90 251 210 245
244 358 50311 44.68 252 220 266
252 356 50350 44.56 252 225 277 89786 89788
279 352 50481 44.24 253 240 314
298 349 50533 44.05 254 250 340
Columns 3 and 4:
e 1  E sp E 
(σ = 37) → S0 ; ∆S = S - S0 ; WYYp = WYYp0 +  e t −  ∆S ; d =  WYYp + sr WYYr 
 2 S  Es Es 
 e
2
e2  E sp E S
IYYp = IYYp0 -   e t −  +  ∆S ; I = IYYp + sr IYYr - Sd2 ; λ = L
 2 12  Es Es I
λ2
Column 5: σcrit = σ0 - 2
σ 20
4∏ E

Curves (σ, λ) and (σcrit, λ):
350
300
sigcrit
250
Sigma
200
150
100
50
43.00 44.00 45.00 46.00 47.00 48.00
Slenderness ratio

V1-1 •4.7-SIMPLIFIED APPROACH TO THE JOHNSON FORMULA
V1-1 •4.7.1-Method
The method consists in using the diagram on the following page to obtain an estimated value
of Pcrit.
This is nothing else than a parametric bundle of Euler-Johnson curves. The parameter
 σ 
100 x 0  is dimensionless. Each curve of the diagram associates the dimensionless variable
 E
σ
λ to crit .
σ0
Proceed as follows:
• Determine the Young's modulus of the equivalent homogeneous material and the
allowable stress at zero slenderness ratio as previously explained (refer to paragraph
4.6.3).
• Determine S, I and λ corresponding to σ0, i.e. σp0.
• Determine σcrit using the diagram.
This method may be used for sizing. For example, it makes it possible to quickly assess the
effect of a change in material.
Column buckling in the transition domain

1.2
J
o
1
h
n
Sigmacrit/Sigma0
0.8 s
0.2 o
0.6 n
0.3
0.6
0.4 0.7 0.5 0.4 E
u
0.8
l
0.2
0.9 e
Each curve is defined for: 100*Sigma0/E=Cte 1 r
0
0 10 20 30 40 50 60 70 80 90 100 110
Lambda

Use the examples in paragraphs 4.4.5 and 4.6.4:
Allowable stress at zero slenderness ratio for the super-stiffener
σ0p = 267 MPa

σ0r = 380 MPa
Effective width and load-carrying section as per paragraph 4.2, for σp = σ0p:
Sp = 159 ; S = Sp + Sr = 345 mm²
Average allowable stress at zero slenderness ratio (refer to paragraph 4.4.3):

σ0 = 328 MPa
Young's modulus of the equivalent homogeneous material:
E = 71456 MPa
Calculation of the column buckling load:
σ0 328
= = 0.46% σ crit
E 71456 → = 0.77
σ0
λ ≈ 44
σcrit = 0.77 x 328 = 252.6 MPa

Pcrit = 345 x 252.6 = 87100N

PAGE INTENTIONALLY LEFT BLANK

V1-1 •5-CALCULATION OF FLAT STIFFENED PANELS UNDER

SINGLE NOMINAL SHEAR
Ty G
x
z
Flat stiffened panels designed to transfer shear in the skin plane are classified in three
categories corresponding to their behaviour mode:
- Panels under stable shear.
- Panels under pure diagonal tension.
- Panels under semi-diagonal tension or incomplete diagonal tension.
There are no special comments to make concerning the stressing of stable panels in shear. This
paragraph is dedicated to incomplete diagonal tension. The approach given here is based on
theoretical analyses, readjusted following tests. it applies to the domain defined in the "Limits"
paragraph.
Flat stiffened panels under shear - Page V1-1•5/1 revision 2

V1-1 •5.1-DIAGONAL TENSION THEORIES
Let us consider a rectangular plate, with thickness e, surrounded by members hinged together
at their ends, subjected to a shear flow t.
The members are infinitely stiff in bending and symmetrical in relation to the centre plane of
the skin. The horizontal members have the same cross section, as do the vertical members.
y
t
t
x B
e
The stable shear behaviour is as follows at the scale of a 2D-element from the skin:
τ σt
CS = Π/4
σc
σc = - σt = - τ
At and after a certain load level, folds appear caused by the diagonal compression stress. The
excess load then causes the augmentation of the diagonal tension stress alone.
The pure diagonal tension theory is from Wagner:
If the mechanic and geometric properties of the skin and the load are such that the diagonal
compression critical stress is nil or negligible compared to the diagonal tension stress, the
panel reacts in pure diagonal tension (TDP).

σTDP
τ
TDP
TDP α
2τ
σTDP =
τ = nominal shear stress sin 2α
The result is that the skin pulls on its edges, in both x and y directions thus inducing
compression stresses in the members.
σTDP σyp
τ
TDP TDP
= σxp
τ
σx = ; σy = τ tan α
tan α
(these values are easily found using the Mohr circle)
The normal forces induced in the members are:

τBe tB τAe tA
N TD
x =- =- cotan α ; N TD
y =- tan α = - tan α
2 tan α 2 2 2
The angle α is such that the variation in internal energy of the system is minimal entailing the
following result:
ε − εx
tan2 α = TDP
ε TDP − ε y
In which εTDP is the diagonal strain of the skin (tension) and εx and εy the horizontal and
vertical members strains (compression).

The incomplete diagonal tension theory was developed by Kuhn, Peterson and Levin for the
NACA(1):
Very often, the critical diagonal compression stress to which the skin shear critical stress
corresponds is not negligible. The skin then reacts in stable shear for part of the load and in
diagonal tension for the remaining part of the load:
τ σTD σt
TDI TD Π/4 CS
α +
σc
2 kτ
σTD = σc = - σt = - (1 - k) τ
sin 2α
k is the "diagonal tension factor":
k ∈ [0;1]
k = 0: CS
k = 1: TDP
The stresses in the rectangular sheet result from overlaying CS + TD on a 2D-element of the
base (x, y) are:
σyp
τ
TDI kτ
σxp σx =
tan α
σy = kτtan α
1
: NACA technical note 2661: "A summary of diagonal tension"
Part 1 - Methods of analysis - May 1952.

And on one 2D-element according to the theoretical folds:
σ1
TDI τ12 2 kτ
σ1 = + (1 - k) τ sin 2α
sin 2α
σ2 = - (1 - k) τ sin 2α
σ2
τ12 = (1 - k) τ cos 2α
In practice, α is close to 45°. Under these conditions, σ1 and σ2 are practically the principal
stresses, giving:
σ1 − σ 2
τmax = =τ
2
The incomplete diagonal tension theory is qualified as "theoretical engineering" by its authors
because it is developed on the basis of theoretical analyses (in particular those of Wagner) and
readjusted following a great number of tests. However from these tests it appears that the skin
continues to be pure shear loaded beyond τcp.
Therefore, the diagonal tension factor k is calculated using an empirical formula. Furthermore,
if the members are fastened to the skin, a portion of the skin must be subjected to the same
compression stresses as these members. The load-carrying width calculation formula is
obtained from this, which is also empirical, and is given in this method.
These load-carrying widths must not, in any case, be confused with those from Karman
(compressed panels). In TDP, they tend to zero, the entirety behaving as a frame on which a
film has been stretched.
In TDI, the stresses of the centre plane of skin bays are not constant especially if these bays
are in an end span (for example, at the edge of an opening). In this latter-mentioned case, the
edge members bend and the diagonal tension tends to concentrate in the tensioned diagonals.

For these reasons, the term "incomplete" diagonal tension is used instead of semi-diagonal
tension considered as the overlaying of a pure shear state and a diagonal tension state (not
pure).
The fineness of the results is based on the two empirical formulas that have just been
mentioned. Considering this, it is easy to understand why the use of this method is limited to
the use domain specified in the following paragraph.
V1-1 •5.2-LIMITS
The member flange on the skin side must not be too thin:
es
≥ 0.6
e
The proportions of the bays must be within the range:

B A
0.2 ≤ ≤ 1 ; 120 ≤ ≤ 1500
A e
Riveting:
p
2.5 ≤ ≤ 10
D
Pay attention to very thick skin bays (outside the domain): possibility of non-conservative
forecasts.
Stiffeners are approximately evenly spaced and the thicknesses of adjacent bays are similar.

V1-1 •5.3-PRINCIPLE OF THE STIFFENED PANEL STRESSING UNDER

INCOMPLETE DIAGONAL TENSION
The theory of the incomplete diagonal tension was developed to calculate thin web beams
consisting of flanges with the same section connected by a constant thickness web, which is
stiffened by identical vertical members:
The shear load is constant along the beam and therefore the web bays, all identical, are
subjected to the same constant shear flow.
Under these conditions, the fold angle is the same in all bays.
As seen in paragraph 5.1, α is such that the internal energy variation is minimum. To
determine α, it is necessary to know the strain of the web in the fold direction, the strain of the
flanges and the strain of the uprights. However, these values themselves depend on α.
In practice, α is close to 45°. To solve this problem, an initial set of strain values is calculated
assuming that the folds are at 45°, then the resulting tanα value is calculated. This calculation
is then reiterated using the angle α calculated in the previous step until convergence. In
general, three iterations suffice.
In a general case, the shear flow varies from one point to another of the panel. In principle, it
is known at the centre of each bay. Also, the thickness of the skin may vary from one bay to
another as well as the sections of the members. However, there is no simple analytical method
to calculate the variations of the tension field. Here it is necessary to use some simplifications
making it possible to return locally to the simple configuration described above.
This results in the calculation point notion. The information below makes it possible to
estimate the static failure loads and stresses in stiffeners and in the skin at stiffeners. The
section area of transverse members is only used to determine the fold angle. In and around the
studied super-stiffener, the diagonal tension field is assumed uniform and the skin thickness,
the section and the pitch distance of members constant.

Calculation point
y
x
To determine the same effects at the transverse members, it may be necessary to repeat the
same type of calculation considering the transverse members as stiffeners and vice-versa. The
choice will be made considering the dimensional variations and the variations of the nominal
shear field in the skin and the margins obtained.
If the safety margins are low (less than 15%) it is necessary to confirm then by structural tests.
This being said, and as we are concerned by what happens at a stiffener, we shall use the
super-stiffener notion as in the compressed panel cases (refer to paragraphs 1.1, 2.2, 4.1 and
4.2).
If the super-stiffener is not symmetrical in relation to the plane (G, x, z) it is necessary to build
an equivalent symmetrical super-stiffener.
Lt1.0 Lt2.0
Y Gp Y
dp
d G
e1 e2
Gr dr

The index 0 concerns the initial section in the following equations:
e1 + e 2 e Lt + e 2 Lt 2.0
e= ; Lt0 = 1 1.0
2 e1 + e 2
k1 + k 2
Diagonal tension factor: k =
2
2 .3
1 + k
Loading rate: Rs =  
1 − k
e τ + e2 τ2
τ= 1 1
e1 + e 2
V1-1 •5.4-LOADING RATE, DIAGONAL TENSION FACTOR
• The loading rate of a mesh Rs is the ratio between the nominal shear stress and the critical
buckling stress.
• The diagonal tension factor formula is as follows:
 τ 
 R s = < 1 ⇒ (k = 0)
 τ cp 
 R s 0 . 4343 − 1
(Rs ≥ 1) ⇒  k = 
 R s 0 .4343 + 1
V1-1 •5.5-LOAD-CARRYING SECTIONS AND WIDTHS

TD TD
Lt Lt
σxp
e
σp
Lt0 Lt0

Index 0 relates to the initial section, i.e. before folding.
• Load-carrying widths (NACA formula):
Lt 0
LtTD = (1 - k)
2
b a
Stiffener direction: Lt0 = ; Transverse member direction: Lt0 =
2 2
• Load-carrying section:
Exactly the same method is used as for the compressed super-stiffeners, i.e. the subtractive
method: instead of rebuilding the load-carrying section from load-carrying widths, remove
the initial section which is not load-carrying.
∆STD = 2 (Lt0 - LtTD) e = (1 + k) Lt0e

STD = S0 - ∆STD
The calculation of the moment of inertia of the load-carrying section is only useful in the
stiffener direction.
INTEGRAL STIFFENER:
e
WTD =  d 0 −  ∆STD
 2
TD
W
dTD = d0 + TD
S
2
TD  e e 2  TD
∆I =   d 0 −  +  ∆S
 2 12 
ITD = I0 - ∆ITD - (d0 - dTD)2 STD

FASTENED STIFFENER:
 e  TD
W TD
YYp = WYYp0 +  e t −  ∆S
 2
2
TD  e e 2  TD
I = IYYp0 -   e t −  +  ∆S
 2
YYp
12 
Then apply the rules of the corrected section property calculation given in paragraph 2.2
(different materials) assuming that materials are of the linear elastic type.
 S TD
p  S
E =  TD  Ep +  TDr  E r
S  S 
 E p  TD  Er 
W TD
YY =   WYYp +   WYYr
 E E
TD
TD WYY
d = TD
S
 Ep   Er  I
I TD =   I TD
YYp +  
YY
 E  E  YYr
2
ITD = I TD
YY - S
TD TD
d
V1-1 •5.6-STRESSES IN THE SUPER-STIFFENER
V1-1 •5.6.1-General method
The load, the average stress and the average compression strain induced by the diagonal
tension in the super-stiffener is formulated as follows:
2 kLt 0 eτ N TD σ TD
N TD
x =- ; σ TD
x = x
; ε TD
x = x
tan α S TD
x E x
In fact, compression is not constant along the stiffener. It is at the minimum level at the
transverse members (gusset effect) and in the middle, at maximum:
σ TD b σ TD
= (1 - k) 1.78 − 0.64
x min x min
 + k ; TD ≥ 1
σ TD
x
 a σx

The same applies in the transverse member direction:
TD TD
N TD
y TD
σ TD
y
N y = - kaeτ tan α ; σ y = ;ε y =
S TD
y Ey
σ TD
y min  a σ TD
y min
= (1 - k) 1.78 − 0.64  + k ; TD ≥ 1
σ TD
y
 b σy
The diagonal tension strain in the skin is:
 2k  τ
εTD =  + (1 − k ) (1 + ν) sin 2 α
 sin 2 α  Ep
The fold angle in relation to the stiffeners is such that:
ε TD − ε TD
tan2 α = x
ε TD − ε TD
y
V1-1 •5.6.2-Simplified approaches
α is equal to 1. This gives:

As the fold angle is close to 45°, it is assumed that sin2α
τ
εTD = ((1 + k) + ν (1 - k))
Ep
ASSUMPTION 1: If the stiffening ratio and the stiffener material are similar to that of
the transverse members, the folds are considered to be at 45° whatever k is.
ASSUMPTION 2: If the transverse members can be considered as being infinitely stiff,

TD
determine σ x using the chart on the following page.
Remark: This assumption is not conservative for the stiffener as the folds axis tends to turn to
the most rigid members direction.

σ TD
x
CHART GIVING , WHEN:
τ
• The transverse members are infinitely rigid.
• ν = 0.33
S 
Plot k on the X axis, and  x0 − 1 on the Y axis. If necessary, interpolate to obtain the
 be 
required value.
σ TD
x
τ
0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.7 0.8
1
0.9
0.9
0.8 1
1.1
0.7
1.2
0.6
(S0/be) - 1
1.4
0.5
1.6
0.4 1.8
2
0.3
2.5
0.2 3
4
0.1 5
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
k

V1-1 •5.7-STRESSES IN THE SKIN
V1-1 •5.7.1-In the pockets
The pockets assumed identical on either side of the stiffener are subjected to in plane stress:
kτ.tan α
τ
TDP
kτ/tan α
y
x
Tresca criterion:
2
τ max  k   ∏  τ max 
= 1+  ; α =  ⇒  = 1
τ  tan 2α   4  τ 
V1-1 •5.7.2-At the stiffener
In the middle between the transverse members, the maximum compression in stiffener
direction is:
Ep
σ TD
xp min = σ TD
x min
E
The skin is also highly stressed in the cut-out plane passing through the skin-stiffener fastener.
Here, the Tresca criterion in the hatched section is:
2
τ' max e  k 
= 1.3 1+ 
τ et 1 + k
et

V1-1 •5.8-STRESSES IN THE STIFFENER
Compression in the stiffener is at the maximum in the middle between the transverse members
and is:
E r TD
σ TD
r min = σ x min
E
Also, the average compression stress of the stiffener is:

E
σ TD
r = r σ TD
x
E
As indicated in paragraph 1.4, these calculations are only applicable to typical areas. There are
no stiffener run-outs and stiffeners react in single compression. The case of stiffener run-outs
is discussed in Chapter V1-4 ("THIN WEB BEAMS").
V1-1 •5.9-ALLOWABLE STRESSES AT UL
The inter-rivet buckling failure mode does not exist when the skin reacts in diagonal tension.
A forced crippling mode substitutes this mode which is described in the next paragraph.
Therefore, natural local buckling can only occur in the stiffener (refer to Chapter V1-3).
V1-1 •5.9.2-Forced crippling
In reality, this is amplified bending: because the stiffener skin side flange is loaded
transversely by skin undulations, its free edge tends to follow these undulations whereas the
edge it forms with the web remains straight. In this case, local bending stresses due to the
transverse loads induced by the buckle pattern of the skin, amplified by compression are added
to the compression stresses due to diagonal tension.
With skin shear failure, forced crippling is one of the two main causes of failure.
Refer to Chapter V1-3 ("STIFFENERS") to determine the allowable compression stress σflf
and the corresponding nominal shear stress τflf.

V1-1 •5.9.3-Super-stiffener column buckling
• Strength with regard to column buckling is checked, in principle, by comparing the average
compression stress σ TD x and the critical column buckling stress. To calculate this, it is
necessary to take the specific features of the flat panels in diagonal tension into account.
When stiffener bending starts, the diagonal tension attempts to oppose it. Reactions are as if
the column was fastened over its entire length to an elastic foundation. The end fixity
coefficient calculation given below takes this phenomenon into account:
1
LTD = KTD A ; KTD =
 Lt 
1 + k2 3 − 4 0 
 A 
TD TD
Then calculate σ TD
xcrit using (S , I , LTD) with the information given in paragraph V1-1•4.
• Depending on the slenderness ratio of the super-stiffeners, it is possible that the column
buckles under a two half wave mode. For this reason, it is advisable to check that σ TD
x min does
LTD
not exceed the buckling stress calculated with the slenderness ratio .
2ρ TD
V1-1 •5.9.4-Skin failure
Generally, the skin is thicker at stiffeners than at the bottom of bays. Failure may then occur
either at the bay-pad blend radii or in the stiffener-skin fastener rows.
σR
• Failure in the pocket (TRESCA): τmax allowable =
2
σR
• Failure at skin-stiffener fasteners: τ'max allowable =
2
V1-1 •5.10-SKIN STIFFNESS
As long as the skin remains stable, the skin shear stiffness is represented by Coulomb’s
modulus, G. In diagonal tension, it is appreciably more elastic. The effective transverse shear
modulus is GTDI.
As we conceive incomplete diagonal tension as overlaying of a single shear state and of a

diagonal tension state, the following can be formulated:
 1 1− k k 
(γTDI = γCS + γTD) ⇔  = + 
 G TDI G G TD 

with:
E 4 2 cot 2 α tan 2 α
= + +
G TD sin 2 2α S0 S mt 1 − k
− (1 − k ) +
Lt 0 e Ae 2
Giving:
−1
G TDI  ε TD 
TD
 4ν  x cot α + ε y tan α
= 1 +  2 − 1 k + νE p 

G   sin 2α  τ 
SPECIAL CASES:
 
 
 G 
• (α = 45°) ⇒  G TDI = 
 σ TD
p 
 1 + ( 4 ν − 1) k + 2 ν 
 τ 
• Infinitely stiff transverse members:

G
GTDI ≈
E p σ TD
1 + (4 ν − 1) k + ν x
cot α
E τ
S0
TD TD
− (1 + k )
σx σx Lt 0 e
with, obtained from the graph and:
τ τ 2k
V1-1 •5.11-GENERAL INSTABILITY, DESIGN
These panels are protected from the general buckling risk for as long as the stiffeners are
designed to rupture under forced crippling or under super-stiffener column buckling under a
nominal shear load slightly less than the shear strength of the skin. Inversely, if the static
margin depends on the skin shear strength, then our panel is protected from any general
instability risk. This condition is satisfied if the stiffeners have sufficient inertia and super
flange thickness. If, in addition, the skin comprises pads of a sufficient thickness, then the
static margin is only governed by the shear strength of the skin material:
σR
τallowable =
2
e
The study condition is satisfied if: ≤ 0.7
et

V1-1 •5.12-EXAMPLE
Use the example in paragraph 2.3 (also see paragraphs 4.5.5 and 4.6.4).
The nominal shear at UL in the bays adjacent to the stiffener is:

τ = 160 MPa
The properties of the skin (2024PLT3) are:

Ep = 70300 MPa ; σRp = 440 MPa ; σ0.2p = 270 MPa
The properties of the stiffeners and transverse members (7075PLT6) are:

Er = 73800 MPa ; σRr = 495 MPa ; σ0.2r = 420 MPa
The two materials have the same Poisson ratio: ν = 0.33
SECTION IN STIFFENER DIRECTION:
170
= =
3
2
35
= =
25
30 2
4
16
Stiffener: Skin:
Sr = 185 mm² Sxp0 = 375 mm²
IYYr = 62186 mm4 IYYp0 = 1485 mm4
dr = 13.9 mm dp0 = 1.86 mm
WYYr = 2572 mm3 WYYp0 = - 698 mm3

SECTION IN TRANSVERSE MEMBER DIRECTION:
Small section: Smt = 509 mm² ; Sy0 = 1604 mm² ; Syp0 = 1095 mm²
24
4
24
4
90 2.5
120
8
6
3 2
3 35
530
The small section is the one where the stiffeners pass. Between the stiffeners, the section of
the transverse member alone is 509 mm². This value is used to calculate the fold angle and the
stresses in the transverse member halfway from the stiffeners, giving:
Smt = 509 mm² ; Sy0 = 1604 mm²
POCKET FOLDS
Refer to the example in paragraph 3.3 (a1 = a2 = A = 530 mm ; b1 = b2 = B = 170 mm):

τcp1 = τcp2 = 52 MPa
LIMITS, AS PER PARAGRAPH 5.2
Stiffener flange:
es 3
= > 0.6
e 2

Skin-stiffener attachment with 4.8 diameter rivets, pitch distance 25 mm:

p
2.5 <  = 5.2 < 10
D 
Skin bay proportions:
B A
0.2 <  = 0.32 < 1 ; 120 <  = 265 < 1500
A  e 
LOADING RATE, DIAGONAL TENSION FACTOR (paragraph 5.4)

τ 160
= = 3.077
τ cp 52
3.077 0. 4343 − 1
k= = 0.24
3.077 0. 4343 + 1
LOAD-CARRYING WIDTHS AND SECTIONS AT UL (paragraph 5.5)
• Stiffener direction:
∆S TD
x = (1 + k) Lt0e = 1.24 x 85 x 2 = 211 mm²
S x = S0 - ∆STD = 560 - 211 = 349 mm²
TD
Sr = 185 mm² ; S TD
xp = 164 mm²
 e
W TD TD
YYp = W YYp 0 +  e t =  ∆S TD = - 276 mm3
 2  x
2
 e e2  4
I TD = I TD
+   e −  + TD
 ∆S x = 572 mm
YYp YYp 0
  t
2  12 
164  185 
Ex =   70300 +   73800 = 72154 MPa
 349   349 
W TD  70300  276 +  73800  2572 = 2361 mm3

YY = - 
 72154   72154 
2361
dTD = = 6.76
349
I TD  70300  572 +  73800  62186 = 64161 mm4
YY = 
 72154   72154 
I = 64161 - 349 x 6.76² ≈ 48200 mm4
TD

• Transverse member direction:

ae
∆S TD
y = (1 + k) = 1.24 x 530 = 657 mm²
2
S TD
y = S0 - ∆S TD
y = 1604 - 657 = 947 mm²
Smt = 509 mm² ; S TD
yp = 438 mm²
438  509 
Ey =   70300 +   73800 = 72181 MPa
 947   947 
STRESSES IN THE SUPER-STIFFENER, TD ANGLE AT UL (paragraph 5.6)
• Compression in stiffener direction:

2 kLt 0 eτ
N TD
x =- = 2 x 0.24 x 85 x 2 x 160 x cot α = - 13019 cot α
tan α
TD N TD TD σ TD
σ x = TD = - 37.3 cot α ; ε x = x = - 0.052%.cot α
x
Sx Ex
• Compression in transverse members:

N TD
y = - kaeτ tan α = - 0.24 x 530 x 2 x 160 x tan α = - 40590 tan α
TD
N TD
y TD
σ TD
y
σ y = = - 42.9 tan α ; ε y = = - 0.059%.tan α
S TD
y Ey
• Diagonal tension (refer to paragraph 5.6.2):

160
εTD ≈ (1 + 0.24 + 0.33 (1 - 0.4)) = 0.339%
70300
• Diagonal tension angle:

 ε = ε TD 0.339 + 0.52 
(α0 = 45°) →  tan 2 α 1 = TD r
TD
= 
 ε TD − ε mt 0.339 + 0.059 
 0.339 + 0.052 / 0.9903 
(α1 = 44.72°) →  tan 2 α 2 = 
 0.339 + 0.059 x 0.9903 
 0.339 + 0.052 / 0.9916 
(α2 = 44.76°) →  tan 2 α 3 = 
 0.339 + 0.0591 x 0.9916 
α = 44.7° ; tan α = 0.99

• Super-stiffener stresses:
37.3
σ TD
x =- = - 37.6 MPa
0.99
σ TD b
x min
= (1 - k) 1.78 − 0.64  + k = 0.76 (1.78 - 0.64 x 0.32) + 0.24 = 1.44
σx TD  a
σ TD
x min = - 54.1 MPa
REMARK 1: The stiffening ratios in the stiffener and transverse member direction are similar
and the panel is a little stiffer in the stiffener direction.
560 − 170 x 2 1604 − 530 x 2

= 0.39 ; = 0.34
560 1604
The use of the following approximation (deviation negligible in relation to the general
solution) is justified:
(α ≈ 45°) ⇒ (σ TD
x = - 37.3 MPa ; σ TD
x min = - 1.44 x 37.3 = - 53.7 MPa)
REMARK 2: Use of the assumption "infinitely rigid transverse members" and therefore the
chart in paragraph 5.6.1 gives the following results (- 6% deviation):
  S0    σ TD 
  − 1 = 0.65 ; k = 0.24 →  x = 0.22 ; σ x = − 0.22 x 160 = − 35.2 MPa
TD
  2 Lt 0 e    τ 
(- 37.3 cot α = - 35.2) ⇒ (α = 46.7°)
POCKET STRESSES (refer to paragraph 5.7.1)

(α ≈ 45°) ⇒ (τmax = τ = 160 MPa)
SKIN STRESS AT THE RIVET ROW (refer to paragraph 5.7.2)

2
2 0.24 
τ'max = 1.3 1 +   x 160 = 0.883 x 160 = 141.2 MPa < τmax
3  1.24 
Therefore, skin failure will occur at the pocket blend radius.
SKIN COMPRESSION IN STIFFENER DIRECTION

E p TD 70300
σ TD
xp min = σ x min = - 54.1 = - 52.7 MPa
Ex 72155

STRESSES IN THE STIFFENER (refer to paragraph 5.8)
E r TD 73800
σ TD
r = σx = − 37.6 = - 38.5 MPa
Ex 72155
E 73800
σ TD
r min = r σ TD
x max = − 54.1 = - 55.3 MPa
Ex 72155
ALLOWABLE STRESSES AT UL
• Local stiffener buckling (refer to paragraph 4.5.5):

flange: 504 MPa, web: 476 MPa
• Forced crippling (refer to example in paragraph V1-3 5.4):
σflf = 144 MPa ; τflf = 283 MPa
• Super-stiffener column buckling (refer to paragraph 5.9.3):

1 1
KTD = = = 0.94
2  Lt 0  2  85 
1 + K 3− 4  1 + 0.24  3 − 4 
 A   530 
LTD = 0.94 x 530 = 498 mm
Johnson formula (refer to paragraph 4.6.4):
S TD
σ0 = 328 MPa ; λ CE TD
= LTD TD x
= 42.4
I
TD 328 2 x 42.4 2
λ = λ CE :σ TD
xcrit = 328 - = 260 MPa
4 ∏ 2 72155
λTD
λ = CE : σ TDxcrit = 311 MPa
2
These values calculated with the slenderness ratio at UL are conservative as λ is
decreasingly monotonic with the applied load.
Therefore, with regards to stability, the allowable is given by the forced

crippling.
• Skin failure at the pocket (refer to paragraph 5.9.4):

σ Rp 440
τmax allowable = = = 220 MPa
2 2

STATIC MARGIN:
τ max allowable 220

= = 1.38
τ max 160
τ flf 283
= = 1.77
τ 160
In our example, the static margin results from the comparison of nominal shear stress at UL
with the failure stress of the skin material. The other margins are not very different, meaning
that the skin pad and the super flange of the stiffener are correctly sized. The risk of general
buckling of the panel is nil and the weight is close to the optimum.

V1-1 •6-STRESSING OF CURVED STIFFENED PANELS IN SINGLE

NOMINAL SHEAR
Ty
G
z x
REMINDER: The curvature of the panel, radius r, is in the plane normal to the stiffeners (refer
to paragraph 1.4).
This paragraph is dedicated to incomplete diagonal tension. For more information, refer to
paragraph 5 (Calculation of flat stiffened panels in single nominal shear). The approach given
here is based on theoretical analysis, readjusted following tests. It applies to the domain
defined in the "Limit" paragraph.
Curved stiffened panels in shear - Page V1-1•6/1 revision 2

V1-1 •6.1-DIAGONAL TENSION THEORIES
The analysis of the diagonal tension in curved panels is based on the methods developed for
flat panels. Therefore, the goal here is to describe the problems posed by the panel curvature.
The typical example of a cylindrical fuselage section subject to a torsion moment is a good
basis for discussion.
• Pure diagonal tension theory:
If the cross section of the fuselage section was polygonal instead of circular, the diagonal
tension theory in flat webs would be applicable with some minor modifications. This will
also be the case if the circular cross section skin was supported by polygonal shaped frames
as the diagonal tension flattens the skin bays. However, in the present case, the frames, i.e.
the transverse members, are circular. Therefore, the diagonal tension in a bay no longer
resides in one plane.
To simplify the problem, Wagner established fundamental relations (equations) derived from
the flat panel theory for two cases:
a) Bays long in the axial direction, stringers (stiffeners) spaced close together:
r
A > 2B ; B <
3
b) Bays long in the transverse direction, frames (transverse members) spaced close
together:
r
B > 2A ; A <
3
r
B
In the first case (a), the major portion of the diagonal tension is in the planes of a polygonal
cylinder. In the second case (b), it is in hyperboloids of rotation.
In the case of flat panels subjected to TDP (refer to paragraph 5.1), the angle α is such that
the variation of internal energy of the system is minimal, giving the following result:
ε TDP − ε TDP
tan2 α = x
ε TDP − ε TDP
y

This principle can be applied to curved panels considering that:
Case a: Flattening is unrestricted between two frames. The change from the circular cross
section shape to the polygonal cross section is equivalent to a decrease in the spacing of
)
stringers, ∆, equal to the difference in length between arc B and chord B :
1  B3 
∆≈  .
24  r 2 
The angle of the folds is then given by:
ε TDP − ε xTDP
tan2 α = 2
TDP 1  B
ε TDP − ε y +  
24  r 
∆
Remark: Strain is fictitious. It does not correspond to a stress in the frames.
B
Case b: In the case of long skin bays in the frame direction, the same type of reasoning
results in:
ε TDP − ε xTDP
tan2 α = 2
1  A
TDP
ε TDP − ε y +   tan 2 α
8r
In practice, the first equation is used when A is greater than B and the second when B is
greater than A.
If the bays are "square bays", the same results should be obtained using the two equations,
which implies:
 tan 2 α = 1  ⇒ (α = 30°)
 3
The latter-mentioned value is quite representative. In practice, the fold angle is from 20° to
45°.
Another difference in behaviour between flat and curved panels due to skin bays flattening:
the "roof" effect at the stringers implies that the stringers are subject to radial loads. The
diagonal tension bends the curved panel stringers whereas it acts against stiffener bending in
flat panels.

Assuming that these radial loads are steadily distributed, there is:
B B q
r A
B BA 2 BA 2
q = τe tan α : M(0) = M(A) = - τe tan α ; M  A  = τe tan α
r 12 r  
2 24 r
With the sign convention: (M > 0: compressed skin).
However, if the frames are fastened to the skin, there is no roof effect around the frames as
they remain curved. And if the bending moment mid-way between the frames is effectively
equal to M(A/2), its modulus is much greater than M (A) at the frames.
• Incomplete diagonal tension theory1:
Diagonal tension develops earlier in curved panels than in flat panels. The diagonal tension
factor k is given by generalising the experimental formula applicable to flat panels.
In stringer direction, the calculation formula of the load-carrying width is identical to the one
applicable to flat panels.
The same applies to the deformation equation εTD.
There are two types of frames: frames directly connected to the skin and frames that are not
connected to the skin, which are named "floating frames".
Normally, frames are connected to the skin and slotted where stringers pass. In this case, the
centripetal load applied to them due to the diagonal tension is uniform. They work in the
vault manner, meaning that they are subjected to compression stresses only. The flanges of
these frames are likely to be destroyed under forced crippling. Here, the load-carrying
section is calculated like in the stringer direction.
Floating frames are only loaded at frame-stiffener intersections. Owing to this, they are
subjected to bending moments. Furthermore, no load-carrying skin reacts and folds are free
to develop on either side of the frames.
1
Part 1 - Methods of analysis - May 1952

The use of floating frames can only be considered for panels with bays that are long in the
stringer direction (a) type).
If the frames are fastened to the skin, the bending moment of super-stringers at frames is
considered as being equal to that mid-way from frames, from a modulus standpoint. If the
frames are floating, this bending moment value is doubled.
V1-1 •6.2-LIMITS
This method can be used in the design of panels made of aluminium alloys normally used in
the aeronautical industry.
Use the rules applicable to flat panels (paragraph 5) if the curvature radius is greater than
3800 mm.
These calculations only apply to typical areas (no stiffener run-outs) as indicated in paragraph
1.4.
The member flange (skin side) must not be too thin:

es
≥ 0.6
e
The bays proportions shall be in the following range:

 B < 1 →  0.2 ≤ B ≤ 1 ; 120 ≤ A ≤ 1500 ; A < r 
A   A e 3
 B > 1 →  0.2 ≤ A ≤ 1 ; 120 ≤ B ≤ 1500 ; B < r 
A   B e 3
Riveting:
p
2.5 ≤ ≤ 10
D
Pay attention to very thick skin bays (outside field of application): possible non-conservative
estimates.
The stiffeners are practically evenly spaced and the thicknesses of adjacent bays similar.

V1-1 •6.3-CALCULATION PRINCIPLE FOR STRESSING STIFFENED

PANELS IN INCOMPLETE DIAGONAL TENSION
To determine α, the strain of the web according to the fold axis, the strain of the stringers and
the strain of the frames needs to be known. However, these values themselves depend on α.
In practice, α is between 20 and 45°. To solve this problem, an initial set of strain values is
calculated assuming fold angle to be α = α0, then the resulting value of tanα is calculated. This
calculation is then reiterated using angle α calculated at the previous pitch up to convergence.
Three iterations suffice if α0 is properly selected.
There is no simple analytical method to calculate the variations in the tension field as indicated
for flat panels. Here it is necessary to use a simple approach making it possible to compare the
present problem to a simple configuration.
The information given below makes it possible to estimate the static failure loads and stresses
in stringers and in the skin at stringers. The frame sections were, in particular, used to
determine the fold angle. In and around the studied super-stiffener, the diagonal tension field
is assumed to be uniform. The thickness and the curvature radius of the skin, the section and
the spacing of members is assumed to be constant.
Calculation point
y
x
Calculations cannot be repeated considering the frames as stiffeners and the stringers as
transverse members due to the curvature. Experimental data on frame failure under diagonal
tension is not available. Frames will always be sized so that failure occurs either in the skin or
in the stringers. Frames are considered to be correctly sized when their local buckling does not
affect the general margin and overall instability of the fuselage.
If safety margins are low (less than 15%) it is mandatory to confirm them by structural
tests.

The super-stiffener notion will be used as for the flat panels (refer to paragraph 5.3). The
curvature of the skin is assumed to be sufficiently small for it not to be taken into account
when calculating the section properties.
An equivalent symmetrical super-stiffener has to be built if the super-stiffener is not

symmetrical in relation to the plane (G,x,z):
Lt1.0 Lt2.0
Y Gp Y
dp
d G
e1 e2
Gr dr
h
In the following equations, the index 0 relates to the initial section.
e1 + e 2 e Lt + e 2 Lt 2.0
e= ; Lt0 = 1 1.0
2 e1 + e 2
r + r2
r= 1
2
k1 + k 2
Diagonal tension factor: k =
2
1
1 + k ϕ
Loading rate: Rs =  
1 − k
e1 τ1 + e 2 τ 2
τ=
e1 + e 2
• The loading rate of a skin bay Rs is the ratio between the nominal shear stress and the
critical buckling stress.

• The general diagonal tension factor k is formulated as follows:
 τ 
 R s = < 1 ⇒ (k = 0)
 τ cp 
 R sϕ − 1
(Rs ≥ 1) ⇒  k = ϕ 
 R s + 1
with:
 0.5 < B < 1 ⇒  ϕ = 0.4343 + 260.6 eA 
 A   rB 
 0.5 < A < 1 ⇒  ϕ = 0.4343 + 260.6 eB 
 B   rA 
 0.2 < B < 0.5 
   e
A
 ⇒  ϕ = 0.4343 + 521.2 
 0.2 < A < 0.5   r
 B  

V1-1 •6.5-LOAD-CARRYING SECTIONS AND WIDTHS
V1-1 •6.5.1-Stiffener direction
Lt Lt
σxp
e
σp
Lt0 Lt0
• Load-carrying widths (NACA formula):
Lt 0
LtTD = (1 - k)
2

• Load-carrying section:
A procedure identical to the one use for compressed super-stiffeners is used, i.e. the
subtractive method. Rather than rebuilding the load-carrying section using load-carrying
widths, remove the initial section which is not load-carrying.
∆S TD
x = 2 (Lt0 - LtTD) e = (1 + k) Lt0e
S TD
x = Sx0 - ∆S TD
x
INTEGRAL STIFFENER:
e
WTD =  d 0 −  ∆S TD
 2
x
W TD
dTD = d0 + TD
Sx
2
 e e2 
∆ITD =   d 0 −  +  ∆S TD
 2
x
12 
ITD = I0 - ∆ITD - (d0 - dTD)2 S TD
x
FASTENED STIFFENER:
 e
W TDYYp = WYYp0 +  e t −  ∆S TD
 2
x
2
TD  e e2 
I YYp = IYYp0 -   e t −  +  ∆S TD
 2
x
12 
Then apply the design rules for the properties of the corrected section given in paragraph 2.2
(different materials) assuming that materials are of the linear elastic type.
TD
 S TD
xp   S 
E x =  TD  E p +  TDr  E r
 Sx   Sx 
 E p  TD  E r 
W TD
YY =  TD  WYYp +  TD  WYYr
 Ex   Ex 
TD
WYY
dTD =
S TD
x
 E p  TD  E 
I TD
YY =  I
TD  YYp
+  TDr  IYYr
 Ex   Ex 
2
ITD = I TD TD
YY - S x dTD

V1-1 •6.5.2-Frame direction
Proceed in the same manner if frames are fastened to the skin:
ae
∆S TD
y = (1 + k)
2
S TD
y = Sy0 - ∆S TD
y
 S TD
yp   S mt 
E TD
y =  TD  E p +  TD  E mt
 Sy   Sy 
Calculation of the moment of inertia of the load-carrying section is not necessary.
It is even simpler if frames are floating:
S TD
y = Smt
E TD
y = Emt
V1-1 •6.6-SUPER-MEMBER STRESSES
The compression force, average stress and average strain in the stringer direction induced by
diagonal tension is formulated as follows:
kbeτ N TD σ TD
N TD
x =- ; σ TD
x = x
; ε TD
x = x
tan α S TD
x E TD
x
Compression is not constant over the entire length of a stringer. Compression is minimum at
the frames (gusset effect) and maximum mid-way between frames:
σ TD
x min  b σ TD
x min
= (1 - k) 1.78 − 0.64  + k ; TD ≥ 1
σ TD
x
 a σx
The same is calculated in frame direction:
TD TD
N TD
y TD
σ TD
y
S TD
y E TD
y
σ TD
y min
= (1 - k) 1.78 − 0.64  + k ; TD ≥ 1
σ TD
y
 b  σy

Diagonal tension strain in the skin is:
 2k  τ
εTD =  + (1 − k )(1 + ν)sin 2α
 sin 2α  Ep
The fold angle in relation to stiffeners is such that:
 
 
 B < 1 ⇒  tan 2 α = ε TD − ε TD
x 
A   1   
B
2
 ε TD − ε TD +
 y
24  r  
 
 TD 
 B > 1 ⇒  tan 2 α = ε TD − ε x 
A   1 A
2

 ε TD − ε yTD +   tan 2 α 
 8 r 
 A = 1 ⇒  tan 2 α = 1 ; α = 30°
B   3 
• Bending moment in the super-stiffener half way from frames:
BA 2
M TDA  = k τe tan α
 
 2 24 r
This moment compresses the stringer skin side flange and the skin.
• Bending moments in the super-stiffener at frames:
BA 2
• Frames fastened to the skin: M (TD
A) = - k τe tan α
24 r
BA 2
• Floating frames: M (TD
A) = - k τe tan α
12 r
This moment compresses the stringer inner flange.
• Bending moments in floating frames mid-way from stringers:
B2 A
M TDB  = - k τe tan α
 
 2 24 r
This moment compresses the frame inner flange.

• Bending moments in floating frames at stringers:
B2 A
M (TD
B) = k τe tan α
12 r
This moment compresses the skin.
V1-1 •6.7-STRESSES IN THE SKIN
V1-1 •6.7.1-In pockets

kτ.tan α
τ
TDP
kτ/tan α
y
Tresca criterion:
2
τ max  k 
= 1+ 
τ  tan 2 α 
V1-1 •6.7.2-At the stringer
In the middle between transverse members, the compression stress is:
M (TD
dTD, if integral stiffener
TD TD A / 2)
σ xp =σ x min - TD
I
E p  TD M (TD
A/2) 
σ TD
xp = TD
σ
 x min − TD
(d TD + e t ) , if fastened stiffener
Ex  I 
The skin is also highly stressed in the cutting plane crossing the skin-stiffener fastener rows.
Here, the Tresca criterion in the hatched part is:
2
τ max e  k 
= 1.3 1+ 
τ et 1 + k

e
et
V1-1 •6.8-STRESSES IN THE STIFFENER
• In the middle between two transverse members, the maximum single compression stress,
the outer flange stress (fastened stiffeners) and the inner flange stress are respectively
equivalent to:
Er
σ TD
r min = σ TD
x min
E TD
x
E r  TD M (TD
A/2) 
σ TD
rs = TD
 σ x min − TD
d TD 
Ex  I 
E r  TD M (TD 
σ TD
rt = TD
Ex 
 σ x min +
I
A/2)
TD ( h − d TD )

• The average compression stress of the stiffener is:
Er
σ TD
r = σ TD
x
E TD
x
• At frames, the outer flange stress is:
TD
E r  TD M ( A ) 
TD (
σ TD
rt = TD
 σ x + h − d TD ) )
Ex  I 
Reminder: If frames are floating, the bending moment is twice as great as the bending
moment at mid-distance from the frames.

V1-1 •6.9-STRESSES IN FRAMES
The single maximum compression stress in frames riveted to the skin mid-way between
stringers is:
E mt TD
σ TD
mt min = σ ymin
E TD
y
If frames are floating, apply the bending moments given in paragraph 6.6. The loads then
apply to the cross sections of frames alone.
V1-1 •6.10-ALLOWABLE STRESSES AT UL
Natural local buckling can only occur in stiffeners and in frames (refer to Chapter V1-3).
V1-1 •6.10.2-Forced crippling
Proceed in the same manner as for flat panels (refer to paragraph 5.9.2). Refer to Chapter V1-3
"STIFFENERS" to calculate the allowable stresses. Check the strength of frame and stringer
outer flanges.
V1-1 •6.10.3-Super-stiffener column buckling
Check the strength with regard to column buckling by comparing the maximum compression
stress σ TD
x min and the critical buckling stress, calculated assuming that the super-stiffener is
clamped at frames.
TD
Calculate σ TD TD
xcrit using (S x ; I ) (refer to the Johnson formula in paragraph V1-1 •4).
V1-1 •6.10.4-Skin failure
Generally, the skin is thicker at stiffeners than in pockets bottoms. Failure may then occur
either in the skin pad-pocket blend radii connection fillets or in the skin-stiffener fastener
rows.

σR
• Pocket failure (TRESCA): τmax allowable = (0.65 + ∆)
2
• Failure at skin-stiffener fasteners: Ditto for τmax allowable
With:
S  S 
∆ = 0.3 tanh  mt  + 0.1 tanh  r 
 Ae   Be 
The skin pads have to be included in member cross sections.
V1-1 •6.11-SKIN STIFFNESS
Skin shear stiffness, for as long as the skin is stable, is represented by Coulomb’s modulus, G.
It is appreciably more compliant in diagonal tension. The effective transverse modulus of
elasticity is GTDI:
G  4ν  ε TD
x cot α + ε TD
y tan α
=1+  2 − 1 k + νE p
G TDI  sin 2α  τ
V1-1 •6.12-GENERAL INSTABILITY, FRAME LOCAL BUCKLING,

DESIGN
The nominal shear stress causing cylinder failure by general buckling τIG is obtained using the
chart on the following page.
Frames must be designed so that the static margin of the panel is given by another criterion.
General instability must be impossible.
Furthermore, frame flanges shall be of a sufficient thickness so that the allowable load is not
limited by local buckling or forced crippling. In fact, it could be potentially dangerous in one
way or another for the general static margin to be the same as the margin of a frame (frames).

USE OF THE CHART:
B
•  < 1 : Calculate ρx and ρy, assuming that the entire sheet interacts with the stringers and
A 
the frames (even floating).
B
•  > 1 : Calculate ρx and ρy, assuming that the sheet does not interact with the stringers
A 
and the frames (member sections only). If ρx > 8.6 mm, take ρx = 8.6 mm.
The solid straight lines correspond to the domains covered by the tests (good correlation). No
explanation has been found for the abrupt transition from one to the other.
Do not extrapolate outside the scope of the chart.
Use the Young's modulus in compression.
General instability criterion for curved panels

τ IG
x 103 in shear
E
2.5
1.5
0.5
0
0 2 4 6 8 10 12
( ρx ρy ) 8
x 104
3
r 4
AB

V1-1 •6.13-EXAMPLE
Use again the example in paragraph 5 with r=2820 mm:
Nominal shear at UL in skin bays adjacent to the stiffener is:

τ = 160 MPa

Ep = 70300 MPa ; σRp = 440 MPa ; σ0.2p = 270 MPa ; ν = 0.33
The properties of the stringers and the frames (7075PLT6) are:

Er = 73800 MPa ; σRr = 495 MPa ; σ0.2r = 420 MPa ; ν = 0.33
ASSOCIATED STRINGER AND SKIN:

170
Stiffener:
= = Sr = 185 mm²
IYYr = 62186 mm4
3 dr = 13.9 mm
2
35 WYYr = 2572 mm3
= =
Skin:
Sxp0 = 375 mm²
IYYp0 = 1485 mm4
dp0 = - 1.86 mm
25
WYYp0 = - 698 mm3
3
30 2
4
16

ASSOCIATED FRAME AND SKIN:
24
4
24
4
90 2.5
120
8
6
3 2
3 35
530
The small section of the frame is the one at the stringer passage. Between the stringer
passages, the section area of the frame itself is 509 mm². This value is used to calculate the
angle of the folds and the stresses in the frame at mid-distance from the stringers, giving:
Smt = 509 mm² ; Sy0 = 1604 mm²
Properties of the small section:
Frame: Frame + skin

Sc = 369 mm² Sy0 = 1464 mm²
Ic = 355232 mm4 I0 = 2415091 mm4
dc = 84.4 mm d0 = 19.9 mm

POCKET FOLDING
Refer to example in paragraph 3.3

(a1 = a2 = A = 530 mm ; b1 = b2 = B = 170 mm ; R = 2820 mm):
τcp1 = τcp2 = 57 MPa
LIMITS AS PER PARAGRAPH 6.2
Stiffener inner flange:

es 3
= > 0.6
e 2
Skin-stiffener connected by 4.8 diameter rivets, pitch distance 25 mm:
p
2.5 <  = 5.2 < 10
D 
Skin bay proportions (A>B):
B A 2820
0.2 <  = 0.32 < 1 ; 120 <  = 265 < 1500 ; A <
A  e  3
 0.2 < B < 0.5 ⇒  ϕ = 0.4343 + 521.2 x 2 

 A   2820 
ϕ = 0.8039
τ 160
Rs = = = 2.807
τ cp 57
2.807 0.8039 − 1
k= = 0.393
2.807 0.8039 + 1
For the identical flat panel subjected to the same nominal shear stress (refer to the example in
paragraph 5), the diagonal tension factor is 0.24. Here, the effect of the diagonal tension is
more significant even though the folds occur a little later.
LOAD-CARRYING SECTION AND WIDTHS AT UL (paragraph 6.5)
• Stringer direction:
∆S TD
x = (1 + k)Lt0e = 1.393 x 85 x 2 ≈ 237 mm²
S x = Sx0 - ∆S TD
TD
x = 560 - 237 = 323 mm²
Sr = 185 mm² ; S TD
xp = 138 mm²

 e
W TD TD
YYp = W YYp 0 +  e t −  ∆S TD = - 224 mm3
 2
x
2
 e e2 
I YYp = I YYp 0 -   e t −  +  ∆S TD
TD TD
= 459 mm4
  2  12 
x
138  185 
E TD =   70300 +   73800 = 72303 MPa
x
 323   323 
W TD  70300  224 +  73800  2572 = 2407 mm3

YY = - 
 72300   72300 
2407
dTD = = 7.45
323
I TD  70300  459 +  73800  62186 = 63920 mm4
YY = 
 72300   72300 
I = 63925 - 323 x 7.45² ≈ 46000 mm4
TD
• Frame direction:
The frames are fastened to the skin, giving:
530 x 2
∆S TD
y = 1.393 x = 738 mm²
2
S TD
y = 1604 - 738 = 866 mm²
509   509  x 73800 = 72357 MPa
E TD = 1 −  x 70300 +  
y
 866   866 
STRESS AND STRAIN, STRINGER AND FRAME DIRECTION, TD ANGLE AT

UL (paragraph 6.6)
• Compression, stiffener direction:
kbeτ
N TD
x =- = - 0.393 x 170 x 2 x 160 x cot α = - 21357 cot α
tan α
N TD σ TD
σ TD
x
x
= TD = - 66.1 cot α ; ε TD
x = x
= - 0.091%.cot α
Sx E TD
x

• Compression, transverse member direction:
N TD
y = - 0.393 x 530 x 2 x 160 x tan α = - 66584 tan α
N TD
y σ TD
y
σ TD
y = = - 76.9 tan α ; ε TD
y = = - 0.106%.tan α
S TD
y E TD
y
• Diagonal tension:
 2 x 0.393  160
εTD =  + (1 + 0.33) (1 − 0.393) sin 2 α
 sin 2 α  70300
0.179%
εTD = + 0.184% x sin 2 α
sin 2 α
 
 TD TD  2
 B < 1 ⇒  tan 2 α = ε − εx  ; 1  B  = 0.015%
A   TD 1  B   24  R 
2
 ε − ε +   
 24  R  
TD y
Considering the small amount of flattening strain, α0 = 45° was selected:

 0.363 + 0.091 
(α0 = 45°) →  tan 2 α 1 = 
 0.363 + 0.106 + 0.015 
→ (α1 = 44.1° ; α2 = 44.3° ; α3 = 44.2°)
α = 44.2° . tan α = 0.974
66.1
σ TD
x =- = - 67.8 MPa
0.974
σ TD b
x min
= (1 - k) 1.78 − 0.64  + k = 0.61 x (1.78 - 0.64 x 0.32) + 0.39 = 1.35
σ TD
x
 a
σ TD
x min = - 91.5 MPa

• Super-frame stresses:
σ TD
y = - 76.9 x 0.974 = - 74.9 MPa
  a   σ TD
y min

 (1 − k ) 1.78 − 0.64  + k = 0.26 < 1 ⇒  TD = 1
 b   σy 
σ TD
y min = - 74.9 MPa
• Bending moments in the super-stiffener:
170 x 530 2
MTD = 0.393 x x 160 x 2 x 0.974 = 86325 N.mm
24 x 2820
SKIN POCKET STRESSES (refer to paragraph 6.7.1)

(α ≈ 45°) ⇒ (τmax = τ = 160 MPa)
SKIN STRESS AT THE ROW OF RIVETS (refer to paragraph 6.7.2)
2
2 0.393
τ'max = 1.3 x 1 +   x 160 = 0.9 x 160 = 144 MPa < τmax
3  1.393 
Skin failure will occur therefore in the pocket blend radius.
SKIN COMPRESSION, STIFFENER DIRECTION, BETWEEN FRAMES

E p  TD M TD TD 
σ TD
p = σ
TD  x max
− TD
(d + e t )
Ex  I 
70300  86325
σ TD
p = x  − 91.5 − x ( 7.45 + 3) = - 108 MPa
72303  46000 

STRESSES IN THE STIFFENER (refer to paragraph 6.8)

• At mid-distance from frames:
E r TD 73800
σ TD
r = TD
σx = − x 67.8 = - 69.2 MPa
Ex 72303
E r TD 73800
σ TD
r min = TD
σ x max = − x 91.5 = - 93.4 MPa
Ex 72303
Outer flange:
E r  TD M 73800  86325
σ TD
rs = σ
TD  x max
− TD d TD  =  − 91.5 − x 7.45 = - 107.7 MPa
Ex I  72303  46000 
Inner flange:
73800  86325
σ TD
rt =  − 91.5 + x (30 − 7.45) = - 50.2 MPa
72303  46000 
• Inner flange stress at frames:

73800  86325
σ TD
rt =  − 67.8 − (30 − 7.45) = - 112.5 MPa
72303  46000 
STRESSES IN FRAMES (refer to paragraph 6.9)
73800
σ TD
mt min = - 74.9 x = - 76.4 MPa
72357
ALLOWABLE STRESSES AT UL (paragraph 6.10)
• Local stiffener buckling (refer to paragraph 4.5.5):

Inner flange: 504 MPa, web: 476 MPa
• Stringer forced crippling (refer to example in paragraph V1-3•5.4):
σflf = - 177 MPa ; τflf = 213 MPa
• Frame forced crippling (refer to paragraph 6.10.2): Discussed in example

V1-3•5.4
σflf = - 171 MPa ; τflf = 258 MPa
No danger from this standpoint.

• Super-stiffener column buckling (refer to paragraph 6.10.3):
KTD = 0.5
LTD = 0.5 x 530 = 265 mm
Johnson formula (refer to paragraph 4.6.4):
TD S TD
TD
σ0 = 328 MPa ; λ = L CE = 22.2
I TD
 TD 328 2 x 22.2 2 
( λ = λ CE )
TD
→ σ
 crit

= 328 − 2
4 ∏ 72303
= 309 MPa

These values calculated with slenderness at UL are conservative as λ is monotonic,
decreasing with the applied load.
With regards to stability, the allowable value is given by the forced crippling.
• Skin pocket failure (refer to paragraph 6.10.4):

σ Rp
τmax allowable = (0.65 + ∆)
2
 544   220 
∆ = 0.3 tanh   + 0.1 tanh   = 0.2
 530 x 2   170 x 2 
440
τmax allowable = x 0.85 = 187 MPa
2
STATIC MARGIN:
τ max allowable 187
= = 1.17
τ 160
τ flf 213
= = 1.33
τ 160
FRAME SIZING (paragraph 6.12)
I0 58759
ρx = = = 10.2 mm (refer to paragraph 2.3)
Sx 0 560
2415091
ρy = = 40.6 mm (small frame section)
1464
7
 
(
 ρx ρy 8
)  τ
x 10 = 16.8 ⇒  IG x 10 3 > 3
4
 3  E 
 r 4 AB 
 

To be conservative, put down E=Ep, giving:

τIG > 211 MPa > τmax allowable
Therefore, the frames are sized in such a manner that general instability is not possible,
which was what was required.


V1-1 •7-STRESSING OF STIFFENED PANELS UNDER COMPRES-

SION AND SHEAR
Ty
G P = - Nx
z x
REMINDER: The curvature of the skin is in the plane normal to the stiffeners (refer to
paragraph 1.4).
To calculate the skin bay folding load and the failure load of a fastened skin-stiffener
panel under combined compression and shear it is first of all necessary to determine
these same values in pure compression (refer to paragraph 4) then in single nominal
shear (refer to paragraph 5 or 6). The following is then obtained:
• skin buckling stresses (σcp0 ; τcp0),
• failure loads (Pallowable,0 ; τflf ; τRp).
It is assumed that the skin failure load in single nominal shear τRp is not modified
significantly by the presence of a compression load. On the other hand, strain induced by
diagonal tension folds in fastened stiffeners outer flange affects the cross sections of the super-
stiffeners and therefore the allowable compression load. In other words, forced crippling
interacts with column buckling.
Naturally, this type of failure does not occur in integral stiffener panels. Here it is
accepted that the initial instability of the stiffener (local buckling or column buckling)
encountered when the panel is subjected to single shear interacts with the column
buckling.
Stiffened panels under compression and shear - Page V1-1•7/1 revision 2

V1-1 •7.1-STRESSING PRINCIPLES
The method developed here is from Kuhn, Peterson and Levin1.
• Under combined loading, the folds in the skin bays occur earlier. The buckling stresses
(σcp;τcp) are obtained using the interaction formula:
2
σ cp  τ cp  σ σ
+   = 1, and before folding: cp = p
σ cp 0  τ cp 0  τ cp τ
• If the failure mode of the super-stiffener in single nominal shear is forced crippling of the
stiffener outer flange, then the failure load is obtained by the interaction formula:
1. 5
Pallowable τ  Pallowable P
+  allowable  = 1 ; and: =
Pallowable .0  τ flf  τ allowable τ
If the failure mode of the super-stiffener under single nominal shear is skin failure, then the
interaction curve above is truncated by the vertical line of abscissa τRp:
Pallowable.0 interaction curve
Pallowable
P
allowable loads
τ τallowable τRp τflf

0 τ
1
Part 1 - Methods of analysis - May 1952

Integrally stiffened panels: The single shear load causing local buckling or column buckling
of the super-stiffener has to be determined. This is an iterative calculation. Then it is possible
to proceed as above replacing τflf by τbuckling.
In all cases, the ideal situation is to determine τflf (or τbuckling) by testing.
• The calculation of stresses in the stiffeners and the transverse members under a given load is
more complex. The premature occurrence of folds affects the diagonal tension factor k. The
presence of a compression stress increases the angle of folds (90° in single compression, 20
to 45° in single shear, depending on the panel curvature). Furthermore, the load-carrying
section under compression is not equal to the effective section under shear. On the other
hand, it is accepted that there is no difference in nature between compression stress due to P
(direct) and the stress due to diagonal tension (indirect). The total stress is simply the
addition of the two.
If the super-stiffener is not symmetrical in relation to the plane (G,x,z) it is necessary to build
an equivalent symmetrical super-stiffener:
Lt1,0 Lt2,0
Y Gp Y
d dp
e1 G e2
Gr dr
h
In the following equations, the index 0 relates to the initial section.
e1 + e 2 e Lt = e 2 Lt 2.0
e= ; Lt0 = 1 1.0
2 e1 + e 2
 σ cp1.0 + σ cp 2 .0   1 1 1 1 
 σ cp 0 =  ⇒ =  + 
 2   R c 2  R c1 R c 2  
e τ + e2 τ2
τ= l 1
2e
k 1.0 + k 2.0
Diagonal tension factors in single shear: k0 =
2
1
1 + k0  ϕ
Rs =  
1 − k0 

V1-1 •7.2-LIMITS
Those given in paragraphs 4, 5 and 6.
V1-1 •7.3.1-Loading rate
• The loading rate in single compression is:

P
Rc =
Pcp0
Remark: A load ratio and not a stress ratio has to be used to eliminate non-linearity
between applied load and stresses. Pcp0 is the beginning of fold load.
• The loading rate in single nominal shear is:

τ
Rs =
τ cp0
• The loading rate under combined loads is:

P τ R + R 2c + 4 R s2
R= = = c
Pcp τ cp 2
τ cp R s σ cp R Pcp
= ; = c =
τ cp 0 R σ cp 0 R Pcp 0
V1-1 •7.3.2-Diagonal tension factor
The general diagonal tension factor is formulated as below:

(R < 1) ⇒ (k = 0)
 R ϕ − 1
(R ≥ 1) ⇒  k = ϕ 
 R + 1
with:
 0.5 < B < 1 ⇒  ϕ = 0.4343 + 260.6 eA 
 A   rB 
 0.5 < A < 1 ⇒  ϕ = 0.4343 + 260.6 eB 
 B   rA 
 0.2 < B < 0.5 
    e
A
 ⇒  ϕ = 0.4343 + 521.2 
 0.2 < A < 0.5   r
 B  
V1-1 •7.4-LOAD-CARRYING CROSS SECTIONS AND WIDTHS
V1-1 •7.4.1-Stiffener direction
Under the combined load (P, τ), the load-carrying widths are:
σ cp
• Compression: LtC = Lt0 ,
σp
Rs Lt
• Shear: LtTD = (1 − k ) 0
R 2
And variations to the corresponding sections are:
• Compression: ∆S Cx = 2 (Lt0 - LtC)e

• Shear: ∆S TD
x = 2 (Lt0 - LtTD)e
The properties of the load-carrying sections are then determined as previously in compression
or single shear (refer to paragraphs 2.2, 4.2 and 5.5).
If the stiffener and the skin are made of different materials, the corrected section design rules
given in paragraph 2.2 are used. Under compression, the secant modulus of the equivalent
homogeneous material is obtained using the secant modulus of the materials determined for
the compression mode alone. Under shear, use the same skin and stiffener secant moduli in the
interpolation law.
C
 S Cxp   Sr  TD
 S TD
xp   S 
E sx =  C  Esp +  C  Esr ; E sx =  TD  Esp +  TDr  Esr
 Sx   Sx   Sx   Sx 

V1-1 •7.4.2-Frame direction
If the frames are fastened to the skin:
ae
∆S TD
y = (1 + k)
2
S TD
y = Sy0 - ∆S TD
y
 S TD
yp   S mt 
E TD
y =  TD  E p +  TD  E mt
 Sy   Sy 
If the frames are floating:
S TD
y = Smt
E TD
y = Emt
V1-1 •7.5-STRESSES IN SUPER-MEMBERS
The compression load, average stress and strain, in the stringer direction, induced by diagonal
tension, are formulated as follows:
kbeτ N TD σ TD
N TD
x = - ; σ TD
x = x
TD
; ε TD
x = x
TD
tan α S x E sx
The stress and strain, in the stringer direction, induced by the single compression load, are
formulated as follows:
C N Cx C σ Cx
σx = C ; εx = C
Sx E sx
The force, the average stress and the strain, in the stringer direction, are formulated as follows:
Nx = N TD
x + N Cx ; σx = σ TD
x + σ Cx ; εx = ε TD
x + ε Cx
The compression due to τ is not constant along a stringer. It is at a minimum level at the
frames (gusset effect) and maximum mid-way from the frames:
σ TD
x min  b σ TD
x min
= (1 - k) 1.78 − 0.64  + k ; TD ≥ 1 ; σx min = σ Cx + σ TD
σx TD  a σx
x min
Frame direction:
TD TD
N TD
y TD
σ TD
y
S TD
y E TD
y

σ TD
y min
= (1 - k) 1.78 − 0.64  + k ; TD ≥ 1
σ TD
y
 b σy
The diagonal tension strain in the skin is:

 2k  τ
εTD =  + (1 − k ) (1 + ν) sin 2 α
 sin 2 α  E sp
The fold angle in relation to the stiffeners is such that:

2
 1 R s  B 
 ε TD − ε y +   
 B < 1 ⇒  cot 2 α = 24 R  r  
A   ε TD − ε x 
 
 
2
 1 Rs  A 
 ε TD − ε y +   tan 2 α 
 B > 1 ⇒  cot 2 α = 8 R r 
A   ε TD − ε x 
 
 
• Bending moment in the super-stiffener, mid-way from the frames:
BA 2
M TDA  = k τe tan α
 
 2 24 r
This moment compresses the stringer outer flange and the skin.
• Bending moments in the super-stiffener, at the frames:
BA 2
• Frames fastened to the skin: M (TD
A) = − k τe tan α
24 r
BA 2
• Floating frames: M (TD
A) = − k τe tan α
12 r
This moment compresses the stringer inner flange.
• Bending moments in floating frames, mid-way from stringers:
B2 A
M TDB  = − k τe tan α
 
 2 24 r
This moment compresses the frame inner flange.

• Bending moments in floating frames, at stringers:
B2 A
M (TD
B) = k τe tan α
12 r
This moment compresses the skin.
Riveted frames: no bending moment in frames.
V1-1 •7.6-STRESSES IN THE SUPER-STIFFENER
V1-1 •7.6.1-In the skin
In the middle between transverse members, the compression in stringer direction is:
M (TD
A/2)
σxp min = σx min - TD
d TD , if integral stiffener
I
E sp  M (TD 
σxp min = C  σ x min − TD ( d TD + e t ) , if fastened stiffener
A / 2)
E sx  I 
V1-1 •7.6.2-In the stiffener
• The average compression stress of the stiffener is:

E
σr = Csr (σx)
E sx
• In the middle between transverse members, the maximum compression stress, the outer
flange stress (fastened stiffeners) and the inner flange stress are respectively:
E
σr min = Csr σx min
E sx
E sr  M (TD
A/2) 
σrs = C
 σ x min − TD
d TD 
E sx  I 
E sr  M (TD 
σ TD
rt = C

E sx 
σ x min +
I
A/2)
TD ( h − d TD )

• At frames, the flange stress is:

E sr  M (TD 
TD (
h − d TD )
A)
σrt = C
 σ x +
E sx  I 

Reminder: If frames are of the floating type, the bending moment is twice as high as the
moment at mid-distance from frames.
V1-1 •7.7-STRESSES IN FRAMES
In frames fastened to the skin, the maximum compression at mid-distance from frames is:
E mt TD
σmt min = TD σ y min
Ey
If frames are floating, apply the bending moments given in paragraph 7.5. The loads are then
applied to the frame cross section alone.
V1-1 •7.8-ALLOWABLE LOADS
V1-1 •7.8.1-At limit load
The folding load/ratio limit load ratio is:

Pcp τ cp P
= =
PLL τ LL PLL R
V1-1 •7.8.2-At ultimate load
Generally, a shear stress (abscissa)-compression load (ordinate) diagram is used. The

combined failure load (τallowable, Pallowable) is, as indicated in paragraph 7.1, at the intersection of
the interaction curve and the straight line passing through points (0,0) and (τ,P).
1. 5
Pallowable τ  P Pallowable
+  allowable  = 1 ; max {τallowable} = min {τRp;τcrit.0}; =
Pallowable .0  τ crit .0  τ τ allowable
The same result is obtained when the following dimensionless diagram is used:
 τ P 
Mark the point  ;  , add, if necessary, the vertical straight line of the abscissa
 τ crit .0 Pallowable .0 
τ Rp
. The allowable load is determined as below.
τ crit.0
FASTENED STIFFENERS: τcrit.0 = τflf
INTEGRAL STIFFENERS: τcrit.0 = τbuckling

Compression-shear (failure) interaction diagram
1
0.9
0.8
0.7
0.6
P
P0 0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
T/T0

V1-1 •7.9-EXAMPLE
Use the example in paragraph 4.6.4 and 5.12 (flat panel: r → ∞).
The load applied to the super-stiffener at UL is:

Nx = - 38000N ; τ = 160 MPa

Ep = 70300 MPa ; σRp = 440 MPa ; σ0.2p = 270 MPa
The properties of the stiffeners and transverse members (7075PLT6) are:

Er = 73800 MPa ; σRp = 495 MPa ; σ0.2r = 420 MPa
The Poisson ratio of the two materials is the same: ν = 0.33
CROSS SECTION, STIFFENER DIRECTION:
170
= =
3
2
35
= =
25
30 2
4
16
Stiffener: Skin:
Sr = 185 mm² Sxp0 = 375 mm²
IYYr = 62186 mm4 IYYp0 = 1485 mm4
dr = 13.9 mm dp0 = - 1.86 mm
WYYr = 2572 mm3 WYYp0 = - 698 mm3

CROSS SECTION, TRANSVERSE MEMBER DIRECTION:
Small section: Smt = 509 mm² ; Sy0 = 1604 mm² ; Syp0 = 1095 mm²
24
4
24
4
90 2.5
120
8
6
3 2
3 35
530
The small section is the one at the stiffener passage. Between the passages, the cross section
area of the transverse member alone is 509 mm². This value is used to calculate the angle of
the folds and the stresses in the transverse member at mid-distance from stiffeners, giving:
Smt = 509 mm² ; Sy0 = 1604 mm²

FAILURE LOAD, MARGIN AT UL (paragraph 7.8.2)
• Column buckling load in single compression: Pcrit.0 = 90200 N
• Forced crippling load (single shear): τflf = 283 MPa
• Skin bay failure shear: τRp = 220 MPa
• Compression-shear interaction:
1.5
Pallowable  τ allowable  P 38000
+  = 1 ; max {τallowable} = 220 MPa ; allowable = = 237.5
90200  283  τ allowable 160
Use of the dimensionless diagram:

τ 160  τ  220 P 38000
= = 0.565 ; max   = = 0.777 ; = = 0.421
τ flf 283  τ flf  283 Pcrit.0 90200
OM
MS = - 1 = 14%
Om
Compression-shear (failure) interaction diagram
1
0.9
0.8
0.7
0.6
P M
P0 0.5
0.4
m
0.3
0.2
0.1
0
0 0.2 0.4 0.6 0.8
T/T0

FOLDING OF POCKETS UNDER SINGLE LOADS
τcp0 = 52 MPa ; σcp0 = 36 MPa ; Pcp0 = 20492 N
160 38000
Rs = = 3.077 ; Rc = = 1.854
52 20492
1.854 2 + 1.854 2 + 4 x 3.077 2
R= = 4.141
2
ϕ = 0.4343
4.1410. 4343 − 1
k= = 0.299
4.1410. 4343 + 1
POCKET FOLDING
 R s 3.077 
 = = 0.743 ⇒ (τcp = 0.743 x 52 = 38.6 MPa)
 R 4.141 
 R C 1.854 0.448 ⇒ (P = 0.448 x 20492 = 9175 N ; σ = 16.1 MPa)
 = =  cp cp
 R 4.141 
 P = 38000 = 25330 N ⇒ folding at: 9175 = 36% LL

 LL 1.5  25330
LOAD-CARRYING SECTIONS AND WIDTHS AT UL (paragraph 7.4)
• Stiffener direction:
Modelling of material curves and load-carrying sections in compression

0.00023 16.1 0.00023 16.8 0.023% 16.1 16.9 16.4
0.00051 36 0.00051 37.8 0.051% 36.0 37.8 36.7
0.00149 104.73 0.00149 110 0.149% 104.7 110.1 107.6
epsilon Sp S P Esp Esr Es
0.023% 375.0 560.0 9164 70300 73800 71456
0.051% 262.4 447.4 16437 70300 73800 71747
0.149% 168.3 353.3 38001 70181 73800 72076
(These tables are compiled in the same way as those on page V1-1•4/19, with σcp = 16.1.)

S Cx = 353 mm²
Sr = 185 mm² ; S TD
xp = 168 mm²
C
E sx = 72076 MPa
Esp = 70181 MPa ; Esr = 73800 MPa
85
LtTD = 0.743 x (1 - 0.299) x = 22,1 mm
2
∆S TDx = 2 (85 - 22.1) x 2 = 251 mm²
TD
S x = 560 - 251 = 309 mm²
Sr = 185 mm² ; S TD
xp = 124 mm²
124   185  x 73800 = 72351 MPa

E sxTD =   x 70181 x  
 309   309 
• Frame direction (fastened to the skin):
530 x 2
∆S TD
y = 1.299 x = 689 mm²
2
S TD
y = 1604 - 688 = 915 mm²
509   509  x 73800 = 72246 MPa
E TD = 1 −  x 70300 +  
y
 915   915 
STRESSES AND STRAIN, STRINGER AND FRAME DIRECTION, TD ANGLE

AT UL (paragraph 7.5)
• Compression, in stiffener direction, due to TD:
kbeτ
N TD
x =- = - 0.299 x 170 x 2 x 160 x cot α = - 16266 cot α
tan α
− 16266 − 52.6
σ TD
x = = - 52.6 cot α ; ε TD
x = = - 0.073% cot α
309 72351
• Single compression load, in stiffener direction:
N Cx = - 38000 N
σ Cx = - 107.6 MPa ; ε Cx = - 0.149%

• Compression, stiffener direction:
Nx = - 16266 cot α - 38000

σx = - 52.6 cot α - 107.6 ; ε TD
x = - 0.073% cot α - 0.149%
• Compression, transverse member direction:
N TD
y = - 0.299 x 530 x 2 x 160 x tan α = - 50730 tan α
TD
N TD
y TD
σ TD
y
σ y = = - 55.4 tan α ; ε y = = - 0.077%.tan α
S TD
y E TD
y
• Diagonal tension:
 2 x 0.299  160
εTD =  + (1 + 0.33) (1 − 0.299) sin 2 α
 sin 2 α  70181
0.136%
εTD = + 0.213% x sin2α
sin 2 α
 B < 1 ⇒  cot 2 α = ε TD − ε y  ⇒ α = 48.6° ; tan α = 1.136

 
A   ε TD − ε x 
52.6
σ TD
x =- = - 46.4 MPa ; σx = σ TD
x + σ Cx = - 46.4 - 107.6 = - 154 MPa
1136
.
σ TD b
x min
= (1 - k) 1.78 − 0.64  + k = 0.7 x (1.78 - 0.64 x 0.32) + 0.3 = 1.4
σ TD
x
 a
σ TD TD C
x min = - 65.1 MPa ; σx min = σ x min + σ x = - 172.7 MPa
• Transverse super-member stresses:
σ TD
y = - 55.4 x 1.136 = - 62.9 MPa
  a   σ TD
y min

 (1 − k ) 1.78 − 0.64  + k < 1 ⇒  TD = 1
 b   σy 
σ TD
y min = - 62.9 MPa

SKIN COMPRESSION, STIFFENER DIRECTION (paragraph 7.6)

C
E sp 70181
σp min = C
σx min = - 172.7 x = - 168.3 MPa
E s 72076
STRESSES IN THE STIFFENER
E srC 73800
σr min = C σx min = - x 172.7 = - 176.9 MPa
Es 72076
STRESSES IN THE FRAMES (refer to paragraph 7.7)
E mt 73800
σ TD
mt min = TD
σy min = - 62.9 x = - 64.3 MPa
Ey 72246


CONTENTS
issue date change
V1-1 STIFFENED PANELS 0 6/1999 Creation

V1-1 8CALCULATING STIFFENED PANELS COMPRESSION AND 0 6/1999 Creation
BENDING 0 6/1999 Creation
V1-1 9ALLOWABLE STRESS 0 6/1999 Creation
V1-1 10EXAMPLE 0 6/1999 Creation
V1-1 8 CALCULATING STIFFENED PANELS

COMPRESSION AND BENDING 1
V1-1 8.1 INTRODUCTION 1
V1-1 8.2 ALLOWABLE STRESSES AT ULTIMATE LOAD 2
V1-1 8.3 POCKET FOLDING LOAD COMPRESSION/BENDING 2
V1-1 8.3.1 Detection of minimum folding stress σcp 2
V1-1 8.3.2 Search for transmitted load Pt 2
V1-1 8.3.3 Example 5
V1-1 8.4 COMPRESSION/BENDING STRESS 8
V1-1 8.4.1 Position of extreme fibres 8
V1-1 8.4.2 Compression + bending stresses at Max moment point 8
V1-1 8.4.3 Compression + bending stresses at Min moment point 9
V1-1 8.4.4 Limits, accuracy 10
V1-1 8.4.5 Sign conventions 10
V1-1 8.4.6 Example 11
V1-1 8.5 AMPLIFIED BENDING MARGIN 14
Issue 1 Contents page V1-1•i

V1-1 8 CALCULATING STIFFENED PANELS COMPRESSION AND

BENDING
V1-1 8.1 INTRODUCTION
M2
G
M1
y
P = - Nx
z
x
A super-stiffener submitted to both axial and bending loads behaves as a beam-column.
Note that this combined loading condition corresponds to several practical utilisation cases for thin sheets:
section 21 upper surface panel, etc.
The super-stiffener instability mode is described in § 1.2 ("behaviour, failure modes"). As the cross
members are sufficiently stiff, the general buckling of the panel cannot occur.
Two compression/bending stress calculation methods are given in § V1-1 8.4. The choice of one or the
other of these two methods depends on the field of use explained in § V1-1 8.4.
Issue 1 Calculating Stiffened Panels Compression And Bending −1

page V1-1•8−
 AEROSPATIALE 1999
V1-1 8.2 ALLOWABLE STRESSES AT ULTIMATE LOAD
Refer to chapter V1-3 ("STIFFENERS") for local bending stress calculation.
Allowable = minflange crippling, flange local buckling, lateral

buckling
Allowable = web local buckling
Allowable = minbase crippling, stiffener inter-rivet buckling
Allowable = skin inter-rivet buckling
The critical crippling stresses will be calculated as follows:
S talon ⋅ σ crit . talon + 1 ⋅ Sâme ⋅ σ crit .âme crip.talon = flange crip.

σ crip. talon = 2 talon = flange
1
S talon + ⋅ S âme crit.talon = flange crit.
2 âme = web
crit.âme = web crit.
1 semelle = base
S talon ⋅ σ crit .semelle + ⋅ S âme ⋅ σ crit .âme crit.semelle = base crit.
σ crip.semelle = 2
1
Ssemelle + ⋅ S âme
2
V1-1 8.3 POCKET FOLDING LOAD COMPRESSION/BENDING
V1-1 8.3.1 Detection of minimum folding stress σcp
To do these calculations, refer to next chapter (V1-2 "Buckling of plates and thin shells").
V1-1 8.3.2 Search for transmitted load Pt
By varying ε, we search for transmitted load Pt for which the compression/bending stress at max moment point of the
panel σcfmax-p is equal to the minimum folding stress σcp.

page V1-1•8−
For a given strain value ε, we calculate:
• The corresponding characteristics of the super-stiffener:
♦ Calculation of the simple patch compression stresses (panel + stiffener) using the modified Ramberg &
Osgood formula:
n
σi  σ  i
ε= + 0,002 ⋅  i  ⇒σ
i ∀i = p , r.
Ei σ 
 0i 
♦ Calculation of the load-carrying section: S

The aim is to obtain a max skin stress equal to the minimum folding stress, therefore the load-carrying
section is equivalent to the initial section.
Sp = Sp0.
S = Sp0 +Sr
♦ Calculation of transmitted load Pt
Pt = Sp.σp + Sr .σr
♦ Calculation of the tangent and secant moduli for each patch i: Esi and Eti
Calculation of the tangent and secant moduli of the equivalent homogeneous material: Es and Et
 σi  r Si
 E si = E s = ∑ ⋅ E si
 ε  i=p S
 ∀i = p, r ⇒ 
n 1 − ni r Si
 1 = i + 
 E ti E si Ei E t = i=
∑
p S
⋅ E ti
 
♦ Calculation of the static moments in relation to YY: WYYi
See § V1-1.2.2
♦ Calculation of the offset in relation to the CG: d
See § V1-1.2.2
♦ Calculation of the moment of inertia in relation to YY: IYY
See § V1-1.2.2
♦ Calculation of the moment of inertia at the CG: I
See § V1-1.2.2

page V1-1•8−
• The resulting end and original moments
We correct the end and original moments applied:
P P
M1 = M1 × t M2 = M2 × t
P P
• The position of the extreme fibre of the panel: Vp
Vp represents the position of the extreme fibre of the panel pockets in relation to the CG.
• The compression + bending stress at Max moment point: σcfmax-p
We calculate the maximum amplified moment Max and its position Xmax in relation to the origin (see beams-
columns §).
The stress is obtained by the normal bending formulas consisting of:
♦ Part due to compression directly related to ε: σp

♦ Part due to bending ("Mmax.Vp/I") calculated with the Vp and I values corresponding to ε.
Vp
⇒ σ cf max −p = σ p + M max×
I
r
Having determined the transmitted load Pt = ∑ S i ⋅ σ i , we say that the folding load is reached when
i=p
Pcp = Pt.

page V1-1•8−
V1-1 8.3.3 Example
Let us take the example of § 2.3

The characteristics of the skin material (2024 PLT 351) are:
Ep = 70300 MPa, σ0.2p = 270 MPa, np = 7.05.
The characteristics of the stiffener material (7075 T 73510) are:

Er = 73800 MPa, σ0.2r = 420 MPa, nr = 13.83.
170
2 3
35
25
3
2
30
4
16
Stiffener: Skin:
Sr = 185 mm² Sp0 = 375 mm²

IYYr = 62186 mm4 IYYp0 = 1485 mm4
dr = 13.9 mm dp0 = -1.86 mm
WYYr = 2572 mm3 WYYp0 = -698 mm3
Definition of loading case at UL:
The compression load is: P = 38000 N

The distributed lateral load is: q = 10 N/mm
The original bending moment is: M1 = -224083 mm.N
The end bending moment is: M2 = -212083 mm.N
q
M1 M2
P P

page V1-1•8−
By varying ε, we try to find the transmitted load Pt for which the compression/bending stress at max moment point of
panel σcfmax-p is equal to the minimum folding stress σcp = 36 MPa.
After numerical resolution, we obtain:
(σ p = 29 MPa ) → (S p = S p0 = 375 mm ² ; S = S0 = 560 mm ² )

 7,05
ε = 29 + 0,002 ⋅  29  ≈ 29 = 0,0413 %
and:  70300  270  70300
σ = 30, 4 MPa
 r
Stress which verifies:
13,83
30, 4  30, 4  30, 4
ε= + 0,002 ⋅   ≈ = 0,0413 %
73800  380  73800
- The compression load corresponding to these stresses is equal to:
Pt = 29X 375 + 30.4 X 185 = 16507 N.
- Deduce the corrected end and original moments:
M1 = −224083 × 16507 = -97341 mm.N.

38000
M 2 = −212083 + 22000 × 16507 = -92128 mm.N.
38000
- Determine the maximum bending moment expressions
Sp S
Et = ⋅ E tp + r ⋅ E tr = 375 × 70300 + 185 × 73800 = 71456 MPa
S S 560 560
EtI 71456 × 58759 = 504 mm.

j= =
P 16507
By using case 6 (see figure V2-6 2.1), we obtain:

M2 − M1 ⋅ cos L − 92128+ 97341× cos
530
j
C1 = j2 ⋅ q ⋅ tan L + = 5042 × 4,3 × tan 530 + 504
.
2j L 2 × 504 530
sin sin
j 504
C1 = 590209 mm.N.
C2 = M1 + j2.q = -97341+5042 X 4.3 = 1007581 mm.N.
f(x) = -q = -4.3 N/mm.

page V1-1•8−
M = 590209 ⋅ sin x + 1007581 ⋅ cos x − 4,3 ⋅ j 2

j j
- Determine the maximum bending moment
C1
x = j ⋅ Arc tan = 267.2 mm.
C2
267, 2 267,2
M max = 590209 ⋅ sin + 1007581 ⋅ cos − 4,3 × 504 2 = 62796.5 mm.N.
504 504
- Calculate total stress at panel
• Stress due to axial load:
σ c = σ p = 29 MPa.
• Stress due to bending moment:
I
( )
σ f = M ⋅ Vp =
62796,5 × 6,5
58759
= 7 MPa
• Maximum compression stress:
σ cf max −p = σ c + σ f .
➾ σ cf max −p = σ c + σ f = 36 MPa
Therefore σ cf max −p = σ cp .

page V1-1•8−
V1-1 8.4 COMPRESSION/BENDING STRESS
V1-1 8.4.1 Position of extreme fibres
Calculation of the positions of the extreme fibres of the pocket under study.
V1-1 8.4.2 Compression + bending stresses at Max moment point
➊ High compression, relatively low bending
We calculate the Maximum amplified moment Mmax and its position Xmax in relation to the origin (see
§ V2.6 "Beam-column").
The stresses are obtained by the normal bending formulas consisting of:
- part due to compression directly related to ε: σp

- part due to bending ("Mmax Vi/I") calculated with the values of Vi and I corresponding
to ε. The reduction in the load-carrying width under the effect of the moment is
therefore neglected (compression load predominant by hypothesis).
➋ High bending
We calculate the Maximum amplified moment Mmax and its position Xmax in relation to the origin (see
- part due to compression: conservatively, we consider, for the calculation of the load-
carrying section, that σp = σlim-p, the load-carrying section is then assumed to be
constant.
- part due to bending ("Mmax.Vi/I") calculated with the values of Vi and I corresponding
to σlim-p
For added Super-Stiffener:
Patch 2: Stiffener
Patch 1: Panel
Base Web Flange
Compression/bending stresses at max.
σcfmax-p σcfmax-s σcfmax-a σcfmax-t
moment point
For integrated Super-Stiffener:
Patch 2: Stiffener
Patch 1:Panel
Web Flange
Compression/bending stresses at max.
σcfmax-p σcfmax-a σcfmax-t
moment point

page V1-1•8−
V1-1 8.4.3 Compression + bending stresses at Min moment point
➊ High compression, relatively low bending
We calculate the Minimum amplified moment Mmin and its position Xmin in relation to the origin (see
- part due to compression directly related to ε: σp

- part due to bending ("Mmin Vi/I") calculated with the values of Vi and I
corresponding to ε. The reduction in the load-carrying width under the effect of the
moment is therefore neglected (compression load predominant by hypothesis).
➋ High bending
We calculate the Minimum amplified moment Mmin and its position Xmin in relation to the origin (see
- part due to compression: conservatively, we consider, for the calculation of the

load-carrying section, that σp = σlim-p, the load-carrying section is then assumed to
be constant.
- part due to bending ("Mmin.Vi/I") calculated with the values of V and I
corresponding to σlim-p
For added Super-Stiffener:
Patch 2: Stiffener
Patch 1: Panel
Base Web Flange
Compression/bending stresses at min.
σcfmin-p σcfmin-s σcfmin-a σcfmin-t
moment point
For integrated Super-Stiffener:
Patch 2: Stiffener
Patch 1: Panel
Web Flange
Compression/bending stresses at min.
σcfmin-p σcfmin-a σcfmin-t
moment point

page V1-1•8−
V1-1 8.4.4 Limits, accuracy
Typically, method ➊ is applicable for Airbus Sect 21 upper surface panel calculations.
In cases of doubt, use method ➋ which is a fast and unfavourable method. For AU4G sections, we obtain the
well-known approximation of the load-carrying width: Lt = 15e.
A minimum margin of 15% is required.
V1-1 8.4.4 Sign conventions
The sign conventions used in the formulas below are as follows:
The compression loads are positive.

A positive bending moment compresses the panel.
A positive distributed lateral load leads to a positive bending moment.
The compression stresses are positive.
Original bending moment End bending moment

(+ compresses the panel) (+ compresses the panel)
Normal load
(+ compression) Panel
Distributed lateral load Stiffener

(+ compresses the panel)

page V1-1•8−
V1-1 8.3.3 Example

Ep = 70300 MPa, σ0.2p = 270 MPa, np = 7.05.

Er = 73800 MPa, σ0.2r = 420 MPa, nr = 13.83.
170
2 3
35
25
3
2
30
4
16
Stiffener: Skin:
Sr = 185 mm² Sp0

IYYr = 62186 mm4 IYYp0 = 1485 mm4
dr = 13.9 mm dp0 = -1.86 mm
WYYr = 2572 mm3 WYYp0 = -698 mm3

q
M1 M2
P P

page V1-1•8−
Method used εp σp εr σr ε σp σr σ
➊ 0.001197 84.1 0.001197 88.4 0.001197 84.1 88.4 85.9
➋ 0.001531 107.4 0.001531 113.0 0.001531 107.4 113.0 110.4
The input data are the skin and stiffener stresses, columns 2 and 4. Column 7, the stiffener stresses corresponding to
ε = εp are obtained by linear interpolations in columns 3 and 4.
np nr
σp  σp  σ  σ 
εp = + 0,002 ⋅  
 ; ε r = r + 0,002 ⋅  r 
 ; σ= P
Ep  σ 0, 2 p  Er  σ 0, 2 r  S
Method used εp Sp S C Pt Esp Esr Es Etp Etr Et

➊ 0.001197 257.4 442.4 1.00 38000 70266 73800 71743 70060 73800 71624
➋ 0.001531 159.2 344.2 1.00 38000 70150 73800 72112 69259 73800 71700
Columns 2 and 4: Simplified method ➊:

 σ cp 
S p = S p 0 − be ⋅ 1 −  ; P = Sp ⋅ σ p + Sr ⋅ σ r
 σp 
 
Simplified method ➋:
 σ cp 
 σ lim −p 
 
Columns 5 to 7: Es = σ
ε
−1
 
Columns 8 and 9: E t =  n + 1 − n 
E
 s E 
Sp Sr
Column 10: Et = ⋅ E tp + ⋅ E tr
S S
Method used σ S WYYp d IYYp I P Pcrit σp σr

➊ 85.9 442.4 - 463 4.96 975 54060 38000 136044 84 88
➋ 110.4 344.2 - 266 6.90 550 47812 38000 120448 107 113
Columns 3 to 6: ∆S = S − S 0 ; WYYp = WYYp 0 +  e t − e  ⋅ ∆S ;

 2
 sp
E E   2 2 
d = 1 ⋅  ⋅ WYYp + sr ⋅ WYYr  ; I YYp = I YYp 0 −   e t − e  + e  ⋅ ∆S
S  Es Es   2 12 
E sp E
I= ⋅ I YYp + sr ⋅ I YYr − Sd 2
Es Es

page V1-1•8−
π2 ⋅ E t ⋅ I
Column 8: Pcrit = (Column).
L2
Method used Vp Vs Vt J q M1 M2 C1 C2 F(x)

➊ - 7.96 - 4.96 25.04 319 10.0 - 224083 - 212083 881482 794854 - 1018938
➋ - 9.90 - 6.90 23.10 300 10.0 - 224083 - 212083 836273 678048 - 902131
Columns 1 to 3:
Vp d
Vs
Vt
EtI
Column 4: j=
P
Columns 8 to 10: See chapter V2-6 "Beams-Columns".
Method used Xmin Xmax Mmin Mmax

➊ 0.0 267.2 - 224083 167993
➋ 0.0 267.2 - 224083 174485
Columns 1 to 2:
ptmax
ptmin
Xmax
C1
X max = j ⋅ Arc tan
C2
Columns 3 to 4: M = C1 ⋅ sin x + C 2 ⋅ cos x − q ⋅ j 2

j j
Method used σcfmax-p σcfmax-s σcfmax-t σcfmin-p σcfmin-s σcfmin-t

➊ 108.8 103.8 10.5 51.1 67.8 192.2
➋ 143.5 138.2 28.7 61.0 80.7 221.3
Columns 1 to 6: σ cfi − j = σ j +
Mi
I
(
⋅ − Vj )

page V1-1•8−
V1-1 8.5 Amplified bending margin
The amplified bending margin is calculated in three steps:
- Calculation of the amplified bending margin at max moment point: MSMmax

- Calculation of the amplified bending margin at min moment point: MSMmin
- Calculation of the minimum margin: MSmin = Min(MSMmax ; MSMmin)
The margins at the max moment points and at the min moment point are calculated in the same way.
A multiplication factor C is applied to the load P, to the distributed load q and to the end and original moments. By
varying C, we try to find the new matrix for which one of the corresponding compression/bending stresses σcfmax(or
min)-j is equal to its limit stress.
• If one of the allowable stresses is reached at the max (or min) moment point, we calculate the margin:
MSMmax(or min) = C – 1.
• If no limit stress is exceeded before the compression load reaches the general buckling load PadmCE of the super-
stiffener, we are in the maximum allowable load case. The margin is calculated in relation to the buckling.
If C x P = PadmCE ➾ MSMmax(or min) = C – 1.
Example:

Ep = 70300 MPa, σ0.2p = 270 MPa, np = 7.05.

Er = 73800 MPa, σ0.2r = 420 MPa, nr = 13.83.


page V1-1•8−
Method used εp σp εr σr ε σp σr σ
➊ 0.002590 174.9 0.002590 191.1 0.002590 174.9 191.1 182.9
➋ 0.002645 177.9 0.002645 195.2 0.002645 177.9 195.2 187.2
The input data are the skin and stiffener stresses, columns 2 and 4. Column 7, the stiffener stresses corresponding to
ε = εp are obtained by linear interpolations in columns 3 and 4.
np nr
σp  σp  σ  σ 
εp = + 0,002 ⋅  
 ; ε r = r + 0,002 ⋅  r 
 ; σ= P
Ep  σ 0, 2 p  Er  σ 0, 2 r  S
Method used εp Sp S C Pt Esp Esr Es Etp Etr Et

➊ 0.002590 189.2 374.2 1.80 68460 67546 73796 70635 54603 73741 64064
➋ 0.002645 159.2 344.2 1.70 64433 67260 73794 70773 53313 73723 64285
Columns 2 and 5: Simplified method ➊:

 σ cp 
 σp 
 
Simplified method ➋:
 σ cp 
 σ lim −p 
 
Pt
Column 4: C=
P
Columns 5 to 7: Es = σ
ε
−1
 
Columns 8 and 9: E t =  n + 1 − n 
E
 s E 
Sp Sr
Column 10: Et = ⋅ E tp + ⋅ E tr
S S
Method used σ S WYYp d IYYp I P Pcrit σp σr

➊ 182.9 374.2 - 326 6.35 680 50548 68460 113780 175 191
➋ 187.2 344.2 - 266 7.06 550 48223 64433 108921 178 195
Columns 3 to 6: ∆S = S − S 0 ; WYYp = WYYp 0 +  e t − e  ⋅ ∆S ;

 2
 E sp E   2 2 
d = 1 ⋅  ⋅ WYYp + sr ⋅ WYYr  ; I YYp = I YYp 0 −   e t − e  + e  ⋅ ∆S
S  Es Es   2 12 

page V1-1•8−
E sp E sr
I= ⋅ I YYp + ⋅ I YYr − Sd 2
Es Es
π2 ⋅ E t ⋅ I
Column 8: Pcrit = (Column).
L2
Method used Vp Vs Vt J C q M1 M2 C1 C2 F(x)

➊ - 9.35 - 6.35 23.65 217 1.80 18.0 - 403704 - 382085 1253077 448482 - 852186
➋ - 10.06 - 7.06 22.94 219 1.70 17.0 - 379957 - 359610 1179311 435836 - 815793
Columns 1 to 3:
Vp d
Vs
Vt
EtI
Column 4: j=
P
Columns 6 to 8: q = C x qCE
Mi = C x MiCE
Columns 9 to 11: See chapter V2-6 "Beams-Columns".
Method used Xmin Xmax Mmin Mmax

➊ 0.0 266.9 - 403704 478730
➋ 0.0 266.9 - 379957 441482
Columns 1 to 2:
ptmax
ptmin
Xmax
C1
X max = j ⋅ Arc tan
C2

page V1-1•8−
Columns 3 to 4: M = C1 ⋅ sin x + C 2 ⋅ cos x − q ⋅ j 2

j j
Method used σcfmax-p σcfmax-s σcfmax-t σcfmin-p σcfmin-s σcfmin-t MS

➊ 263.4 251.2 - 32.9 100.3 140.4 380.0 0.80
➋ 270.0 259.8 - 14.8 98.7 139.6 376.0 0.70
Columns 1 to 6: σ cfi − j = σ j +
Mi
I
(
⋅ − Vj )
Column 7: MS = C – 1.

page V1-1•8−
SSM V1-2 • BUCKLING OF THIN PLATES AND SHELLS
CONTENTS
revision Date
V1-2 Buckling of thin plates and shells 1 07/1996 Original issue

V1-2•1 General information 1 07/1996 Original issue
V1-2•2 Flat plates 1 07/1996 Original issue
V1-2•3 Curved plates 1 10/1996 Original issue
V1-2•4 Cylinders and other shells 1 03/1997 Original issue
1- GENERAL
1.1- Introduction (V1-2•1/1)
1.2- Single and combined loading (V1-2•1/2)
1.3- Plasticity corrections (V1-2•1/3)
1.4- Field of application (V1-2•1/4)
2- FLAT PLATES
A - Summary of cases dealt with (V1-2•2/2)
B - Flat plate analysis procedure (V1-2•2/4)
2.1- Single loading (V1-2•2/5)
2.2- Plasticity correction factors (V1-2•2/6)
2.3- Buckling factors (V1-2•2/7)
2.3.1- Rectangular flat plates
2.3.2- Flat parallelogram plates
2.3.3- Flat triangular plates
2.3.4- Flat plates with evolutive width
2.4- Combined loading (V1-2•2/27)
2.4.1- Combined loading without transverse loading
2.4.2- Combined loading with transverse loading
2.5- Examples (V1-2•2/38)
3- CURVED PLATES
B - Curved plate analysis procedure (V1-2•3/3)
3.1- Single loading (V1-2•3/4)
3.3- Buckling factors (V1-2•3/6)
3.3.1- Rectangular curved plates
3.4- Combined loading (V1-2•3/13)
3.4.1-Shear + compression (or tension)
3.4.2- Compression + side pressure
3.4.3- Shear + side pressure
3.5- Examples (V1-2•3/15)
4- CYLINDERS AND OTHER SHELLS
Revision 1 Contents page V1-2.i
© AEROSPATIALE - 1998
B - Circular cylinder analysis procedure (V1-2•4/3)

C - Elliptical cylinder analysis procedure (V1-2•4/4)
D - Spherical shell analysis procedure (V1-2•4/5)
E - Conical shell analysis procedure (V1-2•4/5)
4.1- Behaviour of circular cylinders under compression (V1-2•4/6)
4.1.1- Introduction
4.1.2- Linear theory and test results
4.3- Buckling factors (single loading) (V1-2•4/9)
4.3.1- Circular cylinders
4.3.2- Elliptical cylinders
4.3.3- Spherical shells
4.3.4- Conical shells
4.4- Interaction equations (combined loading) (V1-2•4/28)
4.4.1- Circular and elliptical cylinders
4.4.2- Conical shells
4.5- Examples (V1-2•4/33)
Revision 1 Contents page V1-2.ii

SYMBOLS USED
a: Bay length, long axis of an ellipse kpl: Buckling factor of flat circular plates
b: Bay width, short axis of an ellipse L: Length
Ci: Constants (curve equations) m: number of half waves
-0.15
C= 0.6*In (3.748*(r/e) ) (constant) MS: Static margin
Cf: 1.3*C (constant in cylinder bending) Mf: Bending moment
UL: Ultimate loads Mf,cr: Critical bending moment
LL: Limit loads Mt: Torsion moment
d: Diameter Mt,cr: Critical torsion moment
e: Plate or cylinder thickness Mapp: Applied moment
E: Young's modulus n: Stress-strain shape factor (Ramberg and
Es: Secant modulus Osgood)
Et: Tangent modulus n: Number of circumferential waves
G: Shear modulus n0: Number of circumferential waves when
Gs: Secant shear modulus L=r
I: Bending stiffness constant (inertia) p: Pressure
J: Torsion stiffness constant Pint: internal pressure
K,k: Buckling factors Pcr: Critical pressure (external)
Kc,kc: Compression buckling factors Pc,cr: Critical compression load
Ks,ks: Shear buckling factors Papp: Applied load
Kf,kf: Bending buckling factors r : Radius or small radius
Kt,kt: Torsion buckling factors r: Critical curvature radius
Ky,ky: Buckling factors under external radial Ri: Loading rate for load i alone
pressure Rc: Loading rate in single compression
Kp,kp: Buckling factors under hydrostatic Rs: Loading rate in single shear
pressure Rf: Loading rate in single bending
k c ,k f ,k p : Buckling factors at the border of Rp: Loading rate under pressure
* * *
two curves Rcs: Loading rate under a shear load

kc0: Buckling factors of column plates under Rct: Loading rate in torsion
single compression
Revision 1 Contents page V1-2.iii

R: Large radius or loading rate under σiadm,0: Buckling stress corresponding to load i
combined loads alone
Sa: Strength modulus of the elliptic cylinder σcc,0: Critical single compression stress
cross section σcf,0: Critical single bending stress
Se: Strength modulus of the elliptic cylinder σcy,0: Critical stress under hydrostatic or radial
cross section pressure
r σcp,0: Critical stress under external pressure
t= 1 − ν 2e (constant)
e σcxe: Elastic longitudinal critical stress
v: Distance to the neutral axis σcye: Elastic transverse critical stress
ycr: Critical curvature radius position from the σcx: Longitudinal critical stress corrected and
neutral axis taking transverse load into account
L2
Z= 1 − ν 2e (constant) σc,cr: Critical compression stress
re
σf.cr: Contrainte critique en flexion
Z* = Z constant at the border of two curves
σc,cr: Critical compression stress
αβγ: Exponents in interaction curve equations
σc,cr: Critical compression stress
α: Also used to designate angles
4Z σt: Longitudinal tension stress
β: 2 (used here as a constant)
π t ∆σcc,0: Increase in critical compression stress
ρ: Minimum curvature radius due to internal pressure
ρmoy: Average curvature radius ∆σcf,0: Increase in critical bending stress due
ε: Strain (expansion) to internal pressure
Φ: Flow τ: Shear stress
η: Plasticity factor τR: Failure shear stress
νe: Poisson's ratio - elastic τcr: Critical shear stress
νp: Poisson's ratio - plastic τcr,0: Critical single shear stress
ν: Poisson's ratio - elastic - plastic τct,0: Critical single torsion stress
σ: Normal stress τcs,0: Critical shear stress due to a single shear
σ0: Reference stress for transverse loading load
σa: Stress at position a τt,cr: Critical torsion stress
σ0.2: Conventional allowable compressive
yield stress
σR: Tension failure stress
σeq: Equivalent normal stress
σcr: Critical stress
Revision 1 Contents page V1-2.iv

BIBLIOGRAPHY
1- SCHILDCROUT & STEIN

Critical combination of shear and direct axial stress for curved rectangular panels.
NACA T.N 1928, August 1949.
2- G. GERARD & H. BECKER

HANDBOOK OF STRUCTURAL STABILITY
PART I: Buckling of flat plates. NACA T.N 3781, July 1957
PART III: Buckling of curved plates and shells. NACA T.N 3783, August 1957.
3- ENGINEERING SCIENCES DATA UNIT

Volume 1: General Data: Plasticity correction factors for plate buckling.
Volume 2: Buckling of rectangular plates and panels in compression.
Buckling of flat isotropic plates under uniaxial and biaxial loading.
Volume 9: Cylinders and Pipes.
Volume 10: Buckling of flat rectangular plates and panels in shear.
4- BRUHN
ANALYSIS & DESIGN OF FLIGHT VEHICLE STRUCTURES
5- R.J. ROARK & W.C. YOUNG

FORMULAS FOR STRESS AND STRAIN
Revision 1 Contents page V1-2.v

Revision 1 Contents page V1-2.vi

V1-2 •1-GENERAL
V1-2 •1.1-INTRODUCTION
The definition and the description of the general behaviour of plates is given in Volume 2 of
this manual. Simply remember that a thin plate is a prismatic body which is thin in relation to
its other two dimensions. Thin plates are items with two dimensions.
The buckling stress of a thin flat plate is the stress at which a change to the stable
configuration of equilibrium occurs. The plate was flat then becomes slightly buckled. An
increase in the external load beyond this limit cause an accelerated increase in deflection. All
external loads, no matter how distributed are contained within the central plane (x,y).
z
y
φy
φxy
φx
x
The flat plate stability theory is flanged on the "thin plates with low deflection" assumptions.
All components of the stress may be expressed by the deflection which is a two dimension
function in the central plane.
Structural items held between sheets (stiffened panel bays, spar webs, etc.) may naturally be
considered as thin plates. Most angle sections may also be considered as consisting of an
assembly of plates. Thus, this Chapter 2 of Volume 1 can be used to determine load buckling
stresses in stringers, frames, stiffeners and other members.
A shell is always thin in relation to its other dimensions and its curvature radii. The surface
bisecting the thickness of a shell is the average surface. A shell is entirely defined
geometrically when the shape of its average surface and its thickness at each point is specified.
To analyse it, an infinitely small plate, limited by two pairs of adjacent planes, normal to the
surface and containing its principal curvatures is cut from it.
Revision 1 General page V1-2•1.1

The theory of small deformations, which gives good results for buckling flat plates in the
linear elastic domain does not provide good results for buckling stresses of curved plates and
shells. For curved plates and shells it is necessary to select the large displacement assumption
and introduce a perturbation. This type of approach requires the use of the finite element
method.
This chapter of the static stressing manual presents the results of the theoretical analysis of flat
plates and a semi-empirical approach for shells and curved plates.
V1-2 •1.2-SINGLE AND COMBINED LOADING
The buckling stress of a plate subjected to a single load, i.e. unidirectional, may be formulated
as follows:
2
 e
σcr (or τcr) = ηKE  
 b
K, the buckling factor (linear elasticity) depends on the supporting conditions at the edges and
the dimensions of the plate. The same applies to the plastic correction factor η.
The plate is stable for as long as its buckling stress is greater than the stress σ (or τ) of which it
φ
is the seat. This stress is then linear with the applied load: σ =
e
To determine the buckling stresses of a plate subjected to combined loads, the interaction
principle is used:
A plate subjected to a system of m single loads, where σi is the stress and σiadm,0 the
buckling stress (or allowable if buckling is not possible) corresponding to the load i alone.
The loading rate for load i alone is as follows:

σi
Ri =
σ iadm,0
Interdependence of m loads is represented by an interaction equation, the general

formulation of which is:
R = R 1α1 + R α2 2 + ... + R αi i + ... + R αmm
The exponents αi are determined by the theory or experimentally. Buckling occurs when
the interaction equation result reaches 1.
R is the loading rate under combined loads. R is such that:

α1 αi αm
 R 1  + ... +  R i  + ... +  R m  = 1
     
 R  R  R 
This equation is the one of the required border. If this plate is subjected to two loads, it
can be graphically illustrated in the form of an interaction curve.

The border is called the interaction curve and the margin with regards to buckling is:
1 σ1 σ2
MS = -1 R= =
R σ 1adm σ 2 adm
σ1adm.0 interaction curve
σ1adm
σ1
Allowable loads
σ2 σ2adm σ2adm,0
0
V1-2 •1.3-PLASTICITY CORRECTIONS
In addition to the type of plate and the boundary conditions, the plastic correction factor η
depends on E, Et, Es and ν. These latter values depend on the stress value to be calculated.
The equation of the buckling stress can be formulated as follows:
σcr = ησcre
Where σcre is the linear elastic buckling stress. Therefore, this calculation is iterative.
In practice, for as long as σcre is less than the proportionality limit σe of the material, the
plastic correction factor may be considered as being 1.
σ
For standard aluminium alloys: σe = 0.2
2
Beyond that, the Ramberg and Osgood model is used:
n
σ  σ 
ε = + 0.002  
E  σ 0 .2 
σ
Es =
ε
1 n 1−n
= +
E t Es E

The elastic-plastic Poisson ratio ν is obtained by interpolation between the elastic Poisson
ratio νe and the plastic Poisson ratio νp:
Es  E 
ν= νe + 1 − s  νp ; νp = 0.5 (conservation of volume)
E  E
Apply the interaction formula at single load rates corrected for plasticity, for combined
loading.
V1-2 •1.4-FIELD OF APPLICATION
The formulas and charts given in this chapter apply to thin plates and shells. They may be used
to calculate most shell and stiffened panel items encountered in flight vehicle structures.

V1-2 •2-FLAT PLATES
z
y
a b
Strictly speaking, a plate is entirely bordered by tracks (for single hinging or fixed ends).
The following figures show the column to plate transition when single supports (hinges) are
added at unloaded edges (the loaded edges are hinged).
COLUMN
FLANGE
PLATE
Revision 1 Flat plates page V1-2•2.1

A - SUMMARY OF CASES DEALT WITH

RECTANGULAR FLAT PLATES
a - Single compression (page 8, curves pages 11 to 14)
FLANGES PLATES COLUMNS
CASE 1 CASE 5 CASE 1
CASE 3 CASE 7
CASE 4 CASE 8
free edge
CASE 9 hinged edge
fixed edge
CASE 10
b - Single shear (page 9, curves page 15)
CASE 1 CASE 2
c - Single bending (page 10, curves page 16)
CASE 1 CASE 2 CASE 3 CASE 4

Combined loading of rectangular flat planes studied on page 27, curves on pages 28 to 32.
FLAT PARALLELOGRAM PLATES
Single loads (page 17, curves 18 to 20)

FLAT TRIANGULAR PLATES
Compression (with or without shear) (page 21, curves pages 22, 23)
b
fixed edges hinged edges fixed edges hinged edges
case 1 case 2 case 3 case 4
Compression, Compression,
no shear with shear
VARIABLE WIDTH FLAT PLATES
Compression: (page 24, curves pages 24 to 226)
σ1 b1 σ2 b2
σ1 σ σ
3 cases are studied: = 0.8 ; 1 = 1.0 ; 1 = 1.2
σ2 σ2 σ2

B - FLAT PLATE ANALYSIS PROCEDURE

Obtain the following data:
- plate geometry: a, b, e
- plate loading
- plate material properties
Identify plate boundary conditions:

free, hinged, fixed edges, etc.
Calculate the buckling factors (paragraph 2.3)

Kc (as per x) if loading direction x
Kc (as per y) if loading direction y
Ks if shear
Kf if bending
Select the plastic correction factor
Calculate the critical buckling stresses

(paragraph 2.2 and 2.3) in:
compression as per x
compression as per y
shear
bending
No Yes
Is the plate subjected to combined loading?
No
Calculate Ri (paragraph 1.2) with transverse compression?
MS = (1/Ri) - 1 Yes
Calculate σcx (paragraph 2.4.2)
Calculate the loading rates (paragraph 2.2

and paragraph 2.4)
Rc if compression
Rs if shear
Rf if bending
use the interaction curves (paragraph 2.4)

V1-2 •2.1-SINGLE LOADING
Three single in-plane loads are possible:
τxy σx
b
a
Compression Shear Bending
The corresponding buckling stresses are formulated as follows:

2
 e
2
 e   e 
2
σcc,0 = ηKcE   ; τcr,0 = ηKsE   ; σcf,0 = ηKfE
 b  min {a, b}  b
REMARK: If a plate is loaded with a transverse compression stress (as per y), replace b by a
in the formula.
The buckling factors of rectangular, parallelogram, triangular plates and plates with an
evolutive width under single loads are given in paragraph 2.3.

V1-2 •2.2-PLASTICITY CORRECTION FACTOR
The plasticity correction is applied to the buckling stresses under single loads, even if the
studied plate is subjected to combined loads.
Loading Boundary conditions Equation

2
Flange with one unloaded 1 − ν  Es
η1 =  e

hinged edge 2
1 − ν  E
Flange with one unloaded  E 
η2 = η1  0.33 + 0.335 1 + 3 t 
fixed edge  Es 
Compression and
bending Plate with unloaded hinged  E 
η3 = η1  0.5 + 0.25 1 + 3 t 
edges  Es 
Plate with unloaded fixed  E 

η4 = η1  0.352 + 0.324 1 + 3 t 
edges  Es 
Et
Compression Column η5 =
E
G
Shear All conditions η6 = s
G
In the particular case of shear, the compression stress-strain curve of the material will also be
used. Proceed as follows:
• Calculate the equivalent normal stress: σeq = τ 3
• Calculate the corresponding Es and ν σeq
Gs  1 + νe  Es 1− ν
• =  = η1
G  1 + ν  E 1 − νe
Remark: η1 > η3 > η4 > η2 > η5

Choose a conservative solution if doubt subsists.

V1-2 •2.3-BUCKLING FACTORS
Except in the case of columns, the plate buckling factors in the linear elastic domain K are
formulated as follows:
∏2
K= k
12 (1 − ν 2e )
Classically, the curves giving k as a function of a/b are wavy. Each wave corresponds to a
buckling mode, the number of half waves of the buckle pattern m increases with a/b. Owing to
this, it is possible to observe relatively high variations of k for low variations of a/b.
m=1
m=2
m=3
smoothed curve
a/b
As one is never absolutely sure to choose the best boundary conditions, it is mandatory to use
smoothed curves to determine k and to select conservative boundary conditions.
The curves show a horizontal asymptote for a/b tending toward the infinite.

V1-2 •2.3.1-Rectangular flat plates
V1-2 •2.3.1.1-Single compression
Processed cases:
FLANGES PLATES COLUMNS
CASE 3 CASE 7
CASE 4 CASE 8
free edge
CASE 9 hinged edge
fixed edge
CASE 10
Smoothed curves, cases 1 to 10, except 7 and 8:
2 2
 0 ≤ a ≤ Lim →  k = C  a  + C  b  + C 
 c 3
 b  
1
 b 2
 a 
 a ≥ Lim → ( k = k )
b  c c∞
π2 k c
Kc =
12 (1 − ν 2e )
a
When tends towards 0, this tends toward the column plate case.
b
a
case 6: if < 0.4
b ⇒ Use the formula for case 12
a
case 10: if < 0.57
b
Smoothed curves, cases 7 and 8:
kc = k c (unloaded hinged edges) . k c (unloaded fixed edges)
π2 kc
Kc =
12(1 − ν 2e )
Smoothed curves, cases 11 and 12:

2
b
kc = C2  
 a
2
π2 kc π2 C2 e
Kc = ⇒ σcc.0 = η E  
12 12  a
C1 C2 C3 Lim kc∞
CASE 1 0 1 0.43 ∞ 0.43
CASE 2 0 4.143 0.384 9.49 0.43
CASE 3 0.136 0.987 0.551 1.64 1.28
CASE 4 0.007 3.923 0.94 4.34 1.28
CASE 5 1 1 2 1 4
CASE 6 0 2.7 4 ∞ 4
CASE 9 4.885 0.965 2.638 0.66 6.98
CASE 10 0 2.9 6.98 ∞ 6.98
CASE 11 0 1 0 ∞ 0
CASE 12 0 4 0 ∞ 0
V1-2 •2.3.1.2-Single shear
CASE 1 CASE 2
Smoothed curves:
2
 a b 
ks = C1  min  ;   + C 2
 b a  
Π2 ks
ks =
12 (1 − ν 2e )
C1 C2 = ks∞
4 hinged edges (case 1) 3.8 5.35
4 fixed edges (case 2) 5.53 8.98

V1-2 •2.3.1.3-Single bending
CASE 1 CASE 2 CASE 3 (*) CASE 4 (*)
free hinged fixed
Smoothed curves:
2 2
 0 ≤ a ≤ Lim →  k = C  a  + C  b  + C 
 b   f 1
 b 2
 a 3

 a ≥ Lim → k = k
b 
( f f∞ )
π2 k f
Kf =
12 (1 − ν 2e )
C1 C2 C3 Validity (a/b) Lim kf∞

CASE1 3.33 1.66 18.1 0.4 - 0.6 0.6 23.9
CASE2 52.3 2.2 20.4 0.3 - 0.45 0.45 41.8
CASE3 1.3 1.96
CASE4 1.3 0.85
(*) the free edges are compressed

no reliable curve in cases 3 and 4 for a/b < 1.3

CASE 1
CASE 3
FLANGES (hinged loaded edges)

CASE 2
CASE 4
FLANGES (fixed loaded edges)

CASE 5
CASE 7
CASE 9
PLATES (hinged loaded sides)

CASE 10
CASE 6
CASE 8
PLATES (fixed loaded sides)

4 hinged sides
4 fixed sides
Rectangular plates - single shear

4 hinged sides
4 fixed sides
Rectangular plates - single bending

V1-2 •2.3.2-Flat parallelogram plates
V1-2 •2.3.2.1-Single loads
Processed cases
a
b
α α α
4 fixed edges 4 hinged edges 4 fixed edges

case 1 case 2 case 3
Smoothed curves, cases 1 to 3:
2 2
 0 ≤ a ≤ Lim →  k = C  a  + C  b  + C 
 c ,s 3
 b  
1
 b 2
 a 
 a ≥ Lim → ( k = k )
b  c ,s c ,s∞
∏ 2 k c ,s
K c ,s =
12 (1 − ν 2e )
Case 1 C1 C2 C3 Validity (a/b) Lim Kc (infinite)

α=0 0 2.9 6.98 0.8 - ∞ ∞ 6.98
α=30 0 4.15 6.98 0.75 - ∞ ∞ 6.98
α=40 0 5.53 6.98 0.9 - ∞ ∞ 6.98
α=45 0 7.5 6.98 1.0 - ∞ ∞ 6.98
α=0 1 1 2 0.3 - 1.0 1 4
α=15 0.5 1.11 2.8 0.3 - 1.2 1 4.26
α=30 - 0.24 1.12 5.9 0.4 - 2.6 2.6 4.35
α=45 - 0.45 4 7.3 0.5 - 2.8 2.8 4.7
α=60 0 15.38 4.87 1.0 - ∞ ∞ 4.87
α=0 0 5.53 8.98 0.8 - ∞ ∞ 8.98
α=45 0 4.65 8.98 0.8 - ∞ ∞ 8.98
Note: k (α=30) = 1.43 * k (α=0) - 3.00

k (α=40= = 1.91 * k (α=0) - 6.33 ⇒ Equations valid only for case 1
k (α=45) = 2.59 * l (α=0) - 11.07
* All angles are in degrees

** For angles < 15°, use the curves corresponding to flat rectangular plates

Parallelogram plates with 4 fixed sides - single compression

Parallelogram plates with 4 hinged sides - single compression

Parallelogram plates with 4 fixed sides - single shear

V1-2 •2.3.3-Flat triangular plates
V1-2 •2.3.3.1-Compression (with or without shear)
Processed cases:
b
fixed edges hinged edges fixed edges hinged edges
case 1 case 2 case 3 case 4
Compression, Compression,
no shear with shear
Smoothed curves, cases 1 to 4:
2 2
 0 ≤ a ≤ Lim →  k = C  a  + C  b  + C 
 c ,s 3
 b  
1
 b 2
 a 
 a ≥ Lim → ( k = k )
b  c ,s c ,s∞
∏ 2 k c ,s
K c ,s =
12 (1 − ν 2e )
C1 C2 C3 Validity (a/b) Lim Kc (infinite)

CASE 1 0 2.8 9.2 0.5 - 2.0 2 9.9
CASE 2 0 2.09 2.7 0.4 - 2.0 2 3.2
CASE 3 0 7.73 28.2 0.5 - 1.03 1.03 35.4
CASE 4 0 5.11 8.2 0.5 - 1.3 1.3 11

hinged edges
fixed edges
Triangular plates - compression without shear

hinged edges
fixed edges
Triangular plates - compression with shear

V1-2 •2.3.4-Flat plates with evolutive width
V1-2 •2.3.4.1-Compression
σ1: compression side b1

σ2: compression side b2
b1 b2 0.91  e
σ1 σ2 σcr = η 2
Kc E  
1 − νe  b2 
CASE:
σ1
= 0.8
a σ2
fixed loaded sides, all

all sides are hinged other sides hinged
Kc Kc
hinged loaded all sides are

sides, other sides fixed
fixed
Kc Kc

σ1
= 1.0
σ2
fixed loaded
all sides are sides, other
hinged sides hinged
Kc Kc
all sides are

hinged loaded fixed
sides, other sides
fixed
Kc
Kc

σ1
= 1.2
σ2
fixed loaded
all sides are sides, other
hinged sides hinged
Kc Kc
hinged loaded sides,

other sides fixed all sides are
fixed
Kc Kc

V1-2 •2.4-COMBINED LOADING
Applicable to the flat rectangular plate case.

V1-2 •2.4.1-Combined loading without transverse loading
The interaction curve of a plate subjected to compression combined with shear and bending is
given in the following form (refer to paragraph 1.2):
σi
R αc + R βs + R γf = 1 where Ri =
σ iadm,0
Exponents α, β, γ are determined by theory or by testing:

compression + shear:
Interaction equation: RC + R S2 = 1
2
Rf = 0 ; α = 1 ; β = 2 ⇒ MS = -1 (Rc > 0)
R C + R 2C + 4 R S2
Tension + shear:
(tension is considered as a negative compression)
Interaction equation: Rs + 0.5 Rc = 1
1
MS = -1 (Rc < 0)
R s + 0.5 R c
Bending + shear:
Interaction equation: R s2 + R 2f = 1
1
Rc = 0 ; β = 2 ; γ = 2 ⇒ MS = -1
R s2 + R f2
Bending + compression:
Interaction equation: R 1f.75 + Rc = 1

MS may be determined graphically using the
Rs = 0 ; α = 1 ; γ = 1.75 ⇒ interaction curve, on page 30 or numerically (refer to
paragraph 1.2)
Bending + compression + shear:

1.75 R c + 2 R s
Interaction equation: Rc + R s2 + Rf =1
Rc + Rs
1.75 R c + 2 R s MS may be determined graphically using the
α=1;β=2;γ= ⇒ interaction curves, on pages 31 and 32 or numerically
Rc + Rs
(refer to paragraph 1.2)
Compression (tension) - shear interaction curve
Rc+Rs^2=1 (if Rc>0 compression) Rs-0.5Rc=1 (if Rc<0 tension)
Revision 1
Flat plates
page V1-2•2.28
Shear - bending interaction curve
Rs^2 + Rf^2 = 1

Compression - bending interaction curve
Rc + Rf^1.75 = 1

Compression - shear - bending interaction curves
as a function of the different values of Rc/Rs indicated directly on the curves
Revision 1
Flat plates
page V1-2•2.31
Compression - shear - bending interaction curves
as a function of the different values of Rs/Rc indicated directly on the curves
Revision 1
Flat plates
page V1-2•2.32
V1-2 •2.4.2-Combined loading with transverse loading
The critical stress in a flat rectangular plate is subjected to a combined loading with transverse
stress (tension or compression) is calculated as follows:
• Without longitudinal compression:

The transverse load is processed as a longitudinal load by replacing the in value b formulas
by a value.
In the presence of other loadings such as shear or bending, the interaction curves in the
previous paragraph are used.
• With longitudinal compression:
Plate stability not only depends on the various type of loads applied but also on the in-plane
dimensions. In fact, the ratio (a/b) defines the buckling mode.
The transverse load, whether tension or compression, affects the longitudinal critical
compression stress. The curves given on the following pages show the critical longitudinal
σcxe and transverse σcye stresses as a function of the a/b (length/width) ratio and the boundary
conditions at plate edges.
In presence of other loads such as shear or bending, the interaction curves of the previous
paragraph are used, putting down:
σx
Rc = where σcx = η x σcxe
σ cx
To use the curves on the following page, note that σ0 is a reference stress:
2
π2 E  e 
σ0 =
3 (1 − ν 2e )  b
without longitudinal stress with longitudinal stress

The 4 sides are hinged

hinged
fixed
hinged

fixed
hinged
fixed

The 4 sides are fixed

V1-2 •2.5-EXAMPLES
A - Use the example in paragraph V1-1, paragraph 2.3
y
pocket
x
BAY
b = 170 mm
e = 2 mm
a = 530 mm
land members
- Plate geometry: a = 530 mm ; b = 170 mm ; e = 2 mm
- Plate loading (at L.L.):

σx = 18 MPa
σy = 9 MPa
τxy = 27 MPa
- Plate material: 2024 PL T3

E = 70300 MPa
σ0.2 = 270 MPa
σR = 440 MPa
n = 7.05
νe = 0.33
- Plate boundary conditions:
The 4 sides are considered as being hinged.

Critical stress in longitudinal compression (paragraph 2.4.2, page 33)
- The reference stress is:

2
π2 E  e
σ0 =   = 36 MPa
3 (1 − ν 2e )  b 
- Then the curves on page 34 are used:
a / b = 530 / 170 = 3.1

 σ x σ cxe  σ σ cye 
 = = 2 ⇒  cxe = 2 
 σ y σ cye   σ0 σ0 
σ cxe a
- It is then possible to plot curve as a function of :
σ0 b
The 4 sides are hinged
 σ cxe σ cye 

 σ0
= 0.5 ;
σ0 
(
= 0,25 ⇒ σ cxe = 18 MPa ; σ cye = 9 MPa )
It is not necessary to calculate the plasticity correction.
σx 18
Rc = = =1
σ cx.0 18

Critical stress in shear (paragraph 2.3.1.2, page 9)
- Rectangular plate simply supported on four edges.

b
- = 0.32
a
2
b
Giving: ks = 3.8   + 5.35 = 5.74
 a
π2 ks
Ks = = 5.3
12 (1 − ν 2e )
2
e
τcr.0 = 5.3 E   ≈ 52 MPa
 b
It is not necessary to calculate the plasticity correction.
τ 27
Rs = = = 0.52
τ cr.0 52
Condition of critical stresses (paragraph 2.4.1, page 27)
According to the compression - shear interaction formula
 R + R 2c + 4 R s2 
R = σ = τ = c = 1.22
 σ cx τ cr 2 
 
The plate folds at 82% of L.L.
The folding stresses are:
σx = 14.8 MPa
σy = 7.4 MPa
τ = 22.1 MPa

B - It is necessary to determine the buckling stress in longitudinal compression of the

same plate, which as a constant thickness of 5 mm.
- This is the same as case 5 of paragraph 2.3.1.1, page 8.
 a = 31
. > Lim ⇒ (kc = kc∞ = 4)
b 
4 π2
Kc = = 3.69
12 (1 − 0.33 2 )
2
 e  σ 
σcce = 3.69 E   = 224.5 MPa  > 0 .2 
 b  2 
The table of plasticity correction factors on page 6 gives:
 E  1 − ν 2e E s
η3 = η1  0.5 + 0.25 1 + 3 t  ; η1 =
 Es  1 − ν2 E
With:
[Es, Et, ν] calculated using the Ramberg and Osgood model as indicated in paragraph 1.3.
This calculation is iterative:

With one stress value σ1, associate η3.
Then σ2 = η3 σcce
Convergence is obtained when: σ1 ≈ σ2
In our case:
(σ1 = 224.5 MPa) ⇒ (ε = 0.003738 ; Es = 60060 MPa ; Et = 31925 MPa ; ν = 0.355 ; η3 = 0.79)
σ2 = 0.79 * 224.5 = 176.5 MPa
σ1 + σ 2
To ensure convergence, repeat by putting down: σ1 =
2
σ1 = 200.5 MPa ⇒ η3 = 0.78 ; σ2 = 195.3 MPa
σ1 = 197.9 MPa ⇒ η3 = 0.878 ; σ2 = 197 MPa
Finally:
η3 = 0.879 ; σcc = 197.4 MPa
It is possible to see the need to calculate the plasticity correction well before σcce = σ0.2


V1-2 •3-CURVED PLATES
z
y
e
b
a
r
The curved plates discussed in this chapter are rectangular plates with a single curvature in the
loaded edge plane.
Revision 1 Curved plates page V1-2•3.1


RECTANGULAR CURVED PLATES
a - Single compression, single shear (page 7, curves pages 8 to 12)
hinged edge
Case 1 Case 2
fixed edge
Case 3 Case 4

B - CURVED PLATE ANALYSIS PROCEDURE
Obtain the following data:

- plate geometry:
- plate loading
- plate material properties
Identify plate boundary conditions:

Determine the buckling factors K

(paragraph 3.3)
Kc (in x direction) if stress in x direction
Ks if shear
Select a plastic correction factor (paragraph 3.2)
Calculate the elastic-plastic buckling critical

stresses (paragraphs 3.2 and 3.3) in:
compression in x direction
shear
No Yes
Is the plate subjected to combined loading?
Calculate Ri (paragraph 1.2) Calculate the loading rates

(paragraphs 1.2 and 3.4)
Rc if compression
Rs if shear
Use the interaction equation to

calculate R and MS (paragraph
3.4)

V1-2 •3.1-SINGLE LOADING
Two single stresses are possible:
τxy
x
b
a
y
Compression Shear
The corresponding buckling stresses are formulated as below:

2
e
2
 e 
σcc,0 = ηKcE   ; τcr,0 = ηKsE  
 b  min{a, b}
REMARKS: The buckling stress equation is the same as that for flat plates with higher
buckling factors Kc and Ks.
The buckling factors of rectangular curved plates are given in paragraph 3.3.

The plasticity correction is applied to buckling stresses under single load even if the studied
plated is subjected to combined loads.
Loading Boundary Flat plate Cylinder Curved plate

conditions
Flange, one  1 − ν 2e  E s Es Et
unloaded edge η1p =   ηc = η1 = (1 - c)η1p + cηc
hinged  1 − ν2  E E Es
Plate with
hinged  E  Es Et
Compression η3p = η1p  0.5 + 0.25 1 + 3 t  ηc = η3 = (1 - c)η3p + cηc
unloaded  Es  E Es
edges
Plate with  E  Es Et
fixed unloaded η4p = η1p  0.352 + 0.324 1 + 3 t  ηc = η4 = (1 - c)η4p + cηc
edges  Es  E Es
Shear All conditions G Gs Gs
η= s η= η=
G G G
1  b2 
Factor c depends on the geometry of the curved plate: c =  − 1
99  re 
The plasticity correction coefficients for a curved plate are calculated by linear interpolation
b2
between the coefficients of the flat plate given with a geometric limitation 1 > and those
re
b2
of the cylinder given for > 100.
re
Thus, the geometric boundarys of the plasticity correction coefficients given in the table above
are:
b2
≤ 1 : η = ηp (flat plate)
re
b2
≥ 100 : η = ηc (cylinder)
re
b2
1< < 100 : η = ηi (curved plate)
re
The stress-strain compression curve is also used in the particular case of shear loading.
Proceed as follows:
• Calculate Es and ν corresponding to σeq.
G  1 + νe  Es 1− ν
• s =  = η1
G  1 + ν  E 1 − νe
Remark: η1 > η3 > η4

Select a conservative solution if any doubt subsists.

V1-2 •3.3-BUCKLING FACTORS
The plate buckling factors K in the linear elastic domain are formulated as follows:
π2
K= k
12 (1 − ν 2e )
b2 1
The curves representing k as a function of Z =
re
( e ) (factor depending on the
1 − ν 2 2
geometry and on the material of the plate) are given in a logarithmic scale.
Note that:
- The k factors corresponding to small values of Z (Z < 1), therefore high values of r must
satisfy the flat plate theory.
- The k factors corresponding to high values of Z, therefore small values of r, must satisfy the
cylindrical shell theory.

V1-2 •3.3.1-Rectangular curved plates
Processed cases:
hinged edge
fixed edge
Case 1 Case 3
The curves for long curved panels representing the kc buckling factor as a function of Z are
given on the next page.
The linear theory for small deflections applicable in the case of single compression of flat
plates, provides theoretical results higher than the results from tests performed on curved
plates.
Use the recommended curves to determine k.
Processed cases:
hinged edge
fixed edge
Case 2 Case 4
For each case, the long panels and the wide panels are processed.
The curves representing the buckling factor ks in function of Z are given on pages 9 to 12.

Revision 1
Theoretical curve Recommended
curves, Reference
N.A.C.A. T.N.
1928
4 hinged edges
4 fixed edges
Curved plates
Buckling factor in axial compression for long curved
plates (*)
* 4 hinged sides:
4 fixed sides:
page V1-2•3.8
Buckling factor in shear for fixed long curved panels

Buckling factor in shear for fixed wide curved
cylinder
panels

Buckling factor in shear for long curved panels,

simply supported

Buckling factor in shear for wide curved panels,

simply supported

V1-2 •3.4-COMBINED LOADING
V1-2 •3.4.1-Shear + compression (or tension)

The interaction curve of a curved plate subjected to axial compression combined with shear is
given in the following form:
σ
Interaction equation: Rc + R s2 = 1 with Rc =
σ cc,0
2 τ
⇒ MS = -1 Rs =
2
Rc + Rc + 4 Rs 2 τ cr,0
where: Rc > 0 ⇒ compression and Rc < 0 ⇒ tension
V1-2 •3.4.2-Compression - side pressure

The existence of internal pressure p increases the axial buckling stress in compression of the
curved sheet panel σcc. The presence of an external pressure reduces it. Here, side pressure is
present and not hydrostatic pressure:
pr
• Side pressure: σx = 0 ; σy =
e
pr pr
• Hydrostatic pressure: σx = ; σy =
2e e
2
The interaction equation is: R c + Rp = 1
σ
in which Rc = c
σ cc
Rp designating the ratio between the pressure effectively applied and the external
pressure that would have to be applied to buckle a cylinder of the same radius
as the panel (refer to paragraph 4.3.1.5, on page 4/15).
This external pressure is calculated using the curve on page 14.
In the above interaction equation, the value Rp is taken as negative in the internal pressure
case.
V1-2 •3.4.3-Shear + side pressure

The existence of an internal pressure p increases the buckling stress in shear of a curved sheet
panel τcr. The presence of an external pressure reduces it. Here, side pressure is present and not
hydrostatic pressure:
pr
• Side pressure: σx = 0 ; σy =
e
pr pr
• Hydrostatic pressure: σx = ; σy =
2e e
2
The interaction equation is: R s + Rp = 1 (Rp < 0 in the case of internal pressure)
τ
in which Rs =
τ cr
Rp designating the ratio between the pressure effectively applied and the external
pressure which would have to be applied to buckle a cylinder of the same
curvature radius as the panel (refer to paragraph 4.3.1.5, page 4/15).
This external pressure will be calculated using the curve on page 14.
Buckling factor under external radial pressure ky

as a function of Z
radial pressure
with:
r
t= 1 − ν 2e
e
a2 1
Z=
re
(1 − ν e )
2 2
k y π2 E  e 
2
σcy =
12 (1 − ν e2 )  a

A - Use the example in V1-1, paragraph 2.3.
Assume that the plate is long and curved with a curvature radius of r.
y
pocket
x BAY
b = 170 mm
e = 2 mm
a=530 mm
land members
- Curved plate geometry: a = 530 mm

b = 170 mm
e = 2 mm
r = 2800 mm
- Plate loading (at L.L.):

σx = 18 MPa
- Plate material:
E = 70300 MPa
σ0.2 = 270 MPa
σR = 440 MPa
n = 7.05
νe = 0.33
- Plate boundary conditions:

The 4 sides are considered as being hinged.

Longitudinal compression:
- Calculation of the buckling factor (using figure page 8):
b2 1
Z=
re
(1 − ν 2e ) 2 = 4,87
Kc = 4.95
r/e = 2800 / 2 = 1400
Thus, the critical longitudinal stress is (refer to paragraph 3.3).

2
Kcπ2E  e  = η x 44.5 MPa
σcc,0 = η
12 (1 − ν 2e )  b
Obviously it is not necessary to calculate the plasticity correction (η = 1).

σ 18
σcc,0 = 44.5 MPa Rc = = = 0.4
σ cc, 0 44.5
 1 
The margin is: MS =   - 1 = 1.5 ⇒ folding at 250% of L.L.
 Rc 
B - Use the example above assuming that in addition to longitudinal compression, the
curved plate is subjected to a 0.01MPa (refer to paragraph 3.4.2) internal pressure.
The buckling stress in compression was calculated above:
σcc,0 = 44.5 MPa
To assess Rp, it is necessary to determine the external radial pressure which would cause
buckling of a circular cylinder with the same curvature as the curved sheet panel: (curves,
page 14).
a2 2 530 2
Z= 1 − νe = 1 −0.33 2 = 47.4
re 2800 x 2
Kyπ2E 2 2
 e  = 8.5 x π x 70300 x  2  = 7.85 MPa
2
ky = 8.5 ⇒ σcr =
12 (1 − ν 2e )  a  12 x (1 − 0.33 2 )  530 
The external radial pressure capable of producing this buckling is:

e 2 0.01
Pcr = σcr = 7.85 x = 0.0056 MPa ⇒ Rp = - = - 1.78
r 2800 0.0056
The interaction equation gives (refer to paragraph 1.2)
σ 2cc σ 2cc
+ Rp = 1 being - 1.78 = 1 Giving σcc = 74.2 MPa
σ 2cc.0 44.5 2
Thus, the effect of a 0.01 MPa internal pressure is to increase the axial buckling stress
from 44.5 MPa to 74.2 MPa.
C - Use the example in A above assuming that in addition to longitudinal compression,

the curved plate is subjected to a 27 Mpa shear.
Longitudinal compression: (already calculated above)
σ 18
σcc,0 = 44.5 MPa Rc = = = 0.4
σ cc, 0 44.5
Shear: (as per paragraphs 3.1 and 3.3.1.2)
2
Ksπ2E  e 
Linear elastic calculation: τcr,0 = η
12 (1 − ν 2 )  b
According to the curves on page 11:

b2 1
a 530
Z= (1 − ν 2e ) 2 = 4.87 and  = .  ⇒ (Ks = 6.2)
= 31
re  b 170 
Thus:
Ksπ2E  e  2
τcr,0 = η x   = η x 55.7 MPa
12 (1 − ν 2e )  b 
It is obviously not necessary to calculate the plasticity correction.
τ 27
τcr,0 = 55.7 MPa Rs = = = 0.48
τ cr, 0 55.7
Compression - shear interaction
 R + R 2c + 4 R s2 
R = σ = τ = c = 0.72 ⇒ Folding 139% of L.L.
 σ cx τ cr 2 
 

D - Use the same example A above but with a 5 mm pocket thickness.
Critical stress in compression
- Calculation of the buckling factor (using figure page 8):
b2 1
Z=
re
(1 − ν 2e ) 2 = 1.95
Kc = 4.19
r 2800
= = 560
e 5
Thus, the critical longitudinal stress is (refer to paragraph 3.3).
Kcπ2E  e 2
σcc,0 = η   = η x 235.2 MPa
12 (1 − ν 2e )  b 
- Calculation of the plasticity correction factor, iterative method:
Use the Ramberg and Osgood formulas (refer to paragraph 1.3) at the first step of the
calculation:
n 7. 05
σ  σ  235.2  235.2 
ε= + 0.002   = + 0.002   = 0.0041
E  σ 0 .2  70300  270 
σ 235.2 E  E 
Es = = = 57366 MPa and ν = s x 0.33 + 1 − s  x 0.5 = 0.361
ε 0.0041 E  E
1 n 1− n 7.05 1 − 7.05
= + = + = 0.00003684 ⇒ Et = 27148 MPa
E t Es E 57366 70300
According to paragraph 3.2, page 3:
 1 − ν 2e  E s  E t   Es Et
η3 = (1 - c) x   x x  0.5 + 0.2 1 + 3   + c x x
 1 − ν 2  E  E s   E Es
1  b2 
with c =  − 1 = 0.01075
99  re 
Et E Et 57366 27148
0.5 + 0.25 1 + 3 = 0.889 and s x = x = 0.561
Es E E s 70300 57366
 1 − 0.33 2  57366 
η = (1 - 0.01075) x  2
x x (0.889) + 0.01075 x 0.561 = 0.74
 1 − 0.361  70300 
The critical stress σcc,0 is then: 0.74 * 235.2 = 174 MPa
To ensure convergence, it is necessary to repeat the operation using as an initial value
for each iteration, the average between the value found and the initial value of the
previous step
σ = (235.2 + 174) / 2 = 204.6 MPa
σ = 204.6 MPa ⇒ ε = 0.00319 ⇒ Es = 64070 MPa ⇒ Et = 41707 MPa ⇒ η = 0.856 ⇒ σcc.0 = 201.3 MPa
σ = 203 MPa ⇒ ε = 0.00316 ⇒ Es = 64334 MPa ⇒ Et = 42509 MPa ⇒ η = 0.861 ⇒ σcc.0 = 202.5 MPa
Thus, after several iterations, convergence is towards
σ 18
η = 0.862 and σcc,0 = 202.7 MPa ⇒ Rc = = = 0.089
σ cc, 0 202.7

- Calculation of the plasticity correction factor, schematic method:
Plot the curve giving σcr corrected according to elastic σcr:
σ 0.2 270
Assume as an initial value: σcr corrected = = = 135 MPa
2 2
n 7. 05
σ  σ  135 135 
ε= + 0.002   = + 0.002   = 0.00194
E  σ 0 .2  70300  270 
σ 135
Es = = = 69752 MPa
ε 0.00194
−1 −1
 n 1 − n 7.05 1 − 7.05 
Et =  +  =  +  = 66609 MPa
 Es E   69752 70300 
Es  E  69752 69752 
ν= νe +  1 − s  νp = x 0.33 + 1 −  x 0.5 = 0.331
E  E  70300  70300 
 1 − ν 2e  E s  E  E Et
η = (1 - c)  2  0.5 + 025 1 + 3 s  + c t = 0.985
1 − ν  E  E Es Es
σ 135
σ elastic = = = 137.1 MPa
η 0.985
Thus, for σ corrected = 135 Mpa ⇒ σ elastic = 137.1 MPa
By varying the corrected σcr, it is possible to calculate the corresponding elastic stress
making it possible to plot the curve σcr corrected according to elastic σcr (refer to the
next page).
Then the required result is obtained directly from this curve:
σ elastic = 235.2 MPa ⇒ σ corrected = 202.5 MPa
σ 18
⇒ Rc = = = 0.089
σ cc, 0 202.5

Plasticity correction
Elastic stress
Corrected stress

V1-2 •4-CYLINDERS AND OTHER SHELLS
e
r
The thin shells discussed in this chapter are circular and elliptical cylinders, spherical caps and
truncated cones.
Revision 1 Cylinders and shells page V1-2•4.1


CIRCULAR CYLINDERS (single load)
compression, bending, shear, torsion, radial pressure, hydrostatic pressure
Compression Bending Shear
Torsion Radial pressure Hydrostatic pressure
ELLIPTICAL CYLINDERS (single load)

compression, bending, shear, torsion
Compression Bending Shear Torsion
SPHERICAL SHELLS (single load)

external pressure
External pressure
CONICAL SHELLS (single load)

compression, bending, torsion, external pressure
Compression Bending Torsion External pressure

B - CIRCULAR CYLINDER ANALYSIS PROCEDURE

Determine the following data:
- cylinder geometry: L, e, r
- cylinder loading,
- properties of the cylinder material
Identify the cylinder boundary conditions:

Determine the buckling factors K (paragraph 4.3.1)

Kc if compression
Kf if bending
Kt if torsion or shear load
Ky if external radial pressure
Calculate the buckling elastic-plastic critical stresses

(paragraphs 4.2 and 4.31):
compression, bending, torsion, shear, radial pressure
No Yes
Is the cylinder subjected to combined loading?
Calculate Ri with internal pressure?
No
Yes
Calculate the loading rate Calculate the loading rate
(paragraph 4.4.1.A) (paragraph 4.4.1.B)
Rc if compression Rc if compression
Rcs if shear Rcs if shear
Rct if torsion Rct if torsion
Rf if bending Rf if bending
Use the interaction equation

Rc + Rct^2 + (Rcs^3 + Rf^3)^0.33 = 1

C - ELLIPTICAL CYLINDER ANALYSIS PROCEDURE

- geometry of the cylinder: L, e, a, b
- cylinder loading
- properties of the cylinder material
Identify the cylinder boundary conditions:


Kc if compression
Kf if bending
Kt if torsion or shear

(paragraphs 4.2 and 4.3.2) in:
compression, bending, torsion, shear
Yes No
Is the cylinder subject to combined loading?
Calculate Ri with internal pressure?
No
Yes
Calculate the loading Calculate the loading
rates (paragraph 4.4.1.A) rates (paragraph 4.4.1.B)
Rc if compression Rc if compression
Rcs if shear Rcs if shear
Rct if torsion Rct if torsion
Rf if bending Rf if bending
use the interaction equation

Rc + Rct^2 + (Rcs^3 + Rf^3) ^ 0.33 = 1

D - SPHERICAL SHELL ANALYSIS PROCEDURE

- shell geometry: d, e, r
- shell loading: external pressure
- properties of the shell material
Determine the buckling factor Kp (paragraph 4.3.3)
Calculate the buckling elastic-plastic critical stress

(paragraphs 4.2 and 4.33) under: external pressure
E - CONICAL SHELL ANALYSIS PROCEDURE

- shell geometry: L, e, r, R, α, ρ
- shell loading
- properties of the shell

Kc if compression
Kf if bending
Kt if torsion
Ky if external pressure

(paragraphs 4.2 and 4.3.4) in:
compression, bending, torsion, external pressure
No Yes
Is the cylinder subject to combined loading?
Calculate Ri Calculate the loading rates (paragraph 4.4.2)

Rc if compression
Rf if bending
Rct if torsion
Ry if external pressure
use the interaction equation

(Rc + Rf + Rct) ^ 1.2 + Ry ^ 1.2 = 1

V1-2 •4.1-BEHAVIOUR OF CIRCULAR CYLINDERS UNDER COMPRES-

SION
V1-2 •4.1.1-Introduction
Circular cylinders may have 4 different types of behaviour under the effect of compression:
• Short cylinders
The deflected shape after buckling is similar to that of column

plates (cylinders with an infinite radius). The buckling waves
are similar to those obtained in flat plates (half sine waves).
The supporting conditions at ends have a considerable

influence on the critical stress.
• Long cylinders
The deflected shape after buckling is of a diamond-like

appearance. The supporting conditions at ends have no affect
on the critical stress.
• Transition length cylinders
The transition length is in between the two extreme lengths

defining the short cylinders and the long cylinders.
The buckle pattern is an interaction of two shapes:

This is a combination of the half sine waves and diamond
shape.
• Very long cylinder
A cylinder is very long when its column buckling stress is less

than its local buckling stress.

V1-2 •4.1.2-Linear theory and test results
Here we shall deal with the first 3 types of behaviour mentioned above.
The energy causing the occurrence of an inward blister is much lower than that required to
obtain an outward blister.
With compressed cylinders, unlike what happens in flat plates, the occurrence of an inward
blister entails the formation of a transverse compression membrane stress. This stress make
the blister even more unstable.
The linear stability theory (low deflection) gives satisfactory results when positive or nil
membrane transverse stresses (tension) accompany folding. This is the case for flat plates,
cylinders subjected to torsion or radial pressure, curved plates in shear or under side pressure.
On the other hand, for curved plates and cylinders in compression or bending and for spheres
or spherical caps subjected to an external radial pressure, the surface stresses can no longer be
considered as linear functions of transverse displacements.
The formulas and charts given in this document are derived from the linear theory readjusted
following a great number of tests.
A good correlation of tests and stresses calculated in this manner is obtained for short or long
cylinders in compression or bending. On the other hand, the results are highly conservative in
the transition range. In this latter case, the test results are located between the theoretical
linear calculations and the semi-empirical calculations.

Even if the studied shell or cylinder is subjected to combined loads, the plasticity correction is
applied to the buckling stresses under single loading.
CIRCULAR AND ELLIPTICAL CYLINDERS
Cylinder loading Plasticity correction factor

Es Et
Compression, Bending η=
E Es
Gs
Shear, Torsion η=
G
E s 1 − ν e2  E 
Radial or hydrostatic pressure η= 2 
0.25 + 0.75 t 
E 1− ν  Es 
SPHERICAL CONES AND TAPERED CONES
Cone loading Plasticity correction factor

Es Et
External radial pressure η=
E Es
In the particular case of shear, the compression stress-strain curve of the material is also used.
Proceed as follows:

• Calculate Es and ν corresponding to σeq,
G  1 − νe  Es
• s = 
G  1 − ν2  E

V1-2 •4.3-BUCKLING FACTORS (SINGLE LOADING)
V1-2 •4.3.1-Circular cylinders
Six different loadings cause the occurrence of stress in the surface of a cylinder:
Compression Bending Shear
Torsion Radial pressure Hydrostatic pressure

The corresponding buckling stresses are given in the paragraphs below in function of the
geometry of the cylinder and the boundary conditions.
The critical buckling stress under single compression is formulated as below:
2
 e π2
σcc,0 = η Kc E   with Kc = kc
 L 12 (1 − ν 2e )
Calculation of kc
L2
The buckling factor depends on the parameter: Z = 1 − ν 2e
re
* short cylinders and transition range
kc0: buckling factor of column plates
2
(0.581 x C x Z )
kc = kc0 + fixed edges: kc0 = 4
k c0
hinged edges: kc0 = 1
* long cylinders
e
kc = 1.16 x C x Z ⇒ σcc,0 = η C E  
 r
◊ C factor
The C factor used in the formulas below is:
−0 .15
  r  r
C = 0.6 x ln  3.748 x    ; C ≤ 0.6 valid for: 8.52 < < 3000
  e   e
r
if < 8.52 take C = 0.6
e
◊ Border between the transition range and long cylinders:
k c0
Z* = and kc* = 2 kc0
0.58 C
Buckling factor in compression kc as a function of Z

loaded sides fixed loaded sides simply supported

The critical buckling stress under single bending is formulated as below:

2
e
σcf,0 = η Kf E  
 L
The buckling factor Kf is formulated as below:
π2
Kf = kf
12 (1 − ν 2e )
Calculation of kf
L2
re
* Short cylinders and transition range
(0.581 x C f x Z ) 2
kf = kc0 +
k c0
Kc0 is the buckling factor of column plates under compression:

fixed edges: kc0 = 4
hinged edges: kc0 = 1
* long cylinders
kf = 1.16 x Cf x Z
e
The critical buckling stress becomes: σcf,0 = η Cf E  
 r
◊ Cf factor:
The Cf factor above is: Cf = 1.3 * C (factor used in compression)

−0 .15
 r 
C = 0.6 x ln  3.748 x    ; C ≤ 0.6
  e 
r
valid for: 8.52 < < 3000
e
r
if < 8.52 take C = 0.6
e
◊ Border between the transition range and long cylinders:

k c0
Z* = and kf* = 2 kc0
0.58 C f

Buckling factor in bending, kf, as a function of Z

loaded sides fixed loaded sides simply supported

V1-2 •4.3.1.3-Single torsion
Unlike what happens in compression, cylinders subjected to torsion do not spontaneously

collapse after the occurrence of folding.
If a cylinder is gradually loaded, first of all slight deflections occur increasing in practically a
linear manner. The curvature gradually flattens around the occurrence of folding then major
deflections occur for a slight variation in the load.
The torsion moment drops when the cylinder collapses, whereas the side deflections continue
to increase.
Collapsing load
The folding stress is approximately 84% of the collapsing stress.
The critical collapsing stress under torsion is formulated as follows:
2
 e π2
τct,0 = η Kt E   with Kt = kt
 L 12 (1 − ν 2e )
Calculation of Kt
L2
The buckling factor depends on parameter: Z = 1 − ν 2e
re
* Short cylinders
For Z ---> 0: the cylinder behaves like an infinitely long flat plate.
Short cylinders have as a Z factor: 0 < Z < 100 1 − ν 2e
- fixed edges: kt = 8.98 + 0.204 Z
- hinged edges: kt = 5.35 + 0.216 Z
* Intermediate cylinders
3
r2
- fixed edges: 100 1 − ν 2e < Z < 63 1 − ν 2e and kt = 0.93 Z 4
e2
3
r2
- hinged edges: 100 1 − ν 2e < Z < 44 2 1 − ν 2e and kt = 0.85 Z 4
e
* long cylinders
r2
- fixed edges: Z > 63 1 − ν e2 0.33 Z
e2 kt = 1
 r 1 − ν 2  2
r2 e e

- hinged edges: Z > 44 1 − ν 2e
e2
τcr is then formulated in the simpler form:
3
−3
e 2
τcr = η x 0.272 (1 − ν 2
) 4 E  
e
 r

Buckling factor under torsion, kt, as a function of Z
fixed
hinged
Collapsing load
The critical collapsing stress under a shear load is formulated as follows:

2
e π2
τcs,0 = 1.25 η Kt E   with Kt = kt
 L 12 (1 − ν 2e )
Calculation of kt
kt is the buckling factor calculated in the torsion case studied in the previous paragraph.

V1-2 •4.3.1.5-External radial pressure
Collapsing load
The critical collapsing stress under external radial pressure is formulated as follows:
2
e π2
σcy,0 = η Ky E   with Ky = ky
 L 12 (1 − ν 2e )
Remember that with a radial pressure p, the longitudinal stress is nil whereas the transverse
stress is equal to pr/e.
Calculation of Ky (chart, page 17)
L2
re
* Very short cylinders
The case of very short cylinders is identical to that of infinitely long flat
plates in axial compression.
Values for Z=0:
- hinged edges: ky = 4
- fixed edges: ky = 6.98
* Short cylinders and long cylinders
The difference between these cylinders is the number of waves obtained in the circumferential
direction:
short cylinders: the number of waves is greater than 2.
long cylinders: only 2 waves.
Let n0, be the number of circumferential waves obtained in the specific case
r
where Z = t, with: t = 1 − ν 2e ; Z = t ⇔ L = r
e
The ky factor calculation procedure (then σcr) is as follows: (refer to the Appendix).
n 20 = 9.3 Z0.5
t 2
n2 = n0
Z
n2 < 4
take n² = 4
4Z n 20
β2 = 2 β2 =
π t π2
ky =
(1 + β )2 2
+
12 Z 2
π 4 β 2 (1 + β 2 )
2
β2
π2 k y E  e  2
σcr =  
12 (1 − ν 2e )  L 

V1-2 •4.3.1.6-Hydrostatic pressure
Collapsing load
The critical circumferential collapsing stress under hydrostatic pressure is formulated as

follows:
2 2
e π
σcy,0 = η Kp E   with Kp = kp
 L 12 (1 − ν 2e )
Remember that with hydrostatic pressure p, the axial stress is equal to pr/2e, whereas the
transverse stress is equal to pr/e.
Calculation of kp:
L2
re
* Very short cylinders
The case of very short cylinders is identical to that of infinitely long

flat plates in axial compression.
Values for Z=0:

- hinged edges: kp = 2
* Short cylinders and long cylinders
The behaviour of long cylinders under hydrostatic pressure is similar to that of long cylinders
under radial pressure. This gives a kp curve that overlays the one obtained under radial
pressure for very high Z values which converge for small Z values towards kp=2 instead of 4.
The kp curve will be conservatively plotted in the form of two straight lines (logarithmic
scale):
kp = 2 for 0 < Z < 10
kp = 0.5514 Z0.56 for Z > 10

Buckling factor under external radial pressure, ky, as a function of Z
radial pressure
hydrostatic pressure

V1-2 •4.3.2-Elliptical cylinders

Elliptical cylinders behave in the same manner as circular cylinders under the effect of
compression.
Therefore, the same formulas and the same charts as those used for the study of circular
cylinders will be used but this time with a curvature radius equal to a²/b instead of r (circular
cylinder radius).

Calculation method
I
1- Calculate the strength modulus of the circumscribed circle: Sa = = π a² e
ν
elliptical cross section σa
e
Compression
a
r Ycr σcr
Tension
b circumscribed circle
2- Calculate the strength modulus of the elliptical cylinder cross section Se = (Se/Sa)*Sa
S
where e is the cross section efficiency.
Sa
2 3
Se  b  + 0.521 *  b  - 0.163 *  b 
Equation of the curve: = 0.424 + 0.218 *
Sa  a  a  a

y r
3- Calculate the following ratios:   and ( r : critical curvature radius)
 a  cr a
−0 . 5
 y  = 0.5 *   b 2   r  a  with 0 ≤ b ≤ 0.866
 a  cr 1 −    and   = 0.649 *
 b
 a   a a
b
if > 0.866 buckling occurs at the end of the long axis ⇒ r = a
a
4- Calculate the stress applied at the critical curvature:

M y
σycr = f *   ; Mf is the bending moment.
S e  a  cr
5- Calculate the allowable stress (long elliptical cylinders):

−0 .15
C f * E *e   r 
σcr = η with Cf = 1.3 * 0.6 * In 3.748 *   
r   e 
6- Calculate the margin between the allowable stress and the applied stress.
Collapsing load
The torsion moment applied around the elliptical section of the cylinder first of all causes folds
at the small axis, moving gradually towards the long axis where collapsing takes place.
As the curvature radius at the long axis is the smallest, this area stabilises the cylinder until the
folds cause collapsing.
b a
folding propagation direction
The critical collapsing stress is calculated as for axi-symmetrical cylinders, putting down r = a
(refer to paragraph 4.3.1.3).
Collapsing load
The critical collapsing stress under a shear load is formulated as follows:
2
e π2
τcs,0 = 1.25 η Kt E   with Kt = kt
 L 12 (1 − ν 2e )
Calculation of kt:
kt is the buckling factor calculated for the torsion case.

V1-2 •4.3.3-Spherical shells
V1-2 •4.3.3.1-External pressure
The critical buckling stress under external pressure is formulated as below:

2
 e
σcp,0 = η Κp Ε  
 d
The buckling factors Kp are formulated as follows:
∏2
Kp = kp
12 (1 − ν 2e )
Calculation of kp
d2
re
* Spherical shells with small Z
(0.581 x C x Z ) 2
kp = kpl +
k pl
kpl is the buckling factor of flat circular plates.
kpl = 3
e
* Spherical shells with long Z r

kp = 1.16 x C x Z d
e
The critical buckling stress becomes: σcp,0 = η C E  
 r
◊ C factor:
The C factor used in the formulas above is:
C = eA with A = - 0.199 * In (r/e) - 0.235

r
valid for: 10 < < 1000
e
r
if > 1000 take C = 0.2
e
◊ Border between two curves
k pl
Z* = and kp* = 2 kpl with kpl = 3
0.58 C

kp as a function of Z

V1-2 •4.3.4-Conical shells
The critical buckling stress under compression is formulated as: Pc,cr

Pc,cr = 2 π r e σc,cr r
Pc,cr: critical compression load.
σc,cr: critical compression stress (calculation, page 24). ρ α
L
ρ: minimum curvature radius.
α: cone half angle.
r: cone small radius.
e: cone thickness. Pc,cr
Mf,cr
V1-2 •4.3.4.2-Single bending r
The critical buckling stress under bending is formulated: ρ α

L
Mf,cr = π r² e σf,cr
Mf,cr: critical bending moment.
σf,,cr: critical bending stress (calculation, page 25).
Mf,cr
Mt,cr
The critical buckling stress under bending is formulated: r
Mt,cr = 2 π r² e τt,cr
Mt,cr: critical torsion moment. L ρ α
τt,,cr: critical torsion stress (calculation, page 26).
V1-2 •4.3.4.4-Single external pressure Mt,cr
The critical external pressure is given by the following formula:
π2 e3 α
pcr = k y r
12 (1 − ν 2e ) ρ moy L2
L Pcr
pcr: critical pressure (calculation, page 27).
ρmoy
ky: buckling factor under external pressure.
R
r+R
ρmoy: average curvature radius =
2 cos α

V1-2 •4.3.4.5-Single loading curves
a) in compression
critical compression stress
Curve equations:
curves valid in the elastic domain with: 5° < α < 11°

−1. 575
L
= 4 : σc,cr * (
1000  ρ
) = 4220 *  
ρ E  e
−1. 575
L 1000 ρ
= 2 : σc,cr * ( ) = 5040 *  
ρ E  e
−1. 575
L
= 1 : σc,cr * (
1000  ρ
) = 5851 *  
ρ E  e
−1. 575
L
= 0.5 : σc,cr * (
1000  ρ
) = 6569 *  
ρ E  e

b) in bending
critical bending stress
Curve equations:

−1. 565
L
= 4 : σf,cr * (
1000  ρ
) = 5361 *  
ρ E  e
−1. 565
L
= 2 : σf,cr * (
1000  ρ
) = 6148 *  
ρ E  e
−1. 565
L
= 1 : σf,cr * (
1000  ρ
) = 6977 *  
ρ E  e
−1. 565
L
= 0.5 : σf,cr * (
1000  ρ
) = 7886 *  
ρ E  e

c) in torsion
critical torsion stress
Curve equations:

−1. 407
L
= 8 : τt,cr * (
1000  ρ
) = 545 *  
ρ E  e
−1.362
L
= 4 : τt,cr * (
1000  ρ
) = 598 *  
ρ E  e
−1.328
L
= 2 : τt,cr * (
1000  ρ
) = 695 *  
ρ E  e
−1. 285
L
= 1 : τt,cr * (
1000  ρ
) = 764 *  
ρ E  e

d) under external pressure
CRITICAL EXTERNAL PRESSURE
L2
Z= 1 − ν 2e
ρ moy e
Curve equation:
0 < Z < 60 : ky = 4 + 0.12241 * Z - 0.0006 * Z²

Z > 60 : ky = 1.456 * Z0.45
π2 E e3
Critical external pressure: Pcr = η k y
12 (1 − ν 2e ) ρ moy L2

V1-2 •4.4-INTERACTION EQUATIONS (COMBINED LOADING)
V1-2 •4.4.1-Circular and elliptical cylinders
A - No external pressure
a - Compression + bending + torsion + shear load
The interaction equation of a circular cylinder subjected to axial compression combined

with bending, torsion and shear due to a shear load is:
⇒ Interaction equation: Rc + R 2ct + 3 R 3cs + R 3f = 1

where:
applied compression load (or applied compression stress)

Rc =
critical compression load (or critical compression stress)
shear stress due to torsion
Rct =
critical torsion stress
shear stress due to the shear load
Rcs =
critical shear stress
bending moment applied (or bending stress)
Rf =
critical bending moment (or critical bending stress)
The intermediate interaction equations are derived from the general information above.
Thus:
1
b - Longitudinal compression + pure bending: Rc + Rf = 1 ; M.S = -1
Rc + Rf
2
c - Longitudinal compression + torsion: Rc + R 2ct = 1 ; M.S = -1
R c + R 2c + 4 R 2ct
1
d - Pure bending + shear load: R 3f + R 3cs = 1 ; M.S = -1
3
R + R 3cs
3
f
e - Longitudinal compression + pure bending + torsion:

2
Rc + Rf + R 2ct = 1 ; M.S = -1
R c + R f + ( R c + R f ) 2 + 4 R 2ct
3
f - Longitudinal compression + pure bending + shear load: Rc + R 3cs + R 3f = 1
refer to the charts on the following page.

Interaction curves under axial compression,

bending, shear load

B - With internal pressure
The internal pressure appreciably improves the cylinder's strength whatever the loading mode.
Thus, the interaction formulas above remain valid provided that the new critical stresses,
improved by the presence of an internal pressure, are taken into account.
a) Compression + internal pressure:
The increase in critical stress in compression due to internal pressure is given below:
Increase in critical stress in compression due to internal pressure
100
ηE e
∆σcc,0 r
pr
An isostatic internal pressure p implies a longitudinal tension at the cylinder σt: σt =
2e
A radial pressure implies nil longitudinal tension: σt = 0

The critical compression load is then:
Pc,cr = 2 π r e (σcc,0 + ∆σcc,0 + σt)

σcc,0, ∆σcc,0, σt are all positive values.
Pc,cr
The safety margin is then: M.S = -1
Papp
b) Bending + internal pressure:
The increase in critical stress in bending due to internal pressure is given below:
increase in critical stress in bending due to

the internal pressure
e
r
∆ct,0
ηE
pr
An isostatic internal pressure p implies a longitudinal tension at the cylinder σt: σt =
2e
A radial pressure implies a nil longitudinal tension: σt = 0

The critical bending moment is then:
Mf,cr = π r² e (σcf,0 + ∆σcf,0 + σt)
σcf,0, ∆σcf,0, σt are all positive values.

M f ,cr
The safety margin is then: M.S = -1
M app

c) Torsion + internal pressure:
The critical stress in torsion due to the internal pressure is given below:
effect of the internal pressure on the critical torsion stress
The interaction equation is:
R 2ct + Rp = 1 (Rp < 0, if p is an internal pressure)
applied torsion stress applied pressure

Rct = ; Rp =
critical torsion stress critical pressure
d) Shear load + internal pressure
R 2cs + Rp = 1 (Rp < 0, if p is an internal pressure)
applied shear stress applied pressure

Rcs = ; Rp =
critical shear stress critical pressure

V1-2 •4.4.2-Conical shells
The interaction equation relating the compression, the bending and the shear load and the
external pressure is:
(Rc + Rf + Rcs)1.2 + R 1p.2 = 1

with:
compression stress
Rc =
critical compression stress
bending stress
Rf =
critical bending stress
shear stress due to a shear load
Rcs =
1.4 * critical torsion stress
applied pressure
Rp = (internal pressure: Rp < 0)
critical pressure
A- Take a circular cylinder made of 2024 PL T3 aluminium alloy with the following
geometrical and mechanical properties:
- Cylinder geometry:
L = 1950 mm
r = 1300 mm
e = 1,3 mm
- Cylinder material: 2024 PL T3

E = 70300 MPa
σ0.2 = 270 MPa
σR = 440 MPa
n = 7.05
νe = 0.33
- Cylinder boundary conditions:

The ends are hinged
- Cylinder boundary conditions:

Compression load = 135 000 N
Bending moment = 156 000 m.N
Torsion moment = 265 000 m.N
Internal pressure = 350 hPa

Longitudinal compression with internal pressure (according to paragraphs 4.3.1.1 and

4.4.1.B):
- Calculation of the buckling factor:
r 1300 L 1950
= = 1000 ; = = 1.5
e 1.3 r 1300
L2 1950 2
Z= 1 − ν 2e = * 1 − 0.33 2 = 2124
re 1300 * 1.3
Consultation of curves, page 10 (or formulas, page 9) ⇒ C = 0.171 and kc = 421
k c0 1
Z* = = = 10 << 2124 ⇒ long cylinder
0.58 * C 0.58 * 0.171
e  1.3  = η * 12 MPa
σcc,0 = η C E   = η * 0.171 * 70300 *
 r  1300 
Shows that it is obviously not necessary to calculate the plasticity correction (η = 1)
The critical axial load is (without taking internal pressure into account):
Pcr = 2 π r e σcc,0 = 127 423 N
The internal pressure improves the critical compression load:
according to page 30:
2 2
 p   r  =  0.035   1300  = 0.5
 
 E   e  70300   1.3 
according to the figure on page 30:
∆σ cc,0  r  70300 * 1.3
  = 0.175 ⇒ ∆σcc,0 = η * 0.175 * = η = 12.3 MPa
η  E e 1300
The global critical compression stress:
σ
σcc,0 + ∆σcc,0 = 24.3 MPa << 0.2 giving η = 1
2
On the other hand, the internal pressure induces a longitudinal tension load (page 30).
pr 0.035 * 1300
σt = = = 17.5 MPa ⇒ Ft = 2 π r e σt = 185 825 N
2e 2 * 1.3
An external compression load should have already cancelled out this tension load before
subjecting the cylinder to a longitudinal compression.
The critical compression load is then:
Pc,cr = 2 π r e (σcc,0 + ∆σcc,0 + σt) = 2 π * 1300 * 1.3 * (12 + 12.3 + 17.5) = 443 857 N
The margin is given by the following formula:
Pc,cr 443 857

M.S = −1= - 1 = 2.29
p 135 000

Bending moment with internal pressure:
r L
= 1000 , = 1.5 , Z = 2124
e r
Consultation of the curve, page 12 (or formulas, page 11) ⇒ Kf = 553.5
k c0 1
Cf = 0.222 ⇒ Z* = = = 7.8 << 2124 ⇒ long cylinder
0.58 * C f 0.58 * 0.222
e  1.3  = 15.6 MPa (η = 1)
σcf,0 = η Cf E   = η * 0.222 * 70300 *
 r  1300 
Thus, the critical bending moment in the absence of internal pressure is:
I
Mf,cr = σcf,0 σcf,0 * π * r² * e = 15.6 * π * 1300² * 1.3 = 107 672 m.N
r
The internal pressure improves the critical bending moment:
according to page 31:

2 2
 p   r  =  0.035   1300  = 0.5
 
 E   e  70300   1.3 
according to the figure on page 31:
∆σ cf ,0  r  70300 * 1.3
  = 0.315 ⇒ ∆σcf,0 = η * 0.315 * = η = 22.1 MPa
E  e 1300
The global critical bending stress:

σ 0.2
σcf,0 + ∆σcf,0 = 15.6 + 22.1 = 37.7 MPa << giving η = 1
2
On the other hand, the internal pressure induces a longitudinal tension load (page 31).
pr 0.035 * 1300
σt = = = 17.5 MPa
2e 2 * 1.3
The critical bending moment is then:

Mf,cr = π r² e (σcf,0 + ∆σcf,0 + σt) = π * 1300² * 1.3 * (15.6 + 22.1 + 17.5) = 381 000 m.N
M f ,cr 381 000

M= −1= - 1 = 1.44
Mf 156 000

Torsion moment with internal pressure:
r
= 1000 , Z = 2124
e
Intermediate cylinder with hinged edges (refer to page 13).

3
kt = 0.85 * Z 4 = 0.85 * (2124)0.75 = 266 (or according to the figure on page 14)
π2
Kt = kt = 245.5
12 (1 − 0.33 2 )
2 2
e 1.3 
τct,0 = η Kt E   = η * 245.5 * 73800 *   = 8.1 MPa (η = 1)
 L  1950 
The critical torsion moment is:
J
Mt,cr = τct,0 = τct,0 * 2π r² e = 8.1 * 2π * 1300² * 1.3 = 111 814 m.N
r
Internal pressure improves the critical torsion moment.

The critical hydrostatic pressure pcr, causing the collapse of the cylinder is according to
the figure on page 17 (or formulas, page 16):
π2
Z = 2124 ⇒ kp = 40.2 ⇒ Kp = kp * = 37.1
12 (1 − 0.33 2 )
2 2
e 1.3 
σcp,0 = η * Kp * E *   = η * 37.1 * 73800 *   = 1.2 MPa (η = 1)
 L  1950 
The critical pressure causing this circumferential stress in the cylinder is:
e 1.3
Pcr = σcp,0 = 1.2 * = 12 hPa
r 1300
The critical torsion stress with internal pressure is, according to the figure on page 32:
Pint - Pcr = 350 - 12 = 338 hPa

τct,0 = 38 MPa
The critical torsion moment is then:
Mt,cr = τct,0 * 2π r² e = 38 * 2π * 1300² * 1.3 = 524 558 m.N
M t ,cr 524 558

M= −1= - 1 = 0.98
Mt 265 000

Compression - bending - torsion interaction
The applied loads are:
- P = 135 * 103 N (compression load)

- Mf = 156 * 103 m.N (bending moment)
- Mt = 265 * 103 m.N (torsion moment)
- Pint = 350 hPa (internal pressure)
The critical loads calculated taking internal pressure into account:
Pc,cr = 443 857 N

Mf,cr = 381 000 m.N
Mt,cr = 524 558 m.N
Giving:
P 135 000
Rc = = = 0.3
Pc,cr 443 857
Mf 156 000
Rf = = = 0.4
M f ,cr 381 000
Mt 265 000
Rct = = = 0.5
M t ,cr 524 558

Rc + Rf + R 2ct = 1
0.3 + 0.4 + 0.5² = 0.95 < 1
The cylinder does not collapse under this combined load.


Static stressing manual V1-3 • STIFFENERS
A - CONTENTS
Chapter Title Issue Date

V1.3 STIFFENERS 1 12 / 1996 Original issue
V1.3.1 GENERAL 1 12 / 1996 Original issue
V1.3.2 DATA PREPARATION 1 12 / 1996 Original issue
V1.3.3 LOCAL BUCKLING 1 12 / 1996 Original issue
V1.3.4 WARPING 1 12 / 1996 Original issue
V1.3.5 FORCED CRIPPLING 1 12 / 1996 Original issue
V1.3.6 INTER-RIVET BUCKLING 1 12 / 1996 Original issue
V1.3.7 CRIPPLING 1 12 / 1996 Original issue
V1.3.8 OPTIMIZATION 1 12 / 1996 Original issue
Pages
V1.3 STIFFENERS V1.3.1

A - Contents V1.3.1
B - Symbols V1.3.3
C - References V1.3.5
V1.3.1 GENERAL V1.3.1/1
V1.3.2 DATA PREPARATION V1.3.2/1
V1.3.3 LOCAL BUCKLING V1.3.3/1
V1.3.3.1 Assumptions V1.3.3/1

V1.3.3.2 Local buckling of the flange V1.3.3/2
V1.3.3.2.1 Flat flange V1.3.3/2
V1.3.3.2.1.1 Critical local buckling stress σflt V1.3.3/3
V1.3.3.2.1.2 Local buckling factor V1.3.3/3
V1.3.3.2.1.3 Plasticity correction factor V1.3.3/4
V1.3.3.2.2 Curved flange V1.3.3/5
V1.3.3.2.2.1 Critical local buckling stress σflt V1.3.3/5
V1.3.3.2.2.2 Local buckling factor V1.3.3/6
V1.3.3.2.2.3 Plasticity correction factor V1.3.3/6
V1.3.3.2.3 Block diagram V1.3.3/7
V1.3.3.2.4 Example V1.3.3/8
V1.3.3.3 Local buckling of the web V1.3.3/11
V1.3.3.3.1 Critical local buckling stress σfla V1.3.3/12
V1.3.3.3.2 Local buckling factor kc V1.3.3/12
V1.3.3.3.3 Plasticity correction factor V1.3.3/13
V1.3.3.3.5 Example V1.3.3/16
Page V1-3 / 1 revision 1 © Copyright AEROSPATIALE 1998

Pages
V1.3.3.4 Sizing of a flanged edge V1.3.3/19

V1.3.3.4.1 Lip V1.3.3/19
V1.3.3.4.2 Bulb V1.3.3/24
V1.3.4 WARPING V1.3.4/1

V1.3.4.2 Stiffness of the equivalent elastic medium V1.3.4/2
V1.3.4.3 Warping stress (linear elasticity) V1.3.4/3
V1.3.4.4 Determination of m0 minimising σ ~ V1.3.4/4
dr
V1.3.4.5 Calculation of σ~ V1.3.4/5
dr
V1.3.4.6 Plasticity correction V1.3.4/5
V1.3.4.7 Block diagram - summary V1.3.4/7
V1.3.4.8 Example V1.3.4/8
V1.3.5 FORCED CRIPPLING V1.3.5/1

V1.3.5.2 Allowable stresses V1.3.5/2
V1.3.5.3 Equivalent thickness of the stiffener skin side flange V1.3.5/5
V1.3.5.4 Example V1.3.5/6
V1.3.6 INTER-RIVET BUCKLING V1.3.6/1

V1.3.6.2 Critical stress V1.3.6/1
V1.3.6.3 Example V1.3.6/3
V1.3.7 CRIPPLING V1.3.7/1
V1.3.8 OPTIMISATION V1.3.8/1

B - SYMBOLS
A: cross section area
b: width
C: buckling coefficient for a curved flange
e: thickness
Ec: Young's modulus in compression
ES: secant modulus
ET: tangent modulus
f: deflection
h: height
I: Inertia
k c: buckling factor for a flat item
k: diagonal tension factor
K: end fixity factor
L: length
m: number of half wave lengths
n: stress-strain shape factor (Ramberg and Osgood)
p: inter-rivet pitch
R: curvature radius
r: radius
ν: Poisson's ratio in the elastic-plastic field
ν e: Poisson's ratio in the elastic domain
νp: Poisson's ratio in the plastic field
η: plasticity correction ratio
β: elastic foundation modulus
λ: column slenderness
σdr: critical flange warping stress
σcrip: critical crippling stress
σfla: critical buckling stress of the web
σflf: critical forced crippling stress
σflt: critical buckling stress of flange
σir: critical inter-rivet buckling stress
σrTDmin: compression stress due to diagonal tension in the stiffener
σc0.2: 0.2% yielding point (conventional allowable compression yield stress)
τ: shear stress in the panel
τcp: critical folding stress of the panel
τflf: critical forced crippling shear stress
σ~: stress calculated before plasticity correction

Meaning of indexes
a: "web"
c: "compression"
bt: "flanged edge"
p: "skin"
r: "stiffener"
s: "skin side flange"
t: "flange"
tp: "pad"
x, y, z: "reference axes"

C - REFERENCES
REF. 1: Theory of elastic stability TIMOSHENKO
REF. 2: Analysis and design of flight vehicle structures BRUHN


V1-3 •1-GENERAL
The purpose of this chapter is to make it possible to:

1 - calculate the various allowable stresses in compression proper to stiffeners
(consequently, the following shall be considered as a complement to chapter
V1-1 STIFFENED PANELS).
2 - optimise a stiffener in the design phase.
V1-3 •1.1-DEFINITIONS
Stiffeners are members secured to the panel to which they are supposed to increase the
stiffness. These members are laid in the normal dominating stress direction. For example, with
a fuselage the stiffeners are called "stringers".
If the panel and the stiffener are two distinct items, the stiffener is then called a "fastened
stiffener" (it is fastened to the panel over its entire length with bolts, rivets or tack welds).
If the panel/stiffener assembly is machined from a thick plate, the stiffener is then called an
"integral stiffener".
In all cases, the length of a stiffener is always much greater than its transverse dimensions.
For local studies, these stiffeners are broken down into skin side flanges (secured to the skin),
webs and flanges.
CAP
FLANGED EDGE
STIFFENER
W
E
B
SKIN SIDE FLANGE
SKIN
Skin side flange: Part of the stiffener connected to the panel by riveting or welding, of which
the centre plane is parallel to the centre plane of the panel.
Cap: Free part of the stiffener of which the centre plane or the axis (curved cap
stringer) is parallel to the centre plane of the panel.
Web: Part of the stiffener connecting the flanges of which the centre plane is
perpendicular or forms an angle with the centre plane of the panel.
Flanged edge: Part of the stiffener connected to the cap and located opposite the web.
General - Page V1-3•1/1 revision 1 © Copyright AEROSPATIALE 1998

V1-3 •1.2-BEHAVIOUR, FAILURE MODES
Stiffeners are capable of handling a normal load Nx, a shear load Tz and a bending moment
My.
panel
My
Nx
Tz
Bending and normal loads result in stresses. The shear load causes shear stresses in the web
but which may be considered as negligible (long beam assumption).
The stiffener is mainly subjected to tension and compression stresses.
* Tension: The static strength limit of the stiffener at UL is given by the properties.
* Compression: This stiffener may be unstable under this type of load. Such instabilities are:
- under a short wave length mode (local):

Local buckling of the flange
Local buckling of the web
- under an average wave length mode:
Stiffener warping
- at the skin side flange and related to the fact that rivets are used:
Forced crippling
Inter-rivet buckling
- related in general to the stiffener:
Crippling resulting in stiffener failure. This phenomenon only occurs after
local buckling.
It is easy to understand the advantage of breaking down the stiffener into small flat or curved
interconnected plates, in order to be able to study each of the phenomena separately.

A stiffener is optimised if no local instability can occur in it before the conventional

compression allowable compression yield stress. Its allowable stress is then σc0.2.
STIFFENER INSTABILITY
LOCAL
BUCKLING CRIPPLING
WARPING OF THE CAP (closed type
cross section)
CAP
FLANGED
W EDGE
LOCAL E STIFFENER
BUCKLING B
OF THE WEB
FORCED
CRIPPLING
SKIN SIDE FLANGE
SKIN
INTER-RIVET
BUCKLING


V1-3 •2-DATA PREPARATION
V1-3 •2.1-MATERIAL DATA
These calculations require the incorporation of the elastic-plastic behaviour of materials. To

this end, the Ramberg and Osgood model is used (refer to V2-2 "Material behaviour"):
n
σ  σ 
ε= + 0.002 .  
Ec  σ c 0.2 
σ
Es =
ε
1 n 1− n
= +
ET Es Ec
Therefore, the following is needed for each type of material used:

• the Young's modulus in compression, Ec,
• the conventional allowable compression yield stress σc0.2,
• the Ramberg & Osgood coefficient, n.
V1-3 •2.2-GEOMETRICAL DATA
As the data required to study the stiffener are specific to each phenomenon, they are detailed
in the paragraphs involved.
Data preparation - Page V1-3•2/1 revision 1 © Copyright AEROSPATIALE 1998


V1-3 •3-LOCAL BUCKLING
For these calculations a stiffener angle section may be broken down into plates, the unloaded
edges of which are either hinged or free. Then the rules given in Chapter V1-2 are used
"Buckling of thin plates and shells".
Local buckling concerns all parts of the stiffener which are not connected with the skin (the
cap, the web, etc.). This is low wave length instability.
The length of the unloaded edges is very long in relation to the width of the parts making up
the angle section and therefore in relation to the half wave length of the pattern. The cap or the
web must be considered as being infinitely long plates in single compression and therefore the
final constant buckling factors must be used.
The dimensions used for local buckling calculation or sizing of the flanged edge are as
follows:
bt bt
Dbt
bbt et
et
ebt
flat flange with a straight flange

lip edge with bead
et rt
ea
ea ha
ha
curved flange web
Local buckling - Page V1-3•3/1 revision 1 © Copyright AEROSPATIALE 1998

V1-3 •3.2-LOCAL BUCKLING OF THE FLANGE
V1-3 •3.2.1-Flat flange
The boundary conditions at unloaded edges depend on two points:

- Stiffness of the link between the flange and the web,
- Whether or not there is a flanged edge and the amount of support provided by it.
* Conservatively in all cases, the web-flange connection shall be considered as a

single support.
* The condition on the extreme edge of the flange depends on the presence or absence
of a flanged edge:
if there is a flanged edge, take into account the amount of support it provides in
relation to the flange; otherwise, it is a free edge.
et et
bt L bt L
without a flanged edge with a flanged edge

V1-3 •3.2.1.1-Critical local buckling stress σflt
To calculate the critical local buckling stress of a flat flange, the results from Chapter V.1.2
(paragraph 2.3.1.1) of this manual (flat plate buckling) are used.
In all cases (whether a flanged edge or not) the equation of the critical local buckling stress of
the flat flange shall be:
2
ηEc π2  et 
σflt = kc   Eq. 1.3.3.2.1.1
12 (1 − ν 2e )  bt 
with:
bt: flange width

et: flange thickness
ν e: Poisson ratio in the elastic domain
k c: local buckling factor (dépendant des C.L. et de la géométrie)
η: plasticity correction factor
To calculate this critical stress, it is therefore necessary to determine the local buckling factor
kc and the plasticity correction factor η, as a function of the boundary conditions and the
geometry of the stiffener.
V1-3 •3.2.1.2-Local buckling factor
- without a flanged edge: kc = 0.43 (one edge supported, one free edge)
# If the flange has a flanged edge, then the factor kc of the flange depends on the amount of
support this flanged edge provides.
To ensure a simply supported condition, it has to satisfy the following conditions: (ref. 2,
paragraph C7.9).
I bt A
2.73 3
− bt ≥ 5 Eq. 1.3.3.2.1.2.1
btet btet

kc is calculated as follows:
I bt A
- with a flanged edge such that: 2.73 3
− bt ≥ 5 then kc = kc∞ = 4
btet btet
I bt A  I bt A  
- with a flanged edge such that: 2.73 3
− bt < 5 then kc = 4 + 0.714  2.73 3
− bt  − 5
btet btet  btet btet  
Eq. 1.3.3.2.1.2.2
with:
bt : flange width
Abt: cross section area of the flanged edge
Ibt: inertia of the flanged edge in relation to the neutral line of the flange
(refer to paragraph 3.4 for the equations of A1 and I1 in the case of a lip or a bulb (bead))
V1-3 •3.2.1.3-Plasticity correction factor
The critical stress noted σ~ is calculated with η=1,

flt
~ ≥ 0.5 . σ
if: σ σ ~
flt c0.2 then flt = η . σ flt,
σ flt ~
This calculation is iterative as η = f (ES, ET, ν) = g(σflt). It converges when: →σ flt
η
According to Chapter V.1.2 (paragraph 2.2) of this manual, there is:

 1 − ν 2e  E S
- For a flange without a flanged edge: η1 =  
 1 − ν2  E c
 E 
- For a flange with a flanged edge: η3 = η1  0.5 + 0.25 1 + 3 T 
 ES 
Eq. 1.3.3.2.1.3
with:
ES: secant modulus
ET: tangent modulus
ν e: Poisson’s ratio in the elastic domain
ν: Poisson’s ratio in the elastic-plastic domain
Remark: η3 < η1 in case of doubt, choose the conservative solution.

V1-3 •3.2.2-Curved flange
For local buckling calculation of a curved flange, the stiffener flange is compared to an
infinitely long cylinder under single compression (refer to paragraph V1.2.4 of this manual:
"Long cylinder buckling").
rt
et
V1-3 •3.2.2.1-Critical local buckling stress σflt
To calculate the critical local buckling stress of the curved flange, the results from Chapter
V.1.2 (paragraph 4.4.1) of this manual (long cylinder buckling) are used.
The equation for the critical local buckling stress of the flange shall be:
e 
σflt = η CEc  t  Eq. 1.3.3.2.2.1
 rt 
with:
rt: flange radius
C: local buckling coefficient
η: plasticity correction factors

To calculate this critical stress it is therefore necessary to determine the local buckling
coefficient C and the plasticity correction factor η.
V1-3 •3.2.2.2-Local buckling factor
According to Chapter V.1.2 (paragraph 4.4.1) of this manual, considering the case of long
cylinders:
−0 .15
r  r  
- For: 8.52 ≤ t ≤ 3000: C = 0.6 * In  3.748  t  
et   e  
 t 
rt
- For: ≤ 8.52: C = 0.6
et
Eq. 1.3.3.2.2.2
V1-3 •3.2.2.3-Plasticity correction factor
Noting the critical stress σ ~ calculated with η = 1,

flt
~ ~ ,
if: σ flt ≥ 0.5 . σc0.2 then σflt = η . σ flt
σ flt ~
This calculation is iterative as η = f (ES, ET, ν) = g (σflt). It converges when: →σ flt
η

E ET
- For a long cylinder in compression: η = S
Ec ES
Eq. 1.3.3.2.2.3
with:

ES: secant modulus
ET: tangent modulus

V1-3 •3.2.3-Block diagram
LOCAL BUCKLING OF THE FLANGE
flat flange curved flange E

CALCULATION OF kc CALCULATION OF C
S
rt rt
yes Flanged no ≤ 8.52 8.52 ≤ ≤ 3000
edge? et et
flanged edge
= hinged edge?
yes I bt A bt
no
S
2.73 − ≥5
3
btet btet
Eq. 1.3.3.2.1.2.1 Eq. 1.3.3.2.2.2

kc = 4    kc = 0.43 kc = 0.6  −0 . 15

I bt A bt r
 2 .73   C = 0.6 • In  3. 748   
E
kc = 4 + 0.714 − − 5 t
 btet
3
btet    e  
 t 
CALCULATION OF σflt CALCULATION OF σflt
N
e 
2
Ec π
2
e 
Eq. 1.3.3.2.1 ~
σ = kc  
t
Eq. 1.3.3.2.2.1
~ = CE
σ flt c   t
r
flt
) b 
2
12 (1 − ν e t t
PLASTICITY CORRECTION
no ~ ≥ 0.5 . σ ?
σ flt c0.2 yes
T
flat flange curved flange
I
flanged
yes no
edge?
~
σflt = η3 . σ ~
σflt = η1 . σ ~
σflt = η . σ
flt flt flt
with: with: with:
 ET  ν 2e
1 −  ES ES ET
η3 = η1  0.5 + 0.25 1 + 3  η1 =   η=
A
2
 ES  1 − ν  Ec Ec ES
and:
Eq. 1.3.3.2.1.3 Eq. 1.3.3.2.2.3
1 −  ES ν 2e
η1 =   2
1 − ν  Ec
Eq. 1.3.3.2.1.3 L
end

Use the stiffener studied in Chapter V1-1:
16
30 2
3
2
25 3
35
The geometrical characteristics are: (refer to paragraph V1.3.1)

flange thickness: et = 4 mm
flange width: bt = 15 mm
The materials properties are: (refer to paragraph V1.3.2)

- stiffener:
7075 T73510 σr = 495 MPa Ec = 73800 MPa
σc0.2 = 420 MPa n = 13.83
νe = 0.33 νp = 0.5
V1-3 •3.2.4.1-Calculation of kc
No flanged edge, therefore: kc = 0.43
~ (linear elastic material)

V1-3 •3.2.4.2-Calculation of σ flf
2
73800 ( π )
2 2
~ =k Ec π2  et   4 ~ = 2082 MPa
σ   = 0.43   let: σ
( )
flt c flt
12 (1 − ν 2e )  b t  12 1 − ( 0.33)  15 
2

V1-3 •3.2.4.3-Plasticity correction

~
σflt = η1 σ flt
with:
 1 − ν 2e  E S
η1 =  
 1 − ν2  E c
ES  E 
ν= νe +  1 − S  νp
Ec  Ec 
1
Es = n −1
1 0.002  σ flt 
+  
Ec σ c 0 .2  σ c 0 . 2 
After numerical resolution there is:
σflt = 504 MPa
Remark:
This critical local buckling stress of the flange is greater than the 0.2% yielding point.


V1-3 •3.3-LOCAL BUCKLING OF THE WEB
The boundary conditions on the unloaded edges depend on the stiffness of the connections
between the web and the cap, between the web and the skin side flange.
* Conservatively and in all cases, the web-skin side flange connection is considered as
a single support.
* The web-cap connection depends on the amount of support the flange provides in
relation to the web.
ha
ea
Remark:
If the stiffener is not provided with a cap, then it is necessary to refer to the previous chapter
(V1.3.3.2) to calculate the local buckling of the web, proceeding as for a flat flange without a
flanged edge (skin side flange-type plate).

V1-3 •3.3.1-Critical local buckling stress σfla
To calculate the critical local buckling stress of the web, the results from Chapter V.1.2
(paragraph 2.3.1.1) of this manual (flat plate buckling) are used.
The equation of the critical local buckling stress of the web is:
2
η Ec π2  ea 
σfla = kc   Eq. 1.3.3.3.1
12 (1 − ν 2e )  h a 
with:
ha: loaded edge length (web height)
ea: web thickness
k c: local buckling factor
η: plasticity correction factor
Therefore, to calculate this critical stress, the local buckling factor kc and the plasticity
correction factor η, have to be determined in function of the boundary conditions and the
geometry of the stiffener.
V1-3 •3.3.2-Local buckling factor kc
For the flange to truly support the web, it has to satisfy the condition given in Eq. 1.3.3.2.1.2.1
(refer to paragraph C7.9):
It At
2.73 3
− ≥5 Eq. 1.3.3.3.2.1
h a ea haea
At: cross section area of the flange

It: inertia of the flange in relation to the neutral line of the web
If the flange satisfies this condition then, according to Chapter V.1.2 (paragraph 2.3.1.1) in
this manual, the web is considered as an infinitely long plate simply supported on unloaded
edges:
It At
kc = kc∞ = 4 if 2.73 3
− ≥5
h a ea haea
 I At   It At
kc = 4 + 0.714  2.73 t 3 −  − 5 if 2.73 3
− <5
 h a ea haea   h a ea haea
Eq. 1.3.3.3.2.2

V1-3 •3.3.3-Plasticity correction factor
The critical stress noted σ~ is calculated with η = 1,

fla
~ ≥ 0.5 . σ
if: σ σ ~
fla c0.2 then fla = η . σ fla,
σ fla ~
This calculation is iterative as η = ƒ (ES, ET, ν) = g(σfla). It converges when: →σ fla
η
To be applied in the case when: σfla ≥ 0.5 . σc0.2

 1 − ν 2e  E S
- For a flange-type plate (no cap): η1 =  
 1 − ν2  E c
 E 
- For a plate simply supported in compression: η3 = η1  0.5 + 0.25 1 + 3 T 
 ES 
Eq. 1.3.3.3.3
with:
ES: secant modulus
ET: tangent modulus
ν: Poisson’s ratio in the elastic-plastic domain
Remark:
If the stiffener does not have a cap, then refer to the previous chapter (V1.3.3.2) for calculation
of the local buckling of the web proceeding as for a flat flange without a flanged edge (side
skin flange-type plate).


LOCAL BUCKLING OF THE WEB

E
CALCULATION OF kc
Is the flange a
S
support for the
yes web? no
S
It At
2.73 − ≥5
3
haea haea
Eq. 1.3.3.3.2.1
kc = 4
 It At  
kc = 4 + 0.714  2.73 −  − 5 
 
3
h a ea haea 
CALCULATION OF σfla
E
2
Ec π  ea 
2
N
2   Eq. 1.3.3.3.1
~
σ = kc
fla
12 (1 − ν e )  h a 
no ~
σ fla ≥ 0.5 . σc0.2 yes
T
~ if flange
• σfla = η3 . σ fla
~
• σfla = η1 . σ fla if no flange
with:

I
ET 
η3 = η1  0.5 + 0.25 1 + 3 
 ES 
and:
η1 = 
 1 − ν 2e  E S

 1 − ν2  E c
A
end
Eq. 1.3.3.3.3
L
Use the stiffener studied in this chapter:

16
30 2
3
2
25 3
35
The geometrical properties are: (refer to paragraph V1.3.1)
cap thickness: et = 4 mm
cap width: bt = 15 mm
web thickness: ea = 2 mm
web height: ha = 26,5 mm
The characteristics of the materials are: (refer to paragraph V1.3.2)
- stiffener:
7075 T73510 σr = 495 MPa Ec = 73800 MPa
σc0.2 = 420 Mpa n = 19.66
νe = 0.33 νp = 0.5
V1-3 •3.3.5.1-Calculation of kc
e t b 3t bt et 4 (15) 3 4 .15
0.91 3
− = 0.91 3
− = 56.8 ≥ 5
h a ea ha ea 26.5 (2 ) 26.5 . 2
The condition Eq.1.3.3.3.2.1.1 is verified and therefore the flange is a sufficient support for the
web.
kc = 4
V1-3 •3.3.5.2-Calculation of σfla

2 2
~ =k E c π2  ea  73800 ( π ) 2  2  ~ = 1552 MPa
σ   =4   let: σ
12 (1 − (0.33) 2 )  26.5 
fla c fla
12 (1 − ν 2e )  h a 

V1-3 •3.3.5.3-Plasticity correction

~
σfla = η3 σ fla
with:
 E   1 − ν 2e  E S
η3 = η1  0.5 + 0.25 1 + 3 T  and η1 =  
 ES   1 − ν2  E c
ES  E 
ν= ν e + 1 − S  ν p
Ec  Ec 
1
Es = n −1
1 0.002  σ fla 
+  
Ec σ c 0 .2  σ c 0 . 2 
1
ET = n −1
1 0.002 n  σ fla 
+  
Ec σ c 0.2  σ c 0.2 
After numerical resolution, the following is found:
σfla = 476 Mpa
Remark:
This critical local buckling stress of the web is greater than the 0.2% yielding point.


V1-3 •3.4-SIZING OF A FLANGED EDGE
Flanged edges are of two types: lips and beads.

bt bt D
et bbt et
ebt
lip bead
The purpose of a flanged edge is to provide, at least, a single support for the flange. However,
for this flanged edge to be efficient, it must not buckle before the flange. These two conditions
are translated as follows:
- The flanged edge must be a support for the flange: (Ref. 2, paragraph C7.9)
I bt A
2.73 3
− bt ≥ 5 Eq. 1.3.3.4.1
bt et bt et
with: Abt: cross section area of the flanged edge

Ibt: inertia of the flanged edge in relation to the neutral line of the flange
- The flanged edge must not buckle before the flange: the critical local
buckling stress of the flanged edge must be greater than that of the flange.
σflbt ≥ σfltalon Eq. 1.3.3.4.2
V1-3 •3.4.1-Lip
For a lip, there is:

bt
Abt = bbt ebt bending axis
1
Ibt =   e bt b 3bt et bbt
 3
ebt
lip

Being, for the first condition (Eq. 1.3.3.4.1):

3
b  b b
0.91  bt  − bt − 5 t ≥ 0 Eq. 1.3.3.4.1.1
 et  et e bt
For the second condition, the limits are set in the elastic domain as, if this condition is satisfied
in this domain, it is all the more so satisfied in the plastic domain.
In fact, if put down:

- η1 and η3 the plasticity correction factors respectively of the flanged edge and the
flange then: η1 ≥ η3 (refer to Chapter V.1.2 (paragraph 2.2)) of this manual.
- σ ~ the critical buckling stress calculated in the elastic domain, such as: σ = η σ ~
fl fl fl
then:
σ~ ~
fl bt ≥ σ fl flange
and ⇒ σ fl bt ≥ σ fl flange
η1 ≥ η3
The sufficient condition is therefore: σ ~ ~
fl bt ≥ σ fl flange which is formulated as follows:
2 2
E c π 2  e bt  EC π2  et  b bt  k cbt   b t 
kcbt   ≥ k   ⇒ ≤   Eq. 1.3.3.4.1.2
e bt  k cflange   e t 
cflange
12 (1 − ν2e )  b bt  12 (1 − νe2 )  b t 
with the kc factors given in the previous chapter:

kc flange = 4: for the flange
kc bt = 0.43: for the flanged edge
The minimal ebt equation is deduced:

 4   et   et 
ebt =     bbt let ebt = 3.05   bbt
 0.43   b t   bt 
Using this equation in Eq.1.3.3.4.1.1 it is possible to obtain the minimum value of the ratio
( b bt ) min. e
as a function of the ratio t of the flange, so that the lip is at least a single support
bt bt
for the flange:
 et 
2   bt 
2 
( b bt ) min.
=
1
  1 + 1 + 18.2  0.43    
bt 1.82  bt    4   et  
 
let: Eq. 1.3.3.4.1.3
2
e 
2  
( b bt ) min.
= 055  t  1 + 1 + 6  b t  
bt  bt    e t  


Having chosen a value for bbt so that (bbt)selected ≥ (bbt)min., it is necessary to determine the
minimum width of the flanged edge so that it does not buckle before the flange. To this end,
Eq. 1.3.3.4.1.2 is used:
(e bt )min.  4   et  (e bt )min. e 
=    let = 3.05  t  Eq. 1.3.3.4.1.4
(b bt )selected  0.43   bt  (b bt )selected  bt 
Finally, the straight flanged edge dimensions should verify that:

bbt ≥ (bbt)min.
ebt ≥ (ebt)min.
with:
Minimum dimensions ebt and bbt of a flanged edge as a function of the flange dimensions.
1 - bbt minimum as a function of bt and et.
2 - ebt minimum as a function of bt, et and bbt selected in 1.
(bbt)selected
bt
(bbt)min.
(ebt)min.
et
bt
Fig.: 1.3.3.4.1.1

Special cases:
b bt e
Case 1: If bbt is imposed then by putting down βt = , such that bt = bt0 - bt and bt0 fixed:
b t0 2
it is necessary: bt0
- according to Eq. 1.3.3.4.1.1:
e bt 5
≥ 3
b t0 3  b t0   b t0 
2.5 + 0.91β t   − βt   et bbt
 et   et  bt
- according to Eq. 1.3.3.4.1.2:
2 ebt
 e bt  e  e
  − 2  bt  + 6.1 . β t t ≤ 0
 b t0   b t0  b t0
e
The second condition is only possible if: t ≥ 6.1 . βt, then in this case it is necessary:
b t0
et e e
1 − 1 − 6.1β t ≤ bt ≤ 1 + 1 − 6.1β t t
b t0 b t0 b t0
Furthermore, the case ebt ≥ bt0 means making the flange thicker, which is not the purpose of
e
this study. Therefore, work is limited to cases: ebt ≤ bt0 ⇒ bt ≤ 1;
b t0
b e
To conclude, if bbt is imposed with βt = bt and bt = bt0 - bt , then it is necessary:
b t0 2
  
  
 et   5   e bt
max 1 − 1 − 6.1β t ;  3   ≤ b ≤1 Eq. 1.3.3.4.1.5
 b t0   b   b 
 2.5 + 0.91β 3  t 0  − β  t 0    t0
  t
 et  
 e t   
 
ebt must be selected in the following non-hatched domain: (case where β = 0.25).
Determination of the thickness ebt of a flanged edge
the height of the flanged edge bbt is imposed in relation to
dimensions et and bt0 of the flange
here b bt = 0.25 b t0
Fig.: 1.3.3.4.1.2

e bt
Case 2: If ebt is imposed, then by putting down αt = , it is necessary that:
et
  5 bt    b bt 
 3 0.55 25 b 2t 0.55  5 b t 25 b 2t b 
    ≤
 αt α e + − 0.163  + 3
α t  α t e t
− − 0.163  ≤  0.328α t t 
α 2t e 2t   e t 
2 2
 t t  αt et et 

Eq. 1.3.3.4.1.6
e bt
For example, this is the case of a bent over edge. Then there is αt = = 1 and bbt has to be
et
selected in the following non-hatched domain:
Dimensions of the straight flanged edge as a function of the dimensions of the flange if
they both have the same thickness et: α = 1
The flanged edge buckles

before the flange
The flanged edge is not a

support for the flange
Fig.: 1.3.3.4.1.3

V1-3 •3.4.2-Bulb
bt
For a bulb flange there is: Dbt
πD 2bt
Abt =
4 et
2
πD 4bt πD 2bt  D bt − e t 
Ibt = +  
64 4  2 
bulb flange
with Ibt inertia of the bead in relation to the
neutral line of the flange.
Assumption: the critical buckling stress of the bead is always greater than that of the flange.
In this case, there is only one condition to verify:
The bead must be at least one support for the flange.
4 3 2
D  D  D  b 
This means, using the Eq. 1.3.3.4.1:  bt  − 1.6  bt  − 0.374  bt  ≥ 7.44  t  ;
 et   et   et   et 
Eq. 1.3.3.4.2
By representing this equation by the following curve (Fig. 1.3.3.4.2.1), it is possible to
determine the minimum diameter of the bead Dbt as a function of dimensions et and bt of the
flange:
Minimum values of the Dbt/ et ratio as a function of the dimensions bt and et

of the flange, so that a bead is a support for the flange
bulb flange
Fig. 1.3.3.4.2


FLANGED EDGE SIZING
I A
1 - 2.73 bt 3 − bt ≥ 5
CONDITIONS
and ~ ≥σ
2- σ ~
E
fl bt fl flange
bt et bt et
lip
Abt = bbt ebt bt

Type of flanged edge?
πD
2
bead
bt Dbt
S
bt
Abt =
 1
Ibt =   ebt b3bt et bbt 4 et
 3
S
4 2 2
lip
ebt
Ibt =
πD bt
+
 D bt − e t 
πD bt
bulb flanged
64 4  2  edge
if ebt and bbt are to be defined

Condition 2 is always verified and condition 1
imposes at least:
e 
2  2
 1 + 1 + 6 b t  
E
( b bt ) min. 4 3 2
0.55 t   D bt  D  D  b 
bt  bt    e t     − 1.6 bt  − 0.374 bt  = 7.44 t 
  et   et   et   et 
refer to Fig. 1.3.3.4.2
( e bt ) min. e 
= 3.05 t 
( b bt )selected  bt 
refer to Fig. 1.3.3.4.1.1
if ebt to be defined and bbt imposed N

b bt e bt
with βt = and bt = bt0 - it is therefore necessary that:
bt0 2
e bt



≥ max  1 − 1 − 6.1β t
et  


; 
5




T
 b t 0  
3
bt0  bt0  3  bt0 
 2.5 + 0.91β t   − β t   
   et   e t   
I

e bt
≤1
b t0
if bbt to be defined and ebt imposed
by putting down αt =
e bt
et
it is necessary that: A
  
0.55  5 b t  0.55  5 b t
2 2
b bt 25 b t 25 b t
≥ 3  + 
− 0.163  + 3  − − 0.163 
 α t  α t e t  
L
2 2 2 2
et αt et  α t  α t et αt et

b bt  b 
≤  0.328α t 
t e
et  t 


V1-3 •4-WARPING
Here the problem involved is the general buckling of the stiffener cap supported by an elastic
foundation: the web. In fact, if the flange subjected to compression load buckles, it may buckle
the entire stiffener. Therefore, the case will be considered as being between two stiffener
cleats to isolate the flange as a beam, supported at its ends and elastically retained by the web
over its entire length.
The assumption taken is that the stiffener flange behaves like a beam with two hinges at the
cleats. The cleats are considered as being rigid supports as defined in Chapter V-1-9 "Cleats".
et
bt
Z
CdGflange
ea
ha
L
Y
X
Fig. 1.3.4.1
Warping - Page V1-3•4/1 revision 1 © Copyright AEROSPATIALE 1998

V1-3 •4.2-STIFFNESS OF THE EQUIVALENT ELASTIC MEDIUM
Fig. 1.3.4.2.1
Let "α" designate the elastic constant of equivalent discrete supports and "a" the equal distance
separating two supports. The stiffness of the equivalent elastic medium is expressed by the
quantity:
α
β=
a
Where "β" called the "foundation modulus" has the dimensions of a force divided by the
square of a length. It represents the intensity of the foundation reaction.
In this case, the bending rigidity around axis "x" of the web characterises the elastic
foundation and is expressed as follows:
F=α*f;
3. E c . I X .web
α=
h 3a
a. e 3a
IX.web =
12
(refer to Figs. 1.3.4.1 & 1.3.4.2.1)
Fig. 1.3.4.2.2

The stiffness of the equivalent elastic medium is:
3
E e 
β = c . a  Eq. 1.3.4.2
4  ha 
with:
Ec: stiffener elasticity modulus
ea: web thickness
ha: web height (refer to Fig. 1.3.4.1)
V1-3 •4.3-WARPING STRESS (linear elasticity)
The energy method is used to determine the critical buckling load (Ref. 1, paragraph 2.10)
considering the deflection of the elastic line as a sinusoidal series with a wave node at each
end of the line, i.e. at the stiffener cleats.
By minimising this deformation energy, the lowest critical buckling load is obtained, which is:
critical warping stress (linear elasticity):

~ = π . E c . I zt .  m 2 + 
2
β. L4
σ dr 2  2 4  Eq. 1.3.4.3
St L  m . π . E c . I zt 
with:
Ec: stiffener elasticity modulus
IZt: flange minimum inertia (*)
St: area of the flange section (*)
L: length of the beam modelling the flange between the cleats (refer to
Fig. 1.3.4.1)
m: number of wave half lengths
β: elastic foundation modulus (refer to Eq. 1.3.4.2)
"m" (number of half wave lengths) shall be an integer and greater than or equal to 1 as the
cleats are located at the wave nodes, in the assumptions. This number of wave half lengths "m"
in which the bar is sub-divided during buckling is determined by the condition that the
equation 1.3.4.3 is minimal.
(*) IZt is the minimum inertia of the flange in relation to the Z axis, therefore passing through
the centre of gravity of the flange section, which is expressed as follows:
• case of a flange without a flanged edge:
e t . b 3t
Izt = (refer to Fig. 1.3.4.1)
12
• case of a flange with a flanged edge:
Zt
Izt = I(Gt;Zt)
(main inertia) Gt

~
V1-3 •4.4-DETERMINATION OF m0 MINIMISING σ dr
~ it is necessary to determine "m" such that:

To minimise σ dr
~
dσ dr
=0
dm
Determination of m giving the lowest values of σ ~

dr
L β
m0 = 4 Eq. 1.3.4.4
π E c . I zt
with:
Ec: elasticity modulus of the stiffener in compression
IZt: inertia of the flange in relation to the Z axis (refer to Fig. 1.3.4.1)
L: length of the beam modelling the flange between cleats (refer to Fig. 1.3.4.1)
m0: number of wave half lengths
β: elastic foundation modulus (refer to Eq. 1.3.4.1)
There is little chance of the value of m0, determined with Equation 1.3.4.4, being an integer
and we have already established that it is necessary to have an integer value of m to have a
~ will be carried out for the two closest
wave node at the cleats. Therefore, the calculation of σ dr
integers bracketing the calculated m and the integer value of m giving the minimum σ ~ will
dr
be used.
~ calculated with m0 is the warping stress when the cleats are infinitely spaced apart
σ dr
3
~ (m ) = E c e 
σ dr 0 I zt  a 
St  ha 
~ (E(m )) > σ
If σ ~ (E(m + 1)) then it is necessary to take m = E(m + 1)
dr 0 dr 0 0
otherwise m = E(m0)
with E(m0) = integer part of m0

~
V1-3 •4.5-CALCULATION OF σ dr
Finally, putting down:

~ (m ) = 2 L β
σ dr 0 E c I zt β , m 0 = 4 , m 1 = E( m 0 ) and m 2 = E( m 0 + 1)
St π E c I zt
~  m  2  m  2 
then: ~ (m ) = σ dr ( m 0 )
σ  i  +  0  
dr i
2  m   m i  
 0
And the critical stress is calculated as follows:

~ = min σ~ ( m ), σ~ ( m ) Eq. 1.3.4.5
σ dr [ dr 1 dr 2 ]
V1-3 •4.6-PLASTICITY CORRECTION
Putting down σ ~ , the critical stress calculated with η = 1, if: σ

~ ≥ 0.5 . σ then the critical
dr dr c0.2
stress after plasticity correction shall be:
~
σdr = η5 σ Eq. 1.3.4.6
dr
σ ~
This calculation is iterative as η5 = ƒ(ET, ν) = g(σdr). It converges when: dr → σ dr
η5
η5 is the plasticity correction factor:
E
η5 = T with ET = tangent modulus
Ec
Ec = Young's modulus in compression of the stiffener
In the design phase, we shall work in such a manner that σcr, (cor) ≥ σc0.2 is obtained, so that
only σc0.2 is used as an allowable compression value.


V1-3•4.7-BLOCK DIAGRAM - SUMMARY
WARPING
E
CALCULATION OF IZt
CALCULATION OF St S
Calculation of β
3
E e  Eq. 1.3.4.2
S
β = c . a 
4  ha 
Calculation of m0
L β Eq. 1.3.4.3
m0 = 4
E
π E c I zt
~ (m )
Calculation of σ dr 0
~ (m ) = 2 Eq. 1.3.4.5
σ dr 0 E c I ztβ
St
Calculation of σ ~
dr
N
with: m1 = E(m0)
m2 = E(m0 + 1)
T
~ 2 2
~ σ dr ( m 0 )   m i   m0  
 
σ dr (mi) =
2   m  +  m   Eq. 1.3.4.5
 0 i 
then:
~ = min σ~ ( m ), σ~ ( m )
σ [ dr 1 dr 2 ]
I
dr
no
~
σ dr ≥ 0.5 . σc0.2 ? yes
~
σdr = η5 σ
A
dr
Eq. 1.3.4.6 ET
with: η5 =
L
Ec
End

V1-3 •4.8-EXAMPLE
Using the stiffener studied in this chapter:

16
30 2
3
2
25 3
35
The characteristics are: (refer to paragraph V.1.3.1)
Material: 7075 T7351 σr = 495 Mpa Ec = 73800 MPa

σc0.2 = 420 Mpa n = 13.83
flange thickness: et = 4 mm
flange width: bt = 16 mm
web thickness: ea = 2 mm
web height: ha = 26,5 mm
Inter-frame distance: L = 530 mm
V1-3 •4.8.1-Calculation of St and of IZt
St = et . bt St = 4 . 16 St = 64 mm²
e t . b 3t 4 . 16 3
Izt = Izt = Izt = 1365,3 mm4
12 12
V1-3 •4.8.2-Foundation stiffness

3 3
E e  73800  2 
β = ca .  a  being therefore β= . 
4  ha  4  26.5 
giving β = 7.93 MPa
V1-3 •4.8.3-Calculation of m0
L β 530 7.93
m0 = 4 giving m0 = 4
π E ct . I zt π 73800 .1365.33
and therefore m0 = 2.83

V1-3 •4.8.4-Critical stress (linear elasticity)
•σ~ (m ) = 2 E I β ~ (m ) = 2 73800 . 1365.3 . 7.93

σ ~ (m ) = 883 MPa
σ
dr 0 c zt dr 0 dr 0
St 64
• m1 = E(m0) m1 = 2
• m2 = E(m0 + 1) m2 = 3
~ 2 2
~ (m ) = σ dr ( m 0 )   m i  +  m 0  
•σ dr i  
2  m 0   mi  
2 2
~ (m ) = 883   2  +  2.83   = 1104.5 MPa
σ dr 1
2   2.83   2  
2 2
~ 883   3   2.83  
σ dr (m2) =   +   = 889 MPa
2   2.83   3  
~ (m ) < σ
As σ ~ (m ) then σ
~ =σ ~ (m = 3) ~ = 889 MPa
let: σ
dr 2 dr 1 dr dr 2 dr
V1-3 •4.8.5-Plasticity correction

~ = η .σ
σ dr 5 dr
ET σ c0.2
with η5 = and if σcr ≤ then ET = Ec
E ct 2
1
otherwise ET = n −1
1 0.002 . n  σ cr ,cor 
+  
Ec σ c 0.2  σ c 0.2 
In this case: ~ = 889 MPa > σ c0.2

σ dr
2
To determine the corrected critical stresses, it is necessary to solve the following equations
numerically.
~
σ dr
σdr = n −1
0.002 . n E c  σ dr 
1+  
σ c 0 .2  σ c 0.2 
889
application: σdr = 12 .83
0.002 .13.83 . 73800  σ dr 
1+  
420  420 
after numerical resolution, the following is obtained σdr = 380 MPa


V1-3 •5-FORCED CRIPPLING
The forced crippling phenomenon may occur when the panels carry loads in diagonal tension
(refer to paragraphs V1-1.5 and V1-1.6: stiffened panels in single nominal shear).
Waves caused by folding of panel bays induce local deflection in the skin side flange of the
stiffener which increases when the thickness of the flange is thinner than that of the panel.
The free edge (AA) of the flange tends to follow the waves of the skin whereas the edge it
makes with the web (BB) remains straight.
Fig. 1.3.5.1.1
Forced crippling - Page V1-3•5/1 revision 1 © Copyright AEROSPATIALE 1998

A panel A.A
A stiffener
Fig. 1.3.5.1.2
Then the local bending stresses amplified by the compression are added to the compression
stresses due to the diagonal tension.
V1-3 •5.2-ALLOWABLE STRESSES
Forced crippling occurs when the maximum compression in the stiffener reaches a value
defined by a empirical equation integrating the plasticity of the material:
STRAIGHT STIFFENER (e.g.: Stringer...R(Curvature radius) > 3800 mm)

E 1/ 3
σ c 0.2 r  e' . E 
S σflf = - 0.051 .
σ c 0.2 r
. k 2/3 .  sr cr  Eq. 1.3.5.2.1 (Ref. 2, Fig. C11.38)
 e p . E cp 
+ 0.002
S E cr
E
with:
N e'sr: equivalent thickness of the stiffener skin side flange
Ecr: compression of the stiffener Young's modulus
T ep: panel thickness
I Ecp: compression of the panel Young's modulus
k: diagonal tension factor (paragraph V-1.1.5.4)
A σc0.2: conventional allowable compressive yield stress of the stiffener
L

CURVED STIFFENER (e.g.: Frame...500 mm ≤ r (curvature radius) ≤ 3800 mm)

1/ 3
E
σ c 0.2 r  e' . E 
σflf = - (0.511r + 3095.3) . 10-5 .
σ c 0.2 r
. k 2 / 3 .  sr cr  S
 e p . E cp 
+ 0.002 S
E cr
Eq. 1.3.5.2.2 (Ref. 2, Fig. C11.38) E
with: N
e'sr: equivalent thickness of the stiffener skin side flange
Ecr: compression Young's modulus of the stiffener
T
ep: panel thickness I
Ecp: compression Young's modulus of the panel
k: diagonal tension factor (§ V-1.1.5.4) A
σc0.2: conventional allowable compression yield stress of the stiffener L
r: stiffener curvature radius
The forced crippling calculation is to be integrated in the general iteration of the calculation of
a stiffened panel when the critical stiffener stress calculation step is reached. At each iteration,
k, according to the applied stress, is set. Like this it is possible to calculate the compression
stress value applied to stiffener σ TD
r min as well as the allowable stress value.
The calculation converges when: σflf ≈ σ TD

r min
The calculated value of σflf must be clipped at σc0.2.
In other words, forced crippling occurs when the shear stress in the panel reaches a critical
value: τ = τflf
To determine the load τ = τflf causing forced crippling of the stiffener skin side flange, it is
necessary to increment τ (single shear stress in the panel) until
σ TD
rmin
= σflf

The following method is used to estimate τflf with sufficient accuracy:

σ flf ( 0 ) τ (0)
RF(0) = TD
RS(0) =
σ r min( 0 ) τ cp
E It is possible to estimate RS(1) (iteration 1) using the following equation

RS(1) = RF(0) (RS(0) - 1) + 1 Eq. 1.3.5.2.3
S Knowing RS(1) and τcp, it is possible to deduce τ and therefore τflf:
S τflf = RS(1) . τcp Eq. 1.3.5.2.4
E τflf(i) = RS(i) . τcp

R ϕ (i ) − 1
N Then to iterate until RF(i) ≈ 1, putting down: k(i) = Sϕ (*)
R S (i) + 1
k (i ). R S (i) σ TD
r min ( i )
T = TD
k (0). R S (0) σ r min (0 )
I
with:
A σflf(i): Critical forced crippling stress at iteration (i)
σrTDmin(i): Compression stress due to diagonal tension in the stiffener at iteration (i)
L τ(i): Shear stress in the panel at iteration (i)
τcp: Critical panel folding stress
τflf: Critical forced crippling shear stress
(*) Reminder:
R ϕS (i ) − 1
According to paragraph V1-1.6.4, if (RS ≥ 1) ⇒ (k(i) = )
R Sϕ (i ) + 1
with
 0.5 < B < 1 ⇒  ϕ = 0.4343 + 260.6 e p A 
  
A   rB 
 0.5 < A < 1 ⇒  ϕ = 0.4343 + 260.6 e p B 

  
B   rA 
 B 
 0.2 < A < 0.5   ep 
  ⇒  ϕ = 0.4343 + 521.2 
B
 0.2 < < 0.5   r
 A  

where:
A: transverse member pitch
B: stiffener pitch
r: panel curvature radius
ep: panel pocket thickness

V1-3 •5.3-EQUIVALENT THICKNESS OF THE STIFFENER SKIN SIDE

FLANGE
The presence of a panel over thickness under the stiffener (PAD) delays the occurrence of
forced crippling. e'sr, the effective thickness of the skin side flange, is then greater than esr
(Eq. 1-3.5.2.1 or 2):
bsr
esr
etp
ep
btp
Fig. 1.3.5.3.1
Equivalent thickness of the skin side flange:

E
Ep S
(e )
2
e'sr = e sr2 + tp − ep limited to esr' ≤ 1.35 . esr Eq. 1.3.5.3.1 S
Er
E
• if etp > 1.5 ep, take etp = 1.5 . ep in the calculation of e'sr N
T
with:
I
etp: pad thickness
esr: stiffener skin side flange thickness
A
ep : panel pocket thickness L

V1-3 •5.4-EXAMPLE
Use the stiffener studied in Chapter V1-1 considering the case when it is straight, then the case
in which it is curved: 16
30 2
3
2
25 3
35
Stiffener material: 7075 T73510 σr = 495 MPa Ecr = 73800 MPa
σc0.2 = 420 MPa

stiffener skin side flange thickness: esr = 3 mm
skin side flange width: bsr = 25 mm
Panel material: 2024 PLT351 σr = 440 Mpa Ecp = 70300 MPa
σc0.2 = 270 Mpa

skin thickness: ep = 2 mm
pad thickness: etp = 3 mm
pad width: btp = 35 mm
frames pitch: L = 530 mm
V1-3 •5.4.1-Straight stiffener
According to the example in paragraph V1-1.5.12, there is:

k = 0.24
τ = 160 MPa
τcp = 52 MPa
σ TD
r min = - 55 MPa
V1-3 •5.4.1.1-Equivalent thickness
Ep
(e )
2
e'sr = e sr2 + tp − ep
Er
70300
application: e'sr = 32 + (3 − 2)2 e'sr = 3.15 < 1.35 . esr = 4.05
73800
e'sr = 3.15

V1-3 •5.4.1.2-Normal critical stress
In the case of a straight stiffener, the critical stress is formulated:

1/ 3
σ c 0.2 r  e ' . E cr 
σflf = - 0.051 . .k 2/3
.  sr 
σ c 0.2 r  e p . E cp 
+ 0.002
E cr
with k = 0.24 (refer to paragraph V1-1.5.12), therefore:
1/ 3
420  315
. . 73800 
σflf = - 0.051 . . 0.24 2 / 3 .  
420  2 . 70300 
+ 0.002
73800
σflf = - 111 MPa
with σ TD
r min (0) = - 55.3 MPa (refer to paragraph V1-1.5.12), therefore:
σ ( 0) − 111
RF(0) = TD flf =
σ r min(0 ) − 55
RF(0) = 2.017
V1-3 •5.4.1.3-Critical shear stress
with τ(0) = 160 MPa and τcp = 52 MPa (refer to paragraph V1-1.5.12), therefore:
τ ( 0 ) 160
Rs(0) = =
τ cp 52
Rs(0) = 3.0769
Calculation of RS(1):
RS(1) = RF(0) (RS(0) - 1) + 1
RS(1) = 2.017 (3.0769 - 1) + 1
RS(1) = 5.189
Calculation of the new panel loading τ:
τ(1) = RS(1) . τcp = 5.189 . 52
τ(1) = 269 MPa
This new shear value in the panel will give a new loading factor k and therefore a new
stiffener stress. By continuing the calculation loop in paragraph V-1-1.5.12, there is:
R 0. 4343 − 1 TD
k (1) R s (1) σ r min (1)
S(1)
k(1) = 0. 4343 = 0.343; ≈ = 2.41 ⇒ σ TD
r min (1) = - 133 MPa
R +1 k (0 ) R s (0 ) σ TD ( 0 )
S(1) r min
and ⇒ σflf (1) = - 141 MPa

σ flf (1) 141
RF(1) = = ⇒ RF(1) = 1.062
σ TD
r min (1) 133

giving the new Rs:
RS(2) = RF(1) (RS(1) - 1) + 1

RS(2) = 1.062 (5.1499 - 1) + 1
RS(2) = 5.45
⇒ τ(2) = 283 MPa
This new shear value in the panel will give a new loading factor k and therefore a new
stiffener stress. By continuing the calculation loop of paragraph V-1-1.5.12, there is:
TD
k (2) R s (2 ) σ r min (2 )
k(2) = 0.352; ≈ TD = 2.6 ⇒ σ TD
r min (2) = - 143 MPa
k (0) R s (0) σ r min (0 )
and ⇒ σflf (2) = - 143 MPa
and therefore a new RF:

RF(2) = 1.001
An RF very close to 1 is obtained after two iterations. The calculation is stopped here and
therefore:
τflf = 283 MPa

V1-3 •5.4.2-Curved stiffener
According to the example in paragraph V1-1.6.13, there is:

r = 2820 mm
k = 0.393
τ = 160 MPa
τcp = 57 MPa
σ TD
r min = - 76 MPa
V1-3 •5.4.2.1-Equivalent thickness
Ep
(e )
2
e'sr = e sr2 + tp − ep
Er
70300
application: e'sr = 32 + (3 − 2)2 e'sr = 3.15 < 1.35 . esr = 4.05
73800
e'sr = 3.15
V1-3 •4.5.1.2-Normal critical stress
With a curved stiffener, the critical stress is formulated:

1/ 3
-5 σ c 0.2 r  e' . E 
σflf = - (0.511 . r + 3095.3) . 10 . .k 2/3
.  sr cr 
σ c 0.2 r  e p . E cp 
+ 0.002
E csr
with k = 0.393 (refer to paragraph V1-1.6.13), therefore:

1/ 3
-5 420  315
. . 73800 
σflf = - (0.511 . 2820 + 3095.3) . 10 . . 0.393 2 / 3 .  
420  2 . 70300 
+ 0.002
73800
σflf = - 137 MPa
with σ r min (0) = - 76 MPa (refer to paragraph V1-1.6.13), therefore:
TD
σ ( 0) − 137
RF(0) = TD flf =
σ r min(0 ) − 76
RF(0) = 1.804
V1-3 •5.4.1.3-Critical shear stress
with τ(0) = 160 MPa and τcp = 57 MPa (refer to paragraph V1-1.6.13), therefore:
τ ( 0 ) 160
Rs(0) = =
τ cp 57
RS(0) = 2.807

Calculation of RS(1):
RS(1) = RF(0) (RS(0) - 1) + 1

RS(1) = 1.804 (2.807 - 1) + 1
RS(1) = 4.26
Calculation of the new panel loading τ:

τ(1) = RS(1) . τcp = 4.26 . 57
τ(1) = 242 MPa
This new shear value in the panel will give us a new loading factor k and therefore a new
stiffener stress. By continuing the calculation iteration of paragraph V-1-1.6.13, there is:
R 0.8039 − 1 TD
k (1) R s (1) σ r min (1)
S(1)
k(1) = 0.8039 = 0.524; ≈ TD = 2.023 ⇒ σ TD
r min (1) = - 154 MPa
R +1 k (0) R s (0) σ r min (0)
S(1)
and ⇒ σflf (1) = - 166 MPa

σ flf (1) 166.97
RF(1) = = ⇒ RF(1) = 1.08
σ TD
r min (1) 154.59
giving the new RS:

RS(2) = RF(1) (RS(1) - 1) + 1
RS(2) = 1.08 (4.26 - 1) + 1
RS(2) = 4.52
⇒ τ(2) = 257 MPa
This new shear value in the panel will gives us a new loading factor k and therefore a new
stiffener stress. By continuing the calculation iteration of paragraph V-1-1.5.12, there is:
TD
k (2) R s (2 ) σ r min (2 )
k(2) = 0.541; ≈ TD = 2.22 ⇒ σ TD
r min (2) = - 169 MPa
k (0) R s (0) σ r min (0 )
and ⇒ σflf (2) = - 170 MPa
and therefore a new RF:

RF(2) = 1.007
An RF very close to 1 is obtained after two iterations. The calculation is stopped here and
therefore:
τflf = 257 MPa

V1-3 •6-INTER-RIVET BUCKLING
The inter-rivet buckling phenomenon occurs when there is local buckling in the panel (or the
stiffener skin side flange) between two consecutive rivets. It may occur if the panel is loaded
in compression. To prevent this phenomenon from occurring, the compression stress in the
panel (or in the stiffener skin side flange) is limited to a value that will be defined here after.
A 2b wide sheet strip is assimilated to a column plate (refer to V1-2).

A.A
A
panel
ep
es
b b
p
stiffener base
A
V1-3 •6.2-CRITICAL STRESS
The critical inter-rivet buckling stress is given by:

2
π2 .Ec  e 
σir = η5   Eq. 1-3.6.2.1
12  Kp 
Inter-rivet buckling - Page V1-3•6/1 revision 1 © Copyright AEROSPATIALE 1998

with:
Ec: the compression modulus of the material of the item involved

E
η5: plasticity correction factor (column plate) η5 = t (refer to paragraph V1-2.2.2)
Ec
K: end fixity coefficient, depending on the type of fastener used
e: thickness of the item involved (skin or flange)
Tests have shown that the boundary conditions at the ends of the column depend on the type of
fasteners used. The ultimate conditions that will be encountered will be located between the
dual-hinged conditions (K = 1) and the dual-fixed conditions (K = 0.5).
Countersunk head bolt (or rivet) K = 0.66:
E  e
2
σir = 1.89 . η5 . Ec   Eq. 1-3.6.2.2

 p
S
S Tack welding or flat head bolt K = 0.54:
E  e
2
σir = 2.82 . η5 . Ec   Eq. 1-3.6.2.3

N  p
T with:
I Ec: compression modulus of the material of the item involved

A e: panel or stiffener skin side flange thickness
E
L η5: plasticity correction factor η5 = t (refer to paragraph V1-2.2.2)
Ec
During the design phase, work shall be organised to have σir ≥ σc0.2 so as to use only σc0.2 as
the allowable compression value.

V1-3 •6.3-EXAMPLE
Skin inter-rivet buckling:
The stiffener skin side flange and the panel are joined by countersunk head bolts, pitch 25 mm.
ep = 3 mm
countersunk head bolt => K = 0.66
p = 25 mm
Ecp = 70300 MPa
np = 7.05
Critical stress:
2
 ep 
Eq. 3.3.6.2 σirp = 1.89 . η5 . Ecp  
 p
2
3
σirp = η5 . 1.89 . 70300   σirp = η5 . 1911 MPa
 25 
Plasticity correction:
Et 1
η5 = with Et = ( np −1)
E cp 1 0.002 . n p  σ irp 
+ . 
E cp σ c 0 .2  σ c 0.2 
giving:
1911
σirp = ( np −1)
0.002 . E cp . n p  σ irp 
1+ . 
σ c 0.2  σ c 0.2 
or:
1911
σirp = 6 . 05
0.002 . 70300 . 7.05  σ irp 
1+ . 
270  270 
after numerical resolution, the following is obtained for the panel σirp = 289 MPa

Stiffener inter-rivet buckling:
er = 3 mm
countersunk head bolt => K = 0.66
p = 25 mm
Ecr = 73800 Mpa
nr = 13.83
Critical stress:
2
e 
Eq. 1.3.6.2 σirr = 1.89 . η5 . Ecr  r 
 p
2
3
σirr = η5 . 1.89 . 73800  
 25 
σirr = η5 . 2008 MPa
Plasticity correction:
Et 1
η5 = with Et = ( nr −1)
E cr 1 0.002 . n r  σ irr 
+ . 
E cr σ c 0.2  σ c 0 .2 
giving:
2008
σirr = ( nr −1)
0.002 . E cr . n r  σ irr 
1+ . 
σ c 0.2  σ c 0 .2 
or even:
2008
σirr = 12 .83
0.002 . 73800 . 13.83  σ irr 
1+ . 
420  270 
after numerical resolution, the following is obtained for the panel σirr = 412 MPa

V1-3 •7-CRIPPLING
Under a compression load, a stiffener may buckle locally (e.g.: local buckling of the flange).
However, after local buckling, this stiffener may still support an increase in the load before
failure occurs. The overload is supported by the stiffer areas not yet subjected to buckling.
Crippling is the failure of the stiffener under a compression load.
The crippling load is the sum of the load-carrying capacity of the elements comprising the
angle section.
The cross section area of the stiffener is divided into single elements:
1
E
S2, σcri2 S
S1, σcri1
S3, σcri3
2
S
3
E
N
The critical crippling stress shall be calculated as follows: T
ΣS i σ cri
σcrip = Eq. 1.3.7
ΣS i I
with:
σcri: critical crippling stress of item i
A
Si: area of the section of item i L
Crippling - Page V1-3•7/1 revision 1 © Copyright AEROSPATIALE 1998

Crippling - Page V1-3•7/2 revision 1 © Copyright AEROSPATIALE 1998

V1-3 •8-OPTIMISATION
Critical stresses in a correctly sized stiffener are very close one to the other and are greater
than the conventional allowable yield stress of the material.
V1-3 •8.1-SKIN SIDE FLANGE
The fastener pitch must be such that the inter-rivet buckling stress is high enough. Ideally, this
stress must be greater than or equal to the conventional allowable compression yield stress of
the stiffener. This gives:
p 1.38
• countersunk head bolts, rivets: ≤
e σ c 0.2
+ 0.002 n
Ec
p 1.70
• flat head bolts, tack welding: ≤
e σ c 0.2
+ 0.002 n
Ec
The maximum fastener pitch shall depend on the thickness of the skin under the stiffeners.
Then, care shall be taken to ensure that the thickness of the stiffener skin side flange is
sufficient, and that the pitch selected is compatible with the diameter of the fasteners:
p
3≤ ≤6
D
p
= 4.5 is a frequently used value ("sealed pitch")
D
Example:
p
• Skin made of 2024 PLT3, riveted stiffeners: ≤ 10.2 (E = 70300 MPa, σc0.2 = 270 MPa, n = 7.05)
e
If the skin thickness is 3 mm, the following is necessary: p ≤ 30.6 mm
p
• Stiffener made of 7075 T73510 ≤ 7.5 (E = 73800 MPa, σc0.2 = 420 MPa, n = 13.83)
e
If the thickness of the base is 3 mm, the following is necessary: p ≤ 22.5 mm
However, if p=30 mm, then it is necessary to have esr ≥ 4 mm
p
In the first case, the following fasteners shall be used: φ 4.8 ( = 4.7),
D
p
In the second case, the following fasteners: φ 6.35 ( = 4.7).
D
Optimisation - Page V1-3•8/1 revision 1 © Copyright AEROSPATIALE 1998

The skin side flange width shall be determined using the diameter of the selected fasteners
(minimum edge distance = 2D, minimum distance between the centreline of the fasteners and
the web).
2D mini
If the skin is subjected to diagonal tension, the pad thickness under the stiffener shall ideally
e
be 1.5 times the thickness of the adjacent pockets: t = 1.5
e
e et
V1-3 •8.2-WEB
h
Ideally, folding should not occur before σc0.2. This is the case if the ratio satisfies the
e
following conditions:
h 0.66
• Plate with a free unloaded edge: ≤
e σ c 0.2
+ 0.002
Ec
h 1.75
• Plate with two hinged unloaded edges: ≤
e σ c 0.2
+ 0.002
Ec
Example:
h
Stiffener made of 7075 T73510 (E = 73800 MPa, σc0.2 = 420 MPa, n = 13.83) with flange: ≤ 20
e
Generally, the height of the web is limited for space allocation reasons.

V1-3 •8.3-FLANGE
The local buckling condition is governed by the rule defined for the web. Ideally, it would also
be better if the warping stress was greater (or equal) to the allowable yield stress of the
material.
If the stiffener is provided with a single flange (Z stiffener), this last condition is practically
impossible to satisfy.
It is necessary, at least, for the flange to be a support for the web.
V1-3 •8.4-CRIPPLING
Imposes a supplementary condition on the skin side flange: refer to paragraph 8.2 (plate with
one unloaded free edge).
Example:
bs
Stiffener made of 7075 T73510: ≤ 7.5. This results in a skin side flange width equal to the
es
fastener pitch.


MCS V1-4 • THIN WEB BEAMS
CONTENTS
issue date revision
V1-4 THIN WEB BEAMS 0 01/1999 Original issue

V1-4 1 GENERAL 0 01/1999 Original issue
V1-4 2 PREPARATION OF BASIC DATA 0 01/1999 Original issue
V1-4 3 GENERAL PRINCIPLES FOR DIAGONAL 0 01/1999 Original issue
TENSION
V1-4 4 WEB FOLDING STRESS 0 01/1999 Original issue
V1-4 5 STUDY OF UPRIGHTS IN STANDARD ZONE 0 01/1999 Original issue
V1-4 6 STUDY OF FLANGES 0 01/1999 Original issue
V1-4 7 END SPANS AND SPANS WITH OPENINGS 0 01/1999 Original issue
V1-4 8 STUDY OF WEB 0 01/1999 Original issue
V1-4 9 STUDY OF FASTENERS 0 01/1999 Original issue
V1-4 10 EXAMPLE 0 01/1999 Original issue
V1-4 1 GENERAL 1–1

V1-4 1.1 DEFINITIONS 1–1
V1-4 1.1.1 Thin web beams 1–1
V1-4 1.1.2 Pocket and super-stiffener 1–2
V1-4 1.2 BEHAVIOUR OF A THIN WEB BEAM 1–3
V1-4 1.3 FAILURE MODES 1–4
V1-4 1.4 SCOPE, LIMITS 1–6
V1-4 2 PREPARATION OF BASIC DATA 2–1

V1-4 2.1 MATERIAL CHARACTERISTICS 2–1
V1-4 2.2 SECTION OF A SUPER-UPRIGHT IN STANDARD ZONE 2–1
V1-4 2.3 END SUPER-UPRIGHT SECTIONS 2–3
V1-4 2.4 POCKET DIMENSIONS, BOUNDARY CONDITIONS 2–4
V1-4 3 GENERAL PRINCIPLES FOR DIAGONAL

TENSION 3–1
V1-4 3.1 INCOMPLETE DIAGONAL TENSION THEORY 3–1
V1-4 3.2 LOADING RATIO, DIAGONAL TENSION FACTOR OF A
POCKET 3–1
V1-4 3.3 LOAD-CARRYING WIDTH 3–2
V1-4 4 WEB FOLDING STRESS 4–1

V1-4 5 STUDY OF UPRIGHTS IN STANDARD ZONE 5–1

V1-4 5.1 GENERAL 5–1
V1-4 5.2 LOAD-CARRYING WIDTHS AND SECTIONS 5–2
V1-4 5.3 STRESSES IN UPRIGHTS 5–3
V1-4 5.3.1 General method 5–3
V1-4 5.3.2 Fold angle 5–4
V1-4 5.3.3 Simplified approaches 5–4
V1-4 5.4 ALLOWABLE STRESSES IN UPRIGHTS 5–5
V1-4 5.4.1 column buckling (refer to V1-1.5.9.3) 5–5
V1-4 5.4.2 forced crippling 5–5
V1-4 5.4.3 natural local buckling 5–5
V1-4 6 STUDY OF FLANGES 6–1

V1-4 6.1 LOADS AND STRESSES IN FLANGES 6–1
V1-4 6.1.1 beam geometrical characteristics 6–1
V1-4 6.1.2 loads and stresses due to primary bending of beam 6–2
V1-4 6.1.3 additional loads and stresses due to the action of the tension field,
mean strain 6–2
V1-4 6.1.4 total normal loads and mean stresses in flanges 6–3
V1-4 6.1.5 secondary bending moment in flanges 6–3
V1-4 6.1.6 stresses due to secondary bending moment in flanges 6–4
V1-4 6.1.7 overall stresses in flanges 6–4
V1-4 6.1.8 allowable stresses and margins 6–5
V1-4 6.2 EXAMPLE 6–6
V1-4 6.3 CONCLUSION 6–10
V1-4 7 END SPANS AND SPANS WITH OPENINGS 7–1

V1-4 7.2 EDGE ELEMENTS 7–1
V1-4 7.4 ADDITIONAL STUDIES FOR OPENINGS IN A THIN WEB
BEAM 7–4
V1-4 8 STUDY OF WEB 8–1

V1-4 8.1 STRESSES IN WEB 8–1
V1-4 8.1.1 In pocket thicknesses 8–1
V1-4 8.1.2 At uprights 8–1
V1-4 8.2 WEB FAILURE 8–1
Issue 0 Contents page V1-4•ii

V1-4 9 STUDY OF FASTENERS 9–1

V1-4 9.1 WEB-FLANGE AND WEB-END UPRIGHT
CONNECTIONS 9–1
V1-4 9.2 STANDARD DISSYMMETRICAL WEB-UPRIGHT
CONNECTIONS (SINGLE ANGLE) 9–2
V1-4 9.3 UPRIGHT-WEB CONNECTIONS 9–2
V1-4 10 EXAMPLE 10–1
Issue 0 Contents page V1-4•iii

SYMBOLS
a: Distance between end rows of web-flange h: Beam height.

fasteners. h1: Beam height between fastener rows.
b,bi: Distances between end rows of web-upright h2: Beam height between flange CGs.
fasteners.
hm: Upright height between centre of rivets attaching
b c: Width of free flange of an angle. upright to flanges.
A: Distance between flanges. I: Inertia (index 0: initial inertia before folding).
B,Bi: Distances between uprights. Is: Flange inertia.
UL: Ultimate loads. Iss: Upper flange inertia.
LL: Limit loads. Isi: Lower flange inertia.
CS: Single shear. Iaa: Beam inertia with web.
Cfs: Attenuation factor. Isa: Beam inertia without web.
d: CG (centre of gravity) offset. IXXa: Inertia of web around an XX axis.
d a: Offset between web CG and total section CG. IXXm: Inertia of upright around an XX axis.
dm: Offset of upright CG and total section CG. k,ki: Diagonal tension factors.
Ks: Shear buckling coefficient.
D: Fastener diameter.
l: Distance between load introduction point and
e, ei: Web thickness on either side of an upright. section studied.
es: Member flange thickness. Le: Reduced effective length.
esi: Lower flange thickness. Lt,Lti: Load-carrying widths.
ess: Upper flange thickness. Lt0: Initial load-carrying width (before folding).
et: Land thickness. TD
Lt : Load-carrying width after folding.
e’: Equivalent thickness. M f: Bending moment.
E: Young's modulus (indexes "a": web, "m": upright). Mf1: Bending moment concerning beam resisting
shear.
Es: Secant modulus.
Mf2: Bending moment concerning beam resisting
Et: Tangent modulus. pure diagonal tension.
fb: Linear load inducing bending in edge elements. Mfs: Secondary bending moment.
fr: Load per length unit acting on the rivets. M y: Bending moment around y-axis.
fmax: Max shear flow. n: Work hardening coefficient (Ramberg and
F1: Mean axial load in a flange due to shear. Osgood).
F2: Mean axial load in a flange due to pure diagonal N: Normal load (index x or y => along x or y).
tension. p: Rivet pitch.
Fc: Compression load induced by shear load under Pm: Load transferred from base to upright.
action of the tension field. q: Length of an opening.
Fsim: Mean axial load in lower flange. Ri: Reactions at flanges.
Fssm: Mean axial load in upper flange. Rs: Loading ratio in single shear.
Fsit: Mean total axial load in lower flange. S: Section (index 0: initial section before folding).
Fsst: Mean total axial load in upper flange. Sa: Web section.
g: Distance between web centre plane and upright Ss: Flange section.
CG. Sss: Upper flange section.
G: Shear modulus (Coulomb).
Ssi: Lower flange section.
Gs: Secant shear modulus.
Issue 0 Contents page V1-4•iv

Saa: Section of beam with web. σ0.2: Conventional yield strength (indices "a": web,
"m": upright).
Ssa: Section of beam without web.
σR: Breaking tensile stress (indices "a": web, "m":
Sm: Section of upright alone.
upright).
Sequi: Equivalent section.
σ TD
x : Mean stress along a flange.
T: Shear load.
σTD
x min : Minimum stress along a flange.
Tcis: Shear load acting in form of pure shear.
σ TD
y : Mean stress along an upright.
TTD: Shear load acting in form of pure diagonal
σTD
y min : Minimum stress along an upright.
tension.
σa: Normal stress in web.
TD: Diagonal tension.
σxa: Normal stress along x-axis in web.
TDI: Incomplete diagonal tension.
σya: Normal stress along y-axis in web.
vi: Dimensions between neutral fibre of a flange
and its extreme fibres. σflf: Allowable forced crippling stress.
W: Static moment (index 0: static moment before σ fss : Primary bending stress at extreme fibre of
folding). 1 upper flange.
 e 4
Wd: Parameter= 0.7 * b *   σ fsi : Primary bending stress at extreme fibre of
 (I ss + I si ) * h 2  lower flange.
Wsem: Static moment around neutral fibre of flange.
σssm: Mean stress in upper flange due to primary
Wame: Static moment of web portion above neutral
bending.
fibre.
σsim: Mean stress in lower flange due to primary
WXXa: Static moment of web around XX axis.
bending.
WXXm: Static moment of upright around XX axis.
σssc: Compression stress in upper flange due to
yi: Dimensions between neutral fibre of beam tension field.
(with or without web) and extreme fibres of
σsic: Compression stress in lower flange due to
flanges (upper or lower).
tension field.
α: Fold angle in pockets.
σsst: Total upper flange mean stresses.
ε: Strain (index x or y => along x or y).
σsit: Total lower flange mean stresses.
εTD: Diagonal tension strain.
σfsi: Secondary bending stresses in extreme fibres
Φy: Transverse flow. (upper or lower) of flanges (upper or lower).
η: Plasticity correction coefficient. σssTi: Overall stress in lower fibre of upper flange.
λ: Slenderness ratio. σssTs: Overall stress in upper fibre of upper flange.
νe: Poisson ratio – elastic. σsiTi: Overall stress in lower fibre of lower flange.
νp: Poisson ratio – plastic. σsiTs: Overall stress in upper fibre of lower flange.
ν: Poisson ratio – elasto-plastic. τ,τi: Shear stresses.
ρ: Radius of gyration of a section. τcr: Critical stress.
σ: Normal stress (index x or y => along x or y). τcr,a: Critical stress in web.
σ0: Allowable compression stress with zero
slenderness ratio.
Issue 0 Contents page V1-4•v

BIBLIOGRAPHY
1- KUHN & PETERSON & LEVIN
METHOD OF ANALYSIS : PLANE - WEB SYSTEMS.
NACA T.N 2661 , MAI 1952.
2- BRUHN
Issue 0 Contents page V1-4•vi

V1-4 1 GENERAL
The aeronautical structure designer is constantly searching for types of structure or calculation methods enabling a gain
in weight whilst meeting manufacturing and cost constraints. A good example is that of thin beams where the buckling
of the web is accepted, the shear loads being balanced by the web diagonal tensile stresses.
The rules for calculating these beams are derived from the principles established for the stiffened panels in chapter
V1-1.
V1-4 1.1 DEFINITIONS
V1-4 1.1.1 Thin web beams
We call "thin web beams" beams where the webs are subjected to diagonal tension (*). In general, they are
distinguished by:
- concentrated longitudinal members called flanges which, practically alone, transmit all normal (tensile or
compression) and bending loads,
- a thin web the only practical role of which is to ensure the transmission of the shear load by diagonal tension
and shear which induces a compression overload in the flanges and uprights,
- transverse members called uprights the function of which is to provide supports, more or less regularly spaced
and more or less rigid, for the flanges and the web. The uprights balance the diagonal tension.
This definition establishing the respective roles of the web and the flanges concerning load transmission in reality
comprises only an "ideal limit". Indeed, we know, from chapter V1-1, that all thin webs participate in the transmission
of the normal loads until its stability limit is reached. We also know that beyond this limit a thin web still participates
but to a lower extent evidenced by the "load-carrying" width notion.
Also, after folding, and whilst continuing to transmit the shear load, the thin web is subjected to strains higher than
those defined by the linear elasticity theory. Indeed, the efficient transverse modulus of elasticity GTDI, representing the
rigidity of the skin to shear in the presence of diagonal tension, decreases as the load increases (refer to § V1-1.5.10).
Thin web beams are significantly more flexible than stable web beams. This marks the limit of their applications.
We learnt, in chapter V1-1, how to evaluate the "additional loads" consecutive to the instability of stiffened panels.
These theories apply generally to all the beams studied below for which we admit the buckling of the web.
The interest of these thin web beams lies in their lightness and flexibility. The stiffer beams where the buckling of the
web is not tolerated are covered by chapter V1-5.
Flange
Web
Uprights
Flange
(*): The diagonal tension theory is described in chapter V1-1.5
Issue 0 General page V1-4•1–1

V1-4 1.1.2 Pocket and super-stiffener
(Refer to §V1-1.1.1 )
The definition of the thin web beams implies that the beam is criss-crossed by longitudinal and transverse members.
These are attached to the web either because they are "sewn" by means of screws, rivets or tack welding ("added
flanges or uprights"), or because the assembly is machined from a plate ("integrated flanges or uprights").
We call:
- "pockets": the web quadrilaterals delimited by the members.
- "super-stiffener": the association of an upright or a flange with an associated web portion (half-pockets on
either side).
Members
Pocket thickness
Land
B B
Uprights
Issue 0 Contents page V1-4•1–2

V1-4 1.2 BEHAVIOUR OF A THIN WEB BEAM
x
T
y
Mf
a a
The external loads which can be applied to a thin web beam are:
• T: Shear load
• Mf: Bending moment
Induced effects:
These loads mainly generate in the beam normal internal loads in the flanges and the uprights and a shear and diagonal
tension flow in the web of the beam.
Below, we show a section of the thin web beam at an upright, representing it associated with a load-carrying web
portion. The internal loads applied to this section are:
Nx: Normal load in the upright and its load-carrying web portion, induced by diagonal tension.
Φy: The normal transverse flow due to diagonal tension in the web.
The stresses resulting from these loads are:
Normal (σx) in the upright. Normal (σx), transverse (σy) and tangential (τxy) in the web.
Section a-a Upright
z Nx
Φy Φy
G y Web
Remark: In certain specific cases, secondary induced loads are added to the loads applied above (e.g.: transfer of
neutral fibre to an upright runout induces a bending moment My in the upright).

The calculating difficulties come from the fact that the behaviour of the beam depends on the loading intensity.
• Tension, the static strength limit at ultimate load is given by the characteristics of the material.
• Shear, the same applies even though the folds occur beyond the critical shear of the pockets. These folds are oblique
to the edges. For this reason, the term "diagonal tension" is used even though inappropriate.
• This property of dependence is further emphasised in the case of compression. It is related to the instability of the
thin web.
V1-4 1.3 FAILURE MODES
There are several types of instability and failure.
• Web instability:
Thin sheets fold under relatively low compression and/or shear loads.
On the compressed flange side, the web is saturated when its buckling stress is reached. At greater values, the excess
load is picked up by the flange and the load-carrying portion of the web.
In shear cases, after the appearance of the folds, the web transfers the excess load partly in the form of shear stresses
and the remainder as diagonal tension. The tangential flows at pocket edges are the same at constant loads with or
without the folds. However, the diagonal tension induces flows normal to these edges which causes overloads,
especially in the flanges and uprights.
The instability of the web does not lead to the failure of the beam and is therefore not a cause for limiting the
ultimate load.
• Instability of super-stiffeners
If the longitudinal or transverse members, spaced respectively at a pitch of A or B, are sufficiently rigid, the super-
stiffeners (upright or flange + load-carrying web) behave as columns of length A or B on single supports.
Unstable failure may occur in the most critical super-stiffener under:
- local buckling,
- column buckling,
- forced crippling of the flanges of the members (inter-rivet buckling does not exist when the web is subject to
diagonal tension).
• Failure of the skin

Generally, the web is thicker at the members (uprights, flanges) than in the pockets. Failure can then occur, either in
the pocket-land blend-in radii or in the web-members fastener rows.
It must be checked that the maximum shear in the web does not exceed half the material tensile breaking stress
(TRESCA criterion).
• Failure of fasteners
Under the combined effect of shear and diagonal tension in the web.
Also, the secondary effects must be taken into account:

- Bending of the flanges due to diagonal tension,
- Secondary bending of the dissymmetrical uprights.

Bent flange (1) Row of fasteners (4)

Section a-a Section b-b
Opening in web (2) Upright run-out (3)
Section c-c
(1) (3) (4)
Secondary bending of Secondary bending of the Fasteners
the flanges due to dissymmetrical uprights
M1 diagonal tension. M1 Section b-b
flange F1
web R
F2
M2 F1 due to the shear flux in the web.

F2 due to the transverse stress in the web
generated by resulting diagonal tension
R.
Secondary bending of the uprights
N
N TD
upright
Web stresses
d
-N
- Nd
N TD
upright
Equilibrium:
N TD TD
x web + N x upright = 0
d N TD
x upright + M f upright = 0

(1): Bent flange

Each flange is forced towards the inside of the web by a load due to the diagonal tension. The web behaves as a
beam clamped at each end at the uprights.
(2): Calculating holes and openings

The presence in a thin web beam of flanged holes or openings of any sort creates overstresses in the sheet around
the holes whereas secondary bending loads appear in the flanges.
(3): Upright run-out

The run-out of an upright overloads the upright by a secondary moment due to the transfer of the neutral fibre
between the web and the centre of gravity of the upright.
(4): Fasteners
After folding, the web pulls on its edges, hence additional loads perpendicular to the alignment of the fasteners are
to be taken into account.
(5): Web
The bending of the flanges leads to a variation in the tension field in the pockets. Their loads are increased in the
vicinity of the flange and upright intersections.
V1-4 1.4 SCOPE, LIMITS
SCOPE
In addition to the thin web beam calculation, this chapter deals with certain specific cases relevant to stiffened panels
which were not studied in chapter V1-1:
- bending of flanges due to diagonal tension or bending of a stiffener bordering on an opening in a stiffened
panel,
- influence of holes in the web or in the skin of a stiffened panel,
- effect induced by a stiffener run-out (secondary bending),
- calculation of the fasteners attaching the web to the flanges or attaching the stiffened panels together,
- sizing of the fasteners attaching the web to the uprights or attaching the stiffeners and the skin of a stiffened
panel.
LIMITS
es
≥ 0.6 (the flange of the members on the skin side must not be too thin)
e
B A
0.2 ≤ ≤ 1; 120 ≤ ≤ 1500 (the proportions of the pockets must be within these ranges)
A e
p
2.5 ≤ ≤ 10 (riveting)
D
Caution: risk of non-conservative predictions with thick pockets (outside range).
The uprights are more or less equally spaced and thicknesses of adjacent pockets are similar.

V1-4 2 PREPARATION OF BASIC DATA
V1-4 2.1 MATERIAL CHARACTERISTICS
The elasto-plastic behaviour of the materials must be taken into account in these calculations. To this end, the Ramberg
and Osgood model (refer to V2-2 "Material behaviour") is used:
n
σ  σ 
ε= + 0.002  
E  σ0.2 
σ
Es =
ε
1 n 1− n
= +
E t Es E
Therefore, for each material used, the following is required:

• Young's modulus of compression, E,
• Compression yield strength, σ0.2
• R.&.O. coefficient, n.
V1-4 2.2 SECTION OF A SUPER-UPRIGHT IN STANDARD ZONE
SECTION (INTEGRATED UPRIGHT)
Lt1 Lt2
e2
d
G X
e1 et
Z
Initial section: Lti = Bi/2
B1 B2
CHARACTERISTICS:
Cross section area: S
Moment of intertia/GX: I
C.G. offset: d
Pocket thicknesses: e1, e2
Issue 0 Preparation Of Basic Data page V1-4•2–1

SECTION (ADDED UPRIGHT):
Lt1 Lt2
e2
X da X
e1 d
G dm
Z
Initial section: Lti = bi/2
b1 b2
b1 and b2 are the distances between outer fastener rows.
WEB CHARACTERISTICS MEMBER CHARACTERISTICS (beam upright)
Cross section area: Sa Sm

Moment of inertia/XX: IXXa IXXm
Static moment/XX: WXXa WXXm
Pocket thicknesses: e1, e2
W WXXm
C.G. offset: da = XXa da =
Sa Sm
Remark: WXXa is negative whereas WXXm is positive.
CHARACTERISTICS OF A SUPER-STIFFENER MADE SYMMETRICAL:

If a super-stiffener is not symmetrical in relation to the (G, Y, Z) plane, an equivalent symmetrical super-stiffener must
be constructed.
Lt1. Lt2.
e2
X da X
e1 d
G dm
Index 0 is relevant to the initial section.

Z
e e e Lt + e 2 Lt 2.0 e τ + e 2τ2
e = 1 2 ; Lt0 = 1 1.0 :τ= 1 1
2 e1 + e 2 e1 + e 2
2.3
k1 + k 2 1 + k 
Diagonal tension factor: k = ; Loading ratio: Rs =  
2 1 − k 

V1-4 2.3 END SUPER-UPRIGHT SECTIONS
SECTION (INTEGRATED UPRIGHT)
B
Lt
d e X
G
End stiffener Z
Initial section: Lt = B/2
CHARACTERISTICS:
Cross section area: S

Moment of inertia/GX: I
C.G. offset: d
Load-carrying width: Lt
Pocket thickness: e
SECTION (ADDED UPRIGHT):
Initial section: Lt = b/2

B
Lt
et
da
X X
e d
dm G
b: is the distance between outer fastener rows.
WEB CHARACTERISTICS MEMBER CHARACTERISTICS (beam upright)
Cross section area: Sa Sm

Moment of inertia/XX: IXXa IXXm
Static moment/XX: WXXa WXXm
Load-carrying width: Lt
Pocket thickness: e
W WXXm
C.G. offset: da = XXa dm =
Sa Sm
Remark: WXXa is negative whereas WXXm is positive.
Issue 0 Preparation Of Basic Data page V1-4•2–3

V1-4 2.4 POCKET DIMENSIONS, BOUNDARY CONDITIONS
Use the conditions given in § V1-1.3 (pocket folding stresses in stiffened panels):
If the members have opened profiles, the edge conditions to be chosen for the pocket thickness buckling stress
calculation are: four hinged edges.
b'
(distance between fasteners)
et
e
b"
Width of pocket "b" is:

b = B for integrated stiffeners.
b = b’ for added stiffeners.
b = b’’ for all types, if e t ≥ 3e
Height of pocket "a" is:

a = A, a’, a’’ (same principle as above)

V1-4 3 GENERAL PRINCIPLES FOR DIAGONAL TENSION
V1-4 3.1 INCOMPLETE DIAGONAL TENSION THEORY
This theory is detailed in §V 1-1.5.1
Let us recall the main principle:
In a normal thin web beam, the stress condition in the web is between the pure shear stress condition and the pure
diagonal tension condition. This theory is based on the assumption that the total shear load in the web can be broken
down into two parts: T = Tcis + TTD
- a pure shear part Tcis = (1 - k)T (where k is the diagonal tension factor)
- a pure diagonal tension part TTD = kT.
Web not buckled: pure shear Web subjected to pure diagonal tension
τ - (1 - k)τ σy = kτ tan α σ2 = - (1 - k) τ sin2α

τ
TDI TD + CS TDI or TDI τ12 = (1 - k) τ cos2α
α (1 - k)τ
2 kτ 45° kτ α 2 kτ
σx = σ1 = + (1 − k ) τ sin 2α
sin 2α tan α sin α
In base (x,y) Facette along theoretical plies
V1-4 3.2 LOADING RATIO, DIAGONAL TENSION FACTOR OF A POCKET
The loading ratio Rs and the diagonal tension factor k are defined by:
τ
Rs = (τcr,a: critical shear in web)
τ cr , a
k = 0 if Rs < 1
R s0.4343 − 1
k= if Rs > 1
R s0.4343 + 1
Dissymmetrical pockets: refer to V1-4.2.2
Issue 0 General Principles For Diagonal Tension page V1-4•3–1

V1-4 3.3 LOAD-CARRYING WIDTH
UPRIGHT ORIENTATION
σya
LtTD LtTD
σp e
Lt0 Lt0
Lt0
Lt TD = (1 − k) Index 0 is relevant to the initial section (before folding)
2
b
Lt 0 =
2
FLANGE ORIENTATION
σx
a
Lt0
σp e
LtTD
TD Lt0
Lt = (1 − k) Index 0 is relevant to the initial section (before folding)
2
a
Lt 0 =
2

V1-4 4 WEB FOLDING STRESS
The folding stress of the web between two uprights is calculated according to the rules given in chapter V1-2
(Buckling of plates and thin shells):
2
π2  e 
τcr,a = η k s E   (Refer to § V1.2.2)
 min {a , b}
2
12 (1 − ν e )
ks: theoretical buckling coefficient for a panel the edges of which simply supported.
b: width of sheet between upright-web fastener rows.
a: height of web between flange-web fastener rows.
e: web thickness.
η: plasticity correction coefficient (Refer to § V1-2.1.3 and § V1-2.2.2)
Hinged edges:
2
  a b 
ks = 3.8  min  ;   + 5.35
  b a 
G  1 + νe  Es
η= s =  
G 1+ ν  E
a e
The dimensions of the pockets a, b and the boundary conditions to be taken into account are defined in § V1-4.2.4.
Issue 0 Web Folding Stress page V1-4•4–1


V1-4 5 STUDY OF UPRIGHTS IN STANDARD ZONE
V1-4 5.1 GENERAL
a a b b
A e
Symmetrical Dissymmetrical
upright upright
Web Web
Section a-a Section b-b
Two types of uprights can be identified on thin web beams:

• symmetrical uprights formed of two identical sections on either side of the web (section a-a),
• single upright on one side of the web (section b-b).
Calculation of standard uprights is almost identical to that of the stiffeners of flat stiffened panels subjected to shear
(Refer to § V1-1.5 ).
Issue 0 Study Of Uprights In Standard Zone page V1-4•5–1

V1-4 5.2 LOAD-CARRYING WIDTHS AND SECTIONS

σya Upright Upright Upright
LtTD LtTD
σa Flange
e α
X X
Lt0 Lt0
y
x
Flange
Z
After the section of an upright has been made symmetrical in relation to the Z-axis (if necessary) as described in
§.V1-4.2.2, the widths, sections and other load-carrying geometrical characteristics are given by the following
formulas:
Index 0 is relevant to the initial section, that is before folding.
Lt 0
• Load-carrying width LtTD = (1 - k)
2
TD
• Load-carrying section let S0 be the initial total section and ∆S the non load-carrying portion of the web (not
shaded)
∆STD = 2 (Lt0 - LtTD) e = (1 + k) Lt0e
STD = S0 - ∆STD
• INTEGRATED UPRIGHT
 e
∆WTD =  d 0 −  ∆STD
 2
W = W0 - ∆WTD
TD
∆W TD
dTD = d0 +
STD
 e
2
e2 
∆ITD =   d 0 −  +  ∆STD
 2 12 

I = I0 - ∆I - (d0 - dTD)2 STD
TD TD
• ADDED UPRIGHT
CHARACTERISTICS OF BEAM WEB
TD  e
WXXa = WXXa 0 +  e t −  ∆STD
 2 
 e
2
e 2  TD
I TD 
XXa = I XXa 0 −  e t −  +  ∆S
 2 12 

CHARACTERISTICS OF CORRECTED SECTION (DIFFERENT MATERIALS)
STD Sm
E = aTD E a + TD Em
S S
TD E TD E
WXX = a WXXa + m WXXm
E E
TD
W
dTD = TD XX
S
E a TD E
I TD
XX = I XXa + m I XXm
E E
2
ITD = I TD
XX − S
TD
d TD

V1-4 5.3 STRESSES IN UPRIGHTS
V1-4 5.3.1 General method
The stress in an upright is not uniform:

- Single uprights, dissymmetrical in relation to the web mean plane are the seat of secondary bending (Refer to
§ 1.3).
- The stresses vary along the upright from a minimum at the beam flanges to a maximum in the centre of the
upright ("gusset effect") (fig “b“).
σ TD
y
a a
σy σ TD
y min
Neutral axis σy
σ TD
Gusset effect y
σ TD
y Section a-a
figure "a" figure "b"
SECONDARY BENDING
Due to the secondary bending caused by loading outside of the neutral fibre, compression stress is maximum at beam
web (fig "a"). The "mean" maximum compression stress (variation of σ along y is not taken into account here) is
determined as if it were a single compression by dividing the compression load by a "equivalent section", STD
equi :
N TD
y
σ TD
y = where:
STD
equi
STD
N TD
y = - 2 k Lt0 e τ tanα and STD
equi = 2
g
1 +  
ρ
e
Integrated upright: g = dTD -
2
e
Added upright: g = dTD + et -
2
I TD
ρ=
STD
α is the fold angle (See next page)

TD
Remark: If the upright is symmetrical, then g = 0 and STD
equi = S
GUSSET EFFECT
The maximum compression is given by the following equation:
σ TD
y min 
TD
b  σ y min σ TD
y min
= (1 − k )  1 . 78 − 0. 64  ; ≥ 1 ; take 1 if <1
σyTD
 a  σy TD
σ TD
y
a: height of the upright between the centres of the rivets attaching the upright to the flanges.

V1-4 5.3.2 Fold angle
The rules for determining the α's described in § V1-1.5.6 (flat stiffened panels - stresses in the super-stiffener) are
applied as follows:
ε − ε TD
tan2α = TD x
ε TD − ε TD
y
 2k  τ
The elongation of the web along the folds is equal to: ε TD =  + (1 − k )(1 + ν ) sin 2α 
 sin 2α  Ea
The mean compression strains induced by the diagonal tension in the upright ( ε TD TD
y ) and the flanges ( ε x ):
σ TD
y
ε TD
y = Calculation of ε TD
x is described in § 6 (Study of flanges).
Ey
V1-4 5.3.3 Simplified approaches
α is equal to 1.
As the fold angle is around 45°, we suppose that sin 2α
τ
Hence: ε TD = [(1 + k ) + ν(1 − k )]
Ea
ASSUMPTION 1: If the stiffening ratio and the materials of the uprights are similar to those of the longitudinal
members (the flanges), we consider that the folds are 45° irrespective of k.
ASSUMPTION 2: If the longitudinal members (the flanges) can be considered as infinitely rigid, determine
σ TD
y and τ using the graph below.
Remark: This assumption is not conservative for the upright as the folds tend to bend over onto the more rigid
members.
STD
equi
Place k on the x-axis and − 1 on the y-axis. If necessary, interpolate to obtain sought value.
eb
STD
equi is the initial geometrical section (∆STD = 0).
Analysis of diagonal tension: σ TD ( ) TD

(
y / τ versus k and Sequi − be / be )
1.0
σ/τ = 0.05 0.10 0.25 0.30 0.45 0.50 0.70 0.80
0.9 0.15 0.20 0.35 0.40 0.55 0.60 0.90
0.8 1.00
1.10
0.7
1.20
0.6
1.40
0.5
1.60
0.4 1.80
2.00
0.3
2.50
0.2 3.00
4.00
0.1 5.00
0.0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
k

V1-4 5.4 ALLOWABLE STRESSES IN UPRIGHTS
3 types of failures can be encountered:

1. Column buckling.
2. Forced crippling.
3. Natural local buckling.
V1-4 5.4.1 column buckling (refer to V1-1.5.9.3)
• Resistance to column buckling is checked, in principle, by comparing the mean stress at the centre of gravity of the
super-upright ( σ yTD ) and the critical buckling stress.
• When upright bending is initiated, the diagonal tension tends to oppose it. This interface action is taken into account
by use of the clamping coefficient below:
1
LTD = KTDA; KTD =
 Lt 
2
1 + k 3 − 4 0 
 A 
TD TD TD
Then calculate σ TD
y crit from (S ; I ; L ) using the rules given in § V1-1.4
(the simplest method consists in using the Johnson formula).
• According to the slenderness ratio of the super-upright, the column may buckle under a two half-wave mode
(especially dissymmetrical uprights).
In addition, it must be checked that the maximum compression stress at the centre of gravity of the super-upright is
LTD
lower than the buckling stress calculated with a buckling length equal to :
2
 σ y min  TD
  σ y ≤ σ  LTD 
 σ y  y crit  
 2 
 
V1-4 5.4.2 forced crippling
Ripples in the web favour the buckling of the flange of the upright attached to the web especially if the flange of the
upright is thinner than the web.
The allowable stress is calculated by means of the rules established for the stiffeners (refer to V1-3.5):
1
2
σ c 0.2 m  e' E 3
σ flf = − 0.051 k  sm cm
3  (indexes => c: compression; m: upright; a: web; s: flange)

σ c 0.2 m
+ 0.002  e a E ca 
E cm
e'sm is the equivalent thickness of the flange.
On account of the local nature of this failure, stress σ TD TD
y min (and not σ y ) must not exceed σflf.
V1-4 5.4.3 natural local buckling
To avoid natural local buckling, the stresses in the upright must be lower than the local buckling stresses calculated by
means of the rules in chapter V1-3.
As added dissymmetrical uprights are the seat of secondary bending, they are practically limited only by the Forced
crippling.
The stresses to be compared with the allowable local buckling are those at the centre of the uprights (on account of the
gusset effect).


V1-4 6 STUDY OF FLANGES
V1-4 6.1 LOADS AND STRESSES IN FLANGES
The loads in the flanges are:
1) The normal loads due to the primary bending of the beam calculated by the elementary bending theory.
2) The additional loads induced by the tension field in the web.
3) The secondary bending loads generated by the tension field (figure "a").
T
figure "a"
V1-4 6.1.1 beam geometrical characteristics
l T
A
e b h h1 h2
v1
v3 esi Y3
v2 ess y1
neutral axis
v4
neutral axis
Sss = upper flange section.
Ssi = lower flange section.
Iss = upper flange inertia. y2 Y4
Isi = lower flange inertia.
v1 = distance between upper extreme fibre of
upper flange and neutral axis.
v2 = distance between lower extreme fibre of
Beam section (A-A) with web Beam section (A-A) neglecting web
upper flange and neutral axis.
Saa = Beam section with web. Ssa = Beam section without web.
v3 = distance between upper extreme fibre of
Iaa = Beam inertia with web. Isa = Beam inertia without web.
lower flange and neutral axis.
v4 = distance between lower extreme fibre of y1 = distance between extreme fibre of y3 = distance between extreme fibre of
lower flange and neutral axis. upper flange and neutral axis. upper flange and neutral axis.
ess = upper flange thickness. y2 = distance between extreme fibre of y4 = distance between extreme fibre of
esi = lower flange thickness. lower flange and neutral axis. lower flange and neutral axis.
Issue 0 Study Of Flanges page V1-4•6–1

V1-4 6.1.2 loads and stresses due to primary bending of beam
1) Bending moment
Introducing a load T at the end of a thin web beam implies the existence of a bending moment changing along the
length of the beam. The architecture of each beam span must be taken into account to evaluate the exact location
where the effect of the bending moment on the flanges is the most significant. Thus, section A-A of the thin web
beam shown on the previous page is the section farthest from the load T introduction point before the end fittings
start to take the load off the flanges.
According to the incomplete diagonal tension theory, a part of the total shear load (and therefore of the bending
moment) acts in the form of pure shear, the other part in the form of pure diagonal tension.
Let Mf be the total bending moment at the section under study: Mf = T * l = Mf1 + Mf2
Mf1 the bending moment concerning the beam stable in shear.
Mf2 the bending moment concerning the beam in the form of pure diagonal tension.
Mf1 = (1 - k) Mf and Mf2 = k Mf where k: diagonal tension factor

2) Bending stresses on extreme fibres
• The bending stress on the extreme fibre of the upper flange is:
M y M y
σfss = − f 1 1 − f 2 3 T > 0 implies that the upper flange is compressed and Mf > 0
Iaa Isa
• The bending stress on the extreme fibre of the lower flange is:
Mf 1y2 M y
σfsi = + f2 4
Iaa Isa
3) Normal loads in the flanges due to bending
Whilst the web is stable in shear, the normal load in a flange can be expressed by:
I  M
F1 =  sa  ∗ f 1 (where h2 is the distance between the centres of gravity of each flange)
 I aa  h2
Mf 2
In the pure diagonal tension state, the normal load in the flange can be expressed by: F2 =
h2
Total normal loads in the flanges:
Upper flange: Fssm = - F1 - F2 Lower flange: Fsim = F1 + F2.
4) Mean stresses due to primary bending in flanges

F F
Upper flange: σssm = ssm Lower flange: σsim = sim
Sss Ssi
V1-4 6.1.3 additional loads and stresses due to the action of the tension field, mean strain
The flanges are submitted to a compression load induced by the diagonal tension field:
T
Fc = - TD * cot α (Fc < 0) where:
2
TTD = kT and α is the diagonal tension angle measured in relation to the flanges.
The stresses generated are:
Fc Fc
Upper flange: σ ssc = Lower flange: σsic =
Sss Ssi
1  σ ssc σ 
The mean strain to be taken into account to calculate angle α is: ε TD
x =  + sic 
2  E ss E si 

V1-4 6.1.4 total normal loads and mean stresses in flanges
Fsst
Upper flange: total load Fsst = Fssm + Fc mean stress σ sst = = σ ssm + σ ssc
Sss
Fsit
Lower flange: total load Fsit = Fsim + Fc mean stress σ sit = = σ sim + σ ssc
Ssi
V1-4 6.1.5 secondary bending moment in flanges
As the web is subjected to diagonal tension, it pulls on the elements comprising the flanges; in other terms, each flange
acts as a continuous beam with the uprights as supports. The transverse load on a flange is equal to the vertical flow in
the web due to the diagonal tension.
1
The maximum secondary bending moment in the flanges is: Mfs = k τ e b2 Cfs
12
1
 e 4
Cfs: attenuation factor given on the curve below versus parameter: Wd = 0.7 b  

 (I ss + Isi ) h 2 
Remark 1: The secondary bending moment is maximum at the uprights.
If k ≈ Cfs ≈ 1, the moment in the centre between the bearing points is equal to half the maximum moment
with opposite sign.
TWsem  2 Wame 
Remark 2: τ = 1 +  in all cases.
Ie  3 Wsem 
Wsem = static moment of the flange.
Wame = static moment of the efficient section of the web above the neutral fibre (for an upper flange).
T
τ = Sufficient approximation when flange height is low compared with the height of the beam
h 2e
and the flanges consist of thin sections.
for calculating I and W, efficient web thickness is taken as being equal to (1-k) e
Cfs versus Wd
1
0.95
0.9
Cfs
0.85
0.8
0.75
0 1 2 3 4
Wd
An approximation of Cfs versus Wd is given by the following polynomial:
Cfs = 0.0058 Wd3 - 0.0676 Wd2 + 0.1252Wd + 0.9366

V1-4 6.1.6 stresses due to secondary bending moment in flanges
UPPER FLANGE
at uprights:
M v
σ fs 2 = − fs 2 stress in lower fibre of upper flange
I ss
M fs v1
σ fs1 = stress in upper fibre of upper flange
Iss
When the upper flange bends on the bearing point, the lower fibre is compressed and the upper fibre stretched.
in the centre between two uprights:
In this case, M’fs is calculated in a conservative manner considering k = Cfs = 1
M 'fs v 2
σ′fs 2 = stress in lower fibre of upper flange
2 * I ss
M 'fs v1
σ′fs1 = − stress in upper fibre of upper flange
2 * I ss
LOWER FLANGE
at uprights:
M fs v 4
σ fs 4 = stress in lower fibre of lower flange
I si
M fs v 3
σ fs 3 = − stress in upper fibre of lower flange
I si
When the lower flange bends on the bearing point, the lower fibre is stretched and the upper fibre compressed.
in the centre between two uprights: (same remark as made above for M’fs)
M 'fs v 4
σ′fs 4 = − stress in lower fibre of lower flange
2 * I si
M 'fs v 3
σ′fs 3 = stress in upper fibre of lower flange
2 * I si
V1-4 6.1.7 overall stresses in flanges
UPPER FLANGE
σssTi = σssm + σssc + σfs2 overall stress, lower fibre of upper flange at uprights
σssTs = σfss + σssc + σfs1 overall stress, upper fibre of upper flange at uprights
σssTi = σssm + σssc + σ'fs2 overall stress, upper flange lower fibre in centre between two uprights
σssTs = σfss + σssc + σ'fs1 overall stress, upper flange upper fibre in centre between two uprights
LOWER FLANGE
σsiTi = σfsi + σsic + σfs4 overall stress, lower fibre of lower flange at uprights
σsiTs = σsim + σsic + σfs3 overall stress, upper fibre of lower flange at uprights
σsiTi = σfsi + σsic + σ'fs4 overall stress, lower flange lower fibre in centre between two uprights
σsiTs = σsim + σsic + σ'fs3 overall stress, lower flange upper fibre in centre between two uprights

V1-4 6.1.8 allowable stresses and margins
The allowable stresses required to calculate the margins are as follows:
* Tensile: Strength of flanges at ultimate loads is given by the characteristics of the material.
* Compression: The instability of the flanges can be caused by:
1) under a low wavelength mode:
a) Local buckling (of the flange for example) of the flanges.
b) Forced crippling of the riveted section of the flanges.
2) under a medium wavelength mode: (this concerns the complete flange)
a) Lateral buckling
b) Column
c) Crippling
The details for calculating the allowable stresses are given in chapter V1-3.
The margins are calculated by applying the following rule:

• The compression stresses calculated taking the secondary effects into account (secondary bending or gusset effect for
example) are compared with the allowable local buckling stresses (natural or forced).
• The mean compression stresses in the flanges are compared with the allowable lateral buckling, column and crippling
stresses.
• The tensile stresses are compared with tensile breaking stress of the material.

V1-4 6.2 EXAMPLE
Let us take the cantilever beam of § 6.1.1 submitted at its free end to a load of 60000 N and with the following
geometrical characteristics:
l = 1270 mm b = 290 mm e = 0,6 mm h = 762 mm h1 = 725,4 mm h2 = 746,4 mm
1270 60 000 N
A
0,6 290 762 725,4 746,4
v1
v3 esi Y3
v2 ess y1
neutral axis
v4
neutral axis
Sss = 430 mm2

Ssi = 245 mm2 y2 Y4
Iss = 57 250 mm4
Isi = 14 780 mm4
v1 = 9,7 mm
v2 = 30,3 mm
v3 = 24,1 mm Beam section (A-A) with web Beam section (A-A) web neglected
v4 = 5,9 mm Saa = 1121 mm2 Ssa = 671 mm2
ess = 4,5 mm Iaa = 110,1*106 mm4 Isa = 86,1*106 mm4
esi = 2,5 mm y1 = 319,4 mm y3 = 277,8 mm
y2 = 442,6 mm y4 = 484,2 mm
The mechanical characteristics of the web of the 2024 PLT3 beam are:
Ea = 70300 MPa; σRa = 440 MPa; σ0,2a = 270 MPa
The mechanical characteristics of the members (flanges and uprights) of the 7075 T73510 beam are:
Em = 73800 MPa; σRm = 495 MPa; σ0.2m = 420 MPa
The two materials have the same Poisson ratio: νe = 0.33
LIMITATIONS
First of all, we must check if the beam is compatible with the limitations of § 3.2
e 4 .5 e si 2.5
• ss = = 7.5 > 0.6 and = = 4.2 > 0.6
e 0.6 e 0.6
b 290 h 762
• 0.2 < = = 0.38 < 1 and 120 < = = 1270 < 1500
h 762 e 0.6

CRITICAL SHEAR STRESS

According to § 4, the critical shear stress is given by the following formula:
2 2
π2  e   h b 
τcr, a = η k s E a   with ks = 3.8 * min  1 ;  + 5.35
(
12 * 1 − ν 2e )  min {h 1 , b}    b h 1 
π2   290 
2
  0 .6 
2
τcr, a = η * 3 . 8 *   + 5 . 35 * 70300 *   = η * 1.7 MPa

(
12 * 1 − 0.332  )  725.4    290 
Obviously, there is no need to calculate the plasticity correction.

τcr, a = 1.7 MPa
LOADING RATIO AND DIAGONAL TENSION FACTOR

The loading ratio Rs and the diagonal tension factor k are defined in § 3.3
τ R 0.4343 − 1
Rs = and k = s0.4343
τ cr , a Rs +1
The flanges consist of tee sections the heights of which are low when compared with the beam.
Hence, according to § 5.1.5:
T 60000
τ = = = 134 MPa
h2 e 746.4 * 0.6
τ 134 790.4343 − 1
Rs = = = 79 and k = = 0.74
τcra 1 .7 790.4343 + 1
BENDING MOMENTS
The maximum bending moment is located at section A-A (refer to § 6.1.2).
Mf = T * l = 60000 * 1270 = 76.2 * 106 mm.N
Mf = Mf1 + Mf2
Mf1 is the bending moment concerning the beam resisting shear.
Mf2 is the bending moment concerning the beam in the form of pure diagonal tension.
Mf1 = (1 - k) Mf = (1 - 0.74) * 76.2 * 106 = 19.8*106 mm.N
Mf2 = k Mf = 0.74 * 76.2*106 = 56.4*106 mm.N
BENDING STRESSES ON THE EXTREME FIBRES OF THE FLANGES

Upper flange
The direction of the load T implies that the upper flange is compressed:
M * y1 M * y3 19.8 * 106 * 319.4 56.4 * 106 * 277.8
σ fss = − f 1 − f2 = − − = − 239 MPa
I aa I sa 110.1 * 10 6 86.1 * 10 6
Lower flange
The lower flange is under tension:
M f1 * y2 M * y4 19.8 * 106 * 442.6 56.4 * 10 6 * 484.2

σ fsi = + + f2 = + + = 397 MPa
I aa Isa 110.1 * 106 86.1 * 10 6

NORMAL MEAN STRESSES IN THE FLANGES DUE TO BENDING
Upper flange
Fssm = - F1 - F2 where:
I  M
F1 =  sa  * f 1 Normal load in a flange of a beam stable in shear.
 I aa  h2
Mf 2
F2 = Normal load in a flange of a beam in pure diagonal tension condition.
h2
 86.1 * 10 6  19.8 * 10 6 56.4 * 10 6
F1 =  6
*
 = 20745 N and F2 = = 75563 N
 110.1 * 10  746.4 746.4
The normal load in the upper flange is a compression load:

Fssm = - F1 - F2 = - 20745 - 75563 = - 96308 N
The mean stress in the upper flange is:

F − 96308
σ ssm = ssm = = - 224 MPa
Sss 430
Lower flange
The normal load in the lower flange is a tensile load:
Fsim = F1 + F2 = 20745 + 75563 = 96308 N
The mean stress in the lower flange is:

F 96308
σ sim = sim = = 393 MPa
Ssi 245
STRESSES DUE TO ADDITIONAL LOADS GENERATED BY THE DIAGONAL TENSION

The upper and lower flanges are submitted to a compression load induced by the shear load under the action of the
tension field:
T
Fc = − TD * cot α where TTD = kT
2
α is the diagonal tension angle the calculation of which is detailed in chapter V1.1 and § 5.3.2.
For this calculation, we will suppose that α = 45°
kT 0.74 * 60000
Fc = − = − = − 22200 N
2 2
Upper flange
F − 22200
σssc = c = = − 52 MPa
Sss 430
Lower flange
F − 22200
σsic = c = = − 91 MPa
Ssi 245

STRESSES DUE TO THE SECONDARY BENDING MOMENT IN THE FLANGES

Secondary bending moment in the flanges (refer to § 6.1.5):
1
M fs = k τ e b 2 C fs where Cfs is the attenuation factor given on the curve on page V1-4.6.3 versus Wd
12
1 1
 e 4  0 .6 4
Wd = 0.7 * b *   = 0.7 * 290 *   = 2.1 ⇒ Cfs = 0.955
 (I ss + Isi ) h 2   (57250 + 14780) * 746.4 
1
Mfs = * 0.74 * 134 * 0.6 * 2902 * 0.955 = 398200 mm.N
12
Upper flange at uprights

M * v2 398200 * 30.3
σ fs 2 = − fs = − = − 211 MPa compression stress in lower fibre.
Iss 57250
M fs * v1 398200 * 9.7
σ fs1 = = = 67 MPa tensile stress in upper fibre.
I ss 57250
Upper flange in the centre between two uprights (k = Cfs = 1 ⇒ M’fs = 563470 mm.N)
M 'fs * v 2 563470 * 30.3
′
σ fs 2 = = = 149 MPa tensile stress in lower fibre.
2 * I ss 57250 * 2
M 'fs * v1 563470 * 9.7
σ′fs1 = − = − = - 48 MPa compression stress in upper fibre.
2 * I ss 57250 * 2
Lower flange at uprights
M fs * v 4 398200 * 5.9
σ fs 4 = = = 159 MPa tensile stress in lower fibre.
I si 14780
M fs * v 3 398200 * 24.1
σ fs 3 = − = − = - 649 MPa compression stress in upper fibre.
I si 14780
Lower flange in the centre between two uprights (k = Cfs = 1 ⇒ Mfs = 563470 mm.N)
M 'fs * v 4 563470 * 5.9
′
σ fs 4 = − = − = − 112 MPa compression stress in lower fibre.
2 * I si 14780 * 2
M 'fs * v 3 563470 * 24.1
σ′fs 3 = = = 459 MPa tensile stress in upper fibre.
2 * Isi 14780 * 2
OVERALL STRESSES IN FLANGES
Upper flange at uprights
σssTi = σssm + σssc + σfs2 = - 224 - 52 - 211 = - 474 MPa overall stress in lower fibre.
σssTs = σfss + σssc + σfs1 = - 239 - 52 + 67 = - 224 MPa overall stress in upper fibre.
Upper flange in the centre between two uprights
σssTi = σssm + σssc + σ'fs2 = - 224 - 52 + 149 = - 127 MPa overall stress in lower fibre.
σssTs = σfss + σssc + σ'fs1 = - 239 - 52 - 48 = - 339 MPa overall stress in upper fibre.
Lower flange at uprights
σsiTi = σfsi + σsic + σfs4 = 397 - 91 + 159 = 465 MPa overall stress in lower fibre.
σsiTs = σsim + σsic + σfs3 = 393 - 91 - 649 = - 347 MPa overall stress in upper fibre.
Lower flange in the centre between two uprights
σsiTi = σfsi + σsic + σ'fs4 = 397 - 91 - 112 = 194 MPa overall stress in lower fibre.
σsiTs = σsim + σsic + σ'fs3 = 393 - 91 + 459 = 761 MPa overall stress in upper fibre.

SUMMARY
Overall stresses without secondary bending in flanges (in MPa)

Flange Lower Upper
upper fibre 302 - 291
lower fibre 306 - 276
Overall stresses with secondary bending in flanges at uprights (in MPa)
Flange Lower Upper
upper fibre - 347 - 224
Overall stresses with secondary bending in flanges at centre between two uprights (in MPa)
Flange Lower Upper
upper fibre 761 - 339
V1-4 6.3 CONCLUSION
The calculation above shows that by taking secondary bending into account, the stresses in the flanges are significantly
modified. Thus, in the centre between two uprights, the lower fibre of the lower flange changes from a tensile stress of
302 MPa to a stress of 762 MPa greatly exceeding the breaking stress of the flange (σR= 495 MPa). Whereas, at the
uprights, the lower fibre of the upper flange drops from - 276 MPa to - 474 MPa exceeding the yield strength of the
material of the flange (- 420 MPa).
A new flange design is therefore required. Modifications can be made without adding weight to the existing flange. For
instance, by extending the height of the vertical section of the flange and by adding a lip, which requires a minute
reduction of the thicknesses to conserve same weight, we obtain a considerable gain for the flange quadratic moment;
this being necessary to reduce the stresses due to the action of the secondary bending moment.
Flange studied in the example Flange with same surface area but
with extended vertical section.
An added lip.
Slightly reduced thicknesses.
In the example dealt with the web was relatively thin; also, the value of 0.74 for the diagonal tension factor k means
that the folding of the web is fairly severe as 74% of the shear load is taken by the diagonal tension. For heavily loaded
and fairly low beams, such as wing spars and bulkheads submitted to high external loads, the webs are much thicker
and factor k much lower. Generally speaking, to avoid the forming of permanent folds, no folding of the web is
accepted before 60% of the limit loads have been applied (at present, the target is 80%).
Web weight saving, in relation to that which is required for a web which must not buckle, could exceed the additional
weight in the web flanges and stiffeners induced by the diagonal tension field. However, the calculations for the rivets
in the semi-tension field case poses more problems as the loads in the rivets are higher and more complex than in a
stable-web beam.

V1-4 7 END SPANS AND SPANS WITH OPENINGS
V1-4 7.1 GENERAL
T Load
applied Flange Upright
Flange
Tension exerted by End upright

Opening in Opening
the tension field framework
web
Figure a Figure b
The "end spans" adjacent to the folded area and the edge elements of an opening comprise a specific case. The
diagonal tension pulls the end upright and the edges of an opening towards the interior; this induces secondary bending
the same as the one acting on the flanges in addition to the normal compression. Figures "a" and "b" above clearly
show this action. The end uprights and the edges of the opening must be much more rigid than the uprights in the
standard area or at least reinforced by additional elements to reduce the stresses induced by bending.
V1-4 7.2 EDGE ELEMENTS
The linear load component inducing bending in the edge elements is given by the following formulas:
fby = k τ e tg α for edge elements parallel to the neutral fibre of the beam (flanges).
fbx = k τ e cot α for edge elements perpendicular to the neutral fibre of the beam (uprights).
The greater the distance between the supports of an edge element submitted to a linear load f, the greater the bending
moment that it must support.
A combination of the three methods proposed below would be a good solution to relieve these elements:
• Simply reinforce the element so that it can take all the loads submitted to it.
• Increase the thickness of the web in the edge area either to prevent buckling or to reduce value k and thus the linear
load which introduces the bending into the edge element.
• Add stiffening elements to support the edge element and thus reduce its bending moment induced by fb.
The calculation for the fold angle α is detailed in §5.3.2.
Issue 0 End Spans And Spans With Openings page V1-4•7–1

V1-4 7.3 EXAMPLE
Let us again take the cantilever beam example of § 5.1.8 with same loads and same geometrical characteristics.
1270 60 000 N
A
0,6 290 fb 762 725,4 746,4
A
End upright
The linear load inducing bending in the end upright is:

fbx = k t e cot α = 0.74 * 134 * 0,6 * 1 = 59.5 N/mm (We suppose that α = 45°)
Let us consider the adding of three equally-spaced horizontal stiffeners as method for reducing the influence of the
bending on the end upright and analyse the inner loads in the end span.
End span
60 000 N
290
upp. flange CG
186,6
186,6
Added 746.4 762
fb stiffeners
186,6
186,6
low flange CG
End upright
The figure above shows the end span reinforced by three stiffeners spaced 186,6 mm apart.
Two loads are applied to it:
• The shear load at beam end: 60 000 N

• The linear load component due to the tension field: fb = 59.5 N/mm.
Below, we will show what effect adding three stiffeners has on the local shear in the span.
The effect of each load applied will thus be studied then summed.

60 000 N 60 000 N
290 290
23312 N 22205 N 45517 N
f = 57.4N/mm f = 137.8N/mm
f = 80.4N/mm f = 19.1N/mm f = 99.5N/mm
60 000 N 746,4 + 59.5 N/mm + 59.5 N/mm

f = 19.1N/mm f = 61.3N/mm
60 000 N
f = 57.4N/mm f = 23N/mm
23312 N 22205 N
1107 N
Figure "a" Figure "b" Figure "c"
1) Figure "a" shows the shear flows taken as constant in each of the end span pockets and due to the load applied
F = 60 000 N.
Equilibrium is shown directly on figure "a" above:
60 000 * 290
The reaction at each of the flanges is : R 1 = = 23 312 N
746,4
F R1
The shear flow inside each pocket is: f1 = = = 80.4 N/mm
746,4 290
2) Figure "b" shows the mean shear flows in the centre of each of the pockets of the end span and due to the linear
load component generated by the tension field: fb = 59.5 N/mm.
The equilibrium is shown directly on figure "b" above: 76.6
The reaction at each flange is:
57.4 upper
59.5 * 746,4
R2 = = 22 205 N stiffener
2 19.1 centre
In actual fact, the shear flow in the end span varies from 0 N/mm in 0 stiffener
the centre of the beam up to a maximum at the centres of gravity of 19.1
the flanges: lower
57.4 stiffener
59.5 * 746,4
f max = = 76.6 N/mm
2 * 290 76.6
3) Figure "c" shows the total shear flows obtained in the centre of each pocket of the end span and due to all loads.
CONCLUSION
We can see that the shear flows in the two upper pockets of the end span increase significantly above the nominal value
(80.4 N/mm).
This means that the effect of the diagonal tension in this area (the pocket is smaller) must be locally taken into account
for checking the upper flange (in our example), the end upright, the added stiffeners, the rivets and the web, etc.
As the thin web beam calculation principle supposes that the shear flow is identical in adjacent spans, the justification
of the upper stiffener by the method detailed in chapter V1.1 requires the use of mean flux f and factor k values on
either side of the stiffener as explained in the characteristics of a super-stiffener made symmetrical on page V1-4.2-2
(ditto for LH upright).
Issue 0 End Spans And Spans With Openings page V1-4•7–3

V1-4 7.4 ADDITIONAL STUDIES FOR OPENINGS IN A THIN WEB BEAM
The presence of any openings in a thin web beam creates overstresses in the sheet around the holes whereas secondary
bending loads may appear in the flanges and uprights.
Let us consider a beam submitted to a shear load and a bending moment variable spanwise and including a rectangular
opening on its horizontal axis.
A vss C
v1 Flange CG lines
N neutral fibre N
h2
v2
B vsi q D
T
Mf (CD)
Mf (AB) Tmean
Mf
FLANGE CALCULATION
1) The primary bending causes the following stresses in the flanges at section AB:
M (AB) * V1 M (AB) * V2
σf1 = f (Upper flange) and σf1 = f (Lower flange)
I(AB) I(AB)
where Mf(AB): primary bending moment at section AB.

V1,2: offset between extreme fibre of the upper or lower flange and the neutral fibre of the beam at AB.
I(AB): inertia of beam section (section AB), opening (not necessarily rectangular) subtracted.
2) To pass along the opening, the shear load creates an additional bending moment in the flanges which is calculated
by comparing these with elementary beams of length q/2 clamped at one end and free at the other and submitted to
a load at end equal to the mean shear load in the opening area.
T
T/2
q/2
q if no load T/2
distributed between
upright upright the uprights upright
The breakdown of the moment induced by the mean shear load of the span (no load distributed between the uprights)
prorata the upper and lower flange inertias (Iss and Isi ) gives:
I ss q Isi q
M fss = *T* and M fsi = * T * . The resulting stresses are:
I ss + I si 2 I ss + I si 2
M fss * v ss T * q * v ss M fssi * v si T * q * v si
σ fss = = and σ fsi = =
Iss 2 * (I ss + I si ) I si 2 * (Iss + I si )
v = dimension between the extreme fibre of the flange (upper or lower) and its neutral fibre.
For a variable-height beam, refer to chapter V2.4.
3) Sum total stresses at section AB:
upper flange side: σss = σf1 + σfss lower flange side: σsi = σf2 + σfsi

V1-4 8 STUDY OF WEB
V1-4 8.1 STRESSES IN WEB
V1-4 8.1.1 In pocket thicknesses
Refer to V1-1.5.7.1.
The pocket thicknesses, assumed to be identical on either side of the upright are subjected to plane stresses.
k τ tan α
τ
TDP
2
τ max  k   π τ 
Tresca criterion: = 1+   ;  α =  ⇒  max = 1
τ  tan 2α   4  τ 
V1-4 8.1.2 At uprights
In shaded section at web - upright fastener row:
et
2
τ max e  k 
Tresca criterion: = 1,3 1+ 
τ et 1 + k 
V1-4 8.2 WEB FAILURE
Generally, the skin is thicker at the uprights than in the pockets. Failure can then occur either in the pocket-land blend-
in radii or at the web-stiffener fastener rows.
σR
• Failure in pocket thickness (TRESCA): τmax adm =
2
σR
• Failure at web - upright fasteners: τ'max adm =
2
Issue 0 Study Of Web page V1-4•8–1


V1-4 9 STUDY OF FASTENERS
V1-4 9.1 WEB-FLANGE AND WEB-END UPRIGHT CONNECTIONS
MINIMUM SHEAR STRENGTH
Upright Upright Upright

Flange
y
A
Lt A-2Lt Lt
Flange
x
The stress is not constant along a pocket:
• As the upright is compressed, a portion of the web 2 Lt long is submitted to compression.
• The centre of the pocket is submitted to diagonal tension as shown above over length A-2Lt.
Thus:
* The flows acting on the fastener rows connecting the flange to the web are:
nx 2k ty
= cot α ; = 1
τe 1+ k τe
These flows are supported by any line of fasteners parallel to the flanges.
For stiffened panels: junction parallel to crossmembers.
* The flows acting on the rows of fasteners connecting the end upright to the web are:
ny 2k t
= tan α ; x = 1
τe 1+ k τe
For stiffened panels: junction parallel to stiffeners.
End upright
ps
Flange
Row of fasteners y
ty
nx pm
tx
ny
Issue 0 Study Of Fasteners page V1-4•9–1

TENSILE STRENGTH
The fasteners must also have a sufficient tensile strength to prevent the buckled sheet from separating from the flanges.
The required tensile strength for a fastener is given by the following criterion:
F(fastener failure) ≥ 0.15 * σR * p * e
where: σR: web tensile breaking stress
p: rivet pitch
e: web thickness
V1-4 9.2 STANDARD DISSYMMETRICAL WEB-UPRIGHT CONNECTIONS (SINGLE

ANGLE)
SHEAR STRENGTH
No criteria concerning the shear strength of the fasteners on a single angle has been established by NACA.
The continuity of the web on either side of the upright means that the standard fasteners are subjected to no loads
(except in cases of separation between the web - upright elements). However, at the ends of the uprights, the fasteners
are submitted to loads.
The tensile strength criterion is probably sufficient to ensure a satisfactory design.
TENSILE STRENGTH
The required tensile strength for a fastener is given by the experimental criterion:
F(fastener failure) ≥ 0.22*σ
σR*p*e
(The tensile strength of a fastener is defined as being the tensile load which will cause any failure; if the sheet is thin,
the failure will consist of the fastener being pulled through the sheet).
Spacing between fasteners on single angles must be sufficiently low to prevent inter-rivet buckling (or buckling of the
flange of the angle if its thickness is lower than that of the web), at a compression stress level equal to σTD
y min
. This
spacing must also be lower than b/4 to justify the supported edge assumption made when determining the critical web
stress.
V1-4 9.3 UPRIGHT-WEB CONNECTIONS
These fasteners must take the load applied by the web and transferred to the upright.
The loads are as follows:
Pm = σ TD
y * Sm for a double angle.
Sm
Pm = σ TD
y * 2
for a single angle.
g
1 +  
ρ

V1-4 10 EXAMPLE
Let us again take the cantilever beam of §5.1.1 submitted at its free end to a load of 60000 N and with the following
geometrical characteristics:
l = 1270 mm b = 290 mm e = 0,6 mm h = 762 mm h1 = 725,4 mm h2 = 746,4 mm
60 000N
0,6 290 762 725,4 746,4
3,2
v1
v3 esi
v2 ess 25
v4 7,4
25
Upper flange Lower flange Upright (single angle)
Sss = 430 mm2 Ssi = 241 mm2 Sm = 150 mm2

Iss = 57 250 mm4 Isi = 14 780 mm4 Im = 8654 mm4
ess = 4,5 mm esi = 2,5 mm em = 3,2 mm
v1 = 9,7 mm v3 = 24,1 mm
v2 = 30,3 mm v4 = 5,9 mm
The mechanical characteristics of the web of the 2024 PLT3 beam are:
Ea = 70300 MPa; σRa = 440 MPa; σ0.2a = 270 MPa
The mechanical characteristics of the members (flanges and uprights) of the 7075 T73510 beam are:
Em = 73800 MPa; σRm = 495 MPa; σ0.2m = 420 MPa
The two materials have the same Poisson ratio: νe = 0.33
REMINDER
The limitations, the critical shear stress, the loading ratio and the diagonal tension factor were checked in § 5.1.8:
e 3,2
From among the limitations, we must check: m = = 5,3 > 0,6
e 0,6
τcr,a = 1.7 MPa
We recall that: τ = 134 MPa
Rs = 79
k = 0.74
Issue 0 Example page V1-4•10–1

CALCULATING UPRIGHT CHARACTERISTICS

Lt 290
Load-carrying width: Lt TD = (1 − k ) 0 = (1 − 0,74 ) = 19 mm
2 4 X X
Load-carrying section:
290 LtTD LtTD
∆STDy = (1 + k) Lt0e = (1 + 0,74) * * 0,6 = 151,2 mm2
2
S0 = Sm + 2Lt0e = 150 + 290 * 0,6 = 324 mm2
STD
y = S0 - ∆STD
y = 324 - 151,2 = 172,8 mm2
Web characteristics:
Sa0 = b * e = 290 * 0,6 = 174 mm2
IXXa0 = 179 mm4
WXXa0 = - 174 mm3
da0 = - 1 mm
e 3
WXXaTD
= WXXa0 + ∆STD
y = - 174 + 0,3 * 151,2 = - 128,6 mm
2
 e 2 e 2  TD  0,6 2 0,6 2 
I TD
XXa = I XXa0 -  +
 4 12  y ∆ S = 179 −  +  * 151,2 = 161 mm 4
   4 12 
Characteristics of the load-carrying section of the upright after correction due to the various materials
 STD   Sm 
Ey =  aTD  E a +  TD  E =  22.8  * 70300 +  150  * 73800 = 73338 MPa
 Sy   S y  m  172.8   172.8 
   
 E  TD  E m   70300   73800 
TD
WXX =  a  WXXa + W = * − 128.6 +   * 1111.6 = 995 mm
3
 Ey   E y  XXm  73338   73338 
   
TD
WXX 995
dTD = TD
= = 5,76 mm
Sy 172.8
 Ea  TD  
I TD
XX =
  I XXa +  E m  I XXm =  70300  * 161 +  73800  * 8654 = 8863 mm 4
 Ey   Ey   73338   73338 
   
2
ITD = I TD TD
XX − S y d
TD
= 8863 - 172.8 * 5.762 = 3130 mm4
Characteristics of the load-carrying section of the lower flange (less rigid) after correction due to the various
materials
h 1e 725,4 * 0,6
∆STD
x = (1 + k ) = (1 + 0.74) * = 189 mm 2
4 4
 h 1e   725,4 * 0,6 
STD
X =  Ssi +
TD
 − ∆SX =  241 +  − 189 = 269,6 mm
2
 2   2 
2
STD TD
Xa = SX − Ssi = 269,6 - 241 = 28,6 mm
 STD  Ssi  28,6   241 

Ex =  Xa  * E a + TD
TD 
* Em =   * 70300 +   * 73800 = 73429 MPa
 SX  SX  269,6   269,6 

CALCULATING THE STRESSES IN THE UPRIGHT (WITH ASSOCIATED WEB)

Web orientation compression:
k Lt 0 e τ 725,4
N TD
x = − = − 0.74 * * 0.6 * 134 * cot α = - 21579 cot α
tan α 2
N TD − 21579 cot α
σ TD
x = x
TD
= = − 80 cot α
Sx 269,6
σ TD − 80 cot α
ε TD
x = x
= = − 0.109%. cot α
Ex 73429
Upright orientation compression:
N TD
y = - k b e τ tan α = - 0.74 * 290 * 0,6 * 134 * tan α = - 17254 tan α
N TD
y − 17254 tan α
σ TD
y = TD
= = − 99.85 tan α
S y 172,8
σTD
y − 99.85 tan α
ε TD
y = = = − 0.136% tan α
Ey 73338
Diagonal tension:
As the fold angle is around 45°, we assume sin 2α = 1
τ 134
ε TD ≈ [1 + k + ν e (1 − k )] = [1 + 0,74 + 0,33(1 − 0,74 )] = 0,348%
Ea 70300
Diagonal tension angle:
 ε − ε TD 0.348 + 0.109 
(α0 = 45°) →  tan 2 α1 = TD X
= = 0.944 ⇒ tan α1 = 0.972 ⇒ α1 = 44.18° 
 TD
ε TD − ε y 0.348 + 0.136 
 
 ε − ε TD 0.348 + 0.109 / 0.972 
(α1 = 44.18°) →  tan 2 α 2 = TD X
= = 0.958 ⇒ tan α 2 = 0.979 ⇒ α 2 = 44.39° 
 TD
ε TD − ε y 0.348 + 0.136 * 0.972 
 
 ε − εX TD
0.348 + 0.109 / 0.979 
(α2 = 44.39°) →  tan 2 α 3 = TD = = 0.955 ⇒ tan α 3 = 0.977 ⇒ α 3 = 44.34° 
 TD
ε TD − ε y 0.348 + 0.136 * 0.979 
 
 ε − εX TD
0.348 + 0.109 / 0.977 
(α3 = 44.34°) →  tan 2 α 4 = TD = = 0.956 ⇒ tan α 4 = 0.978 ⇒ α 4 = 44.35° 
 TD
ε TD − ε y 0.348 + 0.136 * 0.977 
 
finally: α = 44.35°; tan α = 0.978
Mean and minimum compression stresses in the upright.

σ TD
y = - 99.85 tan α = - 99.85 * 0.978 = - 97.7 MPa
σ TD
y min  b   290 
= (1 − k ) 1.78 − 0.64 *  + k = (1 − 0.74) 1.78 − 0.64 *  + 0.74 = 1.14
σ TD
y  h m   725,4 
σ TD TD
y min = 1.14 * σ y = - 111.4 MPa
REMARKS:
The upright orientation and flange orientation stiffening coefficients are close to each other:
324 − 290 * 0,6 466 − 375 * 0,6
upright orientation coefficient: = 0.46; flange orientation coefficient: = 0.52
324 466
We were justified to use the following approximation (difference negligible in relation to the solution above).
(α ≈ 45°) ⇒ ( σ TD
y = - 99.9 MPa; σ TD
y min = 1.14 * - 99.85 = - 113.9 MPa)

STRESSES IN WEB ALONE AND IN UPRIGHT ALONE

Shear in web (refer to V3-1 § 5.7.1)
(α ≈ 45°) ⇒ (τmax = τ = 134 MPa)
Shear in web at rivet row (refer to V3-1 § 5.7.2)

e
τ'max = τmax = τmax = τ = 134 MPa
et
Compression stress in upright orientation web
Ea 70300
σ TD TD
ya min = σ ymin * = − 111.4 * = − 106.8 MPa
Ey 73338
Compression stress in upright alone
E 73800
σ TD
m = σ TD
y * m = − 97.7 * = − 98.3 MPa
Ey 73338
Em 73800
σ TD TD
m min = σ y min * = − 111.4 * = − 112.1 MPa
Ey 73338
ALLOWABLE STRESSES
The upright consists of a single angle; we will calculate below the allowable column buckling and Forced crippling
stresses.
Column buckling stress
• No stress in the upright exceeds the elastic stress of the material: 420 MPa.
• Calculation of allowable stress at zero slenderness ratio supposing, to limit calculations, that the limitation comes
from the web (refer to V3-1 § 4.6.4):
270
σ0a = 270 MPa; ε0 = + 0.002 = 0.584%
70300
13.83
0.584 385.6  385.6 
σ0m = 385.6 MPa because: = + 0.002 *  
100 73800  420 
 STD   Sm 
σ0 =  aTD  σ 0a +  TD  σ =  22.8  * 270 +  150  * 385.6 = 370 MPa
 Sy   S y  0 m  172.8   172.8 
   
The mean allowable stress at zero slenderness ratio is: 370 MPa.
I TD 3130 h 725,4 100 100
ρ = = = 4,3 mm; λ = m = = 84; σ 0 * = 370 * = 0.5
STD
y 172,8 2ρ 2 * 4 .3 Ey 73338
According to the diagram on page V3-1.4/23:
σ * 100 σ
for λ = 84 and 0 = 0.5 we obtain : crit = 0.28 ⇒ σcrit = 0.284 * 370 = 105 MPa
Ey σ0
The critical column buckling stress is - 105 MPa.
The mean stress in the straight section of the upright is – 97.7 MPa and does not exceed - 105 MPa.
Forced crippling stress
(es = 3,2 mm; ea = 0,6 mm; σ0.2m = 420 MPa; Em = 73800 MPa; Ea = 70300 MP; k = 0.74)
1 1
2 2
σ 0.2 m  e E 3 420  3.2 73800  3
σ flf = − 0.051 k 3  s m  = − 0.051 0.74 
3  = − 354.8 MPa
σ 0.2 m
+ 0.002  ea E a  420
+ 0.002
 0,6 70300 
Em 73800
The stress σTD
m min = - 112.1 MPa does not exceed the stress of – 354.8 MPa.

SSM V1-5 • STABLE WEB BEAMS
CONTENTS
issue date change
1 4/1999
V1-5 STABLE WEB BEAMS
1 4/1999
V1-5 1 GENERAL
1 4/1999
1 4/1999
V1-5 3 STABLE WEBS
1 4/1999
V1-5 4 STUDY OF UPRIGHTS
1 4/1999
V1-5 5 JUNCTIONS
1 4/1999
V1-5 6 EXAMPLE
1 4/1999
V1-5 7 STABLE WEB BEAMS WITH ROUND LIGHTENING HOLES.
V1-5 1 GENERAL 1–1

V1-5 1.1 DEFINITIONS 1–1
V1-5 1.2 BEHAVIOUR OF A STABLE WEB BEAM 1–2
V1-5 1.3 FAILURE MODES 1–3
V1-5 2 STUDY OF FLANGES 2–1

V1-5 2.1 ALLOWABLE STRESSES 2–1
V1-5 3 STABLE WEBS 3–1

V1-5 3.1 CRITICAL STRESSES IN WEBS 3–2
V1-5 4 STUDY OF UPRIGHTS 4–1

V1-5 5 JUNCTIONS 5–1
V1-5 5.1 GENERAL FORMULAS 5–1
V1-5 5.2 JUNCTIONS BETWEEN REINFORCING PLATES
AND FLANGES 5–2
V1-5 5.3 WEB SPLICES 5–3

V1-5 7 STABLE WEB BEAMS WITH ROUND
LIGHTENING HOLES 7–1
V1-5 7.2 STANDARDISATION OF FLANGED HOLES 7–1
V1-5 7.3 ALLOWABLE SHEAR FLOW 7–2
V1-5 7.4 REMARKS 7–5

 AEROSPATIALE - 1999
SYMBOLS USED
a: Pocket length (between fastener rows) Lt0: Initial load-carrying width

b: Pocket width (between fastener rows) Lt: Load-carrying width
B: Width of a flange M: Moment
C: Distance between the edges of 2 holes Mf: Bending moment
C’: C-2B Mf,ame: Bending moment in web
d: Diameter of a rivet or of a hole Mmax: Max bending moment
D: Diameter of a hole including the stamping n: Work hardening coefficient (Ramberg and
e: Thickness of the web Osgood), or number of rivets
E: Young’s modulus N: Normal tensile force
Ec: Young’s modulus in compression p: Rivet pitch
Ep: Young’s modulus - panel r: Distance between a rivet and the centre of
the group of rivets
f: Shear flow
Rf: Loading ratio in single bending
fadm: Allowable shear flow
Rs: Loading ratio in single shear
fp: Shear flow in a reinforcing plate
Fadm,cis: Allowable shear force S: Area
FMf: Force induced by the bending moment Sc: Reduced shear section
FR: Resulting force Sp: Area of reinforcing plates
FS: Tensile force on a flange Ss: Total area of a flange
FT: Force induced by shear force Sss: Upper flange area
g: Distance between hole centrelines Ssi: Lower flange area
ha: Height of the web (between datum lines) T: Shear force
H: Height of the beam or of a flange Tcr: Critical shear force
I: Bending rigidity constant (inertia)
v: Calculation dimension in relation to neutral
Im: Inertia required for an upright fibre
Iss: Upper flange inertia W: Static moment
Isi: Lower flange inertia WSemelle: Static moment flange section alone
IA.N: Inertia in relation to neutral fibre y: Distance between the centreline or the
I0: Inertia of an item in relation to its own neutral fibre and the cg of the item
centreline y : Offset between the centreline and the
Ks,ks: Buckling coefficients in shear neutral fibre
Kf,kf: Buckling coefficients in bending
L: Spacing between uprights (between datum
lines)
Issue 1 Contents page V1-5•iii

ε: Strain (expansion) τ: Shear stress

η: Plasticity correction coefficient τR: Breaking shear
νe: Poisson ratio - elastic τcr: Critical shear
σmat: Allowable bearing stress τcr,0: Critical shear stress alone
σ0.2: Conventional yield strength
τh, τc: Collapsing stress in shear for a plate of
σR: Breaking tensile stress length h or c
σcr: Critical stress τmax: Max shear stress
σcrip: Critical crippling stress τnet: Net shear stress at a hole
σcp: Critical panel compression stress
σf: Bending stress
σcf,0: Critical bending stress alone
Issue 1 Contents page V1-5•iv

BIBLIOGRAPHY
1- BRUHN
Issue 1 Contents page V1-5•v

Issue 1 Contents page V1-5•vi

V1-5 1 GENERAL
This chapter mainly deals with the means for lightening thin profile beams without buckling the webs. The
case where the web folds and becomes the seat of a diagonal tension field is studied in chapter V1-4.
The general beam theory is developed in volume 2 of MTS004.
The structural weights of stable web beams are not always optimal when compared with thin web beams.
They must however be preferred each time that the stiffness of the beam is a design criterion.
The first part of this chapter gives an exhaustive study of flat web beams equipped with uprights. The second
part of this chapter deals with beams with webs drilled with flanged holes. This type of beam can provide a gain
in weight when compared with the full web beam stiffened by relatively spaced uprights.
V1-5 1.1 DEFINITIONS
A beam is said to be a stable web beam if while supporting the loads it conserves its initial flatness (no
buckling of the web). The tangential stress remains lower than the critical buckling stress of the webs which are
considered as plates bearing on their members with rigidity sufficient to prevent all global buckling of the web.
We recall that the buckling stress of the web is not necessarily a failure stress. Thus, to be stable, a web is
not loaded to its maximum capacity. Its stiffeners or uprights which are used only to prevent local buckling must
also be stable to avoid any problems due to crushing.
To obtain a satisfactory strength/weight ratio, the flanges must be designed to make the radius of gyration of
the cross section of the beam as large as possible whilst themselves having a cross section resisting buckling.
Also, for long cantilever beams (the spars of a cantilever wing for instance) the flange sections must be tapered,
that is a reduction in the section from the root towards the tip.

V1-5 1.2 BEHAVIOUR OF A STABLE WEB BEAM
Only one bending plane is considered.
x
T
y
Mf
a a
The external forces which can be applied to a stable web beam are:
• T: Shear force
• Mf: Bending moment
The webs of thin web beams, which are unstable, take only the shear and the diagonal tension.
Here, the behaviour is in compliance with the plane bending theory. The shear force is contained in the plane
of the web and the flanges can be compared with plates in the planes almost perpendicular to that of the web.
Therefore, the web takes almost alone the shear force whereas the highest normal stresses are taken by the
flanges.
Schematically, the web is submitted to shear and bending stresses whereas the flanges are submitted to
normal tensile or compression forces. The presence of the uprights stabilises the web and limits the local strains
due to point loads.
Compared with a beam without uprights subjected to the same forces, the thickness of the web can be
reduced thus leading to a significant gain in weight for very high beams.
The uprights are dimensioned to ensure at least correct web supporting conditions.

The causes of failure in a stable web beam are:

• Flange tensile failure.
• Local buckling of flanges in compression.
• Failure of the web in shear.
Tests are required to accurately determine the allowable stresses in the beams. However, in most cases, the
material curves (in tension and in compression) and the limitations given by the buckling calculations in
compression comprise a sufficient database to evaluate the allowable stresses.
It often happens that the stresses calculated at the extreme fibre of a section are above the yield strength of
the material which means that a portion of the section will buckle before the others and, as in certain cases,
several different materials are used in the same section, a general method for calculating the ultimate strength of
the section consists in using the stress-strain laws of the materials.
In a stable web beam, the following must be checked:

• The stability of the web (plate submitted to the combined effects of compression, bending and
shear).
• The stability of the compressed flange.
• The rigidity of the uprights ensuring the stability of the web.
• The non-failure condition of the tensioned flange.
• The strength of the attachments.
• The shear strength of the web.


The flanges of a stable web beam are mainly submitted to normal forces generated by the general bending of
the beam.
The stability of the compressed flange and the non-failure of the tensioned flange must be ensured.
Tensioned flange
Mf
Compressed flange
V1-5 2.1 ALLOWABLE STRESSES
COMPRESSED FLANGE
The allowable stresses in a flange submitted to compression are those guaranteeing good resistance to all
local buckling phenomena.
The calculation of these stresses was studied in chapter V1-3.
TENSIONED FLANGE
The allowable stress is the failure tensile stress.


V1-5 3 STABLE WEBS

The web, designed not to buckle under the action of the design loads is capable of supporting the plane
bending stresses in addition to the shear force.
In the elastic range, the stresses due to the bending moment and to the shear force are:
M v
• σf = f where I is the inertia of the complete cross section of the beam.
I
T
Ie∫
• τ= y dS (BREDT formula).
As integral ∫ y dS is maximum at the neutral fibre => τmax is located on the neutral fibre.
If the material is concentrated in the flanges, the shear stress in the web is almost constant and is around:
T T
τ= =
h.e Sc
Sc: Reduced shear section.
I
h= .
W
W: Static moment of the 1/2 section.
Example:
10
G I 100
50 M
10
50 G
Figure (a) Figure (b)
Figure (a): Shear stress at G:
T T T
τG = = =
I 84.8 × 5 424
.e
W
Figure (b): Shear stress at M:

T T T
τM = = =
I 99.9 × 5 499.3
.e
W
τM
⇒ = 0.85
τG
Issue 1 Stable Webs page V1-5•3–1

V1-5 3.1 CRITICAL STRESSES IN WEBS
The critical stresses in the webs are calculated in the same way as described for a flat plate in chapter V1-2
"Buckling of plates and thin shells".
The web of the beam is considered as simply supported on its edges (flanges and uprights). The uprights
must provide sufficient stiffness to prevent the buckling of the pocket and also so that they can be considered as a
support (see next paragraph).
T
a
Mf
e
b
Let: a = distance between uprights (between fastener rows);

b = height of beam (between fastener rows);
e = web thickness
SHEAR (see V1-2.2.3.1.2)

2
π 2k s E  e  τ
τ cr,0 = η 2   RS =
( )
12 1 − ν e  min{a, b} τcr,0
BENDING (see V1-2.2.3.1.3 ) σ1
π2 k f E  e 2 σf
σ cf,0 = η   Rf = Mf σ1 − σ 2
( )
12 1 − ν 2e  b σcf ,0 σf =
2
σ2
SHEAR + BENDING (see V1-2.2.4.1)
1
R s2 + R 2f = 1 where R.F. = −1
R + R 2f
2
s
Issue 1 Stable Webs page V1-5•3–2

V1-5 4 STUDY OF UPRIGHTS

The moment of inertia to be given to a cross section of an upright intended to prevent the buckling of the
web under shear is:
4
2.29 L  Tcr h a  3 L
Im =  
e  33 E 
Im: Moment of inertia required for the upright.
e ha
L: Spacing between uprights (between datum lines).
e: Web thickness.
Tcr: Critical vertical shear force in section.
ha: Web height (between datum lines).
E: Young’s modulus of elasticity.
2
 e 
As Tcr = τcr h a e and τ cr = K s E   hence:
 min{L, ha }
τcr: Web critical shear stress.

Ks: Buckling coefficient in shear.
The curve below is used in the following cases:
• All flat or curved panels.
• The uprights are attached at both ends (flanges).
• No participation of the effective skin must be included in the calculation of the moment of inertia.
• The curve is only applicable to cases where buckling is elastic or where L = min {L , h a }
The moment of inertia required for the upright can be written in another way:
4
 2
3
 Ks E   h 2a
e
e
4
2.29 L  τ cr h 2a e  3 2.29 L   L  = 0.0217 L e
3
Im 0.0217
Im =   = => =
e  33 E  e  33 E  8
Le 3 8
   L 3  L 3
     
h K  h K 
 a s   a s 
When the upright is used as such and not as a means for transferring an outside load concentrated on the web
of the beam, the question to be asked is what is the minimum number of rivets required to connect the upright to
the web? For stable webs, we suggest two criteria:
1. The upright must be attached to the flanges at both ends.
2. The spacing between the rivets must be at most equal to a quarter of the spacing between the uprights or
a quarter of the height of the web if this is smaller so as to justify the simply supported hypothesis on the
edges of the web panel.
Issue 1 Study Of Uprights page V1-5•4–1

Curve for calculating the minimum inertia

required for the upright
1000
100
lm / (L * e3)
10
0,1
0 0,04 0,08 0,12 0,16 0,2
L / (ha * Ks0,5)
Issue 1 Study Of Uprights page V1-5•4–2

V1-5 5 JUNCTIONS
V1-5 5.1 GENERAL FORMULAS

In general, standard beams consist of separate web and flange elements assembled by rivets, bolts or other
means. We must know the loads that these fasteners will take to give them a suitable strength.
p
C A GFlange
F s + ∆F s rivet Fs
D B Q
fxy
T
d
fyx
∆M
M+∆ neutral M
fibre GComplete section
T
D’ B’
F i + ∆F i rivet Fi
C’ A’
The figure above shows a beam portion equal in length to the riveting pitch "p". The section at (AA’) is
submitted to a bending moment M, the section (CC’) is submitted to moment M+∆M. The distribution of the
bending stresses in the beam sections is given by the triangular profiles assuming that the web takes no bending
stresses.
Let Fs be the total tensile force on the sections comprising the upper flange AB due to the bending stresses
induced by M in section AA’. Also, the total tensile force on the sections comprising the flange CD due to the
bending moment M+∆M is equal to Fs + ∆Fs. Under the action of these two forces, the sections comprising the
upper flange must move towards the left; this movement is prevented by the rivet which attaches these angles to
the web. Therefore, the force on the rivet is equal to ∆Fs (same reasoning for the lower flange).
M
Fs = ∫ σ dS where: σ =
I
x
Semelle sup
M M
Fs =
I ∫ x dS = I
Ws where: Fs: force normal to upper flange.
Semelle sup
Also:
M M
Fi = ∫ x dS = I Wi
I Semelle
where: Fi: force normal to lower flange.
inf
We deduce:
∆M ∆M
∆Fs = Ws and ∆Fi = Wi
I I
Now:
∆M = T × p
T. Ws T. Wi
⇒ ∆Fs = .p and ∆Fi =
.p
I I
∆Fs and ∆Fi represent the horizontal shear flow caused by the flanges over a length p.
Issue 1 Junctions page V1-5•5–1

The fasteners ensuring the cohesion of the beam must be capable of transferring the shear flows given by
T.W
formula: f = .
I
The fasteners attaching the web to the flange are submitted to shear flow fxy, passing from the web into the
flange.
At point Q, the shear flows equal:
T
f xy = fyx = Wsemelle .
I
where: I = Moment of inertia of complete beam section.
Wsemelle = Static moment of flange section alone.
Wsemelle = ∫ y dS = Ssemelle × d
We can deduce the force per fastener:

T. Wsemelle
∆F = ×p
I
Specific case: If the two flanges are identical and the section of the web is low: 1 = 2 x Ssemelle x d2
T
⇒ f=
2d
Approximation: By taking 2d = H the height of the beam, we obtain results close to the general expression.
T
⇒ f=
H
V1-5 5.2 JUNCTIONS BETWEEN REINFORCING PLATES AND FLANGES

The figure below shows typical beam flanges consisting of a main (T-shaped) element reinforced by plates
(a) and (b) for the upper flange and (c) and (d) for the lower flange. The aim of the fasteners is to prevent the
reinforcement plates from sliding along the main (T-shaped) element of the flange by the bending action of the
beam. This horizontal force which tends to make the reinforcing plates slide, this being prevented by the shear of
the rivets, is given by the fundamental shear flow equation:
T
f = ∫ y dS
I
fa1 fa2
a
GComplete section
b
ha
Hd hc
T Shear flow
GComplete section
1st shear plane
Upper flange c d 2nd shear plane
Lower flange
Upper flange (LH portion):
T. Wa1
Flow passing through shear plane: f a1 =
I
Where: Wa1 = S a1 × h a

Lower flange (LH portion):

T.Wc1d1
• Flow passing through 1st shear plane: f c1d1 =
I
Where: Wc1d1 = Sc1 . hc + Sd1 . hd
T.Wd1
• Flow passing through 2nd shear plane: f d1 =
I
Where: Wd1 = Sd1 . hd
The reasoning is the same for the RH portions of the plates.
Where: index 1: LH portion of the plate.

index 2: RH portion of the plate.
V1-5 5.3 WEB SPLICES

Splices are often used in the design of a beam with thin sheet sections either to increase the thickness of the
web or to ensure a junction between two beams. The case shown on figure (a) below uses splices to avoid the
joggling of a relatively heavy web as case on figure (b).
ra Overlap
junction
Splices
Figure (a) Figure (b)

x
fy Calculation of normal flow: fy = σy . e
T.W(x )
fx Calculation of tangential flow: fx =
x2 I
x2
Where: W(x ) = ∫ x dS
x1
y x1
W(x) static moment of the section above x
Remark: If we consider fx as constant (conservative):

T.W
Max. shear flux: f x = where W = 1/2-section static moment.
I

V1-5 5.4 EXAMPLE

Let us check the riveting of the beam below supposing that the vertical shear force on this section is equal to
15000 N.
38 1,6 44
8
Detail of flange
30 1,3
sections
All rivets 19
ASN-A 2051 Neutral fibre
diameter 3.2 mm 14,3
178 Centreline 5,3
19
Web height
175 mm
1,6
8
Surface area = 58 mm2
Inertia = 2020 mm4
Tensioned flange
ITEM AREA (S) Y Sy Sy2 I0 I = I0 + S y 2

mm2 mm mm3 mm4 mm4 mm4
Upper flange 116 83,7 9709 812660 4040 816700
Upper reinforcement
70 89,8 6286 564483 negligible 564483
plate
Web 228 0 0 0 580600 580600
Lower flange 116 - 83,7 - 9709 812660 4040 816700
Lower flange rivet
- 14 - 78 1092 - 85176 negligible - 85176
hole
Total 516 7378 2693307
I0: moment of inertia of an element in relation to its own axis.

y: distance between the centreline and the centre of gravity of the element.
The offset between the centreline and the neutral fibre is given by: y =
∑ Sy = 7378 = 14, 3 mm
∑ S 516
The moment of inertia in relation to the neutral fibre is:
( )
I A.N = I − ∑ S y 2 = 2 693 307 - 516*14, 32 = 2 587 790 mm 4
CHECKING WEB-FLANGE RIVETING

T
I A .N ∫
The shear flow is given by formula: f = y dS
The beam bending load is supposed to be such that the upper flange is compressed.
The first step consists in determining the moment of inertia of the cross section in relation to the neutral fibre
as shown in the calculation details of the table above, by calculating the moment of inertia at the centreline of the
section then transferring it to the neutral fibre.
The force passing through the rivets attaching the upper flange angles to the web:
F = f * p ( p = rivet pitch = 30 mm)

The static moment of the flange in relation to the neutral fibre is:
∫ ydS = 2 * 58* (89 −14, 3 − 5, 3) = 8050 mm

3
The static moment of the reinforcing plate in relation to the neutral fibre is:
∫ ydS = 44 *1, 6 * (89 − 14, 3 + 0,8) = 5315 mm

3
The total static moment of the web - upper flange junction in relation to the neutral fibre is: 13365 mm3
The force in the rivet is then:
 T  15 000
F = f *p = 
 IA .N
∫ ydS * p = 2 587 790 * 13 365* 30 = 2324 N
The shear flow calculated by the simplified equation can be written:
T
f= where H is the height of the beam.
H
T 15 000
The force in the rivet is then: F = f * p = *p = * 30 = 2506 N
H 179.6
This result is correct if compared with the 2324 N obtained by the exact theory.
The web is attached to the two sections of the flange by 3.2 mm diameter aluminium alloy ASN-A 2051
rivets with single shear strength of 180 MPa.
The allowable double shear force in a web-flange junction rivet:
 d2  3, 22 
Fadm ,cis = 2 * τ R *  π  = 2 *180 *  π  = 2895 N
 4  4 
The allowable bearing force of the web (1,3 mm thick 2024-T3 element, σmat = 905 MPa) is:
Fadm ,mat = σmat * e * d = 905 *1, 3 * 3, 2 = 3765 N
The allowable double shear force is therefore more critical; the Reserve Factor is then:
F 2895
R.F = adm ,cis = = 1, 24
F 2324
CHECKING RIVETING OF REINFORCEMENT PLATE.
Two rows of rivets with a rivet pitch of 38 mm.
The load on two rivets is:
 T  15 000
F = f *p = 
 IA .N
∫ ydS *p = 2 587 790 * 5 315*38 = 1170 N i.e. 585 N per rivet
Using the simplified formula, the load on the two rivets is:
 T Sp  15 000 70.4
F = f *p =  *p = * * 38 = 1198 N i.e. 599 N per rivet
 H Ss s  179.6 186.4
The rivets are stressed in single shear:
 d2   3, 22 
Fadm ,cis = τ R *  π  = 180 *  π  = 1447 N
 4   4 
1447
The Reserve Factor is: R. F = = 2.47
585


V1-5 6 EXAMPLE
Let us take an assembled beam with an I-section, bearing simply at 2 points and loaded by three forces
concentrated as shown and which comprise the design loads for the beam.
200
pocket 28
630
Uprights made of sections:
Span 4
Dimensions :13,5*13,5*1
eame = 1,5
pocket 22
11 000 N
pocket 21
630
Span 3
90 eame = 1,8
pocket 15
12 250 N
pocket 14
3,2 mm diameter
rivets in flange,
pitch 25 mm
630 Span 2
eame = 1,8
pocket 8
11 000 N
pocket 7
pocket 6
pocket 5
630
Span 1
pocket 4
eame = 1,5
pocket 3
pocket 2
pocket 1

10,5
19 19 19 2,4
0,6 (sheet)
5,6
neutral fibre
19
Sss = 85 mm2
100
Iss = 2856 mm4
8 upper flange
neutral fibre
200 central line
web height 1,5
196,8
Ssi = 58 mm2
3,2 (rivet hole)
Isi = 2020 mm4
0,5 (skin) 19
11 neutral fibre
5.3
19 1.6
100
lower flange
The characteristics of the materials: 2024 PLT 351 Skin, Sheet and Web; 7075 T73510 flanges
Skin, Sheet and Web: σ R = 440 MPa ; σ 0,2 = 270 MPa ; E c = 70300 MPa ; n = 7.05
Flanges: σ R = 495 MPa ; σ 0,2 = 420 MPa ; E c = 73800 MPa ; n = 13.83
S Y Sy Sy2 I0 I0 + Sy2
ELEMENTS
mm2 mm mm3 mm4 mm4 mm4
upper flange 170 94,4 16048 1514931 5712 1520643
upper skin 22,8 100,3 2287 229370 negligible 229370
web (th=1,5 mm) 295,2 0 0 0 952764 952764
web (th=1,8 mm) 354,24 0 0 0 1143317 1143317
lower flange 116 - 94,7 - 10985 1040298 4040 1044338
lower skin 50 - 100,25 - 5013 502503 negligible 502503
flange-skin rivet holes - 13,44 - 99,45 1337 - 132926 negligible - 132926
flange-web rivet holes - 15,04 - 89 1339 - 119132 negligible - 119132
flange-web rivet holes - 16 - 89 1424 - 126736 negligible - 126736
TOTAL (1,5 mm web) 625,5 5012 3997561
TOTAL (1,8 mm web) 683,6 5098 4180509
Section with 1,5 mm web
Σ sy 5012
y= = = 8 mm
ΣS 625,5
2
IA.N. = (Σ I0 + Sy2) - (ΣS) y = 3997561 - 625,5 * 82 = 3957529 mm4
Section with 1,8 mm web: (with same position of neutral fibre)

2
IA.N. = (Σ I0 + Sy2) - (ΣS) y = 4180509 - 683,6 * 82 = 4136758 mm4

CHECKING BEAM BENDING STRENGTH

11000 12250 11000
17125 17125
630 630 630 630
16000000
14647500
14000000
12000000
10000000
8000000
6000000
4000000
2000000
0 x
630 1260 1890 2520
With the beam simply supported at both ends, with symmetrical loading and with a constant cross section
along complete length.
The maximum bending moment is found in the centre of the beam as shown by the equilibrium of the part
and the bending moment diagram above.
M max = 17125* 1260 −11000 * 630 = 14 647 500 mm.N
The beam is riveted to a skin panel as beam section shows on previous page. We will associate a certain
efficient sheet width to the upper flange which under the loading considered is in compression. This will depend
on the stress in the flange of the beam.
On tensioned flange side, the complete skin is load-carrying which corresponds to the half-distance to the
first skin stiffener on each side (distance between 2 stiffeners: 100 mm).
Details of the cross section in the centre of the beam are shown on the previous page with three rivet holes
on the tensioned flange. The table on the same page gives the geometrical characteristics of the section in relation
to the neutral fibre.
The bending stress at the centre of gravity of the horizontal portion of the compressed flange is:
Mf * v 14 647 500 * 90.8
σf = = = 321.5 MPa
I 4 136 758
The critical stress in the 1 mm thick and 78 mm wide sheet (between fastener rows) is:
2 2
kcπ2Ep  e 4∗ π 2 ∗70300  0,6 
σ cp =η   = η   = 15.4 MPa 7
12(1 − ν 2 )  b  12(1 − 0.332 )  78 
σ cp 15.4
Load-carrying width on a leg of the compressed flange: L t = L t0 = 39 = 8,5 mm
σf 321.5
The total width of each of the leg is 8,5 + 10,5 = 19 mm which is exactly the value taken for the calculation
of the geometrical characteristics of the section of the beam.

CALCULATING ALLOWABLE COMPRESSION STRESS IN UPPER FLANGE

The upper flange consists of two right angle extrusions with two equal legs attached both to the skin and to
the web. The allowable stress of such a beam is the critical crippling stress which is equal to the critical buckling
stress of one of the two equal legs of the section:
∑ S i σ cri
As per §V1-3.7: σ crip = = σ cr of one of the two legs.
∑ Si
2 2
ηE c π 2  e η * 73800 * π 2  2.4 
σ cr = K c 2   = 0.43   = η * 532.5 MPa
12(1 − ν e )  b  12(1 − 0.332 )  1.8 
After plasticity correction σ cr = 417 MPa
417
The Reserve Factor: R.F = = 1.3
321.5
CHECKING LOWER FLANGE IN TENSION

The bending stress at the centre of gravity of the skin is:
M * v 14 647 500 * 108.25
σf = f = = 383.3 MPa
I 4 136 758
440
The Reserve Factor (Skin): R.F = = 1.15
383.5
495
The Reserve Factor (Lower flange): R.F = = 1.29
383.5
The Reserve Factor will be improved if we use the method based on the stress-strain diagram as explained in
§ 2.1.1 and which consists in calculating the ultimate bending moment.
CHECKING BUCKLING STRESS IN WEB

The maximum shear force is found at the bearing points and is equal to: 17125 N.
The thickness of the web at the bearing points is 1,5 mm.
The height of the web is the distance between the rows of fasteners connecting the web to the flanges:
178 mm.
The spacing between the uprights is 90 mm.
The maximum shear stress in the web can be calculated:
T 17125
by the simplified equation: τ = = = 64.1 MPa .
he 178 * 1,5
T 17125
by the exact equation: τ = ∫ ydS =
Ie 3957529 * 1,5
[ ]
170 * 86.4 + 22.8 * 92.3 + 1,5 * 90.4 * 45.2 = 66.1 MPa .
The bending moment is negligible near to the bearing points; therefore, we can consider the panels as being
loaded with pure shear with the edges simply supported at the uprights and the fastener rows.
2
b  90 
a = h = 178 mm. b = 90 mm. = 0, 5 ⇒ k s = 3, 8 *   + 5, 35 = 6, 3 see §V1-2.2.3.1.2
a  178 
2 2
ηk sπ 2 E  e  6,3 * π 2 * 70300  1,5
τ cr,0 =   =η   = η * 113.5 MPa = 107 MPa (h = 0.943).
( )
12 1 − ν 2e  b  12 * (1 − 0,332 )  90 
107
The Reserve Factor is: R.F. = = 1.62
66,1
This value is fairly correct as the real boundary condition for the web is certainly more stabilising than the
one taken, that is simple supported.

CHECKING END SPANS WITH A WEB THICKNESS OF 1,5 MM.

The most critical webs are those of pockets 7 and 22. They are submitted to a shear load of 17125 N
superimposed on a bending moment in the centre of the pocket equivalent to: 17125 * 585 = 10 018 125 mm.N.
As the web is clamped between the sections of the upper flange and those of the lower flange, the buckling
of the web will occur near to the lower edge of the upper flange sections, that is, at a distance of 73 mm from the
neutral fibre.
M * v 10 018 125 * 73
The bending stress is: σ f = f = = 184.8 MPa
I 3 957 529
2
ηk f π 2E  e 
The critical buckling stress in bending: σ cf,0 = 2   .
12 1 − νe  b  ( )
The height of the pocket h=b = 162 mm. The length of the pocket = spacing between uprights = a = 90 mm.
a 90
= = 0, 55 ⇒ k f = 41, 8 (see §V1-2.2.3.1.3) pocket with two sides clamped and two others bearing.
b 162
2
η * 41,8 * π 2 * 70300  1,5 
σ cf,0 =   = η * 232.5 MPa
(
12 * 1 − 0,332 )  162 
The plasticity correction factor for a plate with two clamped edges and two others hinged is given in
§ V1-2.2.2 : η = η4 = 0.856 henceσ cf,0 = 199 MPa .
The shear stress at pocket 7 or 22 is the same as the one calculated at the bearing points, as the shear force is
the same.
Thus τ = 66.1 MPa ; τ cr,0 = 107 MPa.
The interaction equation for the combined bending - shear loads is: (see § V1-2.2.4.1 )
2 2 σ 184.8 τ 66.1
R f + R s = 1 with R f = f = = 0,93 and R s = = = 0.62
σ cf,0 199 τ cr,0 107
1 1
R.F. = = = 0,9
R 2f + R 2s 0.93 + 0.62 2
2
CHECKING SPANS IN CENTRE OF BEAM WITH A 1.8 MM WEB.

The bending moment at the centre of a pocket adjacent to the centre of the beam (pockets 14 and 15) is
equal to:
M f = 17125* 1215 − 11000 * 585 = 14 371 875 mm.N.
The shear force on the same pocket is equal to: T = 17125 − 11000 = 6125 N
M * v 14 371 875 * 73
The bending stress is: σ f = f = = 253.6 MPa
I 4 136 758
The shear stress is:
T 6125
τ = ∫ ydS =
Ie 4136758 * 1.8
[
170 * 86.4 + 22.8 * 92.3 + 1.8 * 4086 = 19.9 MPa ]
2
η * 41.8 * π 2 * 70300  1.8 
The critical buckling stress in bending: σ cf,0 =   = η * 334.8 MPa
(
12 * 1 − 0.332 )  162 
The plasticity correction factor for a plate with two clamped edges and two others hinged is calculated as per
§ V1-2.2.2 : η = η4 = 0.708 hence σ cf,0 = 237 MPa .
The critical shear stress:
2 2
ηk s π 2 E  e  6,3 * π 2 * 70300  1.8 
τ cr,0 =   =η   = η * 163.5 MPa = 133.2 MPa ( η = 0.815).
( )
12 1 − ν 2e  b  12 * (1 − 0.332 )  90 

The interaction equation for the combined bending - shear loads is: (see § V1-2.2.4.1 )
σf 253.6 τ 19.9
R 2f + R 2s = 1 where R f = = = 1,07 and Rs = = = 0.15
σ cf ,0 237 τ cr,0 133.2
1 1
R.F. = = = 0,93
R 2f + R 2s 1.07 2 + 0.152
CHECKING RIGIDITY OF UPRIGHTS.

The web is stiffened on a single side by the uprights: sections with legs of equal size (mm) 13,5*13,5*1.
The formula of § 5 gives the moment of inertia required for an upright to prevent buckling.
4 4
2, 29 l  Tcr h a  3 2, 29 * 90  17125* 196, 8 3 4
Im =   =   = 204 mm
e  33 E  1, 5  33* 70300 
The moment of inertia of the uprights is 460 mm4, therefore greater than the moment of inertia required.
Such uprights are therefore satisfactory.
The moment of inertia required will also be checked by using the curve on page V1-5.5.2:
l l 90
The parameter on the X-axis is: = 0,5 = 0 ,5 = 0.21
h * Ks0 ,5
 π2 * k   2 
π * 6, 3
h * s
178* 
 1 − ν 2 * 12   1 − 0, 332 *12 
 (e )  (  )
Im 3 4
From the curve, we obtain ≈ 1, 4 ⇒ Im = 1, 4 * 90 *1, 5 = 425 mm .
l e3
The rigidity of the uprights is therefore satisfactory.
CHECKING RIVETS ATTACHING THE WEB TO THE FLANGES, END SPANS.
The shear load T = 17125 daN

In the end spans
Web thickness e = 1,5 mm
The shear flow is:

T 17125
by approximate formula: f = = = 85.6 N/mm
H 200
T 17125
by exact formula: f = ∫ ydS = (85 * 2 * 86.4 + 19 * 2 * 92.3) = 78.7 N/mm
I 3957529
All rivets are made of aluminium alloy 2117-T3, diameter 3,2 mm with a pitch of 25 mm and a shear
strength of 180 MPa.
 π * d2 
The allowable single shear force is then: Fadm cis = 180*   = 1448 daN
 4 
As the web-flange junction rivets are submitted to double shear: Fadm cis = 2896 daN
The maximum force in a rivet is: f * pitch = 85.6 * 25 = 2140 daN, which gives R.F = 1.35.
The rivets attaching the skin to the upper flange must have a pitch so as to prevent all buckling of the skin
between the rivets as we retained a certain efficiency of the skin in the beam moment of inertia calculation.
However, the rivets attaching the skin to the lower flange need not be checked for buckling between rivets, as the
skin is tensioned, but simply for shear loading.
The uprights can be checked at the external load application points and at the beam bearing points as they
transfer the concentrated loads to the web of the beam.

V1-5 7 STABLE WEB BEAMS WITH ROUND LIGHTENING HOLES

V1-5 7.1 GENERAL
The type of web analysed in the previous paragraphs includes many elements (uprights) to obtain the
required lightness. To reduce manufacturing costs, the number of parts can be reduced. A trade-off must be found
between the manufacturing costs and excess weight. This trade-off often depends on the type of aircraft under
calculation. Thus, reducing the number of elements enables extensive savings in manufacturing costs and can thus
be more economical.
To reduce the number of uprights, we frequently use a web drilled with round flanged lightening holes with
varied spacing.
However, there is a general limitation to this as we must place an upright at locations where high loads are
applied to the beam.
Webs with holes provide many accesses for hydraulic and electrical conduits.
Load
b
A A
Flange Upright under concentrated load
d C B
C’ H
e
D
g
Section A-A
V1-5 7.2 STANDARDISATION OF FLANGED HOLES
Design standards at Aérospatiale define the geometry of flanged holes according to the diameter of the hole
(Standard ASN 451.01). Thus, we will give below the values of: d, D, H (in mm).
e = 0,6 - 0,8 - 1 - 1,2 - 1,4 e = 1,6 - 1,8 - 2 - 2,5

d D H d D H
15 20 3 15 22,7 4
25 30 3 25 32,7 4
30 36 3 30 37,7 4
40 45,5 3 40 48,3 4
50 55,5 3 50 58,3 4
60 65,8 3,5 60 68,7 4,5
70 75,8 3,5 70 78,7 4,5
80 85,8 3,5 80 88,7 4,5
90 96,1 4 90 98,9 4,5
100 106,3 4 100 109 5
110 120 5 110 123 6
120 133,6 6 120 136,8 7
130 143,8 6 130 146,9 7
140 153,6 6 140 157 7
150 163,7 6 150 166,9 7
160 173,6 6 160 177 7
Issue 1 Stable Web Beams With Round Lightening Holes page V1-5•7–1
V1-5 7.3 ALLOWABLE SHEAR FLOW
The rule-of-thumb formula giving the allowable shear flow (in relation to failure) for webs with these types
of holes is:
 d  C′

( )
fadm = k × e × τ h 1 − (d / b) 2 + τ c ×
b  g
k = 0.85 − 0.0006 × (b / e)
τ h or τ c = shear stress producing the collapse of a long plate with width h or C and thickness e
(obtained for instance by using the curves on the following page).
g = distance between the centrelines of two adjacent holes.
b = height of the web between the lower and upper fastener rows (web - flange fastener).
D−d
B= and C = g - d
2
C’ = C - 2B = C + d - D = g - D
In general, we find, with the rule-of-thumb formula, that the largest holes (diameter around 0.8 h) with a
fairly large spacing g give lighter but less rigid webs.
In addition to the calculation above (collapse), it must also be checked at the holes that the web under
ultimate load f × b does not give net shear stresses above the breaking shear τ R .
Calculation graphs can be plotted from the rule-of-thumb formula (see example on page V1-5.7-4).
Zh, Zc en MPa Zh, Zc en MPa
140
Issue 1
280
266 126
252 112
238 98
224 84
210 70
(e) mm 56
196 0,4
182 42
0,5
0,6
0,8
168 1,0 28
1,3
1,6
14
Stable Web Beams With Round Lightening Holes

154
NACA A.R.R. DEC., 1942
0
140
0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320
CONTRAINTES DE CISAILLEMENT DE RUINE, Zh ou Zc1 POUR LES AMES

EN ALLIAGE D'ALUMINIUM 2024
page V1-5•7–3
Issue 1
Flux de cisaillement admissible (N/mm)
157.5
h = 100
d/h = .50 h = 100
140
d/h = .80
g = 3.0 d
122.5
g = 2.25 d
105
h = 100
h = 100
87.5
175 175
70 h = 100
h = 100
g = 1.5 d 175 250

52.5 175
175 250 175
250
35
Stable Web Beams With Round Lightening Holes

250 250 250
17.5
NACA A.R.R. DEC., 1942

0
0,5 1,0 1,5 0,5 1,0 1,5 0,5 1,0 1,5
épaisseur de l'âme (mm)
page V1-5•7–4
FLUX DE CISAILLEMENT ADMISSIBLE (EFFONDREMENT) POUR DES AMES EN 2024 AVEC TROUS
D'ALLEGEMENT CIRCULAIRES PRESENTANT DES BORDS TOMBES A 45°
V1-5 7.4 REMARKS

3 options are possible:
1. No uprights: Stable web (plates of infinite length).

E.g.: Extrusions.
2. Uprights: Stable and thin iso-loaded web (plates of finite length).

The uprights must be correctly sized.
3. Flanged holes: Lightening - stabilisation trade-off.

For safety reasons, we place the uprights at the concentrated loads and in areas
with high distributed loads to avoid crushing.
V1-5 7.5 EXAMPLE

Let us consider the beam shown below (same geometry and design loads as the one of example §5)
11000 12250 11000
h = 178 Span "A" Span "B" Span "B" Span "A"
630 630 630 630
17125 17125
Let us dimension the webs drilled with round lightening holes.

Using the simplified formula, we can find the shear flux in spans A and B:
T 17125 T 17125 − 11000
fA = = = 85.6 N / mm and f B = = = 30.6 N / mm
H 200 H 200
Thus span "B" will be lighter than span "A". The introduction of an 11000 N load means that at this location
a load upright must make the junction between spans "A" and "B".
As large-diameter holes (and with lower spacings) are more efficient for a gain in weight, let us consider two
holes of diameter d ≈ 0.8 h in each span (of 630 mm) spaced 315 mm apart.
We thus obtain the following values:
h = 178 mm ⇒ d = 140 mm ⇒ D = 153, 6 mm and H = 6 mm (table page V1 - 5.7.1)
g = 315 mm ⇒ C = g - d = 315 -140 = 175 mm and C' = g - D = 315 -153, 6 = 161, 4 mm
Span "A"
Assuming a web thickness of 2 mm.
h 178 h
= = 89 ⇒ k = 0, 85- 0, 0006 * ( ) = 0, 85 - 0, 0006* 89 = 0, 8
e 2 e
h 178 C 175
From the curves on page V1-5.7.3 for = = 89 and = = 87, 5 we obtain:
e 2 e 2
τ h = 92.5 MPa et τ c = 94.5 MPa
The allowable shear flow is then:
   d 2 d   C'    140  2  140  161, 4
fadm = k * e * τ h 1 −   + τ c   * = 0, 8 * 2 *  92, 5 *  1−    + 94, 5 * * = 97, 6 N/mm
  h h  g    178   178  315
f adm 97.6
The Reserve Factor for collapse is: R.F = = = 1.14
fA 85.6
Let us check the pure shear stresses in a section at a hole:
T 17125
τ net = = = 225 MPa
(h − d ) * e ( −140) * 2
178
The breaking shear for 2 mm thick material 2024 PL T351 is 270 MPa.
τR 270
The pure shear Reserve Factor is: R.F = = = 1.2
τ net 225
Span "B"
Assuming a web thickness of 1,3 mm.
h 178
= = 137 ⇒ k = 0, 85 - 0, 0006 * (h/e) = 0, 85 - 0, 0006* 137 = 0, 77
e 1, 3
h 178 C 175
From the curves on page V1-5.7.3 for = = 137 and = = 135 we obtain:
e 1, 3 e 1, 3
τ h = 59.5 MPa and τ c = 60.5 MPa
The allowable shear flow is then:
   d 2 d   C'    140  2  
fadm 
= k * e * τ h 1 −   + τ c   * = 0.77 * 1,3 * 59.5 *  1 − 
   + 60.5 * 140  * 161.4 = 39.2 MPa
   h  h   g
 

  178  
  178  315

f adm 39.2
The Reserve Factor for collapse is: R.F = = = 1.28
fB 30.6
Let us check the pure shear stresses for a section at a hole:
T 6125
τ net = = = 124 MPa
(h − d) * e (178 −140) *1,3
τR 270
The pure shear Reserve Factor is: R.F = = = 2.18
τ net 124
Remarks
a) The stressman can use the procedure above for smaller holes and lower thicknesses to find an
arrangement making the web lighter.
b) We can see that the examples above, for calculating the allowable flows, are confirmed by the curves on
page V1-5.7.4. (the curve for the 2 mm thick web requires an extrapolation).
c) The beam equipped with vertical uprights studied in example of § 5 is much heavier than the beam
studied above.
SSM V1-7 • LUGS
CONTENTS
issue date revision
V1-7 LUGS 1 4/1999
V1-7•1 GENERAL 1 4/1999
V1-7•2 DATA PREPARATION 1 4/1999
V1-7•3 ALLOWABLE AXIAL FORCE 1 4/1999
V1-7•4 ALLOWABLE TRANSVERSE FORCE 1 4/1999
V1-7•5 STRENGTH UNDER AN OBLIQUE FORCE 0 1/1998
V1-7•6 ALLOWABLE BEARING IN BUSH 1 4/1999
V1-7•7 PIN STRESSING 1 4/1999
V1-7•8 NUMERICAL EXAMPLE 1 4/1999
V1-7 1 GENERAL 1-1

V1-7 1.1 INTRODUCTION - GENERAL PRESENTATION 1-1
V1-7 1.2 PRINCIPAL DIRECTIONS 1-2
V1-7 1.3 FAILURE MODES 1-3
V1-7 1.3.1 Failure under axial load 1-3
V1-7 1.3.2 Failure under a transverse load 1-4
V1-7 1.3.3 General information on pin stressing 1-5
V1-7 2 DATA PREPARATION 2-1

V1-7 2.1 MATERIAL PROPERTIES 2-1
V1-7 2.2 GEOMETRICAL PROPERTIES 2-1
V1-7 3 ALLOWABLE AXIAL LOAD 3-1

V1-7 3.1 ALLOWABLE TENSION LOAD 3-1
V1-7 3.1.1 Determination of the elastic over-stress factor 3-1
V1-7 3.1.2 Neuber rule 3-1
V1-7 3.1.3 Application of the Neuber rule to the allowable tension load calculation 3-2
V1-7 3.1.4 Calculation of εs 3-3
V1-7 3.2 ALLOWABLE SHEAR/BEARING LOAD 3-5
V1-7 3.3 SPECIAL CASE CONCERNING EYE ENDS 3-7
V1-7 3.4 BLOCK DIAGRAM 3-11
V1-7 3.4.1 Solid lugs 3-11
V1-7 3.4.2 Eye ends 3-12
V1-7 4 ALLOWABLE TRANSVERSE LOAD 4-1

V1-7 5 STRENGTH UNDER AN OBLIQUE LOAD 5-1
Issue 1 General page V1-7•i

© Copyright AEROSPATIALE 1998
SSM V1-7 • LUGS
V1-7 6 ALLOWABLE BEARING 6-1

V1-7 7 PIN STRESSING 7-1
V1-7 7.1 PIN SHEAR 7-1
V1-7 7.2 PIN BENDING 7-1
V1-7 7.2.1 Calculation of the lever arm "b" 7-1
V1-7 7.2.2 Calculation of the bending modulus 7-3
V1-7 7.2.3 Allowable bending load 7-12
V1-7 8 INTERACTIVE CALCULATION 8-1

V1-7 8.1 PRINCIPLE 8-1
V1-7 8.2 EFFECT ON ALLOWABLE AXIAL LOAD 8-2
V1-7 8.3 EFFECT ON ALLOWABLE TRANSVERSE LOAD 8-3
Issue 1 General page V1-7•ii

SSM V1-7 • LUGS
SYMBOLS USED
Abbreviation Meaning
CG centre of gravity
UL ultimate load
LL limit load
RF reserve factor
Notation Unit Description

a (mm) distance from centre of hole to edge of lug in axial direction
A (mm²) cross section area of eye end ring
Aav (mm²) weighted surface area for calculation of lug subjected to transverse force
Abr (mm²) projected bearing surface area
Abrb (mm²) minimum bearing surface area at the pin/bush/bore interface
As (mm²) sheared surface area
At (mm²) critical tensile cross section area
A1 (mm²) cross section area of lug subjected to transverse force in a plane 45° to pin hole axis
A2 (mm²) 1) (axial strength of eye ends) cross section area of ring at point ②
2) (transverse strength, all lugs) cross section area of lug subjected to transverse
force at pin hole
A3 (mm²) 1) (axial strength of eye ends) cross section area of ring at point ③
2) (transverse strength, all lugs) minimum radial cross section area
A4 (mm²) cross section area of lug, taken opposite force in transverse direction in a plane 45°
to pin hole
b (mm) lever arm
C (mm) curved abscissa (eye ends)
d (mm) pin diameter
di (mm) pin inside diameter
D (mm) hole diameter
e % minimum elongation at break
E (daN/mm²) lug modulus of elasticity
Ea (daN/mm²) pin modulus of elasticity
F (daN) force applied at ultimate load
Fa (daN) projection of F (U.L.) or F/1.5 (L.L.) in axial direction
Fbru (daN/mm²) bearing stress ultimate strength (e/D = 2)
Fbry (daN/mm²) conventional bearing yield strength (e/D = 2)
Ftr (daN) projection of F (U.L.) or F/1.5 (L.L.) in transverse direction
Ftux (daN/mm²) allowable tensile limit at ultimate load
Ftyx (daN/mm²) allowable tensile limit at ultimate load
g (mm) clearance between male and female parts
I (mm4) moment of inertia of a section in relation to its own centre of gravity
I2/I3 (mm4) moment "I" at points ② and ③ for eye ends in tension
k section geometrical factor
K dimensionless factor used for eye end calculations
Kbr shear/bearing factor at ultimate load
Kbrv shear/bearing factor at limit load
Kte elastic tensile overstress factor
Ktru transverse factor at ultimate load
Ktrv transverse factor at limit load
Kt AEROSPATIALE FATIGUE MANUAL overstress factor
Kε real strain concentration factor
Kσ real overstress factor
L (mm) distance from centre of hole to edge of lug in a direction parallel to axis of
symmetry of external profile
Mf (daN.mm) pin bending moment
Issue 1 General page V1-7•iii

SSM V1-7 • LUGS
M3 (daN.mm) bending moment at point ③ for eye ends in tension
N2/N3 (daN) normal force at points ② and ③ for eye ends in tension
Padm (daN) allowable oblique force on lug
Pbru (daN) allowable shear/bearing force at ultimate load
Pbrv (daN) allowable shear/bearing force at limit load
P'brv (daN) allowable bush bearing force
Pfu (daN) allowable bending force at ultimate load
Pfv (daN) allowable bending force at limit load
Psu (daN) allowable pin shear force at ultimate load
Ptru (daN) allowable transverse force at ultimate load
Ptrv (daN) allowable transverse force at limit load
Ptu (daN) allowable tensile force at ultimate load
Ptv (daN) allowable tensile force at limit load
Ptu2 (daN) allowable tensile force at ultimate load conditioned by static strength at point ② (eye
ends)
Ptv2 (daN) allowable tensile force at limit load conditioned by static strength at point ② (eye
ends)
Ptu31 (daN) allowable tensile force at ultimate load conditioned by static strength under normal
stress at point ③ (eye ends)
Ptv3 (daN) allowable tensile force at limit load conditioned by static strength at point ③ (eye
ends)
Ptu32 (daN) allowable tensile force at ultimate load conditioned by static strength under
tangential stress at point ③ (eye ends)
Pu (daN) allowable axial force at ultimate load
Pv (daN) allowable axial force at limit load
r (mm) 1) neutral fibre radius of an eye end ring
2) (peaking) dimensionless quantity used for calculating γ
R (mm) lug nose radius
R3 ratio of Fa over Pu (U.L.) or Fa over Pv (L.L.)
Rtr ratio of Ftr over Ptru (U.L.) or Ftr over Ptry (L.L.)
t (mm) thicknesses of lug parts
t'1/t'2 (mm) reduced active thicknesses of lug
T3 (daN) shear force at point ③ for eye ends in tension
v (mm) distance between CG and extreme fibre of a section
v3 (mm) dimension v at point ③ for eye ends in tension
W (mm) lug width
β, ϕ, ψ (°) angles used in calculating eye ends subjected to axial force

γ peaking coefficient
εMAX maximum strain corresponding to σMAX
εeMAX maximum allowable elastic strain
εpMAX maximum allowable plastic strain
εmax maximum strain corresponding to σmax
εnom strain corresponding to σnom
εs strain corresponding to start of striction
ε0.2 strain corresponding to σ0,2
θ (°) angle between axial direction and direction of application of F
σapp (daN/mm²) bending stress in pin at extreme fibre
σb (daN/mm²) bending modulus
σbrg (daN/mm²) bearing stress in lug
σbru (daN/mm²) allowable bearing stress
σbrv (daN/mm²) 0.2% bearing yield strength
σe max (daN/mm²) maximum elastic tensile stress
σmax (daN/mm²) maximum real tensile stress in lug
σMAX or σM (daN/mm²) maximum bending stress in pin (at neutral fibre)
σnom (daN/mm²) nominal tensile elastic stress in the critical section of lug
Issue 1 General page V1-7•iv

SSM V1-7 • LUGS
σR (daN/mm²) ultimate strength
σ0 (daN/mm²) Cozzone's fictitious stress at neutral fibre in pin
σ0,2 (daN/mm²) conventional tensile yield strength
σc0,2 (daN/mm²) conventional compression yield strength
σ2/σ3 (daN/mm²) normal stress at points ② and ③ for eye ends in tension
τ (daN/mm²) shear stress
τadm (daN/mm²) allowable pure shear stress
τ3 (daN/mm²) tangential stress at point ③ for eye ends in tension
Index Meaning
1 characteristic index of female part of assembly
2 characteristic index of male part of assembly (except eye ends)
Issue 1 General page V1-7•v

SSM V1-7 • LUGS
Page intentionally left blank
Issue 1 General page V1-7•vi

SSM V1-7 • LUGS
REFERENCES
1 - AIRFRAME STRUCTURAL DESIGN, Nyu, Michael C. Y.
2 - MANUEL FATIGUE AEROSPATIALE MTS 005
3 - LUGS AND SHEAR PINS, Melcon, M. A. & Hoblit, F. M. in Product Engineering, June 1953
4 - FORMULAS FOR STRESS, STRAIN AND STRUCTURAL MATRICES, Pilkey, Walter D.
5 - ANALYSIS & DESIGN OF FLIGHT VEHICLE STRUCTURES, Brühn, E. F.
6 - STRESS MEMO No. 1e, Lockheed-California Company.
7 - SUPERSEDES SECT. 7.7, Canadair.
Issue 1 General page V1-7•vii

SSM V1-7 • LUGS
Issue 1 General page V1-7•viii

SSM V1-7 • LUGS
V1-7 1 GENERAL
V1-7 1.1 INTRODUCTION - GENERAL PRESENTATION
Traditionally, lug junctions are used in a certain number of cases:
- attachment of engine pylons under the wing
- rods inside wing boxes
- inter-rail floor beams
- floating floor junctions with adjacent zones
- door fittings
- galley upper attachments
- etc.
Generally, a lug-type junction comprises:

- a male part
- a female part
- a pin
More rarely, a junction consisting of two male parts is found.
The bores in the male and female parts may be fitted with bushes to avoid direct contact with the pin.
Female part of the lug junction
Pin
Male part of the lug junction
Figure V1-7 .1.1-1: Example of a lug assembly
As these junctions often transfer single forces, special care must be paid to sizing, especially by taking into
account the fitting factor, imposed by certification texts. Most often, the following is recommended:
- take either a fitting factor of 1.15 which increases the force applied:
- or a margin of at least 20%.
Issue 1 General page V1-7•1-1

SSM V1-7 • LUGS
Note that these special precautions apply to all items for which mechanical strength is not guaranteed by tests.
This may concern the male and female parts as well as the pin.
A calculation at ultimate load and a calculation at limit load are made. The limit load calculation is especially
important when the removability of the junction must be conserved for maintenance operations.
V1-7 1.2 PRINCIPAL DIRECTIONS

Before calculating this type of junction, it is essential to define two main force directions.
In the simple case where the hole is centred on the bisector of the external profile of the lug (refer to figure V1-
7 .1.2-1):
- the axial direction is defined as being this bisector,
- the transverse direction is defined as being perpendicular to the axial direction.
Bisector
Axial direction
Figure V1-7 .1.2-1: Lug with centred hole
In certain cases, the hole is not located on the bisector. Depending on the distance between the centre of the bore
and the edge of the part, the axial direction may be plotted either in a direction parallel to the bisector or in the
minimum material direction (refer to Figure V1-7 .1.2.2).
R R
Bisector Bisector
Axial direction
Axial
L L
direction
Figure V1-7 .1.2-2: Lugs with asymmetrical holes

SSM V1-7 • LUGS
Example for determining axial direction: Axial
adapter beam end lug direction
Tangent to the
outer profile
Outer
profile
Bisector
Figure V1-7 .1.2-3: Example for determining axial direction
In the example of Figure V1-7.1.2-3, the hole is located on the bisector. Therefore, the axial direction coincides
with this bisector. The transverse direction is perpendicular to the axial direction, orientated in the orthogonal
projection direction of the applied force, downwards or upwards.

V1-7 1.3.1 Failure under axial force
The following may occur under an axial force:
- tensile failure: the concentration of stresses at the edge of the hole cause failure in a section
approximately perpendicular to the axial direction and located on either side of the hole (Figure V1-
7.1.3.1-1).
Figure V1-7 .1.3.1-1: Tensile failure under axial force
- failure by a combination of bearing and shear: the bearing and shear stresses combine to cause
failure in planes located at ± 40° to the axial direction (Figure V1-7.1.3.1.2).
40° F
40°
Figure V1-7 .1.3.1-2: Bearing/shear failure

SSM V1-7 • LUGS
In the specific case of eye ends, failure under axial load occurs generally at the junction between the lug ring and
the remainder of the part (refer to Figure V1-7.1.3.1-3). Essentially, this is a failure caused by annular tensile
stresses.
Figure V1-7 .1.3.1-3: Axial failure of an eye end
V1-7 1.3.2 Failure under transverse force

The failure mode is more complex under a transverse force. Failure is the combination of several modes,
including:
- material shear between the hole and the outer edge of the lug, where the force is applied,
- local bending of the same segment of material,
- circumferential tension in the annular ring of material between the hole and the outer edge of the lug.
Furthermore, a weakness in one of these modes is compensated for by all the others. Locally the load paths may
vary. The generic term "transverse failure" is used to designate this combination.
F
F
Shear at load Bending + tensile failure due to

annular stress
Figure V1-7 .1.3.2-1: Examples of transverse failures
Generally, the force is not purely axial or purely transverse. Therefore, it is called an "oblique force"
comprising both an axial component and a transverse component.
The calculation procedures given here make it possible initially to determine the allowable axial and
transverse forces, with the possibility of combining them to calculate a margin for an oblique force.
Then, we calculate the strength of the lug pin.

SSM V1-7 • LUGS
V1-7 1.3.3 General information on pin stressing

The pin is substantiated in shear and bending.
The calculation of the real lever arm for pin bending takes the "peaking" phenomenon into account: bending
strain causes a concentration of contact pressures close to the outer faces of the male part and the inner faces of
the female part.
F/2 F/2 F/2 F/2
Female part
Inactive sections
on thicknesses
Male part
F F
UNIFORM DISTRIBUTION CONCENTRATION OF CONTACT
OF CONTACT PRESSURES PRESSURES (PEAKING)
Fig. V1-7 .1.3.3-1: Distribution of contact pressures in the lug
The bending calculation also takes into account the plasticization of the extreme fibres of the pin using the
Cozzone method.

SSM V1-7 • LUGS

SSM V1-7 • LUGS
V1-7 2 DATA PREPARATION

V1-7 2.1 MATERIAL CHARACTERISTICS
The following information on the male and female parts of the lug junction is required:
- the ultimate tensile strength: σR
- the ultimate bearing strength (e/D = 2): Fbru
- the conventional bearing yield strength (e/D = 2): Fbry
- the conventional tensile yield strength: σ0.2
- the conventional compression yield strength: σc0.2
- the modulus of elasticity in tension: E
- minimum elongation at break c(%)
As far as possible, the following shall be obtained:

- the factor "n" of the Ramberg & Osgood equation (tensile curve)
For an eye end:

- the allowable pure shear stress: τR
- the ultimate tensile strength: σR
- the conventional tensile yield strength: σ0.2
If the assembly includes a bush, record:

- the conventional compression yield stress: σc0.2
The following information is required for the pin:

- the ultimate strength: σR
- the conventional yield stress: σ0.2
- the conventional compression yield strength: σc0.2
- the allowable pure shear stress: τadm
- the modulus of elasticity in tension: Ea
- the factor "n" of the Ramberg & Osgood equation (tensile curve)
V1-7 2.2 GEOMETRICAL CHARACTERISTICS

In addition to the thickness, the dimensions to be recorded for the lug calculations are shown on Figures
V1-7 2.2-1 to -4.
R
W
Axial direction
L=a
Fig. V1-7 .2.2-1: Lug with parallel edges
Issue 1 Data preparation page V1-7•2-1

SSM V1-7 • LUGS
W/2 R
β Axial direction
W/2
L=a
Fig. V1-7 .2.2-2: Lug with converging edges
Axial direction
W/2
L=a
Fig. V1-7 .2.2-3: Lug with hole off-centred in relation to the outer profile and large edge distance (L>R)
R
W/2
Axial
L direction
Fig. V1-7 .2.2-4: Lug with hole off-centred in relation to the outer profile and small edge distance (L<R)
The following notations are used:

- R: lug nose radius
- W: lug width (see previous figures for details)
- β: outer profile angle of lug
- L: distance from centre of hole to edge of lug, parallel to axis of symmetry of external profile
- a: distance from centre of hole to edge of lug, in axial direction
- D: hole diameter
- t: lug thickness

SSM V1-7 • LUGS
The following indices will be associated with all specific quantities either for the male part or for the
female part:
- "1" for the male part,
- "2" for the female part.
For a "male-male" assembly, index 2 concerns the second male part.
Two characteristic sections will be used for the lug calculations:

- the projected bearing section: Abr = D.t
- the net section: At = (W – D)t =  W − 1 .A br
D 
Another surface called Aav is used for calculations on lugs subjected to transverse forces; as it is complex to
determine, all the necessary details are given in § V1-7 .4.
If the assembly includes a bush, also use the bearing surface area in the bush, called Abrb, defined as follows:
Abrb = d.t
For pin calculations, the following are required:

- the diameter of the pin: d
- the maximum clearance between the male and female parts: g
For a hollow pin, the inside diameter will be called di.

SSM V1-7 • LUGS

SSM V1-7 • LUGS
V1-7 3 ALLOWABLE AXIAL FORCE

The allowable axial force at U.L. is called Pu; this is the minimum value between the allowable tensile force (Ptu)
and the allowable shear/bearing force (Pbru): Pu = min{Ptu; Pbru}
The allowable axial force at L.L., Py is determined globally from Pu.
Ftyx
P y = K y. .Pu
Ftux
Ftyx = 0.593.Fbry
F
Ftux= bru
1.9
Pu Pu
Si ≥ 1.05: Ky = 1.3625 - 0.25.
A br .Ftux A br .Ftux
Pu
Si ≤ 1.05: Ky = 1.1
A br .Ftux
V1-7 3.1 BREAKING FORCE IN TENSION

Axial forces create a concentration of the stresses at hole edge. This concentration can be sufficiently high to
plasticize the metal. Therefore, to calculate the real amplified stress, the elastic overstress factor must be
weighted with a plasticity correction.
The procedure to be followed to calculate the allowable tensile force is given below.
V1-7 3.1.1 Determining elastic overstress factor

The elastic overstress factor, called Kte, is used to calculate the overstress at hole edge on the basis of a purely
elastic behaviour of the material.
Thus: σe max = Kte . σnom
where σnom is the nominal reference stress calculated considering net section At = (W-D).t.
Refer to the fatigue manual (MTS 005 ch.III.3.8) for determining Kt: Kt = KtET.G
The reference section used in the fatigue manual is the gross section, that is Abrut = W.t
Thus:
At
Kte = .Kt
A brut
where Kt is the elastic overstress factor directly obtained from the fatigue manual graphs.
W−D
Kte = .Kt
W
V1-7 3.1.2 Neuber rule

The Neuber rule is used to calculate a real overstress in the plastic range using the reference stress status, the
elastic overstress factor and the stress - strain curve of the material.
Normally, a given force is used to calculate the real stress and possibly calculate a margin in relation to the
ultimate strength of the material.
Let:
F
- σnom be the uniform stress used as reference: σnom =
At
- and εnom the corresponding strain.
Issue 1 Allowable axial load page V1-7•3-1

SSM V1-7 • LUGS
The equation of a hyperbola is defined assuming that the product of the real stress concentration factor Kσ and the
real strain concentration factor Kε is equal to the square of the theoretical stress concentration factor (elastic) Kte,
let:
σ
Kσ . Kε = K 2te where: Kσ = max
σ nom
ε max
Kε =
ε nom
σ e max
Kte =
σ nom
σe max is a fictious amplified stress as it is calculated on the basis of a purely linear elastic behaviour of the
material.
Thus, we obtain: σmax . εmax = K 2te . σnom . εnom
Knowing Kte, σnom and εnom, the previous equality can be written: σmax . εmax = Cte
C te
This equality enables us to define an equation hyperbola: σ = . By plotting this hyperbola in the same plane
ε
as the material curve, we determine σmax and εmax as being the coordinates of the intersection point between the
two curves (refer to Figure V1-7 .3.1.2.-1).
σe max
σmax Neuber hyperbola
Material curve
σnom
εnom εmax
Fig. V1-7 .3.1.2-1: Graphic interpretation of the Neuber rule

SSM V1-7 • LUGS
V1-7 3.1.3 Applying Neuber rule to allowable tensile force calculation

Generally, for lug calculations, we try to express the capability of the lugs in terms of allowable forces which
supposes an approach opposite to the one described above.
The Neuber hyperbola passing via the coordinate point (εs; σR) cuts the straight line of slope E at coordinate point
σ e max
(σe max; ).
E
σ e max
Therefore, σR . εs = σe max .
E
Now σe max = Kte . σnom adm
σ max
The elastic-plastic overstress factor Kσ is defined as: Kσ =
σ nom
σR σR
That is here: Kσ = ⇒ σnom adm =
σ nom adm Kσ
σR
⇒ Kσ = Kte .
E . εs
The value of Kσ only has a meaning if Kσ ≥ 1. Otherwise, we consider that Kσ = 1.
Finally, the force bringing the real stress to the maximum allowable value ( σmax = σR) is:
At . σR
Ptu =
Kσ
V1-7 3.1.4 Calculating εs

εs is the strain corresponding to the start the striction, i.e. the unstable state where strain continues at a point
without it being necessary to increase the total force F. We then obtain σ = σR.
εs is not conventional material data. Nonetheless, it is possible to calculate εs from "n", the Ramberg & Osgood
equation factor:
εs ≈ 1 ≤ e (%) minimum elongation at break for the material

n
If "n" is not material data, solve the following equation:

n
σ 
n = 500 .  0.2 
 σR 

SSM V1-7 • LUGS
Page Intentionally Left Blank

SSM V1-7 • LUGS
V1-7 3.2 SHEAR/BEARING BREAKING STRESS

The allowable shear/bearing force at ultimate load is:
Fbru
Pbru = Kbr . Abr . Ftux where Ftux =
1.9
The factor Kbr depends on ratios a/D and D/t. It is determined using the graph on Figure V1-7.3.2-1.
Figure V1-7 .3.2-1: Determining factor Kbr
On the Figure V1-7 .3.2-1
- curve A is the upper limit to be taken into account for all forged aluminium alloy parts when the long
transverse direction coincides with direction "C" on the drawing of the lug,
- curve B is the upper limit to be taken into account for all aluminium alloy forgings, bars and sheets when
the short transverse direction coincides with direction "C" on the drawing of the lug and for die forged
parts when the lug contains the jointing plane in a direction approximately perpendicular to direction "C".
In addition to the limits of curves A and B, the value of Kbr will be limited to 2.0 for all aluminium alloy
forgings, bars and sheets of a thickness equal to or greater than 12.7 mm.

SSM V1-7 • LUGS
Page Intentionally Left Blank

SSM V1-7 • LUGS
V1-7 3.3 SPECIAL CASE CONCERNING EYE ENDS

Considering their specific geometry, eye ends are mainly subject to normal circumferential stresses. On Figure
V1-7 .3.3-1:
- point ② is the RH point from which the load transmitted in the ring is pure tension (normal force),
- point ③ is the point where the ring is clamped in the remainder of the part.
②
①
β
③ ϕ
Ψ F
Figure V1-7 .3.3-1: Eye end
Experience proves that the critical section of eye ends is located at ② or at ③. The first step consists in
determining the position of ②, knowing the position of ③ and according to a certain number of geometrical
parameters of the ring. This calculation requires the use of the graph on Figure V1-7 .3.3-2. The expression of
factor K required to read the graph is:
C.I
K=
A.r3
π . r . (90° + β)
where: C: circular distance (refer to Figure V1-7 .3.3-1) C=
180°
I: moment of inertia of the section at ① in relation to its own centre of gravity
A: area of the section at ①
r: radius of the neutral fibre of the ring
Issue 1 Allowable axial force page V1-7•3-7

SSM V1-7 • LUGS
Figure V1-7 .3.3-2: Determining angle ϕ

SSM V1-7 • LUGS
By writing the static equilibrium of the ring, the normal load at ② is therefore:
F
N2 =
2 . cos ϕ
By calling A2 the section area at ②, the corresponding normal stress is:
N2 F
σ2 = =
A 2 2 A 2 . cos ϕ
At ultimate load, we will obtain:
σ2 = σR
Let Ptu2 be the allowable tensile force at ultimate load conditioned by the static strength at point ②. We then
obtain from the previous equality:
Ptu2
σR =
2 A 2 . cos ϕ
Hence finally:
Ptu2 = 2A2 . cos ϕ . σR
And at limit load:
Pty2 = 2A2 . cos ϕ . σ0.2
The internal load matrix at ③ is deduced from the value calculated at ②:
- normal force: N3 = N2 . cos Ψ
- shear force: T3 = N2 . sin Ψ
- bending moment: M3 = N2 . r . (1 - cos Ψ)
The maximum normal stress at ③ is then:
N 3 M 3 . v 3 N 2 . cos ψ N 2 . r . (1 − cos ψ ) . v 3
σ3 = + = +
A3 I3 A3 I3
F  cos ψ r . (1 − cos ψ ) . v 3 
σ3 = . + 
2 cos ϕ  A 3 I3 
where: A3: section area at ③

v3: distance between centre of gravity of section at ③ and extreme fibre
I3: moment of inertia of section at ③ in relation to its own centre of gravity
As at ②, at ultimate load, we obtain:
σ3 = σR

SSM V1-7 • LUGS
Let Ptu31 be the allowable tensile force at ultimate load conditioned by the static strength under normal stress at
point ③. We then obtain from the previous equality:
Ptu31  cos ψ r . (1 − cos ψ ) . v 3 

σR = . + 
2 cos ϕ  A 3 I3 
Hence finally:
2σ R . cos ϕ
Ptu31 =
 cos ψ r . (1 − cos ψ ) . v 3 
 + 
 A3 I3 
Remark: The calculation of allowable force Ptu31 does not take into account the plasticization of the
extreme fibres of the section which would "increase" the apparent bending limit. Thus, Ptu31, calculated on
the basis of σR, is underestimated and therefore conservative.
Also at limit load:
2 σ 0.2 . cos ϕ
Pty3 =
 cos ψ r . (1 − cos ψ ) . v 3 
 + 
 A3 I3 
The shear stress at ③ at ultimate load is written:
T3 N . sin ψ F .sin ψ
τ3 = = 2 =
A3 A3 2 A 3 . cos ϕ
At ultimate load, we obtain:
τ3 = τadm
Let Ptu32 be the allowable tensile force at ultimate load conditioned by the static strength under tangential stress at
point ③. We then obtain from the previous equality:
τadm = Ptu32
Hence finally:
2 A 3 . τ adm . cos ϕ
Ptu32 =
sin ψ
Finally, the overall allowable tensile force at ultimate load is:
Ptu = min(Ptu2; Ptu31; Ptu32)
At limit load:
Pty = min(Pty2; Pty3)

SSM V1-7 • LUGS
V1-7 3.4 BLOCK DIAGRAMS
E
V1-7 3.4.1 Solid lugs
AXIAL LOADING
yes
n known?
no
Solve:
Allowable shear - bearing force at
ultimate load:
Pbru = Kbr . Abr . Ftux
S
n
σ  F ( e / D = 2)
n = 500 .  0.2  Ftux = bru
1.9
S
 σR 
+ graph on Figure V1-7 .3.2-1
1 Determine Kt from graphs in

Striction strain: εs = ≤ e(%) FATIGUE MANUAL MTS 005
n
Kte =
W−D
W
. Kt
E
Kσ = Kte .
σR
E . εs N
yes no
Kσ ≥ 1 ?
Kσ = 1
T
At . σR
Ptu =
Kσ
Allowable axial force at ultimate load:

Pu = min (Ptu; Pbru) I
Ftyx = 0.593.Fbry (e / D = 2)
yes Pu
A br . Ftux
≥ 1,.5
no A
Pu
Ky = 1.3625 – 0.25 Ky = 1.1
A br Ftux
Allowable axial force at limit load:

Ftyx
L
Py = K y . . Pu
Ftux

SSM V1-7 • LUGS
V1-7 3.4.2 Eye ends
E AXIAL LOADING
Evaluation of position of point ③ (angle β)
S Circular distance: C=
π . r . (90° + β)
180°
S Determining angle ϕ from K =

C.I
A . r3
and graph on Figure V1-7 .3.3-2
E Allowable tensile force at U.L. conditioned

by the static strength at point ②:
Ptu2 = 2A2 . cos ϕ . σR
Allowable tensile force at L.L. conditioned
by the static strength at point ②:
Pty2 = 2A2 . cos ϕ . σ0.2
N
Allowable tensile force at ultimate load Allowable tensile force at limit load
conditioned by the static strength under conditioned by the static strength at point ③:
normal stress at point ③: Pty3 = 2σ 0.2 . cos ϕ
2 σ R . cos ϕ  cos ψ r .(1 − cos ψ ). v3 
Ptu31 =  + 
 cos ψ r . (1 − cos ψ ) . v 3   A3 I3 
 + 
 A3 I3 
T Allowable tensile force at ultimate load

conditioned by the static strength under
tangential stress at point ③:
I
2 A 3 . τ adm . cos ϕ
Ptu32 =
sin ψ
A Allowable tensile force at ultimate load:

Ptu = min (Ptu2; Ptu31; Ptu32)
Allowable tensile force at limit load:
Pty = min (Pty2; Pty3)
L
SSM V1-7 • LUGS
V1-7 4 ALLOWABLE TRANSVERSE FORCE

In spite of the combination of several distinct failure modes, the transverse capability of the lug is expressed by a
single allowable force.
The allowable transverse force at ultimate load is:
Ptru = Ktru . Abr . σR
The allowable transverse force at limit load is:
Ptry = Ktry . Abr . σ0.2
Factors Ktru and Ktry are obtained from the graph, Figures V1-7 .4-2 to V1-7 .4-11. However, before using these
graphs you must calculate an equivalent section Aav, which takes the various failure modes in the transverse
direction into account. The definition is given in Figure V1-7 .4-1.
A4 R
A3 A4 45°
45°
45°
A1 45°
A1 A2
Transverse direction
= Direction of force
CONVENTIONAL LUG EYE END
Figure V1-7 .4-1: Definition of the equivalent cross section Aav
6
We can write: Aav =
3 1 1 1
+ + +
A1 A 2 A 3 A 4
- A2 is measured at the hole in the transverse force direction

- with a conventional lug: A1 and A4 are measured parallel to the transverse direction along lines offset by
45° (refer to Figure V1-7 .4-1).
- with an eye end: A1 and A4 are measured perpendicularly to the circular neutral line of the lug (refer to
Figure V1-7 .4-1).
- A3 is the smallest radial cross section around the hole.
A av
To read the graph and therefore determine Ktru and Ktry, parameter is used which represents the shape of
A br
lug. Note that unlike Ktru, Ktry does not depend on the material concerned.
Issue 1 Allowable transverse force page V1-7•4-1

SSM V1-7 • LUGS
Ktru or Ktry
Figure V1-7 .4-2: Determining factors Ktru and Ktry

Interpretation of the curves above makes reference to Table V1-7 .4-3. Note that these curves only apply if the
short transverse direction is not in the plane of the lug.
Curve Material concerned

1 steel 25CD4S and 25CD45 86 daN/mm²
3 Ktry for all materials
5 2024-T3 and -T4 plates, thickness ≤ 12.7 mm
6 2014-T6 and 7075-T6 plates, thickness ≤ 12.7 mm
7 2024-T3 and -T4 plates, thickness > 12.7 mm, 2024-T4 bar
8 cast 356-T6
9 2024-T6 and 7075-T6 plates, thickness > 12.7 mm, ≤ 25.4 mm
7075-T6 extrusions
2014-T6 forgings, cross section ≤ 23224 mm²
2014-T6 and 7075-T6 die forged parts
10 2024-T6 plates
2024-T4 and 2024-T42 extrusions
11 2014-T6 and 7075-T6 plates, thickness > 25.4 mm
7075-T6 forgings, cross section ≤ 10322 mm²
12 7075-T6 forgings, cross section > 10322 mm²
2014-T6 forgings, cross section > 23224 mm²
Table V1-7 .4-3: References for curve .V1-7.4-2

SSM V1-7 • LUGS
1.7 1
1.6
2
1.5
1.4
1.3 3
1.2 4
1.1
1.0
Ktru or Ktry 0.9
0.8 5
0.7
0.6
0.5
A
0.4
0.3 6
0.2
0.1
0
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4
A ay
A br

Interpretation of the curves above makes reference to table V1-7 .4-5.

1 steel < 87 daN/mm²
2 steel 105 daN/mm²
3 Ktry for all materials
Table V1-7 .4-5: References for curve V1-7 .4-4

SSM V1-7 • LUGS
1.0
.8 1
A
Ktru 2
.6
4
3
5
.4 6
7
.2 A
0
0 .2 .4 .6 .8 1.0 1.2 1.4
A ay
A br
Figure V1-7 .4-6: Determining factors Ktru


1 Ti-6A1-4V die forged annealed (LT), thickness ≤ 127 mm
Ti-6A1-4V forged annealed (LT), cross section ≤ 10323 mm²
Ti-6A1-4V forged (L, LT), thickness ≤ 51 mm
2 Ti-6A1-4V forged annealed (LT), cross section > 10323 mm²
Ti-6A1-4V die forged (L), thickness ≤ 127 mm
Ti-6A1-4V die forged (LT), thickness ≤ 25 mm
Ti-6A1-4V forged (LT), thickness > 51 mm, ≤ 76 mm
3 Ti-6A1-4V die forged (LT), thickness > 25 mm, ≤ 76 mm
4 Ti-6A1-6V-2Sn plate annealed (LT), thickness ≤ 51 mm
Ti-6A1-6V-2Sn die forged annealed (ST), thickness ≤ 51 mm
Ti-6A1-6V-2Sn forged annealed (LT), thickness ≤ 51 mm
5 Ti-6A1-6V-2Sn plate annealed (LT), thickness > 51 mm, ≤ 101 mm
Ti-6A1-6V-2Sn die forged annealed (ST), thickness > 51 mm, ≤ 101 mm
Ti-6A1-6V-2Sn forged annealed (LT), thickness > 51 mm, ≤ 101 mm
6 Ti-6A1-6V-2Sn die forged (L)
Ti-6A1-6V-2Sn forged (L, LT), thickness ≤ 101 mm
7 Ti-6A1-6V-2Sn die forged (LT)
Ti-6A1-6V-2Sn forged (LT), thickness > 101 mm

SSM V1-7 • LUGS
1.2
1.0 2
Ktru
.8
1
A
.6
.4
3
.2
0
0 .2 .4 .6 .8 1.0 1.2 1.4
A ay
A br
Figure V1-7 .4-8: Determining factors Ktru

1 steel 300M with e% ≥ 6
2 low alloy steels with Ftu ≤ 210 daN/mm² (except 300M)
3 low alloy steels with 3 ≤ e% ≤ 6

SSM V1-7 • LUGS
1
1.4
2
1.3
1.2
1.1 3
1.0 4
.9
Ktru
.8 A
.7 5
.6
.5
.4
.3
A
.2
.1
0
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1.0 1.1 1.2 1.3 1.4
A av
A br


1 steel 15-5PH H1150 with Ftu = 95 daN/mm² and e% = 11 (T)
3 Inconel steel 718 B&F Sta with Ftu = 126 daN/mm² and e% = 10 (LT)
5 Inconel steel 718 B&F Sta with Ftu = 126 daN/mm² and e% = 6 (T)

SSM V1-7 • LUGS
If the material is not in the list given in Tables V1-7 .4-3, V1-7 .4-5, V1-7 .4-7, V1-7 .4-9 or V1-7 .4-11, an
analogy may be made based on:
- the same type of finished product (plate, extrusion, bar, forging, etc.),
- a similar ductility (in particular, expressed by the elongation at break),
- a similar σR/σ0.2 ratio.
If no analogy can be found, a finite element analysis may be used in compliance with the following rules:
- take into account any possible anisotropy due to the manufacturing process and correct orientation of the
model in relation to the longitudinal, long transverse and short transverse directions (the results of the study
above prove that this parameter has an affect on the transverse strength). If the direction of these fibres is
not known, the long transverse direction may be placed unfavourably parallel to the transverse
direction. Remember that the plane of the lug must not contain the short transverse direction,
- nonlinear calculation in the plastic domain at ultimate load.
Globally, if the dimensions of the hole are small compared with the lug, this being quantified by the ratio Aav/Abr,
the local transverse failure phenomena become negligible in relation to the overall bending strength of the lug.
This transition is approximately indicated by curve A on graphs V1-7 .4-2 to V1-7 .4-11.
If Ktru is below this curve, it is advisable to carry out a fixed end beam calculation using a cross section as shown
on Figure V1-7 .4-12. Only take the lower section of the lug into account.
Depending on the shape and the material of the lug, Ktru may be very conservative. Ptru is in no case lower than
the ultimate bending load of the ligament under the transverse load.
Figure V1-7 .4-12: Approximate portion of the lug acting as a fixed end beam
The approach consists in calculating the inertia and bending moment in the most critical cross section, moving
away from the hole (force application point - refer to Figure V1-7 .4-13).
ji
n i 21
Figure V1-7 .4-13: Calculation of bending moment in lug flange
For a critical cross section "i", the maximum bending stress is:
M f i . vi F . ji . v i
σi = =
Ii Ii

SSM V1-7 • LUGS
where:
Mfi: bending moment due to F in cross section "i"
vi: distance between the centre of gravity and the extreme fibre of cross section "i"
ji: lever arm of Mfi
Ii: moment of inertia of cross section "i" in relation to its own centre of gravity
Taking as condition σiMAX = σR , the allowable force at U.L. is therefore:
 I 
Ptru = σR . min  i 
 j i . v i  i =1.. n
And at L.L.
 I 
Ptry = σ0.2 . min  i 
 j i . v i  i =1.. n

SSM V1-7 • LUGS
TRANSVERSE
LOADING
yes
Material referenced?
no
E
yes no
Analogy?
Determine A1, A2, A3 and A4, then Aav

R
Finite element analysis
S
A4
A3 A4 45°
45°
S
where Aav
45°
A1 45°
3 1 1 1 A1 A2
+ + +
A1 A 2 A 3 A 4
Transverse direction
CONVENTIONAL LUG EYE END
Interpretation of graphs:
E
V1-7 .4-2 to V1-7 .4-11
N
yes no
T
Fixed end beam?
Ji
n i 21
F
Determination of Ktru and Ktry I
 I 
Allowable force at U.L. Ptru = σR . min  i 
 j i . v i  i =1.. n
 I 
Allowable force at L.L. Ptry = σ0.2 . min  i 
 j i . v i  i =1.. n
Allowable force at U.L. Ptru = Ktru . Abr . σR
Allowable force at L.L. Ptry = Ktry . Abr . σ0.2

A
L
SSM V1-7 • LUGS

SSM V1-7 • LUGS
V1-7 5 STRENGTH UNDER AN OBLIQUE FORCE
E
In order to check the strength of the lug under an oblique force, the force must be projected in the transverse and
axial directions as shown on Figure V1-7 .5-1.
Transverse
direction
S Axial direction Fa
θ
F
Ftr
S Figure V1-7 .5-1: Lug under an oblique force
Notation:
Fa = F . cos θ the axial component of the force (at U.L.)
E Ftr = F . sin θ the transverse component of the force (at U.L.)
We consider that the force at limit load is obtained by dividing the force at U.L. by 1.5: thus, we obtain:
F
Fa = . cos θ the axial component of the force (at L.L.)
1.5
N Ftr =
F
1.5
. sin θ the transverse component of the force (at L.L.)
In all cases, the interaction equation between the axial and transverse loads is:
T R 1a.6 + R 1tr.6 = 1
where:
(document reference (5) §D.1.8)
Fa Fa
Ra = (at ultimate load) Ra = (at limit load)
Pu Pbry
I Rtr =
Ftr
Ptru
(at ultimate load) Rtr =
Ftr
Ptry
(at limit load)
We can therefore calculate a "reserve factor" for this combined loading using the interaction equation. Finally, we
A
obtain:
1
R.F. =
(R )
1.6 0 .625
a + R 1tr.6
L
Issue 1 Strength under an oblique load page V1-7•5-1
SSM V1-7 • LUGS
OBLIQUE LOADING
Ultimate load calculation Limit load calculation
Axial and transverse projection of F: Axial and transverse projection of F:

Fa = F . cos θ Fa = F/1.5 . cos θ
Ftr = F . sin θ Ftr = F/1.5 . sin θ
Comparative ratios: Comparative ratios:

F F
Ra = a Ra = a
Pu Pbry
F Ftr
Rtr = tr Rtr =
Ptru Ptry
Calculation of "reserve factor"

1
R.F. =
( )
0 .625
R a + R 1tr.6
1.6
Issue 1 Strength under an oblique load page V1-7•5-2

SSM V1-7 • LUGS
V1-7 6 ALLOWABLE BEARING IN BUSH

V1-7 6.1 ALLOWABLE BEARING IN LUGS
The bearing cross section involved (Abrb) will be the minimum area between:
- the bearing area between the pin and the bush,
- the bearing area between the bush and the lug.
The latter may be smaller if, for example, the bush includes an outside chamfer.
The allowable limit force for this type of damage is:
P'bry = 1.85 . σc0.2 . Abrb
with:
σc0.2: 0.2% compression yield strength for the bush material
P'bry will be directly compared with the oblique force at limit load.
Issue 1 Allowable bearing page V1-7•6-1

SSM V1-7 • LUGS
Issue 1 Allowable bearing page V1-7•6-2

SSM V1-7 • LUGS
V1-7 7 PIN STRESSING
V1-7 7.1 PIN SHEAR
The shear stress in the pin can be written:
F
τ=
As
where
As π d2
: sheared surface area of the pin, such that A s = .
4
d: pin diameter.
The calculation of the corresponding allowable forces is given below.
For a male / female part assembly:

π d2
Psu = τ adm × 2 ×
4
(double shear)
π d 2 τ adm
That is, Psu =
2
Where: τ adm : allowable pure shear stress.
Remark: Shear is only calculated at ultimate load.
V1-7 7.2 PIN BENDING

The bending of the lug pins must be studied both at the limit load and at the ultimate load. Indeed, permanent
sets are unacceptable at limit load as they hinder or even prevent the disassembly of the junction. The allowable
bending forces at ultimate load and at limit load are called Pfu and Pfy respectively.
The maximum bending stress will be located at the extreme fibre of the lug pin. By calling σ app the stress applied
(due to bending), we obtain:
Mf ⋅ v Mf ⋅ d b ⋅ F ⋅ d
σapp = = =
I 2I 4I
V1-7 7.2.1 Calculation of lever arm "b"

Conservatively, the calculation of the lever arm to determine the bending moment is often based on the assumption
that the contact pressure between the pin and the bores is constant over the length of the pin.
In actual fact, due to bending strain, the resisting force exerted by the male portion of the lug is concentrated close
to the outer faces and the resisting force exerted by the female portion near to the inner faces (see figure
V1-7 .7.2.1-1). This phenomenon, called "peaking", reduces the lever arm which reduces the bending moment in the
pin.
Moreover, we have demonstrated by tests (document ref. (3)) that the "peaking" does not reduce the strength of the
lugs.
Issue 1 Pin stressing −1

page V1-7•7−
 Copyright AEROSPATIALE 1998
SSM V1-7 • LUGS
t1
g
F/2 F/2 F/2 F/2
Female part
b b Inactive portions of
the thicknesses
Male part
F F
t2
UNIFORM DISTRIBUTION OVER ALL DISTRIBUTION TAKING PEAKING INTO ACCOUNT

THICKNESSES
Figure V1-7 .7.2.1-1: Distribution of the contact pressure between the pin and the bores
In the first case shown on figure V1-7 .7.2.1-1, the lever arm is:
t1 t 2
b= + +g
2 4
where:
t 1: thickness of one of the sides of the female portion of the lug
t 2: thickness of the male portion of the lug
g: clearance between the male and female parts
Taking "peaking" into account gives a new expression for the lever arm:
t t 
b = γ  1 + 2 +g
2 4
t1 t 2
For a "male - male" assembly, the lever arm can be written: b= + +g
2 2
Determining factor γ:
 D
a − 2 
First of all, the quantity r=  . must be determined. "a" is then measured in the direction of the
t2
applied force whether the load is purely axial, purely transverse or oblique.
Padm
It is also necessary to evaluate the ratio called x.
A br ⋅ Ftux
Padm is defined as being the allowable force by writing, in the general case of an oblique force:
F
Padm =
(
R1a,6 + R1tr,6 )
0,625
We then obtain γ from the graph on figure V1-7 .7.2.1-2 or from the following equations:
 r 
 x ≤ 1− 1− 
 0,55  x
  ⇒ γ=
 r  2
 x ≥ 1+ 1− 
 0,55 
  
 1 − 1 − r ≤ x ≤ 1 + 1 − r  ⇒ γ = 1 − r
  0,55 0,55  1,1 x
 

page V1-7•7−
SSM V1-7 • LUGS
Figure V1-7 .7.2.1-2: Determining peaking factor γ
V1-7 7.2.2 Calculating bending modulus
We can see that an item subjected to bending is capable of supporting a load greater than the calculated load
assuming a linear distribution of the stresses over the height of the section. This is due to the fact that the extreme
fibres plasticize which modifies the normal stress profile. Therefore, the real stress in these fibres is lower than that
obtained by a simple linear calculation (refer to Figure V1-7 .7.2.2-1).
zσ z
MAX
z σMAX
x x
y
x σMAX σMAX
STRESS CALCULATED BY REAL STRESS
LINEAR METHOD
Figure V1-7 .7.2.2-1: Bending stress
On the figure above, σMAX is the stress applied to the extreme fibre. The Cozzone method is based on an
approximation of the real stress profile. The real stress profile shown on Figure V1-7 .7.2.2-1 is replaced by a
trapezoidal profile as shown on Figure V1-7 .7.2.2-2.
σ z
σMAX
Equivalent profile σMAX
σ0 Material curve
σ0 x
σ0
ε
εMAX σMAX
Figure V1-7 .7.2.2-2: Cozzone's equivalent stress profile

page V1-7•7−
SSM V1-7 • LUGS
Stress σ 0 is a fictitious stress that is supposed to exist at the neutral fibre of the pin or at null strain which is the
same thing here. We determine σ 0 by stating that σ 0 does not theoretically depend on the shape of the section.
The Cozzone method enables a fictitious allowable bending stress called "bending modulus" to be calculated. This
stress can be compared with the maximum stress calculated by a linear method.
The general expression of the bending modulus is: σb = σMAX + σ0 . (k - 1)

Remarks:
① The calculation of the bending modulus is based on the real desired maximum stress level, that is generally
σ MAX = σ R the ultimate strength for a calculation at ultimate load. At limit load, the stress level must
remain lower than the yield strength.σb = σMAX + σ0 . (k - 1)
② The bending modulus also depends on the geometrical characteristics of the section by means of the term
"k". "k" is determined from figure V1-7 .7.2.2-3 or from the following equation.σb = σMAX + σ0 . (k - 1)
3
d 
1 −  i 
16  d 
k= ⋅
3π d 
4
1 −  i 
 d 
Where:
d: pin outside diameter.
di : pin inside diameter.
1.85
1.75
1.65
k
1.55
1.45
1.35
1.25
0.0 0.2 0.4 0.6 0.8 1.0 1.2
di/d
Figure V1-7 .7.2.2-3: Determining "k"

page V1-7•7−
SSM V1-7 • LUGS
The stress σ 0 is determined from the various graphs shown on figures V1-7 .7.2.2-5 to -19.
If none of the curves corresponds to the material used, σ 0 must be calculated knowing that this stress does not
theoretically depend on the shape of the section. That which is valid for a rectangular section is also valid for all
other sections.
We have checked that the method given below to calculate the bending modulus σ 0 gives the same results as
the graphs shown on figures V1-7 .7.2.2-5 to -19.
σ 0 corresponding to σ R is calculated; this stress level is obtained on the extreme fibre of the pin. Considering this
as pure bending and that the material curve is perfectly symmetrical in tension-compression, we assume that stress
is null at the neutral fibre of the pin. Consequently, a second x-axis can be plotted corresponding to the dimension
between the neutral fibre and a fibre given on the material curve (refer to figure V1-7 .7.2.2-4).
σ
z σR
σ
Zmax = h/2
z σ(z)
0 ε
εs
0 z
b z zmax = d/2
Figure V1-7 .7.2.2-4: Real stress profile
Thus, the bending moment can be written:
h/2
M=2b
∫ z ⋅ σ(z) ⋅ dz
0
 h ε
z = 2 × ε
 M
Where:  Plane bending: the straight sections remain plane. ε is linear with z.
dz = h dε
 2ε M
Also, ε can be written:
n
σ  σ 
ε= + 0,002 
E  σ0, 2 
 
n
 σ 0, 2 
Where: n = 500 

 σR 

page V1-7•7−
SSM V1-7 • LUGS
We deduce the expressions of ε and ε max :
n
σ 1 σ 
ε = +   = εe + ε p

E n  σR 
n
σMAX 1  σMAX 
ε max = +   = εe + ε p MAX
E n  σR  MAX
M⋅v 6
However σb = = ⋅M
I b ⋅ h2
Also:
σb = (k − 1)⋅ σ0 + σMAX

 σ0  σb 
 h⋅W 3 ⇒ = 2 − 1
k = = σMAX σ
 MAX 
 I 2
After resolution, we obtain:
2
σ0 2(n − 1)  ε p MAX    2n + 1   εe MAX  
 + 1
= ⋅   ⋅  ⋅
σMAX 2n + 1  εMAX    n + 2   εp MAX  
 

This expression can also be written:
σ0  n − 1   ε p MAX    n − 1   ε p MAX  
= 2⋅ ⋅  ⋅ 1 −   
σ MAX  n + 2   ε MAX    2n + 1  ⋅  ε 
   MAX  
Where:
σ MAX maximum allowable stress
n
1 σ  
ε p MAX =  MAX  maximum allowable plastic strain
n  σR 
σ
ε MAX = MAX + ε p MAX maximum allowable strain
E
n
 σ 0, 2 
n = 500 
 (numerical resolution)
 σR 
Specific cases:
1 σR 1
① (σMAX = σR ) ⇒ ε p MAX = ; ε MAX =
+
n E n
σ0, 2
② (σMAX = σ0,2 ) ⇒ ε p MAX = 0, 2% ; ε MAX =
E
+ 0,2%

page V1-7•7−
SSM V1-7 • LUGS
Figure V1-7 .7.2.2-5: Curves σMAX = f(ε) and σ0 = f(ε) - 2024-T3 clad sheet (0.25 mm < thickness < 1.6 mm)
Figure V1-7 .7.2.2-6: Curves σMAX = f(ε) and σ0 = f(ε) - 2024-T6 clad sheet (thickness < 1.6 mm)
Figure V1-7 .7.2.2-7: Curves σMAX = f(ε) and σ0 = f(ε) - 2024-T4 clad sheet (6.3 mm < thickness < 12.7 mm)

page V1-7•7−
SSM V1-7 • LUGS
Figure V1-7 .7.2.2-9: Curves σMAX = f(ε) and σ0 = f(ε) - 2024-T3 plate and sheet (thickness ≤ 6.35 mm)
Figure V1-7 .7.2.2-10: Curves σMAX = f(ε) and σ0 = f(ε) - 2024-T3 & -T4 plate and sheet (thickness ≤ 12.7 mm)
Issue 1 Pin stressing page V1-7•7-8

SSM V1-7 • LUGS
Figure V1-7 .7.2.2-12: Curves σMAX = f(ε) and σ0 = f(ε) - 7075-T6 sheet (thickness < 1 mm)
Figure V1-7 .7.2.2-13: Curves σMAX = f(ε) and σ0 = f(ε) - 7075-T6 extrusion (thickness < 6.35 mm)
Figure V1-7 .7.2.2-14: Curves σMAX = f(ε) and σ0 = f(ε) - 7075-T6 forged part (thickness < 10322 mm²)
Figure V1-7 .7.2.2-15: Curves σMAX = f(ε) and σ0 = f(ε) - AZ61A forged magnesium alloy part (direction L)

SSM V1-7 • LUGS
Figure V1-7 .7.2.2-16: Curves σMAX = f(ε) and σ0 = f(ε) - T-A6V
Figure V1-7 .7.2.2-17: Curves σMAX = f(ε) and σ0 = f(ε) - T-M4A4
Figure V1-7 .7.2.2-18: Curves σMAX = f(ε) and σ0 = f(ε) - Z6 CNU 17.04 bar and forged part
Figure V1-7 .7.2.2-19: Curves σMAX = f(ε) and σ0 = f(ε) - 7079-T6 die forged part (L direction)
(thickness ≤ 152.4 mm)

SSM V1-7 • LUGS
V1-7 7.2.3 Allowable bending force

o calculate Pfu and Pfy , we make the linear bending stress expression equal to the Cozzone expression. Thus,
generally:
F⋅b⋅d
= σMAX + σ0 ⋅ (k − 1)
4Ι
At ultimate load, we obtain: σ MAX = σR
Hence:
Pfu ⋅ b ⋅ d 4Ι
= σR + σ0 ⋅ (k − 1) ⇒ Pfu = [σR + σ0 ⋅ (k − 1)]
4Ι b⋅d
Where:
I: Pin moment of inertia
b: bending moment lever arm (refer to §V1-7 .7.2.1)
d: pin diameter
σR: ultimate strength
σo: Cozzone fictitious stress at neutral fibre (refer to §V1-7 7.2 2)
k: σ 0 factor (refer to §V1-7 .7.2.2).
At limit load, the stress levels must remain within the elastic range which means that σ 0 =0 and
σ b = σ MAX = σ0, 2 . Therefore, the allowable force at limit load will be simply:
4 Ι ⋅ σ 0, 2
Pfy =
b⋅d
Where:
σ0, 2 : 0.2% tensile yield strength.

SSM V1-7 • LUGS
PIN BENDING
Determine the dimensionless quantity r =


t2
D
a − 2 
 
and ratios:
Padm
A br ⋅ σR
(ultimate load) and
E
Padm F
(limit load) where Padm =
(R )
S
A br ⋅ σ 0, 2 1,6 0,625
a + R 1tr,6
Determine γ at U.L. and L.L. with the graph on figure V1-7 .7.2.1-2
Calculate the lever arm at U.L. and L.L.:

t
2
t1 t 2
t 
b = γ ⋅  1 + 2  + g
4 
("male / female" assembly) S
b= + +g ("male / male" assembly)
2 2
Calculation at ultimate load Calculation at limit load

E
( D) 3
N
1− d 16
Calculate: k = ⋅
1 − (d ) 3π
4
D
yes no
Material
T
known?
yes no
n
known?
Determine σ 0 from the graphs
Solve:
on figures V1-7 .7.2.2-5 to V1-7 n
I
.7.2.2-19  σ 0, 2 
n = 500 

 σR 
Calculation of stress at neutral fibre:

σ0  n − 1   ε p MAX    n − 1   ε p MAX 
 ⋅ 1 −  
A
= 2 ⋅ ⋅
σ MAX  n + 2   ε MAX    2n + 1   ε 
   MAX 
where:
n
1σ  σ
ε p MAX =  MAX  ; ε MAX = MAX + ε p MAX
n  σR  E
Allowable bending force at U.L.:

4Ι
Allowable bending force at L.L.:
4 Ι ⋅ σ0,2
L
Pfu = [σ R + σ 0 ⋅ (k − 1)] Pfy =
b⋅d b⋅d

SSM V1-7 • LUGS
V1- 8 NUMERICAL EXAMPLE
V1-7 8.1 BASIC DATA

t2 = 18 mm
D = 18 mm
R 20 mm
g = 0.6 mm
t1 = 10 mm
Figure V1-7.9.1-1: Example
The male and female parts are taken from a thick 2024 T351 plate. The characteristics taken into account are:
E1 = E2 = 7380 daN/mm² σR = 44 daN/mm² σ0.2 = 30.5 daN/mm².
Fbru (e/D=2) = 82 daN/mm² Fbry (e/D=2) = 62.5 daN/mm² n = 9 e(%) = 6
The pin is made of TA6V; we consider that the pin diameter is 18 mm.
The characteristics taken into account are:
Ea = 11030 daN/mm² σR = 90 daN/mm² n = 40
σ0.2 = 87 daN/mm² τadm = 52 daN/mm²
According to the geometrical information above, we have:

W = 40 mm
L = a = 20 mm
R = 20 mm
D = 18 mm
d ≈ 18 mm
t1 = 10 mm
t2 = 18 mm
g = 0.6 mm
The critical surface areas considered are:

A t1 = (W − D ) ⋅ t1 = (40 − 18)× 10 A t1 = 220 mm 2
A br1 = D ⋅ t1 = 18 × 10 A br1 = 180 mm 2
A t 2 = (W − D ) ⋅ t 2 = (40 − 18)× 18 A t 2 = 396 mm 2
A br 2 = D ⋅ t 2 = 18 × 18 A br 2 = 324 mm 2
Force applied:
The components of the oblique force to be transferred at ultimate load are:
Fa = 7000 daN
Ft = 3000 daN
We consider that these forces already take a fitting factor into account (1.15 for instance).
Issue 1 Numerical Example page V1-7•8-1

SSM V1-7 • LUGS
V1-7 8.2 LUG STRESSING

Calculating the allowable axial forces
Calculating the allowable tensile forces
We determine the breaking strain:
1 1
= = 0,11 > e(% ) ⇒ ε s = e(% ) = 6 %
n 9
The elastic overstress factors are determined using the graphs in the AEROSPATIALE FATIGUE MANUAL.
Considering the lug geometry and using the notations in the FATIGUE MANUAL, we obtain:
KtET = 5.25
Male part: G ≈ 1.1
Female part: G ≈ 1.2
Therefore:
Male part: Kt = Kt ET ⋅ G = 5,25 × 1,1 = 5,78
Female part: Kt = Kt ET ⋅ G = 5,25 × 1,2 = 6,3
The reference section used in the FATIGUE MANUAL is the gross section whereas it is the net section which is
taken into account in the S.S.M. Therefore, the overstress factors to be considered below are as follows:
W−D 40 − 18
Male part: K te = ⋅ Kt = ⋅ 5,78 = 3,18
W 40
W−D 40 − 18
Female part: K te = ⋅ Kt = ⋅ 6,3 = 3,47
W 40
The overstress factors Kσ are then:
σR 44
Male part: K σ = K te ⋅ = 3,18 × ≈1
E ⋅ εs 7380 × 0,06
σR 44
Female part: K σ = K te ⋅ = 3,47 × ≈ 1,09
E ⋅ εs 7380 × 0,06
Thus, the allowable tensile forces at ultimate load are:
A t 2 ⋅ σ R 396 × 44
Ptu = =
Male part: Kσ 1
Ptu = 17420 daN
A t1 ⋅ σR 220 × 44
Ptu = 2 × = 2×
Female part: Kσ 1,09
Ptu = 17760 daN

SSM V1-7 • LUGS
Calculating allowable shear / bearing forces:

a
For the male and female parts, we obtain: = 1,11
D
According to the graphs on figures V1-7 .3.2-1 and V1-7 .3.2-2,
a 
= 1,11
D 
Male part:  ⇒ K br ≈ 0,98
D
=1 
t2 
a 
= 1,11
D 
Female part:  ⇒ K br ≈ 0,98
D
= 1,8 
t1 
Thus, the allowable forces at ultimate load are:
82
Pbru = K br ⋅ A br 2 ⋅ Ftux = 0,98 × 324 ×
Male part: 1,9
Pbru = 13700 daN
82
Pbru = 2 × K br ⋅ A br1 ⋅ Ftux = 2 × 0,98 × 180 ×
Female part: 1,9
Pbru = 15220 daN
Summary of allowable axial forces at ultimate load:
Male part: Pu = min (Ptu ; Pbru) = min (17420 ; 13700)
Pu = 13700 daN
Female part: Pu = min (Ptu ; Pbru) = min (17760 ; 15220)
Pu = 15220 daN
Ftyx = 0,593 × Fbry = 0,593 × 62,5 = 37,1 daN / mm ²

Pu 13700
= = 0,98 < 1,05
A br 2 × Ftux 82
Male part: 324 ×
1,9
⇒ K y = 1,1
Pu 15220
= = 0,98 < 1,05
2 × A br1 × Ftux 82
Female part: 2 × 180 ×
1,9
⇒ K y = 1,1
Summary of allowable axial forces at limit load:

Ftyx
Male part: Py = K y × × Pu = 12940 daN
Ftux
Ftyx
Female part: Py = K y × × Pu = 14380 daN
Ftux

SSM V1-7 • LUGS
Calculating allowable transverse forces
The geometry of the lugs gives:
6 6
Male part: A av = = × 18 = 227,3 mm2
3 1 1 1 3 1 1 1
+ + + + + +
A1 A 2 A 3 A 4 13,64 11 11 13,64
6 6
Female part: A av = = × 10 = 126,3 mm2
3 1 1 1 3 1 1 1
+ + + + + +
A1 A 2 A 3 A 4 13,64 11 11 13,64
According to the graph of figure V1-7 .4-2 (curves 3 and 7), we obtain the following results for both the male and
female part:
A av A
= av ≈ 0,7 ⇒ K tru = 0,54 and K try = 0,8
A br 2 A br1
Thus, the allowable forces at ultimate load are:
Ptru = K tru ⋅ A br 2 ⋅ σ R = 0,54 × 324 × 44

Male part:
Ptru = 7700 daN
Ptru = 2 × K tru ⋅ A br1 ⋅ σ R = 2 × 0,54 × 180 × 44

Female part:
Ptru = 8550 daN
The allowable forces at limit load are:
Ptry = K try ⋅ A br 2 ⋅ σ 0 , 2 = 0,8 × 324 × 30,5

Male part:
Ptry = 7900 daN
Ptry = 2 × K try ⋅ A br1 ⋅ σ 0, 2 = 2 × 0,8 ×180 × 30,5

Female part:
Ptry = 8780 daN
Calculating the breaking load as if only the ligament subjected to the load was supporting
the transverse force would be more conservative still.
For this, we retain: Ptru = Ptry .
Summary of allowable transverse forces:

At ultimate load:
Male part: Ptru = 7900 daN
Female part: Ptru = 8780 daN
At limit load:
Male part: Ptry = 7900 daN
Female part: Ptry = 8780 daN

SSM V1-7 • LUGS
Calculation of the lug reserve factors
The reserve factors at ultimate load are:
Male part:
1 1 1
R .F. = = = = 1,45
[
R 1a,6 + R 1tr, 6 ]
0, 625
 F
 a

1,6
 F 
1, 6 

0,625
 7000 1,6  3000 1,6 
  +  
0, 625
 +  tr 
 Pu 
  Ptru

   13700   7900  
 
Female part:
1 1 1
R .F. = = = = 1,61
[
R 1a,6 + R 1tr, 6 ]
0, 625
 F
 a

1, 6
 F 
1, 6 

0,625
 7000 1,6  3000 1,6 
  +  
0, 625
 +  tr 
 Pu 
  Ptru

   15220   8780  
 
Let us consider that the limit loads are equal to the ultimate loads divided by 1.5. We then obtain:
Fa = 4667 daN
Ft = 2000 daN
The reserve factors at limit load are:
Male part:
1 1 1
R .F. = = = = 2,09
[
R 1a, 6 + R 1tr,6 ]
0,625

 Fa


1, 6
F 
1, 6 

0,625
 4667 1,6  2000 1,6 
  +  
0, 625
+  tr 
 Pbry   Ptry    12940   7900  
    
Female part:
1 1 1
R .F. = = = = 2,33
[
R 1a,6 ]
+ R 1tr, 6
0, 625

 Fa


1, 6
F 
1,6 

0, 625
 4667 

1,6
 2000 
+
1, 6 

0,625
+  tr   
 Pbry   Ptry    14380   8780  
    
V1.7 8.3 PIN STRESSING

Pin shear:
Here, the pin is subjected to double shear; the allowable force is therefore equal to:
π ⋅ d 2 ⋅ τadm π × 182 × 52
Psu = = = 26465 daN
2 2
The force applied to the pin is:
F = Fa2 + Ftr2 = 70002 + 30002 = 7616 daN

Hence the R.F. of the pin subjected to shear is:
Psu 26465
R.F. = = ⇒ R.F. = 3,47
F 7616

SSM V1-7 • LUGS
Pin bending
At ultimate load:
The allowable oblique force at ultimate load Padm is:
F 7616
Padm = = = 11017 daN
[ ]
R 1a, 6 + R 1tr,6
0,625
 7000 1,6  3000 1,6 
  +  
0, 625
 13700   7900  
The peaking parameter is defined from the graph on figure V1-7 .7.2.1-2
 D  18  
a − 2  20 − 2  
   
r= = = 0,61
t2 18  ⇒ γ = 0,38
Padm 11017 
= = 0,77 
A br ⋅ σ R 324 × 44 
The lever arm to be considered at ultimate load is therefore:
t t   10 18 
b = γ ⋅  1 + 2  + g = 0,38 ×  +  + 0,6 = 4,21 mm
 2 4   2 4
For a solid circular pin, the Cozzone calculation gives:
16
k= ≈ 1,7
3π
We would like that at ultimate load the stress at the extreme fibre of the pin does not exceed the ultimate strength,
that is:
σ M = σR = 90 daN / mm 2
We deduce the bending modulus σ0 :

σ0  n − 1   ε p MAX    n − 1   ε p MAX 
= 2 ⋅  ⋅ 1 −   
σ MAX  n + 2   ε MAX    2n + 1  ⋅  ε 
   MAX 
1 σR 1 90 1
Where: ε p MAX = and ε MAX = + = + = 0,033
n E n 11030 40
σ0 = 80.3 daN / mm²
The allowable force at ultimate bending load is therefore:

π × 18 4
4
4Ι
Pfu = [σ R + σ 0 ⋅ (k − 1)] = 64 × [90 + 80,3 × (1,7 − 1)]
b⋅d 4,21× 18
Pfu = 39769 daN
The corresponding bending R.F. is therefore:

P 39769
R .F. = fu = ⇒ R.F. = 5,22
F 7616

SSM V1-7 • LUGS
At limit load:
The allowable oblique force at limit load Padm is:
F 7616 / 1,5
Padm = = = 10630 daN
[
R 1a,6 + R 1tr,6 ]
0 ,625
 4667 1, 6  2000 1,6 
  +  
0 ,625
12940   7900  

The peaking parameter is defined from the graph on figure V1-7 .7.2.1-2
 D  18  
a − 2  20 − 2  
   
r= = = 0,61
t2 18  ⇒ γ = 0,53
Padm 10630 
= = 1,08 
A br ⋅ σ R 324 × 30,5 
The lever arm to be considered at ultimate load is therefore:
t t   10 18 
b = γ ⋅  1 + 2  + g = 0,53 ×  +  + 0,6 = 5,64 mm
2 4  2 4
The allowable force at ultimate bending load is therefore:
π × 184
4Ι ⋅ σ0, 2 4 × 87
Pfy = = 64
b⋅d 5,64 × 18
Pfy = 17664 daN
The corresponding bending R.F. is therefore:
Pfy 17664
R .F. = = ⇒ R .F. = 3,48
F 7616 / 1,5

Static stressing manual V1-8 • HOLE REINFORCEMENT
A - CONTENTS
Chapters Titles Issues Dates
V1-8 1 General 1 04/1997 Original issue
V1-8 2 Data preparation 1 04/1997 Original issue
V1-8 3 General method 1 04/1997 Original issue
V1-8 4 Load in the fasteners of a fastened reinforcement 2 01/1998 Modification
V1-8 5 Block diagram 1 04/1997 Original issue
V1-8 6 Examples 1 04/1997 Original issue
Pages
V1-8. HOLE REINFORCEMENT V1.8 / 1
A - CONTENTS V1.8 / 1
B - SYMBOLS V1.8 / 2
C - REFERENCES V1.8 / 3
V1-8 • 1. GENERAL V1.8.1/1
V1-8 • 1.1. DEFINITIONS V1.8.1/1
V1-8 • 1.2. BEHAVIOUR AND FAILURE MODE V1.8.1/2
V1-8 • 1.3. PRESENTATION OF THE METHOD V1.8.1/2
V1-8 • 2. DATA PREPARATION V1.8.2/1
V1-8 • 2.1. MATERIAL DATA V1.8.2/1
V1-8 • 2.2. GEOMETRICAL DATA V1.8.2/1
V1-8 • 3. GENERAL METHOD V1.8.3/1
V1-8 • 3.1. CALCULATION OF OVER-STRESS COEFFICIENT Kt V1.8.3/1
V1-8 • 3.1.1. Calculation of η for a compact reinforcement V1.8.3/2
V1-8 • 3.1.2. Calculation of η for a flanged edge V1.8.3/2
V1-8 • 3.1.3. Calculation of η for a wide reinforcement V1.8.3/3
V1-8 • 3.1.4. Taking infinite stress state into account V1.8.3/4
V1-8 • 3.1.5. Conclusion V1.8.3/5
V1-8 • 3.2. CALCULATION OF EQUIVALENT STRESS σequi V1.8.3/6
V1-8 • 3.3. TAKING BLEND RADII INTO ACCOUNT V1.8.3/7
V1-8 • 3.4. LOAD IN THE FASTENERS OF A FASTENED REINFORCEMENT V1.8.3/8
V1-8 • 3.4.1. Circular hole V1.8.3/8
V1-8 • 3.4.2. Elliptical hole V1.8.3/10
V1-8 • 3.4.3. Rectangular hole V1.8.3/11
V1-8 • 4. BASIC DATA V1.8.4/1
V1-8 • 4.1. CALCULATION OF η V1.8.4/1
V1-8 • 4.2. CALCULATION OF Kt V1.8.4/8
V1-8 • 4.2.1. Wide reinforcement: circular hole V1.8.4/9
V1-8 • 4.2.2. Narrow reinforcement: elliptical and circular hole V1.8.4/15
V1-8 • 4.2.3. Narrow reinforcement: square hole V1.8.4/18
V1-8 • 4.2.4. Narrow reinforcement: triangular hole V1.8.4/21
V1-8 • 4.2.5. Narrow reinforcement: circular hole close to an edge V1.8.4/30
V1-8 • 5. BLOCK DIAGRAM V1.8.5/1
V1-8 • 6. EXAMPLES V1.8.6/1
V1-8 • 6.1. CIRCULAR HOLE WITH A WIDE SYMMETRICAL REINFORCEMENT V1.8.6/1
V1-8 • 6.2. ELLIPTICAL HOLE WITH A COMPACT NON-SYMMETRICAL NARROW
REINFORCEMENT V1.8.6/3
V1-8 • 6.3. BUSH INSERTED AS A REINFORCEMENT INSIDE A CIRCULAR HOLEV1.8.6/4
V1-8 • 6.4. REINFORCEMENT FOR A TRIANGULAR HOLE V1.8.6/5
V1-8 • 6.5. EVOLUTIVE THICKNESS REINFORCEMENT V1.8.6/7

B - SYMBOLS
Kt max: allowable theoretical coefficient
d: circular hole diameter
D: circular reinforcement diameter
tr: reinforcement width
hr: reinforcement height
t: plate thickness
br, ar: dimensions defining elliptical, square, triangular holes
ra: corner radius for triangular holes
rr: corner radius for square holes
c: edge-distance of a hole
Kt: over-stress coefficient
σmax: maximum equivalent stress
σequi: gross equivalent stress (Von Mises)
A: equivalent section
Ar: radial section of the reinforcement
Ap: radial section of the plate under the reinforcement
η: geometrical efficiency coefficient of the reinforcement
I∞: inertia of the reinforcement in relation to a centreline passing through its centre of
gravity and parallel to the neutral line of the plate
Ixx: inertia of the reinforcement in relation to the neutral line of the plate
Ep: Young's modulus of the plate
Er: Young's modulus of the reinforcement
σx, σy, τ: stress condition around the hole
σ1, σ2: main stresses
Ktr: over-stress coefficient at the joint of the reinforcement and the plate for a circular
hole
Ktb: over-stress coefficient at the hole edge for a circular hole

C - REFERENCES
REF. 1: Elastic stress concentration factors. Single reinforced and unreinforced holes in
infinite plate of isotropic materials
ESDU DATA ITEM 80027
REF. 2: Stress concentration factors

R.E. PETERSON


V1-8 • 1.GENERAL
This document proposes a method to calculate the maximum stress around a reinforced hole in
an infinite plate or close to a free edge. The plate is assumed stable.
V1-8 • 1.1.DEFINITIONS
Holes are frequently used in aircraft structures. They are used to lighten the structures, route
cables or ducts or even to give access to certain other areas of the structure.
Generally, these holes are reinforced by extra thickness, a fastened reinforcement or an

internal bush:
- the extra thickness is machined in the plate with the hole,
- the fastened reinforcement is a ring secured around the hole by bolts, rivets or tack welds
fastened,
- the bush in the hole is made of a material different from the plate. The advantage of this
technique is to reinforce the hole without using an extra thickness (more compact).
An extra thickness or a fastened reinforcement may (or may not) be symmetrical in relation to
the neutral line of the plate.
V1-8 • 1.2.BEHAVIOUR AND FAILURE MODE
A hole in a structure locally modifies the mechanical behaviour of the structure. In particular,
the presence of a hole creates an over-stress state around it. This over-stress may exceed the
allowable limit of the material and cause local failure in the structure.
Use of a hole reinforcement is recommended in this case.
The stress state around the reinforced hole is taken into account using the Mises-Hencky
criterion. The method is used to calculate the maximum equivalent stress that can be compared
with the selected allowable limit.
Note: In regular structural areas, a theoretical over-stress coefficient is accepted (in relation to
the gross stress):
Kt max = 2.4

V1-8 • 1.3. PRESENTATION OF THE METHOD
To cover as many cases as possible, this study will take different geometries of holes and of
reinforcements under different stress states into account.
The various hole geometries are:

- circular holes
- elliptical holes
- rectangular holes (with rounded corners)
- triangular holes (with rounded corners)
The different types of hole reinforcements are:

- extra thickness
- fastened reinforcement (for example riveted)
- bush inside the hole (bush and plate materials different)
They are defined using two geometrical criteria:

- D/d: the ratio between the hole diameter and the diameter of the reinforcement,
- hr/tr: the ratio between the reinforcement height and its width.
t∞
d h∞
If D/d > 1.05: the reinforcement is considered to be sufficiently "wide" to take the variation in
the stress along a radial cross section of the reinforcement into account. In this case, the Kt max
at the edge of the hole and at the joint between the reinforcement and the plate is calculated.
If D/d < 1.05: the reinforcement is considered "narrow". This means that the stress is assumed
to be uniform along a radial cross section. The reinforcement will be represented by its surface
area and the interface with the plate by a line. This line will be defined by the outer radius of
the reinforcement. The Kt max will be calculated at some point of this line.
In this case, two sub-types of narrow reinforcements are differentiated:
If hr/tr < 3: the reinforcement is considered as being compact.
If hr/tr > 3: the reinforcement is considered as being a flanged edge.
DIFFERENT TYPES OF REINFORCEMENTS

WIDE NARROW
D/d > 1.05 D/d < 1.05
COMPACT: hr/tr < 3 FLANGED EDGE: hr/tr > 3
D
D D
hr hr
d d d
tr tr

In general, a hole reinforcement is a local over-thickness around the hole. For reasons of
overall dimensions another reinforcement principle may be used. This principle consists in
inserting a bush, made of a material different from the plate, in the hole. The method described
in this chapter covers this type of reinforcement as it takes the differences in Young's modulus
and Poisson ratio between the plate and the reinforcement into account.
Also, the fastened reinforcement case (for example riveted) is handled as an extrapolation of
the general method discussed in this chapter.
The different load cases are taken into account by means of the Mises-Hencky
criterion. The method is used to calculate the maximum equivalent stress that can be
compared with the selected allowable limit.
The elementary loads studied here are:

- uniaxial tension,
- biaxial tension,
- shear.
The calculated stress for a combined load will be (by interpolation) a linear combination of
stresses defined for each basic case.
Remarks:
• In the case of a circular hole, it if preferable to use the case of the main stresses in
biaxial tension (refer to paragraph V1-8.3.1.4).
• The real stress is always less than or equal to this calculated stress. In fact, the place
where the maximum stress is located is not the same for each load case. The
combination of these maximum stresses is therefore greater than the real stress.


V1-8 • 2.DATA PREPARATION
V1-8 • 2.2.MATERIAL DATA
The Young's modulus: E must be known for each of the materials used (plate and
reinforcement).
V1-8 • 2.2.GEOMETRICAL DATA
The geometrical data required for the study is specific to each type of reinforcement and hole
discussed in this chapter.
• For a wide reinforcement and a circular hole, the following must be known:
t
t+tr
dD
• For a narrow reinforcement and an elliptical or circular hole, the following must be
known:
t
br
tr
ar
hr

• For a narrow reinforcement and a square hole, the following must be known:
t
2br
tr
rr
hr
• For a narrow reinforcement and a rectangular hole, the following must be known:
t
tr
ra
br hr
• For a narrow reinforcement and a circular hole close to an edge, the following must be
known:
t
tr
ar hr

V1-8 • 3.GENERAL METHOD
A hole in an infinite plate. As this hole is reinforced (over-thickness, bush, added ring) the
goal is to calculate the maximum stress in the hole reinforcement area knowing the stress field
in the plate close to the hole.
To this end, the over-stress coefficient Kt and Mises-Hencky criterion are used to calculate the
equivalent stresses.
The maximum equivalent stress due to the presence of the reinforced hole is:
σmax = Ktσequi
with:
σmax: maximum equivalent stress (to be compared with the limit stress)
Kt: over-stress coefficient
σequi: gross equivalent stress (Von Mises) around the reinforcement area (function
of the stress state in the plate)
Therefore, it is necessary to calculate the over-stress coefficient Kt and the equivalent stress
σequi as a function of the geometry and the stress state in the plate.
V1-8 • 3.1.CALCULATION OF OVER-STRESS COEFFICIENT Kt
This coefficient depends on the geometry of the hole, the type of reinforcement and the stress
state in the plate around the hole.
To characterize the geometry of the hole and the type of reinforcement, the equivalent section
of the reinforcement is defined. To solve the real case, this parameter, noted A , makes it
possible to use the results obtained in basic cases.
The formulation of this equivalent section is:
A = Ap + ηAr
with:
Ap: cross section area of the plate under the reinforcement
Ar: reinforcement cross section area
η: geometrical efficiency coefficient of the reinforcement
The calculation of η differs according to the type of reinforcement. This coefficient takes both
the neutral fibre offset (if the reinforcement is not symmetrical in relation to the plate) and the
differences in the materials of the plate and the reinforcement (if this is the case) into account.
General method - Page V1-8•3/1 revision 1 © Copyright AEROSPATIALE 1998

V1-8 • 3.1.1.Calculation of η for a compact reinforcement
The formulation of η for a compact reinforcement (D/d<1.05 and hr/tr<3) is:

I E
η = 00 r
I xx E p
with:
I00: inertia of the reinforcement in relation to an axis parallel to the neutral line
of the plate and passing through its centre of gravity
Ixx: inertia of the reinforcement in relation to the neutral line of the plate
Example: a non-symmetrical rectangular cross section reinforcement:
tr
hr
d/2
tp
D/2
In this case there is:
2 2
t r h 3r  hr + tp  t r h 3r  hr + tp 
I00 = and Ixx = I00 + Ar   = + trhr  
12  2  12  2 
1 Er
being: η = 2
 tp  Ep
1 + 3 1 + 
 hr 
If in addition the thickness of this plate in relation to that of the reinforcement is not taken into
account, then η = 1/4
V1-8 • 3.1.2.Calculation of η for a flanged edge
The charts in chapter V1-8.4.1 give the values of η for symmetrical and non-symmetrical
flanged edges in biaxial tension.
The thin plate theory has been used to plot these curves and the regions in which this theory is
not truly applicable are shown in dotted lines.
These η values may be used to calculate non-circular holes with any stress state. The
generated errors are low if the point where the maximum stress is located is on a rounded
contour and if the hole is practically circular.

Remark:
The cross section area of the plate under the reinforcement, Ap, participates in the
reinforcement. If the Young's modulus of the reinforcement and that on the plate are different,
then this has to be taken into account in the calculation of η. To this end, it is necessary to
multiply the value read on the chart by the ratio: Er/Ep.
If the plate and the reinforcement are made of different materials, then:
Er
ηtrue = ηchart
Ep
with:
The cases shown in the charts in Chapter 1.8.4.1 are:

- symmetrical flanged edge,
- non-symmetrical flanged edge for hr/tr values between 3 and 10.
V1-8 • 3.1.3.Calculation of η for a wide reinforcement
The results obtained on the large reinforcements proposed in this chapter, do not make it
possible to take a non-symmetrical reinforcement or a difference in material between the plate
and the reinforcement into account. In the case of a large non-symmetrical reinforcement or if
the material is different from that of the plate, two approximations are possible:
- the reinforcement may be considered as being narrow. The advantage is to be able to
take into account the non-symmetry and the difference in material by calculating η (refer
to paragraph V1-8.3.1.1). The disadvantage is having to calculate the stress at joint
between the reinforcement and the plate assuming it to be constant throughout the
reinforcement.
- the reinforcement may be considered as being wide, symmetrical and of the same
material as the plate (refer to the example, paragraph V1-8.6.5). The advantage is that it
takes the development of the stress in the reinforcement into account by calculating it at
the edge of the hole and at the joint of the reinforcement and the plate. The disadvantage
is that the neutral fibre offset or the difference in material cannot be taken into account.
These approximations are also necessary for a non-circular wide reinforcement. In fact, only
the case of a wide circular reinforcement is discussed in this chapter.

V1-8 • 3.1.4.Taking infinite stress state into account
For the calculation of Kt, using the charts in Chapter 1.8.4.2, in addition to the calculation of
the equivalent cross section area A , it is also necessary to take into account the gross stress
state around the reinforcement.
For wide reinforcements (D/d>1.05), it is necessary to calculate the main gross stresses (in
the plate, without a hole and without reinforcement):
σy
τ τ
τ σ2 σ1 τ
σx σx
σ1 σ2
τ τ
τ σy τ
There is:
σ1 =
1
2
(σx + σy) +
1
4 [(
σx − σy )
2
]+ τ 2
σ2 =
1
2
(σx + σy) -
1
4[(
σx + σy ) ]+ τ
2 2
It is then necessary to obtain the linear combination which makes it possible to calculate these
main stresses using the processed basic cases.
Example:
For the case of a circular hole reinforced by a wide ring (D/d>1.05), five basic cases are
proposed:
σ1 σ2
case 1 1 1
case 2 1 0.5
case 3 1 0
case 4 1 -0.5
case 5 1 -1
σ 1
If for the case to be processed, there is: 2 =
σ1 4
K t ( cas3 ) + K t ( cas2 )
then: K  σ 1  = with the Kt(cas i) obtained from the charts.
t 2 =  2
 σ1
4 

For narrow reinforcements (D/d<1.05), unlike wide reinforcements, it is not necessary to

calculate the main stresses. All that is necessary is to obtain the linear combination of the basic
cases making it possible to calculate the real case.
V1-8 • 3.1.5.Conclusion
Knowing the equivalent cross section area A , and the linear combination between the basic
cases, making it possible to obtain the real stress status, it is possible to use the charts in
Chapter 1.8.4.3.2 to determine Kt max.

V1-8 • 3.2.CALCULATION OF EQUIVALENT STRESS σequi
Let us assume any stress state around the hole.
σy
τ τ
τ τ
σx σx
τ τ
τ σy τ
Then:
the formulation of the equivalent stress: (Mises-Hencky criterion) is:
σequi = σ 2x − σ x σ y + σ 2y + 3τ 2
This equivalent stress multiplied by the coefficient Kt makes it possible to calculate the
maximum stress at the edge of the hole or around the reinforcement. This maximum value can
be directly compared to the fixed limit value.
The basic data required to calculate Kt is given in the following chapter.

V1-8 • 3.3.TAKING BLEND RADII INTO ACCOUNT
With machined reinforcements, the presence of blend radii reduces the over-stress at the joint
between the reinforcement and the plate.
The following rule may be used to assess this effect.
1 - Calculation of Kt0 at the reinforcement/plate joint, in the case of a reinforcement

without a blend radius, using the methods and charts given in this chapter.
2 - Taking the effect of the radius on Kt into account, using the following formula:
Kt = β Kt0
with:
- Kt0: over-stress coefficient calculated at the joint between the reinforcement with the
plate without taking the presence of the radius into account.
- β: correction coefficient depending on the geometry of the blend radius. This
coefficient is calculated using the following chart:
1
0.95
0.9
0.85
0.8
0.75
β 0.7
0.65
0.6
0.55
0.5
0.45
0.4
0 0.25 0.5 0.75 1 1.25 1.5 1.75 2
r
t
r t

V1-8 • 3.4.LOAD IN THE FASTENERS OF A FASTENED REINFORCE-

MENT
The determination of the loads transferred by the fasteners of a fastened reinforcement is a

relatively complex problem. Numerous parameters are to be taken into account such as the
geometry of the hole, the type of reinforcement, the quantity, the location and the type of
fasteners, the state of stress around the reinforcement.
The method proposed here is an approximate method. It is simple but conservative. This
means that it may induce very significant loads in fasteners. It is therefore preferable to use a
local finite element model to get closer to reality.
Assumptions:
A hole reinforced in an infinite plate. The stress state around this reinforcement is any value.
The different hole geometries are those proposed in this chapter (circular, elliptical,
rectangular).
Method:
The main difficulty is to take shear stresses into account. An adapted method is proposed for
each type of geometry. These methods make it possible to take each type of stress (normal or
shear) into account separately.
V1-8 • 3.4.1.Circular hole

σy
τ
t
tr
σ2
σ1
σx σx
d D
σ2
σ1
τ
σy
The method is as follows:
1 - Calculation of main stresses as a function of infinite gross stresses:
1
σ1 = (σx + σy) +
2
1
4 [(
σx − σy
2
)]
+ τ2
1
σ2 = (σx + σy) -
2
1
4 [(
σx − σy
2
) ]
+ τ2

2 - Calculation of the total gross load transferred at the reinforcement in the main
directions:
Ftot1 = σ1 (D . t)
Ftot2 = σ2 (D . t)
3 - Calculation of the loads transferred into the reinforcement in proportion to the
thicknesses:
 t 
Frenf1 = Ftot1  r 
 t + tr 
 t 
Frenf2 = Ftot2  r 
 t + tr 
4 - Distribution of each of these loads in each fastener. Distribution will occur at the same
bearing and same shear as indicated in Chapter V1-6 "Splices" in this manual. For each
main direction, a loading area and an unloading area will be considered.
Fasteners subjected to Frenf 1 Fasteners subjected to Frenf 2

σy σy
τ τ τ τ
σ2 σ2
σ1 σ1
σx σx σx σx
σ2 σ2
σ1 σ1
τ τ τ τ
σy σy
In this manner, two loads per fastener F/fix1 and F/fix2 are obtained which then simply have to
be combined to obtain the total load per fastener:
F/fix = F/2fix1 + F/2fix 2

V1-8 • 3.4.2.Elliptical hole
σy
τ
t
br
tr
σx ar σx hr
τ
σy
The main stresses cannot be used for an elliptical hole.
The normal stresses (σx, σy), are taken into account in the same manner as in the case of a
circular hole with main stresses (refer to paragraph V1-8.3.4.1).
The shear stresses (τ) are taken into account using the following method:
1 - Calculation of the total gross load related to the shear at the reinforcement:
τ
Fτ = τ (L . t) L
τ
2 - Distribution of this load between the plate and reinforcement in proportion to the
thicknesses:
 hr 
Fτ renf = Fτ  
 t + hr 
3 - Distribution of Fτ renf (force transferred into the reinforcement). On each fastener,
distribution will occur at the same bearing and the same shear as indicated in Chapter
V1-6 "Splices" in this manual.
A loading area and an unloading area will be considered. The unloading area will be
distributed to the following fasteners:
τ
L
fastener subjected to Fτ renf

V1-8 • 3.4.3.Rectangular hole

σy
τ t
2br
tr
σx hr
σx
2ar
τ
σy
Main stresses cannot be used for a rectangular hole.
Normal stresses (σx, σy) are taken into account using the following method:
1 - Calculation of the total gross load transferred at the reinforcement in σx and σy

directions.
Ftotx = σx (2ar . t)
Ftoty = σy (2br . t)
2 - Calculation of the loads transferred into the reinforcement in proportion to the
thicknesses:
 hr 
Frenfx = Ftot1  
 t + hr 
 hr 
Frenfy = Ftot2  
 t + hr 
3 - Distribution of each of these loads. Distribution will occur at the same bearing and the
same shear as indicated in Chapter V1-6 "Splices" in this manual.
For each main direction, a loading area and an unloading area will be taken into
account:
fasteners subjected to Frenfy
σy
σx σx
fasteners subjected to Frenfx
σy
Note: Unlike the other fasteners, corner fasteners are subjected to both Frenfx and Frenfy.

Shear stresses (τ) are taken into account using the following method:
1 - Calculation of gross loads related to shear at the reinforcement on each side:

2br
τ
Fτa = τ(2ar . t) 2ar

Fτb = τ(2br . t)
2 - Distribution of these loads between the plate and the reinforcement in proportion of the
thicknesses:
 hr 
Fτ renf a = Fτa  
 t + hr 
 hr 
Fτ renf b = Fτb  
 t + hr 
3 - Distribution of Fτ renf a and Fτ renf b (forces transferred to the reinforcement) on each
fastener.
This distribution will occur at the same bearing and the same shear as indicated in
Chapter V1-6 "Splices" in this manual.
The loads are assumed to be distributed on the following fasteners:
fasteners subjected to Fτ renf b
2br
τ
fasteners subjected to Fτ renf a
2ar
fasteners subjected to -Fτ renf a τ
fasteners subjected to -Fτ renf b
Load combinations: Finally, for each fastener, it is necessary to combine the loads due to the
various stresses (σx, σy, τ).

V1-8 • 4.BASIC DATA
V1-8 • 4.1.CALCULATION OF η
Basic data - Page V1-8•4/1 revision 2 © Copyright AEROSPATIALE 1998

hr
hr
tr
rr
Figure 1-8.4.1.1 Symmetrical flanged edge

t
hr
tr
rr
hr
Figure 1-8.4.1.2 Non-symmetrical flanged edge =3
tr

t
hr
tr
rr
hr
tr

t
hr
tr
rr
hr
tr

t
hr
tr
rr
hr
tr

t
hr
tr
rr
hr
Figure 1-8.4.1.6 Non-symmetrical flanged edge = 10
tr

V1-8 • 4.2.CALCULATION OF Kt
The cases dealt with here enable calculation of Kt in all other load or geometry cases by
applying the method described in the previous chapter (paragraph 1.8.3).
These basic cases are as follows:
TYPE OF
CHART
REINFOR- TYPE OF HOLE TYPE OF LOADING
NUMBERS
CEMENT
τ σy
σ1 σ2
τ σ2 σ1 1 1 1-8.4.2.1.1
D/d > 1.05
σx 1-8.4.2.1.2
WIDE
1 0.5
CIRCULAR 1-8.4.2.1.3
σx 1 0 1-8.4.2.1.4
σ1 σ2 τ 1 -0.5 1-8.4.2.1.5
1 -1
σy τ
τ σy
τ σx σy τ
1 0 0 1-8.4.2.2.1
σx 1-8.4.2.2.1
ELLIPTICAL 1 1 0
σx 1-8.4.2.2.2
1 0.5 0 1-8.4.2.2.2
τ 0 0 1
σy τ
τ σy
τ σx σy τ
1 0 0 1-8.4.2.3.1
SQUARE σx
1 1 0 1-8.4.2.3.1
(rounded
corners)
σx 1-8.4.2.3.2
1 0.5 0 1-8.4.2.3.2
τ 0 0 1
NARROW
D/d < 1.05
σy τ
τ σy σx σy τ
τ 1 0 0
1-8.4.2.4.2
TRIANGULAR 0 1 0 1-8.4.2.4.3
(equilateral σx 1-8.4.2.4.4
1 1 0
rounded σx 1-8.4.2.4.5
corners) 1 0.5 0 1-8.4.2.4.6
τ 0.5 1 0 1-8.4.2.4.7
0 0 1
σy τ
CIRCULAR σx
(close to an σx only 1-8.4.2.5.1
σx
edge)

V1-8 • 4.2.1.Wide reinforcement: circular hole
a) Necessary geometrical data:

σy
τ
t
t+tr
σ2
σ1
σx σx
d D
σ2
σ1
τ
σy
b) Use of the charts:

The following charts enable the calculation of the over-stress coefficient at the edge of a hole
(KtA) and at the joint of the reinforcement and the plate (KtB). These coefficients are given as a
t D
function of the following ratios: r and
t d
Remarks:
D
- if < 1.05 then refer to paragraph V1-8.4.2.2
d
- if the reinforcement is not symmetrical, refer to paragraph V1-8.4.2.2 or consider that
the radial cross section of the reinforcement is uniformly distributed on either side of the
reinforcement by calculating an equivalent thickness tr (refer to the example in
paragraph V1-8.6.5).
- if the hole is elliptical, refer to paragraph V1-8.4.2.2 or consider an equivalent circular
hole with an average radius d (refer to the example in paragraph V1-8.6.5).
c) Single loads:
The charts in this paragraph correspond to the following single load cases.
TYPE OF LOADING CHART NUMBERS
σ1 σ2
1 1 1-8.4.2.1.1
1 0.5 1-8.4.2.1.2
1 0 1-8.4.2.1.3
1 -0.5 1-8.4.2.1.4
1 -1 1-8.4.2.1.5
For any other type of load, it is possible to calculate the maximum stress by combining these
basic cases (conservative calculation). This method is used in the example in paragraph V1-
8.6.1.

σy
τ t
t+tr
σ2 σ1
d D
A σx
σx
B
σ1 σ2
τ
σy
Figure 1-8.4.2.1.1

σy
τ t
t+tr
σ2 σ1
d D
A σx
σx
B
σ1 σ2
τ
σy
Figure 1-8.4.2.1.2

t
t+tr
σ1 σ1
d D
A
Figure 1-8.4.2.1.3

σy
τ t
t+tr
σ2 σ1
d D
A σx
σx
B
σ1 σ2
τ σy
Figure 1-8.4.2.1.4

σy
τ t
t+tr
σ2 σ1
d D
A σx
σx
B
σ1 σ2
τ σy
Figure 1-8.4.2.1.5

V1-8 • 4.2.2.Narrow reinforcement: elliptical and circular hole

σy
τ
t
br
tr
ar
σx σx hr
τ
σy

The following charts are used to calculate the over-stress coefficient at the joint between the
reinforcement and the plate (Kt). These coefficients are given as a function of the following
A a
ratios: and r
(a r + b r ) t br
Remarks:
- A is the equivalent section of the reinforcement, which takes the non-symmetry of the
reinforcement or a difference in material between the reinforcement and the plate into
account where applicable.
c) Single load case:

σx σy τ
1 0 0 1-8.4.2.2.1
1 1 0 1-8.4.2.2.1
1 0.5 0 1-8.4.2.2.2
0 0 1 1-8.4.2.2.2
For any other type of loading, it is possible to calculate the maximum stress by combining
these basic cases (conservative calculation). This method is used in the example in paragraph
V1-8.6.1.

ar
σx σx
br
σy = 0
τ=0
σy
ar
σx σx
br
σy
σy
=1
σx
τ=0
Figure 1-8.4.2.2.1

σy
ar
σx σx
br
σy
σy 1
=
σx 2
τ=0
τ
ar
br
τ
σx = 0
σy = 0
Figure 1-8.4.2.2.2

V1-8 • 4.2.3.Narrow reinforcement: square hole

σy
τ t
2br
tr
rr σx hr
σx
τ
σy

reinforcement and the plate (Kt). These coefficients are given as a function of the following
A r
ratios: and r
brt br
Remarks:

σx σy τ
1 0 0 1-8.4.2.3.1
1 1 0 1-8.4.2.3.1
1 0.5 0 1-8.4.2.3.2
0 0 1 1-8.4.2.3.2
V1-8.6.1.

2br
rr
σx σx
σy = 0
τ=0
σy
2br
rr
σx σx
σy
σy =1
σx
τ=0
Figure 1-8.4.2.3.1

σy
2br
rr
σx σx
σy
σy = 1
σx 2
τ=0
τ
2br
rr
τ
σx = 0
σy = 0
Figure 1-8.4.2.3.2

V1-8 • 4.2.4.Narrow reinforcement: triangular hole

σy
τ
t
ra
tr
σx br σx hr
τ
σy

reinforcement and the plate (Kt).
r r
Initially, using the ratio a , it is necessary to determine the theoretical ratio t using the chart
br br
in Figure 1-8.4.2.4.1.
A r
Then the over-stress coefficients are given as a function of the ratios: and t .
brt br
Remarks:

No. σx σy τ
1 1 0 0 1-8-4.2.4.2
2 0 1 0 1-8.4.2.4.3
3 1 1 0 1-8.4.2.4.4
4 1 0.5 0 1-8.4.2.4.5
5 0.5 1 0 1-8.4.2.4.6
6 0 0 1 1-8.4.2.4.7
V1-8.6.1.

d) Location of the over-stress:
y
x
The position of the over-stress and the joint of the reinforcement and the plate depends on the
stress status around the reinforcement.
For cases 1, 3 and 4, the over-stress is at corners B.
For case 5, the over-stress is at corner A.
ra rt
For cases 2 and 6, this position depends on the geometrical ratios of the triangle: , and
br br
A
. To locate this over-stress, regions have been defined on the corresponding charts: Figure
brt
1-8.4.2.4.3 and Figure 1-8.4.2.4.7. Region 1 corresponds to an over-stress at corners B and
region 2 corresponds to an over-stress at corner A.
TYPE OF LOADING LOCATION
No. σx σy τ
1 1 0 0 B
2 0 1 0 B (if region 1)
A (if region 2)
3 1 1 0 B
4 1 0.5 0 B
5 0.5 1 0 A
6 0 0 1 B (if region 1)
A (if region 2)

ra
br
real shape
rt
br
theoretical shape
Figure 1-8.4.2.4.1

ra br
σy = 0
σx σx τ = 0
Figure 1-8.4.2.4.2

σy
ra br
σx = 0
τ=0
σy
Figure 1-8.4.2.4.3

σy
ra
br
σx = σy
σx σx τ=0
σy
Figure 1-8.4.2.4.4

σy
ra
br
σy 1
σx σx σx 2
τ=0
σy
Figure 1-8.4.2.4.5

σy
ra
br
σx 1
=
σx σx σ y 2
τ=0
σy
Figure 1-8.4.2.4.6

τ
ra
br
σx = 0
σy = 0
Figure 1-8.4.2.4.7

V1-8 • 4.2.5.Narrow reinforcement: circular hole close to an edge
tr
ar hr

A c
Depending on the ratios: and , the following chart is used to calculate the over-stress
2a r t ar
coefficient at the joint of the reinforcement and the plate (Ktr) and at the edge of the plate
(Ktb).
Remarks:
- A is the equivalent section of the reinforcement, which takes the non-symmetry

The chart in this paragraph corresponds to the unidirectional load case.
σx σy τ
1 0 0 1-8.4.2.5.1

ar
σx σx
Figure 1-8.4.2.5.1


V1-8 • 5.BLOCK DIAGRAM
E
CALCULATION OF MAXIMUM STRESS AT
THE EDGE OF THE REINFORCED HOLE
CALCULATION OF Kt maxi CALCULATION OF σequi
yes D/d > 1.05 no

S
Calculation of η
yes hr/tr > 3 no

S
Er I 00 E r
ηreal = ηchart η=
E
Ep I xx E p
Calculation of σ1 and σ2
σ1 =
1
2
(σx + σy) +
1
4
[(σ x − σy )
2
]+τ 2
Calculation of the equivalent cross
σ2 =
1
2
(σx + σy) -
1
4
[(σ x − σy )
2
]+τ 2
section area A :
A = Ap + ηAr N
Kt(real) = combination (Kt(basic case))
Kt(basic case) using the charts
σequi = σ 2x − σ x σ y + σ 2y + 3τ 2
T
σmax = Ktσequi
I
τ σy τ
τ σ2 σ1 τ A
σx σx
σ2
L
τ σ1 τ
τ σy τ
Block diagram - Page V1-8•5/1 revision 1

Block diagram - Page V1-8•5/2 revision 1

V1-8 • 6.EXAMPLES
V1-8 • 6.1.CIRCULAR HOLE WITH A WIDE SYMMETRICAL

REINFORCEMENT
Let us assume a 100 mm diameter hole in a 5 mm infinite plate. This hole is reinforced by a
200 mm diameter, 10 mm thick symmetrical over-thickness.
σy
τ
σx σx
σy τ
200 mm
10 mm
100 mm
5 mm
A B
10 mm
The stress state around the hole is as follows:

σx = 185 Mpa
σy = 65 Mpa
τ = 45 Mpa
D 200
There is: = = 2 > 1.05, therefore it is a wide reinforcement.
d 100
In this case, it is necessary to consider a development of the stress according to the
reinforcement width (between A and B).
To determine the over-stress at the edge of the hole (at A) and at the reinforcement/plate joint
(at B), the charts in Chapter V1-8.4.2.1 will be used after calculating the main stresses around
the hole.
Examples - Page V1-8•6/1 revision 1 © Copyright AEROSPATIALE 1998

• Calculation of main stresses and equivalent stress:

1 1 1 1
σ 1 = (σ x + σ y ) +
2
[
4
]
σx − σy + τ2 σ 1 = (185 + 65) +
2 4
[185 − 65] + 45 2
being
1 1 1 1
σ 2 = (σ x + σ y ) −
2
[4
]
σx − σy + τ2 σ 2 = (185 + 65) −
2 4
[185 − 65] + 452
σ = 200 MPa
Giving: 1 50 MPa
σ2 =
Similarly, the equivalent stress is:
σequi = σ 12 − σ 1 σ 2 + σ 22 = 200 2 − 200 . 50 + 50 2 giving: σequi = 180 MPa
• Determination of the linear combination:
σ = 200 MPa σ 1
There is: σ 1 = 50 MPa being: 2 =
2 σ1 4
This load case does not correspond to one of the charts on wide reinforcements. Therefore, the
σ 1
linear combination of basic cases has to be found to obtain 2 =
σ1 4
The basic cases processed are:
σ1 σ2
case 1 1 1
case 2 1 0.5
case 3 1 0
case 4 1 -0.5
case 5 1 -1
K t ( cas3 ) + K t ( cas2 )
Then, the linear combination is: K σ 1
= with the Kt(casi) found in the
t 2 =  2
 σ1 4 
charts.
• Calculation of the over-stress coefficient:
t r 2 . 10
Using the charts in Chapter V1-8.3.2.1, the following is obtained (with = = 4 and
tp 5
D
= 2):
d
K t B (cas2) = 1.06 ⇒ K = 1.06 + 1.22 ⇒ K = 114 .
- At the edge of the hole (pt B): K t B (cas3) = 1.22
tB
2
tB
K t A (cas2) = 1.28
- At the edge of the reinforcement (pt A): 1.28 + 1.63
K t A (cas3) = 1.63 ⇒ K t A = ⇒ K t A = 1.45
2
• Calculation of the maximum stresses at the edge of the hole and at the edge of the
reinforcement:
Using the equivalent stress and over-stress coefficients, the maximum stress values at the edge
of the hole and at the edge of the reinforcement can be deduced:
- At the edge of the hole (pt B): σmax B = Kt B . σequi = 1.14 . 180 ⇒ σmax B = 205 Mpa
- At the edge of the reinforcement (pt A): σmax A = Kt A . σequi = 1.45 . 180 ⇒ σmax A = 261 Mpa

V1-8 • 6.2.ELLIPTICAL HOLE WITH A COMPACT NON-SYMMETRICAL

NARROW REINFORCEMENT
τ
160
80
τ
3
3
9
with: τ = 70 MPa
In this case there is:
D 83
= = 1.037 < 1.05
• and h d 80 consequently, this is a compact narrow reinforcement
6
r
= =2<3
tr 3
As this reinforcement is not symmetrical, it is necessary to take its geometrical effectiveness

coefficient η into account when calculating it.
• Calculation of coefficient η:
• Inertia of the reinforcement in relation to the axis passing through its centre of gravity
t h 3 3. 6 3
and parallel to the neutral line of the plate: I00 = r r = ⇒ I00 = 54 mm4
12 12
• Inertia of the reinforcement in relation to the neutral line of the plate:
Ixx = I00 + Sr . d 2x = 54 + 18 . 4.5² ⇒ Ixx = 418.5 mm4
with: Sr: reinforcement surface area
d x : distance of the reinforcement centre of gravity to the neutral line of the plate
• Geometrical effectiveness of the reinforcement:

I 00 54
η= = ⇒ η = 0.13
I xx 418.5
• Calculation of equivalent cross section area A :
A = Ap + ηAr = (3 . 3) + 0.13 . (3 . 6) ⇒ A = 11.34
A 11.34
being: = = 0.015
(a r + b r ) t p (160 + 80 ) . 3
This ratio makes it possible to use the chart in Chapter V1-8.4.2.2 concerning the calculation
of the Kt coefficient at the edge of the reinforcement of an elliptical hole in the pure shear
case.
There is:
Kt = 2.2
Let us assume an over-stress at the joint between the reinforcement and the plate equal to:
σmax = Ktσequi = 2.2 . (70 3 ) ⇒ σmax = 266 MPa
V1-8 • 6.3.BUSH INSERTED AS A REINFORCEMENT INSIDE A CIRCU-

LAR HOLE
σx σx
D
d t
with:
D = 160 mm
d = 150 mm
t = 5 mm
σx = 120 Mpa
The plate and the bush are made of different materials:
Plate made of 7175 T7351: Ep = 71000 Mpa
Bush made of 2024 T3511: Eb = 74500 Mpa

D 160
In this case there is: = = 1.06 > 1.05. Consequently, this is a wide reinforcement.
d 150
Now, to take the difference in material between the plate and the bush into account, the
reinforcement is considered as being narrow. The geometrical effectiveness coefficient η is
calculated and the charts in paragraph V1-8.4.2.2 are used to calculate Kt.
• Calculation of the geometry effectiveness coefficient η:
I 00 E r E
There is η = = r as in this case: I00 = Ixx.
I xx E p E p
74500
Giving: η = = 1.05
71000
• Calculation of the equivalent cross section area A :
A = Ap + η . Ar = 0 + 1.05 . (10 * 5) ⇒ A = 52.5
A 52.5
being: = = 0.065
(a r + b r ). t p (80 + 80) . 5
This ratio makes it possible to use the charts in Chapter V1-8.4.2.2 concerning the calculation
of coefficient Kt at the edge of a narrow reinforcement in the case of a unidirectional stress
state.
There is:
Kt = 2.2 (instead of 3 without reinforcement)
Being an over-stress at the edge of a hole equal to:
σmax = Ktσequi = 2.2 . 120 ⇒ σmax = 264 MPa
V1-8 • 6.4.REINFORCEMENT FOR A TRIANGULAR HOLE

τ
30
80
3 τ
9
3

As this reinforcement is not symmetrical, it is necessary to take its geometrical effectiveness

coefficient η into account when calculating it.
• Calculation of the coefficient η:
• Inertia of the reinforcement in relation to the axis passing through its centre of gravity
t r h 3r 3 . 6 3
and parallel to the neutral line of the plate: I00 = = ⇒ I00 = 54 mm4
12 12
Ixx = I00 + Sr . d 2x = 54 + 18 . 4.5² ⇒ Ixx = 418.5 mm4
with Sr: surface area of the reinforcement
I 54
η = 00 = ⇒ η = 0.13
I xx 418.5
• Calculation of the equivalent section A :
A = Ap + ηAr = (3 . 3) + 0.13 . (3 . 6) ⇒ A = 11.34
• Calculation of the equivalent triangle:
There is:
ra = 15 mm²
ra 30
br = 15 mm² therefore: = = 0.375
b r 80
rt
With this ratio and the chart in paragraph V1-8.4.2.4, the following is found: = 0.2.
br
• Calculation of the over-stress at the edge of the reinforcement:
To calculate Kt, the chart in paragraph V1-8.4.2.4 concerning the pure shear case is used,
with:
A 11.34
• = = 0.05
b r t 80 . 3
r
• t = 0.2
br
There is: Kt = 3 (instead of 4, without reinforcement)

V1-8 • 6.5.EVOLUTIVE THICKNESS REINFORCEMENT

τ
τ
160
143
A
τ
τ
8 17
143
3 7 11
SECTION AA
There is:
- D = 143 + 25 = 168 mm
- d = 143 mm
D 168
therefore: = = 1.17 > 1.05, this is a wide reinforcement. However, it is not
d 143
symmetrical and therefore the problem is handled with a narrow reinforcement to take the
offset of the neutral fibre between the plate and the reinforcement into account.
• Calculation of coefficient η:
8 17
1
2
3 7 11
• Inertia of the reinforcement in relation to the axis passing through its centre of gravity and
parallel to the neutral line of the plate:
8 . 83
I 001 = = 341.33 mm 4 ; S1 = 64 mm ²
12 ⇒ I00 = (341.33 + 64 . 1²) + (90.66 + 68 . 1²) = 563.9 mm4
17 . 4 3 4
I 002 = = 90.66 mm ; S 2 = 68 mm ²
12


Ixx = I00 + Sr . d 2x = 563.9 + 132 . 4.47² ⇒ Ixx = 3201 mm4
with Sr: reinforcement surface area

I 563.9
η = 00 = ⇒ η = 0.176
I xx 3201
• Calculation of the equivalent cross section area A :
A = Ap + η . Ar = (3 . 25) + 0.176 . (64 + 68) ⇒ A = 98.2 mm²
A 98.2
being: = = 0.0927
(a r + b r ). t p (185 + 168) . 3
This ratio makes it possible to use the chart in Chapter V1-8.4.2.2 concerning the calculation
of coefficient Kt and the edge of a narrow reinforcement in the pure shear case.
There is:
Kt = 1.41 (instead of 2.2, without reinforcement)

Being an over-stress at the edge of the hole equal to:
σmax = Ktσequi = 1.41 . (70 3 ) ⇒ σmax = 171 MPa
Considering that the reinforcement is narrow, only an over-stress on the joint between the
reinforcement and the plate is obtained, and it is assumed that this stress is constant along the
reinforcement width to the edge of the hole.
To take a change to this stress into account, it is necessary to use the method described in the
case of wide reinforcements. However, this method only applies to circular holes with
symmetrical reinforcements. This is not the case here, so it is assumed that the surface area of
the reinforcement is evenly distributed each side of the plate and that the elliptical hole is
replaced by an equivalent round hole.
τ τ
τ τ
160
???
143
equivalent to
A A
A A τ
τ
τ τ
8 17 17
t
143 r'
3 7 11 t'
SECTION AA SECTION AA

with:
- t' so that the radial cross section area of the reinforcement is maintained:
(2 . t' + 3) . 25 = 207 mm²
1 207
⇒ t' =  − 3 ⇒ t' = 2.64 mm

2 25 
a r + b r 160 + 143
- r' such that: r' = = = 151.5 mm
2 2
2 . t' 2 . 2.64 D 2 . (151.5 + 25)

Then there is = = 1.76 and = = 1.16
t 3 d 2 . (151.5)
Using these ratios and the chart in paragraph V1-8.4.2.1.5 on wide reinforcements in pure
shear, the following is found:
- at the edge of the reinforcement at the joint with the plate: Kt = 1.46 being σmax = 177 Mpa
- at the edge of the hole: Kt = 1.63 being σmax = 198 Mpa
Remark: When the reinforcement is considered as being narrow and non-symmetrical, an

over-stress is found at the edge of the reinforcement equivalent to that obtained, assuming that
the reinforcement is wide and symmetrical.
However, the wide reinforcement assumption makes it possible to take into account a change
to the stress of the radial section of the reinforcement. Here, this phenomenon is not negligible
as the stress at the edge of the hole is 10% greater than the stress at the edge of the
reinforcement.

Static stressing manual V1-9 • CLEATS
CONTENTS
Chapter Title Issue Date

V1-9 CLEATS 1 05 / 1997 Original issue
V1-9.1 GENERAL 1 05 / 1997 Original issue
V1-9.2 DATA PREPARATION 1 05 / 1997 Original issue
V1-9.3 CLEAT STIFFNESS 1 05 / 1997 Original issue
V1-9.4 CLEAT EFFECTIVENESS 1 05 / 1997 Original issue
V1-9.5 CHECKING PROCEDURE 1 05 / 1997 Original issue
V1-9.6 CALCULATION EXAMPLE 1 05 / 1997 Original issue
V1-9.7 Appendix 1: Determination of stiffness criterion 1 05 / 1997 Original issue
V1-9.8 Appendix 2: Discussion on stiffness (...) 1 05 / 1997 Original issue
Pages
V1-9 .1 GENERAL .............................................................................................. V1-9•1/1
V1-9 .2 DATA PREPARATION.......................................................................... V1-9•2/1

V1-9 .2.1 MATERIAL DATA ............................................................................................... V1-9•2/1
V1-9 .2.2 GEOMETRICAL DATA....................................................................................... V1-9•2/1
V1-9 3-CLEAT STIFFNESS .................................................................................. V1-9•3/1

V1-9 .3.1 ASSUMPTIONS..................................................................................................... V1-9•3/1
V1-9 .3.2 DISPLACEMENT AND STIFFNESS.................................................................. V1-9•3/1

V1-9 .3.2.1 Cleat bending: Dlbending/Mz ................................................................................ V1-9•3/2
V1-9 .3.2.2 Cleat rotation: ∆lrotation/z .................................................................................... V1-9•3/4
V1-9 .3.2.3 Cleat translation: ∆ltranslation/x ............................................................................ V1/9•3/5
V1-9 .3.2.4 Bending of the flange bearing on the frame: ∆lbending....................................... V1-9•3/5
V1-9 .3.2.5 General equation of cleat stiffness: K.............................................................. V1-9•3/7
V1-9 .3.2.6 Calculation example ........................................................................................ V1-9•3/8
V1-9 .4 CLEAT EFFECTIVENESS.................................................................... V1-9•4/1

V1-9 .4.1 STIFFNESS CRITERION .................................................................................... V1-9•4/1
V1-9 .4.2 NON-RIGID CLEATS........................................................................................... V1-9•4/4
V1-9 .5 CHECKING PROCEDURE................................................................... V1-9•5/1
V1-9 .6 CALCULATION EXAMPLE ................................................................ V1-9•6/1
Page V1-9/1 Revision 1 © Copyright AEROSPATIALE 1998


SYMBOLS USED

cdg centre of gravity
x, y, z local axes associated with the cleat
F (N) load applied to the cleat at the end fastener on the member
ll (mm) distance between the end fastener in the frame and the fasteners in the
stringer
L (mm) distance between each frame ("inter-frame pitch")
W (mm) distance between each stringer ("inter-stringer pitch")
∆l (mm) total deflection at the end of the cleat
∆lbending/z (mm) deflection due to cleat bending
∆lrotation/z (mm) deflection due to cleat rotation
∆ltranslation/x (mm) deflection due to cleat translation
∆lbending (mm) deflection due to cleat flange bending
l2 (mm) width of the cleat flange in contact with the frame
l3 (mm) height of the cleat web (generally evolutive)
Es (MPa) elasticity modulus of the cleat material
Isz (mm4) moment of inertia of the cleat around z in relation to the centre of
gravity of its cross section
Mf (Mfz(y)) (N.mm) bending moment (at abscissa "y" around z)
σy (MPa) normal stress in the cleat
v / v(y) (mm) x dimension between the centre of gravity of the cleat cross section
and the end fibre
σ0 (MPa) ① constant normal stress in the case of a beam of equal strength
② normal stress in the cleat at joint with the stringer
n ① number of segments to discretize the cleat in the bending calculation
② number of elastic supports on the frame flange between two rigid
supports
e (mm) cleat thickness
∆y (mm) length of a discretization segment
Mf0 (N.mm) bending moment around z in the cleat at the root (y = 0)
I0 (mm4) moment of inertia of the cleat around z in relation to the centre of
gravity of its cross section at the root (y = 0)
v0 (mm) x dimension between the centre of gravity of the cleat cross section
and the end fibre at the root (y = 0)
El (MPa) elasticity modulus of the stringer material
Ilz (mm4) moment of inertia of the stringer around z in relation to the centre of


Sl (mm²) stringer cross section
θ (slope) angle of rotation
l4 (mm) length of an equivalent beam for cleat flange bending calculation
l5 (mm) width of an equivalent beam for cleat flange bending calculation
Ibending (mm4) moment of inertia of the cleat flange for bending calculation
G centre of gravity of a cleat cross section
G0 centre of gravity of the cleat cross section at the root (y = 0)
m ① index from 1 to n for elastic support identification between two
rigid supports
② number of spans between two supports
P (N) axial compression load in the frame inner flange
Rm (N) reaction at intermediate support "m"
γ dimensionless parameter used for the calculation of the stiffness
criterion
Ecc (MPa) elasticity modulus in compression of the frame material
Ity (mm4) moment of inertia of the flange around y in relation to the centre of
I'ty (mm4) moment of inertia of the inner flange around y in relation to the centre
of gravity of its cross section, taking the elastic foundation into
account
Xm (mm) displacement of the member inner flange at an elastic support "m"
k parameter used in the stiffness criterion calculation
kcr critical value of k for P = Pcr
r integer
Pcr (N) critical buckling load for the member inner flange
K (N/mm) cleat stiffness
Kmini (N/mm) minimum stiffness of the cleat for it to be considered as a rigid support
in relation to member warping
λ member inner flange end fixity factor

REFERENCES
1. Michael C.Y. Niu: Airframe Structural Design
2. Roark's: Formulas for Stress & Strain
3. Timoshenko: Théorie de la stabilité élastique (Elastic Stability Theory)
4. Bruhn: Analysis and Design of Flight Vehicle Structure
5. Vallat: Résistance des matériaux appliquée à l'aviation (Material Strength Applied to the
Aircraft Industry)


V1-9 .1 GENERAL INFORMATION

Cleats are elements of stiffened panels (refer to paragraph V1-1) with the following functions:
- to join stiffeners and transverse members,
- to prevent premature warping of members (refer to paragraph V1-3).
With fuselage panels, cleats may also provide frame-skin joints.
In the example below (floating frame fuselage panel) the cleats are in the form of angle
sections, one of the flanges of which (evolutive section) is fastened to the stringer and the
other to the frame. These joints are secured by several fasteners (see Figure V1-9 .1-1).
Generally, these fasteners are rivets (stringer side) and bolts (frame side).
Stringer
Frame Cleat
W
y L
W
z
x
Figure V1-9 .1-1: Typical stringer/floating frame/cleat assembly
The purpose of this chapter is to describe a cleat stiffness calculation and verification
procedure to determine whether or not they can be assimilated to rigid supports with regard to
member warping. This chapter is closely associated with the complete warping study
discussed in Chapter V3-3.
The definition of the cleat can be validated when the minimum stiffness criterion is satisfied.
Before reaching this result, a series of iterations is necessary. The problem of not meeting the
General information Page V1-9•1/1 Revision 1 © Copyright AEROSPATIALE 1998

stiffness criterion can be satisfactorily solved in two different and possibly complementary
manners:
1) Redimension a stiffer cleat.
2) Accept that a cleat is flexible and recalculate the critical warping stress of members
using a longer beam segment.
However, these two solutions must not be used inconsiderably:

- in the first case, it may be difficult to obtain the required stiffness, especially as stiffness
does not depend on cleats alone,
- in the second case, there is the risk of lowering the critical warping stress of the frame
flange too much.
Then, it may be possible to envisage reinforcing the inner flange of the members.
The static strength of fuselage panel cleats is checked using the pressurization loads
transferred from the fuselage to the frame. This is discussed in a separate chapter.
The following mechanism must be fully understood with regards to mechanical loads
originating from cleats:
➢ If cleats are very rigid, they perfectly fulfill their role by imposing the wave nodes of the
deformed section of a warped stiffener inner flange. But then, the displacement of the
flange at the cleat is nil. Consequently, the cleat is not the seat of any stress due to stiffener
warping.
➢ If they are too flexible, they do not impose the wave nodes described above. In this case,
stiffener warping would load the cleats.
➢ In one case like in the other, warping is prohibited. Therefore, cleats are parts which are
not loaded, which does not mean that they are useless. It was thus checked in flight that
the keel beam stabilization rod of the A300/310 was not the seat of any stress. However, the
absence of this rod would make the keel beam incapable of bearing ultimate loads.
General information Page V1-9•1/2 Revision 1 © Copyright AEROSPATIALE 1998

V1-9 .2 DATA PREPARATION

In the remainder of this document the study is limited to a typical fuselage panel and frame
warping. Transposing the method to other cases should not cause any problem.
V1-9 .2.1 MATERIAL DATA

The calculations carried out in this chapter only call for the following material values:
- the elasticity modulus,
- the elasticity modulus in compression.
During the cleat checking process it may be necessary to recalculate frame warping. The
material data simply needed for this calculation are not given here. Refer to Chapter V3-3
which specifically deals with this phenomenon.
V1-9 .2.2 GEOMETRICAL DATA

The main dimensions of the frame/stringer bays are known, being:
- the inter-frame pitch: L
- the inter-stringer pitch: W
Stringer
Frame
Cleats
Figure V1-9 .2.2-1: Main dimensions of a bay (2 frames + 2 stringers)
The dimensions of the cleats are given in Figure V1-9 .2.2-2.
Data preparation Page V1-9•2/1 Revision 1 © Copyright AEROSPATIALE 1998

Cleats
l4
l3min
l5
e
l1
Stringer Frame i + 1
l2
l3max L
Frame i
Figure V1-9 .2.2-2: Dimensions required for the cleat calculation
Remarks:
- In Figure V1-9 .2.2-2, the dotted areas correspond to areas of cleats used to join the frame
and the fuselage skin, to transfer pressurization loads, for example. The entire study in this
chapter only concerns the stabilization role of the parts in bold lines. In this document, these
parts are called "cleats" and will be considered as separate parts.
- Dimension l1 is measured between the last fastener on the frame web and the fasteners of the
cleat on the stringer.
- l5 is defined by drawing two lines at ± 45° from the last fastener of the frame web and
measuring the distance between the two intersection points of these lines and the edge of the
cleat web (refer to Figure V1-9 .2.2-3). l4
45°
45°
l5
Figure V1-9 .2.2-3: Definition of distance l5

- The cleat calculated here comprises:

1) a flanged edge bearing on the frame web the width of which is constant (l2).
2) a web with height "l3" linearly evolutive between two values "l3min" and "l3max"
3) a constant thickness "e".
Also, it is necessary to know the cross section area and the moment of inertia of the stringer
around z in relation to the centre of gravity of its cross section.
The specific dimensions of the frame inner flange are given in paragraph V1-3 .4.3.


V1-9 .3 CLEAT STIFFNESS

V1-9 .3.1 ASSUMPTIONS
Frame cleats are fastened to the stringers. Therefore, the stringer is considered as participating
in cleat stiffness.
The portion of the cleat used for the frame - skin joint is not considered.
The connection of the cleat to the frame is by means of several fasteners. Only the end fastener
is taken into account by assuming that it is on the compression load-carrying part of the frame
web (refer to Figure V1-9 .3.2-1). Also, bending of the stringer around y due to the offset of
this load will be neglected.
Stringers and frames are considered as straight beams the lengths of which are long in
comparison with the size of their inner flanges and their webs.
V1-9 .3.2 DISPLACEMENT AND STIFFNESS

Cleat stiffness can be estimated considering the displacement of the point of application of
load F on the last fastener. This displacement is the sum of the displacements due to:
- cleat bending (∆lbending/Mz)
- cleat rotation (∆lrotation/z)
- cleat translation (∆ltranslation/x)
- cleat flange in contact with the frame bending (∆lbending)
Frame
Cleat
F
Frame
ll
Stringer
F
L L
y
x
Stringer
F
Cleat
y ll
Cleat
L Stringer
z
x W
L L
y
x
Cleat stiffness Page V1-9•3/1 Revision 1 © Copyright AEROSPATIALE 1998

Figure V1-9 .3.2.-1: Mechanical model of the cleat/frame/stringer assembly

Displacement of the point of application of load F: ∆l
∆l = ∆lflexion/Mz + ∆lrotation/z + ∆ltranslation/x + ∆lbending Eq. 1.9.3.2.1
The stiffness of the cleat is given, by definition, by the relationship:
F
Stiffness: K K= Eq. 1.9.3.2.2
∆l
V1-9 .3.2.1 Cleat bending: ∆lbending/Mz

∆lbending/Mz
F Cleat F
A A
l1 l1
Stringer
y
y x
L L
x
Section AA
l2
l3
Figure V1-9 .3.2.1-1: Cleat bending
Load F induces bending of the cleat around z. The cleat is assimilated to a beam clamped at
the stringer. In compliance with the sign conventions given in Figure V1-9 .3.2.1-1, the
bending moment at position y along the cleat is:
Mfz(y) = F . (ll - y) Eq. 1.9.3.2.1.1
The normal stress due to Mfz is:
M fz ( y) . v( y)
σy = Eq. 1.9.3.2.1.2
I Sz (y )
with:

v(y): x distance from the cross section CG to the fibre at position y

ISz(y): moment of inertia of the cleat around z in relation to the centre of gravity of the
cross section, at position y.
On Figure V1-9 .3.2.1-1 above, the dimension "l2" is the width of the flanged edge, whereas
"l3" is the web height. Generally, l3 linearly change from one end of the cleat to the other:
l3 = f (y).
To calculate the deflection at the end of this evolutive cross section beam, an approximation
can be made with a beam of equal strength. Remember that a beam of equal strength is a beam
M fz
for which the value σ0 = , i.e. the normal stress due to bending, is constant over its
 I Sz 
 
 v 
entire length (in particular refer to document ref. (6) Chapter X, paragraph 4.7).
I 
Here the bending moment linearly decreases with y, whereas the increasing ratio  Sz 
 v
follows a polynomial law. In our case, the result is that the stress is not constant along the cleat
but decreases towards its free end (refer to Figure V1-9 .3.2.1-2).
σ
Beam of equal strength
σ0
Cleat
y
ll
Figure V1-9 .3.2.1-2: Evolution of normal stress due to bending in the cleat.
The applied stress decreases towards the end where F is applied. The deflection calculated
using an assumption of a beam of equal strength will therefore be greater than the real
deflection.
The procedure consists in discretizing the beam into n segments of equal length ∆y along
which the bending moment and the dimension v(y) are considered as being constant (refer to
Figure V1-9 .3.2.1-3). Reference document (6) gives the equation of the deflection, in
particular as a function of the stress calculated at the clamped end (y = 0).
σ0 n M (y )
∆lbending/Mz = . ∑ i = l fz i . ∆y Eq. 1.9.3.2.1.3
Es F . v( y i )

(refer to Figure V1-9 .3.2.1-3)
with:
Mf0
σ0 = σ(y = 0) = Eq. 1.9.3.2.1.4
 I0 
 
 v0 
(stress at the fixed end)
Mf0: bending moment at y = 0, being Mf0 = F . ll

I0: moment of inertia around z in relation to the cross section CG at y = 0
v0: x distance from the centre of gravity of the cross section to the end fibre at y = 0
Es: elasticity modulus of the cleat material
Mfz (yi): bending moment at y = yi, being Mf (yi) = F . (ll - yi)
v(yi): x distance from the centre of gravity of the cross section to the end fibre at y = yi
l
∆y: length of a cleat segment, being ∆y = l
n
Clamped end F
∆y
(on the stringer)
1 2 3 … i … n-1 n
0 ll y
Cleat
yi
Figure V1-9 .3.2.1-3: Cleat
Therefore, the deflection as a function of the applied load F is formulated as follows:

F . l 2l . v 0 1 n (l − y i )
∆lbending/Mz = . . ∑i = l l Eq. 1.9.3.2.1.5
Es .I0 n v( y i )
A study on the deviations obtained for different values of "n" shows that in a special case
where there is no flanged edge, i.e. when the cross section reduces to a variable height
rectangle, the maximum deviation in relation to an infinite number of segments is obtained for
n = 1. It is then only around 10%.
V1-9 .3.2.2 Cleat rotation: ∆lrotation/z

Mf0 / 2
Mf0 Mf0 / 2
= +
L L L L
Figure V1-9 .3.2.2-1: Super-stringer bending

The clamping moment of the cleat on the stringer is: Mf0 = f.ll;
Application of this moment bends the super-stiffener and therefore entails a rotation of the
cleat assembly. Here the term "super-stringer" includes the stringer plus part of the skin. This
is a super-stiffener as defined in Chapter V3-1.
Mf0 .L
The induced rotation is: θ = with:
6. E l . I lz
- L: inter-frame distance (refer to Figure V1-9 .1-1)
- El: elasticity modulus of the stringer
- Ilz: moment of inertia around z in relation to the super-stringer cross section CG.
The displacement of the point of application of

load F is therefore: ∆lrotation/z
F
∆lrotation/z = ll.sinθ y
x
Being: ll
 F. l l . L  θ
∆lrotation/z = ll.sin  
 6. E l . I lz 
If angle θ remains small, L L

sin θ ≈ θ
Figure V1-9 .3.2.2-2: Cleat rotation
Therefore:
F. l l 2 . L
∆lrotation/z ≈ Eq. 1.9.3.2.2.1
6. E l . I lz
V1-9 .3.2.3 Cleat translation: ∆ltranslation/x
The transverse load at the clamped end of the cleat on the stringer is F.
Application of this load causes tension/compression of the super-stringer and therefore leads
to translation of the whole cleat.
∆ltranslation/x
Translation of the stringer and therefore y
of the cleat is: -F/2 -F/2 x

F. L
∆ltranslation/x =
2. S l . E l
Eq. 1.9.3.2.3.1
where Sl is the cross section of the stringer + Figure V1-9 .3.2.3-1: Cleat translation
the associated fuselage skin
V1-9 .3.2.4 Bending of the cleat flange in contact with the frame: ∆lbending
Load F induces bending around y of the cleat flange in contact with the frame.
Section BB
F
l5
l4
z y
Figure V1-9 .3.2.4-1: Bending of the flanged edge
This flange is assimilated to a beam clamped at one of its ends (fastener side: refer to Figure
V1-9 .3.2.4-2).
Real structure
Beam model
Figure V1-9 .3.2.4-2: Model of the equivalent beam for flange bending
The maximum deflection of the beam is:
F. l 34
∆lbending = Eq. 1.9.3.2.4.1
3. E s . I bending
where Ibending is the moment of inertia around y of section BB of the cleat.

Remark: Remember that the cross section for bending calculation BB can be found by plotting
two straight lines at 45° from the point of application of the load. Giving l5 = 2.l4.
The compliance of the flange of a cleat, made of folded sheet (the thickness of which is
limited) is relatively high and appreciably reduces the stiffness of the cleat.
V1-9 .3.2.5 General equation of cleat stiffness: K
By carrying forward the equations 1.9.3.2.1.5, 1.9.3.2.2.1, 1.9.3.2.3.1 and 1.9.3.2.4.1 into the
equation 1.9.3.2.1, the following is found:
F . l 2l . v 0 (l l − y i ) F. l 2l . L F. L F. l 34
. ∑ i =1
n
∆l = + + + Eq. 1.9.3.2.5.1
n . Es .I0 v( y i ) 6. E l . I lz 2. S l . E l 3. E s . I bending
By carrying forward this equation to Eq. 1.9.3.2.2, the compliance is:
1 l 2l . v 0 n (l l − y i ) l 2l . L L l 34
= . ∑i =1 + + + Eq. 1.9.3.2.5.2
K n . Es .I0 v( y i ) 6. E l . I lz 2. S l . E l 3. E s . I bending
with:
El: elasticity modulus of the stringer
Ilz: moment of inertia around z in relation to the super-stringer cross section CG
Es: modulus of elasticity of the cleat
L: inter-frame pitch
Ibending: moment of inertia around y of the cleat cross section BB (refer to paragraph
V1-9 .3.2.4)
ll: distance between the end fastener on the frame and the row of fasteners on the
stringer
Mf0: bending moment at y = 0, being Mf0 = F . ll
I0: moment of inertia around z in relation to the cross section CG at y = 0
v0: y distance from the centre of gravity of the cross section to the end fibre at y = 0
Es: elasticity modulus of the cleat material
v(yi): y distance of the centre of gravity of the cross section to the end fibre at y = yi
∆y: length of a cleat segment
e: cleat thickness

Cleats
E
l4
l3min
l5
l1
S
Stringer z Frame i + 1
l2
l3 L
S
G0
v0
x Frame i
Figure V1-9 .3.2.5-1: Dimensions required to calculate the cleat
E
Note: G0 is the cross section CG at clamped end (y = 0)
V1-9 .3.2.6 Calculation example
y
l4
N
l5
T
l1
z
I
Frame i + 1
l2
l3 L
G0
x
v0
Frame i A
Figure V1-9 .3.2.6-1: Reminder of the dimensions required for the cleat stiffness calculation L
Cleat characteristics:
F L ll l2 l3 min l3 max l4 l5 e ∆y
N mm mm mm mm mm mm mm mm mm
-1000 530 80 20 20 30 12 24 2 20
Es I0 v0 ISz n
MPa mm4 mm mm4 4
69000 8920 20.25 16
yi l3i vi
i mm mm mm
1 10 28.75 19.525
2 30 26.25 18.057
3 50 23.75 16.564
4 70 21.25 15.039
Stringer characteristics:
El Ilz Sl
MPa mm4 mm²
73800 26540 177
Displacements:
∆lbending/z = -0,46 mm Refer to Eq. 3.9.3.2.1.5
∆lrotation/z = -0,29 mm Refer to Eq. 3.9.3.2.2.1
∆ltranslation/x = -0,02 mm Refer to Eq. 3.9.3.2.3.1
∆lbending = -0,52 mm Refer to Eq. 3.9.3.2.4.1
∆l = -1,30 mm Stiffness: K = 772 N/mm

Refer to Eq. 3.9.3.2.1 Refer to Eq. 3.9.3.2.2
or Eq. 3.9.3.2.5.1 or Eq. 3.9.3.2.5.2


V1-9 .4 CLEAT EFFECTIVENESS

V1-9 .4.1 STIFFNESS CRITERION
To determine the minimum stiffness that cleats must have, it is necessary to take the problem
of frame stability into account. This problem is studied in detail in Chapter V1-3.
The inner flange of the frame is considered as a continuous beam with a constant cross section
supported by an elastic foundation, rigid supports at its ends, and "n" intermediate elastic
supports of the same stiffness K, also equally spaced (all these supports represent the cleats).
The cross section of the beam defined in this manner comprises the frame inner flange (refer
to paragraph V1-3 .4). Ity is the moment of inertia of the cross section.
The elastic foundation represents the stiffness of the frame web. It is quantified by parameter
β, given in paragraph V1-3 .4. This stiffness is assimilated to additional inertia provided to the
basic beam. It is thus possible to define an equivalent beam without elastic foundation the
inertia of which is I'ty.
Rn Rn-1 R2 R1
P P
x W W W W
z
l
Figure V1-9 .4.1-1: Model of a beam under compression on elastic supports
The first goal can be summarized as being that the intermediate supports (cleats) must not
move if the beam buckles (frame warping). They will then be considered as being absolutely
stiff supports.
Let be K, the stiffness of the intermediate supports. It is proven that the supports can be
considered as being absolutely stiff if K satisfies the equation:
Pcr
K > Kmini = (refer to V1-9 .7)
γ. W
where: - Pcr is the critical Euler load for a beam section of length W
- γ is a parameter depending on the number of cleats n
(refer to document ref. (3), paragraph 2.6)
Cleat effectiveness Page V1-9•4/1 Revision 1 © Copyright AEROSPATIALE 1998

n 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 20
γ 0.5000 0.3333 0.2929 0.2764 0.2679 0.2630 0.2599 0.2578 0.2563 0.2552 0.2543 0.2537 0.2532 0.2528 0.2524 0.2529 0.2519 0.2517 0.2514
0.5000
γ = γ(n)
0.4500
0.4000
0.3500
γ
0.3000
0.2500
0.2000
0 2 4 6 8 10 12 14 16 18 20
n
Figure V1-9 .4.1-2: Evolution of parameter γ as a function of the number of cleats
If there are not many cleats (n<5), it may be useful to take the values of γ indicated in the
table. Otherwise, the asymptotic value of the curve γ = 0.25 has to be used (this is the case for
fuselage frames). The minimum allowable stiffness of the cleat is thus maximized.
It can be noted that the greater the number of cleats, the higher their stiffness has to be to
oppose the same buckling force. This can be explained by the fact that on the beam of length l,
as n increases then W decreases and therefore this is a short beam case. Under these
conditions, as n increases, Pcr will increase and thus the resisting load to be provided during
buckling in the event of support displacement, which directly depends on stiffener K, has to be
high.
The new inertia I'ty is defined by formulating the equation of the critical Euler load for the
beam of length W in two forms:
Critical Euler load for a continuous beam on an elastic foundation:

π 2 . E cc . I ty  2 β. W 4 
Pcr = . 
 p +  Eq. 1.9.4.1.1
W 2
 p . π . E cc . I ty 
2 4
with:
W: length of a span between two supports
Ecc: elasticity modulus in compression of the frame material
Ity: minimum moment of inertia of the frame inner flange (refer to paragraph
V1-3 .4)
β: elastic foundation modulus (refer to paragraph V1-3 .4)
p: number of half waves over length W
Critical Euler load for a continuous beam on two supports:

π 2 . E cc . I' ty
Pcr = Eq. 1.9.4.1.2
W2

Concerning the beam of length W, the two equations of Pcr given below may be equalized.
The inertia of the equivalent beam which comprises the frame inner flange plus the effect of
the elastic foundation is therefore:
 β. W 4 
I'ty = Ity .  p 2 + 2 4  Eq. 1.9.4.1.3
 p . π . E cc . I ty 
Consequently, the stiffness criterion if γ = 0.25 becomes:
4Pcr
K > Kmini =
W
With:
π 2 . E cc . I' ty
Pcr =
W3
Therefore the stiffness criterion may also be formulated as follows:
4 π 2 . E cc . I' ty
K> Eq. 1.9.4.1.4
W3
Remark: Cleat and frame inner flange sizing satisfying the stiffness criterion may be linked by
the equation:
K. W 3
Ecc.I'ty =
4π 2

V1-9 .4.2 NON-RIGID CLEATS

If the stiffness criterion is not satisfied and the cleat is not resized, it is possible to accept that a
cleat is "flexible" in relation to the frame inner flange. The problem then becomes one of the
buckling of a beam on elastic supports, studied in particular in reference document (3),
paragraph 2.6.
The general formulation of the critical Euler load for the stiffener involved on elastic supports
is:
c . π 2 . E cc . I' ty
Pcr =
W2
with:
Ecc: elasticity modulus in compression of the frame material
I'ty: minimum inertia moment of the frame inner flange around y in relation to
its centre of gravity, taking the elastic foundation into account
W: beam length (inter-stringer pitch)
c . π²: dimensionless factor determined using the chart and partly depending on
the stiffness of the intermediate elastic supports (refer to below).
To calculate c . π², a chart has be used with the entry on the abscissa line being a noted
dimensionless quantity (refer to Figure V1-9 .4.2-2).
K.W3
for a beam of length "W" where K is the stiffness of an intermediate elastic
E cc . I' ty
support.
The coefficient "cπ²" for a beam of length "W" is obtained from the charts given in Figure
V1-9 .4.2-2. This chart is valid for a number of supports tending towards the infinite but it can
be accepted that it remains valid in all cases.
10
cπ²
5
0
0 20 40 60
K. W3
E cc . I' ty
Figure V1-9 .4.2-1: Determination of coefficient cπ²

K.W3
For > 40, the following may be taken into account c ≈ 1 and c.π² ≈ π²
E cc . I' ty
This curve may be approximated by a polynomial interpolation. Putting down:
K.W3
X=
E cc . I' ty
For X < 40, there is then;:

cπ² = 0.53591 + 1.1737 . X - 9.1984.10-2 . X2 + 4.2197.10-3 . X3 - 9.5732.10-5 . X4 + 8.2719.10-7 . X5
The maximum deviation between the values read on the chart and those obtained by this
equation is 18.8% for X ≈ 1.25. It must be remembered that this error overlays the
interpretation error made when reading the curve.
1
Putting down λ = , the end fixity factor, the new critical Euler load is then:
c
π 2 . E cc . I' ty
Pcr =
(λ . W ) 2
Thus, if the cleats are not rigid with regard to a beam of length W, they become rigid for a
beam of length λW.
Once the new value of the critical load Pcr has been determined, the allowable warping stress
~ , refer to Chapter V1-3) must be recalculated and checked to make sure that it is still
(σ dr
greater than the applied stress.
Using the notations developed above, the allowable warping stress is:
~ = Pcr < σ
σ ~ Eq. 1.9.4.2.5
dr dr initial
St
with:
St: cross section area of the frame inner flange
Calculation details, especially a possible plasticity correction, are given in Chapter V1-3.


V1-9 .5 CHECKING PROCEDURE
Calculation of cleat flexibility:

BEGINNING
E
1 l2 . v n (l − y i ) l2 .L L l 34
= l 0 . ∑i =1 l + l + + Eq. 1.9.3.2.5.2
K n . Es .I0 v( y i ) 6. E l . I lz 2. S l . E l 3. E s . I bending
 β. W  4
S
Calculation of corrected inertia: I'ty = Ity .  p 2 + 2 4  Eq. 1.9.4.1.2
 p . π . E cc . I ty 
Calculation of the required minimum stiffness: Kmin =

4 π 2 . E cc . I' ty
W 3
Eq. 1.9.4.2.3
S
no yes
E
K > Kmini
yes Resize the no

cleat?
K.W3
Determine c.π² from
N
E cc . I' ty
Eq. 1.9.4.2.1
+ Fig. V1-9 .4.2-1:
10
cπ²
5
T
0
0 20 40 60
K.W3
E cc . I' ty
c. π 2 . E cc . I' ty
I
Calculation of the new critical load: Pcr = Eq. 1.9.4.2.2
W2
Calculation of the new σ dr

~ = Pcr
~ : refer to paragraph V3-3 for the initial calculation, or σ
dr
St A
Eq. 1.9.4.2.3
yes no
no Resize the
Warping?
yes
L
frame?
END
Checking procedure Page V1-9•5/1 Revision 1 © Copyright AEROSPATIALE 1998

The calculation of the stiffness specific to cleats is discussed in paragraph V1-9 .3.
The calculation of the stiffness criterion is given in paragraph V1-9 .4.
If the allowable stress becomes insufficient, it may simply be preferable to resize the frame
inner flange. This avoids defining excessive cleat geometries as indicated in paragraph
V1-9 .1.
Checking procedure Page V1-9•5/2 Revision 1 © Copyright AEROSPATIALE 1998

V1-9 6-CALCULATION EXAMPLE

The example of the frame discussed below first of all comprises the calculation of the initial
warping stress as indicated in paragraph V1-3 .4.8. Here, part of the same notations are partly
reused, here being:
et
bt
Z
CdGflange
ea
ha
W
Figure V1-9 .6-1: Dimensions required to calculate frame warping
- Frame material: 7075 T7351

- Ultimate tension allowable stress: σr = 495 Mpa
- Elasticity modulus in compression: Ecc = 73800 Mpa
- Compression yield stress: σc0.2 = 420 Mpa
- Inner flange thickness: et = 4 mm
- Inner flange width: bt = 16 mm
- Web thickness: ea = 2 mm
- Web height: ha = 110 mm
The inter-frame pitch L in example V1-3 .4.8 is replaced by the inter-stringer pitch.
- Inter-stringer pitch: W = 176 mm
Calculation example Page V1-9•6/1 Revision 1 © Copyright AEROSPATIALE 1998

Inner flange section:

St = et . bt = 4 x 16 = 64 mm²
Moment of inertia of the inner flange at its cross section CG and its principal axes:
(refer to V3-3 .4.3)
e . b 3 4 x 16 3
Ity = t t = = 1365,3 mm4
12 12
Stiffness of the elastic foundation:

3 3
E e  73800  2 
β = cc . a  = .  = 0.111 MPa
4  ha  4  110 
Calculation of p0 (critical number of half waves):

W β 176 0.111
p0 = .4 = .4 = 0.323
π E cc . I ty π 73800 x 1365,3
Giving:
p1 = 0
p2 = 1
Calculation of the initial critical stress:
~ (p ) = 2 . E . I . β = 2 x 73800 x 1365,3 x 0.111 = 105 MPa

σ dr 0 cc ty
St 64
As the number of half waves is necessarily equal to or greater than 1, the following cannot
exist p = 0.
Consequently, p = 1 and:
~ ( p )  1  2
σ  105  1  2 
~
(σ dr )initial = dr 0 2
.   + ( p 0 )  = x   + (0.323) 2  = 509 MPa
2  p 0   2  0.323  
The inertia corrected as a function of the elastic foundation is:
 β. W 4   0.111 x 176 4  4
I'ty = Ity .  p 2 + 2 4  = 1365,3 x 12 + 2 4  = 1380 mm
 p . π . E cc . I ty   1 x π x 73800 x 1365,3 
Therefore, the critical compression load is:
π 2 . E cc . I' ty π 2 x 73800 x 1380

Pcr = = = 32450 N
W2 176 2

The minimum inertia required for the cleat to be considered as being rigid is finally:
4 Pcr 4 x 32450
Kmin = = 737 N/mm
W 176
According to the cleat stiffness calculation discussed in paragraph V1-9 .3.2.6, there is
K = 772 N/mm.
Therefore, K > Kmin, the cleat may be considered as a rigid support.
Now assume that K = 600 N/mm. Therefore there is K < Kmin and the cleat is not rigid.
The cleat is not resized.
It is necessary to determine the coefficient " c.π²":

K . W3 600 x 176 3
= = 32 ⇒ c . π² ≈ 9.54
E cc . I' ty 73800 x 1380
The critical compression load therefore becomes:
c . π 2 . E cc . I' ty 9.5 x 73800 x 1380

Pcr = 2
= = 31234 N
W 176 2
Now the critical warping stress has to be recalculated:
~ = Pcr = 31234 = 488 MPa

σ (to be compared to the 509 Mpa initially obtained)
dr
St 64
Note that a plasticity correction is required to obtain the real warping stress. This aspect of the
calculation is outside the framework of this chapter and therefore not discussed here (refer to
Chapter V1-3).


Technical Manual
MTS 004 Iss. B
External distribution authorised: YES X NO
Static stress manual,

1
4
metallic materials
5
2 4
1
2
5
Structural Design Volume 2

Manuals
Purpose Strength of metallic materials
Scope All programmes.
IT tool Not applicable.

supporting this Manual
Contents V2 - 1 Characteristics of sections

V2 - 2 Behaviour of materials
V2 - 3 Strain measurement
V2 - 4 Bending of beams
V2 - 5 Columns
V2 - 6 Beams-columns
Document Dept code: BTE/CC/CM Validation Name: JF. IMBERT

Manager
Name: J. HUET Function: Deputy Department
Group Leader
Dept code: BTE/CC/A
Date: 11/99
Signature
This document is the property of AEROSPATIALE MATRA AIRBUS, no part of it shall be reproduced or transmitted 3page 1
without authorisation of AEROSPATIALE MATRA AIRBUS and its contents shall not be disclosed.
© AEROSPATIALE MATRA AIRBUS - 1999
Static stress manual, metallic materials - Annex
Reference documents C BE 019: Drawing up of the Structural Justification Dossier
Documents to be consulted See bibliography at the beginning of each chapter.
Abbreviations See Lexique Aerospatiale Airbus/ATR

See "General" paragraph of each chapter
Definitions List of words the definitions of which are integrated into the
Lexique Aerospatiale Airbus/ATR:
Highlights
Issue Date Pages modified Justification of the changes made

A 02/98 V2 - 4 New document.
B 11/99 V2 - 4 Changes as per table page V2-4.i
V2 - 5 New chapter.
V2 - 6 New chapter.
© AEROSPATIALE MATRA AIRBUS - 1999 MTS 004 Iss. B 3Ann. page

Static stress manual, metallic materials - Management information
NOT FOR DISTRIBUTION
List of approval
Dept. code Function Name / First name Signature

BTE/CC/CM Chef de Département CAZET G.
Key words Calcul
Bibliography Néant
Distribution list
Dept. code Function Name / First name (if necessary)

BQP/TE Archives Diderot SIBADE Alain
BQP/TE Bibliothèque BQP/TE SIBADE Alain
BTE/SM/MG Bibliothèque Technique BTE BOUTET Fernand
Distribution list managed in real time by BIO/D (Didocost application)
© AEROSPATIALE MATRA AIRBUS - 1999 MTS 004 Iss. B page IG1

SSM V2-4 • BENDING OF BEAMS
CONTENTS
issue date revision
V2-4 BENDING OF BEAMS 1 5/1999 New § 5 and 6

V2-4 1 INTRODUCTION 0 6/1997 Creation
V2-4 2 STRAIGHT BEAMS 1 9/1999 Correction p.2-8
V2-4 3 BEAMS WITH HIGH CURVATURES 0 6/1997 Creation
V2-4 4 BEAMS WITH VARIABLE HEIGHTS 0 6/1997 Creation
V2-4 5 CALCULATING STRESSES IN STRAIGHT BEAMS 0 5/1999 Creation
V2-4 6 CALCULATING STRESSES IN CURVED BEAMS 0 5/1999 Creation
V2-4 1 INTRODUCTION 1-1
V2-4 2 STRAIGHT BEAMS 2-1

V2-4 2.1 PLANE BENDING FORMULA 2-1
V2-4 2.1.1 Conditions for use 2-1
V2-4 2.1.2 Notations and Conventions 2-2
V2-4 2.1.3 Form 2-3
V2-4 2.2 EXAMPLES 2-17
V2-4 3 BEAMS WITH HIGH CURVATURES 3-1

V2-4 3.1 GENERAL 3-1
V2-4 3.2 THEORETICAL BASIS 3-1
V2-4 3.3 PRACTICAL FORMULAS 3-2
V2-4 4 BEAMS WITH VARIABLE HEIGHTS 4-1

V2-4 4.1 CORRECTION DUE TO "M" 4-1
V2-4 4.2 CORRECTION DUE TO "N" 4-1
V2-4 4.3 PRACTICAL FORMULAS AND COMMENTS 4-2
V2-4 5 CALCULATING STRESSES IN STRAIGHT BEAMS 5-1

V2-4 5.1 LINEAR ELASTICITY 5-1
V2-4 5.1.1 Plane bending 5-1
V2-4 5.1.2 Deviated bending 5-1
V2-4 5.1.3 Composite beams 5-4
V2-4 5.2 PLASTICITY 5-6
Issue 1 Contents page •i

V2-4 5.2.1 Plane bending 5-6

V2-4 5.2.2 Deviated bending 5-16
V2-4 6 CALCULATING STRESSES IN CURVED BEAMS 6-1

V2-4 6.1 SYMMETRICAL SECTIONS 6-1
V2-4 6.1.1 Introduction 6-1
V2-4 6.1.2 Plane elastic bending of solid sections 6-1
V2-4 6.1.3 Plane elastic bending of thin profile sections 6-2
V2-4 6.2 ASYMMETRICAL THIN PROFILE SECTIONS 6-8
V2-4 6.2.2 Introduction 6-8
V2-4 6.2.2 Calculating load-carrying section 6-8
V2-4 6.2.3 Example 6-9
Issue 1 Contents page •ii

V2-4 1 INTRODUCTION
The purpose of Chapter V2-4 is two-fold:

- determine internal loads and related distortion values,
- find stress values.
Internal loads
Tables provide algebraic solutions for internal loads and distortion in different cases of straight beams
carrying loads in the elastic range. This concerns simple cases based on the plane bending theory which, when
used with the principle of superposition makes it possible to solve most problems or to obtain a satisfactory
estimate in the most complex cases (refer to the assumptions used in "straight beams"). Stresses are then
determined using conventional formulas associated with the plane bending theory. These simple formulas are
given here.
Stresses
The stress calculation is detailed for beams with a geometry and/or a loading mode, especially the direction
or the point of load application, prohibiting the use of partial or complete results obtained from the plane bending
theory.
For example, this is the case of beams with a non-negligible radius of curvature ("curved beams") or beams
with variable heights.
In these particular cases, stress determination may involve an internal load distribution that is very different
from the one obtained with the plane bending theory.
Issue 0 Introduction page V2-4•1-1

Issue 0 Introduction page V2-4•1-2

V2-4 2 STRAIGHT BEAMS

V2-4 2.1 PLANE BENDING FORMULA
For various loading cases, the following tables give the expressions to be used to obtain, at all points along
the span of a beam, the value of the resulting loads M (bending moment) and T (shear force) and the
characteristics of the deflected beam θ (slope) and y (deflection). The value of the reactions and the value of the
distortion at each end are called "end values". Specific values, determined using specific geometrical or loading
data are called "remarkable values" (for example: minimum and/or maximum values of M…).
Each table covers a loading type case in two steps: the header of the table gives the general expressions,
then the end values and the remarkable values are given for each of type of boundary conditions. The last two
cases concern the imposed point displacement and rotation cases.
V2-4 2.1.1 Conditions for use

The restrictions given below are directly related to the conditions required to qualify the load-carrying mode
of an plane bending beam, these conditions being:
- strain of the beam in a plane common to the loading plane. This plane defines the bending plane,
- no warping and maintaining of the cross sections after strain in a plane normal to the neutral fibre,
- no rotation of sections around a longitudinal axis guaranteeing a pure bending state, i.e. without torsion;
Restrictive conditions:
1) the beam is long (length relatively long in relation to transverse dimensions);
2) the beam may have a solid or thin section but must be stable;
3) the forces applied act in the same longitudinal plane;
4) this plane includes the shear centre of the cross sections (the shear centre is in the longitudinal plane of
symmetry, if there is one. It is comparable with the centre of gravity for solid sections and for thin closed
sections;
5) the cross sections are constant or vary little along the beam;
6) the beam is straight or slightly curved (high radius of curvature in relation to transverse dimensions);
7) the material is homogeneous, isotropic with a linear elastic range and is stressed in this range;
8) the small displacement assumption is used.
Observance of these conditions makes it possible to take into account the deflection at any point due to the
bending moment M (complementary deflection due to T is not taken into account), disregard neutral fibre
elongation and therefore variations in the beam length and makes the stress normal to a point proportional to the
distance separating it from the neutral fibre.
On the contrary, examine the beam taking the following phenomena into account given respectively for the
conditions above when not fulfilled:
1) the interdependency of the phenomena in the three directions as described by the elasticity theory and which
is disregarded in plane bending, may be high when the beam slenderness is low (Poisson effect);
2) general instability due to the longitudinal components of the applied forces (refer to Chapters 5 and 6);
3) the neutral fibres of the various sections are not included in a single plane. This phenomenon generates torsion
(for the normal stress calculation it is acceptable however to break loading down into two bending planes);
4) essentially for thin open sections: lateral buckling, in particular causing an increase in deflections and
contributing to the torsion thus induced in the normal stress calculation.
Issue 1 Straight beams page V2-4•2-1

5) Correction of the shear load in the case of variable height beams;

6) Nonlinear distribution of stresses for a given section and complementary deflection due to non-negligible T in
the case of high initial curvature;
7) Behaviour specific to composite materials or to the plastic range;
8) In general, the problem requires processing by the finite element method.
V2-4 2.1.2 Notations and Conventions

F: applied point load, expressed in N. The terms F, p, Mo, θo, ∆o designate algebraic values.
p: applied load per unit length, expressed in N/mm.
Mo: applied point moment, expressed in mmN.
! All forces, moments and displacements
θo: imposed rotation, expressed in rd.
applied are positive and oriented as shown in the
∆o: imposed deflection, expressed in mm. tables.
Remark: the minimum and maximum algebraic values given in the tables are positive values, i.e., as drawn. For
negative values, the expressions marked MAX and MIN respectively indicate the minimum algebraic and
maximum algebraic value.
RA (RB): reaction at bearing point A (B), expressed in N, positive upwards.
MA (MB): end moment A (B), expressed in mmN, positive when it compresses the top fibres.
M: bending moment, expressed in mmN, positive when it compresses the top fibres.
T: shear force, expressed in N, positive upwards when it concerns the left end of a portion of a beam.
θ: slope, expressed in rd, positive direction given by the axis-system.
y: deflection, expressed in mm, positive direction given by the axis-system.
y All values positive
M T M+∆M
θ y
z
θ x
T+∆T
This convention implies:

ΣFLEFT = T ; ΣFRIGHT = - T ; ΣMT(FLEFT) = M ; ΣMT(FRIGHT) = M ;
in particular: RA = T (x = 0) ; RB = - T (x = l) ; MA = M (x = 0) ; MB = M (x = l)
The stresses obtained are given in MPa in the unit system above.
The formulas given at the top of each table show the binary function 〈x - a〉o.
It is used alone or combined with ordinary algebraic functions and is defined as follows:
Let x be the abscissa of a cross section, δ the location parameter of a load component:
y if x < δ, then 〈x - δ〉o = 0 ;
F if x > δ, then 〈x - δ〉 o = 1 ;
δ
at x=δ, the function 〈x - δ〉o may be undefined as, for example, the
shear force T at position where a point load is applied:
x
x by extension, 〈x - δ〉n is interpreted as the product (x - δ)n 〈x - δ〉o.

V2-4 2.1.3 Form

The form is divided into five tables, each representing a type of load:
1_concentrated load,
2_distributed load,
3_point moment,
4_imposed deflection,
5_imposed rotation.
The formulas used to calculate, at any point along the span, the shear force and the bending moment values
as regards to the resulting forces, and the slope and deflection values as regards to the distorted part, are called
"General Expressions".
Each table is divided into six parts and each part represents a combination of beam boundary conditions:
a_one free end, one clamped end,
b_one end blocked in rotation, one clamped end,
c_one end on a single bearing point, one clamped end,
d_two clamped ends,
e_two free ends,
f_one end blocked in rotation, one end on a single bearing point.
The reactions and the slope and deflection values at ends are called "End Values".
Some significant results are given (not exhaustively) under the title "Remarkable Values".
A complete case is processed by combining the results of one line of the table with the results in the header.

1 - CONCENTRATED LOADS
General expressions
y Shear force: T = RA - F〈x - a〉
o
F
MB Bending moment: M = MA + RAx - F〈x - a〉
a θB
M A x RA x 2 F
x Slope: θ = θA + + − 〈 x - a〉 2
yA MA EI 2EI 2EI
RB
θA M A x 2 RA x 3 F
l Deflection: y = yA + θAx + + − 〈 x - a〉 3
RA 2EI 6EI 6EI
o o o o
If x < a then 〈x - a〉 = 0 ; if x > a then 〈x - a〉 = 1 = ; by extension, 〈x - a〉 is interpreted as the product (x - a)n 〈x - a〉 .
end on a single end blocked in
free end bearing point clamped end rotation
1-a End values
F F -F
a RA = 0 MA = 0 θA = (1 - a)2 yA = (2l - 3l a + a3)
3 2
2EI 6EI
RB = F MB = - F(l - a) θB = 0 yB = 0
A B
Remarkable values
Fl 2 - Fl 3
MMIN = MB ; if a = 0, MMIN = - Fl θMAX = θA ; if a = 0 θMAX = yMIN = yA ; if a = 0, yMIN =
2EI 3EI
1-b End values
F F (l − a) 2
-F
a RA = 0 MA = θA = 0 yA = (l - a)2 (l + 2a)
2 l 12EI
- F (l 2 − a 2 )
RB = F MB = θB = 0 yB = 0
A B 2 l
Remarkable values
 Fl
 M MAX = M A ; if a = 0, M MAX = 2
M 
− Fl
 M MIN = M B ; if a = 0, M MIN =
 2
- Fl 3
yMIN = yA ; if a = 0, yMIN =
12EI
1-c End values
2
F F (l − a) (2l + a) −F a
a RA = MA = 0 θA = (l - a)2 yA = 0
2 l3 4EI l
F a(3l 2 − a 2 ) - F a(l 2 − a 2 )
RB = MB = θB = 0 yB = 0
A B 2 l3 2 l2
Remarkable values
2
 F a(l − a) (2l + a)

M  M MAX = 2 l3
for x = a ; if a = 0.366 l, M MAX = 0.174 Fl
 M MIN = M B ; if a = 0.5773 l, M MIN = − 0.1924 Fl
1/ 2 1/ 2
 −F  a   a 
if a > 0.414 l: y MIN = a( l − a ) 2   for x = l 
 6 EI  2l + a   2l + a 
 − F a( l 2 − a 2 ) 3 l (l 2 + a 2 )
y if a < 0.414 l: y MIN = for x =
 3 EI ( 3l 2 − a 2 ) 2 3l 2 − a 2
3
 −3 Fl
if a = 0.414 l: y MIN = − 9.8 x 10 EI
for x = a


1 - CONCENTRATED LOADS
1-d End values
F
a
(1 − a)2 (1 + 2a) a(1 − a) 2
RA = F MA = - F θA = 0 yA = 0
l3 l2
A B a 2 (3l − 2a) a 2 (1 − a)
RB = F MB = - F θB = 0 yB = 0
l3 l2
Remarkable values
 2 2
a (l − a) l Fl
 M MAX = 2F 3 for x = a ; if a = , M MAX =
M  l 2 8
l l
 if a < : M MIN = M A ; if a = , M MIN = − 0.1481 Fl
 2 3
 l − 2Fa 3 (l − a) 2 2al
 If a < : yMIN = for x =
2 3EI (l + 2a) 2 1 + 2a
yMIN  3
 if a = l : yMIN = − Fl for x = a
 2 192 EI
1-e End values
F
a
(1 − a) − F a(2l − a) (l − a)
RA = F MA = 0 θA = yA = 0
l 6EI l
A B a F a(l 2 − a 2 )
RB = F MB = 0 θB = yB = 0
l 6EI l
Remarkable values
a(l − a) l Fl
MMAX = F for x = a ; if a = , MMAX =
l 2 4
3/2 1/2
 l − F a  l 2 − a2   l 2 − a2 
 If a < : yMIN =   for x = l −  
 2 3EI l  3   3 
yMIN 
 l − Fl 3
 if a = 2 : yMIN = 48EI for x = a
1-f End values
F
a
−F
RA = 0 MA = F(l - a) θA = 0 yA = (l - a) (2l2 + 2al - a2)
6EI
A B F 2 2
RB = F MB = 0 θB = (l - a ) yB = 0
2EI
Remarkable values
MMAX = MA for 0 ≤ x < a ; if a = 0, MMAX = Fl
Fl 2
θMAX = θB ; if a = 0, θMAX =
2EI
− Fl 3
yMIN = yA ; if a = 0, yMIN =
3EI

2 - DISTRIBUTED LOAD
General expressions
pl pl − pa 2
y Shear force: T = RA - pa 〈x - a〉 - x−a
pa 2(l − a)
a MB p p − pa
2 3
θB Bending moment: M = MA + RAx - a x − a − l x−a
x 2 6(l − a)
yA MA
M A x RA x 2 p 3 pl − pa 4
RB Slope: θ = θA + + − a x−a − x−a
θA EI 2EI 6EI 24EI(l − a)
l
RA MA x 2 RA x 3 p 4 pl − pa 5
Deflection: y = yA + θAx + + − a x−a − x−a
2EI 6EI 24EI 120EI(l − a)
If x < a then 〈x - a〉0 = 0 ; if x > a then 〈x - a〉0 = 1 ; by extension, 〈x - a〉n is interpreted as the product (x - a)n 〈x - a〉0.
free end clamped end
bearing point rotation
2-a End values
pl 3pa + pl
RA = 0 MA = 0 θA = (l - a)3
24EI
pa − pa p − pa
a yA = (l - a)3 (3l + a) - l (l - a)3 (4l + a)
24EI 120EI
p + pl 2pa + pl
RB = a (l - a) MB = - (l - a)2 θB = 0 yB = 0
A B 2 6
Remarkable values
Case of an evenly distributed load on l, i.e. a = 0 and pa = pl = p:
− pl 2 pl 3 − pl 4
MMIN = MB = θMAX = θA = yMIN = yA =
2 6EI 8EI
Case of an evenly increasing load on l, i.e. a = 0 and pa = 0:
− pll 2 p l3 − pll 4
MMIN = MB = θMAX = θA = l yMIN = yA =
6 24EI 30EI
Case of an evenly decreasing load on l, i.e. a = 0 and pl = 0:
− pal 2 p l3 − 11pal 4
MMIN = MB = θMAX = θA = a yMIN = yA =
3 8EI 120EI
2-b End values
pl 3pa + pl
RA = 0 MA = (l - a)3 θA = 0
pa 24I
a − pa p − pa
yA = (l - a)3 (l + a) - l (l - a)3 (3l + 2a) θB = 0 yB = 0
24EI 240EI
p + pl − pa p − pa
A B RB = a (l - a) MB = (l - a)2 (2l + a) - l (l - a)2 (3l + a)
2 6l 24I
Remarkable values
 pl 2
 MMAX = M A = − pl 4
M  6 yMIN = yA =
2
 MMIN = M B = − pl 24EI
 3
 pl l 2
 MMAX = M A = − pll 4
M  24 yMIN = yA =
2
 MMIN = M B = − pl l 80EI
 8
 pal 2
 MMAX = M A = − 7 pal 4
M  8 yMIN = yA =
2
 MMIN = M B = − 5 pal 240EI
 24

2-c pl
End values
pa (l − a )3 ( 3l + a ) pl − pa (l − a ) 3 ( 4l + a )
RA = + MA = 0
pa 8 l3 40 l3
a
− pa (l − a ) 3 (l + 3a ) pl − pa (l − a ) 3 ( 2l + 3a )
θA = − yA = 0
48 EI l 240 EI l
A B p + pl 2 pa + pl
RB = - RA + a (l - a) MB = RAl - (l - a)2
2 6
θB = 0 yB = 0
Remarkable values
3  9 3 − pl 3
RA = pl  M MAX = pl 2 for x = l θ MIN = θ A =
8  128 8 48 EI
M 
5 − pl 2 −3 pl
4
RB = pl  M MIN = M B = y = − 5. 4 x 10 for x = 0.4215 l
8  8 MIN
EI
pl l − pl l 3
RA =  M MAX = 2.98 x 10 −2 pl l 2 for x = 0.4472 l θ MIN = θ A =
10  120 EI
2p l
M  − pl l 2
RA = l M
 MIN = M = pl l 4
B
15 y MIN = − 2.39 x 10 −3 for x = 0.4472 l
5 EI
11 − pa l 3
RA = pal  M MAX = 4.22 x 10 −2 p a l 2 for x = 0.329 l θ MIN = θ A =
40  80 EI
9
M  −7 2
RB = pal  M MIN = M B = 120 p a l pl l 4
y MIN = − 3.04 x 10 −3 for x = 0.4025 l
40 EI
2-d End values
pl pa (l − a ) 3 (l + a ) pl − pa (l − a ) 3 ( 3l + 2a )
RA = +
2 l3 20 l3
pa
a − pa (l − a ) (l + 3a ) pl − pa (l − a ) 3 ( 2l + 3a )
3
MA = − θA = 0 yA = 0
12 l2 60 l2
p + pl 2 pa + pl
A B RB = - RA + a (l - a) MB = RAl + MA - (l - a)2
2 6
θB = 0 yB = 0
Remarkable values
 pl 2 l
 M MAX = for x = − pl 4 l
M  12 2 yMIN = for x =
2
 M MIN = M A = M B = − pl 384 EI 2
 12
3 − pll 2 7
RA = pll MA = RB = pll
20 30 20
 M MAX = 2.15 x 10 −2 pl l 2 for x = 0.548 l
 p l4
M  − pl l 2 yMIN = 1.309 x 10-3 l for x = 0.525 l
 M MIN = M B = 20 EI

CONTINUED pl General expressions

y pl − pa 2
pa Shear force: T = RA - pa 〈x - a〉 - 〈x - a 〉
MB 2(l − a )
a p p − pa
θB
Bending moment: M = MA + RAx - a 〈x - a〉2 - l 〈x - a 〉3
x 2 6(l − a )
yA MA
RB M A x RA x 2 p 3 pl − pa 4
θA Slope: θ = θA + + − a x−a − x−a
EI 2EI 6 EI 24 EI (l − a )
RA l
MAx2 RAx3 p 4 pl − pa 5
Deflection: y = yA + θAx + + − a x−a − x−a
2EI 6EI 24EI 120EI(l − a)
2-e End values

2 pa + pl
RA = (l - a)2 MA = 0 yA = 0
pl 6
− pa (l − a ) 2 (l 2 + 2al − a 2 ) pl − pa (l − a ) 2 (7 l 2 + 6 al − 3a 2 )
pa θA = −
a 24 EI l 360 EI l
2 pa + pl 2 pa + pl
RB = − (l − a ) + (l − a ) MB = 0 yB = 0
6 2
A B
pa (l 2 − a 2 ) 2 pl − pa (l − a ) 2 ( 8l 2 + 9 al + 3a 2 )
θB = +
24 EI l 360 EI l
Remarkable values
pl pl 2 l pl 3 − 5 pl 4 l
RA = RB = MMAX = for x = θMAX = θB = yMIN = for x =
2 8 2 24 EI 384 EI 2
pl − 7 pl l 3
RA = l MMAX = 6.41 x 10-2 pll2 for x = 0.5773 l θA =
6 360 EI
pl l pl l 3 p l 4
RB = θB = yMIN = - 6.53 x 10-3 l for x = 0.5195l
3 45 EI EI
2-f End values
2 pa + pl 2
pl RA = 0 MA = (l - a) θA = 0
6
pa − pa p − pa
a yA = (l - a)2 (5l2 + 2al - a2) - l (l - a)2 (9l2 + 2al - a2)
24 EI 120 EI
P + pl p p −p
RB = a (l - a) MB = 0 θB = a (l - a)2 (2l + a) + l a (l - a)2 (3l + a)
A B 2 6EI 24EI
yB = 0
Remarkable values
pl 2 pl 3 − 5 pl 4
MMAX = MA = θMAX = θB = yMIN = yA =
2 3 EI 24 EI
Case of an evenly increasing load on l, i.e. a = 0 and pa = 0
p l2 p l3 − 3 pl l 4
MMAX = MA = l θMAX = θB = l yMIN = yA =
6 8 EI 40 EI
p l2 5 pal 3 − 2 pal 4
MMAX = MA = a θMAX = θB = yMIN = yA =
3 24 EI 15 EI

COMPLEMENT: case of loading non contiguous to B
The case of a load not extended to point B may be processed by applying the principle of superposition.
pl pl
pb pb
pa pa pa
a a b
b
A B A (1) B A (2) B
End values
The end value equations are written as the sum of the equations giving the values in each case (1) and (2), i.e.:
RTotal A = R (A1) + R (A2) MTA = M (A1) + M (A2) θTA = θ (A1) + θ (A2) yTA = y (A1) + y (A2)
RTB = R (B1) + R (B2) MTB = M (B1) + M (B2) θTB = θ (B1) + θ (B2) yTB = y (B1) + y (B2)
l−a
by writing: pl = pa + (pb - pa)
b−a
General expressions
With the end values known, it is possible to calculate the T, M, θo, and ∆o values at any point along the span using the general
equations below:
pb − pa
TTotal = RTA - pa 〈x - a〉 + pb 〈x - b〉 -
2(b − a )
2
[
x−a − x−b ,
2
]
p p p − pa
MT = MTA + RTAx - a 〈x - a〉2 + b 〈x - b〉2 - b
2 2 6 (b − a)
3
x−a − x−b [ ], 3
M TA x RTA x 2
24 EI (b −a ) [ ],
p 3 pb 3 p −p 4 4
θT = θTA + + − a x−a + x−b − b a
x−a − x−b
EI 2EI 6 EI 6 EI
yT = yTA + θTAx +
MTA x 2 RTA x 3
2EI
+
6 EI
p
− a x−a
24 EI
4
+
pb
24 EI
x−b
4
−
pb − pa
120 EI ( b −a ) [ x−a 5
− x−b
5
].

3 - LOADING MOMENT
y General expressions
a MB Shear force: T = RA
Mo θB Bending moment: M = MA + RAx + Mo 〈x - a〉0
x M x R x2 M
yA MA Slope: θ = θA + A + A + o 〈x - a〉
RB EI 2EI EI
θA M x 2
R x3 M
RA l Deflection: y = yA + θAx + A + A + o 〈x - a〉2
2EI 6 EI 2EI
free end bearing point clamped end rotation
3-a End values

a − Mo Mo 2 2
RA = 0 MA = 0 θA = (l - a) yA = (l - a )
EI 2EI
RB = 0 MB = Mo θB = 0 yB = 0
Mo
A B
Remarkable values
− Mol Mol 2
MMAX = Mo θMAX = θA ; if a = 0, θMAX = yMAX = yA ; if a = 0, yMAX = ; ∀x, y > 0
EI 2EI
3-b End values
a l−a Mo
RA = 0 MA = - Mo θA = 0 yA = a(l - a)
l 2EI
a
Mo RB = 0 MB = Mo θB = 0 yB = 0
A B l
Remarkable values
l M l2
MMAX = MB ; if a = l, MMAX = Mo MMIN = MA ; if a = 0, MMIN = - Mo yMAX = yA ; if a = , yMAX = o ; ∀x, y > 0
2 8 EI
3-c End values
a − 3 Mo l 2 − a 2 − Mo (l − a ) (l − 3a )
RA = MA = 0 θA = yA = 0
2 l3 4 EI l
Mo 3 Mo l 2 − a 2 − Mo l 2 − 3a 2
A B RB = MB = θB = 0 yB = 0
2 l3 2 l2
Remarkable values
 Mo 2l 3 − 3l 2 + 3a 3
 M MAX = for x = a + ε ; if a = 0 or a = l, M MAX = Mo
 2 l3
  − M o l 2 − 3a 2 − Mo
M   if a < 0.282 l: M MIN = 2 for x = l ; if a = 0, M MIN =
 M MIN  2 l 2
2 2
  if a > 0.282 l: M MIN = − 3 Mo a(l − a ) for x = a − ε ; if a = 0.577 l, M MIN = − 0.577 Mo
  2 l3
  l
  If a ≤ 3 the case y > 0 does not exist.
 y MAX 
 M (l − a) ( 3a − l ) 3/ 2 3a − l M l2
 y MAX = o for x = l ; if a = 0.721 l (therefore x = 0.475 l ), y MAX = 2.57 x 10 −2 o
  6 EI 3(l + a) 3(l + a) EI

 
y   If a ≥ 0.577 l the case y < 0 does not exist.
 
  − M o (l 2 − a 2 ) x 3 − 2l 3 ( x − a) 2 + l 2 (l − 3a) (l − a) x 2l 3  1
2 4 
1 −  a  a
 y MIN  y MIN = 4 EI for x = 1 − 6  + 9   ;
l 3 2 2
3(l − a )  2 l l 
   
  − Mo l 2
 l
 if a = 0 (therefore x = ), y MIN =
  3 27 EI

3 - POINT MOMENT
3-d End values

a a( l − a ) l 2 − 4al + 3a 2
Mo RA = - 6 Mo MA = - Mo θA = 0 yA = 0
l3 l2
2al − 3a 2
RB = - RA MB = - Mo θB = 0 yB = 0
A B l2
Remarkable values
2 2 3
 4al − 9 a l + 6 a
 MMAX = Mo for x = a + ε ; if a = l, MMAX = Mo
M  l3
3 2 2 3
 MMIN = − Mo l − 4al + 9 a l − 6 a for x = a − ε ; if a = 0, MMIN = − Mo
 l3
  l
  If a ≤ 3 , the case y > 0 does not exist.
 y MAX 
 − M o (l 2 − 4al + 3a 2 ) 3 l( 3a − l ) M l2
 y MAX = for x = if a = 0.768 l ( therefore x = 0.566 l ), y MAX = 1.615 x 10 −2 o
  54EI 2
a (l − a ) 2
3a EI
y 
 2l

  If a ≥ 3 , the case y < 0 does not exist.
 y  − M o ( 2al − 3a 2 ) 3 M l2
MIN l2
  y MIN = 2 2
for x = ; if a = 0.232 l ( therefore x = 0.434 l ), y MIN = − 1.615 x 10 −2 o
  54 EI a (l − a) 3(l − a) EI
3-e End values

a − Mo − Mo 2l 2 − 6 al + 3a 2
RA = MA = 0 θA = yA = 0
l 6 EI l
Mo Mo l 2 − 3a 2
RB = - RA MB = 0 θB = yB = 0
A B 6 EI l
Remarkable values
 l−a
 MMAX = Mo l for x = a + ε ; if a = 0, MMAX = Mo
M 
a
 MMIN = − Mo for x = a − ε ; if a = l, M MIN = − Mo
 l
  If a ≤ 0.423 l, the case y > 0 does not exist.
 
 y MAX  y Mo (− 2l 2 + 6al − 3a 2 ) 3/ 2 2l 2 M l2
 MAX = for x = − + 2al − a 2 ; if a = l (therefore x = 0.577 l), y MAX = 6.42 x 10 −2 o
  9 3EI l 3 EI
y 
 If a ≥ 0.577 l, the case y < 0 does not exist.
  2
 y MIN  − Mo (l 2 − 3a 2 ) 3/2 l2 2 −2 Mo l
  y MIN = for x = l − − a ; if a = 0 ( therefore x = 0.423 l ), y MIN = − 6.42 x 10
  9 3EI l 3 EI
3-f a End values

Mo
RA = 0 MA = - Mo θA = 0 yA = a( 2l − a )
2EI
Mo − Mo a
A B RB = 0 MB = 0 θB = yB = 0
EI
Remarkable values
MMAX = Mo for 0 ≤ x < a
− Mol
θMIN = θB ; if a = l, θMIN =
EI
Mol 2
yMAX = yA ; if a = l, yMAX = ; ∀x, y > 0
2EI

4 - IMPOSED DEFLECTION
y ∆o General expressions
MB Shear force: T = RA - α∆oEI 〈x - a〉 0
a θB Bending moment: M = MA + RAx - α∆oEI 〈x - a〉
x M x R x2 ∆
yA MA Slope: θ = θA + A + A − α o x − a
2
RB EI 2EI 2
θA
l M x2 R x3 ∆ 3
RA Deflection: y = yA + θAx + A + A − α o x − a
2EI 6 EI 6
3
4-a α=
(l − a ) 3
a End values
3∆ o − ∆ o 2l − 3l a + a 3
3 2
RA = 0 MA = 0 θA = yA =
2(l − a ) 2 (l − a ) 3
A ∆o B
EI EI
RB = 3∆o MB = 3∆o θB = 0 yB = 0
(l − a ) 3 (l − a ) 2
Remarkable values
EI 3∆ o
MMIN = MB ; if a = 0, MMIN = 3∆o 2 θMAX = θA ; if a = 0, θMAX = yMIN = yA ; if a = 0, yMIN = - ∆o
l 2l
12l
4-b α=
(l − a ) 2 (l 2 + 2al − 3a 2 )
a End values
l a(l + 2a )
RA = 0 MA = 6∆oEI 2 θA = 0 yA = ∆o 2
l + 2al − 3a 2 l + 2al − 3a 2
A B
∆o 12∆ o EIl − 6 ∆ o EI l+a
RB = MB = 2 θB = 0 yB = 0
(l − a ) 2 (l 2 + 2al − 3a 2 ) l + 2al − 3a 2 l − a
Remarkable values
EI EI
MMAX = MA ; if a = 0, MMAX = 6∆o MMIN = MB ; if a = 0, MMIN = - 6∆o 2 yMIN = yA ; if a = 0, yMIN = - ∆o
l2 l
12l 3
4-c α=
a 2 (l − a )2 ( 3l 2 − 2al − a 2 )
a
End values
2
2l + a l
RA = 6∆oEI MA = 0 θA = - 3∆o yA = 0
A B a 2 ( 3l 2 − 2al − a 2 ) a( 3l 2 − 2al − a 2 )
∆o
6 ∆ o EI ( 3l 2 − a 2 ) − 6 ∆ o EIl(l 2 − a 2 )
RB = MB = θB = 0 yB = 0
2 2
a(l − a ) ( 3l − 2al − a ) 2
a(l − a ) 2 ( 3l 2 − 2al − a 2 )
Remarkable values
 EI 2l + a EI
 M MAX = 6∆o 2 2 for x = a ; if a = 0.366 l, M MAX = 18.1753∆ o 2
M  ( 3l − 2al − a ) a l
 M MIN = M B ; if a = 0.5773 l, MMIN = − 25.6474 ∆ o EIl −2
1/ 2 1/ 2
 l3  a   a 
 if a > 0.414 l: y MIN = − 2∆ o   for x = l  
 a( 3l 2 − 2al − a 2 )  ( 2l + a )   2l + a 
y  if a = 0.414 l: y MIN = − ∆ o for x = a
 l3 (l − a) (l + a) 3 (l 2 + a 2 )
 if a < 0.414 l: y MIN = − 4 ∆ o 2 2 2 2 2 for x = l 2
 a( 3l − 2al − a ) ( 3l − a ) 3l − a 2

4 - IMPOSED DEFLECTION
3l 3
4-d α=
a (l − a) 3
3
a
End values
l + 2a l
B RA = 3∆oEI 3 MA = - 3∆oEI θA = 0 yA = 0
A a (l − a) a 2 (l − a )
∆o 3l − 2a l
RB = 3∆oEI MB = - 3∆oEI θB = 0 yB = 0
a (l − a ) 3 a (l − a ) 2
Remarkable values
M EI l EI
 MAX = 6 ∆ o a(l − a ) for x = a ; if a = 2
, M MAX = 24∆ o 2
l
M 
l l 81 EI
 if a < : M MIN = M A ; if a = , M MIN =− ∆o 2
 2 3 2 l
 l l3 2al
 if a < : yMIN = − 2∆ o 2 for x =
yMIN 2 (l + 2a ) (l − a ) 1 + 2a

 if a = l : yMIN = − ∆ o for x = a
 2
3l
4-e α=
a 2 (l − a ) 2
a End values
l 2l − a
RA = 3∆oEI MA = 0 θA = - ∆o yA = 0
B a 2 (l − a ) 2a(l − a )
A
∆o l l+a
RB = 3∆oEI MB = 0 θB = ∆o yB = 0
a (l − a ) 2 2a(l − a )
Remarkable values
EI l EI
MMAX = 3∆o for x = a ; if a = , MMAX = 12∆o 2
a( l − a ) 2 l
3/ 2 1/ 2
 l − ∆o  l 2 − a2   l 2 − a2 
 if a < : yMIN =   for x = l −  
 2 a (l − a ) 2  3   3 
yMIN 
 if a = l : y
 MIN = − ∆ o for x = a
2
3
4-f α=
(l − a ) 2 (l + 2a )
a End values
1 2l − a
RA = 3∆oEI 2 MA = 0 θA = - ∆o yA = 0
B a (l − a ) 2a(l − a )
A
∆o 1 l+a
RB = 3∆oEI MB = 0 θB = ∆o yB = 0
a (l − a ) 2 2a(l − a )
Remarkable values
EI
MMAX = MA for 0 ≤ x < a ; if a = 0, MMAX = 3∆o
l2
3 ∆o
θMAX = θB ; if a = 0, θMAX =
2 l
yMIN = yA ; if a = 0, yMIN = - ∆o

5 - IMPOSED ROTATION
y θo General expressions
a MB Shear force: T = RA
θB Bending moment: M = MA + RAx + αθoEI 〈x - a〉0
yA
x M x R x2
MA Slope: θ = θA + A + A + αθo 〈x - a〉
RB EI 2EI
θA
M x2 R x3 θ
RA l Deflection: y = yA + θAx + A + A + α o 〈x - a〉2
2EI 6 EI 2
1
5-a α=
l−a
θo
End values
a l+a
RA = 0 MA = 0 θA = - θo yA = θo
2
EI
RB = 0 MB = θo θB = 0 yB = 0
A B l−a
Remarkable values
EI
MMAX = θo for a < x ≤ l
l−a
θMAX = θo for 0 ≤ x ≤ a
l
yMAX = yA ; if a = 0, yMAX = θo ; ∀x, y > 0
2
l
5-b α=
a( l − a )
θo
End values
a EI l
RA = 0 MA = - θo θA = 0 yA = θo
a 2
EI
A B RB = 0 MB = θo θB = 0 yB = 0
l−a
Remarkable values
MMAX = MB ; if a → l, MMAX → + ∞
MMIN = MA ; if a → 0, MMIN → - ∞
yMAX = yA ; ∀x, y > 0

4l 3
5-c α=
θo
[
(l − a ) (l − a ) 3 + 4 a 3 ]
End values
a 2
1+ a l ( l − 3a )
RA = - 6θoEI MA = 0 θA = - θo yA = 0
( 1 − a ) 3 + 4a 3 (l − a ) 3 + 4 a 3
A B l ( l 2 − 3a 2 )
RB = - RA MB =- 2θo EI θB = 0 yB = 0
[
(l − a ) (l − a ) 3 + 4 a 3 ]
Remarkable values
3 2 3
 EI 2l − 3al + 3a EI
 M MAX = 2 θ o 3 3
for x = a + ε ; if a = 0, M MAX = 4 θ o if a → l, M MAX → + ∞
 l − a ( l − a ) + 4 a l
  EI
M  if a < 0.282 l: M MIN = M B ; if a = 0, M MIN = − 2 θ o
 M MIN  l
 EI
if a > 0.282 l: M MIN = R A a for x = a − ε ; if a = 0.577 l, M MIN = − 6.468 θ o
  l
  l
  If a ≤ 3 the case y > 0 does not exist.
 y MAX 
 2 l3 ( 3a − l ) 3 / 2 3a − l
 y MAX = θ o for x = l ; if a = 0.677 l ( therefore x = 0.453 l ), y MAX = 0.244 θ o l
  3 3
(l − a ) + 4 a 3
3(l + a) 3(l + a )

 
y   If a ≥ 0.577 l the case y < 0 does not exist.
 
  −θo  2l 3  2l 3  1
2 4 
1 −  a  a
 MIN  y MIN =
y 3 
(l + a ) x 3 − ( x − a ) 2 + l 2 (l − 3a ) x  for x = 1 − 6  + 9   ;
3
( l − a ) + 4a  1− a 
2
3(l − a ) 2
2 l  l 
   
  l −4
  if a = 0. 219 l ( therefore x = 0. 4 l ), y MIN = − 0 . 253 θ o l ; if a = 0 ( therefore x = ), y MIN = θ o l
  3 27
l3
5-d α=
θo a(l − a ) (l 2 − 3al + 3a 2 )
a End values
EI l ( l − 3a )
RA = - 6θo 2 MA = - θoEI θA = 0 yA = 0
l − 3al + 3a 2 a(l 2 − 3al + 3a 2 )
A B l ( 2l − 3a )
RB = - RA MB = - θoEI θB = 0 yB = 0
(l − a ) (l 2 − 3al + 3a 2 )
Remarkable values
2 2 3
 EI 4al − 9 a l + 6 a
 M MAX = θ o for x = a + ε ; if a → l, M MAX → + ∞
 a (l − a ) (l 2 − 3al + 3a 2 )
M  3 2 2 3
 M MIN = − θ o EI l − ( 4al − 9 a l + 6 a ) for x = a − ε ; if a → 0, MMIN → − ∞
2 2
 a (l − a ) (l − 3al + 3a )
  If a ≤ l , the case y > 0 does not exist.
  3
 y MAX  3
 −θo  l  ( l − 3a ) 3 l ( 3a − l )
 y MAX =   2 for x = ; if a = 0.735 l ( therefore x = 0.547 l ), y MAX = 0.1963 θ o l
  54  a  l − 3al + 3a 2 3a
y  2l

  If a ≥ 3 , the case y < 0 does not exist.

 y MIN 
3
θ  l  ( 3a − 2l ) 3 l2
  y MIN = o   2 for x = ; if a = 0.265 l ( therefore x = 0.453 l ), y MIN = − 0.1963 θ o l
  54  l − a  l − 3al + 3a 2 3(l − a )

CONTINUED y θo General equations

MB Shear force: T = RA
a
θB Bending moment: M = MA + RAx + αθoEI 〈x - a〉0
x M x R x2
yA MA Slope: θ = θA + A + A + αθo 〈x - a〉
RB EI 2EI
θA M x2 R x3 θ
l Deflection: y = yA + θAx + A + A + α o 〈x - a〉2
RA 2EI 6 EI 2

3l
5-e α=
θo l − 3al + 3a 2
2
a End values
EI − θ o 2l 2 − 6 al + 3a 2
RA = - 3θo 2 MA = 0 θA = yA = 0
l − 3al + 3a 2 2 l 2 − 3al + 3a 2
A B
θo l 2 − 3a 2
RB = - RA MB = 0 θB = yB = 0
2 l − 3al + 3a 2
2
Remarkable values
 l−a EI
 M MAX = 3θ o EI l 2 − 3al + 3a 2 for x = a + ε ; if a = 0, M MAX = 3θ o l
M  a EI
 M MIN = − 3θ o EI 2 2 for x = a − ε ; if a = l , M MIN = − 3θ o
 l − 3al + 3a l
 
  If a ≤ 0.423 l, the case y > 0 does not exist.
 
y  θ ( − 2l 2 + 6 al − 3a 2 ) 3/ 2 1
y = o for x = ( − 2l 2 + 6 al − 3a 2 ) ; if a = 0.718 ( therefore x = 0.5 l ),
 MAX  MAX 3 3 l 2
− 3 al + 3 a 2
3
 
 θ l
 y MAX = 0.3257 θ o l ; if a = l ( therefore x = 0.577 l ), y MAX = o
  3 3
y 
 
  If a ≥ 0.577 l, the case y < 0 does not exist.
 
 θ o ( l 2 − 3a 2 ) 3 / 2 1 2
 y MIN  y MIN = 2 2
for x = l − (l − 3a 2 ) ; if a = 0.282 l ( therefore x = 0.5 l ), y MIN = − 0.3257 θ o l ;
  3 3 l − 3 al + 3 a 3
  − θ ol
 if a = 0 ( therefore x = 0.423 l ), y MIN =
 3 3

1
5-f θo
α=
a
a End values
EI θo
RA = 0 MA = - θo θA = 0 yA = (2l - a)
a 2
A B RB = 0 MB = 0 θB = - θo yB = 0
Remarkable values
EI
MMAX = θo for 0 ≤ x ≤ a
a
θMIN = θo for a ≤ x ≤ l
l
yMAX = yA ; if a = l, yMAX = θo ; ∀x, y > 0
2

V2-4 2.2 EXAMPLES

EXAMPLE 1
y pa
pb
a EI = 84.108 Nmm²
x
pa = 20 N/mm pa = 5 N/mm
b l = 800 mm a = 210 mm b = 600 mm
l
A B
CALCULATION OF END VALUES
Table 2-c gives the equations to be used in the case of loading contiguous to B (i.e. b=l):
pa (l − a) 3 ( 3l + a) pl − pa (l − a) 3 ( 4l + a) − pa (l − a) 3 (l + 3a) pl − pa (l − a) 3 ( 2l + 3a)
RA = + θA = − ,
8 l3 40 l3 48 EI l 240 EI l
p + pl 2 pa + pl
RB = - RA + a (l - a) MB = RAl - (l - a)2 ;
2 6
and
yA = 0 MA = 0 θB = 0 yB = 0.
The application of the principle of superposition (refer to table 2-"complement") gives, by writing
l−a
pl = pa + (pb - pa) ,
b−a
with, in case (1)

1
pa(1) = pa = 20 N/mm, pl( 1) = pl = N/mm,
3
and in case (2)

1
pa(2) = - pb = - 5 N/mm, pl( 2) = - pl = - N/mm,
3
at
RTotal A = RA(1) + RA(2) = 1944.85 + (- 7.84) RTB = RB(1) + RB(2) = 4035.48 + (- 365.49)
= 1937 N = 3688 N
θTA = θ (A1) + θ (A2) = - 1.263 x 10-2 + 8.98 x 10-5 MTB = M B(1) + M B(2) = - 784.1 x 103 + 27.5 x 103
= - 1.25 x 10-2 rd (- 0.718°) = - 756.6 x 103 mmN
RTB may also be calculated using the general equations in table 2-"complement", i.e.:
RTB = - TT (x = l), {x = l} ⇒ {x−a n

= ( x − a) n ; x − b
n
= ( x − b)n }
pb − pa
RTB = - RTA + pa (800 - a) - pb (800 - a) +
2( b − a )
[
( 800 − a ) 2 − ( 800 − a ) 2 ]
= - 3688 N.
The sign convention of paragraph 2.1.2 indicates that RTA and RTB are directed upwards, MTB compresses the
fibres at y < 0 and that θTA is oriented clockwise.

CALCULATION OF MTMAX
p b − pa
2( b − a ) [ ] = 0:
2 2
x is calculated such that TT = 0, i.e. TT = RTA - pa 〈x - a〉 + pb 〈x - a〉 - x−a − x−b
- for x ≤ a ; {x−a n
=0; x −b
n
=0 }
e
TT = ct ;
- for x ≥ b ; {x−a n
= ( x − a) n ; x − b
n
= ( x − b)n }
dTT p − pa
= - pa + pb - b [2(x - a) - 2(x - b)] = 0 therefore TT = cte ;
dx 2( b − a )
- for a ≤ x ≤ b ; {x−a n
= ( x − a)n ; x − b
n
=0 }
pb − pa
TT = RTA - pa(x - a) - (x - a)2 ; {TT = 0} ⇔ {x = 316,3 mm} ;
2( b − a )
the equation of the bending moment being

p p p − pa
MT = MTA + RTAx - a 〈x - a〉2 + b 〈x - b〉2 - b
2 2 6 (b − a ) [ x−a 3
− x−b
3
],
the value sought is
pa p − pa
MTMAX = MT (x = 316.3) = MTA + 316.3 RTA - (316.3 - a)2 - b (316.3 - a)3
2 2( b − a )
= 506.3 x 103 mmN ;
MTMAX compresses the fibres at y > 0.
CALCULATION FOR ANY VALUE OF x
θT1, θT2, θT3 are calculated respectively for x = 100 mm, x = 400 mm and x = 700 mm.
M x R x2 p
θT = θTA + TA + TA − a x − a + b x − b −
EI 2 EI 6 EI
3 p
6 EI
3 pb − pa
24 EI (b − a )
4
x−a − x−b [ 4
]
For x = 100 ; {x−a n
=0; x −b
n
=0 , }
MTA x RTA 100 2
θT1 = θTA + +
EI 2EI
= - 1.14 x 10-2 rd (- 0.625°) ;
θT1 is oriented clockwise.
For x = 400 ; {x−a n

= ( x − a)n ; x − b
n
=0 ,}
MTA x RTA 400 2 p pb − pa
θT2 = θTA + + − a ( 400 − a ) 3 − (400 - a)4
EI 2EI 6 EI 24 EI (b − a )
= 0.34 x 10-2 rd (0.195°) ;
θT2 is oriented counter clockwise.
For x = 700 ; {x−a n

= ( x − a) n ; x − b
n
= ( x − b)n , }
2
MTA x RTA 700 p p pb − pa
θT3 = θTA +
EI
+
2EI
− a (700 − a) 3 + b (700 − a) 3 −
6 EI 6 EI 24 EI ( b − a)
(700 − a) 4 − (700 − b) 4 [ ]
= 0.68 x 10-2 rd (0.390°) ;
θT3 is oriented counter clockwise.

GRAPHIC ILLUSTRATION
y Equilibrium
3688 N
20 N/mm
5 N/mm - 756.6 x 103 mmN
1937 N
x
A B
2000 1937
1000
x = 316.3 x (mm)
0
0 100 200 300 400 500 600 700 800
-1000
-2000
-3000
TT (N)
-3688
-4000 x=a Shear force
x=b
TT = 1937 - 20〈x - a〉 + 5〈x - b〉 + 1.667 x 10-2 x − a [ 2
− x−b
2
]
600
506.3
400
x = 316.3
200
0
x (mm)
0 100 200 300 400 500 600 700 800
-200
MT (103 mmN)
-400
-600
-800 x=a Bending moment x=b

3 2
MT = - 756.6 x 10 + 1937x - 10〈x - a〉 + 2.5〈x - b〉 + 5.556 x 10 2 -3
[ x−a 3
− x−b
3
] -756.6
0
θ = -1.25 x 10-2 200 300 400 500 600 700 x (mm) 800
0
-2
θ = 0.68 x 10 rd
-0.5
θ = -1.14 x 10-2 rd
-1 x = 700
-1.5
-2 x = 100
θ = 0.34 x 10-2 rd x = 400
-2.5
x=a
yT (mm)
-3 x=b
-3.5
Deflected shape
EI yT = - 105.34 x 10 x - 378.3 x 10 x + 322.8 x - 0.833〈x - a〉4 + 0.208〈x - b〉4 + 2.778 x 10-4 x − a
6 3 2 3
[ 5
− x−b
5
]
6 3 2
EI θT = - 105.34 x 10 - 756.6 x 10 x + 1937 x - 3.333 〈x - a〉 + 0.833 〈x - b〉 + 1.389 x 10 3 3 -3
[ x−a 4
− x−b
4
]
EXAMPLE 2
P
C
EI = 84 x 108 Nmm2
I J K P = 5000 N C = - 1 x 106 mmN
d1 d = 800 mm d1 = 100 mm d2 = 500 mm
d2
d
CALCULATION OF REACTIONS
Transfer of force P from I to J and application of the principle of superposition as schematised.

P
C
I J K
y y y
F=P Mo = - d1P Mo = C
x x x
A B A B A B
a = 0 l = d - d1 a = 0 l = d - d1 a = d1 - d2 l = d - d1
(1) (2) (3)
giving for span [J, K] at:
RJ = RA(1) + RA(2) + RA(3)  (1) F (l − a) 2 ( 2l + a) − 3 Mo l 2 − a 2

 RA = RA(2)(3) =
= 5000 + 1071.43 + 1443.15 = 7515 N
 2 l3 2 l3
RK = RB(1) + RB(2) + RB(3)  F a( 3l − a 2 )
2
where  RB(1) = RB(2)(3) = − RA(2)(3)
= 0 - 1071.43 - 1443.15 = - 2515 N  2 l3
2 2
 M (1) = − F a(l − a ) − M o l 2 − 3a 2
MK = M B(1) + M B(2) + M B(3) M B(2)(3) =
 B
2 l 2
2 l2
= 0 + 250 000 + 10 204 = 260.2 x 103 mmN 
(table 1-c) (table 3-c)
CALCULATION OF THE DEFLECTED SHAPE AT POINT I
Considering spans [I, J] and [J, K] mutually fixed at J by a continuation of the part, the deflection at I is the
algebraic sum of:
- the deflection due to P if the beam, reduced in span [I, J] was fixed at J,
- and the deflection caused by rotation of the section at J.
y y
P
x θJ x
I J I J K
(4) (5)

Therefore, we obtain:
yI = y (4) (5)
I + y I and
−F
y I(4) = (2l3 - 3l2a + a3) ; table 1-a, where F = P, l = d1, a = 0 ;
6 EI
= - 0,1984 mm
y I(5) = xI θJ = - d1 θJ.
The same calculations give, in the case of rotation at:
θI = θ (4) (5) F
I + θ I and θ (4) = (l - a)2 ; table 1-a, where F = P, l = d1, a = 0 ;
I
2EI
= 2.976 x 103 rd
θ (5)
I = θJ.
The assumptions taken for the calculation of reactions (refer to previous page) entail for the determination of θJ
at:
θJ = θ (2) (3) − M o ( l − a ) (l − 3 a )
A + θ A and θ (2)(3) = ; table 3-c, with case (2): Mo = - d1F, l = d - d1, a = 0 ;
A
4 EI l
case (3): Mo = C, l = d - d1, a = d2 - d1 ;
θJ = 10.417 x 10-3 + (- 6.378 x 10-3) = 4.039 x 10-3 rd (0.231°).
The deflected shape at I is finally characterised by:
yI = - 0.1984 + (- 100) x 4.039 x 10-3 = - 0.602 mm and θI = 2.976 x 10-3 + 4.039 x 10-3 = 7.015 x 10-3 rd (0.402°).
GENERAL EXPRESSIONS
Whatever the type of load, the general expressions are similar. Each equation includes the components due to
boundary conditions at A (end at x = 0) and the component expressing the loading effect:
 F 3
y due to concentrated load:  M A x 2 RA x 3  − 6 EI x − a
 yA + θ A x + + 
y due to applied moment:  2EI 6 EI  + Mo x − a 2
 2 EI
boundary condition components/loading component
For example, writing the deflection equation boils down to summing the effects of the various components, i.e. in
this case, by assimilating the reaction at J to a supplementary loading:
y y
PJ
P PI
C C
x x
I J K I J K
By writing PJ = - RJ (due to sign convention, refer to paragraph V1-4 2.1.2) and PI = P.
Boundary condition components: yI ≠ 0 ; θI ≠ 0 ; MI → 0 ; RI → 0.
Loading components:
P 3 P x3 PJ 3 C 2
PI → − I x − 0 = − I ; PJ → − x − d1 ; C → x − d2 .
6 EI 6 EI 6 EI 2EI

The deflection expression for any x is then:
PI x 3 P 3 C 2
y = yI + θIx - − J x − d1 + x − d2 ,
6 EI 6 EI 2EI
or
y = yI + θIx +
C
2EI
x − d2
2
−
1
6 EI [
Px 3 − RJ x − d1
3
] (1).
By successive derivations, we obtain:

θ = θI +
C
EI
x − d2 −
1
2 EI [
Px 2 − RJ x − d1
2
] (2),
M = C〈x - d2〉0 - Px + RJ〈x - d1〉 (3),
T = - P + RJ〈x - d1〉0 (4).
These expressions make up a system of four equations with three unknowns (yI, θI and RJ) which are
evaluated by solving the system {θ(x = d) = 0 ; y(x = d) = 0 ; y(x = d1) = 0}.
Once the specific values at I and the reaction at J have been established, these equations also make it
possible to calculate T, M, θ and y for any x. For example, for x = d1, the θ equation gives 4.039 x 10-3 and
verifies the calculation in the previous paragraph.
The reactions at K are determined by M(x = d) = MK and T(x = d) = - RK (refer to sign convention, paragraph
V1-4 2.1.2).
GRAPHIC ILLUSTRATIONS
0,75 7.5
θ (10-3 rd)
y (mm)
0,5 5
0,25 2.5
x (mm)
0 0
0 100 200 300 400 500 600 700 800
-0,25 -2.5
-0,5 -5
-0,75 -7.5
Deflected shape
600 5000
M (103 mmN)
400
3000
200
1000
0
0 100 200 300 400 500 600 700 x (mm) 800
-1000
-200
-400 -3000
T (N)
-600 Bending moment and shear force -5000

SSM V2-4 • BEAM BENDING
V2-4 3 HIGH CURVATURE BEAMS

V2-4 3.1 GENERAL INFORMATION
The qualification "curved beam" is adopted as soon as an initial curvature radius exists - radius included in the
bending plane. This geometrical characteristic entails:
- a bending mode which is always the composed bending mode (resulting T, M and N for a straight section),
- a specific distribution for normal stresses.
The purpose of this chapter is to provide practical calculation information to quantify the effect of the initial
curvature on stress values.
In practice, this effect can only be assessed by evaluating the R/h ratio in which R is the initial curvature radius of
the average line and h the height of the section.
R
≤ 5: case of high curvatures, special formulas.
h
R
> 5: case of arcs, usual formulas for in-plane bending.
h
R
= 5: error of around 7% (*).
h
R
= 10: error of around 3.5% (*).
h
R
= 15: error of around 2% (*).
h
(*) given for information purposes for a rectangular section.
V2-4 3.2 BASIC THEORY

The assumptions are those of in-plane bending, indicated in paragraph 2.1.1.
The main difference in relation to straight beams resides in the difference of the initial length of fibres in function
of the radius, entailing:
- a non-linear distribution of normal stresses - the stress remains proportional to the deformation but is not
proportional to the distance to the neutral axis,
- a neutral axis that does not pass by the centre of the section.
a) single bending
e
σ σ
G G y
A.N.
A.N.
r
(S)
M M
rG rN
0' 0 R G: initial curvature radius

rN: neutral fibre radius
r: radius of a calculation point
0
G: centre of the section
A.N.: neutral axis
Revision 0 High curvature beams page V2-4•3-1

Hooke's law, then section (S) equilibrium, lead to a normal stress of the following form:
y
Hooke's law → σx = E εx = E ω ,
rN + y
 M
 σ dS = N = 0

( S )
x ∫ ω = ESe ,

equilibrium →  ⇒ S
rN = .
(S)
∫
 σ x ydS = M 
 ∫
dS
r
 (S)
This stress obeys a hyperbolic distribution and increases very quickly at internal fibres when R/h decreases.
The practical equation is:
M y
σx = .
Se rN + y
b) normal load
In this case, the equilibrium in the section gives:
 σ dS = N
∫
( S )
x

ω=
N
,
 ⇒  ES
∫
 σ x ydS = M = 0
( S )
rN = 0.
Therefore there is no effect of initial curvature on the contribution of the normal load in the calculation of σx.
c) combined bending
By superposition, the stress becomes:
N M y
σx = + .
S S( rG − rN ) rN + y
d) shear load
For curved beams under shear and bending, the negligible effect of curvature on longitudinal shear stress is
proven:
τexact = k x τ(neglected curvature) with 1 < k < 1.1.
Therefore, τ is calculated with straight beam formulas.
V2-4 3.3 PRACTICAL FORMULAS

The position of the neutral axis given by rN is the only unknown factor in the σx equation. The following table
calculates the normal stress value for most cross sections encountered using four distinct formulas.
For tubular sections, the difference between rN(total cross section) and rN(empty cross section) must be determined.
For the case of any type of cross section it is therefore necessary to calculate the integral ∫ dS / r (reducing dS to
the x dr form) or, if this proves to be difficult, to break down the cross section into simple elements to approach
the rN value (refer to composed cross sections below).
It is important to note that rN is independent from the width b of the cross section (refer to "rectangular cross
section" and "elliptic cross section"); for combined sections, rN only depends on the width ratios between the
sections (refer to trapezoidal sections and combined sections). For example, this makes it possible to assimilate a
parallelogram, the base of which perpendicular to r, with a rectangle of the same height. This is only valid for
the calculation of rN, S has to be calculated with the characteristics of the real cross section.

M y
Normal bending stress:
S (rG − rN ) rN + y
Rectangular cross section

S bh h
rN = = r → rN =
h dS r
∫
e
dr ln e
(S)
r b
r ∫
ri
ri
re
Square cross section, side a
h
make h = a
ri b
0 0'
be Trapezoidal cross section

h( bi + be )
rN =
 b r − be ri re 
2 i e ln − ( bi − be )
h  h ri 
h(1 + α ) b
where rN = with α = e
re  re − αri re  bi
2 ln − (1 − α )
 h ri 
ri bi Triangular cross section
make be = 0 or bi = 0
0 0'
Elliptical cross section
G h2
h rN =
4 2 rG − 4rG2 − h 2 
Circular cross section

rG h
make h = diameter
b
0 0'
b3 Combined sections
h3
rN =
∑h b I I
where rN =
∑h α I I
with α =
bI
r r
b2
h2 ∑ b ln r
I
eI
iI
∑α ln r I
eI
iI
b1
re, ri: outer and inner radii

r4 h1
r3 being for the I opposite:
h1 b1 + h2 b2 + h3 b3
r2 r rN =
1 b1 r r r
b b1 ln 2 + b2 ln 3 + b3 ln 4
h r1 r2 r3
0 0'


V2-4 4 VARIABLE HEIGHT BEAMS

A variable height beam is a beam for which a dimension measured parallel to the shear loads varies continuously
over its length. Therefore, this definition excludes beams with an abrupt cross-section variation.
This variation means that a correction has to be made to the longitudinal slide stress calculations, which concern
the shear load(*) effectively found in a cross section:
TW
- bent straight beam: τ = , with W: static moment, I: Moment of inertia, b: width.
Ib
T W
- bent variable height beam: τ = hv with Thv ≠ T.
Ib
(*)
The normal load is not corrected owing to the smallness of the angles involved.
The general assumptions are those of in-plane bending given in paragraph 2.1.1. The sign convention selected is
as follows:
M>0 T>0 N>0
In practice, the following assumptions have to be added:
- the cross sections involved with the area (S), are close to cross sections perpendicular to neutral fibres,
- the stresses due to M and N converge at 0,
N My
- the normal component, perpendicular to (S), of these stresses is given by: σ = + .
S I
V2-4 4.1 CORRECTION DUE TO "M"

λ h: variable height
T 0: point of convergence of outer and inner fibres
λ: distance of the cross section involved to point 0
f1 M: bending moment acting on section (S)
τ1 α1
M T: shear load acting on section (S)
σ(M) σ(M): stress due to M
h 0 T': corrective shear load
σ(M) α2
T' τ1
f1
(S)
Summing on (S) of the tangential component of the stresses due to M, gives:

M
T' = - (tg α1 + tg α2),
h
value to be added algebraically to the shear load T acting on (S)

M h
TREDUCED = T + T' = T - tg α considering α1 and α2 to be sufficiently low so that tg α1 + tg α2 ≈ tg α = ;
h λ
V2-4 4.2 CORRECTION DUE TO "N"

Generally, this correction is not necessary as with a bent beam the contribution of M is by far more preponderant
than that of N.
Revision 0 Variable height beams page V2-4•4-1

This correction may be used when the centre of the cross section is highly offset in height in relation to point 0, a
case in which the studied section is no longer normal to the neutral axis.
λ
h: variable height
0: point of convergence of external and internal fibres
T λ: distance to point 0 of the cross section involved
N N: nominal load acting on the section (S)
σ(N) σ(N): stress due to N
T'' O T": corrective shear load
h y o τ2
f2 βo
G
≠π/2
mean line
(S)
Summing on (S) of the tangential component of the stresses due to N gives:

T'' = N tgβo,
value to be added algebraically to the shear load T acting on (S).
TREDUCED = T + T'' = T + N tgβo.
V2-4 4.3 PRACTICAL FORMULAS AND COMMENTS

1 - Equivalent height: the equation of T' (correction due to M) gives parameter h which is the height of the cross
section involved.
h cannot be the geometrical height: we have seen that T' is the sum of the tangential component of
! normal stresses at (S). However, for a shear bowed beam, the law of distribution of normal stresses on
(S) σ is no longer that of the longitudinal slip stresses τ. Therefore, from the shear standpoint, the beam
height is considered as an equivalent height defined by:
I  I: moment of inertia of the section,
Hequivalent = with 
WMAX Wmax : maximum static moment.
The shear load corrected of the M effect alone is then expressed as:
M Hequivalent
TREDUCED = T - tg α, with tg α = (if tg α1 + tg α2 ≈ tg α).
Hequivalent λ
G1
(*)
For thin web beams with a high web height , compared to that of the flanges, the height
Hequivalent to be considered is the distance d separating the C.G.'s G1 and G2 of the flanges.
(*) In this study, this relative magnitude criterion is true if the following checks out:
I(beam total cross section) ≈ S(flanges)d2. d
The general equation for the shear load corrected for the effect of M and N is:
M
TREDUCED = T - (tg α1 + tg α2) + N tg βo.
Hequivalent G2
2 - Relationship between TREDUCED and T: typically, h increases with M, T' is then in the opposite direction to T
and often less than it. The corrected shear load is then generally in the same direction as T and less than it, which
gives the name TREDUCED.
3 - Specific case: if the resultant of the vertical loads acting on the beam is applied at 0 then T' = -T.
Revision 0 Variable height beams page V2-4•4-2

V2-4 5 CALCULATING STRESSES IN STRAIGHT BEAMS

V2-4 5.1 LINEAR ELASTICITY
A straight, homogeneous and isotropic beam submitted to bending in the elastic range generates linear
strains and stresses.
The stress at a given point along the axis is proportional to the strain at this point and the distance of this
point from the neutral fibre.
V2-4 5.1.1 Plane bending
“Elastic” plane bending occurs when the bending moment applied to a section of the beam acts around a
single main axis of the section generating a maximum stress not exceeding the elastic limit. The neutral axis is
then parallel to the moment axis (see fig. 2-4.5.1.1-1).
+y The bending stress at a given point on the section is:
+x Mz x y
σf =
Iz
Mz σf: bending stress (> 0 if tensile).

Neutral axis +z y: distance from neutral fibre along y-axis.
Mz: bending moment.
Iz: inertia of section around its neutral fibre.
Fig. 2-4.5.1.1-1 Distribution of stresses in a beam

submitted to “elastic” plane bending.
V2-4 5.1.2 Deviated bending
“Elastic” deviated bending occurs when the bending moment, applied to a section of the beam, acts around
an axis other than the main axis of inertia.
In this case, the neutral fibre is no longer parallel to the moment axis.
Deviated bending can be taken into account in complex section cases where the main axes are not known.
However, we know that for beams submitted to bending only, the neutral axis passes through the centre of gravity
of the section (see fig. 2-4.5.1.1-2).
+y
Neutral axis
The neutral axis and the bending moment are in the yz
plane.
M
+z
Fig. 2-4.5.1.1-2 Distribution of stresses in a beam

submitted to “elastic” deviated bending.
Issue 0 Calculating stresses in straight beams page V2-4•5-1

© AEROSPATIALE
The stresses can be calculated by projecting the moment onto any perpendicular axis system passing through
the centre of gravity of the section under study.
y
My = M.sinα
M
My Mz = M.cosα
α My, Mz: components of moment applied along y and z
z
Mz respectively.
α: angle between the moment axis and the z-axis.
Fig. 2-4.5.1.1-3 Determining stresses in a section

submitted to elastic deviated bending.
The bending stress at a given point (y,z) of the section is:
( M y . I yz + M z . I y ) .y ( M z . I yz + M y . I z ) .z
σf = 2
−
I y . Iz − I yz I y . I z − I 2yz
Where: Iy, Iz: moments of inertia around y and z respectively.

Iyz: inertia product.
y, z: coordinates of the point along the y- and z-axis.
Specific case: If the y- and z-axis are the main axes (this case occurs if one of the axes corresponds to a
symmetry axis of the section), the inertia product Iyz is then null.
Mz . y My . z
σf = −
Iz Iy
When a beam is submitted to several loading cases, it is good practice, before starting to calculate the
bending stresses, to locate the main axes, the moments of inertia and the components of the various loads on these
axes.
Below, we recall the analytical method and the Mohr’s circle method to determine the main axes and the
corresponding moments of inertia.
• Moments and inertia products along inclined axes
y
Ix1 = Ixx . cos² φ - Ixy . sin2φ + Iyy . sin² φ
y1
Iy1 = Ixx . cos² φ - Ixy . sin2φ + Iyy . sin² φ
I xx − I yy
x1 Ix1y1 = Ixy . cos² φ + . sin2φ
2
ρ + φ x
x1 and y1 are inclined axes.
φ is the angle between x and x1.

© AEROSPATIALE
• Main axes
V y These are the axes where Ixx and Iyy are minimum and
maximum and Ixy is null. When Ixx, Iyy and Ixy are
known, f, IMAX and IMIN are calculated as follows:
U
ρ + φ x  2 I xy 
φ = 1 tan-1  

2  I yy − I xx 
2
I xx + I yy  I xx − I yy 
IMAX = + I 2xy +  
2  2 
2
I xx + I yy 2  I xx − I yy 
IMIN = + I xy + 
2  2 
• Mohr’s circle
Ixy Graphic construction from values Ixx, Ixy and Iyy.
(Iyy, +Ixy) 1 - determine the two coordinate points (Iyy, +Ixy)

2φ and (Ixx, -Ixy).
IMIN IMAX
Ixx, Iyy 2 - plot the circle passing through the two points
the centre of which is on the Ixx, Iyy axis.
(Ixx, -Ixy) 3 - IMIN and IMAX are the intersection points of the
circle with the Ixx, Iyy axis.
I MAX + I MIN 2φ is the angle shown on the figure.
2
-Ixy
Example: Calculating bending stresses at points A, B, C.
y
25
Thickness of section: 2,5 mm.
A C
B S = 237 mm²
Y
M = 105 mm.N Izz = 89557 mm4
Iyy = 22389 mm4
Z
α = 150° Izy = - 33398 mm4
φ z 50 Caution: Mac Inertie gives Izy > 0.
25

© AEROSPATIALE
 2 I xy 
φ = 1 tan −1 
 I yy − I xx
 = 22.4°

2  
2
I xx + I yy 2  I xx − I yy  4
IMAX = + I xy +  = 103337 mm
2  2 
2
I xx + I yy 2  I xx − I yy  4
IMIN = + I xy +  = 8609 mm
2  2 
Let us project moment M onto the axes (z,y) Let us project moment M onto the axes (Z,Y)
Mz = M . cos α = 105 x cos 150° = - 86600 mm.daN Mz = M . cos (α - φ) = - 61000 mm.daN

My = M . sin α = 105 x sin 150° = 50000 mm.daN My = M . sin (α - φ) = 79230 mm.daN
Coordinates in the z,y axis system. Coordinates in the Z,Y axis system.
ZA = - 23.75 ZA = zA . cos φ + yA . sin φ = - 12.4

YA = 25 YA = yA . cos φ - zA . sin φ = 32.2
ZB = - 23.75 ZB = zB . cos φ + yB . sin φ = - 13.4

YB = 22.5 YB = yB . cos φ - zB . sin φ = 29.9
ZC = 1,25 ZC = zC . cos φ + yC . sin φ = 10.7
YC = 25 YC = yC . cos φ - zC . sin φ = 22.6
Calculating bending stresses at points A, B and C Calculating bending stresses at points A, B and C
( M y . I yz + M z . I yy ) . y ( M z . I yz + M y . I zz ) . z Mz . Y M y . Z
σf = − σf = −
I yy . I zz − I 2yz I yy . I zz − I 2yz Iz Iy
σf = - 4.056 . y - 8.284 . z σf = - 0.59 . Y - 9.203 . Z
σfA = - 4.056 x 25 - 8.284 x (- 23.75) = 95.3 MPa σfA = - 0.59 x 32.2 - 9.203 x (- 12.4) = 95.1 MPa
σfB = - 4.056 x 22.5 - 8.284 x (- 23.75) = 105.5 MPa σfB = - 0.59 x 29.9 - 9.203 x (- 13.4) = 105.7 MPa
σfC = - 4.056 x 25 - 8.284 x 1.25 = - 111.8 Mpa σfC = - 0.59 x 22.6 - 9.203 x 10.7 = - 111.8 MPa
V2-4 5.1.3 Composite beams
A beam is said to be composite if it consists of several materials with different moduli of elasticity.
The strain in the elastic range is assumed to be linear and proportional to the distance between the neutral
fibre and the calculation point. The elongation-stress relation is:
e= σ
E
E = modulus of elasticity of a beam comprising several materials; it is not a constant.

© AEROSPATIALE
y
AiEi
y i'
G
+
yi
y
n
Area: AE = ∑A Ei =1
i i
∑y A E
i =1
i i i
Centre of gravity: y= n
∑A E
i =1
i i
c
Static moment: WiE = ∑y A E
i =1
'
i i i
n n n
∑ ∑ ∑A E
2
Moment of inertia: IE = y i2 A i E i + Eoi Ei − y i i
i =1 i =1 i =1
where:
n = number of elements.
c = number of the element where CG is at yi.
Ioi = inertia of element i itself.
As: yi = y + y i'
n n
Then: IE = ∑i =1
y i' 2 A i E i + ∑I
i =1
oi Ei
The stresses at any point i = j of the section are:
M . E j . y 'j
Bending: σfj = M: bending moment
EI
N . Ej
Normal: σnj = N: normal force
AE
T . Wj . E j
Shear: τj = T: shear force
b j . IE
bj: width of section at point j

© AEROSPATIALE
V2-4 5.2 PLASTICITY
V2-4 5.2.1 Plane bending
V1-5 5.2.1.1 Introduction
Plane “plastic” bending is nonlinear behaviour specific to ductile materials. Indeed, these materials
frequently exceed the maximum bending force calculated in linear elasticity.
The characteristics of beams submitted to “plastic” bending show that the distribution of the strains remains
linear whereas the distribution of the stresses is nonlinear. The latter depends on the shape of the beam and the
stress-strain characteristics (figure V2-4 5.2.1.1-1).
+y
+x
Mz
Neutral fibre +z
Fig. 2-4.5.2.1.1-1 Distribution of stresses in a beam submitted to

plastic bending.
V1-5 5.2.1.2 Calculating bending module
On the figure below, σMAX is the stress applied to the extreme fibre of a symmetrical section. The Cozzone
method is based on an approximation of the profile of the real stresses; the real profile is replaced by a
trapezoidal profile as shown on figure 2-4.5.2.1.2-1.
σMAX
σ0 Real stress distribution

σ0
σMAX
Fig. 2-4.5.2.1.2-1 Cozzone’s equivalent stress profile
Stress σ0 is a fictive stress supposed to exist at the neutral fibre of the axis.
Cozzone’s method enables a fictive allowable bending stress called “bending modulus” to be calculated.
This stress can be compared with the maximum stress calculated linearly.
The general bending modulus expression is: σb = σMAX + σ0 . (k - 1)

© AEROSPATIALE
Remarks:
c The bending modulus calculation is based on the real required maximum stress level, that
is, generally σMAX = σR at ultimate strength for a calculation at ultimate load. At limit load, the stress level must
remain below the elastic limit.
d The bending modulus also depends on the geometrical characteristics of the section by
means of term “k”. “k” is determined by the following equation:
D.W
K= D D
I
Where:
W = static moment of the 1/2 section.
I = moment of inertia of the complete section.
The values of k are given on the figure below for several symmetrical sections:
K 1.0 1.5 2.0 1.33
SECTION
1.85
1.75
d di
1.65
k 1.55
3
d 
1.45 1 −  i 
 d 
K = 16 .
3.π d 
4
1.35 1 −  i 
 d 
1.25
0.0 0.2 0.4 0.6 0.8 1.0 1.2
d/d

© AEROSPATIALE
e2
b b
1.70
e2/h = .5 e1
1.60 h h
e2/h = .4 e1
Section Shape Factor, K
e2/h = .3
1.50 e2
1.40
1.30
K= W
I/D
1.20 e2/h = .2
be 2 ( h − e 2 ) e 1  h 2
e2/h = .1 W= +  − e 2 
2 2 2 
1.10
0.0 .2 .4 .6 .8 1.0 1.2
I/D =
( )
2 be 2 e 2 + 3( h − e 2 )² e 1 ( h − 2 e 2 ) 3
2
e1/b 12 h
b
2.7
2.5 h
e1 e2
Section Shape Factor, K
2.3
e1/b = .1
2.1
e1/b = .2 K= W
e1/b = .3 I/D
1.9 e1/b = .4 e h ² ( b − e 1 ) e 22
e1/b = .5 W= 1 +
1.7 e1/b = .7 8 8
e1/b = .9 e h ² ( b − e 1 )e 32
I/D = 1 +
1.5 12 12 h
0.0 .2 .4 .6 .8 1.0 1.2
e2/h

© AEROSPATIALE
Let us now calculate σ0 knowing that this stress does not theoretically depend on the shape of the section.
That which is valid for a rectangular section is also valid for all other sections.
We wish to find σ0 corresponding to σR; this stress level is obtained at the extreme fibre of the axis. By
considering pure bending and assuming that the material curve is perfectly symmetrical in tension-compression,
we suppose that the stress is null at the neutral fibre of the axis. Therefore, we can plot a second x-axis
corresponding to the dimension between the neutral fibre and a given fibre on the material curve (see figure
2-4.5.2.1.2-2).
σ
σR
z σ
ZMAX = h/2
z σ(z)
0 ε
εs
0 z
z ZMAX = h/2
Fig. 2-4.5.2.1.2-2 Real stress profile
Thus, the bending moment can be written:

h/2
M = 2b ∫ z . σ( z) . dz
0
z= h x ε
2 ε MAX
Where: h Plane bending: the cross sections remain plane. ε linear with z.
dz = dε
2 . ε MAX
Also, ε can be written:

n n
  σ 
ε = σ + 0.002 .  σ  where: n = 500 .  0.2 
E  σ 0 .2   σR 
We deduce the expressions of ε and εMAX:
n
 
ε = σ + 1 .  σ  = εe + εp
E n  σR 
n
σ σ 
εMAX = MAX + 1  MAX  = εeMAX + εpMAX
E n  σR 
M.v
But: σb = = 6 .M
I b . h²
Moreover:
σb = (k - 1) . σ0 + σMAX
σ0  
⇒ = 2 .  σb − 1
h.W 3 σ MAX  σ MAX 
K= =
I 2

© AEROSPATIALE

2
σ0 2 ( n − 1)  εP    2n + 1   ε e  
= .   .    + 1
   n + 2  .  ε P
MAX MAX
σ MAX 2n + 1  
 ε MAX    MAX  
This expression can also be written:
σ0  n − 1   ε P   n − 1   εP 
= 2.  . MAX
 .  1 −  . MAX

σ MAX  n + 2   ε MAX  
   2 n + 1   ε MAX 

Where:
σMAX maximum allowable stress
n
σ 
εPMAX = 1 .  MAX  maximum allowable plastic yielding
n  σR 
σ MAX
εMAX = + εPMAX maximum allowable distorsion
E
n
σ 
n = 500 .  0.2  (numerical resolution)
 σR 
Specific cases:
σ
c (σMAX = σR) ⇒ εPMAX = 1 ; εMAX = R + 1
n E n
σ 0.2
d (σMAX = σ0.2) ⇒ εPMAX = 0.2%; εMAX = + 0.2%
E
V1-5 5.2.1.3 Symmetrical sections
Example: Calculation of allowable moment
Y
S = 825 mm²
Ix = 384444 mm4
Material: 7075 T76510
e1 = 6 σR = 525 MPa
X σ0.2 = 460 MPa
M h = 54.4 σ0.2c = 460 MPa
E = 71700 MPa
Ec = 73800 MPa
n = 23
e2 = 6 nc = 27
50

© AEROSPATIALE
Step c: Determine shape factor k
 e1 6 = 0.12; e 2 = 6 = 0.110  ⇒ K = 1.23

 = 
 b 50 h 54.4 
Step d: Calculate maximum elongation

The beam section submitted to a bending moment, M, has a neutral fibre at 27,2 mm from the two extreme
fibres (symmetry).
The maximum elongation on tensioned side is:
σR
εMAX = + 1 = 525 + 1 = 0.051
E n 71700 23
We verify that εMAX < e% 0.051 < 0.07
If εMAX > e%, the limit is e%.
Plane bending hypothesis: the cross sections remain plane, ε is linear with z. The elongation at the extreme
fibre of the tensioned portion is 0.051. Therefore, on the extreme fibre of the compressed portion it will be:
εMIN = - εMAX = - 0.051
Step e: Calculate the maximum allowable stresses σMIN on compressed side and σMAX on tensioned side
Let us calculate σMIN knowing that εMIN = - 0.051.
nc
σ σ 
εMIN = MIN + 0.002 MIN 
Ec  σ 0.2 c 
27
σ MIN σ 
- 0.051 = + 0.002 MIN  ⇒ σMIN = - 515 MPa
73800  460 
Let us check the allowable stresses of the local buckling phenomena:
η x 73800 x π ²  6  2
n Local buckling of free flange (or fixed flange): σflt = 0.43 .   ≈ 1687 x η MPa
12 . (1 − 0.33²)  25 
1 − υe2 E s Es  E  1
η= . ; υ= . υ e +  1 − s  . υ P ; Es =
1 − υ² E c Ec  E c   σ 
nc
− 1 + 0.002 .  flt 
Ec σ C 0.2  σ C 0.2 
σflt = - 500 MPa

η x 73800 x π ²  6  2
o Local buckling of web: σfla = 4 x .  ≈ 4187 x η MPa
12 . (1 − 0.33²)  48.4 
After plasticity correction:
σfla = - 515 MPa

© AEROSPATIALE
We can therefore deduce the allowable stress in compression:

σMIN = - 500 MPa
The Cozzone method considers, for a symmetrical section, that σMAX = - σMIN, therefore:
σMAX = 500 MPa
Step f: Calculate the bending modulus

σb = σMAX + σ0 . (k - 1)
The expression giving σ0 is:
σ0  n − 1   ε P   n − 1   εP 
= 2. . MAX
 .  1 −  . MAX

σ MAX  n + 2   ε MAX  
   2 n + 1   ε MAX 

n
σ  σ
We take: (σMAX = 500 MPa) ⇒ εPMAX = 1  MAX  ; εMAX = MAX + εPMAX
n  σR  E
σ0 = 405 MPa ⇒ σb = 597 MPa
Step g: Calculate the allowable moment of the section
Ix
Madm = σb x = 597 x 384444 = 8438 x 103 mm.N
Z MAX 27.2
V1-5 5.2.1.4 Asymmetrical sections
Example: Calculating allowable moment.
Y
50 S = 1000 mm²
IXG = Ixx = 557500 mm4
e2 = 8 IyG = Iyy = 95900 mm4
e1 = 8 σR = 525 MPa
X' X σ0.2 = 460 MPa
M σ0.2c = 460 MPa
h = 50 E = 71700 MPa
Ec = 73800 MPa
n = 23
38.8
nc = 27
e2 = 8
25

© AEROSPATIALE
Step c: Divide the section above at the main axis X’X into 2 symmetrical sections
Section a Section b
Y
Y
e1 = 8
X X
h = 77.6 h = 54.4
38.8 e1 = 8
e2 = 8
e2 = 8
50
25
S = 892 mm² S = 1107 mm²

Ix = 642400 mm4 Ix = 472600 mm4
Step c: Determine the shape factors k

e e 
n Section a:  1 = 8 = 0.32; 2 = 8 = 0.103 ⇒ K = 1.3
 b 25 h 77. 6 
e e 
o Section b:  1 = 8 = 0.16; 2 = 8 = 0.147  ⇒ K = 1.24
 b 50 h 54 , 4 
Step d: Calculate the maximum elongations.

The beam section submitted to a bending moment, M, has a neutral fibre at 38,8 mm from the tensioned
extreme fibre and at 27,2 mm from the compressed extreme fibre.
The highest stress is obtained on the tensioned side, that is on the lower side of the beam.
The maximum elongation on tensioned side is:
σR
εMAX = + 1 = 525 + 1 = 0.051
E n 71700 23
We verify that εMAX < e% 0.051 < 0.07
If εMAX > e%, the limit is e%.
Plane bending hypothesis: the cross sections remain plane, ε is linear with z. The elongation at the extreme
fibre of the tensioned portion is 0.051. Therefore, on the extreme fibre of the compressed portion it will be:
εMIN = - εMAX x 27.2 = - 0.036

38.8

© AEROSPATIALE
Step e: Calculate the maximum allowable stress σMIN on compressed side

Let us calculate σMIN knowing that εMIN = - 0.036.
nc
σ σ 
εMIN = MIN + 0.002 MIN 
Ec σ
 0.2 c 
27
σ MIN σ 
- 0.036 = + 0.002 MIN  ⇒ σMIN = - 508 MPa
73800  460 
Let us check the allowable stresses of the local buckling phenomena:
η x 73800 x π ²  8  2
n Local buckling of free flange: σflt = 0.43 .   ≈ 3000 x η MPa
12 . (1 − 0.33²)  25 
1 − υe2 E s Es  E  1
η= . ; υ= . υe +  1 − s  . υP ; Es =
1 − υ² E c Ec  E c   σ 
nc
− 1 + 0.002 .  flt 
Ec σ C 0.2  σ C 0.2 
σflt = - 514 MPa
η x 73800 x π ²  8  2
o Local buckling of web: σfla = 4 x .   ≈ 5184 x η MPa
12 . (1 − 0.33²)  58 
After plasticity correction:
σflt = - 520 MPa
We can therefore deduce the allowable stress in compression:

σMIN = - 508 MPa
The allowable stress on the tensioned side must be limited to a value σMAX = σR.
Step f: Calculate the bending moduli

σb = σMAX + σ0 . (k - 1)
σ0  n − 1   ε P   n − 1   εP 
= 2. . MAX
 .  1 −  . MAX

σ MAX  n + 2   ε MAX  
   2 n + 1   ε MAX 

n
σ  σ
n Section a: We take: (σMAX = 525 MPa) ⇒ εPMAX = 1  MAX  ; εMAX = MAX + εPMAX
n  σR  E
σ0 = 474 MPa ⇒ σb = 667 MPa

n
σ  σ
o Section b: We take:(σMAX = 508 MPa) ⇒ εPMAX = 1  MAX  ; εMAX = MAX + εPMAX
n  σR  E
σ0 = 433 MPa ⇒ σb = 612 MPa

© AEROSPATIALE
Step g: Calculate the allowable moments for sections a and b
Ix
n Section a: Ma = σb x = 667 x 642400 = 11043 x 103 mm.N
Z MAX 38.8
Ix
o Section b: Mb = σb x = 612 x 472600 = 10634 x 103 mm.N
Z MAX 27.2
Step h: Calculate the allowable moment of the real section
Ma + M b
Madm = = 10839 x 103 mm.N
2

© AEROSPATIALE
V2-4 5.2.2 Deviated bending
There are no accurate interaction equations concerning plastic deviated bending. The equation given below
is very conservative. It is given only as a rough tool to enable the user to avoid more complex analysis methods
especially when the load is not critical.
When the load is critical, we suggest use of an analysis method based on the finite difference technique. We
will not develop this method here.
When an applied moment, M, acts around an axis other than one of the main axes, the allowable bending
moment can be determined as follows:
v
u
θ M
Components of the moment applied along each main axis:

Mu = M . cosθ and My = M . sinθ
Let us determine the allowable moments Madmu and Madmv around each main axis as if the section was
submitted to single bending around these axes.
We can then deduce the ratios of the moments around each axis:
Mu Mv
Ru = and Rv =
M admu M admv
Finally, let us calculate the safety factor:
S.F. = 1 -1
Ru + Rv
This equation can be written in different ways especially by expressing the safety factor as a function of the
components of the moment:
S.F. = 1 -1
 cos θ 
M .  + sin θ 
 M admu M admv 
By writing, S.F. = 0, we can obtain the equation giving us the allowable bending moment:
Madm = 1
 cos θ 
 + sin θ 
 M admu M admv 
Specific case: if components of M along u and v are equal, then:
M admu M admv
Madm = =
cos θ + sin θ cos θ + sin θ

© AEROSPATIALE
Example:
y M
σR = 525 MPa
θ = 45° σ0.2 = 460 MPa
σ0.2c = 460 MPa
b = 20 z E = 71700 MPa
Ec = 73800 MPa
n = 23
nc = 27
b = 20
Step c: Determine the shape factor k

For a square section k = 1.5 .
Step d: Calculate the bending modulus

σb = σMAX + σ0 . (k - 1)
σ0  n − 1   ε P   n − 1   εP 
= 2. . MAX
 .  1 −  . MAX

σ MAX  n + 2   ε MAX  
   2 n + 1   ε MAX 

σ
We take: (σMAX = σR = 525 MPa) ⇒ εPMAX = 1 ; εMAX = R + εPMAX
n E
σ0 = 474 MPa ⇒ σb = 762 MPa
Step e: Calculate the allowable moments acting around the main axes as if the section was submitted to single
bending around these axes
Iy Iz
Madmy = σb x and Madmz = σb x
y max z max
Square section: yMax = zMAX and Iy = Iz

Where: yMAX = zMAX = 10 mm
4
Iy = Iz = 20 = 13333 mm4
12
Madmy = Madmz = 1016.103 mm.N
Step f: Calculate the allowable moments really acting around the main axes
S.F. = 1 -1=0
Ry + Rz

© AEROSPATIALE
⇒ Ry + Rz = 1
But, Ry = Rz ⇒ Ry = Rz = 0.5
⇒ My = Mz = Madmy x Ry = Madmz x Rz = 1016 x 103 x 0.5 = 508 x 103 mm.N
Step g: Calculate the allowable moment for the section
Madm = M 2y + M 2z = 718 x 103 mm.N

© AEROSPATIALE
V2-4 6 CALCULATING STRESSES IN CURVED BEAMS

V2-4 6.1 SYMMETRICAL SECTIONS
V2-4 6.1.1 Introduction

Up until now, we have considered only beams with a straight longitudinal axis before deformation.
Beams where the longitudinal axes are not straight are called curved beams.
The formulas established previously do not strictly apply to these beams.
Below, we shall limit ourselves to giving the plane bending analysis methods for symmetrical curved beams
in the elastic range.
The method described in paragraph V2-4 6.1.2 disregards all distortion phenomena in the flanges. These
will be studied in paragraph V2-4 6.1.3.
As the method presented in paragraph V2-4 6.1.2 is simpler, the user must determine the level of accuracy
required for his application then choose one of the two methods.
V2-4 6.1.2 Plane elastic bending of solid sections

When a curved beam is submitted to elastic plane bending, the distributions of the stresses and the
circumferential strains are no longer linear as in straight beams but hyperbolic.
These stresses increase faster on the side nearer to the centre of curvature.
See figure V2-4 6.1.2.1 (where we have used arbitrary signs).
The stress is proportional to the strain but, due to the difference in the lengths of the internal and external
fibres of the beam, the stress and elongation are not proportional to the distance to the neutral fibre. For sections
submitted to bending only, the neutral fibre does not correspond with the main axis of inertia.
Main axis Z
Neutral fibre
y0 M
Figure V2-4 6.1.2.1

As the stress-strain distribution of a curved beam submitted to bending is not linear, formula
M.y
σx = is no longer applicable.
I
The circumferential bending stress obtained at any point of the section is of following form:
M  y 
σx = . 
S . ( rG − rN )  rN + y 
Issue 0 Calculating stresses in curved beams page V2-4•6-1

© AEROSPATIALE
where:
M = Moment applied (positive when it decreases the radius of curvature, negative in the
opposite case).
S = Area of the section.
rG = Radius of curvature (distance from the centre of curvature to the CG of the section).
rN = Radius of the neutral fibres.
y = Distance of a given point from the neutral fibre (positive in direction opposite centre of
curvature).
The analytical expressions of variable rN are given on page V2-4 3-3.
For I- and T-sections and other shapes, refer to the method described in paragraph V2-4 6.1.3 before
calculating the circumferential bending stresses.
V2-4 6.1.3 Plane elastic bending of thin profile sections

The general analysis methods described in paragraph V2-4 6.1.2 may give erroneous results, especially for I- and
T-sections and other standard thin web sections due to the distortions which appear during bending.
When a symmetrical curved beam is submitted to bending, the free edges of the flanges move radially and cause
a reduction in the circumferential stress. This redistribution of the circumferential stresses is shown on figure V2-
4 6.1.3.1 (a).
Note that this reduction in stresses reduces the stiffness of curved beams when compared with straight
beams.
bs
M y1
ea
Ne u σr
tral
axis
Radial stresses in web
(a) (b)
Figure V2-4 6.1.3.1 Distribution of stresses in curved beams submitted to bending
To compensate for the nonuniformity of the stresses in the flanges, we determine a flange load-carrying
width called b:
b = ξ . bs
Where:
bs = Real flange width.
ξ = Load-carrying width flange factor obtained from figure V2-4 6.1.3.2.
The method described in paragraph V2-4 6.1.2 can then be followed using b.

© AEROSPATIALE
The deflection of the flanges generates transverse bending stresses (see figure V2-4 6.1.3.1 (a)). The transverse
stress is maximum at the junction between the flange and the web. This is called σ1:
σ1 = β . σh
Where:
β = Transverse bending stress factor obtained from figure V2-4 6.1.3.3.
σh = Circumferential bending stress at CG of the flange of the load-carrying section.
1.8
β
1.6
1.4
ba
hs
1.2
1.0 bs
rs
= Distance from centre of
curvature to flange CG
0.8
0.6
0.4 ξ
0.0 0.5 1.0 1.5 2.0 2.5 3.0
( b s / 2)²
rs . h s
Figure V2-4 6.1.3.2 Transverse bending stress β and flange efficiency factors ξ

© AEROSPATIALE
In addition to the transverse stresses studied previously, radial stresses appear in the web (see figure V2-4
6.1.3.1 (b)). These are introduced by the curvature of the beam.
The maximum radial stress is located at the junction between the web and the flanges. We call it σr:
σ h . Ss
σr =
rs . b s
Where: ba = Web thickness.

Ss = Load-carrying flange area.
Ri = Distances from centre of curvature to CG of upper and lower flanges, called ri and rs
respectively.
30
Example:
A 2,5
M = - 50.104 mm.N
2,5
40
28,75 2,5
Centre of curvature
σR = 525 MPa
Determine:
• The maximum circumferential bending stress.
• The transverse bending stress in the flange.
• The maximum radial stress induced in the web.
• The combined stress at the intersection of the web and the flange.
• The safety factors.
Step c: Determine the load-carrying section
The properties of the load-carrying section can be determined, for an I-section, as follows:
( b s / 2)² (30 / 2 )²
n Fixed flange: = =3
rs . h s ( 28,75 + 2,5 / 2 ) x 2,5
⇒ figure V2-4 6.1.3.2: ξ = 0,42

⇒ b = ξ . bs = 0,42 x 30 = 12,6 mm
( b t / 2)² (30 / 2)²
o Free flange: = = 1,33
rt . ht ( 28,75 + 40 − 2,5 / 2 ) x 2,5
⇒ figure V2-4 6.1.3.2: ξ = 0,62

⇒ b = ξ . bs = 0,62 x 30 = 18,6 mm

© AEROSPATIALE
p Load-carrying section:
18,6
S = 165,5 mm²
I = 35920 mm4
G 2,5
h1 = 2,5 mm
h2 = 35 mm 2,5
h3 = 2,5 mm 40
r1 = 28,75 mm Neutral fibre
r2 = 31,25 mm
r3 = 66,25 mm rG
r4 = 68,75 mm
b1 = 12,6 mm rN
b2 = 2,5 mm 28,75 2,5
b3 = 18,6 mm 12,6
Centre of curvature
Step d: Determine rN
From page V2-4 3-3, expression for rN is:
h1 . b1 + h 2 . b 2 + h 3 . b 3
rN =
r r r
b1 . ln 2 + b 2 . ln 3 + b 3 . 4
r1 r2 r3
rN = 45,74 mm
Step e: Calculate the circumferential bending stress

The expression for this stress at any point located at a distance y from the main axis is:
M  y 
σx = . 
S . ( rG − rN )  rN + y 
The maximum stress is located at r = 28,75 mm Ö y = r - rN = - 16,99 mm.

− 50.10 4  − 16,99 
σxMAX = .  = 379 MPa (tension).
165,5 . (50.45 − 45.74)  45.74 + 16,99 
Step f: Calculate the transverse bending stress factor of the real section
( b s / 2)² (30 / 2 )²
n Fixed flange: = =3
rs . h s ( 28,75 + 2,5 / 2 ) x 2,5
⇒ figure V2-4 6.1.3.2: β = 1.76

Step g: Calculate the circumferential bending stress at flange CG (r = 30 mm Ö y = - 15,74)
− 50.10 4  − 15,74 
σh = .  = 336 MPa (tension).
165,5 . (50.45 − 45.74)  45.74 + 15,74 

© AEROSPATIALE
Step h: Calculate the max. transverse bending stress

σ1 = β . σh = 1.76 x 336 = 591 MPa
Step i: Calculate the max. radial stress
Here, this stress is located at the junction between the flange and the web.
Data concerning the area of the load-carrying flange:
Ss = 12,6 x 2,5 = 31,5 mm²
The expression giving the radial stress is as follows:
σ h . Ss 336 x 31,5
σr = = = 141 MPa
rs . b a ( 28,75 + 2,5 / 2 ) x 2,5
Step j: Calculate the bending stress on the upper face of the flange
This stress is located at r = 31,25 mm Ö y = r - rN = - 14,49 mm
σx = 297 MPa
Step k: Summary of stresses applied at web-flange junction

σx = 141 MPa
x
y
σx = 141 MPa σz = 141 MPa
Step l: Determine the safety factors

n Web-flange junction:
2 . σR
S.F. = -1 Ö S.F. = 0.32
(σ x − σ y )² + (σ y − σ z )² + (σ z − σ x )²

© AEROSPATIALE
o Extreme fibre (centre of flange lower face)

σR
S.F. = -1 Ö S.F. = 0.38
σx ² + σy ² + σx . σz

© AEROSPATIALE
V2-4 6.2 ASSYMMETRICAL THIN PROFILE SECTIONS
V2-4 6.2.1 Introduction

Asymmetrical curved beams such as the one shown on figure V2-4 6.2.1.1 are submitted, in addition to the
distortion of the flanges, to deflection of the web.
The method developed below is based on: Engineering Science Data Unit, “Flange Efficiency Factors for
Curved Beams Under Bending in the Plane of Curvature”, Chapter 71004.
σ1 σ1
σxt σxt
σ1 σ1
σ1 σ1
σ1 σxs σ1
σxs
V2-4 6.2.1.1 C-section submitted to bending
V2-4 6.2.2 Calculating load-carrying section

All references below are taken from ESDU 71004.
Step c: Evaluate web flexibility coefficients
For each flange, determine the flexibility coefficient of the web, φ, by means of equation 4 or 5 on
page 6. For certain standard sections, it will be easier to use the curves on figures 5 or 6 on pages 11 and
12.
Step d: Evaluate the flange efficiency factors

For each flange, determine the flange efficiency factor, η (ξ in the manual), by means of equation 1 or 2
on page 6. For certain standard sections, it will be easier to use the curves on figures 1 to 4 on pages 9
and 10.
Step e: Determine the transverse stress ratios
For each flange, determine the transverse stress ratios σ1/σxt and σ1/σxs by means of equation 1 or
2 on page 6. For certain standard sections, it will be easier to use the curves of figures 1 to 4 on pages 9
and 10.
Step f: Determine the load-carrying section
By means of the flange efficiency factors, determine the load-carrying widths of the flanges and lips.
Then, evaluate all the geometrical properties of this section such as: area, inertia, etc.

© AEROSPATIALE
V2-4 6.2.3 Example

Determine the circumferential and transverse bending stresses of the section shown below and submitted to
a bending moment of 70000 mm.N.
8,25
S = 30,25 mm²
2,75
Ix = 6650 mm4
40,5 70000 mm.N
0,5
449,75
Centre of curvature
Step c: Evaluate the web flexibility coefficients

ha (( 40,5 − 0,5) / 2 )
n Fixed flange: = = 0.044
rS 449,75 + 0.25
ha (( 40,5 − 0,5) / 2 )
= = 40
hS 0 ,5
ha 0 ,5
= =1
h S 0 ,5
From the curve on figure 5 on page 11 (ESDU), we obtain:
φs = 0.53 x (1)3 x 40 = 3.352
ha (( 40,5 − 0,5) / 2 )
o Free flange: = = 0.0408
rt 449,75 + 40.25
ha (( 40,5 − 0,5) / 2 )
= = 40
ht 0 ,5
ha 0 ,5
= =1
ht 0 ,5
From the curve on figure 5 on page 11 (ESDU), we obtain:
φt = 0.52 x (1)3 x 40 = 3.289

© AEROSPATIALE
Step d: Determine the flange efficiency factors

2 2
 b   0 ,5 
 bs − a   8, 25 − 
2  2 
n Fixed flange:  =  = 0,2844
rs . h s ( 449,75 + 0,5 / 2) x 0,5
A bt ( 2,75 − 0.25) . 0,5

= = 0.3125
 b   0 ,5 
 b s − a  . h s  8, 25 −  . 0 ,5
 2   2 
From the curves on figures 3 and 4 on page 10 (ESDU), we obtain:

σ1
ξs = 0,540 and = 0.560
σ xs
2 2
 b   0, 5 
 bt − a   8, 25 − 
 2   2 
o Free flange: = = 0,2612
rt . h t ( 449,75 + 0,5 / 2 ) x 0,5
A bt ( 2,75 − 0.25) . 0,5

= = 0.3125
 ba   0 ,5 
 bt − .h  8, 25 −  . 0 ,5
 2  t  2 
From the curves on figures 3 and 4 on page 10 (ESDU), we obtain:

σ1
ξt = 0,560 and = 0.550
σ xt
Step e: Calculate the properties of the load-carrying section

The area of the load-carrying section is calculated as follows:
  b     0,5  
n Fixed flange: As = ξs .  A bt +  b s − a  . h s  = 0,540 x  2,5 x 0,5 +  8, 25 −  x 0 ,5  = 2,835 mm²
  2     2  
o Lip on fixed flange side: Abt = 0,540 x (2,5 x 0,5) = 0,675 mm²
  b    
 = 0,560 x  2,5 x 0,5 +  8, 25 −
0,5 
p Free flange: At = ξt .  A bt +  b t − a  . h t  x 0,5  = 2,94 mm²
  2     2  
q Lip on free flange side: Abt = 0,560 x (2,5 x 0,5) = 0,7 mm²
The load-carrying section can then be represented (see next page). Note that only the widths of the fixed
flange, the free flange and the lips are modified.

© AEROSPATIALE
4,73
1,65
40,5 G
20,33
0,5
449,75 1,6
4,57
Centre of curvature
Step f: Determine rN
From page V2-4 3-3, expression for rN is:
rN =
∑h . b i i
r
∑ b . ln r
i
ei
ii
rN = 469,68 mm
Step g: Calculate the circumferential bending stresses

The expression for this stress at any point located at a distance y from the main axis is:
M  y 
σx = . 
S . ( rG − rN )  rN + y 
The max. stress is located at r = 449,75 mm Ö y = r - rN = - 19,93 mm.

− 7.10 4  − 19,93 
σxMAX = .  = 304 MPa (tension).
25.53 . ( 470.08 − 469,68)  469,68 − 19,93 
The min. stress is located at r = 490,25 mm Ö y = r - rN = - 20,57 mm.

− 7.10 4  20,57 
σxMIN = .  = 288 MPa (compression).
25.53 . ( 470.08 − 469,68)  469,68 + 20,57 
Step h: Calculate the transverse bending stresses in fixed flange and free flange
σ1
n Fixed flange: = 0.560 Ö σ1 = ± 0.560 x σ xs
σ xs

© AEROSPATIALE
Where:
− 7.10 4  − 19,68 
σxs = .   = 300 MPa
25.53 . ( 470.08 − 469,68)  469 , 68 − 19 , 68 
Ö σ1 = ± 168 MPa
σ1
o Free flange: = 0.550 Ö σ1 = ± 0.550 x σ xt
σ xt
Where:
− 7.10 4  20,32 
σxt = .   = - 284 MPa
25.53 . ( 470.08 − 469,68)  469 , 68 + 20 , 32 
Ö σ1 = ± 156 MPa

© AEROSPATIALE
MCS V2-5 • COLUMNS
CONTENTS
issue date change
V2-5 COLUMNS 0 6/1999 Creation

V2-5 2 ELASTIC ANALYSIS 0 6/1999 Creation
V2-5 3 EFFECTIVE LENGTH 0 6/1999 Creation
V2-5 4 PLASTIC ANALYSIS 0 6/1999 Creation
V2-5 5 EXAMPLE 0 6/1999 Creation
V2-5 1 INTRODUCTION 1–1

V2-5 1.1 CONDITIONS FOR USE 1–2
V2-5 1.2 NOTATIONS AND CONVENTIONS 1–2
V2-5 2 ELASTIC ANALYSIS 2–1

V2-5 3 EFFECTIVE LENGTH 3–1
V2-5 4 PLASTIC ANALYSIS 4–1


 AEROSPATIALE
Issue 0 Contents page V2-5•ii

 AEROSPATIALE
V2-5 1 INTRODUCTION
A beam is called a "column" when it is subjected to the action of two equal forces acting in opposite directions,
directed towards the inside of the material, applied, along its axis, to the CGs of the end sections.
The critical stress is the stress corresponding to the buckling load:
Fcr
σ cr =
S
The column remains stable up to this critical load, a branch characterised by a bending strain mode corresponds
to this critical load.
The strain and type of failure of the columns depend mainly on:
• the length of the beam.
• the shape characteristics of the straight sections of the column.
In this chapter, we will consider only straight columns with constant sections along their complete lengths and
without a tendency for local buckling: no thin web effect (solid sections).
Also, we shall take the Navier-Bernoulli hypothesis: flat cross sections before deformation, remain flat after
deformation.
Issue 0 Introduction page V2-5•1–1

 AEROSPATIALE
V2-5 1.1 CONDITIONS FOR USE

The equations which follow are based on the flat bending theory. Also, they are applicable to solid section
columns irrespective of their slenderness ratio.
The clamping factor formula is applicable to all stability problems of a structural element comparable to a beam.
V2-5 1.2 NOTATIONS AND CONVENTIONS

The compression stresses and compression loads are taken to be positive.
Distributed linear compression load

Normal load (+ compression)
(+ compression)
Pcr = critical buckling load expressed in daN.

P = compression load expressed in daN.
q = distributed linear compression load expressed in daN/mm.
σcr = critical buckling stress expressed in daN/mm2.
Ec = modulus of elasticity in compression expressed in daN/mm2.
I = moment of inertia expressed in mm4.
S = area of the section expressed in mm2.
K = clamping factor (dimensionless).
L = column effective buckling length expressed in mm.
A = real column length expressed in mm.

 AEROSPATIALE
V2-5 2 ELASTIC ANALYSIS

Let us consider a beam with hinged ends.
P
y
y
x
P
x
Let us calculate Pcr:
d2y Mz − P⋅ y
2
= =
dx EI Gz EI Gz
π 2 ⋅ Ec ⋅ I
Pcr =
L2
The critical stress can therefore be expressed by the Euler equation:
π 2 ⋅ Ec ⋅ I
σ cr =
S ⋅ L2
π 2 ⋅ Ec
This stress can also be written: σ cr =
L2
I S
But, the I S ratio represents the square of the radius of gyration of the section:
I
ρ=
S
Issue 0 Elastic Analysis page V2-5•2–1

 AEROSPATIALE
We can therefore deduce a new expression from the Euler formula:
π 2 ⋅ Ec
σ cr =
(L ρ )2
The L ρ coefficient characterises the general buckling tendency. This is called the slenderness ratio and is
represented by the letter λ:
L
λ=
ρ
Therefore the Euler formula can be written:
π 2 ⋅ Ec
σ cr =
λ2
Issue 0 Elastic Analysis page V2-5•2–2

 AEROSPATIALE
V2-5 3 EFFECTIVE LENGTH

To establish the Euler formula, we used a specific case of a beam perfectly hinged at its two ends spaced apart by
length L.
In other end connection conditions, we agree to designate the real length of the studied beam as A.
We will obtain, in each case, the distance L which should exist between two fictive hinge points to reproduce the
same critical buckling load.
We will call L the effective buckling length of the beam under study. The comparison of L with A will be
expressed by the following general relation:
L = KA where K represents a dimensionless factor called the clamping factor specific to each
case studied.
The critical Euler load will therefore always be given by its primitive expression:
π 2 ⋅ Ec ⋅ I π 2 ⋅ Ec ⋅ I
Pcr = equivalent to Pcr =
L2 K 2 ⋅ A2
The critical Euler stress will therefore be expressed by:
π 2 ⋅ Ec ⋅ I π 2 ⋅ Ec ⋅ I
σ cr = equivalent to σ cr =
S ⋅ L2 S ⋅ K 2 ⋅ A2
The values of K for the various types of columns, boundary conditions and loads are given on figure
V2-5 3.1.
Issue 0 Effective Length page V2-5•3–1

 AEROSPATIALE
Issue 0 Effective Length page V2-5•3–2

 AEROSPATIALE
V2-5 4 PLASTIC ANALYSIS

The Euler formula reflects a state of elastic equilibrium: equilibrium between the bending action of Pcr and the
tendency of the beam to elastically straighten itself. We can therefore conclude that the Euler formula is, at
maximum, valid only during the elastic period of the materials to which it is applied.
When this elastic period is exceeded, we say that we have exceeded the proportionality limit, the original
modulus Ec that we considered up until then and which, in reality, is valid only during the elastic period must be
replaced in the Euler formula by the tangent modulus Et.
We then obtain the following Euler - Engesser equations:
π 2 ⋅ Et ⋅ I π 2 ⋅ Et ⋅ I
Pcr = equivalent to Pcr =
L2 K 2 ⋅ A2
The Euler - Engesser critical stress will therefore be expressed by:
π 2 ⋅ Et ⋅ I π 2 ⋅ Et ⋅ I
σ cr = equivalent to σ cr =
S ⋅ L2 S ⋅ K 2 ⋅ A2
For the critical stress calculation, the difficulty lies in determining the value of the tangent modulus
corresponding exactly to a compression stress equal to the critical stress. Indeed, the tangent modulus Et has no
fixed value but varies according to the position of the point on the stress-strain curve.
This leads us therefore to make a correction to the plasticity the technique of which is described below:
We call η, the plasticity correction factor:
Et
η= where Et = tangent modulus.
Ec
Ec = Young's compression modulus.
σ~cr represents the critical stress calculated with Et = Ec ⇒ η = 1.

This gives the following Euler formula:
π 2 ⋅ Ec ⋅ I
σ~cr =
S ⋅ K 2 ⋅ A2
If σ~cr ≥ 0,5 ⋅ σ c 0, 2 then the critical stress after plasticity correction will be:
σ cr = η ⋅ σ~cr
σ cr
This calculation is iterative as η = f (Et ) = g (σ cr ) . It converges when → σ~cr .
η
Issue 0 Plastic Analysis page V2-5•4–1

 AEROSPATIALE
Bearing conditions Type of load K
P
Isolated load 1
A
Bi-hinged column
Distributed load 0.729

P P
P
Isolated load 0.5
A Bi-clamped column,
fixed transversely
P P
P
Bi-clamped column,
A Isolated load 1
not fixed transversely
P
Isolated load 0.7
A Column clamped at one end and hinged

at the other

P P
P
Isolated load 2
A Column clamped at one end and free at

the other

P P
Figure V2-5 3.1: Clamping factors (continued on next page)

 AEROSPATIALE
Bi-hinged column See figure

Isolated load
A on elastic supports below
11
10,5
10
9,5
9
8,5
8
7,5
7
6,5
6
2
π/K)
5,5
(π
5
4,5
4
3,5
3
2,5
2
1,5
1
0,5
0
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60
k.W3/EI
k represents the rigidity of an elastic intermediary support.

This graph is valid for a number of supports tending towards infinity but we can assume that it remains
valid in all cases.

 AEROSPATIALE
P-qL
Isolated load
A Bi-hinged column, qL
qL fixed transversely
+ 1 − 0,47 ⋅
distributed load P
P-qL
Isolated load
A Bi-clamped column, qL
qL
fixed transversely
+ 0,25 − 0,12 ⋅
distributed load P
P-qL
Isolated load
A qL Column clamped at one qL
+ 0,49 − 0,3 ⋅
end and hinged at the other end P
distributed load
P-qL
Isolated load
A Column clamped at one end and qL
qL + 4 − 2,74 ⋅
free at the other end P
distributed load
Figure V2-5 3.1: Clamping factors.

 AEROSPATIALE
V2-5 5 EXAMPLE
Data:
S = 200 mm2. P
I = 1667 mm4.
Material: 2024 T351

Ec = 7380 daN/mm2. 150
σR = 44 daN/mm2.
σc0.2 = 32.5 daN/mm2.
nc = 8.
P
Question: Calculate the critical buckling stress.
Result:
Step 1: Determine K
K = 0.7 (see figure V2-5 3.1).
Step 2: Calculate the critical buckling stress
We use the Euler formula:

~ π2 ⋅ Ec ⋅ I π 2 ⋅ 7380 × 1667
σ cr = 2 2
= 2 2
= 55 daN / mm 2
S⋅ K ⋅ A 200 × 0,7 × 150
Step 3: Apply a plasticity correction
We call σ~cr the critical stress calculated with η = 1,
Where:
Et
η=
Ec
We use the Ramberg and Osgood model:
 σ cr  σ cr 
nc
ε = + 0,002 ⋅  

 Ec  σ c 0,2 

 σ cr
E s =
 ε
1 nc 1 − nc
E = E + E
 t s c

We obtain, after numerical resolution:
σ cr = 27,1 daN / mm 2

 AEROSPATIALE
MCS V2-6 • BEAMS-COLUMNS
CONTENTS
issue date change
V2-6 BEAMS-COLUMNS 0 6/1999 Creation

V2-6 2 CALCULATION OF AMPLIFIED BENDING 0 6/1999 Creation
MOMENT 0 6/1999 Creation
V2-6 3 ALLOWABLE STRESS 0 6/1999 Creation
V2-6 4 EXAMPLE
V2-6 1 INTRODUCTION 1–1

V2-6 1.1 CONDITIONS FOR USE 1–2
V2-6 1.2 NOTATIONS AND CONVENTIONS 1–2
V2-6 2 CALCULATION OF AMPLIFIED BENDING

MOMENT 2–1
V2-6 3 ALLOWABLE STRESS 3–1
Issue 0 Introduction page V2-6•i

Issue 0 Introduction page V2-6•ii

V2-6 1 INTRODUCTION
A beam is called a "beam-column" when it is submitted both to compression and bending loads.
When a structural element similar to a beam is subjected, simultaneously, to a normal load and a bending moment,
we cannot use the superposition theorem to determine the stresses of which it is the seat. Indeed, a part of the
bending moment is due to the action of the normal load on the deflected beam.
Thus, a normal tensile load the direction of which is constant and therefore parallel to the axis of the non-
deflected beam tends to flatten the deflection. Under the same conditions, a compression load tends to amplify it.
In this chapter, we have not considered the case where the normal load is turning (remaining tangent to the
deflected beam). Let us simply invoke on this point the case of the Beck beam (one end clamped, the other end
free, compressed by an external force which remains tangent to the medium fibre). This system has no critical
buckling load but a branch point characterised by a transition to an oscillating dynamic state.
From a practical viewpoint, we apply the superposition theorem to the bent and tensioned beams as this approach
is conservative. However, for beams-columns, we must calculate the amplified bending moments.
The verification of the good resistance of a beam-column will include two steps:
- The study of the column as described in § V2-5.
- The study of the beam-column if the axial load is lower than the critical load calculated in the previous
step.

V2-6 1.1 CONDITIONS FOR USE

See chapter V2.5 "Columns".
V2-6 1.2 NOTATIONS AND CONVENTIONS

The compression loads are taken to be positive.
A positive bending moment compresses the upper fibres of the beam.
A distributed lateral load leads to a positive moment.
A positive lateral load compresses the upper fibres of the beam.
The compression stresses are positive.
Bending moment at end Bending moment at origin

(+ compresses upper fibres) (+ compresses upper fibres)
Normal load
(+ compression)
Distributed lateral load

(+ compresses upper fibres)
P = compression load expressed in daN.

q = distributed lateral load expressed in daN/mm.
F = applied isolated load expressed in daN.
σcr = critical buckling stress expressed in daN/mm2.
Ec = modulus of elasticity in compression expressed in daN/mm2.
I = moment of inertia expressed in mm4.
S = area of the section expressed in mm2.
K = clamping factor (dimensionless).
L = column effective buckling length expressed in mm.
A = real column length expressed in mm.

V2-6 2 CALCULATING AMPLIFIED BENDING MOMENT

Let us consider a beam where the ends are on single supports with compression loads (Nx) and bending loads
(Mz).
y
y
Mz A Mz B
P P
x
A B
x
L
Beam with compression and bending loads
The bending moment at x is:
 Mz B − Mz A 
Mz = Mz A +  ⋅ x − P⋅ y
 L 
Let us differentiate / x:
d 2 Mz d2y d 2 y Mz
= −P ⋅ 2 where: =
dx 2
dx dx 2 EI
d 2 Mz P
Therefore: + ⋅ Mz = 0
dx 2 EI
EtI
Let us write: j=
P
The solution of this equation is:
x x
Mz = C1 ⋅ sin + C 2 ⋅ cos
j j
At x=0: Mz = Mz A C 2 = Mz A
L
Mz B − Mz A ⋅ cos
j
At x = L: Mz = Mz B : C1 =
L
sin
j
We therefore obtain:
  
 Mz B − Mz A ⋅ cos L  
  j 
Mz =    ⋅ sin  x  + Mz ⋅ cos x 
  j A  j
  L     
sin  
 j 
   
Issue 0 Calculating Amplified Bending Moment page V2-6•2–1

Mz is maximum for:
dMz
=0
dx
C  x C  x
⇒ 1 ⋅ cos  − 2 ⋅ sin   = 0
j  j j  j
x C1
That is for tan =
j C2
Mz maximum for:
 L
 Mz B − Mz A ⋅ cos 
j
x = j ⋅ Arc tan 
 L 
 sin 
 j 
By studying other examples, we can see that the bending moment expression is always in following form:
x x
Beam - column bending moment: Mz = C1 ⋅ sin + C 2 ⋅ cos + j 2 ⋅ f (x )
j j
Values of C1 , C2 and f(x) are given on figure V2-6 2.1.
Remark: it is possible to combine several cases (see example below).
M1 M2 L
M 2 − M 1 cos
j x x
Case ① P
⋅ sin + M 1 ⋅ cos
P L j j
sin
L j
✚ ✚ ✚
q
L x x
Case ② j 2 ⋅ q ⋅ tan ⋅ sin + j 2 ⋅ q ⋅ cos − q ⋅ j 2
P P 2j j j
L
= = =
q  L
M1 M2  M2 − M1 ⋅ cos 
Case ⑥ P
 j2 ⋅ q ⋅ tan L +
 2j L
j

x
j
( x
j
)
⋅ sin + M1 + j2 ⋅ q ⋅ cos − q ⋅ j 2
P sin
 
L  j 

Conditions C1 C2 f(x)
①
L
M1 M2 M 2 − M 1 cos
j
L
M1 0
P P sin
L j
②
q
L
j 2 ⋅ q ⋅ tan j2 ⋅ q -q
P P 2j
L
③
q
j ⋅ qL
j ⋅ qL
2 tan
L -q
P P 2
2j
L
x≤a F ④
a<b a b
b j ⋅ F sin
j
x L
0 0
P P sin
L j
x>a F ⑤
a<b a a
b j ⋅ F sin
j a
− j ⋅ F sin 0
P x P L j
tan
L j
q ⑥ L
M1 M2 M2 − M1 ⋅ cos
L j
j2 ⋅ q ⋅ tan + M1 + j2.q -q
P P 2j L
sin
L j
x≥a ⑦
a<b a b
b M ⋅ cos
j
− 0 0
P x M P L
sin
L j
x>a ⑧
a<b a a
b M ⋅ cos
j a
− M ⋅ cos 0
P x M P L j
tan
L j
Figure V2-6 2.1 (continued on next page)

Conditions C1 C2 f(x)
⑨
q
j2 ⋅ q
− q ⋅ (x − 1)
tan
L j2.q
P P L
j
L
⑩
q
j2 ⋅ q
−q ⋅ x
sin
L 0
P P L
j
L
x ≤ L/2 L/2
11
q
2 j2 ⋅ q
2q ⋅ x
x L
⋅ cos
L 0 −
P P L
j 2j
L
x≤a F 12
a<b a b L b b b L b b 
b  ⋅ cos − ⋅ sin   ⋅ cos − ⋅ sin 
j j j j j j j j
j ⋅ F j ⋅ F 
0
P x P  L L L   L L L 
 ⋅ cos − sin   ⋅ cos − sin 
L  j j j   j j j 
x>a 13
a<b F  L a L a b   L  a L a b 
a  cos  sin − cos +    sin  − sin + cos −  
b j  j j j j  j  j j j j 
j ⋅ F j ⋅ F 0
 L L L   L L L 
P x P  cos − sin   cos − sin 
j j j j j j
   
L
14
q  L L   L L 
 tan −   − tan + 
 2 j 2 j L  2 j 2 j L
 L L 
⋅ j ⋅ q ⋅ L + j 2 ⋅ q ⋅ tan
2j  L L 
⋅ j ⋅ q ⋅ L ⋅ tan + j 2 ⋅ q
j
-q
P x P  tan −   tan − 
 j j   j j 
L
15
q  
 
2  1 L L
j.qL j ⋅q − tan -q
 L j j
P P  cos 
 j 
L
16
F
L
j.F − j ⋅ F ⋅ tan 0
P P j
L
Figure V2-6 2.1

V2-6 3 ALLOWABLE STRESS
Two main cases must be considered when choosing allowable stress:
♦ Solid beams
No local buckling phenomena occur before general buckling, the allowable stress on tensioned side is equal to σR.
We will then consider that the allowable stress on compressed side is also equal to σR.
♦ Thin web beams
A local buckling phenomenon can occur before general buckling. This state does not give rise to an immediate
collapse but can considerably decrease the allowable breaking stress.
The allowable stress on compressed side is equal to:
σ adm = MIN{σ local buckling ; σ R }.
The allowable stress on tensioned side is equal to:
σ adm = σ R .
Summary:
Tensioned section: σ adm = σ R .
Compressed section: Solid beams : σ adm = σ R .
Thin web beams : { }

σ adm = MIN σ local buckling ; σ R .
Issue 0 Allowable Stress page V2-6•3–1

V2-6 4 EXAMPLE
Data:
A = 100 mm. F D c
a = 30 mm. M1 M2
c = 10 mm.
S = 100 mm2. c
I = 833 mm4. A C B
P a D P
F = 70 daN.
P = 1160 daN. A DD
M1 = 1200 mm.daN.
M2 = 4000 mm.daN.

Ec = 7380 daN/mm2. E = 7170 daN/mm2.
nc = 27. n = 23.
σc0,2 = 46 daN/mm2. σ0,2 = 46 daN/mm2.
σR = 52.5 daN/mm2. e% = 7.
Question: Calculate the maximum amplified bending moment and the Reserve Factor.
Result:
Step 1: Calculate the allowable compression stress
There is no local buckling phenomenon as the beam is solid in section therefore σR represents
the allowable compression stress.
σ adm = σ R = 52.5 daN/mm2.
Step 2: Column study (see §V2-5)
We check that the beam does not buckle under the action of P alone.
• Determine K
K = 1 (see figure V2-5 3.1).
• Calculate the critical buckling stress
We use the Euler formula:
2
~ = π ⋅ E c ⋅ I = π ⋅ 7380 × 833 = 60,7daN / mm 2
2
σ cr 2 2
S⋅ K ⋅ A 100 ×12 ×100 2
• Apply a plasticity correction
We call σ~cr the critical stress calculated with η = 1,
Where:
Et
η=
Ec

We use the Ramberg and Osgood model:
  
nc
ε = σ cr + 0,002 ⋅  σ cr 
 Ec σ 
 c 0, 2 

 σ cr
E s =
 ε
1 n 1 − nc
 = c +
 Et Es Ec

We obtain after numerical resolution:
σ cr = 41,1 daN / mm 2
But, the stress applied due to the axial load is:
P
σ= = 11,6 daN / mm 2
S
As σ < σ cr , we can deduce that the column does not buckle.
Step 3: Determine the maximum bending moment expressions
EtI 7380 × 833

j= = = 72,8mm as Et = Ec.
P 1160
• For x ≤ a:
By using a combination of cases 1 and 4 (see figure V2-6 2.1), we obtain:
A b
M 2 − M 1 ⋅ cos j ⋅ F ⋅ sin
j j
C1 = + = 8101 mm.daN.
A A
sin sin
j j
C2 = M1 = 1200 mm.daN.
f(x).j2 = 0.
x x
M AC = 8101⋅ sin + 1200 ⋅ cos
j j
• For x > a:
By using a combination of cases 1 and 5 (see figure V2-6 2.1), we obtain:

A A
M 2 − M 1 ⋅ cos j ⋅ F ⋅ sin
j j
C1 = − = 3431 mm.daN.
A A
sin tan
j j
A
C2 = M 1 + j ⋅ F ⋅ sin = 3241 mm.daN.
j

f(x).j2 = 0.
x x
M CB = 3431 ⋅ sin + 3241 ⋅ cos
j j
Step 4: Determine the maximum bending moment
• At point A:
M A = 8101 ⋅ sin 0 + 1200 ⋅ cos 0 = 1200 mm.daN.
• At point C:
30 30
M C = 8101⋅ sin + 1200 ⋅ cos = 4344 mm.daN.
72,8 72,8
• At point B:
100 100
M B = 3431 ⋅ sin + 3241 ⋅ cos = 4000 mm.daN.
72,8 72,8
• Between A and C:
dM AC 8101 x 1200 x
= ⋅ sin − ⋅ cos =0
dx 72,8 72,8 72,8 72,8
The moment is maximum at abscissa point:
8101
x = 72,8 ⋅ Arc tan = 103.7 mm.
1200
But: 103.7 > xC.

Therefore MAC is maximum at xc.
MAC max = 4344 mm.daN.
• Between C and B:
dM CB 3431 x 3241 x
= ⋅ sin − ⋅ cos =0
dx 72,8 72,8 72,8 72,8
The moment is maximum at abscissa point:
3431
x = 72,8 ⋅ Arc tan = 59.3 mm.
3241
59,3 59,3
M CB max = 3431 ⋅ sin + 3241 ⋅ cos = 4720 mm.daN.
72,8 72,8
Therefore, the maximum moment on beam AB is:
MAB max = MAX ( MAC max ; MCB max ) = 4720 mm.daN

Step 5: Calculate the total stress and the Reserve Factor
• Stress due to the axial load (see step 1):
P
σc = = 11,6 daN / mm 2
S
• Stress due to the bending moment:
M c
σ f =± ⋅ = ± 28,3 daN / mm 2
I 2
• Maximum compression stress:
σ MAX = σ c + σ f = 39.9 daN/mm2.
• Reserve Factor:
52,5
R.F . = = 1,32
39,9


Stress Manual Eads PDF

Uploaded by

Copyright:

Available Formats

Stress Manual Eads PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Stress Manual Eads PDF

Uploaded by

Copyright:

Available Formats

Technical Manual

MTS 004 Iss. C

Static stress manual,

Scope All programmes, static justification of metallic structures.

EDP tool Not applicable.

Contents V1 - 1 Stiffened panels

Document Dept code: BTE/CC/CM Validation Name: JF. IMBERT

Reference documents C BE 019: Drawing up of the Structural Justification Dossier

Documents to be consulted See bibliography at the beginning of each chapter.

Abbreviations See Lexique Aerospatiale Airbus/ATR

Issue Date Pages modified Justification of the changes made

© AEROSPATIALE MATRA AIRBUS - 1999 MTS 004 Iss. C 3Ann. page

NOT FOR DISTRIBUTION

Dept. code Function Name / First name Signature

Key words Calcul

Dept. code Function Name / First name (if necessary)

Distribution list managed in real time by BIO/D (Didocost application)

© AEROSPATIALE MATRA AIRBUS - 1999 MTS 004 Iss. C page IG1

V1-1 STIFFENED PANELS 2 8/1999 Creation V1-1 8

*: Page number between brackets

Issue 2 Contents page •i

4.6- Calculating margin at UL using Johnson formula (V1-1y4/15)

page •ii Contents Issue 2

6.9- Stresses in frames (V1-1y6/14)

Issue 2 Contents page •iii

A: Transverse Member pitch p: Fastener pitch

Page V1-1/1 revision 1

σR: Failure tension stress

Page V1-1/2 revision 1

4- Pablo Rodriguez: "Prédimensionnement d'une partie basse de fuselage en GLARE" (Pre-

8- L. Euler: "Methodus inveniendi lineas curvas maximi minimive proprietate gaudentes",

Page V1-1/3 revision 1

V1-1 •1-GENERAL INFORMATION

General information - Page V1-1•1/1 revision 1

General information - Page V1-1•1/2 revision 1

V1-1 •1.2-BEHAVIOUR, FAILURE MODES

These panels are capable of carrying/transferring two types of loads:

- loads in the skin plane (Nx, Φy, Ty, My),

Stresses resulting from these forces are:

- Normal (σx) and tangential (τxz ; τzx) in the stiffener,

• This property of dependence is further emphasised in the case of compression. This

General information - Page V1-1•1/3 revision 1

Three types of instability are differentiated:

General information - Page V1-1•1/4 revision 1

V1-1 •1.3-CALCULATION METHODOLOGY

- The static failure load, i.e. the load allowable at UL.

- Select any stress value in the skin at the stiffener.

- Calculate the "load-carrying section" corresponding to it.

General information - Page V1-1•1/5 revision 1

V1-1 •1.4-FIELD OF APPLICATION

- The studied stiffener does not edge an opening.

- No stiffener run-outs in the studied area.

- Each rectangular skin bay may be considered as having a constant thickness.

- Stiffeners may be considered as having constant cross sections between two

- The skin curvature is in the plane normal to the stiffeners.

- Other effects induced by the presence of an opening in the skin.

- Secondary bending at run-outs.

- Calculation of the fasteners binding the panels together.

- Sizing of the fasteners binding the stiffeners to the skin.