Академический Документы
Профессиональный Документы
Культура Документы
Mechanics of
Materials –
Formulas and
Problems
Engineering Mechanics 2
123
Mechanics of Materials – Formulas
and Problems
Dietmar Gross Wolfgang Ehlers
•
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
Peter Wriggers
Institute of Continuum Mechanics
Leibniz Universität Hannover
Hannover
Germany
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).
Hence the stress tensor is a symmetric tensor of second order: τij = τji .
Plane stress state 3
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
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 ).
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
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 .
τ
direction of scale: 10 MPa
σ1
σ2
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
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.
σξ = σ0 , ση = σ0 , τξη = 0 .
stress state 9
τmax
σ1 = σ0 σ1 = τ0
σ2 = 0 σ σ2 = −τ0 σ
τ
to b)
σ1 = σ2 = σ0 σ
22
; cos 2ϕb = = 0.297 ; 2ϕb = 72.7◦ ; ϕb = 36.35◦ .
74
stress state 11
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
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
|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
σx = σ1 σy = σ2 .
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 ,
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
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
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
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 .
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.
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 ?
εx = εy = αT ΔT .
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
Δr2 = Δr1 + t .
Eεx = σx −νσy +νp0 , Eεy = σy +νp0 −νσx , Eεz = −p0 −νσx −νσy .
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
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
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
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.
Eεz = 0 = σz − ν(σr + σϕ ) .
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
σy = −ρg(h − y)
to be
1 1
ε1 = 2 1 + √ ε̄ , ε2 = 2 1 − √ ε̄ .
3 3
28 Stresses and strains
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
Assumptions:
N (x)
σ(x) = .
A(x)
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,
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).
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.
A(x) = A0 x/l .
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 .
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
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
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
b
Solution The normal force N = −σ0 a is constant. From the kinema-
2
n = ρω 2 xA(x) = ρω 2 A0 xe−αx/l .
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
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
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
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
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
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 .
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
equilibrium: N1 + N2 = F , EA1 , αT 1
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
to b) After removal of the force Fa new forces SAl and SSt are present.
This leads to the equilibrium condition
Hooke’s law
SAl lAl SSt lSt SSt
ΔlAl = , ΔlSt =
EAAl EASt
and the compatibility condition
ΔlAl + ΔlSt = δ .
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
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
a a
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.
→: S2 cos 45◦ = D , C D
◦
↑: S1 = S2 cos 45 , S1 S2
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
(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 .
↓: 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
Kinematics is expressed by
S1 S2 S3 = S1
Δl1 = Δl2 .
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
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
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
3Q Q
Note: τmax = τ (z = 0) = is 50% larger than τmean = .
2 bh bh
ds
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 .
EIw = −M .
(EIw ) = q ,
or for EI = const
EIwIV = q .
EIw = −(M + MT ) .
Boundary conditions and solution methods 61
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
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).
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
My Mz
EIy w = −My , EIz v = Mz , σ= z− y.
Iy Iz
Normal stress 65
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
2a
Determine the normal a a
stresses at point y at the
2
support. 2a
2a
1
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
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.
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
t2 h
y
t1 z
V S(s)
τ =
I t(s)
h t2
= A1 + (h2 − 4z 2 ) .
2 8
in beams 71
h A2
V t1 s V s
2 A1
τ1 (s) = =
h2 A2
1+
A2 h
(A2 + 6A1 )t1
12 6A1
+ 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
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
σS = E S ε = a E S z , σC = E C ε = a E C z .
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
h3 h2 b
6 − =0 ; b = 2h = 80 cm .
6 2
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
a t
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
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
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
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
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
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
∗∗ 8
; x = 1( + −
) l = 0, 519 l .
15
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
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
EI
l
EI wIV = q = q0 ,
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
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
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
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
q0 3/2
h(x) = x .
σ0 bl
This leads to
q0 q0 9/2
I(x) = x .
12σ0 l bσ0 l
"
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
F F F
I B II I B Fl II
wI B
f wI wII
F l3 F l3 F l2
wII = , wIF = , wI F =
3EI 3(2EI) 2(2EI)
(F l)l2 (F l)l
wIM = , wI M = .
2(2EI) (2EI)
' (
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
q0 q0
X = MA
wq
wX
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
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
wq X wX
X
Δa
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
+ 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
A B C wq wq w1 w1 X2
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
x B
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
3 (q1 − q0 )l
w(l) = 0 ; B= q1 l − .
8 10
q1 l4 (q1 − q0 )l4 q0 l4
= +
24 24 24
' ( ' (
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
Given: F = 2q0 l. l l
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
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.
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 .
q0 q0 wX
wq X
f = wq − wX .
Qs
Δs =
EA
to be V S=Q
ϕ
V l3 Q(1 − cos ϕ)l3 H
w= = . wH V
3EI 3EI Q
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
(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
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
Δl 1
M = hS = EAS h .
l 2 h3 EAS
1+
3 lEI
Frame 103
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
M0 M0 /2 M0 /2 M0 /2 M0 /2
M0 l
ΔϕH = 2ψ = .
3EI
Unsymmetrical bending 105
65 · 103 10 · 1203
Iy = + (42.16)2 (65 · 10) + + (22.84)2 (10 · 120)
12 12
= 322.7 cm4 ,
= −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
My Iz My Iyz
Ew = − , Ev = − .
Δ Δ
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
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
and the displacements follow from the table on page 62 (load case no. 1)
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:
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 .
Torsion
MA A
m(x)
B
x
y MB
z l
Basic equations 113
Prerequisites, assumptions:
Equilibrium conditions
dMT
= −m , m(x) = external moment per unit length.
dx
dϑ
GIT = MT ,
dx
x P
ϑ = angle of twist, y
GIT = torsional rigidity, ϑ(x)
G = shear modulus, z
IT = torsional constant. P´
l l
MT
Δϑ = ϑ(l) − ϑ(0) = ϑ (x)dx = dx .
GIT
0 0
MT l
Δϑ = .
GIT
MT
τmax = , WT = sectional moment of torsion.
WT
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
4A2T
τ (s) IT = 0
AT s ds
t(s)
t(s)
t
a a = const
IT = 2πa3 t
t = const
a
τmax
ellipse
b
a 3 b3
a IT = π
a + b2
2
τmax
Sectional moments for torsion 115
WT Remarks
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
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.
MT
τmax = = τadmis .
WT
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
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
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
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
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
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
MT = MT1 + MT2
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
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
yields F
R2 M2 F
M2 = − M1 .
R1
to b) The critical value of the shear stress is reached in each shaft for:
|M2 | R2 16M1 R2
τmax2 = = = τadmis ; d2 = 3 d1 .
W2 R1 πd32 R1
we compute −ϑ2B
1 2 2
32M1 R2 3 R2
ϑC = l1 + l2 .
Gπd41 R1
124 Torsion
MA + MB = M0
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
1 m < x − a >2 +C x + C .
GIT ϑ = − 2 0 1 2
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
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
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
ϑ1 α
α − ϑ2 = ϑ1
ϑ2
1
2
and
MT MT
β = ϑ1
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.
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
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
0
MTi 1 T
ϑi = = ds .
GITi 2GATi t
i
π 2
AT1 = a , AT2 = 4a2
2
' (
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
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
τadmis a2 t
MTadmis = 10 − 75 .
MT
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
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
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
a 26F a3
wF = wC 1 + ϑ C1 = .
2 81EI
4 aF ,
MA = M1 − V1 a = − 9
2 aF .
MB = M1 + V2 2a = − 9
σA σA 2 arF
σ1 = + + τ12 = ,
2 2 2I
σA 2 5arF
τmax = + τ12 = .
2 18 I
136 Verschiebung
F a2 Ma F a3 M a2
wC = − , wC = − .
4EI EI 6EI 2EI
Ma
ϑC = .
GIT
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
10 35 10
[cm]
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
300 300 · 20
Vz = = 150 kN , My = = 1500 kNm ,
2 4
MT = 300 · 1.5 = 450 kNm .
60 −9000
−80 s s −80
zh − line [cm2 ] Sy − line [cm3 ] −10600
60
and principal stresses 141
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
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
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
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
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
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
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 sik 7 11
(3k1 + 5k2 ) (5 − β − β 2 ) sik sik
12 12 15 30
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
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
F F
i k
fik k fik = fki
i fki
Conservation of energy 149
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
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
F M
F F
2a
“1”
1 1
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
M M =0, fV = 0 .
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
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
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
b
a
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
√
S1 = 2F , F
S2 = −F .
MA = aF
aF
AV = F
MV
√
S 1V = 2,
S 2V = −1 .
a
in mixed systems 157
' (
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
MH
1
S 1H = 0 , “1”
2
S 2H = 1 .
AH = 1 a
MA = a
AV = 0
' (
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
A EI C
a a B
2 2
M V = −R cos ϕ
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)
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
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
“1”
M (2)
B
A
Using
' (
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
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
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
“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
“2”-system:
a a
M (2)
(2)
AH “1”
(2) (2)
AV CV
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
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
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
“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
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)
(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
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
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
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
“1”-system:
a
“1” M (1) (x) = (a − x)
a
4
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
“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
“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
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
−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
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
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
1 2
“1 “1
M̄
S̄1 = 1
S̄2 = −1
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
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.
EI F
F
w wIV + λ2 w = 0 , λ2 =
EI
x
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.
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
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 .
δϕ(ca − 3F ) = 0 .
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
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 .
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
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, . . .) .
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
sin λl = 0 ; λl = π .
elastic systems 189
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
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.
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 ,
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
2 2.1 · 10 · 500
4
EI1
Fb = π 2 = π
l12 5002 l1
= 414.52 kN .
Fk l1 −Δl1
Δl1 = ε1 l1 = − + αT ΔT l1
EA1
= −0.1974 + 6 · 10−3 ΔT . ΔT
Using compatibility
ΔT = 142.5 K .
Buckling load 193
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.
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
p (z) = p0 + g z . p(z)
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.
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
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
4
V = π r23
3
h
V1 = π h2 (r2 − ), h = r2 + r22 − r12
3
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
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?
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 )
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
Given: ρW , mB , hSB , l, a.
l a2 ρ W mB
mA = √ 12 hCB 3 − 1 − mB .
6 l a ρW
204 Floating stability
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
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
0.023 d4
hM = − 1.034 > 0 ; d > 6.333 h .
0.892 h d2
206 Floating stability
h ρE
t
ρW l
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
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 .
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
5 − 3, 02 C
; B = 1, 05 = 0.415 MN . ȳD
5
Fluid pressure 209
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
= 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
RV (a − e) = MBG − MBW
MBG − MBW 31870 − 5625
; e=a− =7− = 0.422 m .
RV 3990
2a
h MD 2a
D
3
a
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 .
↑: 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
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 .