FEM Practice

FEM for Engineering Applications
Exercises with Solutions

Jonas Faleskog
KTH Solid Mechanics
August 2008
1. Elastic Energy and Energy principles

2. Matrix formulated structural mechanicsdirect method
3. Strong/weak form and FEM-equations
4. FEM: trusses and beams
5. FEM: planar frames of trusses and beams
6. FEM: 2D/3D solids
7. FEM: heat conduction
FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog
1. Elastic Energy and Energy principles

1.1 A simply supported beam of length 3L and with
bending stiffness EI is subjected to an external
bending moment M0, see the figure to the right.
Determine the rotation by use of Castiglianos
theorems. Neglect possible contributions from
shear forces when evaluating the complementary
elastic energy.
2L
M0
A,
E,
EI, L
1.4 A plane structure is composed of four equal beams

as shown to the right. The beams are of the length L and
have the bending stiffness EI. A moment M0 is applied
at the centre point of the structure. Use a suitable
energy method based on complementary elastic energy
to determine the rotation of the centre point.
M0
1.5 A cantilever beam of length L and with bending stiffness EI is

clamped at its left end, and attached to a spring at its right end. The
3
spring constant is k = EI L . A point force, P, is applied on the
beam according to the figure to the right. Determine, by use of a
suitable energy method, the deflection of the right end of the beam.
1.6 A beam of length 2L and with bending stiffness EI is

subjected to a bending moment applied at its right end,
see the figure to the right. The left end of the beam is
clamped, and the mid point is attached to a vertical spring
3
with spring constant k = 6EI L . Evaluate the rotation
of the right end of the beam by use of an energy method.
1.1 (10)
EI, L
45o
L
45o
A,
1.3 Determine the horizontal displacement at

point B in the beam structure to the right. All
beams are of length L and have the bending
stiffness EI. Use complementary elastic energy
and Castiglianos theorems in the analysis.
E,
1.2 Two beams, each with bending stiffness EI and

length L, are connected at B, see the figure to left.
The left beam is clamped to a rigid wall at A and a
point force is applied on the right beam at point C.
Determine the rotation of the corner B by use of an
energy method (neglect possible contributions
from normal forces and shear forces to the energy
expression).
EI, L
k
L
EI
L
M0

1.7 A beam with bending stiffness EI and total length 2L,
L
is simply supported at its mid point. The left end of the
beam is attached to a linear spring with the spring constant
3
k = EI L . The beam is subjected to a point force P0
k
and a moment M0. Determine M0 such that the deflection
of the right end of the beam becomes zero. Carry out the
analysis using an energy method, based on complementary elastic energy.
L
P0
P
(b)
EI, L
EI, L
kN
2
EI, L
(a)
EI, L
1.8 A system of two beams, each with

bending stiffness EI and length L, and a
spring kN, see Figure (a) to the right, has
proven to be to compliant in an application.
The system is therefore made stiffer by use
of a torsion spring kM, see Figure (b). The
complementary elastic energy for the system is
M0
kM
kN
2
L 1 2
M 2
1 ( P M L ) - 1--- ------------------(M L) W = ------ --- P + P ----- + --- ---------------------------+
,
EI 6
2 k M L EI
L
2 k L3 EI
N
where M is the moment arising in the torsion spring when the system is loaded by a point force
3
P. The stiffness of the springs can be expressed as k N = EI L and k M = EI L , where
= 3 in the current application. Determine , such that the stiffness of the system increases by a
factor of two, i.e. such that the displacement at P in (b) becomes half compared to the case in
(a). Hint: problem (b) is statically indeterminate.
1.9 The constraint and the boundary conditions of a beam

with bending stiffness EI and length 2L is shown in the
right hand figure. Determine the vertical displacement of
the beam at the point force. Use an energy based method.
1.10 A beam with bending stiffness EI and total

length 3L is subjected to a uniformly distributed load
with the resultant Q, see the figure to the right.
Determine all reaction forces acting on the beam.
Use an energy method, based on the complementary
elastic energy, for statically indeterminate quantities.
1.11 A freely supported beam with bending stiffness EI and total length 3L is loaded by a point
force, P, according to the right hand figure. Determine the deflection at point B.
1.2 (10)
P
L
Q
2L
P
A
1.12 A plane frame is composed of three beams connected

at the stiff joints B and C, see the figure to the right. The
bending stiffness of all beams are EI. The frame is loaded
by a point force. Determine the vertical displacement of
point C and evaluate the distribution of bending moment in
the frame.
2L
B
D
L
2L
C
P
The examples below are taken from Exempelsamling i Hllfasthetslra , Eds. P.-L. Larsson & R.
Lundell, KTH, Stockholm, january 2001. The solutions to these problems (in Swedish) are based on
Castiglianos theorems.
L
1.13 Determine the horizontal displacement at point B. The
bending stiffness of each beam in the planar frame is EI.
B
L
A
1.14 A planar frame constructed by two beams, each

with bending stiffness EI, is loaded by a uniformly
distributed load with the resultant P and a point force
P according to the right hand figure. Calculate the
vertical displacement at the point force.
1.15 A beam with circular cross section (diameter

d) is shaped as a U, see the figure to the right. The
beam is clamped at point A and loaded by a point
force P, acting perpendicular to the plane of the
beam, at point D. Calculate the displacement at
point D in the direction of the point force.
D
P
C
L
1.3 (10)
1.16 A planar frame, according to the right hand fig- A

ure, is clamped at point A and point B. The frame is
loaded perpendicular to its plane by a couple of point
forces, each of magnitude P. The beam has a circular
tube shaped cross section with mean radius r and
thickness t (assume that t r ). The material has the
elastic modulus E and the shear modulus G. Calculate
the out-of-plane displacement at points B and C.
Consider the case when L = l.
A
B
L
P
B
l
L
C P
1.17 A rectangular planar frame with measurements

according to the figure is clamped at point A and point D.
Two point forces is applied perpendicular to the frame in
opposite directions at points B and C, respectively. The
bending stiffness of the frame is everywhere equal to EI
and its torsional stiffness is GK, with EI ( GK ) = ,
where being a non-dimensional constant. Calculate the
displacements perpendicular to the plane at the points B
and C, respectively.
C
L
D
2L
1.18 A planar frame of quadratic shape is freely supported at the

corner points A, B, C and D, such that only reaction forces perpendicular to the frame may occur. The side of the frame is of
length L. The bending stiffness and the torsional stiffness are EI
and GK, respectively. The frame is subjected to a uniformly distributed load with the resultant Q acting perpendicular to the
frame on the side AB. Calculate the displacement at the mid
point between A and B.
L, EI, GK
1.4 (10)
Solutions
1.1
Introduce reaction forces: RA
RB Equilibrium R = M ( 3L )
B
0
gives:
RA = R B = M ( 3L )
M0
M
RB = ------03L
M0
M1 M0 M2
M2
M1
RA = ------3L
Equilibrium requires:
M 1 = M0 3
2
M 2 = 2M 0 3
2
LM
2LM
M0L
M0 L
W
- = ---------Complementary elastic energy: W = ----------1- + -------------2- = ---------= ---------6EI
6EI
6EI
M 0
3EI
1.2
Introduce a fictitious bending moment, MF
and reaction forces, R & MR
Equilibrium:
M1 = M F + PL 2
MR
MF
P
Equilibrium:
R=P
MR = MF
M1
M2
P
M1
M2
MF
Complementary elastic energy:

2
M2 = PL 2
P
P
MF
2 3
MF L M F PL
P
L-----------------------------------+
+
W =
2EI 2 2EI 6EI
Rotation at B:
W
B = ----------MF
MF = 0
PL
= ---------------2 2EI
1.3
1 statically indeterminate, chose e.g. MR
MR
M0
2
L
2 2
W = --------- ( M R M R PL + 2P L )
6EI
M0
WPL
---------= 0 M R = ------MR
2
M0
Displacement at point B:
Equilibrium: M 0 = PL
1.5 (10)
W
7 PL
= -------- = ------ --------P
12 EI

1.4
Cut and use the symmetry
properties!
M0/4
V
MR
M0/4
M0
------- M R VL = 0
4
=> 1 statically indeterminate, e.g. MR
V M0/4
Equilibrium:
M0/4 V
M0
Total complementary elastic energy:
M0/4
M
M 2
2
L
W = 4 --------- MR + M R ------0- + ------0-
6EI
4 4
W---------=0
M R
M
M R = ------0- ;
8
Rotation:
M0L
W
= ---------- = ----------M0
16EI
1.5
P
Equilibrium, beam:
MR
Total complementary elastic energy:

2
L 2 N
W = Wbeam + Wspring = --------- MR + -----V
6EI
2k
N
: MR PL + NL = 0
2
L
N
2
=> 1 statically indeterminate, = --------- ( PL NL ) + ----------------------36EI
EI
2EI L
chose e.g. N
k = -----33
L
3
L
2
2
= --------- P + N 1 + --- 2PN
6EI
WP
------Principle of least work:
= 0 N = ------------- ;
N
3+
3
W
PL
1
P = -------- = --------- ----------------Deflection at the right end (Castiglianos 2nd theorem):
P
EI ( 3 + )
: V N + P = 0
1.6
6EI
k = -------3
L
Equilibrium gives
R
MR
M0
M0
M0
M0
M0 M R
N = -------------------L
Complementary
elastic enerrgy:
Note! one statically

indeterminate exists!
2
2
2
2
2
2
L
L
N
L
W = --------- ( MR + M R M0 + M0 ) + --------- ( 3M0 ) + ------ = --------- ( MR + 3M 0 )
6EI
6EI
2k
4EI
WL
---------Condition to determine the unknown
= --------- 2M R = 0 M R = 0
M
4EI
R
reaction force:
W
3 M 0 LRotation at the point where the external moment is applied :
= ---------- = --- ---------M 0
2 EI
1.6 (10)

1.7
R
Free body diagram:
M0
M1
P0
M2
Equilibrium:
EI
k = -----3L
Equilibrium:
: M 0 P0 L + NL = 0
M1 = NL = P 0 L M 0
N = P0 M0 L
Complementary elastic energy:
P0
M2 = P0 L
3
3
2 2P M
2 2P M
M1 L M2 L N2
L - 2 M
L - 2 M
0
0 0
0
0 0
- + ----------- + ------ = -------------------------------------------W = ---------2P
+
+
P
+
---------------6EI 6EI 2k
6EI 0 L
L 2EI 0 L
L
W
P0 = --------- = 0 (given condition)
P0
3 + 2
M0 = ---------------- P0 L
3+
1.8
Case (a): no torsion spring (M = 0), kN = 3EI / L3
W
= -------P
M=0
PL
= --------EI
Case (b): statically indeterminate problem,

where M is an internal indeterminate quantity, thus
3
W
2 -------- = 0 M = PL --------------M
3 + 2
W
L
2M
PL 9 + 2
b = -------- = ------ P --- ----- = --------- ---------------P
EI
3L
EI 9 + 6
9
1
b = --- a = --2
2
According to the given conditions:
1.9
2P
P
P
M0
M0
M0
L
MTotal complementary elastic energy: W = 2 -------dx =

2EI
0
: M 0 = PL
Equilib:
(statically indeterminate probl.)
M( x)
Displacement at the point force (Castiglianos 2nd theorem):
1.7 (10)
2 3
x
P L
= PL --- = -----------L
3EI
3
W
2 PL
P = -------- = --- --------P
3 EI

1.10
Q
RA
MA
RB
Equilibrium: RA + RB Q = 0
M A + RB L Q2L = 0
Qx/L
QL x 2
M ( x ) = -------- ---
4 L
M(x)
x
Note! one statically

indeterminate, choose
for instance MA.
2L M ( x )
L
2
Complementary elastic energy: W = --------- ( M A + MA QL + ( QL )2 ) + --------------- dx
6EI
2EI
0
3Q
5Q
QL
W
----------- = 0 M A = -------- ; R A = ------- ;
R B = ------2
2
2
M A
1.11
Introduce a fictitious force Q when the comple-
mentary elastic energy, W , is calculated.

L
2
2
The comp. elastic energy in the beam becomes: W = ------------ ( 4P + 7PQ + 4Q )
18EI
3
W
7- PL
------------------=
Displacement at B (Castiglianos 2nd theorem):
B =
Q Q = 0
18 EI
1.12
RA
RD
2 equilibrium Eqs. and 3 unknown reaction forces (RA, RD and MD). Thus, the
problem has one statically indeterminate. Treat RA as known when calculating the complementary elastic energy.
MD
P
L
2
2
Complementary elastic energy in the beam: W = --------- ( 76RA 40RA P + 8P )
6EI
W
5
--------The unknown RA is given by:
= 0 RA = ------ P
RA
19
3
--------Displacement in point C (Castiglianos 2nd theorem): C = W
-------- = 52
------ PL
P
57 EI
Bending moment diagram:
10
------ PL
19
18
------ PL
19
10
------ PL
19
10
------ PL
19
1.8 (10)
1.13
1.9
1.14
1.15
1.16
1.9 (10)
1.18
1.17
1.10 (10)
2. Matrix formulated structural mechanicsdirect method
k3
k1
2.1 A system of five springs are connected as

shown in the figure to the right. All the spring
constants ki are in the current application
equal to k. Furthermore, D1 = D4 = 0, F3 = 0
and F2 = P. Determine the displacements and
reaction forces.
k4
k5
k2
D1, F1
D2, F2
D3, F3
D4, F4
4a
2.2 Three springs are connected according

to the figure to the right, also showing the
applied external force P. The spring constants are: k1 = 5k, k2 = k and k3 = 2k.
Determine the displacements at the points
where the springs are connected and evaluate all the reaction forces.
k2
k3
k1
3a
rigid beam
P
2.3 Determine the displacement at the

point force in the spring system shown to
the right.
45o
k1
45o
k1 = k
k2
45o
k2 = 2k
k3
k3 = 2k
2.4 The plane structure to the right consists of four

springs with spring constants: k1 = k2 = k4 = 2k and
k3 = 4k. The springs are attached to five nodes with
coordinates shown in the right hand figure. The structure is loaded by two point forces acting in node 4
and node 5, as shown in the figure. Calculate the displacement at each node and the normal force acting
in spring element number 4.
(1,1)
y/L
(1,1)
k1
k2
x/L
P
5
3
(1,0)
k3
k4
(1,1)
2.1 (12)

1
2
2.5 A plane truss structure consists of three truss elements connected to four nodes, as shown to the right. All trusses have cross
sectional area A and elastic modulus E. The length of each truss
element is evident by the figure. A point force, P, is acting on node
4. Calculate the displacements at the nodes and the reaction forces
at nodes 1 and 2, respectively. Show also that global equilibrium is
satisfied in the vertical direction.
L/2
3
L
4
-
2
30o
30o
k1
2.8 A mass m0 is attached to three similar springs. The
division between the springs is 120o and the spring constants are k1 = k2 = k3 = k. The springs are attached to a y
rigid ring of radius R. The coordinate system shown in the
figure is fixed to the ring, where the y-axis is located in
k1
k
the direction of spring k1. The circumferential position of
m0 3
the ring is determined by the angle . Calculate the displacement of m0 in the x- and y-directions for an arbitrary
k2
angle. The acceleration of gravity g (see the figure) is
assumed to be known.
Hint: derive the equation system with reference to the given xy-coordinate system.
2.2 (12)
EA, L
L
A,
E,
2.7 The plane structure in the figure to the right contains

two truss elements and two spring elements. The truss
elements have the same length L, cross sectional area A
and elastic modulus E. The stiffness of the left spring is
k 1 = 0.75 EA L . A point force P is applied on the structure, acting at an angle , as shown in the figure. Determine the stiffness of the right spring, k2, such that the
displacement always is aligned with the force P, i.e. in the
direction of the angle .
E,
A,
L
2.6 The plane frame structure to the right contains two

truss elements and two spring elements. The spring constant for both springs is k = EA L . The truss elements are of length L, have cross sectional area A and
elastic modulus E. The structure is subjected to a point
force P according to the figure. The displacement will
for the present structure always be in the direction of the
force P. Determine the relation between and P.
EA,L
k2
x
rigid
ring

2.9 An adjustable crane consists of two rods which are connected at node
3, see the figure to the right. The elastic modulus of the rods is E. Rod e1
has a cross sectional area A and a length l. Rod e2 is composed of two
cylindrical tubes to facilitate adjustment of its length by use of a hydraulic actuator, where its length L is given by the angle . The effective cross
sectional area of rod e2 is 3A . The relations = 2 and L = 2l cos
are valid here. Determine the displacement of node 3 and the normal
force acting in rod e2 (the force acting on the hydraulic actuator), when
the crane is loaded by a mass m for the case = 30 . The acceleration
of gravity is known and denoted g.
Hint: the normal force can be determined by use the reaction forces acting on node 2.
e1
l
e2
2P
E, 2A, 2L
E, A, L
2.10 A structure of three truss elements is loaded by two point

forces (P and 2P), see the figure to the right. The elastic modulus, the cross sectional area and the length of each truss are
shown in the figure. Analyse the structure by use of a matrix
formulated method and determine the reaction forces at all the
nodes.
E, A, L
.
2.11 The truss structure to the right contains two truss elements and
one spring element, with spring constant k = 2EA L . The structure
is loaded by a point force P according to the figure. The truss elements are of length L, have cross sectional area A and elastic modulus
E. Determine the displacements at the nodes where the elements are
connected. Evaluate also all reaction forces.
90o
45o
EA
L
45o
EA, L
The examples below are taken from Exempelsamling i Hllfasthetslra , Ed. P.-L. Larsson & R.
Lundell, KTH, Stockholm, january 2001.
2.12 Determine the displacements and the reaction

forces at the nodes.
Node
x/L
y/L
1
3 2 1 2
32
3
2
3k
2k
y
P
12
2.3 (12)

2.13 The planar truss structure to the right consists of four spring elements, each of length L and
with spring constant k. All springs are oriented
with a 45o angle with respect to the horizontal
plane. Determine all displacements and reaction
forces at the nodes.
2
y
P
2.14 Determine the displacements and the reaction forces at the nodes, and the normal forces in
the springs.
Node
x/L
y/L
-1
k
y
2
x
2.15 Determine the displacements and the reaction forces

at the nodes.
Node
x/L
y/L
y
1
2
--k2
3
stel
2.16 Determine the displacements and the reaction

forces at the nodes, and the normal forces in the
spring elements. The stiffness and length of each
spring is shown in the figure to the right.
k-----2
k-----2
k
4
Q
4L, k
3
3L, 2k
5L, 5k
stel
x
1
2.4 (12)
Solutions
2k
2.1 The equation system: k
k
0
R1
k 0 0
P- 3
k k D 2 = P . Solution: D 2 = ----8k 1
0
D3
3k k D 3
R4
k 2k 0
k
3k
k
k
The reaction forces are obtained from Eqs. (1) and (4) as: R1 = R4 = P 2 .
2.2
D6
D4
k
Boundary conditions & prescribed forces:

D1 = D4 = D5 = D6 = 0 and F2 = P, F3 = 0
D3
D5
Equationsystem
2k
5k
D2
D1
D2
Solution:
P
= --------- 7
17k 4
D3
16
12
--k- 16
5 12
0
0
12
19
12
9
0
10
16
12
21
12
5
0
12
9
12
9
0
0
0
0
5
0
5
0
Reaction forces: R1 = 4P 17
R 5 = 4P 17
R1
0
0
P
10 D 2
0 D3 = 0
R4
0
0
0
R5
0
10 0
R6
R4 = 3P 17
R 6 = 14P 17
2.3
D4
D2
D1
e1
Element stiffness matrix: k e
D3
e2
D7
B.C.:
D6
45o
Element 2:
e3
D5
D1 = D 2 = 0
D 3 = D4 = 0
D5 = D 6 = 0
D8 = 0
Assembly:
1 0
0 0
0
K = k1 + k2 + k 3 = k
Element 3:
1 1
1 1
1
0
0 1
1 0 1 1 1
0 0 1 1 1
0
0
1
1
1
1
1
0
1
1
1
1
3
0
0
0
1
1
1
1
0
2
k 3 = 2k,
r= 1 0
0 0
1 2 1 2
1 2 1 2
r=
2
45o
1
0
k 1 = k,
Element 1:
y
D8
= ki r r
r r
k 2 = 2k,
r = 12 12
12 12
x
R1
R2
R3
F =
Equation (7) gives
R4
R5
3kD 7 = P
R6
P
R8
2.5 (12)
Reactionsforces
P
= -----3k

2.4
Boundary conditions:
D3 D1 = D2 = D3 = D4 = D5 = D6 = D7 = D10 = 0
D4
D2
D1
e1
F8 = P, F9 = P
e2
D10
D6
Element stiffness matrices:
D9
D5
e3
e4
D8
D7
2
c = cos
Ke = k i a a , a = c sc ,
2 s = sin
a a
sc s
e1& e4: k 1 = k 4 = 2k,
1
a = { = 45 } = --- 1 1
2 1 1
Assembly of global stiffness matrix:
e2: k 2 = 2k,
K = k
1 1
1 1
1 1
1 1
0
0
0
0
1 1 1 1
1 1 1 1
0
0
0
0
0
4
0
4
0
0
1
1
1
1
0
4
1
5
1
1
1
1
1
1
1
1
1
1
0
1
1
3
1
1
1
1
3
1
a = { = 45 } = --- 1 1
2 1 1
e3: k 3 = 4k,
a = { = 90 } = 0 0
0 1
Eqs. (8) & (9):

D8
k 5 1
= P
1 3 D9
P
D8
D9
P
= --------- 2
14k 4
Normal force in element 4: N = k 4

where k 4 = 2k and = u 2 u 1
D9
u1
u2
D 10
1
1 D
= Tde = ------- 1 1 0 0
= ------- 9
2 0 0 1 1 D7
2 D8
2
-P
N = 3--------7
2.6 (12)
D8

2.5
Boundary conditions: D1 = D2 = D3 = D4 = D5 = D7 = 0; F6 = 0; F8 = P
D2
D1
e1
2
c = cos
Element stiffness matrices: K e = k i a a , a = c sc ,
2 s = sin
D3
a a
sc s
D4
e2
D6
EA
e1: k 1 = -------,
L
1
a = { = 45 } = --- 1 1
2 1 1
2EA
e2: k 2 = -----------,
L
a = { = 90 } = 0 0
0 1
D5
e3
D8
EA
e3: k 3 = -------,
L
1
a = { = 45 } = --- 1 1
2 1 1
Eqs. (6) & (8) (reduced system of equations):
D7
Assembly of stiffness:
1 1
1 1
0 0
EA
0
K = ------2L 1 1 0
1 1 0
0
4
0
4
1
1
0
0
2
0
1
1
1
1
0
4
0
6
1
1
0
D6
EA
------- 6 1
= 0
2L 1 1 D
P
8
0
1
1
1
1
D6
D8
PL
= ----------- 2
5EA 12
1
1
1
1
Reaction forces in node 1 & 2 ( D.O.F.:s 1 - 4):

EA
P
EA
P
R 1 = ------- D 6 = --R 2 = ------- ( D 6 ) = --2L
5
2L
5
0
0
4EA
4P
R3 = 0
R 4 = ----------- ( D 6 ) = ------2L
5
P
4P
------Global equilibrium in vertical dir.: R 2 + R4 + F6 + F 8 = +
+ 0 P = 0 OK!
5
5
2.6
Boundary conditions: D1x=D1y=D2x=D2y=D3x=D3y=D4x=D4y=0
The element stiffness contribution to node 5:
2
2
e2
1
1
e1
2
5
---
e4
2
4
Assembly:
(only node 5)
Eq. system:
D5y
Thus:
2
e3
e1: EA
------- 1 0 ;
L 0 0
e2: EA
------- 0 0 ;
L 0 1
2
c = cos
e3: EA
------- c sc
L
2
s = sin
sc s
cos --- = sin

2
e4:
sin --- = cos

2
EA
------- s s c
L sc c
2
2
EA
EA
K = ------- 1 + ( c + s ) ( sc sc )
= ( 1 + ) ------- 1 0
L
L 0 1
2
2
( sc sc ) 1 + ( c + s )
D 5x
EA
( 1 + ) ------- 1 0
= P cos
L 0 1 D
sin
5y
D5x
PL
= ------------------------( 1 + )EA
2.7 (12)
D 5x
D 5y
PL
= ------------------------- cos
( 1 + )EA sin

2.7
D4
D6
D3
e4
D2
D10 e3
Element sttiffness matrix:

2
c = cos
K e = k i a a , a = c sc ,
2 s = sin
a a
sc s
e1
D1
Boundary Condistions:
D5 D1 = D2 = D3 = D4 = D5 = D6 = 0; F7 = Pcos; F8 = Psin
D9 e2
3
e1: k 1 = --- k 0,
4
a = { = 0 } = 1 0
0 0
e3:
k 3 = k 0,
1
a = { = 60 } = --- 1 3
4
3 3
1
a = { = 0 } = 1 0
e4: k 2 = k 0, a = { = 120 } = --- 1 3
4
0 0
3 3
EA
where k 0 = ------L
Assembly Ki and implementation of B.C. gives the reduced Equation system:
e2: k 2,
1.25k 0 + k 2
0
D9
1.5k 0 D 10
= P cos
sin
Eq. 9: 1.25k 0 + k 2 = P D 0
Eq. 10: 1.5k 0 = P D 0
But D should aligned

with the external force
1.25k 0 + k 2 = 1.5k 0
D9
D 10
= D 0 cos
sin
EA
0.25k 0 = ------4L
2.8
D2
Boundary Conditions: D1 = D2 = D3 = D4 = D5 = D6 = 0
D1
F7 = mg sin , F 8 = mg cos
y
e1
D8
D7
D3
a i ai
a i a i
1
1
a 1 = 0 0 , a 2 = --- 3 3 , a 3 = --- 3 3
4
4
0 1
3 1
3 1
D6
Ki = k i
where
e3
e2
D4
Element stiffness matrix:
D5
mg
Assembly of the stiffness matrix:
K = k
a1
a1
Eq. (7) & (9):
a2
a2
a3
a 3
3--- 1 0 D 7
k
= mg sin
2 0 1 D
cos
8
a1 a2 a3 a1 + a2 + a3
2.8 (12)
D7
D8
2mg
= ----------- sin
3k cos

2.9
D6
o
Boundary Cond.: D1 = D2 = D3 = D4 = 0, F 5 = 0, F6 = mg
D5
60 e1
D4
D3
mg
30o e2
a i ai
Ki =
EA
EA
where a 1 = ------- 3 3 , a 2 = ------- 1 3
4l
4l
3 1
3 3
ai ai
Assembly of stiffness matrix:
D2
x
D1
K =
a2
a2
a1
a1
3
EA
= ------4l
a2 a1 a1 + a2
3
1
1 3 3 3
1 3
3 3
3 3
3 1
4
3 3 3 1 2 3
EA- 2 3 D 5
-----= mg 0
2l
1
3 2 D6
Eq. (5) & (6):
D5
D6
2 3
4
2mgl
= ------------- 3
EA 2
Reaction forces
EA
3
EA
3
R1 = ------- ( 1D 5 3D 6 ) = ------- mg, R 2 = ------- ( 3D5 3D 6 ) = --- mg
at node 1:
4l
2
4l
2
R2
N2
Equilibrium in y-dir. gives: R2 + N 2 cos 30 = 0
N 2 = 3mg
R1
2.10
2
2
B.C.: D1 = D2 = D3 = D6 = 0, F4 = 2P, F5 = P
D4
D3
2
e2
c
Element stiffness matrices: K e = k i a a , a = c sc ,
2 s
a a
sc s
EA
1 D6
e1: k 1 = -------, a = { = 0 } =
L
D
e3
D2
D1
e1
2 3
EA
e2: k 2 = -------, a = { = 90 } = 0 0
L
0 1
Assembly of
stiffness:
Reaction forces:
2
0
EA
K = ------- 0
2L 0
2
0
0
2
0
2
0
0
0
0
1
1
1
1
7P
EA
= ------- ( 2D ) = ----------
(1):
5P
EA
R 1 = ------- ( 2D 5 ) = ------4
2L
(2):
R2
2L
0
2
1
3
1
1
2
0
1
1
3
1
0
0
1
1
1
1
= cos
= sin
1 0
0 0
EA
1
e3: k 2 = -------, a = { = 135 } = --- 1 1
L
2 1 1
Eqs. (4) & (5):
EA 3 1 D 4
------= 2P
2L 1 3 D
p
5
D4
D5
PL
= ----------- 7
4EA 5
(3):
P
EA
R 3 = ------- ( D 4 D 5 ) = --4
2L
(Node 2)
(6):
P
EA
R 6 = ------- ( D 4 D 5 ) = --4
2L
(Node 3)
(Node 1)
2.9 (12)

2.11
D6
3
1
D4
D5
1
e3
2
e1
2
D3
e2
D2
2
c = cos
Element stiffness
K e = k i a a , a = c sc ,
matrices:
2 s = sin
a a
sc s
1
D1
Boundary conditions: D2 = D3 = D5 = D6 = 0,
F1 = P, F4 = 0
EA
1 1 1
e1: k 1 = -------, a = { = 45 } = --L
2 1 1
EA
1 1 1
e2: k 2 = -------, a = { = 45 } = --L
2 1 1
2EA
k 3 = -----------, a = { = 0 } = 1 0
L
0 0
e3:
2
0
EA 1
K = ------2L 1
1
1
Assembly:
Reaction forces:
0
2
1
1
1
1
1
1
5
1
4
0
1
1
1
1
0
0
1
1
4
0
5
1
1
1
0
0
1
1
Eqs. (1) & (4):

EA 2 1 D 1
------= P
2L 1 1 D
0
4
D1
D4
2PL
= ---------- 1
EA 1
(2):
EA
R 2 = ------- ( D 4 ) = P (Node 1)
2L
(5):
EA
R 5 = ------- ( D 1 ) = P
2L
(3):
EA
R 3 = ------- ( D 1 + D 4 ) = 0 (Node 2)
2L
(6):
EA
R 6 = ------- ( D 1 ) = P
2L
(Node 3)
(Node 3)
2.12
6
3
4
3
k
K = --4
8
7
0 0
0 0 3 1
0
0
0
0
0
0
2 3
0
0
3 3
3 1
6
2 3
2 3 0 0
2
0
0
6 2 3
0 0 2 3 2
0 0
0
0
0 12 0 12
2 3 0 0
2
3 3
D 1 = D2 = 0
D 3 = D4 = 0
D 5 = D6 = 0
F 7 = P, F 8 = 0
0 12 3 15
F1
F2
Equationsystem:
D = K F
D7
D8
P 15
P
= --------= --- 0.4545
33k 3
k 0.0525
Reactionforces:
F3
F4
F5
F6
2.10 (12)
12
0.3636
4 3
0.2099
21
P
= -----= P 0.3664
33 7 3
0.3674
0
0
0.1575
3 3

2.13
4
1
3
10
6
9
8
1
1
0
0
k 0
K = --2 0
0
0
5
1
1
1
1
0
0
0
0
0
0
1
1
0
0
1
1
0
0
0
0
1
1
0
0
1
1
0
0
0
0
1
1
0
0
0
0
1
1
1
1
0
0
0
0
0
0
1
1
1
1
0
0
0
0
0
0
1
1
2
0
1
1
0
0
0
0
1
1
0
2
1
1
1
1
1
1
0
0
1
1
3
1
1
1
1
1
0
0
1
1
1
3
D1 = D2 = 0
D3 = D4 = 0
D5 = D6 = 0
D8 = 0
F 7 = 0, F 9 = P, F 10 = 0
F1
F2
Equationsystem:
D = K F
D7
D9
D 10
Reactionforces:
2
P
= ------ 5
6k
1
F3
F4
F5
F6
F8
2
2
3
P
= --- 3
6
1
1
2
2.14
2
1
1
e1
6
2
Equationsystem:
5 e2
1
2
1
D = K F
1
1
k
K = --- 0
2 0
1
1
1
1
0
0
1
1
0
0
2
0
2
0
0
0
0
0
0
0
1
1
2
0
3
1
1
1
0
0
1
1
D1 = D2 = 0
D3 = D4 = 0
D6 = 0
F 5 = P
F1
Reactionforces:
2P
D 5 = ------- 1
3k
F2
F3
F4
F6
The normal force, N, in one element is given by f e = kTD e ,
1
1
P
= --- 2
3
0
1
where N = f2 .
D1
Element 1:
D2
1
k = k 1 1 ; T = ------- 1 1 0 0 ; D e =
2 0 0 1 1
1 1
D5
f = ------k2- DD +DD +DD +DD N = ------32-P

1
D6
Element 2:
D5
D6
k = k 1 1 ; T = 1 0 0 0 ; D e =
1 1
00 1 0
D3
f = k DD +DD N = 2--3- P
D4
2.11 (12)

2.15
4
3
8
1 1 + 2 0
0
0
0
0
0
0
0
2
0
1
0
1
0
k
K = ---------- 0
2 2 0
7
6
stel
0
0
0 2 0
1
1
0
Equationsystem:
D1 = D2 = 0
(p.g.a. stel bom)
F1
F2
D6
D7
F4
F5
3 24
0 2 2
1
1
1 1 + 2 1
1 1
2
1
1
0
0 2+2 2
6+4 2
0.3431
Q
Q
= ---- 3 + 2 2 = ---- 0.1716
k
k
0.6569
54 2
3 2+4
6 28
Q
= ------0
2
85 2
F3
0
0
0
1
0 2 2 1
D8
Reactionforces
F 6 = F 7 = 0, F 8 = Q
D = K F
D3 = D4 = 0
D5 = 0
0 2 2 0
0 0
1
0.1716
0.3431
= Q
0
0.6569
0.1716
2.16
4
6
2
e3
1 2
rigid
Equationsystem: D
= K F
5
1 e2 2
e1
1
1
D2
D3
16
12
3
k 16
K = --5 12
0
0
12
19
12
9
0
10
16
12
21
12
5
0
12
9
12
9
0
0
0
0
5
0
5
0
0
10
0
0
0
10
D1 = 0
D4 = 0
D5 = D 6 = 0
F 2 = Q, F 3 = 0
F1
Reactionforces
Q
Q
= --------- 7 = ---- 0.4118
17k 4
k 0.2353
(rigid support)
F4
F5
F6
20
0.2353
Q 15
= -----= Q 0.1765
85 20
0.2353
70
0.8235
The normal force, N, in one element is given by f e = kTD e ,

where N = f2 .
D1
Element 1:
D2
1
k = 5k 1 1 ; T = --- 4 3 0 0 ; D e =
5
1 1
0 0 43
D3
4D
5
f = k 3D
N = ------ Q
17
3D + 4D
2
D4
D5
Element 2:
D6
1
k = k 1 1 ; T = --- 1 0 0 0 ; D e =
5 0 0 1 0
1 1
D3
f = k DD N =
3
4
------ Q
17
D4
D1
Element 3:
D2
k = 2k 1 1 ; T = 0 1 0 0 ; D e =
1 1
0 00 1
D5
D6
2.12 (12)
f = 2k DD N = 14-----17- Q
2
3. Strong/weak form and FEM-equations

3.1 The solution to a specific one-dimensional problem is governed by the differential equation (strong form)
d- d
----D ------ q + Q = 0
dx dx
for x 1 x x 2 ,
where the primary variable depends on x. Also D, q and Q may depend on x. Derive the
weak form and identify the essential and natural boundary conditions.
d
d
3.2 The weak form to ------ D ------ q + Q = 0 is
dx dx
x2
x2
x2
dv
------ D d
------ dx + vqdx =
dx dx
d
vQdx + vD -----
dx
x1
x1
x1
x2
,
x1
where v(x) is an arbitrary weight function. Derive the FEM-equation for one element. Use a
linear interpolation for the primary variable and use Galerkins method, regarding the choice
of the weight function.
3.3 The figure to the right shows a rod with elastic

modulus E and cross sectional area A. The rod is
loaded by a body force, Kx [N/m3]. The displacement, u, in the rod is given by the solution to the differential equation
Kx
E, A
x
x=L
x=0
d- du
----EA ------ + AK x = 0 .
dx dx
L
dv du
(a) Show that the weak form is ------ EA ------ dx =
dx dx
0
vK x Adx + [ v ( A ) ]0 ,
where denotes the normal stress and v is an arbitrary weight function.

(b) Derive the FEM-equation (use Galerkins method) to the weak form above for one element, i.e. identify the quantities in the equation
k e d e = fe .
Kx
(c) The rod shown to the right is of length 3L and

loaded by K x = Q ( 2AL ) ( x L 1 ) , where
Q corresponds to the total axial force acting
on the rod. Both ends of the rod are clamped.
Divide the rod into two elements of lengths L
and 2L respectively and determine the node
displacements and the reaction forces. Compare with the exact solution. Redo the analysis with more elements!
3.1 (15)
E, A
x=0
x=L
Exact
soln. u ( x ) =
2 QL x
--- -------- --
9 EA L
1- QL
------------ 3 --x 36 EA
L
x = 3L
0xL
x 2
x 3
+ 9 --- 3 ---
L
L
2Q
7Q
N ( x = 0 ) = ------- , N ( x = 3L ) = ------9
9

3.4 The right figure shows a uniaxial bar coupled to
a set of continues springs with spring constant per
unit length kx [(N/m) / m]. The bar has elastic modulus E, cross sectional area A and is loaded by a body
force Kx [N/m3]. The displacement, u, in the bar is
given by the solution to the differential equation
Kx
E, A
x
kx
x=0
x=L
d
------ EA du
------ k x u + AK x = 0 .
dx dx
(a) Show that the weak form is
L
L
dv
------ EA du
------ dx + vk x udx = vK x Adx + [ v ( A ) ] 0 ,
dx
dx
0
where denotes the normal stress and v is an arbitrary weight function.

ke d e = f e .
(c) Divide the bar into two linear elements of the same length and analyse the problem.
Evaluate the node displacements. Apply the boundary conditions: u = 0 for x = 0 and
A = Q for x = L. Assume that E and A are constants and that the spring constant
2
k x = 3EA ( 2L ) and the body force K x = Q ( AL ) .
3.5 Figure (a) to the right shows a uniform bar loaded

by its dead weight, g, where is the density of the bar
and g is the acceleration of gravity. The bar has elastic
modulus E and cross sectional area A. The displacement, u, is given by the solution to the differential
equation
d
du
------ EA ------ + A g = 0 .
dx
dx
(b)
EA
x
g
x=L
P
x = 2L
x=L
(a) Assume that E, A, and g are constants and

show that the weak form is
L
(a)
x = 3L
L
dv du
EA ------ ------ dx = [ v ( A ) ]0 + gA vdx ,
dx dx
0
where is the normal stress in the bar and v an arbitrary weight function.
ke d e = f e .
3.2 (15)

(c) In an application, a bar (E, A) is connected to a linear spring with spring constant k, see
Figure (b) above. The bar is loaded by a point force applied at the x = L and by its dead
weight. Divide the bar/spring structure in three elements with nodes placed in the
points: x = 0, L, 2L and 3L. Thus, the bar should be divided into two equal elements.
Let k = 2EA L , where EA is constant, and calculate the displacements at the nodes.
3.6 The figure to the right shows a beam with bending
stiffness EI attached to an elastic foundation characterized by a spring constant per unit length kz [(N/m) / m]. A
distributed load per unit length q [N/m] is applied on the
beam. The deflection of the beam w is given by the solution to the differential equation
2
z,w
q
x
kz
x = L
d - --------d w
------ EI
+ kz w q = 0 .
2
2
dx dx
x=L
T
M
(a) Show that the weak form is

L
w
d v- d--------------EI
dx +
2
2
dx
dx
vk z wdx = [ vT ]
dv
------ M
dx
+
L
vqdx ,
where v is an arbitrary weight function, T is a shear force and M is a moment. The relations T = ( EIw ) and M = EIw have been utilized at the boundaries.
(b) Derive the FEM-equation (use Galerkins method) to the weak form above for one element, i.e. identify the quantities in the equation ke d e = f e .
(c) Divide the beam into a two-node beam element (2
D.O.F. per node) and determine w ( L ) . Assume
4
that EI is constant, k z = EI L and q = 0. The
boundary conditions are shown in the figure to the x = L
right.
z,w
x
M0
x=L
(d) The beam shown below is subjected to a uniformly distributed load q = Q ( 2L ) ,

where Q is the resultant of the total distributed load acting on the beam. The total
length of the beam is 2L and its bending stiffness is EI. The left end of the beam is
clamped, whereas the right end support is flexible, here modelled by a combination of a
tension spring with stiffness kw [N/m] and a torsion spring with stiffness k [Nm].
Model the beam with one beam element (2 node element with 2 D.O.F. per node) and
calculate the deflection and rotation of the right end of the beam. Use the values of the
spring constants shown in the figure.
Q
2L, EI
3.3 (15)
kw
3EI
EI
k w = --------3- k = -----L
2L

3.7 The figure to the right shows a beam, which is loaded
by its dead weight per unit volume g, where is the density and g acceleration of gravity. The beam has the elastic modulus E, moment of inertia I and cross sectional
area A. The deflection of the beam w is given by the solution to the differential equation
2
z,w
g
x = L
x=L
T
M
d - --------d w
------ EI
+ gA = 0 .
2
2
dx dx
(a) Assume that , g and A are constants and show that the weak form is
L
L
w
dv
d-------v- d-------------- M
EI
dx
=
[
vT
]
gA vdx ,
2
2
dx
L
dx
dx
L
L
where v is an arbitrary weight function, T is shear force and M is moment (defined

according to the figure above). The relations T = ( EIw ) and M = EIw have
been utilized at the boundaries.
(c) The figure to the right shows a cantilever beam
attached to a linear spring with spring constant
g
3
k = EI L . The beam is loaded by its dead weight
and by a point force P. Analyse the beam by use of a
two-node beam element and calculate its deflection
and rotation (slope) at x = L for the special case that x = L
= 1 2 , P = gAL and EI = constant.
3.8 The figure to the right shows a beam with bending stiffz,w
ness EI subjected to a distributed triangular load acting downwards with a total resultant force equal to Q. The beam has
elasticity modulus E and moment of inertia I. The deflection
x = L
of the beam (vertical displacement), w, is given by the soluT
tion to the differential equation
2
z,w
k
x
P
x=L
x
x=L
T
d - --------d w
x Q
------ EI
+ 1 + --- ------ = 0
2
2
L 2L
dx dx
(a) Show that the weak form to the differential equation is
L
L
d-------v- d--------w
dv
EI
dx
=
[
vT
]
------ M
Q
2
2
dx
L
dx
dx
L
1
x dx
v --- 1 + --- ------ ,
2
L L
where v is an arbitrary weight function, T is a shear force and M is a moment (see the
figure above). The relations T = ( EIw ) and M = EIw have been used at the
boundaries.
(b) Derive the FEM equation of the weak form above for one element, i.e. identify the
quantities in the equation ke d e = f e (use Galerkins method).
3.4 (15)

x=L
(c) In an application, the beam is clamped at x = L and
z,w
x
=
L
the rotation is prevented at x = L, see the figure
Q
x
to the right. The beam is subjected to the triangu2L, EI
lar load and a point force according to the figure.
P
Analyse the beam by use of one 2-node beam element and calculate the deflection w(), where is natural coordinate defined as
= x L.
3.9 A one dimensional model of a cooling fin is shown

to the right. The cooling fin has a cross sectional area A
[m2], length 3L/2 and coefficient of thermal conductivity k [W/m/oC]. The convection coefficient is h [W/m2/
o
C] and the perimeter of the fin is P [m]. The temperature distribution in the fin T [oC] at steady state conditions is given by the solution to the differential equation
Convection
Heat conduction
T(x)
x = 0
x = 3L 2
d- dT
----kA ------ + qA hP ( T T ) = 0 .
dx dx
Here, q [W/m3] is a continues distributed heat source and T is the ambient temperature (the
last term represents convection to the surrounding medium).
3L 2
3L 2
dv
------ kA dT
------ dx +
dx dx
3L 2
vhPTdx = [ v ( Q ) ]
3L 2
0
v ( qA + hPT )dx ,
where v is an arbitrary weight function and Q is heat flow, where Q = k AdT dx

(Fouriers law) has been used at the boundaries.
(c) Divide the cooling fin into three linear elements (two nodes per element and one temperature d.o.f. per nod) and determine the temperature at the nodes. The boundary conditions are described by T = 4T at x = 0 and T = T at x = 3L 2 . Assume that k,
2
A and P are constants, q = 0 and that hP = 12kA L .
(d) Change the boundary conditions in x = 3L 2 from prescribed temperature to convection. Assume that the relation between the surface of the perimeter and the surface of
the end of the fin is LP = A 96 .
3.10 The figure to the right shows a model for heat conduction in a one-dimensional rod, where heat exchange
Convection
by convection between the surface of the rod and the
Heat conduction
x
surrounding medium is taken into consideration. The
ambient temperature of the surrounding medium is T . x = x
x = x2
1
T(x)
The rod has cross sectional area A [m2], perimeter P
[m], thermal conductivity k [W/m/oC] and convection
heat transfer coefficient h [W/m2/oC]. The temperature T [oC] in the rod as a function of position at steady state conditions is given by the solution to the differential equation
3.5 (15)
d- dT
----kA ------ hP ( T T ) = 0 .
dx dx
x2
x2
x2
x
dv dT
------ kA ------ dx + vhPTdx = [ v ( Q ) ] x 2 + vhPT dx ,
dx
1
dx
x1
x1
x1
where v is an arbitrary function of x and Q is heat flow, where Q = k AdT dx (Fouriers law) has been used at the left and right boundary of the rod.
(b) Derive the FEM equation of the weak form above for one element, i.e. identify the
quantities in the equation ke T e = fe (use Galerkins method).
(c) Assume that the total length of the rod is L and that k, h, A and P are constants, related
as hPL = 16kA L . Divide the rod into two equal linear elements (two nodes per element with one temperature d.o.f. per node) and calculate the temperature at the nodes.
The boundary conditions are described by a prescribed heat flow Q = hPLT at
x = 0 and a prescribed temperature T = T at x = L .
FORMULAS
1D:
( ) = N1 1 + N 2 2 = N1 N2
N1 = 1
N
L
Beam element:
Deflection:
d3
2L, EI
d2
N2 = L ( 1 + ) 4
N4 = L ( 1 + + ) 4
N 3 = ( 2 + 3 ) 4,
d N
1d N
B = ---------2- = -----2 ---------2dx
L d
w ( ) = N 1 d 1 + N 2 d 2 + N 3 d 3 + N 4 d 4 = Nd e ,
N 1 = ( 2 3 + ) 4,
d4
0
NdN
--1- 1 1
---------- ---- dx dx dx = L
1 1
L
--- 2 1
N Ndx = { dx = Ld } = 6
1 2
T
d1
N2 =
1
3 3L
3 3L
T
1 - 3L 4L 3L 2L
------- B Bdx =
3
2L 3 3L 3 3L
3L 2L 3L 4L
9
x--- dx
1---
1
4L
---------- N 2 1 + L L = 30
21
L
6L
L
3.6 (15)
78
22L
27 13L
T
L - 22L 8L 13L 6L 2
-------N
Ndx
=
105 27 13L 78 22L
13L 6L 22L 8L
Solutions
x2
3.1
(i) Weighted residual:
d
d
v ------ D ------
dx dx
x1
x2
q + Q dx = 0 , where v is an arb. weight fcn.
d
d
d
(ii) Integration by parts v ------ D ------ dx = vD ----- dx dx
dx
x2
x2
dv d
------ D ------ dx gives
dx dx
x1
x1
x1
Weak form:
x2
x2
x2
dv
------ D d
------ dx + vqdx =
dx dx
d
vQdx + vD -----
dx
x1
x1
x1
x2
x1
d
d
= 0 eller D ------ = D -----dx
dx
Essential B.C.
3.2 Approximation function: () = N e , N = 1
p x = x i
0
Natural B.C.
, e =
1
2
d
------ = d
------ d
------ = 1--- dN
------- = B e , where l is the element length
dx d dx l d e
T
Weight fcn. (Galerkins method): v ( ) = N = N where = 1 2 (arbitrary)

T
dv
dv- ----dT 1 dN
T T
------ = ----= --- ---------- = B
dx d dx
l d
Inserted into the weak form gives
1
T
1
T
B DBld + N qNld e =
T
T d
N Qld + N D -----
dx
1
0
kD
T is an arbitrary vector
fs
fQ
kq
k e e = f e dr k e = kD + kq
och f e = fQ + f s
The element matrices becomes:

1
kD
1
1
1
= B DBld = --- 1 D --- 1 1 ld = --- 1 1 Dd
l 1
l
l 1 1
T
kq =
1
T
N qNld =
0
1
fQ =
N Qld = l 1 Qd
1 q
( 1 ) ( 1 ) qd
1 ld = l
2
0 ( 1 )
3.3 (a) See solution to 3.1 and 3.2
3.7 (15)

L
3.3 (b)
k e de = f e , where k e =
3.3 (c)
B EAB dx and fe =
Node/element division:
D1
D2
2 QL
D 2 = --- -------9 EA
2
R 1 = --- Q
9
7
R 3 = --- Q
9
D3
R1
2 2 0 0
EA
------- 2 3 1 D 2 =
Q3
2L
R 3 + 2Q 3
0 1 1 0
Eqn. system:
N K x A dx + [ N ( A ) ] .
3.4(a)
L
Weighted residual:
d
du
v ------ EA ------ k x u + AKx dx = 0 (1)
dx
dx
0
L
du
d
du
Integration by parts: v ------ EA ------ dx = v EA ------
dx
dx dx
0
(2) inserted into (1)

gives the weak form:
dv du
------ EA ------ dx (2)
dx dx
0
L 0
dv
------ EA du
------ dx + vk x udx =
dx
dx
du
vAK x dx + v EA ------
dx
L
0
du
------ = dN
------- de = Bd e
dx
dx
T
T T
T dN
T T
dv
------------- = be B
v = Nb e = b e N
= be
dx
dx
Displacement interpolation: u = Nd e
3.4(b)
Weight function:
T
be
Le
B EAB dx +
N k x N dx de = b e
N K x A dx + [ N ( A ) ] 0
Le
Le
e
ke
T
but b e is arbitrary
k e d e = fe

fe
3.4(c)
Le = L/2
Le = L/2
D1
D2
Elementmatrices:
D3
L
T
2EA
B EAB --- d = ----------- 1 1
2
L 1 1
0
Assembly:
T 3EA L
- N --- d = EA
------- 2 1
N ---------2 2
8L 1 2
2L
1
T Q L
Q 1
EA
ke = ------- 18 15 ; N ------- A --- d = ---AL 2
4 1
8L 15 18
0
reaction force
18 15 0 D 1
1
R
EA
------- 15 36 15 D 2 = Q
---- 2 + 0
8L
4
0 15 18 D 3
1
Q
Equation (2) & (3):
D2
D3
2
B = --- 1 1
L
N = (1 )
QL
QL
= ----------------- 74 = -------- 0.525
141EA 140
EA 0.993
3.8 (15)
point force

3.5(a)
L
d
du
Weighted residual: v ------ EA ------ + Ag dx = 0
dx
dx
(1)
0
L
d
du
du
v ------ EA ------ dx = v EA ------
dx
dx
dx
Integration by parts:
dv du
------ EA ------ dx
dx dx
0
(2)
0
0
L
L
(2) inserted into (1)
dv
du
du
L
with = E ------ gives the weak form EA ------ ------ dx = [ v ( A ) ] + Ag vdx
0
dx
dx dx
0
3.5(b)
du
------ = dN
------- de = Bd e
dx
dx
Displacement interpolation: u = Nd e
v = Nb e =
Weight function:
Inserted into the
weak form gives:
T
be
T
dv
------ = b Te dN
---------- = b Te B
dx
dx
T T
be N
EA B Bdx d e =
T
be
be
is arbitrary
k e de
L
T
[ N ( A ) ] 0 + Ag N dx
0

ke
but

fe
= fe
3.5(c)
Diskretization:
Element matrices:
Truss element:
L1
L2
L3
D1
D2
D3
Element lengths: L1 = L2 = L3 = L
D4
EA
k e = EA B BLd = ------- 1 1
L 1 1
T
Spring element:
Assembly:
Boundary conditions: D1 = D4 = 0
(reaction forces: R1 & R4)
ke = k 1 1
1 1
1
EA
K = ------- 1
L 0
0
1
2
1
0
where
0
1
3
2
0
0
2
2
T
AgL
fe = Ag N Ld = -------------- 1
2 1
0
EA
k = 2 ------L
R1
1
AgL
F = P + -------------- 2
2 1
0
R4
0
D2
EA
Eqs. (2) & (3): ------- 2 1
= P + AgL
L 1 3 D
AgL 2
3
3.9 (15)
D2
D3
PL
AgL
= ----------- 3 + ---------------- 7
5EA 1
10E 4

3.6(a)
L
Weighted residual:
v [ ( EIw ) + k z w q ]dx = 0
(1)
L
L
v [ ( EIw ) ]dx = [ v ( EIw ) ]
v [ ( EIw ) ]dx =
= [ v ( EIw ) ]
[ v ( EIw ) ]
vEIwdx
(2)
(2) inserted into (1) with T = ( EIw ) and M = EIw gives the weak form:
L
vEIwdx +
vkz wdx = [ vT ]
[ vM ]
L
T
B EIB dx +
T
d-------vT d N
------------ = b Te B
=
b
e
2
2
dx
dx
dN
---------- M
L
dx
L
L

fe
ke d e = f e
vqdx
d--------w
d N
= ---------2- d e = Bd e
2
dx
dx
T T
be N
v = Nb e =
Weight function:
Displacement interpolation: w = Nde
T
be
3.6(b)
N k z N dx d e =

ke
T
but b e is arbitrary
T
be
N q dx + [ N T ]
3.6(c)
Stiffness
matrix
3L
78
22L
R1
F =
R2
R3
Reaction
forces/moments
Eq. (4):
27 13L
2
T
T EI
EI
EI
K = B EIBdx + N ---------Ndx = --------- 3L 4L 3L L + --------------- 22L 8L 13L 6L
4
3
3
27 13L 78 22L
L
2L 3 3L 3 3L 105L
3L L
Load
vector
3 3L
3L 4L
2EI
-------- L
M0
3.10 (15)
13L 6L 22L 8L
8EI
+ ------------- d 4 = M 0
105L
M0 L
105
w ( L ) = d 4 = -------------------------- ----------( 210 + 8 ) EI

3.6(d)
Beam (dof 1 to 4)
Element
stiffness
matrices
Tensile & torsion springs

(only dof 3 and 4)
3EI
EI
k w = --------3- k = -----L
2L
3 3L 3 3L
2
2
T
EI
K beam = B EIBdx = --------- 3L 4L 3L 2L
3
2L 3 3L 3 3L
L
3L 2L 3L 4L
3 3L 3
The total stiffness matrix is obtained by assembly

of the stiffnesses from the beam and the springs:
Force vector:
3L
2
EI
K = --------3- 3L 4L 3L 2L
2L 3 3L 6 3L
2
3L 2L 3L 6L
F = F b + Fs
Reaction
force & moment
R1
1
Q
Q
T
where Fb = ------ N dx = ---- L 3
2L
2
1
L
L 3
L
Fs = R 2
External point
0
force & moment
0
Displacement boundary conditions: w1 = 1 = 0. The reduced equation system becomes
Q
EI- 6 3L w 2
-------= ---- 1
3
2
2 L 3
2L 3L 6L 2
w2
QL
= ------------ 5L
27EI 1
3.7(a)
L
Weighted residual:
v [ ( EIw ) + gA ]dx = 0 (1)
v [ ( EIw ) ]dx = [ v ( EAw ) ]
= [ v ( EAw ) ]
v [ ( EIw ) ]dx =
[ v ( EAw ) ]
L
L
vEIwdx (2)
L
vEIwdx = [ vT ]
[ vM ]
gA vdx
L
3.7(b)
Weight function:
v = Nb e =
L
T
be
B EIB dx d e =

ke
T
be
d--------w
d N
= ---------2- d e = Bd e
2
dx
dx

T T
be N
dN
T
[ N T ] ---------- M
gA N dx
dx L
L
L

fe
T
3.11 (15)
T d N
T T
d-------v------------ = b e B
=
b
e
2
2
dx
dx
but
be
is arbitrary
k e d e = fe

3.7(c)
Element
stiffness
matrices:
Beam (d.o.f 1 to 4)
L
Spring (only d.o.f. 3)
3L 3 3L
EI
K fjder = k = --------32L
2
2
T
EI
K balk = B EIBdx = --------- 3L 4L 3L 2L
3
2L 3 3L 3 3L
L
3L 2L 3L 4L
3 3L
Assembly of total stiffness matrix

(the spring only contributes to d.o.f. 3):
3 3L
2
EI
K = K balk + K fjder = --------3- 3L 4L 3L 2L
2L 3 3L 4 3L
2
Force vector: F = F distributed + Fpoint
3L 2L 3L 4L
Reaktion
force/moment
R1
1
T
where Fdistributed = gA N dx = gAL L 3
1
L
L 3
P = gAL inserted gives:
L
F point = R 2
P
0
External point
force/moment
Displacement boundary conditions: w1 = 1 = 0
R 1 gAL
Reduced equation system (Eq. (3) & (4)):
F = R2 gAL 3
2gAL
EI 4 3L w 2
--------= gAL 2
3
2
L3
2L 3L 4L 2
gAL 3
gAL
= ---------------- 2
EI 4 3
w2
3.8 (a)
L
Weighted residual:
x Q
v ( EIw ) + 1 + --- ------ dx = 0
L 2L
L L
v [ ( EIw ) ]dx = [ v ( EAw ) ]
(1)
L
= [ v ( EAw ) ]
v [ ( EIw ) ]dx =
[ v ( EAw ) ]
vEIwdx
(2)
L
vEIwdx = [ vT ]
[ vM ]
1
x dx
Q v --- 1 + --- -----2
L L
L
2
3.8 (b)
v = Nb e = b e N
Weight function:
Inserted into the weak form gives
T
d-------vT d N
------------ = b Te B
=
b
e
2
2
dx
dx
T
L
T
be
d--------w
d N
= ---------2- d e = Bd e
2
dx
dx
B EIB dx d e =

ke
T
be
L
but b e
L
T
dN
T1
x
dx
[ N T ] ---------- M
Q N --- 1 + --- -----is arbitrary
dx L
2
L L
L
L
k d = f
e e
e
fe
T
3.12 (15)

3.8 (c)
d1
FEM, discretization:
(one element)
d3
2L, EI
d2
d4
3L 3 3L
2
2
T
EI
K = B EIBdx = --------3- 3L 4L 3L 2L
2L 3 3L 3 3L
3L 2L 3L 4L
Displacement boundary conditions: d2 = d3 = d4 = 0 => Reaction forces

Nodal force vector:
F = Fb + F point
P
9
R2
Q
T1
x dx
where F b = Q N --- 1 + --- ------ = ------ 4L Fpoint =
30 21
2
L L
R3
L
6L
R4
External point force
Reaction
force/moment
3
EI3Q
2 PL 1 QL
-------3d 1 = P ------- d 1 = --- --------- --- ---------3
10
3 EI 5 EI
2L
Reduced equation system, Eq. (1):
The deflection of the beam is obtained by the displacement interpolation (approx.) as:
3
2 PL
1 QL 2 3 +
w ( ) = N 1 ( )d 1 = --- --------- + --- ---------- --------------------------- ,
3 EI 5 EI
4
where
x
= --L
3.9(a)
3L 2
Weighted residual:
d
dT
v ------ kA ------ + qA hP ( T T ) dx = 0
dx
dx
0
3L 2
dT 3L 2
d
dT
Integration by parts: v ------ kA ------ dx = v kA ------

dx
dx
dx 0
0
dT
(2) inserted into (1) with Q = kA ------ gives the weak form:
dx
3L 2
3L 2
vhPTdx = [ v ( Q ) ]
v = Nb e =
3L 2
dTdN
----= ------- T e = BT e
dx
dx
3.9 (b) Temperature interpolation: T = NT e

Weight function:
3L 2
dv
------ kA dT
------ dx (2)
dx dx
3L 2
dv
------ kA dT
------ dx +
dx dx
(1)
v ( qA + hpT )dx
T
T T
dv
------ = b e dN
---------- = b e B
dx
dx
T T
be N
Inserted into the weak form gives:

x2
x2
T
be
B kAB dx + N hPN dx T e =
x1
but
x2
T
T
be
x1

ke
T
b e is arbitrary
ke T e = f e
3.13 (15)
x2
N ( qA + hpT ) dx + [ N ( Q ) ]
x1
x1

fe

3.9 (c)
1
e1
FEM - model:
T1
For one element applies:
e2
T2
e3
T3
T4
1
1
L
L
k e = ---------- 1 kA ---------- 1 1 --- d + 1 hP 1 --- d =
L2 1
L2
2
2
hPL
kA
kA
= 12 ------ = ------ 4 1
L
L 1 4
kA
kA
= 0 ; hPL = 12 ------ = 3 ------ T 1
L
L
1
4 1 0 0
kA 1 8 1 0
K = -----and the r.h.s
L 0 1 8 1
0 0 1 4
L
fe = 1 ( qA + hPT ) --- d =
2
Assembly of system matrix gives

Equation system:
4
kA 1
-----L 0
0
1
8
1
0
0
1
8
1
0
0
1
4
4T
T2
kA
= 3 ------T
L
T3
T
Q R1
4 1
1
2 +
0 1 8
0 1
2
0
0 0
1
Q R4
0
1
8
1
kA
F = 3 ------ T
L
Q R1
1
2 +
0
2
0
1
Q R4
4
Q R1 L ( kAT )
0
3
T
T
0
0
2
= 6 +
1 T3 T
0
6
4
Q
3
L
1
R4 ( kAT )
Reaction flux
Reduced Eq. system (2) & (3):

8 1 T2 T = 6 { 1 4 + 0 1 }
1 8 T3 T
6 {0 4 1 1 }
T2
T3
T
= ------ 29 = T 1.3809
21 22
1.0476
3.9 (d)
2
[ N ( Q ) ]1 c =
Convection at x=3L/2 gives for element 3:
Q1
0
Q 2
0
Consider only the contribution from element node 2, since the contribution from
the element node 1 is cancelled by the contribution from element node 2 in element 2,
furthermore use that Q 2 = hA ( T2 T )
0
T
kA 0 0 T 1 kA
where hA = h ( PL 96 ) = kA ( 8L )
[ N ( Q ) ] = -----+ -----2
1
The equation system is modified according to:

4 1 0 0
1 8 1 0
0 1 8 1
4
T2 T
T3 T
1
-0 0 1 4 +
8 T4 T
3
6
6
8L 0 1 T
2
Q R1 L ( kAT )
+
1
3 + --8
0
0
0
8L T
T2
1.3811
T 3 = T 1.0491
1.0119
T4
3.10 (a)
x2
Weighted residual:
d
dT
v ------ kA ------ hP ( T T ) dx = 0
dx
dx
x1
x2
d
dT
dT
Integration by parts: v ------ kA ------ dx = v kA ------
dx
dx
dx
x1
x2
(1)
x2
dv
dT
------ kA ------ dx
dx dx
x1
x
1
dT
----(2) inserted into (1) with Q = kA
gives the weak form:
dx
x2
x2
x2
x2
dT
dv
------ kA ------ dx + vhPTdx = [ v ( Q ) ]x + vhpT dx
dx
1
dx
x1
x1
3.14 (15)
x1
(2)

3.10 (b)
dTdN
----= ------- T e = BT e
dx
dx
Temperature interpolation: T = NT e
T
Weight function:
v = Nb e = b e N
T
T T
dv
------ = b e dN
---------- = b e B
dx
dx
Inserted into the weak form gives:

x2
x2
T
be
x1
x2
T
B kAB dx + N hPN dx T e =
T
be
x1

ke
T
but be is arbitrary
x2
N hpT dx + [ N ( Q ) ]
x1
x1

fe
ke T e = fe
3.10 (c)
1
e1
FEM - model:
For one element applies:
T1
e2
T2
T3
L
L
1
1
k e = ---------- 1 kA ---------- 1 1 --- d + 1 hP 1 --- d =
2
2
L2
L2 1
hPL
kA
2kA
= 16 ------ = ---------- 7 1
L
3L 1 7
hPLT
L
fe = 1 hPT --- d = ----------------- 1
2
4
1
Assembly of the stiffness matrix gives:

System matrix:
7 1 0
2 kA
K = --- ------ 1 14 1
3L
0 1 7
and r.h.s.
hPLT
F = ----------------4
Equation system:
1
hPLT 1
0
F = ----------------- 2 + hPLT
=
4
Q R
----------------1
hPLT
5
2
=
QR
----------------1
hPLT
kA hPL T
------ -------------- ------
L kA L 4
= 16
20
8
kA
= -----T
L
4Q R
----------------4
hPLT
20
30
T1
7 1 0 T1
7 1 0
8
12
kA
2 kA
--- ------ 1 14 1
= -----T 1 14 1 T 2 =
T
T2
L
3L
4Q R
6Q R
4 ----------------6 ----------------0 1 7 T
0 1 7 T 3 = T
hPLT
hPLT
Eq. (1) & (2) gives:
1- 433
7 1 T 1 = 30 ( 0 ) T T 1 = ----T 4.46 T
97 121
1 14 T 2
12 ( 1 )
1.25
T2
3.15 (15)
4. FEM: trusses and beams

4.1 A uniaxial bar is modelled by a linear truss element.
For a certain applied load, the node displacement
shown in the figure results. (a) Show that the strain
developing in the element is ( x ) = 0 and (b) show
that if 0 = 0, the element is subjected to a rigid body
motion equal to 0.
4.2 Derive the four shape functions for

the uniaxial cubic element shown to
the right. Express the shape functions
using the natural coordinate .
x=L
x=0
x
u1 = 0
u 2 = 0 + 0 L
Nod 1
Nod 3
Nod 4
Nod 2
= 1
= 1/3
= 1/3
=1
4.3 The figure to the right shows a uniaxial isoparametric element where a 2nd degree polynomial is used for
Nod 1
Nod 3
the interpolation of the displacement. The coordinates of
the nodes can be seen in the figure, where is a nonL
L
dimensional parameter in the interval: 1 < < 1 .
Assume that the node displacements {u1, u2, u3} are
known and calculate the strain in the element.
Hint: express the strain as a function of the natural coordinate , see below.
Coordinates:
x() =
N k xk
k=1
x1
x3
x2
Nod 2
x
L
N 2 = ( 1 + ) 2
Nk k
N3 = 1
k=1
K x = q0
4.4 The bar in the figure to the right is subjected to an
uniformly distributed axial load Kx = q0 and a point force
P. Analyse the bar by use of the finite element method
with (a) one linear element and (b) two linear element.
Compare the solutions with the exact solution given
below the figure.
N 1 = ( 1 ) 2
Primary variable:
( ) =
x=0
E, A
P
x=L
Exact solution:
2
Px q 0 L x 1 x 2
u ( x ) = ------- + ----------- --- --- ---
EA
E L 2 L
P
( x ) = --- + q 0 L ( 1 x L )
A
4.1 (6)

4.5 Carry out a finite element analysis of the uniaxial bar problem shown in figure (a) below.
Divide the bar in two linear elements of the same length. A linear element with shape functions is shown in figure (b) below.
1
2
L
N
K x = q0 ------3u2
u1
m
E, A
0
1
3
x=0
x = 2L
x = --- L
N 1 = 1 , N 2 = dx = Ld
2
(a)
(b)
4.6 One requirement the displacement

2
interpolation in an element must satisfy,
1 = 2 =
is that it should be able to model an arbi
w1 = L
trary rigid body motion. For a plane 2w2
1
node beam element with 2 degrees of
w2 = + L
w1
freedom at each node, this means that
x
the deflection of the beam must be able
L
0
L
to take the form
w(x ) = + x ,
where and are parameters describing an arbitrary rigid body motion as illustrated in the
figure to the right. Show that the displacement interpolation of the element can satisfy a rigid
body motion as described above.
4.7 The figure to the left
w1 = a b c
2
0
shows an initially straight
1 0
1 = b L + 2c L
w2
beam element that is subR 0 = ----1
jected to a deformation state

0
w2 = a + b c
w1
that results in a constant curvature 0 = 1 R0 , where R0
x 2 = b L 2c L
is the radius of curvature.
L
0
L
Curvature is here defined as
w = (small deformations is assumed). The displacements of the two nodes of the element are shown in the figure. (a) Calculate the curvature 0 and (b) find the slope (angle 0)
and the displacement 0 at the midpoint of the element (x = 0).
4.8 A cantilever beam is loaded by a point force P, a
moment M and a uniformly distributed force per unit
length q ( x ) = Q ( 2L ) (Q is the total resultant
force) according to the figure to the right. The bending stiffness of the beam is EI and its length 2L. Analyse the beam with FEM and use a 2-node beam
element (3rd degree polynomial for the interpolation
of the deflection). Carry out the analysis using (a)
one element and (b) two elements. Compare the
results with the exact solution shown below the figure.
4.2 (6)
q ( x ) [N/m]
P
M
x = 2L
x
Exact solution:
3
x 2
PL
x 3
ML x 2
w ( x ) = --------- 6 --- --- + ----------- ---
L
6EI L
2EI L
3
QL 1 x 2 1 x 3 1 x 4
+ ---------- --- --- --- --- + ------ ---
48 L
6 L
EI 2 L

4.9 The figure below shows a cantilever beam, which is subjected to a distributed force, a
moment and a point force. The bending stiffness of the beam is EI. The beam is during a FEManalysis modelled by two 2-nodes beam element. The coordinates of the three nodes used in
the FEM-model are: x = { 0, 2L, 3L } . Determine the force vector, where also the reaction
forces should be indicated.
Q x
q(x) = ------ --2L
L
z
N--m
P
x
M
x = 2L
x=0
x = 3L
FORMULAS
1D:
Deflection:
N1
---------- ----dx = --- 1 1
dN
L 1 1
dx dx
T
d4
N2 = L( 1 + ) 4
N4 = L( 1 + + ) 4
N 1 = ( 2 3 + ) 4,
N 3 = ( 2 + 3 ) 4,
d N
1d N
B = ---------2- = -----2 ---------2L d
dx
w ( ) = N 1 d 1 + N 2 d 2 + N 3 d 3 + N 4 d 4 = Nd e ,
d3
2L, EI
L
N Ndx = { dx = Ld } = --- 2 1
6 1 2
T
Balkelement:
N2 =
d2
N1 = 1
d1
( ) = N1 1 + N2 2 = N 1 N 2
1
3 3L 3
3L
1 3L 4L 3L 2L
B Bdx = -------2L 3 3L 3 3L
T
3L 2L 3L 4L
1
T
N d = L 3
1
1
L 3
1
78
L
N Ndx = -------105
T
6
1
T
N d = ------ L
15 6
1
L
1
4.3 (6)
22L
27 13L
22L 8L 13L 6L
27 13L 78 22L
2
13L 6L 22L 8L
Solutions
4.1
Displacement in the element: u = N 1 u 1 + N 2 u 2 = 0 + 0 L = { x = L } = 0 + 0 x
du
(a) Strain in the element: ( x ) = ------ = 0
dx
(b) The case 0 = 0 results in the rigid body motion since u ( x ) = 0
4.2 Use Lagrange interpolation:
C = 16
N (1) = 1 C = 16
N ( 1 3) = 1 C = 16 9
N (1 3) = 1 C = 16 9
N 1 = ( 1 ) ( 1 3 ) ( 1 + 3 ) C 1
N 1( 1) = 1
N 2 = ( 1 + ) ( 1 3 ) ( 1 + 3 ) C 2
N 3 = ( 1 + ) ( 1 ) ( 1 3 ) C 3
N 4 = ( 1 + ) ( 1 ) ( 1 + 3 ) C 4
4.3
Strain: du
du d
= ------ = ------ -----dx
d dx
3
dx =
=
N
--------- x d = L ( 1 2 )d
u 2 u1 ( u1 + u 2 2u 3 )
1
---------------+ ------------------------------------- ----------------------2L
L
( 1 2 )
k=1
Note! singular for 2 = 1
4.4
(a) One element solution, discretization:
D1
D2
Uniform load contribution to the nodal force vector:
1
1
ALq
T
f b = N K x ALd = ALq0 1 d = ------------0- 1
2 1
EA
k e = ------- 1 1
L 1 1
Boundary conditions
Reaction force
Eq.
system:
PL- q---------0L
-----D
=
+
Eq.
2
gives:
ALq
=
0
D
2
EA
0
1
EA 2E
------- 1 1
= R + ------------- 1
2 1
L 1 1
D2
P
Eq. 1 then gives: R = P ALq 0
e1
(b) Two element solution, discretization:

EA
k1 = k 2 = ------- 1 1
L 1 1
fb1
e2
D1
D2
D3
Uniform load contribution to the nodal force vector:
1
1
ALq 0
L
AL
T
= f b2 = N K x A --- d = ------- q 0 1 d = ------------- 1
2
2
4 1
Eq.
1 1 0 D1 = 0
R ALq 1 Eq. 2 & 3: D 2
=
syst.: 2EA
----------- 1 2 1
= 0 + ------------0- 2
D2
D
4
L
3
1 Eq. 1 then gives:
0 1 1
P
D3
L
PL- 1 q------------------+ 0 - 3
8E 4
2EA 2
R = P ALq0
Note! The point force solution is exact and independent of the number of element used,
whereas the distributed load solution is approximate. The forces acting at the nodes
are in global equilibrium, i.e. external loads are in balance with internal (reaction) forces.
4.4 (6)

4.5
Element stiff- k = k = EA
------- 1 1
2
ness matrices: 1
L 1 1
e2
e1
D1
D2
D3
1
Element 2:
1
T
fQ =
N K x ALd = ALq 0
12
1 d = ALq
------------0- 1
8 3
Reaction forces
B.C.
R1
1 1 0 D 1 = 0
ALq 0 0
EA
------- 1 2 1
= ------------- 1 + 0
D2
L
8
0 1 1 D 3 = 0
3
R3
Assembly:
3ALq
7ALq
EA
R3 = ------- ( D 2 ) ---------------0- = ---------------0L
8
16
ALq
EA
Reaction forces: R1 = ------- ( D 2 ) = ------------0L
16
2L
( R1 + R 3 ) =
Note!
3L 2
4.6
q0L
D 2 = ---------16E
ALq 0
K x Adx = ------------2
w ( x ) = w ( L ) =
N i di = N 1 ( L ) + N 2 + N 3 ( + L ) + N 4
i=1
= ( N 1 + N 3 ) + ( N 2 + N 4 + L ( N 3 N 1 ) ) = + L = + x
4.7
2
N 1
N 2
N 3
N 4
d Ni
- d i = -------- w 1 + -------- 1 + -------- w 2 + -------- = 2c
----- 0 = w = ---------2
2
2
2
2 2
2
dx
L
L
L
L
L
i=1
0 = w ( x = 0 ) =
N i ( = 0 )d i = ( N 1 w 1 + N 2 1 + N 3 w 2 + N 4 2 )
i=1
4
0 = w ( x = 0 ) =
=0
N 1
N 2
N 3
N 4
dN i ( = 0 )
--------------------------- d i = ------- w 1 + ------- 1 + ------- w 2 + -------
dx
L
L
L
L 2
i=1
= a
=0
b
= --L
4.8
D1
D3
(a) Discretization, one element:

D2
L
ke =
D4
3 3L 3 3L
2
2
T
EI
B EIBdx = --------3- 3L 4L 3L 2L
2L 3 3L 3 3L
2
3L 2L 3L 4L
4.5 (6)
Distributed load contribution

to the nodal force vector:
3
1
TQ
Q
fb = N ------ Ld = ---- L
2L
6 3
1
L

4.8 cont.
3 3L 3 3L
Eq.
system:
EI- 3L 4L 3L 2L
-------3
2L 3 3L 3 3L
2
3L 2L 3L 4L
Eq. 3 & 4 give:
D1 = 0
D2 = 0
R
MR
3
Q
+ ---- L
6 3
P
L
M
D3
D4
PL
ML
QL
= --------- 8 3 + ----------- 2 + ---------- 3
EI 2 L
EI 2 L
3EI 1 L
D3
D4
Eq. 1 & 2 then give the reaction forces: R = P Q

D1
(b) Discretization, two elements:
D3
D2
k1 = k2 =
L 2
Eq.
system:
3L 2
D5
D4
D6
Distributed load contribution
to the nodal force vector:

L2
M R = 2PL M QL
3L
2
4EI 3L 2 L
3L 2 L 2
B EIBdx = -------3
3 3L 2
3
3L 2
L
T
3L 2
3L 2 L 2 3L 2
3L
6
TQL
Q
= N ------ --- d = ------ L
2L 2
24 6
1
L
1
f b1 = fb2
D1 = 0
R
D2 = 0
3L 2 L
3L 2 L 2
0
0
MR
D3
6
0
3
3L 2
Q
4EI 3 3L 2
--------= 0 +-----3
2
2
2
24
D
0
L 3L 2 L 2
0
2L 3L 2 L 2
4
P
0
0
3 3L 2
3
3L 2
D5
2
2
M
D
0
0
3L 2 L 2 3L 2 L
2
6
L
12
0
6
L
Eq. 3 - 6 give:
D3
D4
D5
D6
Eq. 1 & 2 then give the reaction

forces:
R = PQ
5
1
17
2
3
ML
QL
PL
9
L
2
L
28
L
= --------+ ----------+ -----------6EI 16
2EI 4
48EI 48
12 L
4L
32 L
3
M R = 2PL M QL
Note! The solutions for P and M are exact independent of the number of beam elements
used, whereas the distributed load solution is approximate. Also note that the forces
acting at the nodes are in global equilibrium, i.e. external loads are in balance
with internal (reaction) forces.
4.9
D1
e1
D2
D3
e2
D4
Element 1: x = L ( 1 + )
1
f b1 =
D5
D6
B.C. & kinematical constraint: D1 = D2 = D3 = 0

(give reaction forces/moments: R1, R2 & R3)
dx = Ld ,
(------------------------------2 3 + )- ----Q3Q
( 1 + )Ld = ------4
2L
10
fb =
N qLd
2QL
f b2 = ----------15
where
7Q
f b3 = ------10
1
T
= R + 3Q
- R2 + 2QL
----------- R3 + 7Q
------- M QL
-------1 -----10
15
10
5
4.6 (6)
P 0
Q
q = ------ ( 1 + )
2L
QL
f b4 = -------5
5. FEM: planar frames of trusses and beams

5.1 The figure to the right shows a structure with three linear
elastic truss elements with elastic modulus E. Element 1 and 3
have cross sectional area A and length L. Element 2 has cross
sectional area 2A and length 2L . The structure is subjected to a point force Q and a force per unit volume (body
force) K x = Q ( AL ) according to the figure. Calculate the
displacement at node 2, all reaction forces and the normal
stress in each one of the truss elements.
y/L
(0,1)
(1,1)
2
Kx
1
(1,0)
5.2 A truss structure containing three trusses, all with elastic

modulus E and cross sectional area A, is shown to the right. The
structure is loaded by one point force Q/2 and a body force of
total magnitude Q acting on the vertical truss member downwards. Model the structure by use of three linear elements and
calculate the displacements and possible reaction forces at the
nodes. Note, the displacements at the nodes will in the current
case agree with the exact solution. Will the numerical solution
deviate from the exact one? If so, how?
L
Q
2L
Q2
5.3 Analyse the cantilever beam to the right by use of FEM.

Use a 2-node element, which allows for both axial deformation and development of curvature (bending). The elastic modulus of the beam is E and the cross section is shown in the
figure. Note that with the load applied in the present case, the
FEM solution will agree with the exact solution. Especially,
evaluate the solution for the case h L = 1 10 .
5.4 A force per unit length q(x) is applied on a beam

with elastic modulus E and a cross sectional area A
and a moment of inertia I. The left end of the beam
is clamped and the right end rests on an elastic sup- x = 2L
port, here modelled by a vertical spring with spring
x/L
45o
h
h
2P
q(x)
P
x
L
constant k = EI L .
x=2L
(a) Carry out a finite element analysis, where
the beam is modelled by one two-node element, and evaluate the deflection of the
beam. Here: = 3 2 and q ( x ) = Q ( 2L ) ( x L ) .
(b) Divide the beam into two element of equal to length and redo the analysis.
Note that the deflection at the nodes will in the current case always coincide with the exact
solution. The deflection between the nodes for 0 x 2L will deviate somewhat from the
exact solution due to distributed load q(x).
5.1 (6)

5.5 Analyse the linear elastic planar frame work shown to the
right by use of FEM. Use a 2-node combined truss/beam element allowing for both axial and bending deformation. The
cross section of the frame is displayed in the figure and the elastic modulus is E. Will the FEM solution agree with the exact
solution, i.e. with the Euler-Bernoulli beam theory, in the present
case?
h
h
5.6 The figure below shows a circular ring, which is an integral part of a flexible machine
member. The ring is subjected to diametrically opposed forces according to the figure. Determine the spring constant defined as k = P by use of FEM. If the symmetry of the problem
is fully utilized, only a quarter of the ring needs to be modelled. The problem can for instance
be analysed by the Matlab based FEM program frame2D, available at the home page of the
course. If the displacement, , primarily is due to bending deformation (a good approximation
if R h ), the spring constant of the ring can analytically be expressed as
4 EI
- ------ ,
k = -----------------2
3
( 8)R
where E is the elastic modulus and I area moment of inertia. Note that in order for the FEM
solution to come close to this result, the FEM model requires that R h .
2
R
R
symmetric quarter
P2
2
R
x
5.2 (6)
FORMULAS
Frames of truss/spring members (based on 2-node elements):
Global forulation for truss & spring elements
D2
D3
a = c sc
2
sc s
where
D1
c = cos
EA
spring constant for a truss element k = ------L
Ke = k a a
a a
D4
s = sin
alternativly a =
Local/global transformation of nodal force vector:

T
dr
Fe = T fe
T =
l 12 m 12 0
l12
l 12 m 12
l12 m 12
m 12
l 12 = cos x = ( x 2 x 1 ) L
m 12 = cos y = ( y 2 y 1 ) L
L =
( x2 x1 ) + ( y2 y1 )
0 l 12 m 12
Frames of beam members (based on 2-node elements):

T
d e = u 1 w 1 1y u 2 w 2 2y
Local formulation
(local coordinate system)
z
fe = f 1x f 1z M 1y f 2x f 2z M 2y
w2, f2z
EA
------2L
u2, f2x
w1, f1z
2y, M2y
0
u1, f1x
ke =
1y, M1y
2L
EA
------2L
D2, F2
D5, F5
D6, F6
D3, F3
D1, F1
{X1, Z1}
Transformation matrix:
T =
T2 0
0 T2
l x mx 0
dr
T2 =
lz m z 0
0 0 1
3EI
--------3
2L
3EI
---------
3EI
--------2
2L
2EI-------L
2L
Global formulation
(global coordinate system)
z
3EI 3EI
--------3- --------22L 2L
3EI EI
--------2- -----L
2L
EA
------2L
0
0
3EI 3EI
--------3- --------22L 2L
3EI
EI
--------2- -----L
2L
EA
------2L
0
0
3EI
--------3
2L
3EI
--------22L
3EI
--------22L
2EI
--------L
x
T
D4, F4
De = D1 D2 D3 D4 D5 D6
{X2, Z2}
F e = F1 F2 F3 F4 F5 F 6
Transformation scheme:
d e = TD e
fe = k e TD e
T
fe = ke de
F e = T ke T D e = K e D e
F e = T fe
X2 X1
l x = cos xX = ------------------ = cos
2a
Z2 Z1
l z = cos zX = ----------------- = sin
2a
Z2 Z1
m x = cos xZ = ----------------- = sin
2a
X2 X1
m z = cos zZ = ------------------ = cos
2a
5.3 (6)
Direction
cosines
2 2
lx + lz = 1
m2 + m 2 = 1
z
x
Solutions
5.1
D6
D4
D5

D3
1 0
EA 0 0
k 1 = ------L 1 0
0 0
D2
D8
1
1
0
1
0
0
1
0 k = EA
------- 1
2
2L 1
0
0
1
1
1
1
1
1
1
1
1
1
1
1
1
0 0
EA 0 1
k 3 = ------L 0 0
0 1
0
0
0
0
0
1
0
1
D1
D7
Q
f b1 = NK x ALdx = ---- 1
2 1
0
1
T
Q
= T f b1 = ---- 0
with T = 1 0 0 0
2 1
00 1 0
0
Distributed load contribution to nodal force vector (el. 1):

Transformation to global coordinate system gives:
Eq.
system:
EA
------2L
2 0
0 0
0
1 1
1 1
0
0
2 0 1 1 0
0 0 1 1 0
0
0
0
2
0
2
2
0
1
1
0
0
3
1
0
0
1
1
0
2
1
3
F b1
D1
R1 Q 2
D2
R2
D3
Give rise to the reaction forces R1 t.o.m. R6
R3
Eq. 7 and 8 give:
D4
R4
D7
D8
D5
R5
D6
R6
D7
Q 2
Q
D8
Bound. cond.: D1=D2=D3=D4=D5=D6=0

QL
= ----------- 1
8EA 5
-------------, R 2 = 0, R 3 = 3Q
Eq. 1 - 6 then give: R 1 = 5Q
8
5Q
3Q
R 4 = -------, R 5 = 0, R 6 = ------8
8
The normal stress in one element: = EBde = EBTDe

D7
D7
D8
Q
= ------ 1 = E --1- --1- 1 0 0 0
8A
L L 0 0 1 0 D1
D2
5.2
D6
D4
D5
D3
D2
D1
1 1
------- ------- 0 0
1
1
2 2
2 = E ---------- ---------1 1
2L 2L
0 0 ------- ------2 2
D1
D6
D7
D8
D3
D8
5Q
3 = E --1- --1- 0 1 0 0
= ------8A
L L 0 0 0 1 D5
D6
3Q
= ------8A
D4
Displacement B.C.: D2 = D3 = D4 = D5 = 0
1
1
1
1 + --------- ---------- 1 0 ---------
2 2 2 2
2 2
Eq.
system:
EA
------L
QL
QL
= ----------- 3 2 = -------- 0.793
EA -0.207
2EA
1 2
1
---------2 2
1
0
1
---------2 2
1
---------2 2
1
---------2 2
0
0
1
---------2 2
1
---------2 2
1
0 0 ---------2 2
1 0
0
0 1
0
1
0 0 ---------2 2
1
0 1 ---------2 2
1
---------2 2
Q
D1
1
---------R2
2 2
0
0
0 = Q 2 + R3
1
0
R4
1
0
R5
---------D6
2 2
0
1
1 + --------
2 2
R 2 R 3 R 4 R 5 = Q -0.207 -1.293 0.207 -0.207
The numerical solution results in a linear displacement variation in vertical element,

whereas the exact solution yields a quadratic displacement variation, due to the
distributed load.
5.4 (6)

5.3
y
D2
E, I, A, 2L
Equivalent
problem:
h
2
I = ------, A = h
12
Discretization:
D4
D1
D3
D5
x
D6
Eq.
system:
0 0 0
0 3 3L 0 3
0
3L
AL
L 2
= --------- = 12 ---
I
h
Here
D1 = 0
Rx
D2 = 0
Rz
Eq. 4 - 6 give: D
4
2PL ( EA )
2
3
PL
= P ( 2L ) 3 3
D 5 = --------- 8 3
EI
2
8 ( 4L )
D6
P ( 2L ) 2
D3 = 0
EI
--------- 0 3L 4L 0 3 L 2L
= My
3
0
0
0
0
D4
2L
P
0 3 3L 0 3 3 L
D5
P
2
2
0 3L 2L 0 3 L 4L
0
D6
Eq. 1 - 3 then give the reaction forces:

R x = P,
R y = P,
M y = P2L
5.4
D2
One element solution:

Eqn.
system:
D3
EI, EA
Q x
q ( x ) = ------ --2L L
D1
D5
D4
D6
D8
EA
------2a
3EI-------3
2a
3EI
--------2
2a
3EI
--------2
2a
2EI
--------a
3EI
--------3
2a
3EI
---------
0
0
2a
0
0
3EI
--------2
2a
2EI
--------a
0
0
K = EA
------2a
0
EA
------2a
0
0
EA
------2a
0
0
0
0
0
0
0
3EI
--------32a
3EI
--------22a
0
D7
0 0
R1
3EI-------0 0
2
2a
2EI
--------- 0 0
a
0
0
0
0
D4
0 0
D5
3EI
3EI
--------- + k --------- 0 k D 6
3
2
2a
2a
0
3EI
2EI
--------- --------- 0 0 0
2
a
2a
0
k
0
0
3EI
where a = 2L och k = -------3
Q
R 2 -----10
7QL
R 3 ----------60
P
=
9Q
------10
23QL
-------------60
0
R8
0 0
0 k
2L
D4
D5
D6
225
L P4L 1
QL
= ---------- --------------+ ---------- 0
45
EA
EI
0
1--5
2
5.5
y
D2
D1
D3
el. 1
D5
D4
D6
Discretization:
Z
Element stiffness
matrices (local):
el. 2
y
x
2
2
A C
EI
k e 1 = k e2 = --------3- 0 3L 4L 0 3 L 2L =
T
0
0
0
0
2L
C B
0 3 3 L 0 3 3 L
0 3L 2L
D8
D9
Element stiffness
matrix (globally):
0 3 L 4L
here
D7
0 0 0
0
3 3L 0 3 3L
K e 1 = k e1
K e 2 = T k e1 T =
Transformation matrix:
T2 0
A C T2 0
T
T
0 T2 C B
0 1 0
T2 = 1 0 0
0 0 1
5.5 (6)
AL
L 2
= --------- = 12 ---
h
I
0 T2
T 2 AT 2 T 2 CT 2
T
T 2 C T 2 T 2 BT 2

5.5 cont.
Considering the displacement boundary conditions (D1=D2=D3=0)
the reduced equation system becomes:
3 + 0 3L 3 0 3L
0 3 + 3L 0 0
EI-------3
2L
3L
3
0
3L
D4
D5
D6
3L 8L 3L 0 2L
0 3L 3 0 3L D 7
0
0 0 D
8
0
2L 3L 0 4L
D9
0
0
= 0
P
0
0
D4
( h L ) 24
1
1L
D5
D6
D7
4PL
= ------------EI
D8
D9
8 3 + ( h L ) 24
1
3 ( 2L )
The reaction forces are given by Eqs. 1 - 3 as R1 = P, R2 = 0, R3 = 2PL

The FEM solution coincides with the exact solution, since the frame only is subjected
to a point force.
5.6
y
Ne = 4
The FEM analysis is carried out by use of a 2

node combined truss/beam element. To facilitate a
comparison with the analytical solution, the
2
geometry must be chosen such that ( h R ) 1 .
Ne = 1
Since, h I A , the cross section of the ring has
Ne = 2
been choose such that R I A .
Ne = number of elements
k FEM k analytic
------------------------------------k analytic
k FEM k analytic
1.3
1.2
1.1
1.0
0
12
16
Number of elements
1
0.1
1
2
0.01
0.001
10
100
Number of elements
5.6 (6)
6. FEM: 2D/3D solids

6.1 Derive the element stiffness matrix
for the CST element shown in the figure
to the right. Assume that the material is
isotropic, linear elastic with elastic modulus E and Poissons number = 0.
N1 = 1
y/L
1
N 2 = c
3
1
N3 =
= x L, = y L
x/L
1
6.2 One way to satisfy compatibility across element

boundaries between regions of high to low order elements
is to use transition elements. A plane triangular transition
element is shown in Figure 4. The shape functions for the
vertex nodes of the element are displayed below the element. Determine the shape function associated with node 4
and show that it fulfil standard requirements put on shape
functions.
1.0
2
1.0
N1 = 1 ( 3 2 ( + ) )
N2 = ( 2 ( + ) 1 )
N3 =
6.3 A plate containing a circular hole with radius R is modelled by use of plane 8-noded bi-quadratic isoparametric
elements. The figure to the right shows one such element
located at the hole of the plate. The element is symmetrically located with respect to the y-axis and extends one
quarter of the circumference of the hole, i.e. the straight element sides: 2-6-3 and 1-8-4, respectively, form 45o angles
with respect to the y-axis. The nodes 1, 5 and 2 are placed at
the border of the hole. Determine the distance from the centre of the hole to the point { x 0, y 0 } , defined by the natural
y
7
4
3
5
8
1
6
2
coordinates = 0 = 1 2, = 0 = 1 , i.e. calcu2
late ( x 0 + y 0 ) . How much does this point deviate from the geometric boundary (the radius)
of the circular hole?
The shape functions of the element are:
1
N 1 = --- ( 1 ) ( 1 ) ( 1 + + ) , N 2 =
4
1
N 3 = --- ( 1 + ) ( 1 + ) ( 1 )
4
1
2
N 5 = --- ( 1 ) ( 1 )
2
1
--- ( 1 + ) ( 1 ) ( 1 + )
4
1
N 4 = --- ( 1 ) ( 1 + ) ( 1 + )
4
1
2
N 6 = --- ( 1 + ) ( 1 )
2
1
2
N 7 = --- ( 1 ) ( 1 + )
2
6.1 (24)
1
2
N 8 = --- ( 1 ) ( 1 )
2

6.4 The Figure to the right shows a plane bi-linear isoparametric
4-node element. The order of node numbering is opposite the
one used in the natural coordinate system (, ). As a consequence, the determinant of the Jacobi matrix becomes negative.
The element stiffness matrix, which is calculated by area integration in the natural coordinate system, will then also become
negative. A FEM-analysis with such an element will crash.
Calculate the determinant of the Jacobi matrix for the element
with the erroneous node numbering to the right.
6.5 The figure to the right shows a three dimensional
20 node element of serendipity type. Assume that the
shape functions associated with the nodes 1 to 19 (N1,
..., N19) are known functions of the natural coordinates: , , . The element is shaped as a cube in the
natural coordinate system, defined in the interval 1 to
1 for each coordinate. Determine the shape function
associated with node 20, i.e. N20(,,).
6.6 Figure (a) to the right shows a plane
(2D) isoparametric element with its shape
functions given in figure (b) to the right.
J = x y
x y
B = B1 B2 B3 B4 .
14
10
13
16
20
7
17
18
2
8
1
N1 = ( 1 ) ( 1 ) 4
N2 = ( 1 + )(1 ) 4
1
1
a + 3l
N3 = ( 1 + ) ( 1 + ) 4
N4 = ( 1 ) ( 1 + ) 4
6.7 Model the plate (thickness h) shown to the right by use of a

4-noded bi-linear isoparametric element. Determine
(a) the coordinate transformation x ( , ) and y ( , ) ,
y/L
6.8 The plate in the above problem is loaded by a pressure p0

(uniform traction) acting on the side 1-4 and by its dead weight
(density ). The dead weight can be modelled as a force per unit
volume (body force) K y = g . Let = 1 2 and determine p
0
the contributions to the nodal force vector from
(a) the pressure p0 and (b) the force per unit volume Ky.
(b) the Jacobi matrix J and its determinant J ,

(c) the strain vector (assume that the displacement vector de
is known),
(d) an expression for the element stiffness matrix ke.
6.2 (24)
x
L
12
1
1
a
(b)
19
(b) Determine the sub-matrix B1 in the

B-matrix of the element
15
(a) Determine the Jacobi matrix of the

element, i.e
11
b+l
(a)
1 x/L
2
1
1

6.9 Calculate the contribution from the force per unit volume Ky to the nodal force vector in
the above problem by use of numerical integration based on Gauss-Legendre quadrature. Use:
(a) 1 1 and (b) 2 2 point integration scheme in the element.
6.10 A traction vector t (force per unit surface) is acting between points A and B located on
the edge of a plate of thickness h. The segment between A and B is straight and of length 2L.
Consider a linear variation of the traction vector according to
1
s
1
s
t = --- 1 --- tA + --- 1 + --- tB ,
2
L
2
L
where s is a natural coordinate, tA and tB are the traction vectors at the points A and B, respectively, see the figure below. Determine the contribution to the total nodal force vector if the
plate is modelled by
(a) one isoparametric 4-node quadrilateral element,
(b) one isoparametric 8-node quadrilateral element, where the mid nodes are placed in the
middle between their corresponding corner nodes.
Assume that the traction vectors along AB are composed of a constant normal stress 0 and a
constant shear stress 0, such that
tA = t B =
tx
0 cos 0 sin
0 sin + 0 cos
ty
Use the results in (a) and (b) to evaluate the contribution to the total nodal force vector along
AB if the boundary is modelled by
(c) three equal isoparametric 4-node quadrilateral elements and
(d) three equal isoparametric 8-node quadrilateral elements, see the figure below.
tb
(a)
(b)
(c)
B 7
3
0
y
B 4
(d)
ta
A
7
2
1
6.11 A bi-linear rectangular element is

loaded by its dead weight (gives rise to a
body force), see the figure to the right.
Determine the nodal force vector.
Assume that the acceleration of gravity,
g, is known.
A
1
y
b
thickness h
x
a
6.3 (24)
6
5
4
3
2
1
Body force
Kx = 0
Ky = g

6.12 A triangular 2D domain is modelled by one plane triangular CST element according to
the figure below. The bottom side of the triangle is rigidly supported and the left side (x = 0) is
T
subjected to a linearly varying pressure, p(y), described by the traction vector t = [ p ( y ), 0 ] .
The material is isotropic, linear elastic with elastic modulus E and Poissons ratio = 1 3 .
Determine the node displacements and the reaction forces. The shape functions of the element
and the element stiffness matrix are given in the figure below.
y/L
(0,1)
p( y) =
y
p 0 1 ---
Shape functions:
N 1 = 1 ,
dr = x L
N 2 = ,
(1,0)
x/L
N3 =
= yL
4 2 3 1
2 4 1 1
3Eh
K e = ---------- 3 1 3 0
16 1 1 0 1
1 1 0 1
1 3 1 0
Node displacement vector:
T
y/L
1
3
2
(a) Show that the shape functions of the element is:

N 1 = 1 ; N 2 = 2 ; N 3 = , where
= x L and = y L .
(b) Calculate the stresses in the element. The displacement vector of the element, d e , is given in the figure,
where 0 is a reference strain.
Load cases:
1
3
1
0
0
3
d e = d 1x d1y d 2x d 2y d 3x d 3y
6.13 A CST element for 2D linear elastic analysis is shown

to the right. The material is isotropic, linear elastic with
elastic modulus E and Poissons ratio = 1/3.
6.14 Consider a thin quadratic sheet metal of size

l l and thickness h of a linear elastic material (E,
). Model the sheet metal by use of two linear triangular elements (CST-element) and carry out
FEM analyses for the three different load cases
(a), (b) and (c). Introduce appropriate displacement boundary conditions, where symmetry conditions can be utilized, and determine node
displacements and stresses in the elements. For
simplicity, let Poissons ratio be = 0 .
1
1
0
1
1
0
x/L
1/2
T
1- -----1- L
d e = 0 0 1--- -------0
0
4 108
6
y/l
4
3
e2
e1
1
(a)
(b)
x/l
2
(c)
0
0 0
(a) uniaxial tension,
(b) pure shear,
(c) dead weight, where is the density and g acceleration of gravity.
6.4 (24)

6.15 A rectangular sheet metal of thickness h is subjected to an uniaxial load corresponding to
a normal stress 0. An exact analysis can for this case be carried out using a plane CST element according to the FEM model shown in the figure below. Here, the uniaxial load is
applied as a traction vector, t, acting on the element side (edge) between node 2 and node 3.
The material is isotropic, linearly elastic with elastic modulus E and Poissons ratio = 1 3 .
The shape functions of the element and the element stiffness matrix Ke (plane stress) are given
in the figure.
(a) Calculate/evaluate the nodal force vector F, where also the reaction forces should be
marked. Use the coordinate s (see the figure) when calculating the consistent nodal
forces. Note that s = 0 at node 2 and s = 2L at node 3, which give the relations
x = L s 2 and y = s 2 .
(b) Calculate the node displacements, de, and the reaction forces.
(c) Calculate the strains in the element and show that they agree with the exact solution, i.e.
x = 0 E , y = 0 E and xy = 0 .
0
y
1 0
t = ------2 0
L
L
s
1
Shape functions:
x y
x
N 1 = 1 --- ---, N 2 = ---,
L L
L
4 2 3 1
2 4 1 1
3Eh
K e = ---------- 3 1 3 0
16 1 1 0 1
1 1 0 1
1 3 1 0
Node displacement vector:
y
N 3 = --L
1
1
0
1
1
0
1
3
1
0
0
3
de = d 1x d 1y d2x d 2y d 3x d 3y
6.16 A rectangular sheet metal of thickness h is subjected to an uniaxial stress 0. An exact

analysis of the problem can for instance be carried out by use of only one plane bi-linear 4node quadrilateral element as in the FEM model shown in the figure below. The position of the
element nodes are x L = 1 and y L = 1 . The uniaxial load is introduced in the model as
a traction vector applied on the element side (edge) between node 2 and node 3. The material
is isotropic, linearly elastic with elastic modulus E and Poissons ratio = 1 3 .
T
(a) Introduce the displacement vector D e = d 1x d 1y d 4y and define the displacement boundary conditions.
(b) Calculate/evaluate the nodal force vector Fe. Mark the reaction forces according to f1x = R1x etc. (the shape functions of the element, isoparametric formulation, are given in the figure below).
(c) Calculate all the reaction forces and also check that the nodal forces due to the
traction t agrees with the answer in (b) above. The element stiffness matrix
Ke and the resulting node displacement vector De are given below.
6.5 (24)
y/L
2L
t = 0 1
0
x/L
2L
8
3
5
Eh
K e = ------- 0
16 4
3
1
0
Shape functions:
N1 = ( 1 ) ( 1 ) 4
3
2
N2 = ( 1 + ) ( 1 ) 4
N3 = ( 1 + ) ( 1 + ) 4
N4 = ( 1 ) ( 1 + ) 4
6.17 Figure (a) to the right shows a plane

rectangular plate of thickness h, subjected
to a pure bending moment. Utilizing the
symmetry in the problem, only half of the
plate needs to be modelled. A crude FEM
model consisting only of one 4-node (bi-linear) quadrilateral element is shown in figure
(b). The coordinates of the nodes are evident from the figure. The moment is
replaced by an equivalent traction vector, t,
prescribed on the boundary x = 2L, as
shown in the figure. The material is isotropic, linear elastic with elastic modulus E
and Poissons ratio . Here plane stress conditions are assumed to prevail.
(a) Calculate/evaluate the nodal force
vector F, where also the reaction
forces should be indicated.
(b) Calculate the normal stress in the xdirection as a function of position.
Make use of the node displacement
solution is given in figure (b).
3
8
0
1
3
4
0
5
5
0
8
3
1
0
4
3
0
1
3
8
0
5
3
4
4
3
1
0
8
3
5
0
2L
(a)
3
4
0
5
3
8
0
1
1
0
4
3
5
0
8
3
0
5
3
4
0
1
3
8
0
2 3
2
0 L 2 3
D e = --------E
2
0
0
0
2L
2L
M
(b)
y/L
4
(2, 1) 3
t = 0 y L
0
x/L
(2,1) 2
D = D 1x D 1y D 2x D 2y D 3x D 3y D 4x D 4y
4
= --- -----0- L 0 0 1 1 1 1 0 0
3E
6.18 A quadratic plate with edges of length 2L and of thickness h is loaded by its dead
weight and rotates around its diagonal with a constant angular velocity , see figure (a) below.
The lower corner of the plate is mounted on a bearing. The plate has density and the material
is isotropic, linear elastic with elastic modulus E and Poissons ratio . Assume further that
plane stress conditions prevail and that = 0 . A rather coarse finite element model that to
some extent utilizes the symmetry of the problem is shown in figure (b) below. The model
consists of only one triangular element with a linear interpolation for the displacements (CST
6.6 (24)

element). The inertia forces due to the angular velocity and the dead weight can be
2
addressed by introducing the body forces K x = x and K y = g into the FEM analysis.
The shape functions Ni and the element stiffness matrix ke are given in the figure below.
(a) Calculate/evaluate the nodal force vector F, where also the reaction forces should be
included symbolically.
(b) Calculate all the displacements at the nodes de and the reaction forces.
(c) Calculate the stresses in the element.
y/L
(a)
(b)
(0, 1)
3
2L
x/L
(1, 0)
2L
g
Shape functions:
1
x y
N 1 = --- 1 --- --- ,
2
L L
(0, 1)

x
N 2 = --L
1
x y
N 3 = --- 1 --- + ---
2
L L
Nodal displacement vector:
T
d e = d 1x d 1y d 2x d 2y d 3x d 3y
3
1
Eh
ke = ------- 4
8 2
1
1
1
3
0
2
1
1
4
0
8
0
4
0
2
2
0
4
2
2
1
1
4
2
3
1
1
1
0
2
1
3
6.19 A rectangular plate of thickness h rotates with a constant angular velocity , see Figure
(a) below. The material of the plate is isotropic linear elastic with elasticity modulus E, Poissons ratio and has density . Plane stress is assumed to prevail in the plate and that = 0. If
the symmetry of the problem is considered and utilized, the plate can be modelled by only two
bi-linear isoparametric elements according to Figure (b) below. The node coordinates is evident from the figure. The inertia forces due to the angular velocity can be addressed by
2
introducing the body forces K x = x and K y = 0 into the FEM analysis.
(a) Determine the contribution from the body force to the node force vector in element 2.
Hint: the coordinate transformation x = ( 3 + )L , y = L is useful in element 2.
(b) The resultant node displacements, DT, corresponding to the current load is given in the
figure below. Calculate the stresses at the three points: {x = 0, y = 0}, {x = L, y = 0}, {x
= 2L, y = 0}. The exact solution for the normal stress in the x-direction can be
2 2
2
expressed as xx ( x ) = ( L 2 ) ( 16 ( x L ) ) . How much does the FEM solution
deviates from the exact solution at the three points? At which point do the solutions
deviate the least?
Hint: the Jacobi matrix of the coordinate transformation is J = L I , where I is a unit
6.7 (24)

matrix of dimension 2.
y/L
(a)
(b)
4 (2, 1)
3 (4, 1)
(0, 1)
, E, = 0
Elem.1
Elem. 2
x/L
4L
(0, -1)
D = D 1x D 1y D 2x D 6x D 6y
4L
2 3
4 L
= --- ---------------- 11 0 16 0 16 0 11 0 0 0 0 0
3 E
6.20 A plane solid of thickness h is subjected to a liny

x
early varying pressure, p(x), acting on a segment of
p ( x ) = p 0 --L
the boundary as shown in the figure to the right. The y = 2L
3
6
material is isotropic, linear elastic with elastic modue4
lus E and Poissons ratio . The plane solid is mode3
e7
2
5
8
elled by triangular CST elements. The mesh is y = L
indicated in the figure, where also the displacement
e6
e2
vector D and the nodal force vector F are shown.
e1
e5
1
4
7 x
y=0
(a) Define the displacement boundary conditions.
(b) Calculate/evaluate the nodal force vector. Mark
x=0
x=L
x = 2L
the presence of possible reaction forces as F1x
T
D = D 1x D 1y D 2x D 2y
D 8x D 8y
= R1x and so on.
T
(c) Calculate the stresses in element e1. Assume
F = F 1x F 1y F 2x F 2y
F 8x F 8y
plane stress prevail and that the displacement
vector D is known, where the zero displacement boundary conditions may be enforced.
6.21 A cantilever beam is modelled by 20 plane 4-node quadrilateral elements arranged
according to the figure below, where the global numbering of nodes and elements are indicated. All elements are rectangular, of equal size and of thickness h. The beam is made of a
linear elastic, isotropic material with elastic modulus E and Poissons ratio = 0. Plane stress
conditions is assumed to be valid. The left end of the beam is welded on a wall, which in the
present model is assumed to be rigid. The beam is subjected to a shearing load acting on its
right end as illustrated by the figure. The components of the displacement vector and the force
vector is also indicated in the figure.
(a) Define the displacement boundary conditions.
(b) Calculate the external load contributions to the node force vector, i.e. the contributions
from the traction vector.
(c) Calculate the stresses at the centroid of element e1. Assume that the displacement vec-
6.8 (24)

tor D is known.
y
2a
e1
e5
11
e9
16
e13
21 e17 26
e2
e6
12 e10
17
e14
22 e18 27
e3
e7
13 e11
18
e15
23 e19 28
e4
e8
14 e12
19
e16
24 e20 29
15
20
10
D = D 1x D 1y D 2x D 2y
D 30x D 30y
25
30
F = F 1x F 1y F 2x F 2y
6.9 (24)
2b
F 30x F 30y
FORMULAS
Plane (2D) triangular linear element:
d 1y
d3y
u ( x, y ) = N 1 0 N 2 0 N 3 0 d = Nd
e
e
v ( x, y )
0 N1 0 N2 0 N3
Displacements:
d3x
Ae
d 2x
de =
d 2y
d 3x
d2y
d1y
d 1x
1
N 1 = --------- [ ( y 2 y 3 ) ( x x 2 ) + ( x 3 x 2 ) ( y y 2 ) ]
2A e
d2x
d 3y
1
N 2 = --------- [ ( y 3 y 1 ) ( x x 3 ) + ( x 1 x 3 ) ( y y 3 ) ]
2A e
d1x
x
1
N 3 = --------- [ ( y 1 y 2 ) ( x x 1 ) + ( x 2 x 1 ) ( y y 1 ) ]
2A e
Strains:
xx
yy = B de
B = B1 B2 B3
Bi =
x y
Stresses:
N i x
N i y
N i y N i x
xx
yy
xy
xx
1
0
E
= ----------------- yy
0
2 1
(1 )
0 0 ( 1 ) 2 xy
(Plane stress)
Plane (2D) quadrilateral bi-linear element:

4
d3y
y d
4y
d1x
y =
Deformation:
N 2 = ( 1 + ) ( 1 ) 4
N 3 = ( 1 + ) ( 1 + ) 4,
N4 = ( 1 ) ( 1 + ) 4
u ( , ) = N 1 0 N 2 0 N 3 0 N 4 0 d = Nd
e
e
v ( , )
0 N1 0 N2 0 N 3 0 N 4
N i x
N i y
Bi =
N i y N i x
where
Stresses:
N i x
N i y
= J
N i
N i
J =
x y
x y
xx
xx
1
0
E
- 1
yy = ----------------- yy
0
2
(1 )
0 0 ( 1 ) 2 xy
xy
6.10 (24)
Ni yi
N 1 = ( 1 ) ( 1 ) 4,
B = B1 B 2 B 3 B 4
i i
d2y
d2x
x
Displacements:
N x
i
4
d3x
d4x
d1y
x =
(Plane stress)
Solutions
6.1
1 0 1 0 0 0
1 0 0
1
B = --- 0 1 0 0 0 1 ; plane stress with = 0 gives C = E 0 1 0
L
1 1 0 1 1 0
0 0 12
Eh
hL T
Element stiffness matrix: k e = B CBhdA = --------- B CB = ------4
2
T
Ve
3 1 2 1
3 0 1
2 0
1
Sym.
1
1
0
1
1
0
2
0
0
0
2
6.2 Use that N 1 + N 2 + N 3 + N 4 = 1 N 4 = 1 ( N 1 + N 2 + N 3 ) = 4 ( 1 )

N 4 ( , ) must satisfy the condition:
(i) zero in nodes 1, 2 and 3, i.e. N 4 ( 0, 0 ) = N 4 ( 1, 0 ) = N 4 ( 0, 1 ) = 0
(ii) unity in node 4, i.e. N 4 ( 1 2, 0 ) = 1
Both conditions are satisfied!
6.3
21
2+1
1
For 0 = 1 2, 0 = 1 we obtain N 1 = ----------------, N 2 = ----------------, N 5 = --- , and
4
4
2
N3 = N4 = N6 = N7 = N8 = 0.
R
R
Coordinates at the nodes: ( x 1, y 1 ) = ( 1, 1 ) ------- , ( x 2, y 2 ) = ( 1, 1 ) ------- , ( x 5, y 5 ) = ( 0, 1 )R
2
2
R
x 0 = x ( 0, 0 ) = N i x i = --- ,
Coord. in xy-plane:
2
y 0 = y ( 0, 0 ) =
Distance from origin:
(1 + 2)
N i y i = --------------------- R
2 2
2
2
+ 2 2 0.989R , i.e. the deviation of the FE-mesh
x 0 + y 0 = R 5------------------8
from the boundary of the hole is 1%.
6.4
The coordinates in the isoparametric element is given by the interpolation (coord.-transform):
4
x ( , ) =
L
N i x i = L ( N 2 N 4 ) = --- ( )
2
i =4 1
y ( , ) =
L
N i y i = L ( N 1 N 3 ) = --- ( )
2
i=1
x
L x
L y
L y
L
Partial derivatives: ------ = --- ; ------ = --- ;------ = --- ;------ = --
2
2
2
2
L
The Jacobi matrix and its determinant becomes: J = --- 1 1
2 1 1
6.11 (24)
L
J = ----2

20
6.5
Use the property
19
N i = 1 N 20 = 1
i=1
N i , Note that N1, ..., N19 are known!
i=1
Alternatively, derive the shape functions by use of the properties:

1
2
N20 = 0 in nodes 1 to 19 and that N20 = 1 N 20 = --- ( 1 ) ( 1 ) ( 1 )
4
6.6 (a)
x1
3
----= --- [ ( 1 ) ( x 2 x 1 ) + ( 1 + ) ( x 3 x 4 ) ] = --- l ;
4
2
y
------ = 0
x
y
1
1
-----= 0 ; ------ = --- [ ( 1 ) ( y 4 y 1 ) + ( 1 + ) ( y 3 y 2 ) ] = --- l
4
2
(b)
N 1 x
N 1 y
= J
B1
x =
6.7(a)
l
= --- 3 0
2 0 1
N 1
2
1
= --- 1 3 0 ( 1 ) 4 = ----- ( 1 ) 3
l 0 1 ( 1 ) 4
2l ( 1 )
N 1
1
= "" ----2l
(1 ) 3
0
0
(1 )
(1 ) (1 ) 3
N i x i = L ( 1 + ),
y=
N i y i = L
2
J = x y = L 1 + 0 and J = L ( 1 + )
x y
1
(b)
(c)
Strain: = Bde, where the B-matrix is given by
N i x
0
B = B 1 B 2 B 3 B 4 where B i =
0
N i y
with
N i y N i x
Inverse of the Jacobi matrix becomes
where
N 1 x
1
= -----4L
N 1 y
------------------( 1 -)
1 +
( 1 )
( 1 ) + -----------------------1 +
1 + --------------1
1 +
= -----4L
( 1 + )
N 3 y
( 1 + ) -----------------------1 +
N 3 x
N i x
N i y
= J
N i
N i
1
= --- 1 ( 1 + ) 0
L ( 1 + ) 1
N 2 x
1
= -----4L
N 2 y
1 --------------1 +
( 1 )
( 1 + ) -----------------------1 +
-------------------( 1 + )1
1 +
= -----4L
( 1 + )
N 4 y
( 1 ) + -----------------------1 +
N 4 x
6.12 (24)

(d)
1 1
= ------------------- , sin =
= 2L 1 + , cos
Se
N tdS =
B CBhdA
=
1
Ae
1+
l14
dS = h ------ d
2
= 1
f 1x = f4x = p 0 hL,
V = h J dd } =
1 1
B CBhJ dd =
ke =
14
B 1 CB 1 B 1 CB 2 B 1 CB 3 B 1 CB 4
6.8
1 1
------------------2
1+
T
1 1 B 3 CB 1
T
T
B 3 CB 2
T
B 3 CB 3
T
B 3 CB 4
hL ( 1 + )dd
B 4 CB 1 B 4 CB 2 B 4 CB 3 B 4 CB 4
B 2 CB 1 B 2 CB 2 B 2 CB 3 B 2 CB 4
= 1
thL 1 + d
f 1y
1 1
= f 4y = p 0 hL
N f V h J dd
1 1
2
J = L ( 1 + )
2
g,
fb3y = f b4y
1
one obtains with = --- that
2
7 2
= --- h L g
6
Note!
05 05070 7
f
biy
= 4hL g = Ve g
6.9 Numerical integration with Gauss-Legendre quadrature

n n
1 1
f bIy =
f ( ,
I
)dd =
1 1
f ( ,
I i
i
j )w i w j where fI = N I ( , ) ( g )hL ( 1 + )
(a) 1 x 1 scheme: n = n = 1 , 1 = 1 = 0 and w 1 = 2 .

fb1y
= N 1 ( 0, 0 ) ( g )hL ( 2 2 ) = h L g
2
in the same way we obtain f b2y = f b3y = f b4y = h L g

1
1
(b) 2 x 2 scheme: n = n = 2 , 1 = 1 = ------- , 2 = 2 = ------- and w 1 = w 2 = 1 .
3
3
fb1y
= ghL [ N 1 ( 1, 1 ) ( 1 + 1 ) ( 1 1 ) + N 1 ( 1, 2 ) ( 1 + 2 ) ( 1 1 )
5
2
+ N 1 ( 2, 1 ) ( 1 + 1 ) ( 1 1 ) + N 1 ( 2, 2 ) ( 1 + 2 ) ( 1 1 ) ] = --- ghL
6
5
2
7 2
in the same way we obtain f b2y = --- ghL and f b3y = f b4y = --- h L g ,
6
6
i.e. the numerical integration are exact!
6.13 (24)

6.10
Contributions to the nodal force vector: fs =
Se
N tdS , with dS = hds = h dx + dy .
x
y
x
y
Here dx = ------ d + ------ d and dy = ------ d + ------ d , but = 1 along 2-3 and thus
d = 0 , which give dS = h ( x ) + ( y ) d .
(a) 4-node element: N 2
= ( 1 ) 2 and N 3
=1
x = ( N 2 x2 + N3 x3 )
y = ( N 2 y2 + N3 y3 )
=1
=1
which give dS = h
f 2x =
N 2 tx hLd =
= ( x 3 x 2 ) 2 ,
= ( y 3 y 2 ) 2 ,
y3 y2 2
+ ---------------- d = hLd .
2
1
1
hL
N 2 ------------ t Ax + ------------ t Bx hLd = ------ ( 2t Ax + t Bx ) ,
2
2
3
hL
hL
hL
= ------ ( 2t Ay + t By ) , f 3x = ------ ( t Ax + 2tBx ) , f 3y = ------ ( t Ay + 2t By ) .
3
3
3
and f 2y
(b) 8-node element: N 2
=1
x = ( N 2 x2 + N3 x 3 + N 6 x6 )
y = ( N 2 y2 + N3 y 3 + N 6 y6 )
x
-----
( 1 ) ( )
= ----------------------------- , N 3
2
=1
=1
=1
N 2 tx hLd =
( 1 + )
= --------------------- , N 6
2
=1
= 1
x6
x2 + x3
= ---------------- =
2
N
2
N6
N6
+ ------ x 2 + N 3 + ------ x 3
2
2
y6
y2 + y3
- =
= --------------2
N
2
N
N
+ -----6- y 2 + N 3 + -----6- y 3
2
2
x 3 x 2 y y 3 y 2
= ----------------, ------ = ---------------- dS = h
2
f 2x =
= (1 + ) 2
x 3 x 2
--------------
2
=1
x 3 x 2
--------------
2
=1
=1
y3 y2 2
+ ---------------- d = hLd
1
1
hL
N 2 ------------ t Ax + ------------ t Bx hLd = ------ t Ax .
2
2
3
hL
hL
hL
In the same way we obtain f2y = ------ t Ay , f 3x = ------ t Bx , f 3y = ------ t By ,
3
3
3
hL
hL
f 6y = ------ ( 2t Ax + 2t Bx ) and f 6y = ------ ( 2t Ay + 2t By ) .
3
3
6.14 (24)

6.10 cont.
In (c) and (d) we have that t Ax = t Bx = tx = 0 cos 0 sin and
t Ay = t By = t y = 0 sin + 0 cos
(c) For a 4-node element this give f 2x = f3x = hL e t x and f 2y = f 3y = hL e ty , where
L e = L 3 . The nodal force vector becomes (nodes 1 to 4, global node numbering)
T
after assembly: F s = hLe t x ty 2t x 2t y 2tx 2t y t x t y

hL
hL
hL
(d) For a 8-node element this give f 2x = f 3x = --------e t x , f2y = f 3y = --------e t y , f6x = --------e 4t x
3
3
3
hLe
L
and f 6y = -------- 4t y , where L e = --- . The nodal force vector becomes (nodes 1 to 7,
3
3
global node numbering) after assembly:
hL
T
F s = --------e t x t y 4tx 4t y 2t x 2t y 4t x 4t y 2t x 2t y 4t x 4t y t x ty
3
1 1
6.11 f b =

1 1
N 1 0 N 2 0 N3 0 N4 0
T
ab
0 .
N fv h ------ dd where N =
and f v =
4
0 N 1 0 N2 0 N3 0 N4
g
N1 0
T
N fv =
0 N1
1 1
ghab
f
= 1y ; f1y = ----------------
4
g
0
1 1
T
fb
ghab
= ---------------- 0 1 0 1 0 1 0 1
4
6.15 (24)
(--------------------------------1 ) ( 1 ) ghab
dd = ---------------4
4

6.12
y/L
t = p0 1
0
Displacement boundary conditions: d1x = d1y = d2x = d2y = 0

give rise to the reaction forces: R1x, R1y, R2x, R2y
The external load acts on side x = 0, between nodes 1 and 3,
there, the shape functions reduces to
N 1 = 1 ,
x/L
N 3 = , N 2 = 0
The consistent nodal force vector becomes (use subst. = x/L, = y/L):
1
fs = N
T
=0
0 1
f 1x =
thLd
f 2x 0,
0 hL
(---------------1 )hLd = ------------2
3
f 3x 0
others equal to zero!

1
f 3x =
0 hL
(----------------1 + )
hLd = ------------2
3
R1x + p 0 hL 3
Reduced Eq. syst. (5) & (6):
R 1y
Load vector:
F =
R 2x
R 2y
p 0 hL 1
d
3Eh
---------- 1 0 3x = ----------16 0 3 d
6 0
3y
p 0 hL 6
0
Reaction forces:
p 0 hL
p0 hL
3Eh
Eq. (1): R1x = ---------- ( d 3x ) ------------ = -----------3
2
16
Eq. (3): R2x = 0
6.13(a)
With
d 3x
d 3y
8p
= --- ----0- L 1
9E 0
p 0 hL
3Eh
Eq. (2): R 1y = ---------- ( d 3x ) = ----------6
16
p 0 hL
3Eh
Eq. (4): R2y = ---------- ( d 3x ) = ----------16
6
2
x1 = 0
x2 = L 2
x 3 = 0 the shape functions
L
A e = ----4
y1 = 0
y2 = L 2
y 3 = L are obtained as
1
N 1 = --------- [ ( y 2 y 3 ) ( x x 2 ) + ( x 3 x 2 ) ( y y 2 ) ] = 1
2Ae
1
N 2 = --------- [ ( y 3 y 1 ) ( x x 3 ) + ( x 1 x 3 ) ( y y 3 ) ] = 2
2Ae
1
N 3 = --------- [ ( y 1 y 2 ) ( x x 1 ) + ( x 2 x 1 ) ( y y 1 ) ] =
2Ae
6.16 (24)

6.13(b) The stresses in the element are given by = CBde
N 1 x = 1 L ; N 1 y = 1 L
N 2 x = 2 L ; N 2 y = 0
N 3 x = 1 L ; N 3 y = 1 L
20
1 0
1
1
B = B 1 B 2 B 3 B 1 = --- 0 1 B 2 = --- 0 0
L
L
02
1 1
1 0
1
B 3 = --- 0 1
L
1 1
C =
P.S.
1---
= 3
d1
3 1 0
3E
-----=
8 1 3 0
0 0 1
14 d =
0
d e = d 2 where d 1 = 0 d2 =
3
0
1 108
1 6
d3
xx
Thus
d1
= yy = C B 1 B 2 B 3 d 2 = C [ B 1 d1 + B 2 d 2 + B 3 d3 ] = E 0
xy
6.14
Stiffness
matrix:
3
1
2
Eh 1
K = ------4 0
0
1
0
1
3
0
1
0
0
1
2
d3
2
0
3
0
1
1
0
1
1
1
0
3
0
2
1
0
0
0
1
0
3
1
2
1
0
0
1
2
1
3
0
1
1
1
0
1
2
0
3
0
0
2
1
0
1
1
0
3
12
0
1 18
Displacements are given by: D = K F

Stresses are given by: = CBD
(a) Displ. B.C. (symmetry & remove rigid body motion): D1x = D1y = D2y = D4x = 0
Force vector:
0 hl
0 hl
= F 1x F 1y F 2x F 2y F 3x F 3y F 4x F 4y = R R ---------- R 2y ---------- 0 R 4x 0
1x 1y 2
2
0
= ------ l 0 0 1 0 1 0 0 0
E
el. 1 & el. 2: = 0 1 0 0
exact!
(b) Displ. B.C. (remove rigid body motion): D1x = D1y = D2y = 0
Force vector:
0 hl
0 hl 0 hl
0 hl 0 hl 0 hl 0 hl 0 hl
---------- R 1y ---------- ---------- R 2y + ---------- ---------- ---------- ---------- ---------2
2
2
2
2
2
2
2
0
T
T
el. 1 & el. 2: = 0 0 0 1 exact!
D = ---------- l 0 0 0 0 1 0 1 0
E2
F =
1x
Note! E/2 = G (shear modulus) in this case!
(c) Displ. B.C. (remove rigid body motion): D1x = D1y = D2y = 0
Force vector:
T
ghl
ghl ghl
ghl
F = R
------------0 R 2y -------------- 0 -------------- 0 -------------1x R 1y
6
3
6
3
2
gl
= "" ----------- 0 0 1 0 3 9 2 15
24E
T
gl
el. 1: = -------- 1 15 1
24
6.17 (24)
T
gl
el. 2: = -------- 1 9 1
24

6.15(a)
2L
Consistent nodal force vector:
fs =
2L
f2y =
N 2 t y h ds =
0
2L
f3y =
N 3 t y h ds =
th ds
23
2L
s
= L ------- =
2
s
= ------- =
2
2L
f 2y 0, f3y 0
0 Lh
s - 0
--1- L ------------ h ds = -----------L
2
2 2
0
0 Lh
s - -------1- ------ h ds = -----------L 2 2
2
Zero displacement B.C.: d1x=d1y=d2y=d3x=0 => reaction forces: R1x, R1y, R2y, R3x
T
0 Lh
0 Lh
Node force vector: F = R R 0 R +
------------ R 3x -----------1x 1y
2y
2
2
6.15(b)
Eq. (3) and (6) gives the reduced equation system
0 Lh 0
d
3Eh
---------- 3 1 2x = -----------16 1 3 d
2 1
3y
Reaction forces:
Eq. (1):
Eq. (2):
Eq. (4):
Eq. (6):
d 2x
d 3y
= L -----0- 1 3
E
1
3Eh
R1x = ---------- ( 3d 2x d 3y ) = 0
16
0 Lh
3Eh
R1y = ---------- ( d 2x 3d 3y ) = ------------2
16
0 Lh
0 Lh
R2y + ------------- = 0 R2y = -----------2
2
R3x = 0
6.15(c)
The strains in the element are given by: = Bd e
N1 = 1 x L y L
N 1 x = 1 L ; N 1 y = 1 L
N 2 = x L N 2 x = 1 L ; N 2 y = 0
N 3 = y L N 3 x = 0 ; N 3 y = 1 L
x
=
y
xy
0
0
1 0 1 0 0 0
1 3
1
= --- 0 1 0 0 0 1 L -----0- 1 3 = -----0- 1
L
E
E
0
1 1 0 1 1 0
0
0
1
6.16(a) Displacement boundary conditions: d1x=d3y=d4x=d4y=0
6.18 (24)
B = B1 B2 B3 =
1 0 1 0 0 0
1
= --- 0 1 0 0 0 1
L
1 1 0 1 1 0
Agrees with the

exact solution
when = 1/3

6.16(b)
L
Consistent nodal force vector: fs =
f 2x =
N 2 tx h dy =
1
1
N 3 tx h dy =
f 2x 0, f3x 0 others are

equal to
zero!
1---------- 0 hL d = 0 hL
2
23
1
L
L
f 3x =
th ds
1-----------+ 0 hL d = 0 hL
2
Displacement boundary conditions give the reaction forces: R1x, R3y, R4x, R4y
T
Node force vector: Fe = R1x 0 0 Lh 0 0 Lh R 3y R 4x R 4y

6.16(c)
Force vector (external forces + reaction forces) can be calculated as Fe = K e D e
F1x = R 1x = 0 hL ; F2x = 0 hL ; F3x = 0 hL ; F4x = R4x = 0 hL

F 1y = 0 ; F2y = 0 ;F 3y = R3y = 0 ; F4y = R 4y = 0
6.17(a) Nodal force vector:

y
4
On the boundary x = 2L, between nodes 2 and 3,

the shape functions take the values:
(1 )
(1 + )
N 1 = N 4 = 0 och N 2 = -----------------, N 3 = ----------------2
2
Consistent nodal force vector becomes (subst. = y/L):
t = 0 y L
0
x
fe =
1
f 2x =
T
=1
thLd
0 hL
(---------------1 )hLd = ------------2
3
Inclusive of reaction forces, the

nodal force vector becommes:
f 3x =
f 2x 0,
f 3x 0 others zero!
0 hL
(----------------1 + )
hLd = ------------2
3
1
T
0 hL
0 hL
F = R 0
------------ 0 ------------ 0 R4x R4y
1x
3
3
6.19 (24)

6.17(b) Stresses are calculated as: = C where = Bd e
Nodal force vector (given):
T
de = dT d T dT d T
1 2 3 4
B-matrix:
B = B1 B2 B3 B4
4
where d 1 = d 4 = 0 and d 2 = --- -----0- L 1 ,
3 E 1
Strains:
x
=
y
xy
1
0
0
(1 + ) ,
( 1 + ) 1
y
xy
1
B 3 = -----4L
1+ 0
0 1+
1+ 1+
1- 4---
2---
-----0- L [ ( 1 ) + ( 1 + ) ]
-----0-
----3E
4L 3 E
= B 2 d2 + B 3 d 3 =
=
y
y
xy
Stresses:
x
J = x y = L 0
x y
0 L
with Jacobi matrix
1
B 2 = -----4L
4
d3 = --- -----0- L 1
3 E 1
2--2

--- -----0-
E 0 0
3 0
3E
= C = 0 E 0
=
y
y
0 0 E2
xy
xy
plane stress ( = 0.3)
6.20 (24)
xy
Note!
The solution based on one
FEM element deviates from the
exact solution: x = 0 y L
y = xy = 0

6.18 (a)
The contribution from the volume force is: F b =
Here, both f bix 0 and
L (L x)
f b1x =
f b2x =
0 ( L x )
L (L x)
0 ( L x )
L (L x)
f b1y =
0 ( L x )
f b3y =
0 ( L x )
N KdV =
0
L
h
2
hL
= ------------- 2x ( L x )dx = ------------------L
6
0
1---
hL
--------------------x- + --y- 2 xhdy dx =
------------h- ( L x ) 2 xdx =
1
12
2
L L
L
1---
x y
gh
2
ghL
1 --- --- ( g )hdy dx = ---------- ( L x ) dx = ---------------
L
2
L L
3
dx
-(
g
)hdy
0 ( L x )
L (L x)
Ve
N Khdy dx

0 ( L x )
2
h
2
hL
1---
x y
1 --- --- xhdy dx = ------------- ( L x ) xdx = ------------------
L
2
L L
12
L (L x)
f b2y =
i = 1, 2, 3
x- 2
dx
-
xhdy
0 ( L x )
L (L x)
f b3x =
f bix 0
L (L x)
2 gh
ghL
= ------------- x ( L x )dx = ---------------L
3
0
gh
ghL
1---
x y
2
1 --- + --- ( g )hdy dx = ---------- ( L x ) dx = ---------------
L
3
2
L L
The displacement B.C:s : d1x = d1y = d3x = 0 give rise to reaction forces!
R1x
1
0
R1y
0
1
2
3
2
hL
ghL
The total nodal force vector becomes: f = ------------------- 2 ---------------- 0 + 0
12
3 1
0
0
1
0
R3x
0
1
0
6.18 (b)
Displacement boundary conditions: d1x = d1y = d3x = 0
Reduced equation system with
respect to the boundary:
conditions, Eq. (3,4,6):
8 0 0 d 2x
2
3 2
2 0
Eh
hL ghL
------- 0 4 2 d 2y =
-------------------------------- 1
0
12
3
8
0
1
0 2 3 d 3y
d 2x
d 2y
d 3y
6.21 (24)
2 3 1
2 0
gL
= ---------------- 0 ------------ 5
6E
3E
0
6

6.18 (c) The stresses in the element is given by = C ,
where = Bd e = B 1 d 1 + B 2 d2 + B 3 d 3
Here we have
1 0 0
2 3
2
2
L- 1
gL 0
gL 0
------------------------------------C = { = 0 } = E 0 1 0 ; d 1 = 0 ; d2 =
; d3 =
6E 0
3E 5
3E 6
0 0 12
N 2 x
N 3 x
0
1 0
1 0
1
1-----B2 =
=
;
B
=
=
0
N 2 y
0
N 3 y
3
L 0 0
2L 0 1
0 1
0 1
N 2 y N 2 x
N 3 y N 3 x
Thus
y
xy
1 0 0
= E 0 1 0
1
0 0 --2
0
1 0 2 3
2
1 L - 1
gL 0
--------------------------+
0 L 0 0 6E
3E 5
0
0
0 1
x
y
xy
1 0
2
1
gL
+ ------ 0 1 ------------ 0
2L
3E 6
0 1
0
2 2 1
L
gL
= ---------------- 0 ---------- 3
6
3
0
1
Note that the stress solution is approximate and rather poor in this case!
To obtain a better solution, many more elements are needed!
6.19(a)
Body force vector in element 2:

1 1
f b2 =
N KdV =
Ve
1 1
f 1x =
J = L
f 4x = f 1x
(---------------------------------1 + ) ( 1 )- 2
2
L ( 3 + )hL dd
4
f 3x = f 2x

1 1
T
f b2
(--------------------------------1 ) ( 1 -) 2
2
L ( 3 + )hL dd
4
1 1
1 1
f 2x =
T
N K h J dd where K = x = L ( 3 + )

0
0
1 1
hL
= ------------------- 8 0 10 0 10 0 8 0
3
6.22 (24)
6.19(b) Stresses in the element is given by = C = CBd e

1 0 0
Here, C = E 0 1 0 d T = d T d T d T d T
1 2 3 4
0 0 12
e1
d1 = d4 = 0
where
d2 = d3 = x
0
2 3
44 L
x = ------ ---------------3 E
= C ( B 1 d 1 + B 2 d 2 + B 3 d 3 + B 4 d 4 ) = C ( B 2 d2 + B 3 d 3 )
1
B i = --L
N i,
0
0
N i,
1
B 2 = -----4L
N i, N i,
xx
1
0
1
0
1----;
B
=
0
( 1 + )
0
(1 + )
3
4L
( 1 + )
1
(1 + )
1
x + x
E x ( 2L )
1
1 0 0
1
22
2 2
--------= E 0 1 0
=
=
L 0
0
0
4L
3
0 0 12
( 1 + )x + ( 1 + ) x
0
0
yy
xy
Comparison with the exact solution:

FEM
x=0
x = 2L
2.2 %
22.2 %
Exact
xx xx
100%
Relative eroor = ---------------------------------Exact
xx
6.20(a) Boundary conditions:
x=L
8.3 %
D 1x = D 1y = D 2x = D 3x = D 3y = D 4y = D 7y = 0
6.20(b) For element e4 yields:

y
x y
x
Shape functions: N 1 = 2 ---, N 2 = --- + --- 2, N 3 = 1 --L
L L
L
t =
x
L
p 0 --L
T
f e = N y = 2L thdx f 2y 0, f 3y 0 others zero!
0
e4
0
L
f 2y
hLp 0
x
x
= --- p 0 --- hdx = -----------L
L
3
0
f 3y =
hLp 0
x
x
--- p0 --- hdx = -----------L
L
6
Assembly of global nodal force vector including the reaction forces gives:
F1x = R 1x ; F1y = R1y ; F2x = R2x ; F3x = R3x ; F4y = R4y ; F7y = R7y
F3y = hLp 0 6 ; F 6y = hLp 0 3
6.23 (24)

6.20(c) Stress in element 1 is given by = CBde
Shape functions:
N1 = 1 x L y L
N 1 x = 1 L ; N 1 y = 1 L B = B 1 B 2 B 3 =
N 2 = x L N 2 x = 1 L ; N 2 y = 0
N 3 = y L N 3 x = 0 ; N 3 y = 1 L
1 0 1 0 0 0
1
= --- 0 1 0 0 0 1
L
1 1 0 1 1 0
1
0
d1
E
1
0
C = { P.S. } = -------------2d e = d 2 d = 0 (R.V.); d 2 = D 4x ; d 3 = 0
1
D 2y
(1 )
1
0
0 0 ---------------d3
2
xx
D L
E ( D 4x + D 2y ) L
Thus = yy = C [ B 1 d1 + B 2 d 2 + B 3 d3 ] = C 4x
= -------------2D 2y L
1 ( D 2y + D 4x ) L
xy
6.21(a) Boundary conditions:
D 1x = D 1y = D 2x = D 2y = D 3x = D 3y = D 4x = D 4y = D 5x = D 5y = 0
6.21(b) For one element yields:

1
0
t =
0
fs =
T
=1
thbd
1
1
f2y =
=1
thbd = 0 hb
1 )(---------------d = 0 hb
2
f 2y 0, f 3y 0 others are equal

to zero!
f 3y =
=1
( 1 + )
thbd = 0 hb ----------------- d = 0 hb
2
Assembly of consistent nodal forces from elements gives non-zero components

F 26y = F 30y = 0 hb och F27y = F 28y = F 29y = 2 0 hb
according to:
6.21(c) Stresses in element 1 is given by = CBde

D6y
4
d1
D6x
D7y
de =
D7x
d2
d3
where d 1 = d4 = 0 (B.C.); d2 =
D 7x
D 7y
d4
C =
P.S.
= 0
1 0 0
= E 0 1 0 ; B = B 1 B 2 B 3 B 4 where
0 0 12
B2 and B3 evaluated in the centroid of element 1 ( = = 0) becomes:

1
B 2 = --4
Thus
1a
0
0 1 b ;
1 b 1 a
xx
= yy
xy
1
B 3 = --4
; d3 =
N
--------i
x
N i
-------y
D 6x
D 6y
1 N i
= --- -------a
1 N
= --- --------i
b
1a 0
0 1b
1b 1a
E
= CB 2 d 2 + CB 3 d 3 = --4
( D 6x + D 7x ) a
( D 6y D 7y ) b
( D 6x D 7x ) ( 2b ) + ( D 6y + D 7y ) ( 2a )
6.24 (24)
7. FEM: Heat conduction

7.1 The Figure to the right shows a one dimensional
a
model of a cooling fin. At the left boundary the T = 80C
o
x=L
b
temperature is constant at 80 C. Along the remaining boundary, heat is lost to the surrounding air by
convection. Determine the cooling effect of the fin,
Convection
i.e. calculate the heat flow across its left boundary.
x
Also determine the displacement in the fin due to
T = 20C
thermal expansion, where it can be assumed that
Data:
the fin is undeformed at 20o C. Analyse the proba = 1 cm, b = 0.4 cm, L = 8 cm,
lem by FEM and use two linear elements. Carry out
5
E = 80 GPa, = 1.4 10
the analysis in two steps: (a) calculate the temperak = 3 W/cm/oC, h = 0.1 W/cm2/oC
ture distribution in the fin and the heat flow at x = 0,
and (b) calculate the displacement in the fin.
The material has elasticity modulus E, thermal expansion coefficient , thermal conductivity k
and convection coefficient h. The material data and the geometry is shown in the figure.
7.2 A wall made of two layers of different material is shown in the

right hand figure. The temperature at the right side of the wall is
kept constant at 20o C. At the left side, a heat flux arises due to
convection, where the ambient temperature is equal to 5o C. The
thermal convection coefficient is h and the thermal conductivity of
the two materials is k1 and k2, respectively. Determine the temperature distribution through the wall by FEM-analysis. In the present
case it suffice to model each layer by one linear element.
L1
L2
Data: k1 = 0.2 W/cm/oC, k2 = 0.06 W/cm/oC, h = 0.1 W/cm2/oC, L1 = 2 cm and L2 = 5 cm.
7.3 A bar of copper is clamped between two rigid

T = 20C
walls. At point B, the bar is influenced by a point force
b
Konvektion
P0 = 2kN and a heat source keeping the temperature
b C
A
constant at 100o C. The right boundary (point C) is
B P0
insulated and at the left boundary (point A) the temperature is kept constant at 20o C. Between the end
L
2L
points of the bar, a heat flux occurs by convection,
where the ambient temperature of the surrounding
medium is equal to 20o C. Conduct a FEM-analysis to determine the distribution of temperature, displacement and normal stress in the bar. Use two linear elements in each of the intervals A-B and B-C, respectively. Carry out the analysis in the two consecutive steps: (a)
calculate the temperature and (b) calculate the displacement and normal stress.
Data: L = 10 cm, b = 1 cm, k = 3.9 W/cm/oC, h = 0.01 W/cm2/oC, E = 125 GPa and
5
= 1.8 10 .
7.1 (3)
Solutions
7.1
(a) Temperature distribution
x = 0: T = 80 oC
T
x = L: kA ------ = hA ( T T )
x
FEM-analysis: 2 linear elements

T1
T2
Equation
system
T3
T 1 = 80C
11.20 + Q R
0.6733 0.1133
0
=
T2
0.1133 1.3467 0.1133
22.40
0
0.1133 0.7133
T3
12.00
unit [W]
unit [W/oC]
T2
T3
= 25.12 C
20.81
Reaction
heat flow
= 39.82 W
(b) Displacement
D1
D2
D3
x = 0: u = 0
x = L: A = 0
3
1.459 10 + R
1 1 0 D1 = 0
6
3
80 10 1 2 1
=
D2
1.326 10
0 1 1
3
D3
0.133 10
unit [N/m]
unit [N]
D
D
2
3
3
= 0.01823 10 m
0.01989
R = 0
Comparison between the exact solution and FEM-solutions based on 2, 4 and 8 elements
80
60
50
40
30
20
Displacement / m
Exact solution
2 element
4 element
8 element
Temperature / C
70
15
10
Exact solution
2 element
4 element
8 element
20
10
4
6
x / cm
7.2 (3)
4
6
x / cm

7.2
l1
l2
T1
T2
Equation
system
k1
----- + h
l1
A
T
Boundary conditions, x = 0: kA ------ = hA ( T T )
x
x = L:
T = -5 oC
T3
k
----1l1
k1 k1 k2
----- ----- + ----l1 l 1 l 2
k2
----l2
= A 0

Q A
Q
0
k2
----l2
k2
----l2
T1
T1
hT
T2
T 3 = 20C
T2
2.58 C
0.161
2
A = 0.242 [ W cm ]
7.3
T1
D1
e1
T2
D2
e2
T3
T4
e3
D3
Prescribed values:
x = xA: T = 20o C
x = xB: T = 100o C
T
x = xC: Q = kA ------ = 0
x
T5
e4
D4
D5
(a) Temperature distribution, divide into two separate analysis, since T3 is prescribed.
T1
T2
T3
T3
T 1 = 20C
2 + Q R1
0.8467 0.7467
0
=
T2
0.7467 1.6933 0.7467
4
0
0.7467 0.8467 T 3 = 100C
2 + Q R3
T4
T5
T 3 = 100C
4 + Q R3
0.5233 0.3233
0
=
T4
0.3233 1.0467 0.3233
8
0
0.3233 0.5233
4
T5
= 55.27C
T4
= 50.54 C
38.87
T5
(b) Displacement and stress calculations

Valid for each element:
2
2
6
125 10 0
0
0
EA
k e = ------- 1 1 ;
li 1 1
2
4
2
0
0
0
2
3
1
0
0
0
1
2
1
T i = average temperature
f T = EAT i 1
1
change in element i
T 4
D1 = 0
R1
0
3.968
T 1 T 2
D2
0
0
9.00
3
= P 0 + EA T T = 2.532 10 +
D3
0
2
3
1
6.877
D4
0
T 3 T 4
1 D =0
5.559
R5
5
T 4
Stress in the elements:
D2
D3 =
D4
14.08
6
7.84 10 m
31.43
R1
R5
D 2 D1
1 = E ------------------- T 1 = 74.9 MPa,
l1
D 3 D2
2 = E ------------------- T 2 = 74.9 MPa
l2
D 4 D3
3 = E ------------------- T 3 = 94.9 MPa,
l3
D5 D 4
4 = E ------------------- T 4 = 94.9 MPa
l4
7.3 (3)
7487 N
9487
R1
0
0
0
R5

FEM Practice

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

FEM Practice

Загружено:

Авторское право:

Доступные форматы

FEM for Engineering Applications

Exercises with Solutions

1. Elastic Energy and Energy principles

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

1. Elastic Energy and Energy principles

1.4 A plane structure is composed of four equal beams

1.5 A cantilever beam of length L and with bending stiffness EI is

1.6 A beam of length 2L and with bending stiffness EI is

1.3 Determine the horizontal displacement at

1.2 Two beams, each with bending stiffness EI and

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

1.8 A system of two beams, each with

1.9 The constraint and the boundary conditions of a beam

1.10 A beam with bending stiffness EI and total

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

1.12 A plane frame is composed of three beams connected

1.14 A planar frame constructed by two beams, each

1.15 A beam with circular cross section (diameter

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

1.16 A planar frame, according to the right hand fig- A

1.17 A rectangular planar frame with measurements

1.18 A planar frame of quadratic shape is freely supported at the

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

Complementary elastic energy:

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

Total complementary elastic energy:

Total complementary elastic energy:

Note! one statically

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

Free body diagram:

Case (b): statically indeterminate problem,

According to the given conditions:

MTotal complementary elastic energy: W = 2 -------dx =

Displacement at the point force (Castiglianos 2nd theorem):

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

Note! one statically

mentary elastic energy, W , is calculated.

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

2. Matrix formulated structural mechanicsdirect method

2.1 A system of five springs are connected as

2.2 Three springs are connected according

2.3 Determine the displacement at the

2.4 The plane structure to the right consists of four

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

2.8 A mass m0 is attached to three similar springs. The

2.7 The plane structure in the figure to the right contains

2.6 The plane frame structure to the right contains two

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

2.10 A structure of three truss elements is loaded by two point

2.12 Determine the displacements and the reaction

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

2.15 Determine the displacements and the reaction forces

2.16 Determine the displacements and the reaction

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

Boundary conditions & prescribed forces:

Element stiffness matrix: k e

Equation (7) gives

FEM for Engineering ApplicationsExercises with Solutions / August 2008 / J. Faleskog

Element stiffness matrices:

e1& e4: k 1 = k 4 = 2k,

Eqs. (8) & (9):

Normal force in element 4: N = k 4