Dietmar Gross · Wolfgang Ehlers

Peter Wriggers · Jörg Schröder

Ralf Müller

Mechanics of
Materials –
Formulas and
Problems
Engineering Mechanics 2

123
Mechanics of Materials – Formulas
and Problems
Dietmar Gross Wolfgang Ehlers
•

Peter Wriggers Jörg Schröder

•

Ralf Müller

Mechanics of
Materials – Formulas
and Problems
Engineering Mechanics 2

123
Dietmar Gross Jörg Schröder
Division of Solid Mechanics Institute of Mechanics
TU Darmstadt Universität Duisburg-Essen
Darmstadt Essen
Germany Germany

Wolfgang Ehlers Ralf Müller

Institute of Applied Mechanics Institute of Applied Mechanics
Universität Stuttgart TU Kaiserslautern
Stuttgart Kaiserslautern
Germany Germany

Peter Wriggers
Institute of Continuum Mechanics
Leibniz Universität Hannover
Hannover
Germany

ISBN 978-3-662-53879-1 ISBN 978-3-662-53880-7 (eBook)

DOI 10.1007/978-3-662-53880-7
Library of Congress Control Number: 2016956827

© Springer-Verlag GmbH Germany 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation,
reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms
or in any other physical way, and transmission or information storage and retrieval,
electronic adaptation, computer software, or by similar or dissimilar methodology
now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks,
etc. in this publication does not imply, even in the absence of a specific statement,
that such names are exempt from the relevant protective laws and regulations and
therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and
information in this book are believed to be true and accurate at the date of
publication. Neither the publisher nor the authors or the editors give a warranty,
express or implied, with respect to the material contained herein or for any errors or
omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer-Verlag GmbH Germany
The registered company address is: Heidelberger Platz 3, 14197 Berlin, Germany
Preface
This collection of problems results from the demand of students for sup-
plementary problems and support in the preparation for examinations.
With the present collection ’Engineering Mechanics 2 - Formulas and
Problems, Mechanics of Materials’ we provide more additional exercise
material.
The subject ’Mechanics of Materials’ is commonly taught in the se-
cond course of Engineering Mechanics classes at universities. The pro-
blems analyzed within these courses use equilibrium conditions and ki-
nematic relations in conjunction with constitutive relations. As we want
concentrate more on basic concepts and solution procedures the focus
lies on linear elastic material behavior and the small strain regime. Ho-
wever, this covers a wide range of elasto-static problems with relevancy
in engineering applications. Special attention is given to structural ele-
ments like bars, beams and shafts as well as plane stress and strain
situations.
Following the warning in the first collection, we would like to make
the reader aware that pure reading and trying to comprehend the pre-
sented solutions will not provide a deeper understanding of mechanics.
Neither does it improve the problem solving skills. Using this collec-
tion wisely, one has to try to solve the problems independently. The
proposed solution should only be considered when experiencing major
problems in solving an exercise.
Obviously this collection cannot substitute a full-scale textbook. If
not familiar with the formulae, explanations, or technical terms the rea-
der has to consider his or her course material or additional textbooks
on mechanics of materials. An incomplete list is provided on page IX.

Darmstadt, Hannover, Stuttgart, Essen and D. Gross

Kaiserslautern, Summer 2016 P. Wriggers
W. Ehlers
J. Schröder
R. Müller
Table of Contents

Literature, Notation .................................................. IX

1 Stress, Stain, Hooke’s Law.......................................... 1

2 Tension and Compression in Bars ................................. 29

3 Bending of Beams ..................................................... 57

4 Torsion ................................................................... 111

5 Energy Methods ....................................................... 143

6 Buckling of Bars ....................................................... 181

7 Hydrostatics ............................................................ 195

Literature
Textbooks
Gross, D., Hauger, W., Schröder, J., Wall, W., Bonet, J., Engineering
Mechanics 2: Mechanics of Materials, 1st edition, Springer 2011
Gross, D., Hauger, W., Wriggers, P., Technische Mechanik, vol 4:
Hydromechanics, Elements of Advanced Mechanics, Numerical
Methods (in German), 9th edition, Springer 2014
Beer, F.P., Johnston, E.R., DeWolf, J.T., Mazurek, D.F., Mechanics of
Materials, 7th edition, McGraw-Hill Education 2012
Hibbeler, R.C., Mechanics of Materials, 10th edition, Pearson 2016
Geer, J.M., Goodno, B.J., Mechanics of Materials, 8th edition, Cengage
Learning 2013
Ghavami, P., Mechanics of Materials, 1st edition, Springer 2015

Collection of Problems
Schaum’s Outlines Strength of Materials, 6th edition, McGraw-Hill
Education 2013
Beer, F.P., Johnston, E.R., DeWolf, J.T., Mazurek, D.F., Mechanics of
Materials, 7th edition, McGraw-Hill Education 2012
Hibbeler, R.C., Mechanics of Materials, 10th edition, Pearson 2016

Notation
The following symbols are used in the solutions to the problems:
↑: short notation for sum of all forces in the direction of the
arrow equals zero.

A: short notation for sum of all moments with reference to
point A equals zero.
; short notation for it follows.
Chapter 1
Stress, Strain, Hooke’s Law
1
 2 Stress
1.1
1.1 Stress, Equilibrium conditions
Stress is related to forces distributed over the
area of a cross section. The stress vector t is
defined as
dF
dF dA
t= ,
dA n
where dF is the force acting on the area ele-
ment dA (unit: 1 Pa = 1 N/m2 ).

Note: The stress vector and its components depend on the orientation
of the area element (with its normal n).

Components of the stress vector:

t
τ
σ – normal stress (perpendicular to the plane)
σ n
τ – shear stress (in plane)

Sign convention: Positive stresses at a positive (negative) face point

in positive (negative) coordinate directions.

Spatial stress state: is uniquely defined

by the components of the stress vectors
in three mutually perpendicular sections.
The stress components are the components z
of the stress tensor σ
τzx z τzy
⎛ ⎞ τxz τyz
σx τxy τxz σy
⎜ ⎟
σ = ⎝τyx σy τyz ⎠ σx τxy τyx
τzx τzy σz x y

Equilibrium of moments yields the following relations

τxy = τyx , τxz = τzx , τyz = τzy .

Hence the stress tensor is a symmetric tensor of second order: τij = τji .
 Plane stress state 3

Plane stress state: is uniquely defined σy

by the stress components of two mutual- τyx
ly perpendicular sections. The stress com-
ponents in the third direction (here z- σx τxy
direction) vanish (σz = τyz = τxz = 0)
σx
 τxy
σx τxy y
σ= . τyx
τxy σy σy
x
τxy = τyx
Coordinate transformation
y
σx + σy σx − σy τξη σξ
σξ = + cos 2ϕ + τxy sin 2ϕ , σx
2 2 ξ
σx + σy σx − σy ϕ
ση = − cos 2ϕ − τxy sin 2ϕ , τxy
2 2
σx − σy ϕ
τξη =− sin 2ϕ + τxy cos 2ϕ . η
2 τyx x
σy
Principal stresses

σ2
σx + σy  σ − σ 2
x y
σ1,2 = ± 2
+ τxy σ1
2 2
2τxy
tan 2ϕ∗ = σ1
σx − σy
y ϕ∗1
ϕ∗2
Note: • The shear stresses vanish σ2
in these directions! x
• The principal directions are
perpendicular to each other: τmax
ϕ∗2 = ϕ∗1 ± π/2. σ0 σ0

Maximum shear stresses

 σ − σ 2 π σ0
x y σ0
τmax = 2 ,
+ τxy ϕ∗∗ = ϕ∗ ± . y
2 4 τmax
ϕ∗∗
In these sections the normal stresses x
reach the value σ0 = (σx + σy )/2.

Invariants
I σ = σx + σy = σξ + ση = σ1 + σ2 ,
IIσ = σx σy − τxy
2
= σξ ση − τξη
2
= σ1 σ2 .
4 Equilibrium conditions

Mohr’s circle
direction of
τ σ2 σ1 center:
τmax
σm = 12 (σx + σy ) ,

2ϕ τ =0

τxy
radius:
2ϕ∗ τξη
ϕ∗

 σ − σ 2
σ2 ση σy σm σx σξ σ1 x y
σ 2
+ τxy
2ϕ ∗∗ 2
y
η
ϕ
ξ
x

• The construction of Mohr’s circle is always possible, provided three

independent quantities are known (e. g. σx , σy , τxy or
σx , σy , ϕ∗ ).
• The shear stress τxy is plotted over σx (τξη over σξ ).
• The angle of transformation ϕ is doubled in the circle (2ϕ) and ori-
ented in opposite direction.

Equilibrium conditions
⎫
∂σx ∂τxy ∂τxz
in space (3D) + + + fx = 0 ,⎪
⎪
⎪
∂x ∂y ∂z ⎪
⎪
⎪
⎪
∂τyx ∂σy ∂τyz ⎬
+ + + fy = 0 , divσ + f = 0 .
∂x ∂y ∂z ⎪
⎪
⎪
⎪
⎪
⎪
+ fz = 0 ,⎪
∂τzx ∂τzy ∂σz ⎭
+ +
∂x ∂y ∂z
⎫
∂σx ∂τxy
in plane (2D) + + fx = 0 ,⎪
⎪
⎬
∂x ∂y
divσ + f = 0 .
∂τyx ∂σy ⎪
⎪
+ + fy = 0 ,⎭
∂x ∂y
where
  ∂σix ∂σiy ∂σiz

divσ = + + ei .
i
∂x ∂y ∂z
 1.2 Strain 5
1.2
1.2 Strain
The strains describe changes in the edge lengths (stretching) and in the
angles (shearing) of a cubic volume element.

Displacement vector P
P u
u = uex + vey + wez

u, v, w = displacement components z
x y
Uniaxial strain state

du
strain ε=
dx
dx du
Biaxial strain state
normal strains shear strains
∂u
dy
∂v ∂y
dy
∂y
dy dy dy ∂v
dx
y ∂x
dx ∂u dx dx
dx x
∂x
∂u ∂v ∂u ∂v
εx = , εy = , γxy = + .
∂x ∂y ∂y ∂x
∂u ∂v ∂w
Triaxial strain state εx = , εy = , εz = ,
∂x ∂y ∂z
⎛ ⎞ ∂u ∂v
εx 1
γ 1γ γxy = γyx = + ,
2 xy 2 xz ∂y ∂x
⎜ ⎟
⎜ ⎟ ∂v ∂w
strain tensor: ε = ⎜ 12 γyx εy 21 γyz ⎟ γyz = γzy = + ,
⎝ ⎠ ∂z ∂y
1
γ 1
γ εz ∂w ∂u
2 zx 2 zy γzx = γxz = + .
∂x ∂z
Remark:
• The strains are, like the stresses, components of a symmetric tensor
of second order. Thus all properties (coordinate transformation, princi-
pal values etc.) of the stress tensor can be used analogously. σx → εx ,
τxy → γxy /2, . . .
• In a plane strain state the following holds: εz = 0, γxz = 0, γyz = 0.
 6 Hook’s law
1.3
1.3 Hooke’s law
Hooke’s law describes the experimentally observed linear relation bet-
ween stresses and strains. The validity of Hooke’s law is restricted by
the proportionality limit (uniaxial σp ). In elastic-plastic materials this
limit frequently conincides with the yield limit (uniaxial σY ).

Uniaxial stress state (bar, beam)

σ
ε= + αT ΔT .
E
E – Young’s modulus,
αT – coefficient of thermal expansion,
ΔT – temperature change.
Plane stress state
1
εx = (σx − νσy ) + αT ΔT ,
E
1
εy = (σy − νσx ) + αT ΔT ,
E
1
γxy = τxy ,
G
E
shear modulus: G= , Poisson s ratio : ν.
2(1 + ν)
Triaxial stress state
1 1
εx = [σx − ν(σy + σz )] + αT ΔT , γxy = τxy ,
E G
1 1
εy = [σy − ν(σz + σx )] + αT ΔT , γyz = τyz ,
E G
1 1
εz = [σz − ν(σx + σy )] + αT ΔT , γzx = τzx .
E G

Selected material data

material E [MPa] ν αT [1/◦ C]

steel 2, 1 · 105 0, 3 12 · 10−6

aluminium 0, 7 · 105 0, 3 23 · 10−6
copper 1, 2 · 105 0, 3 16 · 10−6
concrete 0, 3 · 105 0, 15 . . . 0, 3 10 · 10−6
wood 0, 1 · 105 3 . . . 9 · 10−6

Remark: 1MPa = 103 kPa = 106 Pa , 1Pa = 1N/m2

 1.3 Hooke’s law 7

Problem 1.1 In a thin metal sheet σy P1.1

the stresses σx , σy , τxy are given. De- τxy
termine value and direction of the
principal stresses. σx τxy

Given: σx = 20 MPa, σy = 30 MPa, τxy σx

y
τxy = 10 MPa.
τxy
x σy

Solution We start with the analytical method. The principal stresses

are computed by

σx + σy  σ − σ 2 √
x y
σ1,2 = ± 2 = 25 ±
+ τxy 25 + 100 = 25 ± 11.18
2 2
leading to
σ1 = 36.18 MPa , σ2 = 13.82 MPa .

For the principal directions, we obtain according to σ1

σ2
2τxy
tan 2ϕ∗ = = −2
σx − σy σ2
σ1
the results y

ϕ∗1 = 58.28◦ , ϕ∗2 = 148.28◦ . ϕ∗1

x

To illustrate the results an element loaded by the principal stresses is

sketched.
We can also solve the problem graphically by using Mohr’s circle:

τ
direction of scale: 10 MPa
σ1
σ2

τxy We read off the

σ2 ϕ∗1 2ϕ∗1 σ1 results:
σx σy σ
σ1 ∼
= 36.5 MPa ,
σ2 ∼
= 14 MPa ,
ϕ∗1 ∼
= 59◦ .
 8 Plane

P1.2 Problem 1.2 Determine the stress components, the principal stresses,
and the principal directions, as well as the maximum shear stress in
any cross section for the given special cases of plane stress states :
a) σx = σ0 , σy = 0, τxy = 0 (uniaxial tension),
b) σx = σy = σ0 , τxy = 0 (biaxial, equal tension),
c) σx = σy = 0, τxy = τ0 (pure shear).

Solution to a) The stress components are obtained for any cross sec-
tion which has the angle ϕ to the x- and y-
direction by inserting σx , σy and τxy into the
transformation relations
σ0 σ0
σξ = 12 (σ0 + 0) + 1
2
(σ0 − 0) cos 2ϕ + 0 · sin 2ϕ
= 12 σ0 (1 + cos 2ϕ) ,

ση = 12 (σ0 + 0) − 1
2
(σ0 − 0) cos 2ϕ − 0 · sin 2ϕ τξη
= 12 σ0 (1 − cos 2ϕ) , σξ

τξη =− 12 (σ0 − 0) sin 2ϕ + 0 · cos 2ϕ

= 12 σ0 sin 2ϕ . ση

Due to τxy = 0 the stresses σx , σy are principal y

ϕ
stresses, and the x- as well as y-direction are prin-
x
cipal directions:
τmax
π σ0 σ0
σ1 = σx = σ0 , σ2 = σy = 0, ϕ∗1 = 0, ϕ∗2 = ± .
2 2 2

The maximum shear stress and the corresponding

direction is determined by the following relations
y 45◦
1 1 ∗∗ π
τmax = |σ1 − σ2 | = σ0 , ϕ =± . x
2 2 4

Remark: A plate made from a material that supports only limited shear-
stresses will fail along lines under an angle of ±45◦ to the
x-axis.

to b) Inserting the given values into the coordinate transformation

yields

σξ = σ0 , ση = σ0 , τξη = 0 .
 stress state 9

Therefore the normal stress σ0 σ0

is acting in any section, and the
shear stress vanishes. There is σ0
no distinguished principal direc-
tion, any section is a principal σ0
direction: σ0 y ϕ
σ1 = σ2 = σ0 . x

to c) In this case the coordinate transformation yields

τ0
σξ = τ0 sin 2ϕ , ση = −τ0 sin 2ϕ , τξη = τ0 cos 2ϕ .

The principal stresses and directions are

π π
σ1 = +τ0 , σ2 = −τ0 , ϕ∗1 = , ϕ∗2 = − . σ1 = τ0
4 4
For the maximum shear stress and the correspon-
ding directions we obtain y σ1
45◦ |σ2 | = τ0
τmax = τ0 , ϕ∗∗
1 = 0 , ϕ∗∗
2 = π/2 . x

Remark: A plate made from a material that supports limited normal

stresses will fail along lines under an angle of ±45◦ to the
x-axis.
The results of all three stress states can be illustrated by the correspon-
ding Mohr’s circles:
to a) τ to c) τ

τmax
σ1 = σ0 σ1 = τ0
σ2 = 0 σ σ2 = −τ0 σ

τ
to b)

σ1 = σ2 = σ0 σ

Note: In case b) Mohr’s circle degenerates to a single point along

the σ-axis!
 10 Plane
σ1
P1.3 Problem 1.3 In a plane section the fol-
lowing principal stresses are present
σ2 σ2
σ1 = 96 MPa and σ2 = −52 MPa .

a) Determine the stresses in sections

which are inclined by ϕa = 60◦ with σ1
regard to the principal axes?
b) In which section ϕb does the normal stress vanish? What are the
values of the shear and normal stresses in a direction perpendicular to
the direction ϕb ?
c) In which directions do the maximal shear stresses appear, and what
are the corresponding normal stresses?

Solution to a) According to the sketch we σy = σ1

use a coordinate system x, y that coincides
with the principal axes. The stresses in the σx = σ2
cross sections inclined by ϕa = 60◦ follow
from the coordinate transformation y
τxy = 0
σ2 + σ1 σ2 − σ1 1
σξa = + cos 2ϕa = 22 + 74 · x
2 2 2
= 59 MPa ,
a
σ2 + σ1 σ2 − σ1 1 τξη
σηa = − cos 2ϕa = 22 − 74 · σηa σξa
2 2 2
= −15 MPa ,
σηa
σ2 − σ1 1√ σξa
a
τξη = − sin 2ϕa = 74 · 3 a
2 2 τξη
60◦
= 64.1 MPa . x

to b) For the normal stress σξ to vanish the following must hold

σ2 + σ1 σ2 − σ1
σξb = + cos 2ϕb = 0
2 2

22
; cos 2ϕb = = 0.297 ; 2ϕb = 72.7◦ ; ϕb = 36.35◦ .
74
 stress state 11

For σηb and τξη

b
we obtain
σηb b
σ2 + σ1 σ2 − σ1 τξη
σηb = − cos 2ϕb = 44 MPa ,
2 2
b
τξη
b σ2 − σ1 b
τξη
τξη = − sin 2ϕb = 74 · 0.955
2 σηb
= 70.7 MPa . x
36, 35◦

to c) The maximum shear stress occurs in directions of ±45◦ with

regard to the principal axes. This results in
σ1 − σ2
τmax = = 74 MPa . σm τmax σm
2
The corresponding normal stresses are

σ1 + σ2 σm σm
σm = = 22 MPa
2 τmax
for the given data. x
45◦

All informations can be illustrated by use of Mohr’s circle for the given
stress state

scale: 50 MPa τ

τmax
σξa ∼
b
= 59 MPa , τξη
a
τξη
σηa ∼
= −15 MPa , 2ϕb 2ϕa
a ∼
τξη = 64 MPa , σ
σx = σ2 σηa σξb = 0 σm σηb σξa σy = σ1
ϕb ∼
= 37◦ ,
σηb ∼
= 44 MPa ,
b
τξη ∼
= 71 MPa ,
τmax ∼
= 74 MPa ,
σm ∼
= 22 MPa .
 12 Plane

P1.4 Problem 1.4 The following stresses σy

are acting in a panel.
σx = 20 MPa, σy = 60 MPa and τxy
τxy = −40 MPa. σx
Determine analytically and graphical- σx
ly the principal stresses, the maximum y
shear stress, and the corresponding di- τxy
rections. Sketch the related sections. x σy

Solution The principal stresses and their directions are calculated ana-
lytically by

σx + σy  σ − σ 2 σ1
x y
σ1,2 = ± 2
+ τxy |σ2 |
2 2
 |σ2|
=40 ± (20)2 + (40)2 , y
σ1
x
; σ1 = 84.72 MPa , σ2 = −4.72 MPa , ϕ∗2

2τxy
tan 2ϕ∗ = =2 ; ϕ∗1 = 121.7◦ , ϕ∗2 = 31.7◦ .
σx − σy
To determine which principal stress is associated with which direction,
the transformation relations or Mohr’s circle has to be used.
For the maximum stress the following result is
obtained τmax σm

σm
 σ − σ 2
x y
τmax = 2 = 44.72 MPa ,
+ τxy σm
2
ϕ∗∗ = ϕ∗ ± 45◦ = 31.7◦ ± 45◦ . σm τmax

x
The graphic solution by Mohr’s circle is sket- ϕ∗∗
ched below:
τ
scale: 20 MPa
τmax
ϕ∗1 2ϕ ∗∗ 2ϕ∗1
σ1 ∼
= 85 MPa , σx σ1
σ2 ∼
= −5 MPa , σ2 σm σy σ
τmax ∼
= 45 MPa ,
τxy

ϕ∗1 ∼
= 122◦ ,
ϕ∗∗ ∼
= 77◦ . σ2
σ1
direction of
 stress state 13

Problem 1.5 A thin-walled tube B x P1.5

is loaded by a bending moment, an
internal pressure, and a torsional
moment. In points A and B the s
following stresses occur due to the A
loading:

σxA,B = ±50 MPa , σsA,B = 100 MPa , A,B

τxs = 100 MPa .

Determine value and direction of the principal stresses in the points A

and B.
σs
Solution For point A the principal stresses τxs
are computed by σx

σ1,2 = 12 (σx + σs ) ± [ 12 (σx − σs )]2 + τxs
2 σx
s
 τxs
= 75 ± (−25)2 + 1002 x
σs
= 75 ± 103.08 |σ2 |
σ1
yielding
σ1 = 178.08 MPa , σ2 = −28.08 MPa .
|σ2 |
σ1
For the principal directions we obtain 52◦
∗
tan 2ϕ = 2τxs
σx −σs
= 2·100
50−100
= −4 ; ϕ∗1 = 52.02 , ◦
ϕ∗2 = −37.98◦ .

From the coordinate transformation it is obvious that direction ϕ∗1 is

associated with the principal stress σ1 :
σξ = 12 (σx + σs ) + 12 (σx − σs ) cos 2ϕ∗1 + τxs sin 2ϕ∗1
σs
= 75 − 25 · (−0.242) + 100 · 0.970
τxs
= 178.08 MPa = σ1 . |σx |

|σx |
In an analogous way the principal stresses and
s
their directions in point B are obtained: τxs
 x σs
σ1,2 = 25 ± (−75)2 + 1002
|σ2 | σ1
= 25 ± 125
; σ1 = 150 MPa , σ2 = −100 MPa . |σ2 |
2 · 100
tan 2ϕ∗ = = −1.33 σ1
−50 − 100
∗ ◦ 63◦
; ϕ1 = 63.4 , ϕ2 = −26.6◦ .
∗
 14 Plane stress state

P1.6 Problem 1.6 In a thin aluminium sheet

(E = 0.7 · 105 MPa , ν = 0.3) the fol-
lowing strains εx = 0.001, εy = 0.0005, y P
γxy = 0 are experimentally measured in ξ
η
point P .
x
ϕ = 30◦
What are the principal stresses, the ma-
ximum shear stress, and the stresses in
a sections, that are inclined by ϕ = 30◦
with regard to the principal directions?

Solution In the aluminium sheet a state of plane stress is present. From

Hooke’s law

Eεx = σx − νσy , Eεy = σy − νσx , Gγxy = τxy

the following stresses can be computed

E 0.7 · 105
σx = (εx + νεy ) = (0.001 + 0.00015) = 88.5 MPa ,
1 − ν2 1 − 0.09
E 0.7 · 105
σy = (ε y + νε x ) = (0.0005 + 0.0003) = 61.5 MPa ,
1 − ν2 1 − 0.09
τxy = 0 .

As the shear stress τxy is equal to zero, σx , σy are principal stresses,

and axes x, y are principal axes:

σx = σ1 σy = σ2 .

Therefore, the maximum shear stress is

1 1
τmax = (σ1 − σ2 ) = (σx − σy ) = 13.5 MPa .
2 2
For the sections inclined by ϕ = 30◦ , the stresses follow with τxy = 0
from the transformation relations
σx + σx σx − σy
σξ = + cos 2ϕ = 75 + 13.5 cos 60◦ = 81.75 MPa ,
2 2
σx + σy σx − σy
ση = − cos 2ϕ = 75 − 13.5 cos 60◦ = 68.25 MPa ,
2 2
σx − σy
τξη = − sin 2ϕ = −13.5 sin 60◦ = −11.69 MPa .
2
 Displacements 15

Problem 1.7 In a thin sheet the following P1.7

plane displacement field was obtained by A
measurements: v
y
u(x, y) = 3.5 · 10−3 x + 2 · 10−3 y , A u
v(x, y) = 1 · 10−3 x − 0.5 · 10−3 y . x
a) Determine the state of strain.
b) What are principal strains, and under which angle to the x-axis do
they appear?
c) What is the maximum shear strain γmax ?

Solution to a) The strains are computed by differentiation of the dis-

placement components:
∂u ∂v
εx = = 3.5 · 10−3 , εy = = −0.5 · 10−3 ,
∂x ∂y
∂u ∂v
γxy = + = 2 · 10−3 + 1 · 10−3 = 3 · 10−3 .
∂y ∂x
The strains are constant in the entire sheet (=homogeneous strain
state).
to b) The principal strains and their corresponding directions are cal-
culated from the relations for the principal stresses by using the repla-
cements (σx → εx , τxy → γxy /2 etc.). This yields the principal strains

εx + εy  ε − ε 2  γ 2
x y xy
ε1,2 = ± +
2 2 2

= 1.5 · 10−3 ± (2 · 10−3 )2 + (1.5 · 10−3 )2 = 1.5 · 10−3 ± 2.5 · 10−3

; ε1 = 4 · 10−3 , ε2 = −1 · 10−3 ,

and the principal directions

γxy 3
tan 2ϕ∗ = = ; ϕ∗1 = 18.4◦ , ϕ∗2 = 108.4◦ .
εx − εy 4
to c) The maximum shear strain is

γmax = ε1 − ε2 = 4 · 10−3 + 1 · 10−3 = 5 · 10−3 .

It occurs at angles, which are inclined by ±45◦ with regard to the

principal directions.
 16 Plane stress state
p
P1.8 Problem 1.8 An elastic panel A fits in-
to the rigid socket B of height h
(Young’s modulus E, Poisson’s ratio R vR
ν > 0). B A h
y
Determine the stress σx and the value
x
of the displacement vR at the top edge
R for a constant pressure p. It is ass-
umed that the elastic panel can move
frictionless in the socket mounting.
p = −σy

Solution In the panel a uniform plane

stress state is present, where the stress σx σx
component σy is known: σy = −p. Thus h
Hooke’s law yields

Eεx = σx − νσy = σx + νp ,
y
Eεy = σy − νσx = −p − νσx . p = −σy
x
As the panel cannot expand in x-direc-
tion, it holds

εx = 0 .

Inserting this into Hooke’s law provides the stress σx and the normal
strain in y-direction:

1 − ν2
σx = −νp , εy = −p .
E
Knowing the strain εy we compute the displacement v by integration:

∂v 1 − ν2
= εy ; v(y) = εy dy = −p y+C.
∂y E
The lower edge of the panel does not experience a displacement, i. e.
v(0) = 0, and C = 0. For the value of the displacement at the top edge
we obtain
1 − ν2
vR = |v(h)| = ph .
E
 Plane stress state 17

Problem 1.9 Two quadratic panels made P1.9

of different materials have both the edge
length a in the unloaded state. As sket- 1
a
ched, they are inserted into a rigid socket, 
1 E1 , ν1
which has an opening l, that is smaller l
than 2a.

2 2
What are the stresses and the changes a
E2 , ν2
of the edge lengths, if it is assumed that
the panel can slide frictionless in the rigid a
socket?
σy2
Solution After force fitting into the rigid
socket the panels experience a uniform 
1
plane stress state. Equilibrium in vertical y
x
direction yields σy1 = σy2 = σy . Consi-
dering the condition σx1 = σx2 = 0 in

2
Hooke’s law for both panels provides


1 E1 εy1 = σy , E1 εx1 = −ν1 σy ,
σy1

2 E2 εy2 = σy , E2 εx2 = −ν2 σy .
Δv
With the strain-displacement relation
(constant strains) a y
x
Δu1 Δv1 Δu2 Δv2 Δu
εx1 = , εy1 = , εx2 = , εy2 = a
a a a a
and the kinematic compatibility

(a + Δv1 ) + (a + Δv2 ) = l

we obtain for the normal stress in y-direction

2a − l E1 E2
σy = − .
a E1 + E2

This stress leads to the following length changes

E2 E1
Δv1 = −(2a − l) , Δv2 = −(2a − l) ,
E1 + E2 E1 + E2

Δu1 = −ν1 Δv1 , Δu2 = −ν2 Δv2 .

 18 Thin-walled pressure vessel

P1.10 Problem 1.10 A thin-walled diving sphere

(radius r = 500 mm, wall thickness t = pW water
12.5 mm) is submerged 1000 m under the r
water surface (pressure pW = 10 MPa).
Compute the stresses in the wall of the
sphere. t

Solution We cut the sphere with a section

perpendicular to the surface of the sphere,
σt σt
resulting in two hemispheres. The equilibri-
um conditions r
t
↑ : σt 2πrt + pW r π = 0
2

provide for any section (spherical symme- pW

try) the stresses
r 500
σt = −pW = −10 = −200 MPa .
2t 2 · 12, 5

P1.11 Problem 1.11 A spherical steel tank is heated by r

a hot gas (ΔT = 300 ◦ C) and additionally subjec- ΔT
ted to an internal pressure (p = 1.5 MPa). p
t
Compute the change of the radius.
Given: r = 2 m, t = 10 mm, E = 2.1 · 105 MPa,
ν = 0.3, αT = 12 · 10−6 ◦ C−1 .

Solution For any cross section perpendicular to σt

the surface of the sphere equilibrium provides σϕ

r
σt = σϕ = p .
2t
The strain is computed by the change of circumference
p
2π(r + Δr) − 2πr Δr
εt = εϕ = = .
2πr r σt
Using Hooke’s law
ϕ
Eεt = σt − νσϕ + EαT ΔT

yields
   
p r(1 − ν) 1.5 · 10−3
Δr = r + αT ΔT = 2000 + 3.6 · 10−3 = 8.25 mm .
2Et 3
 Thin-walled pressure vessel 19

Problem 1.12 A thin-walled cylindrical t P1.12

pressure vessel made of steel is subjected
to an internal pressure p. r
p
What is the maximum value of the pres-
sure such that the normal stresses in the l
central part do not to exceed the limit
stress σlim?
Compute for this case the change of radius r and length l.
Given: l = 5 m, r = 1 m, t = 1 cm, E = 2.1 · 105 MPa,
ν = 0.3, σlim = 100 MPa. ϕ
σx
Solution The stresses are determined by x
equilibrium conditions at suitable sections: p
r
→: pr 2 π − σx 2rπt = 0 ; σx = p ,
2t p
r σϕ
↑ : σϕ 2d t − p2rd = 0 ; σϕ = p .
t

These stresses are principal stresses, as the

shear stress vanish in these sections. The lar-
d 2r
gest normal stress exceeds the limit stress for
t
σϕ ≤ σlim ; p ≤ σlim = 1 MPa ; pmax = 1 MPa .
r
The related hoop strain εϕ results from the circumferential change:
2π(r + Δr) − 2πr Δr
εϕ = = .
2πr r
Hooke’s law Eεϕ = σϕ − νσt provides
pmax r  ν
Δr = r 1− = 0.41 mm .
Et 2
In an analogous way the strain εt = Δl/l and Hooke’s law Eεt =
σt − νσϕ provide the length change
 
pmax r 1
Δl = l − ν = 0.47 mm .
Et 2
Note: The caps at the ends of the pressure vessel are excluded-
i. e. the solution for the stresses is only valid in a sufficient distance
from the caps.
 20 Thermal strains

P1.13 Problem 1.13 The rails of a train track are installed at a temperature
of 15◦ C such that no internal forces are present.
Determine the stress at a temperature of −25◦ C, if it is assumed that
the rails cannot experience any length change?
Given: E = 2.1 · 105 MPa, αT = 12 · 10−6 ◦ C−1 .

Solution In the rail exists a uniaxial stress state and Hooke’s law
provides

E ε = σ + E αT ΔT .

As displacements are suppressed, ε has to be zero. Using ΔT = −40◦ C

yields for the stresses

σ = −E αT ΔT = 2.1 · 105 · 12 · 10−6 · 40 = 100.8 MPa .

Note: In rails the stresses due to temperature changes can become

considerably large.

P1.14 Problem 1.14 A thin copper ring of radius r is heated due to the tem-
perature difference ΔT .
What are the changes in radius and circumference if it is assumed that
the ring can deform freely?
Given: r = 100 mm, αT = 16 · 10−6 ◦ C−1 , ΔT = 50◦ C.

Solution A uniform, stress-free uniaxial strain state exists in the ring

after heating. The strain is determined by the change in circumferencial
direction (change in length) Δl:
Δl 2π(r + Δr) − 2πr Δr
ε= = = . Δr
l 2πr r
Using the Hooke’s law for uniaxial states r
σ
ε= + αT ΔT
E
and the stress-free condition σ = 0, leads to

Δr = r αT ΔT = 100 · 16 · 10−6 · 50 = 0.08 mm ,

l
Δl = Δr = 2πΔr = 0.50 mm .
r
 Thermal strains 21
δ
Problem 1.15 A rectangular plate (a > b) P1.15
is inserted into a rigid oversized opening, E , ν , αT
y
such that spacings of size δ are present. b
Subsequently the plate is heated. It is as- x
sumed, that the plate can move frictionless
along its edges. a δ
a) Which temperature increase ΔTa is re-
quired to just close the spacing on the right?
b) For which temperature increase ΔTb is the upper spacing just clo-
sing? What is the value of σx in this situation?
c) What are the stresses in the plate for ΔT > ΔTb ?

Solution to a) For ΔT < ΔTa the plate expands in a stress-free way.

With σx = σy = 0 Hooke’s law reduces to

εx = εy = αT ΔT .

The spacing on the right is closing, if the condition εx = δ/a is met.

Introducing this result yields the temperature increase:
δ
ΔTa = .
αT a

to b) At a temperature increase ΔTa ≤ ΔT ≤ ΔTb the plate can only

expand freely in y-direction, while the strain in x-direction remains
constant. With σy = 0 and εx = δ/a it follows
δ σx σx
= + αT ΔT , εy = −ν + αT ΔT .
a E E
The upper spacing is closing, if the the condition εy = δ/b is satisfied.
All above relations provide
δ a + νb E δ(a − b)
ΔTb = , σx = − .
αT a (1 + ν)b 1+ν ab

to c) For ΔT > ΔTb the strains in both directions remain constant:

εx = δ/a, εy = δ/b. Then

Eεx = σx − νσy + EαT ΔT , Eεy = σy − νσx + EαT ΔT

provide the stresses

   
δ(νa + b) αT ΔT δ(νb + a) αT ΔT
σx = E − , σy = E − .
(1 − ν 2 )ab 1−ν (1 − ν 2 )ab 1−ν
 22 Thermal stresses

P1.16 Problem 1.16 A thin-walled bushing

has to be heated by the temperature
difference ΔT ∗ , to be mounted on a t
shaft. r
E , ν , αT
What are the stresses in the bushing, x
and what is the pressure p between rigid
bushing and shaft after cooling? It is
assumed that the shaft is rigid and
that the displacements of the bushing
in x-directions are blocked by friction.

Solution Before cooling the bushing is stress-free. The stresses after

cooling are obtained by equilibrium, Hooke’s law, and kinematics. The
equilibrium condition provides
r
p · 2rd = σϕ 2t d ; σϕ = p .
t
Hooke’s law with ΔT = −ΔT ∗ (coo-
ling!) states x
ϕ
∗
Eεϕ = σϕ − νσx − EαT ΔT ,
p σϕ
σϕ
d
Eεx = σx − νσϕ − EαT ΔT ∗ .

During cooling the strains in the bushing (shrinking) are blocked by

the shaft and friction. Thus the kinematic relations are given by

εϕ = 0 , εx = 0 .

Combining the above relations and solving for stresses and pressure
yields
E t E
σx = σϕ = αT ΔT ∗ , p= αT ΔT ∗ .
1−ν r 1−ν

Note: • In the bushing a plane stress state is present with equal nor-
mal stresses: σx = σϕ .
• If the bushing can deform freely in x-direction (no friction,
εx = 0), then σx = 0 and σϕ = EαT ΔT ∗ follow.
 Thermal stresses 23

Problem 1.17 On the thin-walled elastic shaft  1 

2 P1.17
r t
a pipe 
2 will be mounted by heat shrinking. Before
heat shrinking both parts have identical geometri- E2 , αT 2
cal dimensions, but are made of different materi-
als. 
1
r t
Which temperature difference is required to the
mount pipe  2 on the shaft  1 ? E1
What is the pressure p between the shaft and the
pipe after cooling, if it is assumed that no stresses
are present in axial direction?

Solution For the pipe 2 to be mounted on the shaft 

1 its radius has
to increase by thermal expansion by t. Thus in the heated state, the
hoop strain has to assume the value
2π(r + t) − 2πr t
εϕ2 = =
2πr r
Now Hooke’s law yields with σϕ2 = 0 (the pipe is stress free in the
heated state!)
1 t
εϕ2 = αT 2 ΔT ; ΔT = .
αT 2 r
The pressure after cooling is obtained from the equilibrium equations
r r
σϕ1 = − p , σϕ2 = + p , 
2
t t p
Hooke’s laws, σϕ2

E1 εϕ1 = σϕ1 , E2 εϕ2 = σϕ2 , p


1
the strains
Δr1 Δr2 σϕ1
εϕ1 = , εϕ2 =
r r
and the kinematic compatibility

Δr2 = Δr1 + t .

Combining the above equations yields

 2
E1 E2 t
p= .
E1 + E2 r
 24 Hooke’s law

P1.18 Problem 1.18 A block is subjected to

a pressure p0 in z-direction by a rigid z
press. E ,ν
y x
Determine the strains and stresses, if
a) the deformations in x- and y-direction are restrained,
b) only the deformation in y-direction is restrained,
c) the deformations in x− and y−direction are not restrained?

Solution In the above cases a homogeneous, triaxial stress and strain

state is present in the plate. With σz = −p0 Hooke’s law yields (there
are no shear stresses present!):

Eεx = σx −νσy +νp0 , Eεy = σy +νp0 −νσx , Eεz = −p0 −νσx −νσy .

For case a) we have εax = εay = 0, and from

0 = σxa − νσya + νp0 , 0 = σya + νp0 − νσxa , Eεaz = −p0 − νσx − νσy

it follows
1 − ν − 2ν 2 p0 ν
εaz = − , σxa = σya = − po .
1−ν E 1−ν
In case b) εby = 0 and σxb = 0 holds (free deformation, i. e. no stresses
in x-direction). With Hooke’s law

Eεbx = −νσyb + νp0 , 0 = σyb + νp0 , Eεz = −p0 − νσyb

we obtain
p0 p0
εbx = ν(1 + ν) , εbz = −(1 − ν 2 ) , σyb = −ν p0 .
E E

In case c), both σxc = σyc = 0, because the deformations in these direc-
tions are not restrained. Therefore Hooke’s law reduces to

Eεcx = ν p0 , Eεcy = ν p0 , Eεcz = −p0 ,

and we have
p0 p0
εcx = εcy = ν , εcz = − .
E E

Note: For ν > 0 we have |εaz | < |εbz | < |εcz |. Especially for ν = 1/3 it
follows
εaz = − 6p0 /(9E) , εbz = − 8p0 /(9E) , εcz = − 9p0 /(9E) .
Due to the deformation constraints in x- and y-direction the plate be-
haves rather stiff in case a)!
 Thick-walled cylinder 25

Problem 1.19 In a thick-walled cylinder P1.19

with a restrained deformation in longi-
tudinal direction (plane strain state) the b
following stresses are present due to loa- r
ding by an internal pressure p: ϕ
a
p
 2 
a2 b
σr = −p 2 − 1 ,
b − a2 r 2
σϕ
 
a2 b2 σr
σϕ = p 2 + 1 . σr
b − a2 r2 σϕ
Determine the stress σz and the resulting
force Fz in axial direction of the cylinder.

Where does the maximum normal stress occur, and what is its value?
Given: p = 50 MPa, a = 100 mm, b = 200 mm, ν = 1/3.

Solution As the deformation in axial direction of the cylinder is res-

trained, we have εz = 0. Hooke’s law in this direction provides

Eεz = 0 = σz − ν(σr + σϕ ) .

Inserting the known relations yields the stress

a2 2
σz = ν(σr + σϕ ) = 2ν p = p = 11.1 MPa .
b2 − a 2 9
As σz is constant across the section, the resulting force is computed by
multiplication of σz with the cross section area:

Fz = σz π(b2 − a2 ) = 2πν p a2 = 1.05 · 106 N .

The absolute values of the stresses σr and σϕ are maximum at the inner
boundary of the cylinder (r = a). There we have
5 2
σr (a) = −p , σϕ (a) = p, σz = p.
3 9
Thus the hoop stress σϕ on the inside is the largest normal stress.
 26 Stresses and strains

P1.20 Problem 1.20 A rigid box with

g
quadratic cross section is filled
with clay (volume V = a2 h, den- Δh
sity ρ). The material behavior h
of the clay is approximated by
Hooke’s law (Young’s modulus y
E, Poisson’s ratio ν).
x a
Determine the settlement Δh z a
of the clay as a consequence of the
weight of the clay and the horizontal pressure distribution at the box
walls as a function of y.

Solution Due to the given loading situation only normal stresses σx , σy ,

and σz are present in the three coordinate directions x, y, and z. Ex-
cept for the strain εy no other strains occur. For σy it holds according
to Hooke’s law with εx = εz = 0
 
E ν E 1−ν
σy = εy + εy = εy .
1+ν 1 − 2ν 1 + ν 1 − 2ν
With the stress distribution

σy = −ρg(h − y)

the settlement Δh is computed by

dv
εy = .
dy
By integration we obtain Δh:
 h  h
(1 + ν)(1 − 2ν)
Δh = v(h) = εy dy = − ρg(h − y) dy
0 0 E(1 − ν)
  h
(1 + ν)(1 − 2ν) y2 1 (1 + ν)(1 − 2ν)
= − ρg hy − =− ρgh2 .
E(1 − ν) 2 0
2 E(1 − ν)

The horizontal pressure distribution as a function of y follows from

Hooke’s law:
Eν (1 + ν)(1 − 2ν)
σx = σz = εy , εy = −ρg(h − y)
(1 + ν)(1 − 2ν) E(1 − ν)
−ν
; σx (y) = σz (y) = ρg(h − y) .
1−ν
 Stresses and strains 27

Problem 1.21 In a sheet metal (Young’s P1.21

modulus E and Poisson’s ratio ν) the
three strains εA = ε̄, εB = 3 ε̄ und B
εC = 2 ε̄ are measured by strain gauges
in the sketched directions.
A
a) Determine the principal strains ε1 C
and ε2 . 30◦
b) Compute the principal stresses σ1 and
σ2 under the assumption of a plane state
of stress.
c) Calculate the principal directions.
η y

Solution to a) We introduce a x, y- and ξ

a ξ, η- coordinate system in direction of
the strain gauges. Then it holds for the
measured strains
x
εx = ε̄ , εy = 3ε̄ , εξ = 2ε̄ . ◦
ϕ = 30

To compute the principal strains we have to determine the shear strain

γxy . According to the transformation relations for ϕ = 30◦ we have
1 1 1
εξ = (εx + εy ) + (εx − εy ) cos 2ϕ + γxy sin 2ϕ
2 2 √ 2
1 1 3
= (εx + εy ) + (εx − εy ) + γxy ,
2   4 √ 4
1 3
2ε̄ = 2ε̄ + − ε̄ + γxy .
2 4
This yields the result
2
γxy = √ ε̄ .
3
With this at hand, the principal strains can be calculated via

εx + εy  ε − ε 2  1 2
x y
ε1/2 = ± + γxy
2 2 2

to be
   
1 1
ε1 = 2 1 + √ ε̄ , ε2 = 2 1 − √ ε̄ .
3 3
28 Stresses and strains

to b) Using the assumption of a plane stress state, Hooke’s law formu-

lated in principal directions provides the principal stresses
 
E 2 E ε̄ 1−ν
σ1 = (ε1 + νε2 ) ; σ1 = 1+ν + √ ,
1 − ν2 1 − ν2 3
 
E 2 E ε̄ 1−ν
σ2 = (ε2 + νε1 ) ; σ2 = 1+ν − √ .
1 − ν2 1 − ν2 3

to c) The principal directions follow either from the stress or from the
strain components. Here we use the strain components to obtain from
the general formula

√2
γxy
∗ 3 1
tan 2ϕ = = = −√
εx − εy −2 3
the solutions

ϕ∗ = −15◦ und ϕ∗ = 75◦ .

In order to decide, which direction corresponds to the principal strain

ε1 or ε2 , respectively, we use the angle ϕ∗ = −15◦ in the coordinate
transformation. This yields with the given strain components
1 1 1
εξ = (εx + εy ) + (εx − εy ) cos(−30◦ ) + γxy sin(−30◦ )
2 √ 2   2
3 ε̄ 1 1
= 2ε̄ − ε̄ −√ =2 1− √ ε̄ = ε2 .
2 32 3
The smallest principal strain ε2 occurs at the angle ϕ∗ = −15◦ , while
the largest principal strain ε1 is related to the direction of ϕ∗ = 75◦ .
Chapter 2
Tension and Compression in Bars
2
30 Stress

Tensile or compressive loading in bars

Assumptions:

• Length l of the bar is large compared to

characteristic dimensions of the cross sec- x,u
tion A(x).
• Axis of the bar (line connecting centroids
of the cross sections) is a straight line. n
• Common line of action (external loads F l
and n(x) are aligned with the axis of the A
bar).
• Cross section A(x) can only vary slightly.

Stress: Assuming a constant stress σ across

the section A the following relation with the
normal force N holds: F

N (x)
σ(x) = .
A(x)

Basic equations of a deformable bar:

dN
equilibrium condition = −n ,
dx
σ
Hooke’s law ε= + αT ΔT ,
E
du
Kinematic relation ε=
dx

E = Young’s modulus,
αT = coefficient of thermal expansion,

ΔT = temperature difference with respect to a reference state,

u(x) = displacement of a point x within the bar.

The basic equations lead to a single differential equation for the displa-
cements ( {·} := d{·}/dx ):

(EAu ) = −n + (EAαT ΔT ) .
 Tension and compression 31
 l
Elongation of a bar: Δl = u(l) − u(0) = ε dx .
0
special cases:
 l
N
Δl = dx (ΔT = 0) ,
0 EA

Fl
Δl = (N = F = const, EA = const, ΔT = 0),
EA
Δl = αT ΔT l (N = 0, EA = const, αT ΔT = const).

Superposition: The solution of a statically indeterminate problem can

be achieved by superposition of solutions of associated statically deter-
minate problems considering the compatibility conditions.
(0) (1)
“0”-System uB “1”-System uB
F F
X =B
A B (0) (1)
uB + uB = 0.

Rotating bar: A bar rotating with the angluar velocity ω experiences

an axial loading per unit length of

n = ρA xω 2 . ω A ρ
Here ρ is the density and x represents
the distance of the cross section A from n
x
the center of rotation.

Elastic-plastic bar: For an elastic-ideal-plastic material behavior,

Hooke’s law is valid only until a certain
yield limit σY : σ
⎧ σY
⎨E ε , |ε| ≤ εY ,
σ=
⎩ εY ε
σY sign(ε) ,|ε| ≥ εY .
−σY

System of bars: The displacements are obtained by “disconnecting”

and “reconnecting” of the bars from the nodes using a displacement
diagram.

Note: In areas with rapidly changing cross sections (notches, holes)

the above theory for bars is not applicable.
 32 Stress

P2.1 Problem 2.1 Determine the stresses distri-

bution σ(x) in the homogeneous bar due to A0
g
its weight. The bar has constant thickness ρ
and a linear varying width. Furthermore,
identify the location and value of the smal- h
lest stress. x
F F a

It is reasonable to introduce the x-coordinate at the intersection of the

extended edges of the trapeziod. The x dependent cross section area
follows then as

A(x) = A0 x/l .

With the weight

 x
x 2 − a2
W (x) = ρgV (x) = ρg A(ξ)dξ = ρgA0
a 2l
of the lower part equilibrium provides
N(x)
x 2 − a2 A(x)
N (x) = F + W (x) = F + ρgA0 .
2h
This leads to the stress W (x)
  x
N (x) F h + ρg A20 x2 − a2 ξ
σ(x) = = . a F
A(x) A0 x

The location x∗ of the minimum is determined by condition σ  = 0:

  
 Fh 1 ρg a2 ∗ 2F h
σ =− + 1+ 2 =0 ; x = − a2 .
A0 x2 2 x ρgA0

The value of the minimum stress is


∗ 2F h
σmin = σ(x ) = ρg − a2 = ρgx∗ .
ρgA0

Note:
• For ρg = 0 (“weight-less bar”) no minimum exists. The largest stress
occurs at x = a.
• The minimum will be located within the bar, only if a < x∗ < h or
ρgA0 a2 /(2h) < F < ρgA0 (h2 + a2 )/(2h) holds.
 Stress 33
a
Problem 2.2 The contour of a light- P2.2
house with circular thin-walled cross W
section follows a hyperbolic equation
y
b2 − a 2 2
y −
2
x = a2 . x
h2 t
h
Determine the stress distribution as a
consequence of weight W of the light-
house head (the weight of the structure
can be neglected).
Given: b = 2a, t
a. b

Solution As the weight W is the only acting external load, the normal
force N is constant (compression):

N = −W .

The cross section area A is changing. It can be approximated by (thin-

walled structure with t
y)

b2 − a 2 2
A(x) = 2πyt= 2πt a2 + x
h2 y

a 2
= 2πt a2 + 3 2 x2
h

t
x2
= 2πat 1 + 3 2 .
h

The stress follows now as

N W
σ(x) = =−
.
A 2
2πat 1 + 3 x2
h

Especially at the top and bottom position we get

W W
σ(x = 0) = − bzw. σ(x = h) = − .
2πat 4πat

Note: The stress at the top is twice as large as the stress at the bottom,
which is a inefficient use of material. This situation changes if the
weight of the thin-walled structure is included in the analysis.
 34 Elongation

P2.3 Problem 2.3 Determine the

elongation Δl of the conical
shaft (Young’s modulus E) F F
d D
under the application of a
tensile force F .

Solution The normal force N = F is constant, while the cross section

area A varies. With σ = N/A the elongation is computed by
l l l l
1 1 N dx F dx
Δl = ε dx = σ dx = = .
E E A E A(x)
0 0 0 0

To describe the change of the cross section area A(x) we start the x-
axis at the peak of the frustum. Using the intercept theorem and the
auxiliary variable a we obtain for the diameter
x x
δ(x) = d
a
and for the area d δ

π 2 π x2
A(x) = δ (x) = d2 2 . a
4 4 a
Introducing this in the relation for the elongation, then integration
provides (integration limits!):

a+l  a+l
F dx 4F a2 1 
Δl = = − .
E π d2 x 2 πE d2 x a
a 4 a2
With
a+l a d l
= ; a=
D d D 1− d
D
the elongation is
4F l
Δl = .
πEDd

4F l Fl
Test: For D = d (constant cross section) we obtain Δl = 2
=
. πEd EA
 Displacements 35

Problem 2.4 A homogeneous frustum of σ0 P2.4

a pyramid (Young’s modulus E) with a
square cross section is loaded on its top a x
surface by a stress σ0 .
Determine the displacement field u(x) of h
a cross section at position x.

b
Solution The normal force N = −σ0 a is constant. From the kinema-
2

tic relation ε = du/dx and Hooke’s law ε = σ/E = N/EA we obtain a

differential equation for the displacement u
du
EA(x) = −σ0 a2 .
dx
The area A(x) follows from the intercept theorem: a
x
A(x) = [a + (b − a) ]2 . x
h
Thus we have  
h
 b−a
b − a 2 du x+a
E a+ x = −σ0 a2 . h
h dx
Separation of variables yields
b
2 
u(x)
2 x
σ0 a dx σ0 a dξ
du = −   ; du = −   .
E b−a x+a 2 E b−a ξ+a 2
h u(0) 0 h
Using the substitution z = a + (b − a) ξ/h, dz = (b − a) dξ/h leads to
  b−a x+a
σ0 a 2 h 1  h σ0 a 2 h  1 1 
u(x) − u(0) = − − =− − .
E b−a z a E b−a a b−a x+a
h
The displacement u(0) of the top surface follows from the bounda-
ry condition that the displacement has to vanish on the bottom edge
x = h:
 
σ0 a 2 h 1 1 σ0 ah
u(h) = 0 ; u(0) = − = .
E b−a a b Eb
From this relation the displacement follows

σ0 a 2 h 1 1
u(x) = − + .
E b−a b b − ax + a
h
 36 Rotating bar

P2.5 Problem 2.5 The cross section of A0

A(x) A0 /2
a solid helicopter blade (density ρ,
ω
Young’s modulus E) is described by
the equation A(x) = A0 e−αx/l .
Determine the stress distributi- a
on σ(x), if the blade is rotating x
with a constant angular velocity ω. l
Compute the elongation Δl under the assumption a = 0.

Solution First, the sketched geometry A(l) = A0 /2 yields

A0 e−α = A0 /2 ; eα = 2 ; α = ln 2 = 0.693 .

The rotation causes a distributed load per unit length

n = ρω 2 xA(x) = ρω 2 A0 xe−αx/l .

The equilibrium condition N  = −n provides the normal force by inte-

gration
 !
ρ ω 2 A0 l 2 αx −αx/l −αx/l
N = − n dx = − − e − e + C .
α2 l
The integration constant C is determined by the boundary condition:

N (l) = 0 ; C = (1 + α)e−α = 0.847 .

Introducing the dimensionless coordinate ξ = x/l yields

ρ ω 2 A0 l 2 σ/(ρω 2l2 )
N (ξ) = [(1 + αξ)e−αξ − C] ,
α2 σmax
and for the stress distribution 1−C
α2
N ρ ω 2 l2
σ(ξ) = = [1 + αξ − Ceαξ ] .
A α2

The elongation is calculated from ξ0 1 ξ

 l  1  1
l ρ ω 2 l3 αξ 2 C
Δl= εdx = σdξ = 2 ξ+ − eαξ
0 E 0 α E 2 α
  0
ρ ω 2 l3 α C α C ρω 2 l3
= 1+ − e + = 0.258 .
Eα2 2 α α E

Note: Due to the varying cross section the maximum stress occurs at
the position ξ0 = −(ln C)/α = 0.24 and attains the maximum value
σmax = −(ρω 2 l2 ln C)/α2 = 0.347 ρω 2 l2 .
 Thermal stresses 37

Problem 2.6 A massive bar (weight W0 , P2.6

cross section area A, thermal expansion
coefficient αT ) is fixed at x = 0 and just E x
touches the ground in a stress-free man- W0
ner. h

Determine the stress distribution σ(x) in αT

the bar after a uniform heating by ΔT .
Which ΔT causes compression everywhe-
re in the bar?
Solution We investigate the ”two load cases”, weight und heating. The
weight causes a a normal force
h−x  x
N (x) = W (x) = W0 = W0 1 −
h h
x σ(x)
which is related to the stress distribution
N (x) W0  x W (x)
σ1 (x) = = 1−
A A h

The heating produces an additional strain, which is blocked by the sup-

port on the bottom. The relation
σ2 (x)
ε= + αT ΔT = 0
E
yields
σ2 (x) = −EαT ΔT .
Thus the total stress is computed by
W0  x
σ(x) = σ1 + σ2 = 1− − EαT ΔT .
A h
Due to the blocked temperature strain, there exists a compressive stress
at the end of the bar (x = h) at all times. As the stress distribution
is linear, the stress will be compressive everywhere, if compression is
present at the top edge. Thus the relation
W0
σ(x = 0) < 0 bzw. − EαT ΔT < 0
A
provides the necessary temperature difference
W0
ΔT > .
EAαT
 38 Thermal stresses
x
P2.7 Problem 2.7 An initially stress- E , αT
free fixed bar (cross section area
A) experiences a temperature
increase varying linearly in x.
l
Determine the stress and strain
ΔT
distribution.

ΔT0 ΔT (x) ΔT1

x
Solution The bar is supported in a statically indeterminate way. Thus
we use equilibrium, kinematics and Hooke’s law for the solution of the
problem. With n = 0 and σ = N/A these equations read
σ
σ = 0 , ε = u , ε= + αT ΔT (x)
E
with
x
ΔT (x) = ΔT0 + (ΔT1 − ΔT0 ) .
l
Combining the above relations renders the differential equation for the
displacements
αT
u = αT ΔT  = (ΔT1 − ΔT0 ) .
l
Integrating twice yields
αT
u = (ΔT1 − ΔT0 ) x + C1 ,
l
αT x2
u= (ΔT1 − ΔT0 ) + C1 x + C2 .
l 2
The two integration constants follow from the boundary conditions:
αT
u(0) = 0 ; C2 = 0 , u(l) = 0 ; C1 = − (ΔT1 − ΔT0 ) .
2
We obtain the displacement field
αT l  x2 x
u(x) = (ΔT1 − ΔT0 ) 2 −
2 l l

together with the (constant) stress

αT
σ = E(u − αT ΔT ) = − (ΔT1 + ΔT0 )E .
2
Note: With constant heating ΔT1 = ΔT0 the displacement u(x)
vanishes. In this situation the stress is σ = −αT ΔT0 E.
 Statically indeterminate problems 39

Problem 2.8 A bar with a constant P2.8

cross section A is fixed at both ends. C
The bar is made of two different a l−a
materials, that are joint together at
point C. F
a) What are the reaction forces, if an Steel Aluminium
external force F is applied at point
C ?
b) Determine the normal force that is caused by a pure heating by ΔT ?
Given: ESt /EAl = 3, αSt /αAl = 1/2 .

Solution We treat the system as two joint bars with constant normal
forces. NA NB
to a) F
St Al
equilibrium: −NA + NB = F , x

kinematics: ΔlSt + ΔlAl = 0 ,

NA a NB (l − a)
Hooke’s law:ΔlSt = , ΔlAl = .
ESt A EAl A
The 4 equations for the 4 unknowns (NA , NB , ΔlSt , ΔlAl ) yield with
the given numerical values
3(l − a) a
NA = −F , NB = F .
3l − 2a 3l − 2a

to b)
ΔT
NA NB
equilibrium: NA = NB = N ,
St Al
kinematics: ΔlSt + ΔlAl = 0 , x
Na
Hooke’s law:ΔlSt = + αSt ΔT a ,
ESt A
N (l − a)
ΔlAl = + αAl ΔT (l − a) .
EAl A
Solving the system of equations for the normal force N yields with the
given numerical values
2l − a
N =− ESt αSt A ΔT .
3l − 2a
 40 Static indeterminate

P2.9 Problem 2.9 Solve Problem 2.8 by superposition.

Solution to a) We choose the reaction force NB as statically redundant

quantity.
“0”-System “1”-System
NB = X
F St
St Al Al

u(0) u(1)
Hooke’s law provides
Fa X(l − a) Xa
u(0) = , u(1) = + .
ESt A EAl A ESt A
As the right edge is fixed compatibility requires
u(0) = u(1) .
This condition yields
Fa a
NB = X = =F .
ESt A
a + (l − a) E 3l − 2a
Al A

From equilibrium we have

3(l − a) NA NB
NA = NB − F = −F . F
3l − 2a St Al

to b) In the free body diagram we choose the normal force N as sta-

tically redundant quantity X. From Hooke’s law
Xa
uSt = + αSt ΔT a , N =X
ESt A
St Al
X(l − a)
uAl = + αAl ΔT (l − a)
EAl A uSt uAl
and the compatibility

uSt + uAl = 0

we obtain
αSt a + αAl (l − a) 2l − a
N =X=− =− ESt αSt A ΔT .
a + (l − a) 3l − 2a
ESt A EAl A
 problems 41

Problem 2.10 An elastically sup- c1 n c2 P2.10

ported bar (c1 = 2c2 = EA/2a)
is loaded by a constant axial EA
x
load n.
a
Compute the distribution of the
normal force N (x) in the bar.

Solution Using the free body dia- B C

gram with the forces B and C at
the ends of the bar, the equilibri-
um conditions can be formulated n
B C
B+C = na , N (x) = B−nx .

The elongation/shortening of the

springs is given by n
B
N(x)
B C
Δu1 = , Δu2 = .
c1 c2 x
The elongation of the bar is computed from
a a
N
ΔuSt = ε dx = dx .
EA
0 0

With N = B − nx we obtain
Ba na2
ΔuSt = − .
EA 2 EA
Finally, the kinematic relation
B Ba na2 C
Δu1 + ΔuSt = Δu2 ; + − =
c1 EA 2 EA c2
with C = −B + na and the given value for c1 and c2 yields
   
2a 4a a a 4a 9
B + + = na + ; B= na
EA EA EA 2 EA EA 14
and the distribution of the normal force follows 5
− na
9 14
N (x) = na − nx .
14
9 N
na
14
 42 Statically indeterminate problems

EAC C
P2.11 Problem 2.11 Determine the com-
pression ΔlC of a casing C of length
l, if the nut of screw S (lead h) is EAS S
turned by one revolution.
EAC 4 l
Given: = .
EAS 3

Solution After the revolution of the nut we cut the system of screw and
casing and introduce the statically
indeterminate force F between the
two parts.
The casing experiences a compres- X X
sion
Xl
ΔlC = . l
EAC
For the screw we obtain an elonga-
X
tion
Xl l−h
ΔlS = .
EAS

The length changes have to be adjusted in such a way that casing and
screw have the same length. Therefore compatibility can be written as

h = ΔlC + ΔlS .

Inserting the length changes yields the force

h 1
X=
l 1 + 1
EAC EAS
and the compression of the casing
Xl 1 1 3
ΔlC = =h =h = h.
EAC EA
1 + EA C 4
1+ 3 7
S

Note: As the axial rigidity of the casing is larger than the one of
the screw, the compression is only 3/7 of the lead. If equal axial
rigidities are present EAC = EAS , the length change of both parts
will be equal, i. e. ΔlC = ΔlS = h/2 .
 Displacements 43

Problem 2.12 A rigid quadratic √ P2.12

plate (weight W , edge length 2 a) 
1
is supported on 4 elastic posts. The EA 
2
2EA
posts are of equal length l, but pos- 

3
4
sess different axial rigidities. 3EA√
l 4EA
Determine the weight distribution 2a
on the 4 posts?
Determine the displacement f in
the middle of the plate.

Solution The system is statically indeterminate of degree one (a table

on 3 posts rests in a statically determinate way!).
II I
Equilibrium yields
S1 S2
↑: S1 + S2 + S3 + S4 = W ,

I: aS4 = aS1 , I W II
 S3 S4
II : aS2 = aS3 .
The displacement f in the middle is obtained from the average value of
the displacements ui (= length change of the posts) at opposite corners
(rigid plate). Accordingly the compatibility reads:
1 1
f= (u1 + u4 ) = (u2 + u3 ) . u1 u2
2 2
With Hooke’s law u3 f u4
Si l
ui =
EAi
and S1 = S4 , S2 = S3 we obtain as intermediate result
S1 l S1 l S2 l S2 l 5 5
+ = + ; S1 = S2 .
EA 4EA 2EA 3EA 4 6
Inserting this into the first equilibrium condition yields
3 3 1 3
S1 + S1 + S1 +S1 = G ; S1 = S4 = G, S2 = S3 = G.
2 2 5 10
form which the displacement follows:
 
1 S1 l S1 l 1 Gl
f= + = .
2 EA 4EA 8 EA
 44 Composite material

P2.13 Problem 2.13 A column of steel rein- F

forced concrete is loaded by a tensile
force F . C C
What are the stresses in the concre-
h
te and the steel as well as the height
change Δh of the column, if we assu-
me
a) a perfect bonding between steel C−C
and concrete?
b) the concrete is cracked and does
not carry any load?
EC , AC
Given: ESt /EC = 6, ASt /AC = 1/9. ESt , ASt

Solution to a) We consider the composite as a system of two ”bars”

of different materials, which experience under load F the same length
change Δl. With this the basic equations of the system are:
F
equilibrium: NSt + NC = F ,
kinematics: ΔhSt = ΔhC = Δh ,
NSt h NC h
Hooke’s law:ΔhSt = , ΔhC = . NC
EASt EAC NSt

Solution of the system of equation yields –

with the stiffness ratio EAC /EASt = 3/2
– the normal forces
EAC
1 2 EASt 3
NSt = F = F, NC = F = F
EAC
1 + EA 5 EAC
1 + EA 5
St St

and the height change

Fh Fh 1 2 Fl
Δh = i. e. Δh = = .
EASt + EAC EASt 1 + EAC 5 EASt
EASt
The stresses result from A = AC + ASt and ASt = A/10 and AC =
9A/10
NSt F NC 2 F
σSt = =4 , σC = = .
ASt A AC 3 A

to b) If only the steel carries load, we will obtain with NSt = F

F F Fh
σSt = = 10 , Δh = .
ASt A EASt
 Composite material 45

Problem 2.14 A laminated bar made P2.14

of bonded layers of two different ma-
terials (respective axial rigidities EA1 , 
2 
1
EA2 and coefficients of thermal expan-
sion αT 1 , αT 2 ) is to be replaced by a
bar made of a homogeneous material.
Determine EA and αT such that the
homogeneous bar experiences the sa-
me elongation as the laminated bar
under application of a force and a tem-
perature change ? EA, αT

Solution For the laminated bar, subjected to a force F and a tempe-

rature increase ΔT , the basic equation yield

equilibrium: N1 + N2 = F , EA1 , αT 1

kinematics: Δl1 = Δl2 = Δllam , N1

N1 l
Hooke’s law:Δl1 = + αT 1 ΔT l , F
EA1 N2
N2 l EA2 , αT 2
Δl2 = + αT 2 ΔT l .
EA2

This yields
Fl EA1 αT 1 + EA2 αT 2
Δllam = + ΔT l .
EA1 + EA2 EA1 + EA2
For a homogeneous bar under identical loading conditions, we have
Fl
Δlhom = + αT ΔT l .
EA
The length changes Δllam and Δlhom agree for arbitrary F and ΔT
only, if
EA1 αT 1 + EA2 αT 2
EA = EA1 + EA2 , αT = .
EA1 + EA2
 46 Forces in bars

P2.15 Problem 2.15 In the depicted support a a

construction for the rigid body B the
lower support bar is too short by the
length δ. In order to assemble the struc- aluminium
ture a force Fa is applied, such that the lAl
end of the bar just touches the ground. Fa
After assembly the force Fa is removed.
The diameters of all bars di are identical. B
a) Compute the required assembly force
steel lSt
Fa .
b) Determine the displacement vB of the
body and the forces in the bars after as- δ
sembly.
Given: lAl = 1 m, dAl = 2 mm, EAl = 0.7 · 105 MPa, lSt = 1.5 m,
dSt = 2 mm, ESt = 2.1 · 105 MPa, δ = 5 mm .

Solution to a) Each aluminium bar carries half of the assembly force

(equilibrium) and elongates by the amount delta δ. This yields
Fa SAl lAl Fa lAl
SAl = , ΔlAl = = = δ,
2 EAAl 2EAAl
δ 5
; Fa = 2 EAAl =2· · 0, 7 · 105 · π · 12 = 2200 N .
lAl 1000

to b) After removal of the force Fa new forces SAl and SSt are present.
This leads to the equilibrium condition

SSt = 2SAl , SAl SAl

Hooke’s law
SAl lAl SSt lSt SSt
ΔlAl = , ΔlSt =
EAAl EASt
and the compatibility condition

ΔlAl + ΔlSt = δ .

Solving the 4 equations yields

δ EAAl 5 0, 7 · 105 · π · 12
SAl = = = 550 N ,
lAl 1 + 2 lSt EAAl 1000 1 + 2 · 3 · 1
lAl EASt 2 3
SAl lAl
SSt = 2SAl = 1100 N , vK = ΔlAl = = 2.5 mm .
EAAl
 Forces in bars 47

Problem 2.16 Two rigid beams P2.16

are connected by two elastic a a
bars. The first beams is fixed
at point A, while the second is
simply supported at point B. EA EA, αT a
Bar 2 is heated by a tempera- A 1 2
ture ΔT .
Compute the forces in the two B
bars. a a

Solution We cut the system S1 S2

and use the following free body
diagram to formulate the equili- S1 S2
brium conditions

B : 2aS1 + aS2 = 0 ,

Hooke’s law
S1 S2
S1 a
Δl1 = ,
EA
S1 S2

S2 a B
Δl2 = + αT ΔT · a
EA
and the compatibility condition

Δl1 = 2Δl2 . B
Δl1
Δl2
Solving for the unknown forces
in the bars yields

2 4
S1 = EA αT ΔT , S2 = − EA αT ΔT .
5 5

Note: In the heated bar compressive forces are generated due

to the constrained deformations.
 48 Displacements

P2.17 Problem 2.17 In the depicted two bar

system both bars have the same axial
rigidity EA.
Determine the displacement of point C 2
where the load is applied.
60◦

C
1
F
l
Solution From equilibrium we have
S2
2√
↑ : S2 sin 60◦ = F ; S2 = 3F,
3 C
◦ 1√ S1
→: −S1 − S2 cos 60 = 0 ; S1 = − 3F. F
3

Thus the elongation and shrinking of the bars follow as

2 √3 l F
√ √
Δl2 =
S2 l2
= 3 cos 60◦ = 4 3 F l , Δl = S1 l1 = − 3 F l .
1
EA EA 3 EA EA 3 EA
To determine the displacements of 2
point C we construct the displace- 1 Δl1
ment diagram. In this diagram the C
length changes are introduced. As the Δl2
length changes are small Δli
l they
are not drawn to the scale. In this 30◦
v
example Δl1 is a shrinkage (to the
left) and Δl2 an elongation. Consi-
dering that the bars can only rotate
around the hinge points we introdu- u
ce the right angles and read off the C
displacement diagram:
√
3 Fl
u = |Δl1 | = ,
3 EA
√ √
Δl2 u 4 3 Fl 1 3 Fl 1 Fl
v= + = √ + √ =3 .
cos 30◦ tan 60◦ 3 EA 1 3 3 EA 3 EA
2
 Displacements 49

Problem 2.18 A rigid weightless D 1 P2.18

triangle is supported by 3 bars
with the axial rigidity EA. The
triangle is loaded in point B by a
the force F . F
2
a) Determine the forces Si in the
3 bars and their elongations Δli . B C
b) Compute the displacement of 3
a
point C.

a a

Solution to a) The system is statically determinately supported. The

forces in the bars follow immediately from the equilibrium conditions:
E

C : aS1 = aF ; S1 = F , S1

E : aS2 = 0 ; S2 = 0 , F
√ S2
↑ : S3 sin 45◦ + F = 0 ; S3 = − 2 F .
C
S3
Related to these forces are the following elongations
S1 l1 Fa
Δl1 = = , Δl2 = 0 ,
EA EA
√ √
S3 l3 2F· 2a Fa
Δl3 = =− = −2 .
EA EA EA

to b) The displacement of point C is

C 2
sketched in the displacement diagram.
As bar 2 experiences no force and thus Δl3
no length change, the horizontal dis-
placement vanishes. From the displace- vC
ment diagram we obtain for the vertical 3
displacement vC :
√ √ Fa
vC = 2 |Δl3 | = 2 2 . C
EA
 50 Deformation

P2.19 Problem 2.19 In the depicted

truss the members have the axi- EA2
al rigidities EA1 , EA2 and the αT2
coefficients of thermal expansion
EA1 EA1 h
αT 1 , αT 2 . β β
αT1 αT1
Determine the axial forces in the
trusses, if the system is heated
by ΔT ? C

Solution As the system is statically indeterminate, we have to use all

basic equations. We start S2
with the equilibrium
β β
2S1 cos β + S2 = 0
S1 S1
and continue with Hooke’s law
C
S1 l1
Δl1 = + l1 αT 1 ΔT ,
EA1
S2 l2
Δl2 = + l2 αT 2 ΔT ,
EA2
where
2
h 1 1
l1 = , l2 = h .
cos β β Cβ

The compatibility of the displacements is according Δl1

to the the displacement diagram:
Δl2
Δl1 = Δl2 cos β .
C

Solving the 4 equations for the two truss forces and the two elongations
yields

αT 2 cos2 β − αT 1
S1 = EA1 ΔT , S2 = −2 cos β S1 .
1 + 2 cos3 β EA
EA2
1


Note: For cos β = αT 1 /αT 2 we obtain S1 = S2 = 0: the trusses can
than expand without causing forces! (special case αT 1 = αT 2
; β = 0)
 of systems of trusses 51
δ 3
Problem 2.20 Truss member 3 P2.20
was produced too short to be as- D D
sembled between two identical C
trusses. 2 2
1 1 a
a) Determine the required as-
sembly force D?
b) Calculate the normal force S3
after the assembly (D = 0)? a a a
√
Given: EA1 = EA3 = EA, EA2 = 2 EA.

Solution to a) The force D has to move point C by δ/2 in horizontal

direction during assembly. From equilibrium

→: S2 cos 45◦ = D , C D
◦
↑: S1 = S2 cos 45 , S1 S2

kinematics (S1 was positively introduced

as compressive force!!) with the prescribed
displacement
√ δ
uC = Δl1 + Δl2 2 , uC = , Δl2
2
C
and Hooke’s law
√ 2
S1 a S2 a 2 Δl1
Δl1 = , Δl2 = √
EA 2 EA
we obtain C uC
1
1 δ
D= EA .
6 a
to b) Equlibrium, kinematics and Hooke’s law are as in a), but D has
to be replaced by S3 . With the known compatibility condition
S3 a
2uC + Δl3 = δ and Δl3 = S3
EA C
it follows S1 S2
1 δ
S3 = EA .
7 a
 52 Statically indeterminate

P2.21 Problem 2.21 A centric F

D
loaded rigid beam is sup- B
ported by 4 elastic bars of 1 2
equal axial rigidity EA. 3 a 4
30◦ 30◦
Determine the forces in the
bars? l l

Solution a) First, we solve the statically indeterminate system by

applying all basic equations simultaneously. Using equilibrium
F
B
S1 S2 S3 S4
→: S1 = S2 ,

↑ : (S1 + S2 ) sin 30◦ + S3 + S4 = F ,


B : lS3 + 2lS4 = lF ,
Hooke’s laws
S1 2a
Δl1 = Δl2 = ,
EA
S3 a S4 a
Δl3 = , Δl4 =
EA EA
and the geometry of the deformation
B
Δl1 Δl2 B D
1 2 v Δl3 Δl4
60◦ v
B

D
B

Δl1 1
v= Δl3 = (v + Δl4 )
cos 60◦ 2
we obtain as solution
2 5
S1 = S2 = S4 = F , S3 = F.
9 9
 truss system 53

b) Now, we solve the problem by superposition. The system is divided

into two statically determinate basic systems:
“0”-System “1”-System
F

(0) (1)
(0) (0) S4 (1)
S1 (1)
S2 X S4
S1 S2

Equilibrium yields
(0) (0) (0) F (1) (1) (1) X
S1 = S2 = S4 = , S1 = S2 = S4 = .
2 2

B C D
(0) (1) (1)
vB (0)
vC (0)
vD
(1)
vB vC vD

B C D
From geometry and Hooke’s laws it follows
(0)
(0) Δl1 F 2a (1) X 2a
vB = = , vB = ,
cos 60◦ EA EA
(0) Fa
(0) (1) Xa
vD = Δl4 = , vD = ,
2EA 2EA
(0) 1  (0) (0)
 5 Fa (1) 5 Xa
vC = vB + vD = , vC = ,
2 4 EA 4 EA
(1) Xa
Δl3 = .
EA
The kinematic compatibility requires the total displacement of point C
to coincide with the shortening of truss 3:
(0) (1) (1)
vC − vC = Δl3 .

Inserting the displacements yields

5
X = S3 = F
9
and
(0) (1) 2 (0) (1) 2
S1 = S1 − S1 = F, S4 = S4 − S4 = F.
9 9
 54 Truss system

P2.22 Problem 2.22 The depicted truss

system (axial rigidity EA) is loaded
by the external force F and additio- F
nally pinned at point C. C
a) Determine the reaction force at
point C. α
b) Calculate the vertical displace- l 2 1
ment of point C.

Solution to a) Using equilibrium F

↓: F + S2 + S1 cos α = 0 , C
S1
→: C + S1 sin α = 0 , S2
Hooke’s laws C
S1 l1 S2 l2
Δl1 = , Δl2 = ,
EA EA
Δl2
and kinematics
α
Δl1 = Δl2 cos α Δl1
C
1
yields 2

sin α cos2 α cos2 α 1

C= F, S1 = − F, S2 = − F.
1 + cos3 α 1 + cos3 α 1 + cos3 α

to b) Knowing S2 the vertical displacement of point C follows as

S2 l 1 Fl
vC = Δl2 = =− .
EA 1 + cos3 α EA

In contrast to the displacement diagram, in which tensile forces (elon-

gations) are assumed, compressive force occur in the system. Due to
shortening point C moves in downwards direction.
Test: α = π/2 yields S1 = 0 and S2 = −F .
α = 0 yields S1 = S2 = −F/2.
 Plasticity 55

Problem 2.23 A rigid beam is sup- F P2.23

ported by three bars of elastic-ideal-
plastic material.
el
a) At what force Fmax and at which E, A, σY l/2
location in the bars is the yield stress
σY reached at first?
pl
b) At what force Fmax occurs plastic l/2
yielding in all bars of the system?

Solution to a) The system is statically indeterminate. Using symmetry

equilibrium provides
F
2S1 + S2 = F

Kinematics is expressed by
S1 S2 S3 = S1
Δl1 = Δl2 .

Until plastic yielding Hooke’s law can be used

S1 l S2 l
Δl1 = , Δl2 = .
EA 2EA
The solution provides forces and stresses in the bars
F F F F
S1 = , S2 = ; σ1 = , σ2 = .
4 2 4A 2A
As the stress in bar 2 is the highest, the yield limit is reached first there
during load increase:
el
σ2 = σY ; Fmax = 2 σY A .

el
to b) For a load increase above Fmax bar 1 and bar 3 still respond
elastically, while bar two undergoes plastic deformation: σ2 = σY . Thus
with Si = σi A it follows from equilibrium F
2σ1 A + σY A = F
F σY
; σ1 = − .
2A 2 S2 = σY A S = σ A
S1 = σ1 A 3 1

All bars are undego plastic deformation if

F σY pl
σ1 = σY ; − = σY ; Fmax = 3 σY A .
2A 2
 56 Plasticity

P2.24 Problem 2.24 In the depicted symme-

tric system all bars are made of the C F
same elastic-ideal-plastic material, but
have different cross sections. E, A, E, 2A,
σY 1 2 σY h
el
a) At what force Fmax and at which
45◦ 45◦
location in the bars is the yield stress
σY reached at first? Determine the
reaction force at C for this situation.
pl
b) Determine the force Fmax when both bars deform plastically?
el
c) Calculate the displacement umax of point C for case a)?
el
Solution to a) Until reaching the force Fmax the system responds
elastically. Therefore the equilibrium conditions are given by
√ √ √ √
2 2 2 2 C
→: S1 − S2 = F , ↑ : S1 + S2 = C ,
2 2 2 2 F
together with Hooke’s law S1 S2
√ √
S1 2 h S2 2 h
Δl1 = , Δl2 = Δl1
EA 2EA
and the kinematics (bar 2 will shorten) C
u C
Δl2
Δl1 = −Δl2 .
1 2
From the above relation we obtain
√ √
2 2 2 F 2F h
S1 = F , S2 = − F, C = − , Δl1 = −Δl2 =
3 √ 3 3 √ 3EA
S1 2 F S2 2 F
; σ1 = = , σ2 = =− .
A 3 A 2A 3 A
The absolute value of the stresses is identical in both bars. Yielding
will occur if
√
el 3√ el 2
σ1 = |σ2 | = σY ; Fmax = 2 σY A, ; Cmax =− σY A.
2 2
el
to b) As at Fmax plastic yielding occurs in both bars, we have
el pl
Fmax = Fmax .
to c) Until the yield limit is reached the displacement of C is given by
√
√ 2 2 Fh σY
u = 2 Δl1 = , ; uel el
max = u(Fmax ) = 2 h.
3 EA E
Chapter 3
Bending of Beams
3
 58 Ordinary bending

Beam = straight structural element, length l large compared to

dimensions of the cross section, perpendicular loads.

q(x)
F

y
x
l x
z z

3.1
3.1 Ordinary bending
nomenclature and assumptions:
• x = axis of cross section centroids; y, z = principal axis of the se-
cond moment of area (moment of inertia).
• kinematic assumption: plane cross sections remain plane

w = w(x) , u = z ψ(x) ,

w = displacement in z-direction,
u = displacement in x-direction,
ψ = rotation angle of cross section.
• stress resultants: M

y, ψ x, u
V = Vz = shear force,
V
M = My = bending moment. z, w

Normal stress

neutral axis
M x
σ(z) = z
I zmax
z σmax

I = moment of inertia with respect to y-axis,

z = distance to neutral axis (= axis of centroids).
The largest absolute value of the stress occurs in the extreme fibre:

M I
σmax = , W = = section modulus.
W |zmax |
 3.1 Ordinary bending 59

Shear stress
a) thin-walled, open profile
τ (s)
t(s)
V S(s)
τ (s) = , y
I t(s)
z
S(s) = static moment of A∗ with regard to A∗
y-axis,
t(s) = thickness of profile at position s. s

b) compact cross section

b
V S(z) τ (z)
τ (z) = .
I b(z) y x
h
special case: rectangle
A∗
 
3Q 4z 2 z z
τ = 1− 2 .
2A h

3Q Q
Note: τmax = τ (z = 0) = is 50% larger than τmean = .
2 bh bh

Shear center M of singly symmetrical cross sections.

moment of V with regard to 0
= moment of distributed shear
stresses with regard to 0:
M rM
0

r⊥
rM Q = τ (s) r⊥ (s) t(s) ds V τ

ds

Position of centriod C und shear center M for selected profiles:

full circle
M semi circle with slit

C M C M C
M =C M =C C
C M
r
M
0, 273 r r
60 Differential equation of the deflection curve

Basic equations

dV dM
equilibrium conditions = −q , =V ,
dx dx
Hooke’s law, kinematics M = EIψ 

V = GAS (ψ + w ) ,

EI = bending stiffness,
GAS = shear stiffness,
AS = κA = shear area (κ = shear correction factor).
Rigid with respect to shear (Bernoulli beam): If we additionally assume,
that cross sections perpendicular to the undeformed beam axis remain
perpendicular to the deflection curve during the deformation, it follows
from Hooke’s law for the shear force (GAS → ∞)

ψ = −w .

Differential equation of the deflection curve for the Bernoulli

beam: Inserting into Hooke’s law for M yields

EIw = −M .

This leads with the equilibrium conditions to

(EIw ) = q ,

or for EI = const

EIwIV = q .

Temperature induced moment

A linearly, across the height h, varying temperature field (= tempera-
ture gradient) can be treated by a temperature moment : Tt
Tb − Tt x
MT = EIαT , h
h

αT = coefficient of thermal expansion. z Tb

In this case, the differential equation for the deflection curve yields

EIw = −(M + MT ) .
 Boundary conditions and solution methods 61

Table of boundary conditions

support w w M V

0 = 0 0 = 0

0 0 = 0 = 0

= 0 = 0 0 0
free end

= 0 0 = 0 0

Solution methods

1. For continuous functions of q(x) or M (x), four or two times integra-

tion of the corresponding differential equation yields the deflection
curve w(x). The four or two integration constants are obtained by
the boundary conditions (see table of boundary conditions).
2. For several regions (discontinuities in the loads, deformation, con-
centrated forces or concentrated moments), the integration has to be
performed piecewise. The integration constants are determined from
boundary and matching (continuity) conditions. The computation
can by simplified by using the Macauley bracket (see Engineering
Mechanics 1):
⎧
⎨0 f ür x < a ,
n
<x−a> =
⎩(x − a)n f ür x > a .

3. Statically indeterminate problems can be solved by using superposi-

tion of known deflections and rotations. For this purpose, deflection
and rotations of the most frequent load cases and support situations
can be found in the table on page 62/63.
4. Statically indeterminate problems can also be solved by using the
principle of virtual forces (energy method) (see chapter 5).
62 Table of end rotations

 
no. load case EIwA EIwB
x F
a b
F l2 F l2
1 (β − β 3 ) − (α − α3 )
6 6
A l B
x
q0
q0 l3 q0 l3
2 −
24 24
A l B

x
qB 7 1
3 qB l3 − qB l3
360 45
A l B
x M0
M0 l M0 l
4 (3β 2 − 1) (3α2 − 1)
A a b B 6 6
l
x F
F a2
5 B 0
A a 2
l
x
q0 q0 l3
6 0
B 6
A l
x
qA
qA l3
7 0
B 24
A l
x M0
8 0 M0 l
A l B

explanations: x a b d 1 d
ξ= , α= , β= , ( ) =
" ()= (),
l l l dx l dξ
 and deflections 63

EIw(x) EIwmax

F l3 F l3
[βξ(1 − β 2 − ξ 2 )+ < ξ − α >3 ]
6 48
for α = β = 1/2

q0 l4 5
(ξ − 2ξ 3 + ξ 4 ) q0 l4
24 384

qB l4
(7ξ − 10ξ 3 + 3ξ 5 ) see problem 3.13
360

M0 l2 M0 l2 √
[ξ(3β 2 − 1) + ξ 3 − 3 < ξ − α >2 ] 3
6 27
for a = 0

F l3 2 F l3
[3ξ α − ξ 3 + < ξ − α >3 ]
6 3
for a = l

q0 l4 q0 l4
(6ξ 2 − 4ξ 3 + ξ 4 )
24 8

qA l4 qA l4
(10ξ 2 − 10ξ 3 + 5ξ 4 − ξ 5 )
120 30

x2 l2
M0 M0
2 2

< ξ − α >n =
" Macauley bracket
 64 Biaxial bending
3.2
3.2 Biaxial bending
x = axis of centroids,
y, z = arbitrary ortho-
gonal axis.

y
x shear forces Vy , Vz
Vy
z Vz

x
and
w
z

x bending moments My , Mz
My (positive when positive right-
v
y Mz hand screw at positive intersec-
tion).

Differential equation of the deflection for shear rigid beams:

1
Ew = (−My Iz + Mz Iyz )
Δ
1
Ev  = (Mz Iy − My Iyz )
Δ

Δ = Iy Iz − Iyz
2
,
Iy , Iz , Iyz = second order area moments.

Normal stress

1
σ= [(My Iz − Mz Iyz )z − (Mz Iy − My Iyz )y] .
Δ

Special case: If y, z are principal axis (Iyz = 0), then

My Mz
EIy w = −My , EIz v  = Mz , σ= z− y.
Iy Iz
 Normal stress 65

Problem 3.1 A cantilever beam with the P3.1

depicted cross section (constant wall thick-
ness t, t
a is subjected to a concentra-
ted force F at one end. F
Determine the maximum stress in the cross
section at the support.
F 2a
2a t

40a a a

Solution The distance of the centroid ξC from the top surface is ob-
tained from the sub-areas by using t
a
II III
I
# $% & # $% &
ξC
Σξi Ai 2 (2at · a) +2 (at · 2a) 8a2 t
ξC = = = II C
ΣAi 2at +2 · %&#$
%&#$ 2at +2 · %&#$
at 8at
ξ
I II III
=a . III
The second moment of area with regard to the
y-axis is computed by using the parallel-axis
theorem. I
II a
I #' $% (& III C II
# $% & # $% *&
)
t(2a)3 16 3 y
Iy = a · 2at +2
2
+2 a · at =
2
ta , a
12 3
z III
Thus we obtain for the section modulus
16 3
Iy ta 16 2
W = = 3 = ta .
zmax a 3

The stress in the cross section at the support is calculated using the
bending moment at this position

M = −40 aF
to be
|M | 40aF 30 F
σmax = = =
W 16 2 4 at
ta
3
(the upper fibre is in tension, the lower under compression).
 66 Computation of

P3.2 Problem 3.2 A cantilever 

2

beam with the sketched F

cross section is loaded by
the force F at point x . 
1

2a
Determine the normal a a
stresses at point y at the 
2
support. 2a

2a


1

Solution As the neutral axis is passing trough the centroids of the

cross sections, we first determine the position of the centroid:
I II
# $% & # $% *&
) ξ
ΣAi ξi 8a2 · a +2 2a2 · 3a 5 II
ξC = = = a.
ΣAi 8a2 + 4a2 3
C I ξC
The second moment of area with respect to the
y-axis is computed by summing up the contri-
butions of the sub-areas:
+  2 ,
4a(2a)3 2 2 II
Iy = + a 8a +
12 3 5
+  2 , y C I a
a(2a)3 4 44 4 3
2
+2 + a 2a = a .
12 3 3 z
The following stress resultants are present in the cross section at the
support 5
N = −F and My = − aF .
3
The associated stresses are (σN due to normal force, σM due to bending
moment)
N F My 5 aF z 5 Fz
σN = =− and σM = z=− =− .
A 12a2 Iy 3 44 4 44 a3
a
3
7
At point y superposition with z2 = − a yields
3
F 5 F 7 2 F
σ = σN + σM (z2 ) = − + a= .
12a2 44 a3 3 11 a2
 normal stresses 67

Problem 3.3 The column with F P3.3

a star-shaped cross section
(t
a) is loaded by a force F , t
applied off center.
Determine 60◦
a) the maximum absolute value l b
a/2
of the stress, F
b) the maximal value of b such
that nowhere in the cross secti-
on tensile stresses occur.

I
Solution to a) Due to the load and the
symmetry of the cross section it is convenient III
to introduce the following y, z-coordinate
system. This yields y II
ta3
IyI = .
12 η
z
The second moments of area for the sub-areas
II and III with respect to the y-axis are −ϕ S
determined by the transformation equations y II
at3 ta3
Iη = , Iζ = , Iηζ = 0 , ϕ = −30◦ . ξ
12 12
z
Using t
a we obtain
Iη + Iζ Iη − Iζ ta3 ta3 1 ta3
IyII = IyIII = + cos 2ϕ+Iηζ sin 2ϕ = − = .
2 2 24 24 2 48
This leads to
ta3 ta3 ta3
Iy = IyI + 2IyII = +2 = .
12 48 8
Together with the stress resultants N = −F and My = −bF it follows
N My F 8bF
σ= + z=− − 3 z.
A Iy 3at ta
The largest stress (compression) occurs at z = a/2:
 
F 1 b
σmax = − +4 .
at 3 a

to b) Tensile stress occurs first at z = −a/2:

a F Fb a
σ(− ) = 0 ; − +4 2 = 0 ; b= .
2 3at ta 12
 68 Inhomogeneous cross section

P3.4 Problem 3.4 A column is clam- Fv

ped at the bottom and is carrying h h
a vertical load Fv at the center of 6 6
Fh
the top cross section and a horizon- 
2 
1 
2
tal load Fh in the middle of edge b
b. The column is made of 3 layers z

4E
4E
E
with different Young’s moduli. l
x y
Determine the normal stress distri- h
bution in the cross section at the
z
clamping.

Solution We consider the different load cases independently.

to a) With the vertical load Fv , we obtain from z

equilibrium σ1 A1 + σ2 A2 = −Fv , ε
Hooke’s law σi = E i ε i
and geometry ε1 = ε2 = ε σ2 σ1
the strain
2 1 Fv
E1 ε1 A1 + E2 ε2 A2 = Eε bh + 4Eε bh = −Fv ; ε=−
3 3 2Ebh
and the associated stresses
Fv Fv
σ1 = − , σ2 = −2 .
2bh bh
to b) Fh causes a moment MS = −Fh l at the support. Then geometry
(assume: cross sections remain plane)
z
u=ψ·z ; ε = ψ · z , ε(z)
Hooke’s law σ(z) = E(z)ε(z)
and
- - 2
h/3 - 2
h/2
σ(z)
M = σzdA = 2bψ  [E1 z dz + E2 z dz]
0 h/3
1 h 4 h h 7
= 2bψ  E[ ( )3 + (( )3 − ( )3 )] = bψ  Eh3
3 3 3 2 3 27
lead to (using M = MS )
27 Fh l
ψ = − .
7 Ebh3
Finally, the stresses follow as
27 M h 9Fh l
σ1 = E 1 ψ  z = E z ; σ1 ( ) = − ,
7 Ebh3 3 7bh2
27 M h 54Fh l
σ2 = E2 ψ  z = 4E z ; σ2 ( ) = − .
7 Ebh3 2 7bh2
 Shear stresses 69
F
Problem 3.5 A wooden cantilever P3.5
can be assembled from 3 beams
(dimensions of the cross section 
1 
2
b = a and h = 2a) in different b
b
ways.
y b h y b
What is the maximal force F for b
b
the two variants x and y , if the
z z
maximal allowed shear stress in the
bonding layer is given by τallow ? h h

Solution With V = F the shear stress in the bonding layer becomes

in general (z = zl )
F S(zl )
τ (zl ) = .
I b(zl )
This yields with τ (zl ) = τallow the maximal load Fmax
τallow I b(zl )
Fmax = .
S(zl )
For variant x we obtain
bh3 hb3  h b 2 ! zl 1
I= +2 + + bh = 10 a4 , (h + b)
12 12 2 2 2
b(zl ) = b = a , z
 A∗
1
S(zl ) = zdA = (h + b)bh = 3 a3 .
A∗ 2
which leads to the force
10a4 · a 10
F1max = τallow = τallow a2 .
3a3 3
Analogously we obtain for variant y
h(3b)3 9
I= = a4 , b(zl ) = h = 2a , zl b
 12 2
S(zl )= zdA = b · bh = 2a 3
A∗ z
A∗

and the force

9a4 · 2a 9
F2max = τallow = τallow a2 .
2 · 2a3 2
Note: The shear stresses in the cross section at z = zl and in the
corresponding perpendicular bonding interface are equal (associated
shear stresses!).
 70 Shear stresses
b
P3.6 Problem 3.6 Determine the shear stress
due to an applied shear resultant force
V in the depicted thin- walled I-profile. V

t2 h
y

t1 z

Solution The shear stresses are com-

puted from

V S(s)
τ =
I t(s)

Thus we need to determine the second

moment of area I with regard to the
y-axis. With t1
b and t2
h we
obtain
 2
h h3
I= I1 + I2 = 2 t1 b + t2
2 12
y
h2 h2
= (t2 h + 6t1 b) = (A1 + 6A2 ) . A∗
12 12
The static moment of sub-area A∗ for a z
position s in the lower sub-area is given s
by
h
S(s) = t1 s
2

and for a position z in the second sub-

area it follows y
 
h z h
  + z h  h + z /2
S(z)= 2
h b
t1 + 2 − z t2
−z 2
2 2 2 2 2

h t2
= A1 + (h2 − 4z 2 ) .
2 8
 in beams 71

These relations yield the shear stress in the upper sub-area

h A2
V t1 s V s
2 A1
τ1 (s) = =
h2 A2
1+
A2 h
(A2 + 6A1 )t1
12 6A1

and in the second sub-area

+  2 ,
  A2 2z
h t2 2 1+ 1−
V A1 + (h − 4z ) 2
4A h
2 8 V 1
τ2 (z) = 2
= .
h A2 A2
(A2 + 6A1 ) t2 1+
12 6A1

The maximum shear stress occurs at the center of the profile,

A2
1+
V 4A1
τ2 max = τ2 (z = 0) = ,
A2 A2
1+
6A1

it depends on the area ratio A2 /A1 . τ τ2 max

The maximum shear stress in the first
sub-area is given by

A2
V A1 b τ1 max
τ1 max = τG (s = b/2) = .
A2 A2 2h
1+
6A1

15 V
For example A1 = A2 and b = h yields τ2 max = at the center
14 A2
6 V
and τ1 max = . For this situation the smallest value in the vertical
14 A2
sub-area
V 1 12 V
τ2 min = τ2 (z = h/2) = = ,
A2 A2 14 A2
1+
6A1
is only 20% smaller than τ2 max . As a rough estimate we can use the
average shear stress τave = V /A2 in the central sub-area.
 72 Stresses

P3.7 Problem 3.7 A composite beam b

consists of an upper concrete slab
h
and a steel I beam. The structure
is loaded by a bending moment M .
2h
a) Determine the width b of the M
concrete slab, such that compres-
sive stresses occur only in the h
concrete part, while the tension is
present in the steel part. Given : M = 1000 kNm
EC = 3.5 · 104 N/mm2
b) For this case compute the
ES = 2.1 · 105 N/mm2
stresses in the extreme fibres of the
two materials. h = 40 cm
AS = h2 / 6
IS = h4 / 18

Solution to a) For the case

that compression occurs only in
the concrete and tension only in
the steel sub-area the strain in y
z
the bonding layer has do be zero
(=neutral fibre). With the chosen
coordinate system we have
zC CC
ε = az , zS
CS

where a is not yet determined. The

stresses in steel and concrete are

σS = E S ε = a E S z , σC = E C ε = a E C z .

As the beam is loaded only by a bending moment, the normal force N

has to vanish:
   
N= σS dA + σC dA = 0 ; ES z dA + EC z dA = 0 .
AS AC AS AC

With
 
h2 h3 h h2 b
z dA = zS AS = h = , z dA = zC AC = − hb = −
6 6 2 2
AS AC
 in composite beam 73

and ES /EC = 6 the required width b is obtained:

h3 h2 b
6 − =0 ; b = 2h = 80 cm .
6 2

to b) The unknown factor a follows from the prescribed bending mo-

ment.
From the definitions
   
M= z σS dA + z σC dA = a ES z 2 dA + a EC z 2 dA .
AS AC AS AC

and the evaluation of the integrals


h4 h4 2
z 2 dA= IS + h2 AS = + = h4
18 6 9
AS

bh3 2
z 2 dA= = h4
3 3
AC

it follows
 
ah4 EC ES M
M= 2 + 6 = 2ah4 EC ; a= .
9 EC 2h4 EC
With this result the stresses in the steel and concrete are

ES M M M
σS = z=3 4 z, σC = z.
2EC h4 h 2h4
For the top extreme fibre in concrete (z t = −h) and the bottom extrem
fibre in steel (z b = 2h) we obtain

−7.8125 N/mm2
t M
σC =− = −7.8125 N/mm2 ,
2h3

M
σSb = 6 = 93.75 N/mm2 ,
h3

93.75 N/mm2
 74 Shear stresses

P3.8 Problem 3.8 Determine the shear 2a V

stresses due to a shear force V t
for the depicted thin-walled beam
cross section (t
a). y
C
2a t
z

a t

Solution At first we compute the cross section area, the location of

the centroid and the second moment of area: I
ȳ s1
A = 4at + 2 · 2at + 2at = 10 at ,
4 b s2 y C
bA = 2a · 2at + 2a · 2at ; b= a,
5
t(2a)3 40 3 II
Iȳ = (2a)2 2at + 2 = ta , z
3 3
104 3
I = Iy = Iȳ − b2 A = ta . III
15
Due to symmetry of the cross section the 3 s
shear stress is symmetric to the z-axis.
Thus only half of the cross section has to be considered. With the
coordinantes s1 to s3 we obtain for the static moments in the sub-areas
I to III
4
SI = b s 1 t = at s1 ,
5 
b − s2 48 2 1
SII = b 2at + s2 + (b − s2 ) t = a t − t s22 ,
2 25 2
6
SIII = (2a − b)t s3 = at s3 .
5 3
26
Q/at
These relations result in the shear stresses τI

Q SI 3 Q s1 3
Q/at
τI = = , 13
It 26 at a 18
Q/at
 
65

Q SII Q 18 15 s22 τ
τII = = − ,
It at 65 208 a2 τ II

Q SIII 9 Q s3
τIII = = . 9
Q/at
It 52 at a 52
τIII
 Shear center 75

Problem 3.9 Locate the shear center b P3.9

for the depicted thin-walled (t
b, h)
box profile with a slit.
t

Solution We start by computing the

static moments with respect to the
y-axis of the three sub-areas:
y
s2 h2 h
SI = t 1 , SII = t + ts2 , s1
2 8  2 III I

h2 h h s3 s3 II
SIII = t + bt + s3 t − .
8 2 2 2
Thus the shear stresses become s2
τII
Q s21
τI = ,
I 2  τI
Q h2 h
τII = + s2 , M rM 0
I 8 2 
Q h2 h s3 Q τI
τIII = + b + (h − s3 ) . τIII
I 8 2 2
The equivalency of moments with re- τII
spect to 0 provides
 h/2  b  3 
h Qt h 1 1
Q rM = 2 τI bt ds1 + 2 τII t ds2 = b + bh3 + h2 b2
0 0 2 I 24 8 4
2
 
Qtbh 1 1
= h+ b .
I 6 4
With the second moment of area for the thin-walled profile
 
th3  h 2 ! h b
I=2 + bt = th2 +
12 2 6 2
we obtain the distance rM of the shear center M to the reference point 0
1 1
tbh2 6 h + 4 b 2h + 3b
rM = =b .
th2 1 1 2h + 6b
h+ b
6 2
 76 Bending along two axes

P3.10 Problem 3.10 The cantilever

with thin-walled box cross l
section is loaded by two ben-
ding moments My = F l and x
Mz = 2F l.
My
Determine the distribution of b Mz
the normal stresses in the
t
cross section for b = 2h. h t  b, h

Solution Because of symmetry y and z are principal axes. The stress

distribution is computed from

My Mz
σ= z− y.
Iy Iz
y
With
th3  h 2 1 2
z
Iy = 2 · +2· tb = th (h + 3b) ,
12 2 6
tb3  b 2 1 2
Iz = 2 · +2· ht = tb (b + 3h)
12 2 6
and the given bending moments we find
Fl 2F l 6F l  z y 
σ= z− y= 3 − .
1 2 1 th 7 10
th · 7h t 4h2 · 5h
6 6
The equation of the neutral axis (line of neutral axis
zero stress) is computed from σ = 0
7
z= y. σ
10 y
z
To clarify the representation the stresses 36 F l
due to the two loading cases are depicted seperately: 35 th2

21 F l 15 F l
due to Mz 35 th2 35 th2
due to My
 Biaxial bending 77

Problem 3.11 A beam, simply P3.11

supported at both ends, with a F l
thin-walled profile (t
b) is loa- 2
ded by a force F in the middle.
Determine the stress distributi- F b
on under the load as well as the t
location and value of the maxi-
mum stress. S
y 2b

b
z

Solution For the unsymmetrical profile the principal axes are not
known. We have to use the equations for biaxial bending. Thus we
obtain for the stresses with Mz = 0

My
σ= (Iz z + Iyz y) .
Δ
The moment due to the load is given by
Fl
My = Mmax = .
4
Together with the geometric quantities of the cross section
t(2b)3 8 tb3  b 2 ! 2 3
Iy = + 2 · b2 (bt) = tb3 , Iz = 2 + bt = t b ,
12 3 12 2 3
b
Iyz = −2 · b · · bt = −tb , 3
2
16 2 6 7
Δ = Iy Iz − Iyz2
= t b − t 2 b6 = t 2 b6
9 9
we obtain the stress
 
Fl 2 3 3 Fl
σ= t b z − t b3 y = (2z − 3y) .
7 2 6 3 28 t b3
4· t b
9
The neutral axis follows from the condition
3 neutral
σ=0 ; z = y. axis
2
compression
The maximal stresses occur at points 2 S
with the largest distance to the neutral y 3
axis (y = 0, z = ±b): tension
3 Fl
σmax = ± . z
14 t b2
 78 Computation of

P3.12 Problem 3.12 A cantilever beam

with thin-walled profile (t
a) is q0
subjected to a constant line load q0
and a concentrated force F .
F
Determine the distribution of the nor-
l
mal stress in the cross section at the
support.
Given: F = 2q0 l. a
t

Solution We place a y, z-coordinate system at a

the not yet known centroid. By symmetry to
the 45◦ -axis the distance ξC to both sub-areas
is identical. As the static moment vanishes with
regard to the symmetry axis, we have
a  a y C
ξC at = − ξC a t ; ξC = .
2 4 ξC
With regard to the symmetry axis we find z
ta3  a 2  a 2 5 3
Iy = Iz = + at+ at= ta ,
12 4 4 24
a a  a  a  1
Iyz =− at− − − a t = − ta3 .
4 4 4 4 8
This yields
 5 2 2 6 1 2 6 1 2 6
Δ = Iy Iz − Iyz
2
= t a − t a = t a .
24 64 36
The internal moments at the support are given by
q0 l2
My = − and Mz = F l = +2q0 l2 .
2
Finally we obtain for the stress
1
σ= {[My Iz − Mz Iyz ] z − [Mz Iy − My Iyz ] y}
Δ
'  
36 q0 l2 5 3 ta3
= 2 6 − ta − 2q0 l2 − z
t a 2 24 8
   (
5 q0 l2 ta3
− 2q0 l2 ta3 + − y
24 2 8
3 q0 l2
= (7z − 17y) .
4 ta3
 stresses in beams 79

Alternatively we can describe the stress distribution with respect to

the principal axes y ∗ , z ∗ , which we know from symmetry considerati-
ons. The principal values of the second moments of area follow with
Iy = Iz and ϕ = 45◦

Iy + Iz 5 3 1 3 1 3
Iy∗ = + Iyz = ta − ta = ta , C
2 24 8 12 y

Iy + Iz 5 3 1 3 1
Iz∗ = − Iyz = ta + ta = ta3 . z
2 24 8 3 y∗ z∗
Decomposition of the loading in the principal directions yields
π
q0 l2 ϕ=
My∗ = − cos ϕ + F l sin ϕ My 4
2
 
1 1√
= q0 l2 2 − 2,
2 2

q0 l2 Mz
Mz∗ = sin ϕ + F l cos ϕ
2
 
1 1√
= q0 l2 +2 2,
2 2

which leads to the stresses in the principal directions

√
My∗ ∗ Mz∗ ∗ 3 2 q0 l2
σ= ∗ z − ∗ y = (12z ∗ − 5y ∗ ) .
Iy Iz 4 ta3

To check the result we transform with

1√
z ∗ =−y sin ϕ + z cos ϕ = (z − y) 2,
2
1√
y ∗ =y cos ϕ + z sin ϕ = (z + y) 2
2
back and find by re-substitution

3 q0 l2 3 q0 l2
σ= 3
[12(z − y) − 5(z + y)] = (7z − 17y) .
4 ta 4 ta3
neutral axis
The neutral axis satisfies the equation compression
17 y
z= y. tension
7
z
 80 Computation of the deflection

P3.13 Problem 3.13 The beam is simply suppor- x

ted at both ends. Determine q0
a) location and value of maximal moment,
b) location and value of maximal deflection,
c) the slope of the deflection curve at both EI
supports. l

Solution Bending moment and deflection curve can be computed in-

dependently, because the beam is statically determinate.
to a) The given loading provides
x
q = q0
l
by twice integration
x2
V = −q0 + C1 ,
2l
3
x
M = −q0 + C1 x + C2 .
6l
With the static boundary conditions
q0 l
M (0) = 0 ; C2 = 0 , M (l) = 0 ; C1 =
6
we obtain
q0 l  x 2 ! q0 l2 x  x 3 !
V = 1−3 , M= − .
6 l 6 l l
Location and value of the maximal moment are determined by the con-
dition M  = 0 :
 x∗  2 1√
M = V = 0 ; 1 − 3 = 0 ; x∗ = 3 l = 0, 577 l ,
l 3
1√  1 1√
Mmax = M (x∗ ) = 3 q0 l2 1 − = 3 q0 l2 .
18 3 27

to b) With the known function of the moment

 
q0 l2 x  x 3
M= −
6 l l

we derive from EI w = −M by twice integration

q0 l2  x2 1 x4 
EI w = − − + C3 ,
6 2l 4 l3
2 3 5
q0 l x 1 x
EI w = − − + C3 x + C4 .
6 6l 20 l3
 by integration 81

The new integration constants are determined from the geometric boun-
dary conditions

w(0) = 0 ; C4 = 0 ,
 
q0 l3 1 1 7
w(l) = 0 ; C3 = − = q0 l3 .
6 6 20 360
Finally we obtain (cf. table on page 62, load case no. 3)
  x 3  x 5 
q0 l4 x
EI w = 7 − 10 +3 .
360 l l l

The maximal deflection is computed by using the condition w = 0 :

 ∗∗ 2  ∗∗ 4
x x
EI w = 0 ; 7 − 30 + 15 =0
l l
 ∗∗ 4  ∗∗ 2
x x 7
; −2 + = 0,
l l 15


∗∗ 8
; x = 1( + −
) l = 0, 519 l .
15

(The (+)-sign provides an x-value outside of the range of validity.) Thus

we have


q l 4
8 
8 
8 2 !
wmax = w(x∗∗ ) =
0
1− 7−10 1− + 3 1−
360EI 15 15 15

q0 l4
= 0, 0065 .
EI

to c) The slope of the deflection curve follows as

C3 7 q0 l3
w (0) = = ,
EI 360 EI
 
q0 l2 l l 7 q0 l3 8 q0 l3
w (l) = − − + =− .
6EI 2 4 360 EI 360 EI

Note: Maximal moment and maximal deflection occur at different

locations: x∗ = x∗∗ .
 82 Computation of the deflection

P3.14 Problem 3.14 Determine the func- x

tion of the bending moment for the q0
depicted beam.

EI
l

Solution The beam is statically indeterminate. Thus the function of

the moment needs to be computed with help of the deflection curve.
From the differential equation we derive by integration

EI wIV = q = q0 ,

−EI w = Q = −q0 x + C1 ,

x2
−EI w = M = −q0 + C1 x + C2 ,
2
x3 x2
EI w = q0 − C1 − C2 x + C3 ,
6 2
x4 x3 x2
EI w = q0 − C1 − C2 + C3 x + C4 .
24 6 2
The 4 integration constants follow from the 4 geometric boundary con-
ditions:

w (0) = 0 ; C3 = 0 ,

w(0) = 0 ; C4 = 0 ,
⎫
q0 l3 l2 ⎪ q0 l
w (l) = 0 ; − C1 − C2 l = 0 ⎪
⎬ C1 =
6 2 2
;
q0 l4 l3 l2 ⎪
⎪ q0 l2
w(l) = 0 ; − C1 − C2 = 0⎭ C2 = − .
24 6 2 12
This yields
  x 2 
q0 l2 x
M =− 1−6 +6 .
12 l l

q0 l2 √l q0 l2
x 12
12 12
l
M q0 l2
24
 by integration 83

Problem 3.15 Determine the deflec- P3.15

tion of the depicted beam. The left x q0
end of the beam is elastically sup-
ported by a spring, the right end is
clamped, and the load has the shape c EI
of a quadratic parabola.
l

Solution We start by computing the quadratic equation for the line

load. From the general equation q = A + Bx + Cx2 and
q(0) = 0 ; A=0,
⎫
q(l) = 0 ; Bl + Cl2 = 0 , ⎬ B q0
l l l2 ; C=− , B=4
q( ) = q0 ; B + C = q0 ,⎭ l l
2 2 4
.x x /
it follows q(x) = 4q0 − ( )2 .
l l
Four times integration of EI wIV = q yields
 x2 x3 
−EI w = V = −4q0 − 2 + C1 ,
2l 3l
 x3 x4 

−EI w = M = −4q0 − + C1 x + C2 ,
6l 12l2
 x4 x 5 
x 2
EI w = 4q0 − − C1 − C2 x + C3 ,
24l 60l2 2
 x5 x6  x3 x2
EI w = 4q0 − 2 − C1 − C2 + C3 x + C4 .
120l 360l 6 2
The boundary conditions provide
M (0) = 0 ; C2 = 0 ,
C4
V (0) = c · w(0) ; C1 = c ,
3
EI 2
q0 l l
w (l) = 0 ; − C1 + C3 = 0 ,
10 2
q0 l4 l3
w(l) = 0 ; − C1 + C3 l + C4 = 0 .
45 6
The 3 equations for C1 , C3 , and C4 yield with the abbreviation
Δ = 1 + cl3 /3EI
7 c q0 l4 q0 l3  1 cl3  7 q0 l4
C1 = , C3 = − 1− , C4 =
90 Δ EI 10Δ 18 EI 90 Δ
which leads to the final result
 
q0 l4 1  x 5 1 x 6 7 cl3  x 3  1 cl3  1  x  7
w= − − − 1− + .
10EI 3 l 9 l 54 ΔEI l 18 EI Δ l 9Δ
 84 Beams
x1 x2
P3.16 Problem 3.16 A cantilever beam is q0
subjected to a constant distributed
load q0 . EI
Determine the deflection at the free
end. 2a a

Solution We solve the problem in two different ways.

1st solution: Due to the discontinuity of q(x) we have to consider two
domains:

0 ≤ x1 < 2a q1 = 0 ,
V1 = C1 ,
M1 = C1 x1 + C2 ,
x21
EI w1 = −C1 − C2 x 1 + C3 ,
2
3
x x2
EI w1 = −C1 1 − C2 1 + C3 x1 + C4 ,
6 2
0 < x2 ≤ a q2 = q0 ,
V2 = −q0 x2 + C5 ,
x2
M2 = −q0 2 + C5 x2 + C6 ,
2
 x32 x2
EI w2 = q0 − C5 2 − C6 x 2 + C7 ,
6 2
x42 x32 x2
EI w2 = q0 − C5 − C6 2 + C7 x 2 + C8 .
24 6 2
The 8 integration constants Ci follow from:
⎧ 
4 boun− ⎨w1 (0) = 0 ; C3 = 0 , w1 (0) = 0 ; C4 = 0 ,
dary q0 a2
⎩
conditons Q2 (a) = 0 ; C5 = q0 a , M2 (a) = 0 ; C6 = − 2
⎧
⎪
⎪ M1 (2a) = M2 (0) ;C1 2a + C2 = C6 ,
⎪
⎪
and 4 ⎪
⎪ (2a)2
contin− ⎪ ⎨w1 (2a) = w2 (0) ; −C1 − C2 2a + C3 = C7 ,
uity 2
⎪ 3
(2a)2
condi− ⎪ ⎪
⎪w1 (2a) = w2 (0) = 0 ; −C1
(2a)
− C2
⎪
⎪ 6 2
tions ⎪
⎩
+C3 2a + C4 = C8 = 0
3 1 1
; C1 = − q0 a , C2 = q0 a2 , C7 = q0 a3 , C8 = 0 .
8 4 4
(For the shear force no continuity condition is available because it expe-
 with different areas 85

riences a jump related to the unknown reaction force B). The deflection
at the free end yields
' (
q0 a4 a4 a4 a4 3 q0 a4
w2 (a) = − + + = .
EI 24 6 4 4 8 EI

2nd solution: Using the Macauley bracket we can describe both do-
mains by a single equation. We introduce x from the left end and have
to consider the jump in the shear resultant at B (assumed to be positive
in upward direction):

q = q0 < x − 2a >0 ,
V = −q0 < x − 2a >1 +B < x − 2a >0 +C1 ,
1
M = − q0 < x − 2a >2 +B < x − 2a >1 +C1 x + C2 ,
2
1 1 1
EI w = q0 < x − 2a >3 − B < x − 2a >2 − C1 x2 − C2 x + C3 ,
6 2 2
1 1 1 1
EI w = q0 < x − 2a > − B < x − 2a > − C1 x3 − C2 x2 + C3 x + C4 .
4 3
24 6 6 2
The 5 unknowns Ci and B follow from
⎧
⎪
⎪ w (0) = 0 ; C3 = 0 ,
⎪
⎪
⎪
⎪
⎪
⎪w(0) = 0 ; C4 = 0 ,
4 boun− ⎪
⎨
dary condi−
⎪Q(3a) = 0 ; −q0 a + B + C1 = 0 ,
tions and ⎪ ⎪
⎪
⎪
⎪
⎪
⎪
⎪ 2
⎩M (3a) = 0 ; −q0 a + Ba + C1 3a + C2 = 0
2
'
1 reaction (2a)3 (2a)2
w(2a) = 0 ; −C1 − C2 + C3 2a + C4 = 0 .
condition 6 2
Solving yields:
3 1 11
C1 = − q0 a , C2 = q0 a2 , C3 = 0 , C4 = 0 , B= q0 a .
8 4 8
Thus the deflection at the free end is given by
 
q0 a4 11 a3 3 (3a)3 1 (3a)2 3 q0 a4
w(3a) = − a + a − a2 = .
EI 24 8 6 8 6 4 2 8 EI

Note: The computation of displacements at designated locations is less

complex with methods discussed in chapter 5.
 86 Computation of the deflection curve by Macauley bracket

P3.17 Problem 3.17 The depicted beam x

is loaded on its cantilever part by q0
a constant line load. EI
Compute the deflection at the hin- A
ge and determine the slope diffe- a a a
rence at the hinge. 2 2

Solution With the help of the Macauley bracket the entire domain can
be descibed by a single equation. During integration the jump in the
slope Δϕ at the hinge has to be considered separately.
a
q = q0 − q0 < x − >0 ,
2
a a
V = −q0 x + q0 < x − >1 +A < x − >0 +C1 ,
2 2
x2 q0 a 2 a
M = −q0 + < x − > +A < x − >1 +C1 x + C2 ,
2 2 2 2
 x3 q0 a 3 A a 2 x2
EI w = q0 − < x − > − < x − > −C1 − C2 x
6 6 2 2 2 2
+EIΔϕ < x − a > +C3 ,0

x4 q0 a A a x3 x2
EI w = q0 − < x − >4 − < x − >3 −C1 − C2
24 24 2 6 2 6 2
+EIΔϕ < x − a >1 +C3 x + C4 .
The 4 integration constants Ci , the unknown reaction force A, and the
slope difference Δϕ at the hinge are determined from the following 6
conditions

V (0) = 0 ; C1 = 0 , M (0) = 0 ; C2 = 0 ,
3 a 1 a
M (a) = 0 ; A = q0 a , w( ) = 0 ; q0 a4 + C3 + C4 = 0 ,
4 2 384 2
4 27 27
w (2a) = 0 ; q0 a3 − q0 a3 − q0 a3 + EIΔϕ + C3 = 0 ,
3 48 32
2 81 81
w(2a) = 0 ; q0 a4 − q0 a4 − q0 a4 + EIΔϕ a + C3 2a + C4 = 0.
3 384 192
This yields the solution
5 39 9
C3 = − q0 a3 , C4 = q0 a4 , EIΔϕ = q0 a3 .
24 384 32
Thus we obtain for the deflection at the hinge
1 q0 a4
wH = w(a) = −
12 EI Δϕ
wH
and for the slope difference
9 q0 a3
Δϕ = .
32 EI
 Beam with variable cross section 87

Problem 3.18 A leaf spring with P3.18

constant thickness t and variable b0
width b = b0 l/(l + x) is fixed at x
one side and loaded at one edge E
by F .
Determine the deflection at the b
position of the load. t F
l

Solution The system is statically determinate. Hence the function of

the moment follows from equilibrium considerations:
V = F = const , M = Fx + C .

The condition M (l) = 0 yields C = −F l and thus

M = −F (l − x) .

Use of the differential equation EI w = −M yields with

t3 b0 t 3 l
I(x) = b(x) =
12 12 l + x
and the abbreviation I0 = b0 t3 /12 :
F (l − x)(l + x) F
w = = (l2 − x2 ) .
EI0 l EI0 l
By integration we obtain
F 2 x3 
w = l x− + C1 ,
EI0 l 3
F  2 x2 x4 
w= l − + C1 x + C2 .
EI0 l 2 12
The boundary conditions
w (0) = 0 ; C1 = 0 , w(0) = 0 ; C2 = 0

render the solution

5 F l3
w(l) = wmax = .
12 EI0

Note: For a beam with constant width b0 the same load results in a
smaller deflection
F l3 4 F l3
w(l) = = .
3EI0 12 EI0
 88 Beam with variable cross section

P3.19 Problem 3.19 A cantilever beam with x

rectangular cross section (width b, q0
height h(x)) is subjected to a linear va-
rying load such that the extreme fibre
experiences a stress σ0 . E
Determine the deflection of the left end. l

Solution First we have to compute the unknown cross section height.

Using
|M |
σmax = = σ0
W
together with
q0 x3 b h3 (x) I b h2 (x)
M =− , I= , W (x) = =
6l 12 h/2 6
yields h(x)

q0 3/2
h(x) = x .
σ0 bl
This leads to

q0 q0 9/2
I(x) = x .
12σ0 l bσ0 l

Integration of EI w = −M provides together with the boundary con-

ditions w (l) = w(l) = 0 :
 
 M q0 x3 12σ0 l bσ0 l −9/2 σ0 bσ0 l −3/2
w =− = x =2 x ,
EI 6lEq0 q0 E q0

 σ0 bσ0 l  
w =2 −2x−1/2 + 2l−1/2 ,
E q0

σ0 bσ0 l  
w=2 −4x1/2 + 2l−1/2 x + 2l1/2 .
E q0

Evaluation at x = 0 yields the deflection at the left end


σ0 bσ0 l2
w(0) = 4 .
E q0

"
As a test we check the physical dimensions (F =force, "
L=length):

−2 −2 2
FL LF L L
[w] = = L.
F L−2 F L−1
 Superposition 89

Problem 3.20 The depicted beam is F P3.20

assembled from two parts with diffe- 2EI EI
rent bending stiffness.
I II
Determine the deflection at the free
end. l l

Solution We use superposition together with the tabulated results on

page 62. First we assume that beam II is fixed at point B and compute
the defection wII . To this we have to add the deflection wI of the left
beam I due to F and M = F l. Finally we have to consider the slope
wI , that appears at the left beam. This slop has to be multiplied by
the length l and added as an additional deflection at the right end:

F F F
I B II I B Fl II
wI B
f wI wII

f = wII + wI + wI l = wII + (wIF + wIM ) + (wI F + wI M )l .

According to load case no. 5

F l3 F l3 F l2
wII = , wIF = , wI F =
3EI 3(2EI) 2(2EI)

and load case no. 8

(F l)l2 (F l)l
wIM = , wI M = .
2(2EI) (2EI)

superposition yields the deflection at the end

' (
F l3 1 3 3 3 3 F l3
f= 1+ + + + = .
3EI 2 4 4 2 2 EI
 90 Superposition
q0
P3.21 Problem 3.21 Determine the
deflection curve for the depic-
A
ted beam.
EI
l

Solution The beam is statically indeterminate. We free the support

moment at the left end and introduce the unknown moment X:

q0 q0
X = MA

wq 
wX

From the table on page 62 we obtain for the slope:

q0 l3
load case no. 2 wq = ,
24EI

 Xl
load case no. 4 (with β = 1) wX = .
3EI

The total slope at the left support has to vanish. Thus compatibility
provides

1
wq + wX

=0 ; X = MA = − q0 l2 .
8

Superposition of the deflection curves in table on page 62 yields the

deflection curve of the system

EI w = EI(wq + wX )

q0 l4 1 l2
= (ξ − 2ξ 3 + ξ 4 ) − q0 l2 (2ξ + ξ 3 − 3ξ 2 )
24 8 6
q0 l4
= (3ξ 2 − 5ξ 3 + 2ξ 4 ) .
48
 Static indeterminate system 91

Problem 3.22 A pole is clamped at C P3.22

A and supported at B by an elastic EI
rope. The pole is subjected to a ho- a a
rizontal linearily varying load. B
EA
Compute the horizontal displace-
EI 1
ment v at point C for 2 = . 2a
a EA 3
q0
A
Solution We disconnect rope and pole:
vq vX

wq X wX
X

Δa

Compatibility at the connection of the rope requires

Xa
wq − wX = Δa , where Δa = (see chapter 2) .
EA
With the table on page 62 we obtain:
q0 (2a)4 8 q0 a4
load case no. 7 wq = = ,
30EI 15 EI
X(2a)3 8 Xa3
load case no. 5 wX = = .
3EI 3 EI
Using these values in the compatibility condition provides
1
8 q0 a4 8 Xa3 Xa q0 a 8
− = ; X= 5 = q0 a .
15 EI 3 EI EA 3 EI 45
1+
8 a2 EA
The displacement v results from superposition (for the linear varying
load we have to consider the displacement wq and the slope wq : vq =
wq + wq a):
+  3 ,
q0 (2a)4 q0 (2a)3 X(3a)3 2 1
EI v = EI(vq + vX ) = + a− 3· −1+
30 24 6 3 3
% &# $
load case no. 5 with α = 2/3
13 14 q0 a4
= q0 a4 − Xa3 = .
15 3 27
 92 Static indeterminate system

P3.23 Problem 3.23 Two parallel beams

(bending stiffness EI, length a) ha-
ve a distance of l and are clamped
at the left support. An elastic bar
(axial rigidity EA) of length l + δ l+δ δl l
is force fitted at a/2 between the
two beams.
a) Determine the force in the bar?
b) Compute the change e by which a/2 a/2
the distance l at the beam ends is
changed.

Solution to a) From geometry (compatibility)

l + 2wX = (l + δ) − Δl
X wX
fX 
; 2wX + Δl = δ wX
Δl
we obtain (see table on page 62,
X
load case no. 5)
 a 3 X
X Xl
wX = 2 und Δl =
3EI EA
X
and the force in the bar (compression)

δ EA 1
S=X= =δ .
l a3 l a3 EA
+ 1+
EA 12EI 12 l EI

to b) The opening e is computed with help of the table on page 62

from load case no. 5

+  3 ,
Xa3 1 1 5 a3 EA δ
e = 2 fX =2 3·1· −1+ = .
6 EI 2 2 24 l EI a3 EA
1+
12 l EI

EA
Note: In the limit case EI → ∞ one obtains S = δ and
l
e = 0.
 Superposition 93

Problem 3.24 Compute the P3.24

q0
reaction forces for the depic-
ted beam. EI
l/2 l/2

Solution The system is twice statically indeterminate. We treat the

support moment MA = X1 and the reaction force B = X2 as static
redundant quantities and use superposition:
q0 q0
X1 w2
w2

A B C wq wq w1 w1 X2

Considering the (arbitrary chosen) directions yields for the compatibi-

lity

wq + w1 − w2 = 0 ,

wq + w 1 − w 2 = 0 .

From the table on page 62 (no. 2, 4 and 1) we obtain

q0 l3 X1 l X2 l2
+ − = 0,
24 3 16
5 1 X2 l3
q0 l4 + X1 l2 − = 0,
384 16 48
which yields
1 4
X1 = − q0 l2 , X2 = q0 l .
56 7
The support reactions are determined by superposition of the 3 load
cases

q0 X1 X2 13
A= − − = q0 l ,
2 l 2 56
4
B = X2 = q0 l ,
7
q0 l X1 X2 11
C= + − = q0 l ,
2 l 2 56
1
MA = X1 = − q0 l2 .
56
 94 Static indeterminate system

P3.25 Problem 3.25 Determine the

deflection curve for the depic- q0
ted beam subjected to a tra- q1
pezoidal load.
A EI B
l

Solution The beam is statically indeterminate. We choose B as the

static redundant quantity and use superposition of 3 load cases (the
trapezoidal load is replaced by an equivalent constant and linearly va-
rying load)
q1
q1 − q0

x B

The table on page 62 (load case no. 6, 7 and 5) provides

q1 l4
EI w(x) = (6ξ 2 − 4ξ 3 + ξ 4 )
24
(q1 − q0 )l4 Bl3
− (10ξ 2 − 10ξ 3 + 5ξ 4 − ξ 5 ) − (3ξ 2 − ξ 3 ) .
120 6

The support condition at B yields the reaction force B

3 (q1 − q0 )l
w(l) = 0 ; B= q1 l − .
8 10

By recasting the above equations

q1 l4 (q1 − q0 )l4 q0 l4
= +
24 24 24

we determine the deflection curve

' ( ' (
q0 l4 5 3 3 2 (q1 − q0 )l4 9 3 7 2
EI w(x) = ξ4 − ξ + ξ + ξ5 − ξ + ξ .
24 2 2 120 2 2
 Superposition 95

Problem 3.26 For the beam with two q0 l P3.26

F
domains determine the support reacti- 2
ons and the deflection at the center of
each domain. A EI B C

Given: F = 2q0 l. l l

Solution We divide the beam into 2 separate (hinged at both ends)

beams and introduce the moment at the central support as statically
redundant quantity:
F
X

 (0)  (0)  (1)  (1)

w1 w2 w1 w2
(0) (0) (1) (1)
A (0)
B1 B2 C (0) (1)
A B1 B2 C (1)

Equilibrium yields
1
(0) (0) F
A(0) = B1 q0 l ,
= B2 = C (0) = ,
2 2
(1) (1) X
A(1) = C (1) = −B1 = −B2 = .
l
The table on page 62 provides
 (0) q0 l3  (0) F l2  (1)  (1) Xl
w1 =− , w2 = , w1 = −w2 =− .
24EI 16EI 3EI
Compatibility can be formulated as
 (0)  (1)  (0)  (1)
w1 + w1 = w2 + w2

which yields together with the tabulated results

1 3 1
X=− q0 l2 − F l = − q0 l2 = MB .
16 32 4
The support reactions are computed by superposition
1 1 1
A = A(0) + A(1) = q0 l − q0 l = q0 l ,
2 4 4
(0) (1) (0) (1)
B = B1 + B1 + B2 + B2 = 2q0 l ,
F 1 3
C = C (0) + C (1) = − q0 l = q0 l .
2 4 4
For the deflections at the center of the domains we compute
 
(0) (1) 5 q0 l4 Xl2 1 1 q0 l4
f1 = f1 + f1 = + − =− ,
384 EI 6EI 2 8 384 EI
 
(0) (1) F l3 Xl2 1 1 5 q0 l4
f2 = f2 + f2 = + − = .
48 EI 6 EI 2 8 192 EI
 96 Temperature load

P3.27 Problem 3.27 A beam (rectangular cross Tt

section, width b, height h) that is clam-
ped at both ends is subjected along its A x Tb B
length l to a constant temperature diffe- l
rence Tt − Tb .
Determine the defection of the beam and the maximum stresses.

Solution The beam is twice statically indeterimante. We choose as

statically redundant quantities the reaction moment X1 = MB and the
reaction force X2 = B. We use superpostion of the three (statically
determinate) systems:
 (0)  (1)  (2)
“0”-System wB “1”-System wB “2”-System wB
(0)
wB (2)
wB
(1)
wB MB B
The deflection in the “0”-System is computed by the temperature mo-
ment
MΔT = EIαT (Tb − Tt )/h

using the differential equation w (0) = −MΔT /EI and considering the
boundary conditions w(0) (0) = 0, w (0) (0) = 0:
MΔT MΔT x2
w (0) (x) = − x, w(0) (x) = − .
EI EI 2
Due to the clamping at B compatibility requires
(0) (1) (2)   (0)  (1)  (2)
w B = w B + wB + wB = 0 , wB = wB + wB + wB = 0.

From the table on page 62 we obtain

MΔT MB l Bl2 MΔT l2 MB l2 Bl3
− l− − =0, − − − = 0,
EI EI 2EI EI 2 2EI 3EI
with the solution
B = 0, MB = −MΔT .

As MB = M is constant along the entire length of the beam the deflec-

tion becomes
M + MΔT
w = − =0 i. e. w ≡ 0.
EI
The maximum stress is computed with the section modulus W = bh2 /6
|M | MΔT
|σmax | = =6 .
W bh2
 Frame 97
q0
Problem 3.28 Determine the P3.28
support reactions for the de- EI
picted frame.
a a

Solution We free the right support and use B as static redundant

quantity
q0 q0

A B vB
vq
B
The individual displacement components are determined from the table
on page 62 and superposition:
q0 q0 a2 q0
2 vq2
ϕ vq1
ϕ

vq
q0 a2 q0 a4 5
vq = vq1 + vq2 = ϕ · a + vq2 = ·a·a+ = q0 a4 ,
2 8 8
Ba ψ v B1 v B2

B
ψ
vB

B
a3 4
vB = vB1 + vB2 = ψ · a + vB2 = Ba · a · a + B = Ba3 .
3 3
The compatibility at B provides the reaction force B:
15
vq = vB ; B= q0 a .
32
The other support reactions follow from equilibrium
17 1
A= q0 a and MA = − q0 a2 .
32 32
 98 Superposition
l l
P3.29 Problem 3.29 An auxiliary bridge, q0
that is resting on the river banks, is
supported in the middle by an addi- EI
tional pontoon (block with cross sec-
tion A at the water line). The bridge ρ A
is subjected to a constant load q0 .
Given: water density ρ, EI/Al3 ρg = 1/24 .
Determine the immersion depth f of the pontoon due to q0 .

Solution The system is statically indeterminately supported. We use

the pontoon force as statically redundant force and apply superposition:

q0 q0 wX

wq X

For the immersion of the pontoon we obtain FA

f = wq − wX .

Archimedes’ principle yields the buoyant force FA that is equal to the

weight of displaced fluid (see also chapter 7), i. e. we have
X
X = FA = ρgf A ; f= .
ρgA
The table on page 62 provides
5 q0 (2l)4 X(2l)3
no. 2 : wq = , no. 1 : wX = .
384 EI 48EI
Using the above results
5 q0 l4
4 3
X 5 q0 16l X8l 24 EI
= − ; X= = q0 l .
ρgA 384 EI 48EI 1 l3 1
+
6 EI ρgA
the immersion depth is given by
X q0 l EI l3 1 q0 l4
f= = = .
ρgA ρgA EI l3 24 EI
 Superposition 99

Problem 3.30 An elastic rope P3.30

(length s) is fixed to the wall
and in C frictionless redirected ϕ EA
by a pulley. The pulley is atta-
ched to a beam (axial rigidity
→ ∞), C
Determine the displacement of EI
Q
the load Q.
l

Solution The displacement of Q is computed by the length change

Qs
Δs =
EA

of the rope and a contributions δ of the deflection of the pulley. The

deflection is calculated by the vertical load on the beam

V = Q − S cos ϕ = Q(1 − cos ϕ)

to be V S=Q
ϕ
V l3 Q(1 − cos ϕ)l3 H
w= = . wH V
3EI 3EI Q

The deflection δ of the load Q follows from

δ = w + an − av
ϕ bv
= w + (s − bn ) − (s − bv )
bn
= w + bv − bn C
w av
with
C Q
bn − bv = w cos ϕ (for w
bv ) . an δ

Q
This leads to the deflection of Q

 
Qs s l3 (1 − cos ϕ)2
vQ = δ + Δs = w(1 − cos ϕ) + =Q + .
EA EA 3EI
 100 Statically indeterminate system

P3.31 Problem 3.31 The depicted struc-

ture consists of a beam and bars a a a a
with stiffness ratio α = EI/a2 EA.
EI B
The structure is loaded by the force
F. A 2 EA
 a
a) Determine the forces in the bars- 
1 1
√
for α = 1/8 C 2EA
b) For which value of α vanishes the F
force S2 ?
c) For which α follows MB = 0?

Solution The system is statically indeterminate in the interior. We free

the middle bar (basic system):

(0) (0)
S1 S1
A B
C
C F F F
2 2
F
(0) √
Equilibrium in C yields S1 = 2F/2. The beam is loaded by the
components F/2. With the table on page 62 (load case no. 1) the dis-
placement at A is given by
 
(0) F (4a)3 3 1 9 1  1 1 1 1 2
EI wA = · 1− − + · 1− − = F a3 ,
2 6 4 4 16 16 4 4 16 16 3
and at location B
 
(0) F (4a)3 1 1 1 1 11 3
EI wB = 2 · · 1− − = Fa .
2 6 4 2 16 4 12
Due to the truss elongation Δl1 point C experiences the displacement
1√ √ C
(0)
√ S1 l1 √ 2
2 Fa 2√ F a Δl1
wC = Δl1 2= √ 2= √ 2= .
2EA 2 EA EA
(0)
Hence the total displacement of C is given by wC
C
(0) (0) (0) 2 F a3 Fa
vC = wB + wC = + .
3 EI EA
Now we load the system by the unknown normal force S2 = X and
consider the two load cases independently:
 Superposition 101

I II
A B A B
X X X

In sub-system I the deformation is analogous to the basic system, if F

is replaced by −X, i. e.

(I) 2 Xa3 Xa (I) 11 Xa3

vC = − − , wB = − .
3 EI EA 12 EI
The displacement in sub-system II is again determined from the table
on page 62

(II) X(4a)3 4 Xa3

wB = = ,
48EI 3 EI
' (
(II) (II) X(4a)3 1 1  1 1 11 Xa3
vC = wA = 1− − = .
6EI 2 4 4 16 12 EI
Compatibility requires that the difference in the total displacement at
points C und B are equal to the elongation of bar 2:
! Xa
(0) (I) (II) (0) (I) (II)
vC + vC + vC − wB + w B + w B =
EA
or
2F a3 Fa 2Xa3 Xa 11Xa3  11F a3 11Xa3 4Xa3  Xa
+ − − + − − + =
3EI EA 3EI EA 12EI 12EI 12EI 3EI EA
1
α− 4
; X= F.
1
2α + 6

With this result the answers to the questions are:

1−1
to a) X = S2 = 8 4 F = − 3 F , S = 1 √2 (F − X) = 13 √2 F ,
1
1+1 10 2 20
4 6
1
to b) S2 = X = 0 ; α = ,
4
F F X
to c) MB = 2a − − a = 0 ; X = −F ,
2 2 2
1
α− 4 1
; F = −F ; α= .
1
2α + 6 36
 102 Superposition principle EAS
D
P3.32 Problem 3.32 The two depic- F
ted posts have to be connected
by a rope. The rope has to be EI EI
fixed at points A and B. The h
Δl
rope is too short by Δl. A B
a) Determine the horizontal
force F at the right post
that is required to fix the
rope stress-free.

b) The force F is removed after assembly. Determine the force in

the rope and the moments at both supports.

Solution to a) The force F has to bend the post by Δl to the left.

From the table on page 62 (load case no. 5) we obtain

F h3 3EI
Δl = ; F = Δl .
3EI h3

to b) The length Δl follows from the extension ΔlS of the rope due to
a yet unknown force S in the rope and the deflection fS of both posts
due to the same unknown force S. Compatibility states

Δl = ΔlS + fS + fS

which yields

Sl Sh3 Sh3 Δl 1
Δl = + + ; S= EAS .
EAS 3EI 3EI l 2 h3 EAS
1+
3 lEI

Finally the moments at the support follow from equilibrium

Δl 1
M = hS = EAS h .
l 2 h3 EAS
1+
3 lEI
 Frame 103

Problem 3.33 A plane frame a P3.33

is loaded in C and D by two
2
forces. a
F 3
C
Determine the reciprocative EI
horizontal displacement Δu of 2a
C und D. D 2
F a
3

Solution To apply the table on page 62 we have to separate the defor-

mation of the individual beams and use superposition.
2
2 M =F a
M =F a ϕ 3
3
ϕ ψ
ϕ F
2 2 w
ϕ a ψ a
3 3 w
F

M M
2 2
C is moved by ϕ· a+ψ· a+w to the right,
3 3
2 2
D is moved by ϕ· a+ψ· a+w to the left.
3 3
Thus, the reciprocative displacement follows
 
2 2
Δu = 2 ϕ · a + ψ · a + w .
3 3
With the table on page 62 it follows:
2  2a  2  2a 2
load case no. 2 EI ϕ = Fa − Fa = F a2 ,
3 3 3 6 9
2  2
load case no. 8 EI ψ = F a a = F a2 ,
3  3
2 3
F a 8
load case no. 5 EI w = 3 = F a3 ,
3 81
which yields
 
4 4 8 F a3 112 F a3
Δu = 2 + + = .
27 9 81 EI 81 EI

Note: Due to the antisymmetry of the system the vertical displace-

ments of C and D are the same.
 104 Frame
l l
P3.34 Problem 3.34 The depic- 2 2 M0
ted frame is loaded by a
moment M0 .
EI
Determine the reciproca-
tive rotation ΔϕH at the
hinge. l

Solution It is reasonable to split the loading into a symmetric and

antisymmetric contribution:

M0 M0 /2 M0 /2 M0 /2 M0 /2

The antisymmetric loading causes no M0 /2

reciprocative rotation at the hinge.
For the symmetric loadign it suffices ψ
ψ
to consider half of the frame struc-
ture. The rotation ψ results solely
from the bending of the vertical post
(only a normal force occurs in the ho-
rizontal beam). Thus from the table
on page 62 (load case no. 4 with β = 1
and α = 0) we obtain
M0
l M0 l
ψ= 2 = .
3EI 6EI

Hence the reciprocative rotation follows

M0 l
ΔϕH = 2ψ = .
3EI
 Unsymmetrical bending 105

Problem 3.35 Determine for the a P3.35

depicted beam with a thin-walled F
t
profile the displacement at the
point where the load is applied.
x
F y 2a
E ta
a
l z
Solution Due to the unsymmetrical profile oblique bending occurs. The
displacements are computed using the two related differential equati-
ons. The bending moments are given by
My = −F (l − x) , Mz = 0 ,
and the second moments of area for the thin-walled profile follow from
t(2a)3 8 2
Iy = + 2(at)a2 = ta3 , Iz = ta3 ,
12 3 3
a 7
Iyz = −2(ta)a = −ta3 , Δ = Iy Iz − Iyz
2
= t2 a6 .
2 9
Thus the two differential equations can be integrated for the z-direction
My Iz 6 F
Ew = − = (l − x) ,
Δ 7 ta3
3 F
Ew = − (l − x)2 + C1 ,
7 ta3
1 F
Ew = (l − x)3 + C1 x + C2
7 ta3
and the y-direction
My Iyz 9 F
Ev  = − =− (l − x) ,
Δ 7 ta3
9 F
Ev  = (l − x)2 + C3 ,
14 ta3
3 F
Ev = − (l − x)3 + C3 x + C4 .
14 ta3
The boundary conditions at the support yield
9 F l2 3 F l2
v  (0) = 0 ; C3 = − 3 , w (0) = 0 ; C1 = ,
14 ta 7 ta3
3 3
3 Fl 1 Fl
v(0) = 0 ; C4 = , w(0) = 0 ; C2 = − .
14 ta3 7 ta3
Thus the displacements at the point, where the load is applied x = l, are
2 F l3 3 F l3
w(l) = , v(l) = − .
7 Eta3 7 Eta3
Note: Although the load is acting in vertical direction a displacement
in horizontal direction occurs. The profile preferably deforms in
the direction which is related to the smaller second moment of area.
 106 unsymmetrical bending

P3.36 Problem 3.36 The simply x

supported beam is loaded by q0
a constant distributed load.
Determine the displacement E
of the centroid of the cross l
section in the middle of the q0
beam (only deformation due 10
to bending).
Given: l = 2 m ,
130 65
E = 2.1 · 105 MPa , [mm]
q0 = 104 N/m .
10

Solution We compute the geometric quantities of the cross section:

ηC
A = 65 · 10 + 120 · 10 = 1850 mm ,
ζC
(65 · 10) · 5 + (120 · 10) · 70 C
ζC = = 47.16 mm , y
1850
z
(65 · 10) · 32.5 + (120 · 10) · 5
ηC = = 14.66 mm ,
1850

65 · 103 10 · 1203
Iy = + (42.16)2 (65 · 10) + + (22.84)2 (10 · 120)
12 12
= 322.7 cm4 ,

10 · 653 120 · 103

Iz = + (17.84)2 (65 · 10) + + (9.66)2 (10 · 120)
12 12
= 55.8 cm4 ,

Iyz = −(−17.84)(−42.16)(65 · 10) − (22.84)(9.66)(10 · 120)

= −75.4 cm4 ,

Δ = Iy Iz − Iyz
2
= 12321.5 cm8 .
The loading causes only a moment along the y-axis:
q0 l x2
My (x) = x − q0 .
2 2
 unsymmetrical bending 107

The basic equations simplify to

My Iz My Iyz
Ew = − , Ev  = − .
Δ Δ

Integrating twice yields

 
Iz q0 x2 x3
Ew = − l − + C1 ,
Δ 2 2 3
 3 
Iz q0 x x4
Ew = − l − + C1 x + C2 ,
Δ 2 6 12
 2 
 Iyz q0 x x3
Ev = − l − + C3 ,
Δ 2 2 3
 3 
Iyz q0 x x4
Ev = − l − + C3 x + C4 .
Δ 2 6 12

The boundary conditions

w(0) = 0 ; C2 = 0 , v(0) = 0 ; C4 = 0 ,

l3 l3
w(l) = 0 ; C1 = − , v(l) = 0 ; C3 = −
12 12
x
together with the abbreviation ξ = yield
l
q0 l4 ) 4 * Iz
Ew = ξ − 2ξ 3 + ξ ,
24 Δ

q0 l4 ) 4 * Iyz
Ev = ξ − 2ξ 3 + ξ .
24 Δ

In the middle of the beam (ξ = 1/2) the curly brackets attain the value
5/16 which leads with the given numerical values (converted to cm) to

5 55.8 1
w = 102 · 2004 · = 0.45 cm ,
384 12321.5 2.1 · 107
C v
5 −75.4 1
v = 102 · 2004 · = −0.61 cm ,
384 12321.5 2.1 · 107 f
w

f= w2 + v 2 = 0.76 cm .
 108 unsymmetrical bending

P3.37 Problem 3.37 In the middle of F l

x
a beam the force F is applied. 2
The thin-walled profile is pro-
duced from an aluminium sheet E
of 2 mm thickness. l
F 4
Compute the deformation at
the point where the force is
2
applied.
Given: l = 2 m , C 10
E = 7 · 104 MPa , y
F = 1200 N .
2 [cm]
4
z
Solution The displacement can be determined with regards to the y, z-
axes, or with regard to the principal axes. We want to consider both
possibilities.
1st solution: The position of the centroid is known. With regard to
the y, z-axes we find
0.2 · 103  0.2 · 103 0.2 · 63 
Iy = + − + 2 · 52 · 0.2 · 4 = 69.73 cm4 ,
12 12 12
0.2 · 83
Iz = + 2 · 42 · 0.2 · 2 = 21.33 cm4 ,
12
Iyz = −2{5 · 2 · 0.2 · 4 + 4 · 4 · 0.2 · 2} = −28.8 cm4 ,

Δ= Iy Iz − Iyz
2
= 657.9 cm8 .
F
With the bending moments My = x , Mz = 0 für 0 ≤ x ≤ l/2
2
(symmetry) the differential equations are given by
F Iz F Iyz
Ew = − x, Ev  = − x.
2Δ 2Δ
After integration and incorporation of the boundary conditions we ob-
tain in the middle of the beam (see also table on page 62):
F l 3 Iz 1200 · 2003 21.33
w= = · = 0.93 cm ,
48E Δ 48 · 7 · 106 657.9

F l3 Iyz 1200 · 2003 (−28.8)

v= = · = −1.25 cm ,
48E Δ 48 · 7 · 106 657.9
√
f = w2 + v 2 = 1.56 cm .
 unsymmetrical bending 109

2nd solution: We refer to the principal axes. The principal directions

and values of the second moment of area are given by
2Iyz
tan 2ϕ∗ = = −1.19 ; ϕ∗ = −24.98◦
Iy − Iz η
ψ ∗ = −ϕ∗
91.06 
I1,2 = ± 24.22 + 28.82 y
2
ζ
; I1 = Iη = 83.15 cm4 , I2 = Iζ = 7.91 cm4 . z

Decomposition of the load into principal directions yields

Fζ = F cos ψ ∗ = 0.906 F , Fη = −F sin ψ ∗ = 0.422 F ,

and the displacements follow from the table on page 62 (load case no. 1)

Fη l3 1200 · 0.422 · 2003

fη = =− = −1.52 cm ,
48EIζ 48 · 7 · 106 · 7.91

Fζ l3 1200 · 0.906 · 2003

fζ = = = 0.31 cm ,
48EIη 48 · 7 · 106 · 83.15

f = fη2 + fζ2 = 1.55 cm .
η

ψ∗ v
y ψ∗
fη
f
ζ w fζ
z

For comparison with the 1st solution we transfer the displacements into
the y, z-coordinate system:

|v| = |fη | cos ψ ∗ − fζ sin ψ ∗ = 1.25 cm ,

w = |fη | sin ψ ∗ + fζ cos ψ ∗ = 0.93 cm .

Note: We used in the computations numerical values up to the second

digit. Thus the numerical value for the total displacement f differs in
the second digit.
 110 Inhomogeneous cross section

P3.38 Problem 3.38 A beam compo- x

b
sed of two different materials
(a bi-metal beam to measu- E1 , α1 h
re temperature) is heated uni- E2 , α2 y h
formly by a temperature diffe-
rence ΔT . l z
Determine the deformation at the free end.

Solution We assume a linear stress distribution in each material and

replace the stresses by a resultant force Fi and a resulting moment Mi .
If we suppose α2 > α1 the lower
part wants expand more. As this F1
M1
is prevented by the upper part, h
the lower part is under compres- M2
sion, while tension prevails in F2

the upper part. F1 and F2 cause

a moment in the composite beam which is in equilibrium with M1 and
M2 (no external loads). Thus the following equations hold:

statics N =0 ; F1 = F2 = F ,
M =0 ; F h = M1 + M2 ,
M1 12 M2 12
Hooke’s law w1
=− , w2 = − .
E1 bh3 E2 bh3
Kinematic compatibility demands
w1 = w2 = w .

Additionally the strains have to match at the interface. They consist of

three contributions: temperature αi ΔT , normal force F/EA and ben-
ding M/EW . Considering tension and compression we formulate
F M1 6 F M2 6
α1 ΔT + + = α2 ΔT − − .
bhE1 E1 bh2 bhE2 E2 bh2
Eliminating the moments Mi and rearrangement to get w yields
12E1 E2 (α2 − α1 )ΔT
w = − = −C .
h(E12 + 14E1 E2 + E22 )
Integration, by incorporating the boun-
ΔT
dary conditions at the left end, provides
the displacement at the free end w
l2
w = −C .
2
Chapter 4
Torsion
4
112 Torsion

Torsion

If an external load causes an internal moment Mx along the longitu-

dinal axis, the bar is loaded by torsion (twisting). In the following we
refer to the moment Mx as torque or torsional moment MT .

MA A
m(x)
B
x
y MB
z l
 Basic equations 113

Prerequisites, assumptions:

• Warping of the cross sections is not constrained (pure tor-

sion),
• The shape of the cross sections does not change during rotation.

Equilibrium conditions

dMT
= −m , m(x) = external moment per unit length.
dx

Differential equation for the angle of twist

dϑ
GIT = MT ,
dx
x P
ϑ = angle of twist, y
GIT = torsional rigidity, ϑ(x)
G = shear modulus, z
IT = torsional constant. P´

Twist of end sections

l l
 MT
Δϑ = ϑ(l) − ϑ(0) = ϑ (x)dx = dx .
GIT
0 0

Special case: GIT = const, MT = const

MT l
Δϑ = .
GIT

Maximum shear stress

MT
τmax = , WT = sectional moment of torsion.
WT

The location of the maximum shear stress is provided in the following

table.
114 Torsional moments of area

Cross section IT

ri
π 4
τ (r) IT = Ip = (r − ri4 )
2 a
τmax

π 4
ra ri = 0 (full circle) IT = r
2 a

thin-walled, closed profile

4A2T
τ (s) IT = 0
AT s ds
t(s)
t(s)

t
a a = const
IT = 2πa3 t
t = const

thin-walled, open profile

t(s) h
s 1
IT = t3 (s)ds
τ 3
0
h
h
t
1 3
t = const IT = ht
3
ti hi 1 3
ti = const IT = hi ti
3
square
τmax a
IT = 0, 141a4

a
τmax
ellipse
b
a 3 b3
a IT = π
a + b2
2

τmax
 Sectional moments for torsion 115

WT Remarks

The shear stresses are distributed

IT π ra4 − ri4 linearly across the cross section:
WT = =
ra 2 ra MT
τ (r) = r.
IT
π 3 Cross sections remain plane
WT = r
2 a during deformation.

τ is constant across the wall-thickness t.

The shear flow
WT = 2AT tmin MT
T = τt =
2AT
is constant.
τmax occurs at the smallest wall-
thickness tmin .
WT = 2πa2 t AT is the area encircled by the central
line of the profile.

IT
WT =
tmax
τ is linearly distributed across the
wall-thickness.
1 2 τmax occurs at the largest wall-
WT = ht
3 thickness tmax .
IT
WT =
tmax

τmax occurs at in the middle of

WT = 0.208 a3
the lateral lengths.

π 2 τmax occurs at the ends of the smaller

WT = ab
2 semi-axis.
 116 Twist
2R
P4.1 Problem 4.1 A shaft with cir-
cular cross section is clamped x F
at one end and loaded by a
pair of forces.
2r
Determine F such that the F
admissible shear stress τadmis
l
is not exceeded. Compute for
this case the twist of the end
section.

Given: R = 200 mm, r = 20 mm, l = 5 m, τzul = 150 MPa,

G = 0.8 · 105 MPa.

Solution The torque (torsional moment)

MT = 2RF

is constant along the bar. The maximum shear stress in the cross section
is given with
π 3
WT = r
2
by
MT 4RF
τmax = = .
WT πr 3
In order not to exceed the admissible shear stress,
πr 3
τmax ≤ τadmis ; F ≤ τadmis .
4R
must hold and we obtain
πr 3 π · 8000 · 150
Fmax = τallow = = 4712 N .
4R 4 · 200
For this load the twist (in radians) can be computed using
π 4
IT = r and MT = 2RFmax .
2
Inserting yields
MT l τzul l 150 · 5000
Δϑ = = = = 0.47 .
GIT Gr 0.8 · 105 · 20
This value is equivalent to an angle of 27◦ .
 and stress 117
2b
Problem 4.2 A shaft has to carry 
1 
2
P4.2
the torque MT = 12 · 103 Nm. Select a
2b
a cross section from the depicted
group.

Dimension the cross sections such 

3 
4

that the admissible shear stress c d

τadmis = 50 MPa is not exceeded.
Which cross section is the most effi- c d
cient in terms of material usage? t= t=
10 10

Solution The admissble shear stress is reached for

MT
τmax = = τadmis .
WT

With the section moment for torsion

π 3
WT1 = a , WT2 = 0.208 · 8 b3 = 1.664 b3 ,
2
π 3 2π 2 π 3
WT3 = 2πc2 t = c , WT4 = dt = d
5 3 150

we determine with the given numerical values

3 2MT 3 MT
a= = 53.5 mm , b= = 52.4 mm ,
πτzul 1.664 τzul

3 5 MT 3 150 MT
c= = 72, 6 mm , d= = 225.5 mm .
π τzul π τzul

The cross section areas are

A1 = πa2 = 89.8 cm2 , A2 = 4b2 = 110.0 cm2 ,

π 2 π 2
A3 = c = 33.1 cm2 , A4 = d = 319.4 cm2 .
5 5

Therefore, the third cross section (i. e. the thin-walled closed profile) is
the most material efficient profile.
 118 admissible

P4.3 Problem 4.3 Determine the MT MT

maximum admissible torque
(torsional moment) and the l
corresponding admissible
twist for the closed profile t t
and the profile that is slit at 2t 2t 2t 2t
A. a
Given: a = 10 cm, t = 2 mm, t t A
τadmis = 20 MPa,
l = 5 m,
G = 0.8 · 105 MPa. a

Solution The admissible torque and the admissible twist are computed
for both profiles via

MTadmis l τadmis WT l
MTadmis = τadmis WT , Δϑadmis = = .
GIT GIT
In the case of the closed profile with t
a it holds
0 a
ds a a
AT = a2 , =2 + =3 ,
t(s) 2t t t
4A2T 4
IT = 0 = ta3 , WT = 2AT tmin = 2a2 t
ds 3
t(s)
and we obtain
MTadmis = τadmis 2a2 t = 800 Nm ,

3τadmis l
Δϑallow = = 0.01875 " 07◦ ) .
(=1,
2Ga
If the profile is open (slit at position A), we compute with
1 3 IT
IT = t hi = 6t3 a , WT = = 3t2 a
3 i i tmax

the torque and twist

MTadmis = τadmis 3t2 a = 24 Nm ,

τadmis l ◦
Δϑadmis = "
= 0.3125 (=17.9 ).
2Gt

Note: The closed profile is much stiffer with respect to torsion than the
open profile.
 twist 119

Problem 4.4 A shaft is loaded P4.4

by a pair of forces. The shaft
is assmbled from two different
2b
thin-walled cross sections (t

a) of the same material (shear

modulus G).
F
Determine in both cases the ad- l F
missible forces and the corre-

1 t 
2 √
sponding twist such that the t 2a
a t
shear stress τadmis is not excee-
t
ded.
t
2a
Solution The torque MT = 2bF is constant along the length of the
shaft. Stress and twist are determined from

MT 2bF MT l 2bF l
τ = = , Δϑ = = .
WT WT GIT GIT
The admissible shear stress will not be exeeded for
WT τadmis WT τadmis
τ ≤ τadmis ; F ≤ ; Fadmis = ,
2b 2b

2blFadmis τadmis WT l
Δϑadmis = = .
GIT GIT
With the values for the two different cross sections
0
π ds a π2 3
x A T = a2 , = (2 + π) , WT = πa2 t , IT = a t,
2 t t 2+π
0 √
ds a 2
y A T = a2 , = (2 + 2 2) , WT = 2a2 t , IT = √ a3 t
t t 1+ 2
we obtain
π a2 t a2 t
Fadmis1 = τadmis , Fadmis2 = τadmis ,
2 b b
2 + π lτadmis √ lτadmis
Δϑadmis1 = , Δϑadmis2 = (1 + 2) .
π aG aG

Note: The admissible force is larger for the first profile, while the
admissible twist is larger for the second profile.
 120 Warping t
2t 2t
P4.5 Problem 4.5 The thin-walled box girder y
is loaded by a torque MT .
z
Determine the warping of the cross section.
t
2a
Solution The warping u(s) (displacement
in longitudinal direction) is computed from

4
the shear strain
1 
3
∂u ∂v ds
γ= +
∂s ∂x x
of the wall segments. With 2
dx s
τ MT
γ= = ,
G G2AT t(s)
∂v dϑ MT r⊥
= r⊥ = r⊥ (s) ,
∂x dx GIT
α
4 · 16a4 32 3 s
2
AT = 4a , IT = = a t r
4a + 4a 3 dϑ
t 2t P
we obtain dv = r⊥ dϑ
P´
  α
∂u MT t 3r⊥ (s)
= − .
∂s 8Ga2 t t(s) 4a

Integration in region x provides (t(s) = 2t, r = a) with u(s = 0) = 0

(then u vanishes on average)
 
MT 1 3 MT
u1 (s) = − s=− s.
8Ga2 t 2 4 32Ga2 t

Analogously, we obtain in regions y , z , {


4
MT
u2 (s) = [s − 2a] , 
32Ga2 t 3

1
x
MT
u3 (s) = − [s − 4a] ,
32Ga2 t s MT
MT 32Gat
MT − 
u4 (s) = [s − 6a] . 32Gat 2
32Ga2 t
 and stress 121

Problem 4.6 A tube y is mounted MT P4.6

by heat shrinking on a shaft x
with circular cross section of diffe- l
rent material.
Determine the maximum shear
stresses in x and y as well as the 
2
twist under the application of a tor- R2
que MT . R1

1

Solution First we consider shaft x and pipe y independently. For the

angle of twist and the stress it yields
MT1 l MT1
ϑ1 = , τmax1 = ,
G1 Ip1 WT1
MT2 l MT2
ϑ2 = , τmax2 =
G2 Ip2 WT2
with
π 4 π 4  I p1 I p2
I p1 = R , I p2 = R2 − R14 , WT1 = , WT2 = .
2 1 2 R1 R2
Together with equilibrium

MT = MT1 + MT2

and geometric compatibilty

ϑ1 = ϑ2 = ϑ

we obtain
G1 Ip1 G2 Ip2
MT1 = MT , MT2 = MT
G1 Ip1 + G2 Ip2 G1 Ip1 + G2 Ip2
and

MT G1 r1 MT G2 r2
τmax1 = , τmax2 = ,
G1 Ip1 + G2 Ip2 G1 Ip1 + G2 Ip2

MT l
ϑ= .
G1 Ip1 + G2 Ip2
 122 Twist

P4.7 Problem 4.7 A conical shaft

with varying radius is loaded by r(x) MT
a torque MT . 4a 2a
x
Determine the twist and the
peripheral stress as a function l
of x.

Solution The differential equation for the twist angle is given with
 x π 4 π  x 4
r(x) = a 2 − , Ip (x) = r = a4 2 −
l 2 2 l
by
MT 2MT 1
ϑ = =  4 .
GIp πGa4
2− x
l
Integration with respect to x yields
2MT l 1
ϑ(x) =  3 + C .
3πGa4
2− x
l
The integration constants are determined from the boundary conditions
2MT l 1
ϑ(0) = 0 ; C=− .
3πGa4 8
Thus the twist results in
⎧ ⎫
⎪
⎨ ⎪
⎬
MT l 1
ϑ(x) =   − 1 .
12πGa4 ⎪ ⎩ 1− x
3 ⎪
⎭
2l

The peripheral shear stress is computed with

Ip π  x 3
WT (x) = = a3 1 −
r 2 l
as
MT 2MT
τP (x) = =  3 .
WT
πa3 2− x
l

Twist and stress have a maximum at x = l:

7MT l 2MT
ϑ(l) = , τP (l) = .
12πGa4 πa3
 and stress 123
l1
Problem 4.8 The depicted gear- P4.8
system consists of two shafts d1
(lengths l1 , l2 ) of identical mate- R1 M1
rial, that are connected by two 
2
gear wheels (radii R1 , R2 ). The C
shaft x is loaded by an external M2 d2 R2
torque M1 . 
1

a) Determine M2 such that

A l2 B
equilibrium is fulfilled.
b) Choose the diameters d1 and d2 such that the admissible shear stress
τadmis is not exceeded?
c) Compute the angle of twist at position C, if shaft y is fixed at
position A.

Solution to a) Equilibrium of moments

M1
M1 = R1 F , M2 = −R2 F

yields F
R2 M2 F
M2 = − M1 .
R1

to b) The critical value of the shear stress is reached in each shaft for:

|M1 | 16M1 3 16M1 ,

τmax1 = = = τ admis ; d 1 =
W1 πd31 πτadmis

|M2 | R2 16M1 R2
τmax2 = = = τadmis ; d2 = 3 d1 .
W2 R1 πd32 R1

to c) For the twist angle in x and y we obtain

l1 M1 32M1 l1 32M2 l2
Δϑ1 = = , Δϑ2 = ϑ2B = .
GIT1 πGd41 πGd42
With the continuity of the rotations
ϑ1B R1 = −ϑ2B R2
and
ϑC = ϑ1B + Δϑ1 R1 ϑ1B

we compute −ϑ2B
1  2 2
32M1 R2 3 R2
ϑC = l1 + l2 .
Gπd41 R1
 124 Torsion

P4.9 Problem 4.9 A homoge-

neous, graded shaft with
circular cross section is clam- 2r1 2r2
ped at both ends and loaded A M0 B
by the torque M0 .
Compute the torques at the a b
support positions A and B as
well as the twist at the point
where M0 is applied.

1

2
MA MB
Solution The system is sta- M0
tically indeterminate because A C B
the support torques MA and
MB cannot be computed
solely from the equilibrium
conditions.

MA + MB = M0

By cutting the shaft at C constant torques are obtained in the regions

x and y . This results in the following twists

MA a MB b
ϑ1 = , ϑ2 = .
GIp1 GIp2

Geometric compatibility requires that the two angles of twist are iden-
tical:

ϑC = ϑ1 = ϑ2 .

Together with

π 4 π 4
I p1 = r , I p2 = r
2 1 2 2

we obtain

1 1
MA = M0 4 , MB = M0 4 ,
1 + r24a 1 + r41 b
r1 b r2 a

2M0 ab
ϑC = .
πG (br14 + ar24 )
 Circular cross section 125

Problem 4.10 A shaft is clam- P4.10

ped at both ends and loaded
l m0
along part b of its length l
by a constant distributed tor-
que m0 . x

Determine the function of twist a

angle and torque. b

Solution The external torque m(x) has a jump at position x = a. We

use the Macauley bracket to incorparate the discontinuous function.
With

m(x) = m0 < x − a >0

the differential equation for the twist angle follows

GIT ϑ = −m(x) = −m0 < x − a >0 .

Integrating twice yields

GIT ϑ = MT = −m0 < x − a >1 +C1

1 m < x − a >2 +C x + C .
GIT ϑ = − 2 0 1 2

The constants folllow from the boundary conditions

ϑ(0) = 0 ; C2 = 0 ,

1 m 0 b2
ϑ(l) = 0 ; C1 = .
2 l
Finally we obtain

m0 b2
' ( MT 2l m0 b(a + l)
b <x−a> 1
MT (x) = m0 b − , 2l
2l b
x a
quadr.
parabola
ϑ
' ( linear
1 m 0 b2 x < x − a >2
ϑ(x) = − .
2 GIT l b2
x
 126 Twist
r
P4.11 Problem 4.11 The depicted shaft y y
x R
with ring-shaped cross section is P l z P
clamped at one end. At the other end z
G
a rigid beam is attached. The beam A
is supported by two springs and uallow
B c
loaded by the forces P . Determine c
a) the maximum force Pmax for a
l/2 l/2
prescribed admissible displacement
uadmis (in z-direction) at point A, Given : uadmis = 2 cm , l = 2 m
b) position and value of the maxi-
mum shear stress in the cross section r = 5 cm , R = 10 cm
of the truss for P = Pmax . c = 106 N/m
G = 8 · 1010 N/m2
Solution to a) The system is statically indeterminate. We free the
system at point B leading to the twist of the shaft
MT l GIp
Δϕ = ; MT = Δϕ
GIp l MT
with (small twist angles)
P P
uadmis MT
Δϕ = = 0.2 .
l/2
Equilibrium of moments for the beam provides B

B : MT = lPmax − lFc , where Fc = c uadmis . Fc Fc

Eliminating Δϕ , MT and Fc yields

 GI 
p
Pmax = 2 3 + c uadmis .
l
With Ip = π(R4 −r 4 )/2 = 1.47·10−4 m4 and the given numerical values
we obtain
 
2 · 8 · 1010 · 1.47
Pmax = + 106
2 · 10−2 = 78.7 kN
104 · 8

to b) The shear stress assumes its maximum value at the outer pe-
rimeter of the cross section. The absolute value is computed by

MT = Pmax l − c uadmis l
= (78.7 − 103 · 0.02) 2 = 117.4 kNm
and
MT R 117.4 · 0.1 τmax
τmax = = = 79.8 MN/m2 .
Ip 1.47 · 10−4
 and displacement 127

Problem 4.12 The hollow shaft A P4.12

x and the solid shaft y are joint
by a bolt at A. 
1 GIT
1
2

GIT2
Determine the torque MT and
the twist angle β of the bolt af- a b
ter assembly for the case that
the ends of the shafts have an α β
angular difference of α in the
stress-free state.

Solution In the assembled state both shafts are loaded by the torque
MT . We cut the system at position A and determine the angle of twist
of x and y separately:

MT a MT b
ϑ1 = , ϑ2 = .
GIT1 GIT2

From the geometric compatibility in the assembled state

ϑ1 α
α − ϑ2 = ϑ1
ϑ2

1

2
and
MT MT

β = ϑ1

we obtain for MT and β

α 1
MT = GIT1 ,
a 1 + b IT1
a IT2

α
β = ϑ1 = .
I
1 + ab IT1
T2
 128 Distributed torque

P4.13 Problem 4.13 The thin-walled spar with ring-shaped cross section
(length l, shear modulus G, radius r, thickness t
r) is located in
the interior of an airplane wing. It is loaded by a distributed torque
mT (x) with mT (0) = 2m0 and mT (l) = m0 . The spar is clamped at
the fuselage.

2m0

m0

t
x
r
l

Determine
a) the torque MT (x) in the spar,
b) the distribution of the shear stress τ (x) and the maximum shear
stress τmax due to torsion,
c) the angle ϑl , by which the end of the wing at x = l rotates with
regard to the fuselage.

Solution to a) The distributed torque is given by

 x
mT (x) = 2 − m0 .
l
The torque follows by integration
  2 
x
MT (x) = − mT (x) dx + C1 = − 2x m0 + C1
2l
 and angle of twist 129

which leads with the boundary condition

MT (l) = 0

 
l 3
; − 2l m0 + C1 = 0 ; C1 = m0 l
2 2
to
 
x2 x 3
MT (x) = −2 + m0 l .
2l2 l 2

to b) For the thin-walled spar cross section the shear stresses are com-
puted using the second moment of area for torsion IT = 2πr 3 t:
 2 
MT m0 l x x 3
τ (x) = r= − 2 + .
IT 2πr 2 t 2l2 l 2

The maximum shear stress occurs at position x = 0 and its value is

given by
3 m0 l
τmax = .
4 πr 2 t

to c) With the second moment of area for torsion IT and the shear
modulus G we obtain for the twist
 2 
MT (x) m0 l x x 3
ϑ (x) = = − 2 +
GIT 2Gπr 3 t 2l2 l 2
as well as for the edge rotation
 3 
mo l x x2 3
ϑ(x) = − + x + C2 .
2Gπr 3 t 6l2 l 2
The integration constant is determined from the boundary condition
ϑ(0) = 0 to be C2 = 0. Thus the edge rotation ϑl at the end of the
wing yields (x = l):
 
mo l 2 1 3 m0 l 2
ϑl = ϑ(l) = − 1 + ; ϑ l = .
2Gπr 3 t 6 2 3Gπr 3 t
 130 Shear stress and

P4.14 Problem 4.14 A shaft with the

depicted thin-walled profile is
loaded by a torque MT .

a) Determine the shear stress in t

different sections of the profile. a
t t 2a
b) Compute the maximum t
admissible torque, such that
the admissible shear stress
t
τadmis is not exceeded.
2a

Solution The profile consists B

of two parts. For each part the
1 S 
2
following holds:
MT1 MT2
B
MTi
T = τ (s) · t(s) = , A B
2ATi

0
MTi 1 T
ϑi = = ds .
GITi 2GATi t
i

With the given values

π 2
AT1 = a , AT2 = 4a2
2

we obtain by considerating that the shear flux in section S is composed

of the contributions from the torques MT1 and MT2 :

'   (
1 MT1 πa MT1 MT2 2a
ϑ1 = + − ,
πa2 G πa2 t πa2 8a2 t
'   (
1 MT2 6a MT2 MT1 2a
ϑ2 = 2 + − .
8a G 8a2 t 8a2 πa2 t

Inserting this result into the geometric compatibility

ϑ = ϑ1 = ϑ2
 admissible torque 131

yields

MT1 2+π
=
MT2 10 + 16
π
with

MT = MT1 + MT2

the torques

2+π
MT1 = MT = 0.254 MT , MT2 = 0.746 MT .
12 + π + 16
π
Now the stresses in the sections A, B and S follow

MT1 MT
τA = = 0.081 2 ,
2AT1 t a t MT MT
0.081 0.093
a2 t a2 t
MT2 MT
τB = = 0.093 2 ,
2AT2 t a t

MT MT
τS = τB − τA = 0.012 . 0.012
a2 t MT
a2 t 0.093
a2 t

Equalizing the maximum shear stress with the admissible shear stress

MT
τmax = τB = 0.093 = τadmis ,
a2 t

provides the maximum admissible torque

τadmis a2 t
MTadmis = 10 − 75 .
MT

Note: Inserting MT1 and MT2 in ϑ determines the second moment

of area for torsion IT = 13.7a3 t. Neglecting the section S, we obtain
IT = 13.6 a3 t. Thus section S only contributes a small amount to the
torsional rigidity.
 132 Displacement
F
P4.15 Problem 4.15 The fixed leaf spring l
(t
b) is eccentrically loaded by a
force F . A
x E, G
Compute the deflection at the point
F b
loading. Determine the maximum
normal and shear stress. x t
y
z
Solution The leaf spring is subjected to a bending and a torsion load.
Due to bending the deflection is given by the table on page 62.
F l3 bt3 F
wB = with I= .
3EI 12
The constant torque
wB
MT = F b/2 ϑ
wT
causes a rotation at the end of the spring
MT l 1 3
ϑ= with IT = bt
GIT 3

and the corresponding displacement wT = 2b ϑ. The total deflection is

thus obtained by
 
4F l3 3Eb2
w = w B + wT = 1 + .
Ebt3 16Gl2

Bending and torsion cause stress in the extreme fibre of the fixed cross
section σB , τT
M 6lF MT 3bF
σB = = 2 , τT = = . y
W bt WT 2bt2
z σB , τT
An area element at the top surface (z = −t/2) τT
is loaded as sketched. Thus the maximum nor-
σB σB
mal and shear stress follow
y

τT

 
x
σB σB  2 3F l b2
σ1 = + + τT2 = 2 1+ 1+ ,
2 2 bt 4l2



σ 2
B 3F l b2
τmax = + τT2 = 2 1+ .
2 bt 4l2
 and stress 133
10b
Problem 4.16 An element of P4.16
a bridge is constructed as F
a thin-walled (t
b) box
girder. During construction
the box girder is eccentrically F
loaded. 2t

Determine the location and

t t b
value of the maximum normal
and shear stress.
t
Solution Section properties 2b
of the profile are

2b2 t + 2 · 2b (b · t) 3 5 5
zs = = b, Sy (zmax ) = b t b = b2 t
8bt 8 8 8
 3   2  2
tb tb3 3 5
Iy = 2 + + 4bt b + 2bt b
12 64 8 8
37 3 zs
= tb , y
24 x
Iy 37 2 zmax
W = = tb ,
zmax 15
C
WT = 2AT tmin = 4b2 t . z

Using bending moment, torque, shear force in the clamped support

MB = −10 b F , MT = b F , Vz = F
τT
yields for the lower section
σB σB
MB 150 F
σB = =− ,
W 37 bt y
τT
MT 1 F V z Sy 15 F x
τT = = , τQ = = .
WT 4 bt Iy t 37 bt
The largest absolute value for the normal stress and the shear stress
are obtained by τ = τT + τQ at location C


σB σB  2 F
σ2 = − + τ 2 = −4.16 ,
2 2 bt


σB  2 F
τmax = + τ 2 = 2.13 .
2 bt
 134 Displacement
a C 2a
P4.17 Problem 4.17 The depicted
cantilever with thin-walled A B
circular cross section is clam- x F G/E = 3/8
ped at both ends and loaded
eccentrically at point C. C
Determine the deflection at y x
the point where the load is ap-
F r
plied and compute the normal
stress and the shear stresses z
due to torsion. a/2

1
M1 F aF M1
Solution The cantilever 
1 M2 2  2
is cut at point C. Equili-
brium yields M3
V1 V1 V2 V2
1
M2 = M3 + aF , V1 = V2 + F .
2
The deflection, the angle of bending, and the angle of twist are given
at point C by (see table on page 62):

V1 a 3 M1 a2 8V2 a3 4M1 a2
wC 1 = − , wC 2 = − − ,
3EI 2EI 3EI 2EI
 V1 a 2 M1 a  4V2 a2 2M1 a
wC 1
= − , wC 2
=+ + ,
2EI EI 2EI EI
M2 a 2M3 a
ϑ C1 = , ϑ C2 = − .
GIT GIT
Compatibility demands
 
w C 1 = wC 2 , wC 1
= wC 2
, ϑ C1 = ϑ C2

which renders
20 7 8
V1 = F , 2 =− F , M1 = aF ,
27 27 27
1 1
M2 = aF , M3 = − aF .
3 6
The second moments of area and the elasticity constants
G 3
IT = 2I = 2πr 3 t und =
E 8
 and stress 135

yield the deflection at the point of loading

a 26F a3
wF = wC 1 + ϑ C1 = .
2 81EI

To compute the stresses, we need the bending moments at A and B:

4 aF ,
MA = M1 − V1 a = − 9

2 aF .
MB = M1 + V2 2a = − 9

The maximum normal stresses due to bending in A, B and C are given

with the section modulus W = I / r

|MA | 4arF 2arF

σA = = , σB = ,
W 9I 9I
|M1 | 8arF
σC = = .
W 27 I

The shear stresse in secion x or y are calculated with WT = 2W = 2I

r :
M2 arF M3 arF
τ1 = = , τ2 = = .
WT 6I WT 12 I
τ1
The largest stresses occur at the point σA σA
A. An area element at the top surface
(analogously on the bottom surface) is y
loaded as sketched. For the principal τ1
stress and the maximum shear stress
x
we obtain


σA σA  2 arF
σ1 = + + τ12 = ,
2 2 2I


σA  2 5arF
τmax = + τ12 = .
2 18 I
 136 Verschiebung

P4.18 Problem 4.18 The depicted

cantilever is fixed at both a x y a
ends and bent by 90◦ . The
cantilever is loaded at point
C by the force F . z
A EI, GIT EI, GIT B
Compute the deflection at C
the point C. F

Solution To solve the pro-

blem we use superposition.
We cut the system at point 
1 
2
C and apply symmetry ar-
guments for the depicted
loading with respect to ben- M M M M
ding and torsion. At this sta- F F
ge the moment M is un- 2 2

known. From the table on

page 62 we deduce

 F a2 Ma F a3 M a2
wC = − , wC = − .
4EI EI 6EI 2EI

The angle of twist due to torsion at C is given by

Ma
ϑC = .
GIT

The geometric compatibility


wC1 = ϑC2

yields 
1 
2

Fa GIT C
M=
4 EI + GIT
wc
and the final result

wC1
F a3 4EI + GIT
wC = . ϑC2
24EI EI + GIT
 Verschiebung 137

Problem 4.19 The depicted semi- top view

P4.19
circular support is loaded at point F
A by a force F . A
Determine the deflection at the side view EI, GIT
point A.
a
F
Solution Equilibrium of moments
provides the bending moment MB
and the torque MT MT MB
MB (ϕ) = −aF sin ϕ ,
MB
s MT a
MT (ϕ) = a(1 + cos ϕ)F .
ϕ ϕ F
The angle of twist is given by
a cos ϕ a sin ϕ
dϑ MT
= mit ds = adϕ .
ds GIT
The twist dϑ at position ϕ causes the deflection at A
dwT A = a sin ϕ dϑ .

Combining the previous results and integration yields the deflection

due to torsion
 π
F a3 2F a3
wT A = dwT A = sin ϕ(1 + cos ϕ)dϕ = .
GIT GIT
0
The deflection due to bending is follows from
d 2 wB d 2 wB F a3
EI 2
= −MB ; 2
= sin ϕ ,
ds dϕ EI
dwB F a3 F a3
= (− cos ϕ + C1 ) , wB (ϕ) = (− sin ϕ + C1 ϕ + C2 )
dϕ EI EI
and the boundary conditions

wB (0) = 0 ; C1 = 1 , wB (0) = 0 ; C2 = 0 .

Using these constants yields

F a3
wB (ϕ) = (ϕ − sin ϕ) .
EI
Finally the total deflection at A is given at position ϕ = π
 
F a3 EI
wA = wT A + wB (π) = π+2 .
EI GIT
 138 Shear stresses
q = 20 kN/m
P4.20 Problem 4.20 A cantilever beam
with the depicted profile is subjec-
ted to an eccentric line load q. De- x
termine at the clamped support l=6m
a) the largest shear stress due to z
the shear force and its position, 3.5
b) the shear stress due to torsion. q
1.2
c) the distribution of the shear
stresses due to shear force and tor- 1.2 y0 z0 1.2
sion across the profile. Determine y C 20
position and value of the largest 1.2 1.2
shear stress. z

10 35 10
[cm]

Solution We start by computing the stress resultants at the clamped

support:

Vz = q l = 20 · 6 = 120 kN ,
2 2
ql 6
My = − = −20 · = −360 kNm ,
2 2
MT = q l · 3.5 cm = 20 · 6 · 0.035 = 4.2 kNm .
With the geometric data of the profile we calculate the position of the
centroid C and the second moment of area Iy :
3
zi A i 2 · (20 · 1.2) · 10 + 2 · (10 · 1.2) · 20
zo = 3 = = 8.42 cm ,
Ai 35 · 1.2 + 2 · 20 · 1.2 + 2 · 10 · 1.2
zu = 20 − zo = 11.58 cm ,
 bi h3i  zo C
Iy = + Ai z̄i2 zu y
12 z
203 · 1.2
= (35 · 1.2) · 8.42 + 2 ·
2
12
+2 · (20 · 1, 2) · 1.582 + 2 · (10 · 1.2) · 11.582

= 7915.8 cm4 .
 due to shear force and torsion 139

to a) The shear stress due to the shear force is obtained by

V z Sy 120
τ = = Sy = 0.01263 Sy .
Iy h 7915.8 · 1.2
The static moment Sy reaches its maximum at z = 0:
35 1
Sy max = S(z = 0) = 8.4 · 1.2 · + 8.42 · 1.2 = 218.7 cm3 .
2 2
From this result the maximum shear stress due to shear force follows

τV max = 0.01263 · 218.7

; τV max = 2.76 kN/cm2 = 27.6 N/mm2 .

to b) The shear stress due to torsion is calculated using the second

moment of area for torsion respectively the torsion modulus of the
profile:
1 3 1
IT = hi ti = (35 + 2 · 20 + 2 · 10) · 1.23 = 54.7 , cm4
3 3
3 3
1 hi ti 54.7
WT = = = 45.6 cm3 .
3 tmax 1.2
With the already calculated torque MT we obtain
MT 4.2 · 102
τT = =
WT 45.6

; τT = 9.21 kN/cm2 = 92.1 N/mm2 .

to c) The largest shear stress oc-

curs at the position z = 0. It τV
is distributed linearly across the
wall thickness with the following
extreme values:
τT
τinside = 27.6 − 92.1 = −64.5 N/mm2 ,

τoutside = 27.6 + 92.1 = 119.7 N/mm2

τV +T
; τmax = 119.9 N/mm . 2
 140 Shear stresses 300 kN
P4.21 Problem 4.21 A thin-walled box
girder is loaded by a force of 300 kN. 
A
Determine for the cross section at x
position  A
z
10 m
a) the stress distribution (normal 20 m
and shear stresses) due to shear for- 300 kN
ce and torsion,
1.5
b) the position of the maximum 2.0 2.0
principal stress and y 80
c) the value and direction of the 
a z 1.5
principal stress at the vertex a of 300 [cm]
the profile. 2 cm
Remark: Assume for the torsional
load case a fork bearing at the left
end.
1.5 cm

a

Solution The second moment of area is given by

 bi h3i  2 · 803
Iy = + Ai z̄i2 = 2· +2·(1.5·300)·402 = 1.611·106 cm4 .
i
12 i
12
The stress resultants at position 
A (or directly left of it) are

300 300 · 20
Vz = = 150 kN , My = = 1500 kNm ,
2 4
MT = 300 · 1.5 = 450 kNm .

to a) The normal stress is linear across the height of the

cross section and reaches in point 
a the value

My 1500 · 1000 · 1000

σx = za = · 40 · 10 = 37.25 N/mm2 .
Iy 1.611 · 106 · 104
37.25 N/mm2
The shear stresses due to Vz are determined by the zh-line and Sy -line.

60 −9000
−80 s s −80
zh − line [cm2 ] Sy − line [cm3 ] −10600
60
 and principal stresses 141

By using the Sy -line we obtain

V z Sy 150 Sy Sy
τV = = = 9.3 · 10−5 kN/cm2 .
Iy h 1.611 · 106 h h

At position  5.6
a they assume the value
4.2
150 · 9000
τV a = 4.9 τV [N/mm2 ]
1.611 · 106 · 1.5
4.2
= 0.56 kN/cm2 = 5.6 N/mm2 .
5.6

The shear stresses due to torsion are given by

MT
τT = , AT = 300 · 80 = 24000 cm2
2AT h
450 · 103 · 103
; τT a = = 6.25 N/mm2 .
2 · 24000 · 1.5 · 103

to b) The maximum shear stresses oc- 

b

cur at points a and b , the maximum

normal stresses at point  a . Thus the
principal stresses assume the largest 
a

value at 
a.

to c) In point 
a the shear and normal
stresses are: τa
2
τa = τV a +τT a = 5.6+6.25 = 11.85 N/mm , σx

σx = 37.25 N/mm2 . y
The principal stresses are given by
x

σx σx
σ1 = + ( )2 + τa2 = 40.7 N/mm2 ,
2 2

σx σx 2
σ2 = − ( ) + τa2 = −3.45 N/mm2 .
2 2
For the direction of the principal stress σ1 we compute
2τ
tan 2α0 = = 0.636 ; α0 = 16.23◦ .
σx
 142 Bending, shear force and torsion

P4.22 Problem 4.22 A cantilever beam F

with thin-walled T-profile (t
a)
is eccentrically loaded by a force F . x
The clamped support is designed l
such that warping is allowed. z
Determine the maximum stresses
due to bending, shear force and tor- F
sion. At which position do they oc- t b
cur?
C
Given: t = a/10, l = 20 a y
z 2a
t
a a

Solution We start by determining the following geometric properties

of the profile:
a
b= ,
2  
t(2a)3 1 I 1
I = b2 2at + + b2 2at = a4 , W = = a3 ,
12 6 3a/2 9
b at 9 3
SC = b 2at + = a ,
2 2 80
1 4 IT 4 3
IT = 2(2a)t3 = a4 , WT = = a .
3 3000 t 300
The bending moment reaches its maximum at the clamped support
(x = 0), while shear force and torque are constant along the beam:

Mmax = −lF = −20aF , V =F, MT = aF .

We compute the maximum bending stress (compression, at the lower

surface, at x = 0), the maximum shear stress due to shear force (at the
centroid C), and the shear stress due to torsion (at the outer boundary
of the flanges):
|Mmax | 20aF F
σmax = = 1 3 = 180 2 ,
W 9
a a
C SC F 9 a3 27 F
τVC = = 1 480 1 = ,
It 6
a 10
a 4 a2
MT aF F
τ MT = = 4 3 = 75 2 .
WT 300
a a

Note: The shear stress due to shear force is small compared to the
shear stress due to torsion.
Chapter 5
Energy Methods
5
144 Energy methods

Energy theorem
The work W done by the external forces (moments) during loading of
an elastic body is equal to the strain energy Π stored in the body:

W =Π.

The specific strain energy can be written in index notation for three-
dimensional problems of elastostatics:
E ν ! 1 . /
Π∗ = εik εik + ε2ii = (1 + ν) σik σik − ν σii
2
2(1 + ν) 1 − 2ν 2E
3
3 3
3 3
3
with εik εik := εik εik and εii := εii .
i=1 k=1 i=1

The following expressions hold for bars and beams:

strain strain-
loading energy per
length unit energy

1 N2 1 N2
tension / compression Π∗ = Π= dx
2 EA 2 EA
l

1 M2 1 M2
bending Π∗ = Π= dx
2 EI 2 EI
l

2 V2
shear Π∗ = 1 V Π= 1 dx
2 GAS 2 GAS
l

1 MT2 1 MT2
torsion Π∗ = Π= dx
2 GIT 2 GIT
l

Total strain energy (tension + bending + shear + torsion):

   
N2 M2 V2 MT2
Π= dx + dx + dx + dx .
2EA 2EI 2GAS 2GIT
l l l l

N 2l
special case: bar (N = const, EA = const): Π= .
2EA

 Si2 li
special case: truss system Π= .
i
2EAi
 Principle of virtual forces 145

Remark: For slender beams, the shear contribution can be neglected

compared to the bending contribution.

Principle of virtual forces

The displacement of a point due to tension, bending, shear, and torsion
can be computed from
   
NN MM VV MT M T
fi = dx + dx + dx + dx .
EA EI GAS GIT
l l l l

where
fi = displacement (rotation) at position i,
N, M, V, MT = stress resultants due to the external loads,
N, M, V , MT = stress resultants due to a virtual force (mo-
ment) “1”at position i in direction of fi .
Since the shear contributions are usually small compared to the other
contributions, they will be neglected in the following problems.
Special case truss:
 Sk S k
fi = lk ,
EAk
k

Special case bending of beams:


MM
fi = dx .
EI
l

Application to statically determinate problems

F
To compute the displacement i
fi at an arbitrary position i,
fi
the bending moment due to
the external loads (M ) and F
due to the virtual load (M ) i
have to be determined.
-
The integral M M dx can be M
evaluated by resorting to the “1
i
tabulated values on page 146.
M
146 Table of

Mk
k k k k
Mi s s s

i i 1 1
sik sik sik
s 2 2

i 1 1 1
sik sik sik
s 2 3 6

i1 i2 s s s
(i1 + i2 )k (i1 + 2i2 )k (2i1 + i2 )k
s 2 6 6

i 2 1 1
sik sik sik
s 3 3 3
quadratic parabola

i 2 5 1
sik sik sik
s 3 12 4

i 1 1 1
sik sik sik
s 3 4 12

i 1 1 1
sik sik sik
s 4 5 20
cubic parabola

i 3 11 1
sik sik sik
s 8 40 10

i 1 2 7
sik sik sik
s 4 15 60

Quadratic parabola: –◦– "

= apex of the parabola,
Cubic parabola: –◦– "
= root of the linear load q(x).
 -
integrals Mi Mk dx 147

αs βs
quadratic parabola
k1 k2 k k k
s s s s

si 1 2 2
(k1 + k2 ) sik sik sik
2 2 3 3

si 1 1 1
(k1 + 2k2 ) sik(1 + α) sik sik
6 6 3 4

s sk
(2i1 k1 + i1 k2 [(1 + β)i1 sk sk
6 6 (i1 + i2 ) (5i1 + 3i2 )
+2i2 k2 + i2 k1 ) +(1 + α)i2 ] 3 12

si (k + k ) 1 sik(1 + αβ) 8 sik 7 sik

1 2
3 3 15 15

si sik 7 11
(3k1 + 5k2 ) (5 − β − β 2 ) sik sik
12 12 15 30

si (k + 3k ) sik (1 + α + α2 ) 1 sik 2 sik

1 2
12 12 5 15

si sik 2 1
(k1 + 4k2 ) ( 1+α)(1+α2) sik sik
20 20 15 12

si sik α3 11 29
(4k1 + 11k2 ) (1+α+α2 − ) sik sik
40 10 4 15 120

si (7k + 8k ) sik (1+α)(7 −α2) 1 sik 1 sik

1 2
60 20 3 5 6

Trapezoids: individual i- or k-values can be negative.

148 Method of Castigliano

Application to statically indeterminate problems

The statically redundant (un-
F
known) force X = B is computed
B from the kinematic constraint, that
A the displacement has to vanish at
= F
point i of X (support):
A
X=B
fi = 0 .

MM
The relation fi = dx
“0”-System EI
F l
yields with
M (0) M = M (0) +X M (1) , M = M (1)
“1”-System
the statical redundant force
“1
x - (0) (1)
M (1) M M dx
X = B = − - (1) (1) .
M M dx

The integrals can be evaluated by resorting to the values tabulated on

page 146.
Remark: In n-fold statically indeterminate problems, n statically red-
undant (unknown) forces/moments Xi occur. They are determined from
n kinematic constraints (e. g. fi = 0).

Method of Castigliano
The derivative of the strain energy with respect to the external force
(moment) Fi is equal to the displacement (rotation) fi in the direction
of the force (moment) at the point where the force (moment) is applied.

∂Π
fi = .
∂Fi

Reciprocity theorem of Maxwell and Betti

F F
i k
fik k fik = fki
i fki
 Conservation of energy 149

Problem 5.1 The depicted sys- P5.1

tem is made of trusses with
a
identical axial rigidity EA.
Determine the vertical displace-
ment f of the force F . F
a a a
Solution The problem is solved using conservation of energy W = Π .
To assume the value f the force has to do the work W = 12 F f . The
strain energy Π is calculated by

1  Si2 li 1  2
Π= = Si li .
2 EAi 2EA
Knowing the reaction forces Si2 li
A = F/3 and B = 2F/3 the i li Si
F 2a
normal forces in the truss sys- √ √ √
tem can be tabulated 1 2a − 2F/3 2 2/9
4 2 a F/3 1/9
3 a F/3 1/9
1 8
4 a −F/3 1/9
3 5 7 √ √ √
5 2a − 2F/3 2 2/9
2 6 9
6 a 2F/3 4/9
A F B 7 a F 9/9
√ √ √
Thus we compute 8 2a −2 2F/3 8 2/9
√ 9 a 2F/3 4/9
4 (5 + 3 2)  2 4 √
f= Fa . Si li = (5 + 3 2)F 2 a
9 EA 9

Alternatively the method of Castigliano can be applied. Using the strain

energy
√
1  Si2 li 2 (5 + 3 2) 2
Π= = F a
2 EAi 9 EA
and the condition f = ∂Π/∂F we get
√
∂Π 4 (5 + 3 2)
f= = Fa .
∂F 9 EA
 150 Conservation of energy

P5.2 Problem 5.2 A beam (flexural rigidity

EI, axial rgidity EA → ∞) is loaded EA
by the force F and supported by the a
inclined rope (axial rigidity EA).
F
Compute the vertical displacement f
in force direction. EI
a a

Solution The problem can be solved by using the energy theorem

W = Π.

The work of the external force F is given by

1
W = F f.
2
The strain energy consists of beam bending and tension in the rope:
Π = ΠS + ΠB .
S
F
With
√ 45◦
 2 4 A
A : 2aF − aS = 0 ; S= √ F B
2
√ 2 x
2 aF
↑ : AV + S − F = 0 ; AV = −F
2
and
M
M (x) = −F x (0 ≤ x ≤ a)

we obtain for the rope

S2 l √ F 2a
ΠS = =4 2
2EA EA
and for the beam (using the symmetry of M (x))
 a 2 2
M2 F x 1 F 2 a3
ΠB = dx = 2 dx = .
2EI 2EI 3 EI
0
Finally, the energy theorem yields
2 F a3 √ Fa
f= +8 2 .
3 EI EA

Note: In section AB of the beam exists a compressive normal

force N = −2F . The corresponding contribution to the strain energy
is zero, because the beam is assumed to have infinite axial rigidity.
 Computation of displacements 151

Problem 5.3 In the depicted truss I II III P5.3

is loaded by the force F . All 5 6
trusses possess the same axial
rigidity EA. 1 4 9 a
2 7
Compute the vertical and hori-
zontal displacement of node III. 3 8 VI
IV V F
a a

Solution Using the principle of virtual forces the displacements follow

from
 Si S i 1 
f= li = Si S i li .
EAi EA
“1”
As the system is statically determinate,
we can compute the forces in the truss S i (V )
members Si due to the load F from
equilibrium considerations.
Loading node III by virtual forces “1” “1”
in vertical or horizontal direction, pro- S i (H)
vides the forces S i (V ) or S i (H) in the
truss members, respectively.

i li Si S i (V ) Si S i (V ) li S i (H) Si S i (H) li
1 a −F −1 Fa 0 0
√ √ √ √
2 2a 2F 2 2 2F a 0 0
3 a −2F −2 4F a 0 0
4 a 0 0 0 0 0
5 a F 1 Fa 1 Fa
6 a F 1 Fa 1 Fa
√ √ √ √
7 2a − 2F − 2 2 2F a 0 0
8 a 0 0 0 0 0
9 a F 0 0 0 0
 (V )
√  (H)
Si S i li = (7 + 4 2)F a Si S i li = 2F a

This leads to the vertical and horizontal displacements:

√ Fa Fa
fV = (7 + 4 2) , fH = 2 .
EA EA
 152 Principle of F

P5.4 Problem 5.4 The depicted plane C

frame (flexural rigidity EI) is
subjected to two point forces F . F
B 2a
Compute for the rigid corner C
a) the horizontal displacement, A
b) the vertical displacement,
c) the rotation. 2a

Solution We use the principle of virtual forces, neglecting shear, ten-

sion and torsion contributions:

MM
fi = dx
EI
l

The bending moment M due to the external forces F is sketched below

F
Fa

F M

F F

to a) Horizontal displacement of the corner: We apply a horizontal

virtual force “1” at the corner C and determine the associated bending
moment.

2a
“1”

1 1

Using integration in sections together with the tabulated values on page

146 yields:
 virtual forces 153

  a 2a
1 2
fH = M M dx = M M dx + M M dx
EI EI
a
 0

2 1 a
= (a)(a)(F a) + (a + 2a)F a
EI 3 2
11 F a3
= .
3 EI

to b) Vertical displacement of the corner: Application of a vertical

virtual force “1” yields no loading due to bending and thus no displa-
cement:
“1”

M M =0, fV = 0 .

to c) Rotation of the corner: The virtual moment “1”, applied at corner

C, yields the following bending moment M :

1
“1”

M
1
2a
1
2a
For the rotation ψ of the corner we obtain by use of the table on page
146:
 a 2a
1
ψ= M M dx + M M dx
EI
0 a
 2 
1 Fa 1 1 1
= (1 + ) + a F a
EI 2 2 3 2
2
11 F a
= .
12 EI
 154 Computation of displacements

P5.5 Problem 5.5 Determine the vertical q0 B

displacement fB and the rotation ψB
in point B at the end of the frame. EI
All beams in the frame are rigid in a
axial direction EA → ∞.

A
a

Solution Using the principle of virtual forces we can compute the dis-
placement and the rotation from

MM
f= dx .
EI
For the original and the auxiliary system we get:

“1” “1”
q0

q0 a2
2 a 1

M MV MW

Using the table on page 146 we compute

  
1 1 a q0 a2 q0 a2 5 q0 a4
fB = M M V dx = · ·a+a· ·a = ,
EI EI 4 2 2 8 EI

  
1 1 a q0 a2 q0 a2 2 q0 a3
ψB = M M W dx = · ·1+a· ·1 = .
EI EI 3 2 2 3 EI
 by the principle of virtual forces 155

Problem 5.6 The depicted frame P5.6

is constructed with beams having
identical flexural rigidity EI. F b
a
Compute the vertical and hori-
zontal displacement at the loading
point.

b
a

Solution According to the principle of virtual forces the displacements

are obtained from

MM
f= dx .
EI

The bending moments M due to the load and M V and M H due to the
auxiliary loads are given below:

“1
F “1

a a a
F
b b 1 b
a−b a−b a
F
b b b

Fa a

M MV a MH
b a−b
Fa a

Using the table on page 146 we compute

' (
1 1 1 F a3  4b 
fV = a(−F a)(−a) + b(−F a)(−a) + b(F a)a = 1+ ,
EI 3 3 3EI a
' (  
1 1 1 F a2 b 3b
fH = b(−F a)(−b) + b(F a)a = 1+ .
EI 2 3 3EI 2a
 156 Computation of displacements

P5.7 Problem 5.7 The sketched system a

consists of a clamped beam (EA →
∞) with flexural rigidity EI and 1
a
two bars of identical axial rigidity
EA. EA
EI
2
Compute the vertical and horizon-
F a
tal displacement at the point of load
application.

Solution The beam is subjected to bending, while the bars experience

tension or compression. We compute the displacements based on the
principle of virtual forces
  Si S i
MM
f= dx + li .
EI i
EAi

As the system is statically determinate, we obtain M and Si from equi-

librium conditions:
M

√
S1 = 2F , F

S2 = −F .
MA = aF
aF
AV = F

The vertical displacement can be calculated by loading the system with

a force “1” in vertical direction. By replacing F by “1”the above results
can be used:

MV

√
S 1V = 2,

S 2V = −1 .
a
 in mixed systems 157

Using the table on page 146 yields

' (
1 1
fV = a(aF )a + a(aF )a
EI 3
1 4√ √ √ 5
+ 2F · 2 · 2a + (−F )(−1)a
EA
√
4 F a3 (1 + 2 2)F a
= + .
3 EI EA

To compute the horizontal displacement we use the following auxili-

ary system:

MH
1

S 1H = 0 , “1”
2

S 2H = 1 .
AH = 1 a
MA = a
AV = 0

The displacement fH follows from evaluation

' (
1 1 1
fH = a(aF )a + 0 + {0 + (−F ) · 1 · a}
EI 2 EA

F a3 Fa
= − .
2EI EA

EI 1
Note: For = it holds fH = 0. For a rigid beam (EI → ∞)
a2 EA 2
the load application point moves to the left (fH < 0).
 158 Computation of displacements

P5.8 Problem 5.8 Determine the verti- sinusoidal q0

cal displacement f in the center of load
the beam.

A EI C
a a B
2 2

Solution The principle of virtual forces yields the vertical displacement


MM
f= dx .
EI
The bending moment M due to the given loading is computed via in-
tegration:
π  a π  a2 π 
q(x) = q0 sin x , V = q0 cos x , M = q0 2 sin x .
a π a π a
For the virtual load “1” it follows:
“1”
1 1
Ax = x, x ≤ a/2
M= 2
a−x A C B
B (a − x) = , x ≥ a/2
2 x

The vertical displacement ist then obtained

 1 a/2 a   2
MM 1 x a−x
f= dx = M dx + M dx .
EI EI 2 2
0 a/2
Integration with

sin cx x cos cx
x sin cx dx = −
c2 c
renders the result
⎧⎡ ⎤a/2
⎪ π π
2 ⎨ sin( x) x cos( x)
q0 a ⎢ a a ⎥
f= ⎣ − π ⎦
2 EI π 2 ⎪⎩ π 2
a
2
a
⎤a ⎫
0
⎡
π π ⎪
⎪
sin( x) x cos( x) ⎬
⎢ −a
2
π a a ⎥ a4 q0
+⎣ cos( x) − + π ⎦ = .
π a π 2
⎪
⎪ π 4 EI
2 a ⎭
a a/2
 by the principle of virtual forces 159

Problem 5.9 The sketched arch is F P5.9

clamped and subjected to a verti-
cal force F .
EI
Compute the vertical and horizon-
tal displacement at the point of
loading. Only deformations due to R
bending shall be considered. s

Solution We use the principle of

virtual forces. The displacements F
are computed from
Q R(1 − sin ϕ)

MM M
f= ds .
EI
N
The bending moment M is obtai- ϕ
ned from equilibrium considerati-
ons
R cos ϕ
M = −F R cos ϕ . “1”

To determine the vertical displacement we ap-

ply a force “1” in vertical direction. This yields

M V = −R cos ϕ

together with ds = R dϕ the displacement

π/2 π/2
R F R3 πF R3
fV = M M V dϕ = cos2 ϕ dϕ = .
EI EI 4EI
0 0
“1”
The auxiliary force in horizontal direction cau-
ses the bending moment

M H = −R(1 − sin ϕ)

and the displacement


π/2 
π/2
R R3 F F R3
fH = M M H dϕ = (cos ϕ − sin ϕ cos ϕ) dϕ = .
EI EI 4EI
0 0

1 (1 + cos 2ϕ) and

Note: For the integration the two relations cos2 ϕ = 2
1
sin ϕ cos ϕ = 2 sin 2ϕ were used.
 160 Forces and displacements

P5.10 Problem 5.10 The depicted A

truss is made of members 2 7
with identical axial rgidity 1 3 5 6
8 a
EA.
Determine the normal forces B C
4 9
in the truss members and F
the vertical displacement
under the load. a a

Solution The support of the truss is statically indeterminate. We use

the reaction force C as statically redundant force and compute its value
from the support constraint
 Si S i li 1 
fC = = Si S i li = 0 .
EAi EA
Here we only compute the normal forces in the “0”-system. The compu-
tation of these forces in the “1”- and “2”-system follow by an analogous
procedure.
“0”-system: “1”-system:
I III I III
A 0 A 0
2 7 2 7
0 0
1 3 5 0 6 8 1 3 5 6 8
0
B 0 0 C B C
4 II 9 4 II 9
F “1”

For example at node I (“0”-system):

√ √
↑: S3 = − 2S5 = − 2F S2
S3 F
→: S2 = F

For example at node B (“0”-System):

√ S1
2
↑: S1 = − S3 = F F S3
2
(0) (1) (1)
With Si = Si + C · Si and S i = Si it follows
 (0) (1) √
Si Si li 3+2 2
C = − (1) (1)
= √ F .
Si Si li 7+4 2
 in statically indeterminate systems 161

(0) (1) (0) (1) (1) (1) (2)

i li Si Si Si Si li Si Si li Si
1 a F −1 −F a a 1
2 a F −2 −2F a 4a 1
√ √ √ √ √ √
3 2a − 2F 2 −2 2F a 2 2a − 2
4 a 0 1 0 a 0
5 a F 0 0 0 1
√ √ √
6 2a 0 − 2 0 2 2a 0
7 a 0 0 0 0 0
8 a 0 0 0 0 0
9 a 0 1 0 a 0
  √   √ 
= −3 − 2 2 F a 7+4 2 a

From this table we deduct the forces in the truss members

√ √
4+2 2 1 4+4 2
S1 = √ F , S2 = √ F , S3 = − √ F,
7+4 2 7+4 2 7+4 2
√ √
3+2 2 4+3 2
S4 = S9 = √ F, S5 = F , S6 = − √ F, S7 = S8 = 0.
7+4 2 7+4 2

To compute the vertical displacement at the loading point we consider

the system as a statically determinate system loaded by F and C, which
fulfills the support constraint fC = 0. For this situation we know the
forces Si .
“2”-system:
0
2 7
0
1 3 5 0 6 8

0 0
4 9
F C
“1”
(2)
With the forces S i = Si of the auxiliary system “2” we obtain
1 
fF = Si S i li
EA
Fa √ √ √ √ √ !
=  √  (4 + 2 2) + 1 − (4 + 4 2)(− 2) 2 + (7 + 4 2)
EA 7 + 4 2
√
20 + 14 2 F a
= √ .
7 + 4 2 EA
 162 Support reactions
2a
P5.11 Problem 5.11 Determine the ben-
q0
ding moment and the horizontal
displacement fH of the support EI a
B of the depicted frame structure EI
2a
B
A
Solution The system is statically
indeterminate. To determine sup-
port reactions we use the principle of virtual forces, where we consider
the moment X = MA as statical redundant reaction. Thus we obtain
the following bending moments and support reaction in the “0”- and
“1”-system:
1
q a2
2 0
“0”-system: q0
(0)
AH =0,

M (0)
(0)
AV = q0 a ,
(0)
AH B (0)
(0)
B = q0 a .
(0)
AV

1
“1”-system:
(1)
AH = 0 ,
M (1)
(1) 1 (1)
AV = − , AH B (1)
2a “1
1 (1)
B (1) = . AV
2a

The condition that the rotation at position A has to vanish


MM
ϕA = 0 = dx ,
EI
yields with

M = M (0) + X M (1) and M = M (1)

 and displacements 163
 
- (0) (1) 1 1
· 2a q0 a2 · 1
M M dx 3 2 q0 a2
X = MA = − - (1) (1) =− =− .
M M dx 1 8
· 2a · 1 · 1 + 2a · 1 · 1
3
The support reactions and the bending moment follow as

15
(0) (1) a
AH = AH + X · AH = 0 , 1
q a2
8 0
16

(0) (1) 17
AV = AV + X · AV = q0 a ,
16 0, 44 q0a2
15
B = B (0) + X · B (1) = q0 a . M
16

To compute the horizontal displacement at B, the frame is considered

as a statically determinate system loaded by q0 and X = MA . At this
system an auxiliary force is applied (“2”-system) rendering the followi-
ng bending moment:
“2”-system:
2a a

“1”
M (2)
B

A
Using

M = M (0) + X · M (1) and M = M (2)

we compute with the table on page 146

 '  (
1 1
fH = M M dx = M (0) M (2) dx + X M (1) M (2) dx
EI EI
'  (
1 2a q0 a2 q0 a2 1 1
= (2a+a) − ·2a·1·(2 ·2a + a) + ·2a·1 · 2a
EI 3 2 8 6 2

13 q0 a4
= .
24 EI
 164 Support reactions

P5.12 Problem 5.12 Compute the re- F

action forces and the deflection EI
at points D and G.
Now an additional force of 2F A D B G C
is applied at D. How does the a a
deflection at G change? a a
2 2

Solution We apply the principle of virtual forces and use the reaction
force B as statically redundant force. Together with the bending mo-
ments in the “0”- and “1”-System
F

“0”-system: M (0) 1
aF 2
aF
A(0) = 13 F B (0) = 32 F 3 3

2
3
a 1
a
3
“1”-system:
“1” M (1)
A(1) = − 23 B (1) = − 13

the kinematic constraint fB = 0 yields the reaction force B:

 
1 a aF  2a  a aF  2a 
M (0) M (1) dx − + 2 −
EI  3 3 3 6 3 3
X = B =− =−
1 (1) (1) a 2a 2a 2a 2a 2a
M M dx +
EI 3 3 3 3 3 3

aF  a  2aF  a  2aF  2a  a 2aF  a 
− +2 − + − + −
3 3 3 3 3 3 3 3 3 7
+ = F .
a 2a 2a 2a 2a 2a 8
+
3 3 3 3 3 3
Furthermore, we compute
1 7 2 F 3
A = A(0) + X · A(1) = F− F · =− , C= F .
3 8 3 4 8

To determine the deflection we consider the beam as a statically deter-

minate system loaded by F and B. From the two auxiliary systems
“2”-system:
“1”

1 2 1
3 3 3
a M (2) 2
a
3
 and displacements 165

“3”-system: “1”

1
5 1 M (3) 1 6
a
6 6 5 3
a
12
a

we obtain

1
fG = [M (0) + X · M (1) ]M (2) dx
EI
'  (
1
= M (0) M (2) dx + X M (1) M (2) dx
EI
'   
1 2a 2aF 2a 7 a 2a a
= + F −
EI 3 3 3 8 3 3 3
  (
a 4a a 2a 2a 2a 2a a a a 2aF 2a
+ − − − − +
6 3 3 3 3 3 3 3 3 3 3 3

5 F a3
= ,
48 EI


1
fD = fDG = [M (0) + X · M (1) ]M (3) dx
EI
'  (
1 1 F a3
= M (0) M (3) dx + X M (1) M (3) dx =− .
EI 64 EI

The deflection at G due to the additional load 2F is computed from

the reciprocity theorem of Maxwell- fDG F
Betti. Based on this theorem the
deflection fDG at D due to the for-
D G
ce F in G is equal to the deflection
fGD at G due to the force F in D. F fGD
As a consequence of the force 2F
at D we obtain at G the deflection
2fGD . Thus, the total deflection at D G
G is given by
fGD = fDG

f = fG + 2fDG 2fGD =2fDG

2F
 
5 1 F a3 7 F a3
= −2 = .
48 64 EI 96 EI G
D
 166 Statically indeterminate

P5.13 Problem 5.13 The depicted q0

frame is loaded by a constant
line load q0 . The frame is
made of beams with identical a
flexural rigidityEI.
A B C
Determine the reaction forces
in the supports. a a

Solution The frame is two q0

times statically indetermina-
te. We consider the reaction
force B and the horizontal f1
force CH as statically redun- f2
dant forces to obtain the de- A
X1 = B X2 = C H
picted system. The unknown
forces X1 = B and X2 = CH
are computed from the kine-
matic constraints f1 = 0 and
f2 = 0.
Using the principle of virtual forces we construct the following basic
and auxiliary systems :

“0”-system:
q0
quadr. par.

1
M (0) q a2
2 ◦
(0)
AH

(0) (0)
AV CV
(0) (0) (0)
AV = q0 a , AH = 0 , CV = q0 a .
1
“1”-system: 2
a

M (1)
(1)
AH
(1) (1)
AV “1” CV

(1) (1) 1 (1)

A V = CV = − , AH = 0 .
2
 frame 167

“2”-system:
a a

M (2)
(2)
AH “1”

(2) (2)
AV CV

(2) (2) (2)

A V = CV = 0 , AH = 1 .

From conditions

1
f1 = [M (0) + X1 M (1) + X2 M (2) ]M (1) dx = 0 ,
EI

1
f2 = [M (0) + X1 M (1) + X2 M (2) ]M (2) dx = 0
EI
we obtain by using the table on page 146
 
5a q0 a2 a 5q0 a4 a3
M (0) M (1) dx = −2 =− , M (1) M (1) dx = ,
12 2 2 24 6
 
a3 2
M (1) M (2) dx = , M (0) M (2) dx = − q0 a4 ,
2 3

a 8
M (2) M (2) dx = 2 (−a)(−a) + 2a(−a)(−a) = a3
3 3
the following two equations
5q0 a4 a3 a3 2q0 a4 a3 8a3
− + X1 + X2 =0, − + X1 + X2 =0.
24 6 2 3 2 3
The solution is given by
8 1
X1 = B = q0 a , X2 = CH = q0 a
7 28

and the remaining support reactions follow

(0) (1) (2) 8 1 3
AV =AV + X1 AV + X2 AV = q0 a − q0 a · + 0 = q0 a ,
7 2 7
(0) (1) (2) 1
AH =AH + X1 AH + X2 AH = q0 a ,
28
8 1 3
CV =q0 a − q0 a · = q0 a .
7 2 7
 168 Moment and displacement
F
P5.14 Problem 5.14 An elastic circular arc
is loaded by two opposing forces F . R EI
Determine the bending moment and ϕ
the compression in the circular arc.
Assume that the arc is rigid with
respect to an axial deformation.
F
Solution We cut the arc at the mid
plane (at ϕ = 0, π) and realize that F
the system is internally statically
indeterminate (the stress resultants
MA A X = MA
cannot be determined from equilibrium
conditions). The unknown bending F F
moment X = MA is computed from 2 2
the fact that the slope at A has to vanish (symmetry!). Use of the prin-
ciple of virtual forces yields:
“0”-system: F “1”-system:

ϕ ϕ
F F
2 2 “1” “1”
1
M (0) = F R(1 − cos ϕ) , M (1) = 1 .
2
From

1
ψA = M M ds = 0
EI
we obtain with
M = M (0) + X · M (1) , M = M (1) , ds = R dϕ

for MA : -
π/2
FR
- (0) (1) 2 (1 − cos ϕ)R dϕ  
M M ds 2 1 1
X = MA = − - (1) (1)
0
=− = −F R − .
M M ds -
π/2 2 π
2 R dϕ
0

These intermediate results yield the bending moment in the range 0 ≤

ϕ ≤ π/2
 
1 2
M = M (0) + X · M (1) = F R − cos ϕ .
2 π
 for a circular arc 169

To compute the vertical displacement at the loading point we consider

the semicircular arc as being loaded by the force F and the moment MA
and to be simply supported (statically determinate). For this system
the bending moment M is known. Form the related auxiliary system
we obtain
“1”

1
M= R(1 − cos ϕ) . ϕ
2
With this result the displacement follows 1 1
π/2
2 2
π/2  
1 F R3 2
fF = 2 M M R dϕ = − cos ϕ (1 − cos ϕ)dϕ
EI 2EI π
0 0
 2  π/2
F R3 2 ϕ 1 F R3  8
= ϕ− + 1 sin ϕ + + sin 2ϕ = π− .
2EI π π 2 4 0 8EI π
The compression of the circular arc yields
F R3 π 2 − 8
Δv = 2fF = .
EI 4π

Using the theorem of Castigliano to solve the problem, we derive with


1 M2
M = F R(1 − cos ϕ) + MA and Π= ds
2 2EI
and by using the fact that the slope at A has to vanish
∂Π
ψA = =0
∂MA
the result
 π/2 
∂M 1
M ds = 0 ; 2 F R(1 − cos ϕ) + MA · 1 · R dϕ = 0
∂MA 2
0
   
1 1 1 2
; MA = −F R − and M = FR − cos ϕ .
2 π 2 π

The displacement fF is computed from


∂Π 1 ∂M
fF = = M ds
∂F EI ∂F
π/2
2

FR2  R  2  F R3  8
= − cos ϕ − cos ϕ R dϕ= π− .
EI 2 π 2 π 8EI π
0
 170 Support reactions

P5.15 Problem 5.15 The depicted pi- A

pe x is clamped on one side 2r B
and supported by the additional 
2
rope y . x 
1
l2
EA2
y
Determine the support reac-
tions in A and B, if the pipe is GIT 1
loaded by the force F . z EI1
F
l1
G1 3 EI1 l2 1 r 1
Given: = , = , = .
E1 8 EA2 l13 100 l1 10

Solution The system is statically indeterminate. We choose the reacti-

on force in B as statically redundant force. This leads to the following
“0”- and “1”-system:
“0”-System:
(0)
(0) MA
A(0) = F , MT 
1
x
x y
(0)
MA = −l1 F , F
A(0) z F z
(0)
MT = rF .
(0)
M (0) MT N (0)
l1 F
rF

2 
2 
2
N (0) =0

1 
1 
1

“1”-System:
“1” “1”
A(1) = −1 , (1)
(1) MA 
2 
2
MA = l1 ,
(1) MT 
1
x
y
(1) x
MT =r. A(1) z z
(1)
M (1) MT N (1) “1”
l1
r

2 
2 
2


1 
1 
1
 in combined systems 171

From the constraint, that the displacement at B has to vanish,

  
MM MT M T NN
fB = 0 = dx + dx + dx ,
EI GIT EA

we obtain with

(0) (1)
M = M (0) + X · M (1) , MT = MT + X · MT ,

(1)
N = N (0) + X · N (1) , M = M (1) , M T = MT , N = N (1)

the unknown force X = B

  (0) (1) 
M (0) M (1) MT MT N (0) N (1)
dx + dx + dx
EI1 GIT 1 EA2
X = B = −  (1) (1) 
M (1) M (1) MT MT N (1) N (1)
dx + dx + dx
EI1 GIT 1 EA2

1 1 1
l1 (−l1 F )l1 + l1 (rF )r + 0
EI1 3 GIT 1
=− .
1 1 1 1
l1 l1 l1 + l1 r r + l2 · 1 · 1
EI1 3 GIT 1 EA2

Using IT 1 = 2I1 (circular cross section!) and the given relations leads to

96
X=B= F .
107

The support reactions at A are given by

96 11
A = A(0) + X · A(1) = F − F = F ,
107 107

(0) (1) 96 11
MA = MA + X · MA = −l1 F + l1 F = − l1 F ,
107 107

(0) 96 11
MT A = MT + X · MT (1) = rF − rF = rF .
107 107
 172 Computation of displacements via

P5.16 Problem 5.16 Determine the second

moment of area Iy for the depicted stati-
cally indeterminate structure, such that q
the vertical displacement at point K is
exactly wK = 1 cm. EA = ∞ a

Given: E = 21 · 107 kN/m2 , K

a = 3 m, a a a
q = 5 kN/m.

Solution To determine the displacement at point K we first have to

compute the stress resultants in the statically indeterminate system.
For this a hinge is introduced at K.
“0”-system:

4
1
qa 3
qa
3
4 1
3
qa 3
qa
5
6
qa2
2
3
qa
2
3
qa
Q0 M0
2
3
qa2
“1”-system:

“1

1
a

1 +1
a

1
a

1 +1
a

Q1 M1
 the principe of virtual forces 173

The rotations at point K in the “0”- and “1”-system are given by

1 2 1 5 1 1 7 3
EIy δ10 = a · 1 · qa2 + a · 1 · qa2 + a · 1 · qa2 = qa ,
6 3 6 6 3 8 24
1  4
EIy δ11 =4· · a · 12 = a.
3 3
With
δ10 + X1 δ11 = 0 − 32
7
qa2
we determine the statically
redundant quantity
5
(bending moment at K) 6
qa2
7
X1 = − qa2 , − 32
7
qa2
32
2
which leads to the total bending moment. 3
qa2 M

To compute the displacement of point K we apply in the statically de-

terminate “0”-system a force “1”and compute the bending moments.

4 2
3 3

2
2 4
3
a
3 3

1 M̄
1 3
a
3 1

1 a 2  1 a 7 
EIy δ1K = 2 · a qa2 − a − qa2
3 33 6 3 32
1 2a  7 2   1 2a 5   1 2a 1 
+ a − qa + 2 a qa2 + 2 a qa2
6 3 32 3 3 6 3 3 8
4 7 14 10 1 
= qa4 − − + +
27 576 576 27 18
929 4
= qa .
1728
The required second moment of area results from condition δ1K = wK
1 929 1 929 5
Iy = qa4 = 3004 = 10368 cm4 .
EwK 1728 21 · 103 1728 100
 174 Principle of virtual forces

P5.17 Problem 5.17 The depicted beam with q0

flexural rigidity EI is statically indeter-
minately supported. A B
x
Compute the deflection at the center of a
the beam.

Solution We regard the reaction force in B as statical redundant force

and use the principle of virtual forces to determine B from the cons-
traint

MM
fB = dx = 0 .
EI
For the “0”- and “1”-system we obtain:
“0”-system:
q0 M (0) (x) = − 21 q0 (a − x)2
1
q a2
2 0

“1”-system:

a
“1” M (1) (x) = (a − x)

With the help of the table on page 146 we deduct

- (0) (1)
M M dx 3
X = B = − - (1) (1) = q0 a .
M M dx 8
To determine the vertical displacement, the beam is considered as a
simply supported beam on two supports. For this situation we compu-
te the bending moment due to the (“2”-system) under a virtual load.
“2”-system: “1”
M

a
4

With M = M (2) and M = M (0) + XM (1) it follows


1
fV = (M (0) + XM (1) )M (2) dx
EI
 
1 X
= (M (0) M (2) )dx + (M (1) M (2) )dx
EI EI
 
1 7 X 3 qa4
= − q0 a4 + a = .
EI 384 16 192 EI
 Combined structures 175

Problem 5.18 The sketched frame P5.18

(axial rigidity EA → ∞, flexural
rigidity EI) is closed by an elastic EA
truss (axial rigidity EA). The a
system is subjected to a constant EI EI
line load q0
q0
Compute the force in the truss. 2a

Solution The system is internally statically indeterminate. We choose

the force in the truss as statically redundant force. From the principle
of virtual forces follow the basic and auxiliary system:
“0”-system: q0 quadr. Par.

M (0)
AH =q0 a
1
q a2
2 0
AV = 14 q0 a B= 14 q0 a

“1”-System:
“1”
S (1) =1
M (1)

a a
a a
From the condition, that difference in the displacement of the frame
and the end of the truss has to vanish,

1 SS2a
Δf = M M dx + =0,
EI EA
follows together with

M = M (0) + X · M (1) , M = M (1) , S=X, S = S (1) = 1

the force in the truss


−
1
M (0) M (1) dx 1 1  1 1 
EI 2a q0 a2 (−a) + a q0 a2 (−a)
X=  = − 2 2  4 2
1 2a 1 2aEI
M (1) M (1) dx + 2 a(−a)(−a) + 2a(−a)(−a) +
EI EA 3 EA
15 1
= q0 a .
64 3EI
1+ 2
4EAa
 176 Principle of virtual forces

P5.19 Problem 5.19 The semicircular arc (fle-

xural rigidity EI) is supported by a bar EI
(axial rigidity EA) and loaded by a force a
F. EA
F
Compute the force in the bar and the
deflection at the connection point of bar A C B
and arc.
Solution The system is statically indeterminate. We use the principle
of virtual forces and the force in the bar as statically redundant force.
This leads to the following “0”- and “1”-system:
“0”-system:

M (0) (ϕ)=F a sin ϕ , a

AH = F ϕ F
(0)
S =0 .

“1”-system: “1

1 “1
M (1) (ϕ)= a(1 − cos ϕ) ,
2 ϕ
(1)
S =1 . (1) C (1) = 1
AV = 1/2 B (1) = 1/2

The difference in displacements of arc and bar has to vanish:


1 S Sa
Δf = M M dx + = 0 , with
EI EA

M = M (0) + X · M (1) , M = M (1) , S=X, S = S (1) = 1 .

This condition provides the force in the bar

 -
1 F a π/2
− M (0) M (1) dx 2 sin ϕ(1 − cos ϕ)dx
EI 2 0
S=  =−
1
M (1) M (1) dx +
a -
a2 π/2 aEI
EI EA 2 (1 − cos ϕ)2 dx +
4 0 EA
4F
=− .
EI
(3π − 8) + 8
EAa2

The deflection f of the arc is given by the deformation of the bar:

Sa 4Fa
f =− = .
EA EI
(3π − 8)EA + 8
a2
 Combined structures 177

Problem 5.20 The depicted frame x EI B P5.20

(axial rigidity EA → ∞, flexural EA
rigidity EI) is loaded by a force F . y a
F
The bar x has the axial rigidi-
rigid
ty EA, while bar y is considered
to be rigid.
EI a
Determine the force S2 in bar y A
and the vertical displacement vB at
point B. a a a

Solution The system is statically determinate supported, but due to

the rigid truss internally statically indeterminate. To compute the for-
ce S2 in bar y we use the following “0”- and “1”-systems. The reaction
forces and the force in bar x can be determined from equilibrium con-
ditions.

“0”-system: “1”-system:
S1 S̄1

F “1”

AH ĀH

AV ĀV
A V = F , A H = F , S1 = F ĀV = ĀH = S̄1 = 0 , S̄2 = 1
The bending moments M0 and M̄1 of both systems are sketched below:

“0”-System: “1”-System:
√
2
−2F a − a
2

M0 M̄1
178 Principle of virtual forces

When evaluating the principle of virtual forces we have to consider that

the bar is rigid. This yields
 √ √
1 1 1 2 5 2 3
α10 = M0 M̄1 dx = · a(−F a−4F a) a(−1)·2 = Fa ,
EI EI 6 2 6EI


1 1 1 1 a3
α11 = M̄12 dx = · a · a2 · 2 = ,
EI EI 3 2 3EI
and for the force in bar y we obtain
√ √
α10 5 2 3 3EI 5 2
S2 = X = − =− Fa · 3 = − F.
α11 6EI a 2

At point B the vertical displacement vB and the force F have the same
direction, hence we can use the energy theorem:

1 1 M2 1  Si2 li
F vB = dx + .
2 2 EI 2 i EA

Its application is based on the bending moment in the total system

M = M0 + X M̄1 :
1
Fa
2

−F a

M
Evaluation the integrals using the bending moment M yields
  
M2 1 1 2 2 1 2 2 2 1 1 1 2 2 7
dx = aF a + · aF a + · a F a ·2 = F 2 a3
EI EI 3 3 3 3 3 4 6EI
Furthermore, with the force S1 in the bar (note, bar y is rigid)
 Si2 li F 2a
=
i
EA EA

we compute the vertical displacement

 2 
7a 1
vB = + Fa.
6EI EA
 Combined structures 179

Problem 5.21 A trapezoidal, P5.21

frame (axial rigidity EA,
flexural rigidity EI) with two √
3
bars (axial rigidity EA) is EI l
2
loaded by a force F . EA EA

Compute the horizontal

and vertical displacement at F
the loading point. l l l l
2 2 2 2

Solution Since the vertical displacement vF of the load has the same
direction as load F , we can determine vF by the conservation of energy:

1 1 M2 1  Si2 li
F vF = dx + .
2 2 EI 2 i EAi

The structure is statically determinate supported. Thus reaction forces,

stress resultants, and the forces in the bars can be determined from
equilibrium conditions.
Fl Fl
4 4

1 2

F
F F M
2 2
F
S1 = S2 = √
3
Using the bending moment and the bar forces yields
+    2 , +  2 ,
2
1 1 Fl Fl 1 F
F vF = 2 l+ l + 2 √ l .
EI 3 4 4 EA 3

Thus we compute for the vertical displacement

11 F l3 Fl
vF = + .
12 EI 3 EA
180 Principle of virtual forces

The horizontal displacement at the loading point follows by loading

with a virtual “1”force in horizontal direction. Bending moment and
forces in the bars are computed from equilibrium
√ conditions.
3
l
2

1 2

“1 “1
M̄

S̄1 = 1
S̄2 = −1

Using the “0”- and “1”-system we determine the horizontal displace-

ment
  √ √ 
M M̄  Si S̄i li 1 1 Fl 3 1 Fl 3
uF = dx + = l·l+ l·l
EI i
EAi EI 3 4 2 2 4 2

with the final result

√
5 3 F l3
uF = .
48 EI

Note: The deformation of the bars and the right part of the frame
do not contribute to the horizontal displacement.
Chapter 6
Buckling of Bars
6
182 Stability

The total potential of elastic systems loaded by conservative forces

consists of an external potential Π(e) of the applied forces and the
potential (strain energy) Π(i) of the internal forces:

Π = Π(e) + Π(i) .

In an equilibrium state,

δΠ = 0

holds.
Formal application of the stability conditions for rigid bodies (see
book 1, chapter 7) to elastic systems yields
⎧
⎪
⎪ > 0 stable equilibrium,
⎪
⎨
2 2 (e) 2 (i)
δ Π=δ Π +δ Π = 0 indifferent equilibrium,
⎪
⎪
⎪
⎩
< 0 unstable equilibrium.

The critical load of an elastic system is reached, if the equilibrium is

indifferent. Besides the original equilibrium state, neighboring equili-
brium states exist related to deformation (“buckling”). Critical loads
and associated equilibirum states can be determined from equilibrium
conditions in the deformed state or by investigating δ 2 Π.
For an elastic bar under compression equilibrium conditions in the de-
formed state provide the differential equation for Euler’s column

EI F
F
w wIV + λ2 w = 0 , λ2 =
EI
x

with the general solution

w = A cos λx + B sin λx + Cλx + D .

The constants A, B, C and D are determined from the boundary condi-

tions for the kinematic and static quantities. Note that these conditions
have to be formulated in the deformed state. For example, under the
assumption of small rotations an elastic support at position x = 0 is
described by
 Buckling 183
x 
F F w (0)
w(0)
w  (0)
F w  (0)
c cw(0) N
Q(0)

Q(0) = c w(0) − F w (0) ; EIw (0) + c w(0) − F w (0) = 0 ,

M (0) = 0 ; EIw (0) = 0 .

Four characteristic boundary conditions establish the Euler buckling

cases:
F
11
00
F F
1
0 F

00
11
11
00 11
00
2
l Fcrit = π EI2 π 2 EI π 2
√ EI π 2 EI
(2l) l2 (l/ 1.43)2 (l/2)2
11
00
11
00
11
00
1. 2. 3. 4.

Approximate solutions for the critical load can be obtained by

using the ansatz w̃(x) in the energy functional for buckling (Rayleigh-
quotient):
-l 2
l EI w̃ dx
1  2 2

EI w̃ dx − F̃crit w̃ dx = 0
0
Π= ; F̃crit = .
2 -l
0 w̃2 dx
0

To determine F̃crit , the ansatz w̃(x) has to satisfy the essential (kinema-
tic) boundary conditions (note that the result for F̃crit improves, if w̃(x)
satisfies also the static boundary conditions). The approximate soluti-
on is in general on the unsafe side, because the inequality F̃crit ≥ Fcrit
holds.
Individual springs at position xi are included in the nominator by
c[w̃(xi )]2 , while torsion springs are incorporated by cT [w̃ (xi )]2 :

x
cT -l 2
F EI w̃ dx + cT [w̃ (lD )]2 + c[w̃(lF )]2
00
11
110101l 11
00 00
D
l
1
0
1
0
c
00
11010100
11
00 11
110101
00 F̃crit =
0
-l
.
F
l
w̃2 dx
0
 184 Buckling in

P6.1 Problem 6.1 Both depicted 1)

systems consist of rigid bars F
supported by elastic springs. c c
Determine the critical loads a a
Fcrit . 2)
F
c

a a a
F
Solution to 1) We consider 2aδϕ
the system in the deflected δϕ
state. From equilibrium c2aδϕ A
caδϕ

A: a(caδϕ) + 2a(2caδϕ) − 2aδϕF = 0

we obtain

δϕ(5ca − 2F ) = 0 .

Thus a neighboring equilibrium state (δϕ = 0) related to the equilibri-

um state ϕ = 0 is only possible for
5
Fcrit = ca .
2
G 
aδϕ δϕ 
2 1
to 2) Equilibrium conditions aδϕ δϕ
for the defelcted state A
F B caδϕ

A: a(caδϕ) − 2aB + aδϕF = 0 ,

y G: 2aδϕ F − aB = 0
provide after elimination of B

δϕ(ca − 3F ) = 0 .

This results in the critical force

ca
Fcrit = .
3
 systems of rigid bars 185

Problem 6.2 The depicted frame P6.2

q
consists of four rigid bars connec-
ted by hinges and torsional cT cT
springs with stiffness cT . b

11
001
00
01
0
Determine the critical load qcrit . cT cT

01
1 10
1 2a 10010101
Solution From the sketched deflected state follows the geometric rela-
tion:

f = b (1 − cos ϕ) .
2aq
Hence the potential energy is given by
f
Π = Π(i) + Π(a)
1
= 4 cT ϕ2 − 2qaf
11
00101010
00
11
2 ϕ

= 2cT ϕ − 2qab(1 − cos ϕ) .

2
00
11
10
01
11
00 00
11
00
11 10
The system is in equilibrium, if
dΠ
δΠ = δϕ = (4cT ϕ − 2qab sin ϕ)δϕ = 0 .
dϕ
Thus for equilibrium in the deflected state with δϕ = 0, we must have

4cT ϕ − 2qab sin ϕ = 0/; .

The trivial equilibrium state is related to ϕ = 0.

Using the second variation of the potential energy we can determine

the type of the equilibrium
⎧
⎪
⎪
2 ⎨>0 stable,
d Π
2
δ Π= (δϕ) = (4cT − 2qab cos ϕ)(δϕ)
2 2
=0 indifferent,
dϕ2 ⎪
⎪
⎩
<0 unstable

At the trivial equilibrium state (ϕ = 0) the system is indifferent for the

critical load
2cT
qcrit = .
ab
 186 Buckling of rigid

P6.3 Problem 6.3 The depicted

00
11
1010
F
system consists of rigid
10
010
00
111 11 00
11
1 11
001010
00
11
c
00
0
11
00
11
00
c
bars that are elastically
supported. a a a

Determine the critical loads and sketch the associated buckling figures.
δϕ2 
2 F
G
Solution The system has 
1 aδϕ1 + 2aδϕ2
two degrees of freedom. aδϕ1
The equilibrium conditions δϕ1 c(aδϕ1 + aδϕ2 )
in the deflected state are A caδϕ1

A : ca2 δϕ1 + 2ca2 (δϕ1 + δϕ2 ) − a(δϕ1 + 2δϕ2 )F = 0 ,

y G: ca2 (δϕ1 + δϕ2 ) − 2aδϕ2 F = 0 .

Using λ = F/ca we obtain the homogeneous system of equations

(3 − λ)δϕ1 + 2(1 − λ)δϕ2 = 0 ,

1 · δϕ1 + (1 − 2λ)δϕ2 = 0 .
For a non-trivial solution the determinant of the coefficient matrix has
to vanish:
√
  5 + 17
  λ = ,
(3 − λ) 2(1 − λ) 1
  = 0 ; λ2 − 5 λ+ 1 = 0 ; 4
√
  2 2
 1 (1 − 2λ) 5 − 17
λ2 = .
4
Re-substituting provides
√ √
5 + 17 3 + 17 F1

01
0
10
11
00
01010
F1 = ca, δϕ1 = δϕ2 δϕ2
4 2

and
11
00
11 δϕ1

√ √
5− 17 17 − 3
F2 = ca, δϕ1 = − δϕ2 . δϕ1
110
00
00
11101
4 2 F2
δϕ2

The originally straight equilibrium state can buckle into two neigh-
boring states, because the system has two degrees of freedom. Since
F2 < F1 , force F2 is the critical load: Fcrit = F2 .
 and elastic bars 187

Problem 6.4 Determine for the F EI P6.4

depicted elastic bar the con-
ditions for buckling and the x
l
critical load.

Solution From the general solution of the Euler coloumn

w = A cos λx + B sin λx + Cλx + D , F ,

λ2 = EI
w = −Aλ sin λx + Bλ cos λx + Cλ ,
w = −M/EI = −Aλ2 cos λx − Bλ2 sin λx ,
w = −Q/EI = Aλ3 sin λx − Bλ3 cos λx
in conjunction with the boundary conditions, we derive

w(0) = 0 : ; A+D =0 ; D = −A ,

w (0) = 0 : ; B+C =0 ; C = −B ,
w (l) = 0 : ; −A sin λl + B cos λl + C = 0 ,
Q(l) = 0 : ; A sin λl − B cos λl = 0 .
Inserting C = −B yields for the last two equations

A sin λl − B(cos λl − 1) = 0 ,

A sin λl − B cos λl = 0 .
This leads to B = 0, and the condition of buckling is given by

sin λl = 0 ; λn l = nπ (n = 1, 2, 3, . . .) .

The smallest eigenvalue λ1 l = π provides the critical load

EI
Fcrit = π 2 .
l2

Inserting the constants and the eigenvalue yields the buckling shape
πx
w = A(cos − 1) ,
l
where A remains undetermined.
Note: The critical load for the considered case is identical to the
2nd Euler buckling case.
 188 Buckling load of

P6.5 Problem 6.5 The depicted

beam is subjected to axial cT EI cT
compression and is supported F
at both ends elastically by
torsional springs.
10
010
00
111 x
l 010101
00
11
Given.: EI = l cT .
a) Determine the critical load.
b) Compute with the ansatz w̃1 (x) = a(l −x)x and w̃2 (x) = a sin(πx/l)
the approximate solution for the critical load via the Rayleigh-quotient.

Solution to a) The general solution of the buckling problem

F
w = A cos λx + B sin λx + Cλx + D , λ2 =
EI
yields with the boundary conditions
w(0) = 0 , M (0) = −EIw (0) = −cT w (0) ,
w(l) = 0 , M (l) = −EIw (l) = +cT w (l)
and with the abbreviation κ = EI/lcT the following system of equations
A+D = 0 ,
κAλ2 l = −Bλ − Cλ ,
A cos λl + B sin λl + Cλl + D = 0 ,
κAλ2 l cos λl + κBλ2 l sin λl = −Aλ sin λl + Bλ cos λl + Cλ .
Elimination of the constants yields an equation for the eigenvalues

2 − 2(1 + κλ2 l2 ) cos λl − λl[1 − (κλl)2 − 2κ] sin λl = 0 .

For κ = 1 we deduce from this equation (e. g. by a graphical solution

method) the first eigenvalue and the associated critical load
EI
λ1 l = 3.67 ; Fcrit = λ21 EI= 13.49 .
l2

Note: We obtain the eigenvalue equations for a beam clamped on both

ends as special case (κ = 0 or cT → ∞)
2 − 2 cos λl − λl sin λl = 0 ; λl = 2π ,

and similarily for the simply supported beam κ → ∞ or cT → 0)

sin λl = 0 ; λl = π .
 elastic systems 189

to b) To determine the critical load with the first ansatz we need to

compute the derivatives:

w̃1 (x) = a (lx − x2 ), w̃1 (x) = a (l − 2x), w̃1 (x) = −2a .

Substituting this into the formula of the Rayleigh-quotient yields:

-l
EI · (−2a)2 dx + cT [a (l − 0)]2 + cT [a (l − 2l)]2
0
F̃crit 1 = .
-l
[a (l − 2x)] dx
2

Integration and rearrangement yields

!l
4a2 EIx + cT a2 l2 + cT a2l2 4a2 lEI + cT a2 l2 + cT a2 l2
0
F̃crit 1 =
4 !l
=
4
.
a2 l2 x − 2a2 lx2 + a2 x3 a2 l3 − 2a2 l3 + a2 l3
3 0 3
Inserting lcT = EI leads to the final result
EI
F̃crit 1 = 18 .
l2

Analogously the second ansatz renders step by step

π π   π 2 π 
w̃2 (x) = a cos x , w̃2 (x) = − a sin x ,
l l l l
-l .  π 2  π /2  π 2 . 2  π   π /
EI − a sin x dx + cT a cos 0 + cos2 l
0 l l l l l
F̃crit 2 =
- .π
l  π /2
a cos x dx
0 l l
 π 2 -l π  . /
EI sin2 x dx + cT cos2 (0) + cos2 (π)
l 0 l
= π 
-l
cos2 x dx
0 l
   
 π 2 1 1 l  π  l  π 2 1
EI − sin 2 x + 2cT EI l − 0 + 2cT
l 2 4π l 0 l 2
=  l = ,
1 1 l  π  1
l−0
+ sin 2 x 2
2 4π l 0
 2 
EI π + 4 EI
; F̃crit 2 = = 13.87 2 .
l2 l
 190 Buckling load of

P6.6 Problem 6.6 A beam is clamped EI B F

at the left end and is elastically c
supported at B by a spring (spring x
constant c). l
Deduce the condition for buckling.

Solution The general solution for the buckling problem is given by

w = A cos λx + B sin λx + Cλx + D , F ,
λ2 = EI
w = −Aλ sin λx + Bλ cos λx + Cλ ,
w = −M/EI = −Aλ2 cos λx − Bλ2 sin λx ,
w = −Q/EI = Aλ3 sin λx − Bλ3 cos λx .
The boundary conditions
Q(l)
w(0) = 0 ,
B w(l)
w (l) F
w (0) = 0 , N
B 
w (l)
M (l) = 0 , Q(l) cw(l)
Q(l) = −c w(l) + F w (l)
lead to the homogeneous system of equations

A+D = 0 ,
B+C =0,
−A cos λl − B sin λl = 0 ,
A cos λl + B sin λl + C(λl − EIλ3 /c) + D = 0 .

Eliminating the constants yields the equation for the eigenvalues (buck-
ling condition)
EI
tan λl = λl − (λl)3 . 2
tan λl
cl3 tan λl
π/2 λ1 l π
0
The solution of this transcendental λl
equation can by obtained graphi-
−2
cally. The special case EI/cl3 = 1 λl − (λl) 3

yields the first eigenvalue

−4
EI
λ1 l ∼
= 1.81 ; Fcrit ∼
= 3.27 2 .
l
 elastic systems 191

00
11
11
00
0010101010
rigid EI F
Problem 6.7 The depicted beam P6.7
consists of a rigid and an elastic
part (flexural rigidity EI).
10
010
00
111 a
x
a 11
00
11
Determine the bucking condition and the critical load.

Solution The general solution of the buckling problem is given by

F
w = A cos λx + B sin λx + Cλx + D , λ2 = .
EI
From the boundary and transmission conditions

w(a) = 0 , ϕ = w  (0)
A x B F

M (a) = −EIw (a) = 0 , ϕa
 N(0)
w(0) = ϕ a = w (0) a , w  (0)
M(0) Q(0)
Q(0) = F w (0)
we derive
A cos λa + B sin λa + Cλa + D = 0 ,

A cos λa + B sin λa = 0 ,

A + D = Bλa + Cλa ,

EI Bλ3 = F (Bλ + Cλ) .

This yields the constants C = 0, D = 0, A = Bλa and the buckling
condition

tan λa = −λa .

tan λa
tan λa
The graphical (or numerical) solution 1
renders the first eigenvalue π/2 λ1 a π
0
λa
λ1 a ∼
= 2.03 −1
−λa
and hence the critical load −2
EI
Fcrit ∼
= 4.12 2 . −3
a
 192 Buckling due to temperature loadings
l2

11
00
P6.8 Problem 6.8 Consider the sketched half-

00
11
frame with different cross section data in the

regions x and y .
2

Given: l1 = 5.0 m ,
l2 = 1.0 m , l1 
1

E = 2.1 · 104 kN/cm2 ,

αT = 1.2 · 10−5 1/K ,
A1 = 50.0 cm2 ,
I1 = 500 cm4 ,
I2 = 10000 cm4 .

How much can region x be heated until

buckling occurs?

Solution We choose a substitute system for

region x according to Euler case 2 with Fk
length l1 . For this case the buckling load is:

2 2.1 · 10 · 500
4
EI1
Fb = π 2 = π
l12 5002 l1

= 414.52 kN .

The displacements of shaft x and beam y

are given by:

Fb l23 414.52 · 1003 011010

10
f
f= =
3EI2 3 · 2.1 · 104 · 104
Fk
= 0.658 cm ,

Fk l1 −Δl1
Δl1 = ε1 l1 = − + αT ΔT l1
EA1
= −0.1974 + 6 · 10−3 ΔT . ΔT

Using compatibility

f = Δl1 ; 0.658 = −0.1974 + 6 · 10−3 ΔT

we can determine the required temperature difference

ΔT = 142.5 K .
 Buckling load 193

Problem 6.9 The depicted sys- F P6.9

tem consists of bars with diffe-
rent flexural rigidity.
EI2 EI1
Assign the individual bars to a
EI1 
2 
1
the corresponding Euler cases
and determine which bar is 3
first buckling for the ratio 2a a a
EI2 = 2EI1 .

Solution The Euler cases are determined from the table on page 183:
Bar x and bar y correspond to the second Euler buckling mode, be-
cause these bars are hinged at both ends.
Bar z is clamped at the right side and simply supported at the left
side. This corresponds to the third Euler case.

The forces due to the load F are given by

F
S1 = − √ ,
2
F
F
S2 = − √ ,
2
S2 S1
F S3
S3 = − .
2

Thus we obtain the following critical forces

F1 crit π 2 EI1 1 π 2 EI1
√ = ; F1 crit = √ ,
2 2a2 2 a2
F2 crit π 2 EI2 √ π 2 EI1
√ = ; F2 crit = 2 ,
2 2a2 a2

F3 crit π 2 EI1 π 2 EI1

= 2.04 ; F3 crit = 1.02 .
2 4a2 a2

Because F1 crit < F3 crit < F2 crit , bar x buckles first. Hence force F1 crit
is crucial for the failure of the entire system.
 194 Spatial buckling A z
z A A
P6.10 Problem 6.10 The depicted
a A F y
construction is assmbled from 
1 y A
two bars with double symmetric A

2
profile (Iy = 2Iz for both bars). x
Determine the maximal height a
a
a, such that no buckling occurs.

Solution Due to the symmetry of the structure and the load the follo-
wing compressive normal forces appear in the two bars
F
S1 = S2 = √ .
2
To investigate the stability behaviour we consider the different support
conditions and the different flexural rigidities. Bar x is simply sup-
ported at the lower end. The upper end is fixed by a rigid slider and
connected to bar y by a hinge. This corresponds to the Euler case no.
3. For buckling along the local y-axis of the profile we compute

F π 2 EIy EIy
S1 = √ = 2.04 ; a1 y = 1.20 π
2 2a2 F
and for buckling along the local z-axis with EIz = 0.5 EIy

F π 2 EIz EIy
S1 = √ = 2.04 ; a 1 z = 0.85 π .
2 2a2 F
follows. Bar y is hinged with one rotation axis in y-direction at the
lower end. With regard to rotation along the x-axis the support is ri-
gid. The support at the upper end is analogous to bar x . Buckling
along the local y-axis corresponds to the third Euler buckling mode.
With S2 = S1 we obtain

a2 y = a1 y .

Buckling along the local z-axis is equivalent to the Euler case no. 4 and
yields with EIz = 0.5 EIy

F π 2 EIy EIy
S2 = √ = 2.04 2
; a2 z = 1.19 π .
2 2a F
Since a1 z is the smallest value, the critical length is given by

EIy
acrit = 0.85 π .
F
Chapter 7
Hydrostatics
7
196 Fluid pressure

Prerequisite: The density ρ (unit: kg/m3 ) of the fluid is constant.

Pressure: The pressure p (unit: Pa ≡ N/m2 ) is a force per area, that

is identical for all cross sections and always acts normal to the cross
section (hydrostatic stress state).
Pressure in a fluid under the acti- p0
on of gravity and a surface pressure
p0 is given by:
z

p (z) = p0 +  g z . p(z)

The buoyancy force acting on a body (volume V ) immersed in a fluid

is equal to the weight of the displaced fluid volume.
Buoyancy force:

FA = ρ g V .
FA
The line of action related to the CF
buoyancy force passes through the V
center of gravity CF of the displa-
ced fluid volume.

Fluid pressure on plane surfaces

Resulting force

F = p (yC ) A = ρ g hC A .

hC Center of pressure D
F yC
x Ix
yD yD = ,
A xC Sx
xD Ixy
xD =− .
C Sx
D
y
 Buoyancy 197

Fluid pressure on curved surfaces

dĀ dFV = p dA cos α = ρ g dV

z dV
dF
dFH = p dA sin α = p dA∗
∗
dA
α C∗ Integration yields
ρ
dA A∗
FV = ρ g V ,
A
FH = pC ∗ A∗ .

The resulting horizontal component of the fluid pressure FH is equal

to the product of the vertically projected area A∗ and the pressure pC ∗
in the centroid of the projected area.

Stability of a floating body: The equilibrium state is stable if the

meta center M is above the centroid CB of the body:

x
1
> 0 : stable
hM =
y
< 0 : unstable

hM M y
with the position of the meta
CB
center
e
CF
Ix
hM = −e.
V
z
water line

Here the following data are used

Ix : second moment of area defined by the water line,
V : volume of the displaced fluid,
e: distance of the centroid of the body centroid CB
from the centroid of the displaced fluid CF .
 198 Buoyancy

P7.1 Problem 7.1 A container is closed r1

during filling by a ball valve.
Determine the density ρB of the
ball, such that no air remains in
the container when the ball closes r2
the valve. ρB
Given.: ρF , r1 , r2 .
ρF

Solution The ball has to submerge

to a depth that just closes the
opening when the container is full.
The buoyancy force is than ρF g V1 ,
where V1 is the volume of the
displaced fluid (spherical segment). r1
The buoyancy force has to be equal
to the weight of the ball
r2
ρB h
ρF g V1 = ρB g V.

With the volume of a sphere

4
V = π r23
3

and the spherical section


h
V1 = π h2 (r2 − ), h = r2 + r22 − r12
3

we compute for the density of the ball

h  2  
V1 π h2 (r2 − )
ρB = ρF = ρF 3 =ρ 3 h
1−
h
.
F
V 4 4 r2 3 r2
π r23
3
 Buoyancy 199
B
Problem 7.2 The design of the de- P7.2
picted valve of a water basin
ensures that the valve opens if the z̄
water level reaches the hinge at z
point B. The flap valve is assumed 111
000
000
111
to be massless. 000
111
r a
ρ 000
111 C
Determine z̄ for the valve to func-
tion in the described way.
Given: ρ, a, r.

Solution The thickness of the flap valve is irrelevant for the following
considerations, as all forces are assumed per unit length.
BV
We compute the resulting horizontal for-
ce from the linear pressure distribution:
BH
1 2 z
FH = ρ g (z̄ + a)
2
with C=0
2 FH
z = (z̄ + a) . FV
3
r
The vertical buoyancy force can be computed from the weight of the
displaced water by using the area of the dashed region:
 π 
FV = ρ g 2 a r − r 2 .
2
The flap valve just opens if the reaction force in C vanishes. Equilibrium
of moments with regard to B provides:

B : −rFV + zFH = 0

 π 2 1 2
; −ρ g 2ar − r r + ρ g (z̄ + a)2 (a + z̄) = 0 .
2 2 3
The solution of this equation with respect to z̄ yields the water level

π
z̄ = 3 3 (2 a r − r 2 ) r − a .
2
 200 Buoyancy

P7.3 Problem 7.3 The depicted cross

section of a tunnel is immersed
in water saturated “liquid” sand ρS Sand h
(density ρSA ). Above resides a layer
of dry sand (density ρS ).
ri
Determine the thickness x of the
concrete base (density ρC ), such h
that a safety factor η = 2 against ρSA
lifting is reached. It is assumed that ρC x
the weight of the dry sand is acting
on the cross section of the tunnel. l

Given: ρB = 2.5 · 103 kg/m3 , ρS = 2.0 · 103 kg/m3 ,

ρSA = 1.0 · 103 kg/m3 , l = 10 m, ri = 4 m, h = 7 m.

Solution The weight (per unit length) of the tunnel cross section and
sand load is given by
l  π  l2 !
G = ρC g x l + − ri 2 h + − ri2 + ρS g l h .
2 2 4
With the buoyancy force (per unit length)

π l2 !
B = ρSA g (h + x) l +
2 4
we can determine the height of the concrete base, such that a safety
factor against lifting
G
η=2=
B
is achieved. Solving for x yields:
l  π  l2 !
(2ρSA l − ρB l)x = ρS lh + ρB − ri 2h + − ri2
2 2 4
 π l2 
−2 ρSA hl + .
2 4
With the given data we get
π !  π 
(20 − 25) x = 2 · 70 + 2.5 14 + (25 − 16) − 2 70 + 25
2 2
; −5 x = 210.34 − 218.54

; x = 1.64 m .
 Buoyancy 201
AS
Problem 7.4 A cylindrical plug P (cross sect- P7.4
ion AP , length a) is elastically supported and S
closes straight with the bottom of a basin for
the water line h0 . In this situation the force ρ
vanishes in the rope (length l) to which a h0
l
floater S is attached (cross section AS > AP ). AP
a) Determine the weight GS of the floater.
P a
b) Which maximal water height h1 can be c
reached before leaking occurs?

Solution to a) The weight GS of the floater GS

is computed from equilibrium and geometry in t0
the reference situation:
2
ρgAS t0 = GS ρgAS t0
; GS = (h0 − l)ρgAS . h0
h0 = l + t0 l

to b) For a water line h the plug is elevated

by a distance y due to the force in the rope S.
The equilibrium conditions for the floater, for
the plug, and the geometric conditions are

ρgAS t = GS + S , S − Fp = cy ,
h = l+t+y. GS
t
In the equilibrium expression, Fp is the diffe-
rence in the pressure force in the displaced and
the reference situation (the forces due to lateral ρgAS t
pressure are in equilibrium): S l h
Fp = ρg(h−y)Ap−ρgh0 Ap = ρg(h−y−h0 )Ap . Fp
y
Eliminating GS , S, Fp , and t yields
a
c !
h − h0 = y 1 + . cy
ρg(AS − AP )

The maximal heighth = h1 is reached, if y = a is attained:

c !
h1 = h0 + a 1 + .
ρg(AS − AP )
 202 Fluid pressure
y
P7.5 Problem 7.5 A dam of length l has a
surface of parabolic shape with a a
horizontal tangent at the bottom of
the water basin.
Determine the force resulting from ρ
the pressure, the position of the point h
of action, and the line of action for a
water height h. x
Given: h, l, a = h/4, ρ.

y y
dA
Solution The vertical
component of the force
is FV = ρ g V with the
volume V = l A. The area is h-y F
α FV
determined by the function yF
y(x) = 16 x2 /h of the para- FH
bola A x x
 a
xF
A= (h − y) dx dx
0 a
16 2 16 3 !a h2
= (h − x ) dx = h x − x = .
0 h 3h 0 6
Thus the vertical component of the pressure force becomes:
1
FV = ρ g h2 l.
6
The vertical force acts at the centroid C of the area
 a  a
1 16 2 x2 16 x4 3
xF = x (h − x ) dx = h − = h.
A 0 h 2 h 4 0 32

The horizontal component of the fluid pressure is computed by the

projected area A∗ = h l and the pressure pS ∗ = 12 ρ g h in the centroid
of the projected area:
1 1
FH = ρ g h2 l with yF = h.
2 3
By the theorem of Pythagoras, we obtain the resulting force, its line of
action passes through the point (xF , yF ) and forms an angle α to the
y-axis:

1√ FH
F = FH 2
+ FV2 = 10 ρ g h2 l, α = arctan = arctan 3 = 71.5o .
6 FV
 Floating stability 203

Problem 7.6 A prismatic body with P7.6

the mass mB , width a, and length l
mB CB ρW
is floating in the water. Its centroid
CB is in the height hB . mA hB
Determine the additional point mass
mA , such that the body floats in a
stable manner. a

Given: ρW , mB , hSB , l, a.

Solution Stable floating of the body is defined by the position of the

meta center hM = Ix / V − e > 0 . For hM = 0 the limit of the stable
state is reached.
The volume V of the displaced fluid is obtained by equilibrium (buoyan-
cy = weight of the body and added mass):
1
ρW g V = (mB + mA ) g ; V = (mB + mA )
ρW
The second mom-
ent of area is CB x
CC y
l a3 e l
Ix = . CF hB
12 hC
hF
For e = hC − hF
we need the cen- a a
ter of gravity hC
of the floating construction and hF of the displaced fluid. They are
determined by
mB
hC (mB + mA ) = hB mB ; hC = hB ,
mB + mA
mB + mA
V = a l (2 hF ) ; hF = .
2 a l ρW
The limit for stable floating is reached if hM = 0:
mB 12 (mB + mA )2
1 − 12 hC + =0.
l a3 ρ W 2 l 2 a4 ρ W 2
Solving for the required additional mass mA yields

l a2 ρ W mB
mA = √ 12 hCB 3 − 1 − mB .
6 l a ρW
 204 Floating stability

P7.7 Problem 7.7 A cone-shaped floa-

ting device is made of two materials d
with densities ρ1 and ρ2 .
Determine the diameter d of the h1 ρ1
cone, such that it floats stable in a
fluid of density ρF . ρF
ρ2
Given: h2 g
2 1
ρ1 = ρF , ρ2 = ρF ,
3 3
h1 = 2 h , h2 = 4 h .

Solution The cone has a stable floating position, if the following con-
ditions are met:

(1) : G = A,

Ix
(2) : hM = − e > 0.
V
d
(1) Floating condition:
h1
d d1 h2 2
= ; d1 = d = d. h2
h1 + h2 h2 h1 + h2 3
d1
The force due to weight is

G = V1 ρ 1 g + V2 ρ 2 g

1 1
= π h1 (d2 + dd1 + d21 ) ρ1 g + π h2 d21 ρ2 g
12 12
23
= π h d2 ρF g = 0.892 h d2 ρF g .
81
The immersion depth t and the diameter dT = d t/(h1 + h2 ) of the cone
at the water line of the fluid follows the buoyancy force
1
A= π t d2T ρF g
12

1 d2
= π 2 ρ F g t3 .
432 h
 Floating stability 205

For G = A we obtain

368 3
t3 = h ; t = 4.969 h .
3

t
(2) Stability condition:
The volume of the displaced fluid is given by dT

1 d2 3 23
V = π t = π h d2 = 0.892 h d2 ,
432 h2 81

and the second moment of area Ix is

d4T π (0.828 d)4 π

Ix = = = 0.023 d4 .
64 64

The distance of the centroid of the body from the centroid of the dis-
placed fluid is provided by
3
e = xS − t
4
with
3 1 3 1
(h1 + h2 ) ρ1 π d2 (h1 + h2 ) + h2 (ρ2 − ρ1 ) π d21 h2
xS = 4 16 4 16
1 1
ρ1 π d2 (h1 + h2 ) + (ρ2 − ρ1 ) π d21 h2
16 16

16
18 h − h
= 9 = 4.761 h
16
4−
27

3
; e = 4.761 h − · 4.969 h = 1.034 h .
4

For the diameter of the cone we finally obtain

0.023 d4
hM = − 1.034 > 0 ; d > 6.333 h .
0.892 h d2
 206 Floating stability

P7.8 Problem 7.8 A block-shaped iceberg of dimensions a × h × l calves of

a floating ice shelf. It is assumed that a  h. The density of the water
9
is ρW , the density of the ice ρI = 10 ρW .
Eisberg Schelfeis
a

h ρE
t

ρW l

For which length l does the iceberg float in a stable way?

Solution We start by determining the immersion depth t of the iceberg.
Equilibrium between iceberg and buoyancy force renders for the given
density ratio the immersion depth
9
ρI ghla = ρW gtla ; t= h.
10
To analyze the floating stability we consider the position hM of the meta
center:
Ix a
hM = −e, x
V
3
al
Ix = ,
12
9 y
V = alt = alh , h t ρE
10
h t h
e= − = . ρW z
2 2 20 l
By combining all relations we derive

5 l2 h
hM = − .
54 h 20
We consider the limit of floating stability (hM = 0). This determines
the length l0 :

27 2 27
l02 = h ; l0 = h ≈ 0.735h .
50 50
In a stable floating state, we must have hM > 0. Thus, the iceberg
floats stable for l > l0 . For l < l0 the iceberg tips over.
 Fluid pressure 207

Problem 7.9 A circular shaped hatch P7.9

closes the outflow of a tank.
a) Determine the mass m, such b
that the hatch opens if m is m a
attached in the distance c from d
the hinge point. ρ e
b) Determine the distance by S
which the mass m has to be B c
shifted, for the hatch to open 45◦
when the water level reaches
the height b.
Given: a, b, c, d, e, m, ρ.

Solution zu a) The force acting on the hatch is √

2
yD − 2a
πd
F = ρ g A hS = ρ g (a + e) . F
4
The point of action of F is determined by B S mg

Iξ √ d2
yD = yS + = 2 (a + e) + √ .
yS A 16 2 (a + e)
The hatch opens, if B = 0. Equilibrium of moments provides
√
F ( yD − 2 a ) − m g c = 0 .

From this we compute the required mass

 
π d2 √ d2
m=ρ (a + e) 2e + √ .
4c 16 2 (a + e)

to b) For the water level b the force acting on the hatch is

π d2
F = ρ g A hs = ρ g (b + e) .
4
With the point of action
√ d2
yD = 2 (b + e) + √
16 2 (b + e)
√
of F the equilibrium condition F ( yD − 2 b ) − m g c = 0 yields the
distance c:
 
π d2 √ d2 1
c=ρ (b + e) 2e+ √ .
4 16 2 (b + e) m
 208 Fluid pressure

P7.10 Problem 7.10 A trapezoidal hatch

closes the outflow of the depicted
basin. 7m
ρW
Determine the resulting force on
the hatch together with the support 2m
B
reactions in point B.
kg m C 3m
Given: ρW = 103 , g = 9.81 2 .
m3 s 4m

Solution The area A = 10 m2 , the

centroid of the hatch B

 2  1 35
ȳs = 5·2, 5+5· ·5 = m ȳ ȳs
3 10 12
ξ S
and the pressure 5
 
3 35 43 η
p (ȳs ) = ρ g 9 + · = ρg
5 12 4

are used to compute the resulting C

force [m]
1 1 1
43
F = ρ g A p (ȳs ) = 103 ·9, 81·10· = 1.05 MN .
4

The position of the line of action follows from

 2  2
53 · 1 35 53 · 1 35 10
Iξ = +5·1 − 2, 5 +2 +5·1 − = 19.1 m4 ,
12 12 36 12 3
ys = ȳs + 15 m and yD = ȳD + 15 m to be
Ix y 2 A + Iξ Iξ 35 19, 1
yD = = s ; ȳD = ȳs + = + 35
Sx ys A ys A 12 ( 12 + 15) 10
= 3.02 m .

The support reaction is determined by equilibrium of moments with

regard to the hinge point C of the hatch
 F
C : B · 5 − F ( 5 − 3, 02 ) = 0 B

5 − 3, 02 C
; B = 1, 05 = 0.415 MN . ȳD
5
 Fluid pressure 209

Problem 7.11 A concrete dam P7.11

(density ρC ) closes a basin
that is filled up to the height 3m
h = 15 m.
Determine ρW
12m
a) the safety factor against 2m ρC
sliding at the bottom (adhesion
coefficient μ0 ),
b) the safety against tilting, 3m 3m
c) the stress distribution at the
bottom, if it assumed to be a 6m 8m μ0
linear distribution.

Given: ρC = 2.5 · 103 kg/m3 , ρW = 103 kg/m3 , μ0 = 0.5 , g = 10 m/s2

Solution to a) To determine the safety factor against sliding we com-

pute the horizontal forces due to the water pressure and compare them
to the adhesion forces acting at the bottom. The horizontal force due
to water pressure is computed from
1 1
FH = ρW g h A = 103 · 10 · 15 · 15 · 1 = 1125 kN/m .
2 2
The resulting force due to the weight of the concrete and the water
pressure is
1
FV = 2.5 · 103 (3 · 2 + 4 · 18 + 3 · 8 + · 12 · 8) + 103 (2 · 12) = 3990 kN/m .
2
Using Coulomb’s friction law we determine the safety factor ηS against
the onset of sliding
μ0 FV 0.5 · 3990
ηS = = = 1.77 .
FH 1125

xBi
to b) The dam can tilt around
point B. The safety against
tilting is determined by compa-
ring the moment of forces. The FH
moment of the water pressure h
is given by 3 FV i

h 15 B
MBW = FH = 1125 · = 5625 kNm .
3 3
210 Fluid pressure

The moment related to the weights is


MBG = FV i xBi
i

= 2.5 · 103 (3 · 2 · 13 + 4 · 18 · 10 + 3 · 8 · 4
1 2
+ · 12 · 8 · · 8) + 103 (2 · 12 · 13) = 31870 kNm .
2 3
This results in a safety factor against tilting
MBG 31780
ηT = = = 5.67 .
MBW 5625

to c) To compute the stress distribution in the bottom gap3of the dam

we determine the excentricity of the resulting force RV = i FV i . The
vertical component of the force acting in the gap yields, according to
the sketch below,

RV (a − e) = MBG − MBW
MBG − MBW 31870 − 5625
; e=a− =7− = 0.422 m .
RV 3990

With the introduced coordinate-

system we compute the normal
stresses in the bottom gap (like in e
a beam cross section)
N My z MBW
σ= + x. MBG
A Iy x
C B
Here we have to insert the fol- σC
RV σB
lowing data: A = 14 m2 , Iy =
1 · 143 / 12 = 288.67 m4 , N =
−RV = −3990 kN, My = N · e = a=7m a=7m
−1685 kNm. As a result we obtain
for the stress distribution
−3990 −1685
σ= + x = −285 − 7.37 x kN/m2 .
14 228.67

For the selected points C and B evaluation yields

σC = −0.23 MPa and σB = −0.34 MPa .

 Fluid pressure 211
Problem 7.12 A rectangular plate P7.12
of width b closes the outlet of a
basin. It is hinged at point D.
a) Determine the water height t, ρ D t
for which the plate starts to rotate h
around point D.
b) Compute the bending moment
at point D for this situation.
l
Given: b, l, h, ρ.

to a) The plate starts to rotate, if the resulting force R of the water

pressure is above point D. In the limit case the resulting force of the
water pressure passes through point D. From this we can determine the
water height
R R̄
t = 3h .
pD

2a
h MD 2a
D
3
a

to b) To compute the bending moment in the plate we start with the

moment at point D. With the resultant R̄ of the upper plate and the
pressure at point D,
1
R̄ = pD 2 a b , pD = ρ g 2 h ,
2
we obtain
2 2 4
MD = −R̄ a = − pD b a2 = − ρ g (l2 + h2 )h b .
3 3 3

The distribution of the bending moment is cubic for a linearly varying

load. The maximum occurs at the hinge point D.

MD

D
 212 Fluid pressure

P7.13 Problem 7.13 The pressure p in gases depends on the density ρ. The
relation between the two state variables is provided by the universal
gas equation p = ρRT (universal gas constant R, temperature T ). E.
g. for air at sea level and at T = 0◦ it holds: p0 = 101325 Pa and
ρ0 = 1.293 kg/m3 .

Determine the dependency of air pressure on height for the case of

a constant temperature (barometric height relation).
Solution First, we apply the universal gas law at sea level. This yields
p0
p0 = ρ0 RT or RT = .
ρ0 p + dp
Equilibrium of an infinitesimal air
column with cross section A and ρgA dz
height dz dz

↑: pA − ρgA dz − (p + dp) A = 0

leads to z p

dp
= −ρg .
dz
Using the universal gas equation yields
dp pg
=− .
dz RT
By separation of variables and integration we obtain:
 p  z
dp g dp̄ g p g
=− dz ; =− dz̄ ; ln =− z.
p RT p0 p̄ 0 RT p0 RT

This renders the air pressure as a function of the height

gz
−
p = p0 e RT .

The air pressure decreases exponentially with the height. From the
relation RT = p0 /ρ0 and the gravity constant g = 9.80665 m/s2 we
deduce
z
−
p = 101325 Pa e 7991 m .

Note: In a height of 5, 5 km the pressure has dropped to one half of

its original value.