Академический Документы
Профессиональный Документы
Культура Документы
2L
M0
A,
E,
EI, L
M0
1.1 (10)
EI, L
45o
L
45o
A,
E,
EI, L
k
L
EI
L
M0
L
P0
P
(b)
EI, L
EI, L
kN
2
EI, L
(a)
EI, L
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.
1.11 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
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
D
P
C
L
1.3 (10)
B
L
P
B
l
L
C P
C
L
D
2L
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
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
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
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
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:
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)
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
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
1.9
2P
P
P
M0
M0
M0
L
: M 0 = PL
Equilib:
(statically indeterminate probl.)
M( x)
1.7 (10)
2 3
x
P L
= PL --- = -----------L
3EI
3
W
2 PL
P = -------- = --- --------P
3 EI
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
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-
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)
k3
k1
k4
k5
k2
D1, F1
D2, F2
D3, F3
D4, F4
4a
k2
k3
k1
3a
rigid beam
P
45o
k1
45o
k1 = k
k2
45o
k2 = 2k
k3
k3 = 2k
(1,1)
y/L
(1,1)
k1
k2
x/L
P
5
3
(1,0)
k3
k4
(1,1)
2.1 (12)
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
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,
E,
A,
L
EA,L
k2
x
rigid
ring
e1
l
e2
2P
E, 2A, 2L
E, A, L
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.
3 2 1 2
32
3
2
3k
2k
y
P
12
2.3 (12)
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
x/L
y/L
y
1
2
--k2
3
stel
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
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
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 =
R4
R5
3kD 7 = P
R6
P
R8
2.5 (12)
Reactionsforces
P
= -----3k
D4
D2
D1
e1
F8 = P, F9 = P
e2
D10
D6
D9
D5
e3
e4
D8
D7
2
c = cos
Ke = k i a a , a = c sc ,
2 s = sin
a a
sc s
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
D8
D9
P
= --------- 2
14k 4
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
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
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
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
D6
D3
e4
D2
D10 e3
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
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
D5
mg
Assembly of the stiffness matrix:
K = k
a1
a1
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
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
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
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)
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
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
Boundary conditions:
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
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
Boundary conditions:
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
Boundary conditions:
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
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
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)
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
Boundary conditions:
Equationsystem:
D1 = D2 = 0
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
Boundary conditions:
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
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
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.
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 ,
Kx
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
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
dx
dx
(b)
EA
x
g
x=L
P
x = 2L
x=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.
(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 .
3.2 (15)
z,w
q
x
kz
x = L
d - --------d w
------ EI
+ kz w q = 0 .
2
2
dx dx
x=L
T
M
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
2L, EI
3.3 (15)
kw
3EI
EI
k w = --------3- k = -----L
2L
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
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)
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).
(a) Show that the weak form to the differential equation is
3L 2
3L 2
dv
------ kA dT
------ dx +
dx dx
3L 2
vhPTdx = [ v ( Q ) ]
3L 2
0
v ( qA + hPT )dx ,
3.5 (15)
d- dT
----kA ------ hP ( T T ) = 0 .
dx dx
(a) Show that the weak form to the differential equation is
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
=
13L 6L 22L 8L
Solutions
x2
3.1
d
d
v ------ D ------
dx dx
x1
x2
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.
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
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
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 )
3.7 (15)
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
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
D2
D3
2
B = --- 1 1
L
N = (1 )
QL
QL
= ----------------- 74 = -------- 0.525
141EA 140
EA 0.993
3.8 (15)
point force
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
Weighted residual:
v [ ( EIw ) + k z w q ]dx = 0
(1)
L
L
Integration by parts:
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:
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
Element
stiffness
matrices
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
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:
Integration by parts:
= [ v ( EAw ) ]
v [ ( EIw ) ]dx =
[ v ( EAw ) ]
L
L
vEIwdx (2)
(2) inserted into (1) with T = ( EIw ) and M = EIw gives the weak form:
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
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
Beam (d.o.f 1 to 4)
L
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
3 3L
2
EI
K = K balk + K fjder = --------3- 3L 4L 3L 2L
2L 3 3L 4 3L
2
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
R 1 gAL
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
Integration by parts:
(1)
L
= [ v ( EAw ) ]
v [ ( EIw ) ]dx =
[ v ( EAw ) ]
vEIwdx
(2)
(2) inserted into (1) with T = ( EIw ) and M = EIw gives the weak form:
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)
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
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
Reaction
force/moment
3
EI3Q
2 PL 1 QL
-------3d 1 = P ------- d 1 = --- --------- --- ---------3
10
3 EI 5 EI
2L
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
3L 2
vhPTdx = [ v ( Q ) ]
v = Nb e =
3L 2
dTdN
----= ------- T e = BT e
dx
dx
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
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
e1
FEM - model:
T1
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
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
T2
T3
T
= ------ 29 = T 1.3809
21 22
1.0476
3.9 (d)
2
[ N ( Q ) ]1 c =
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
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)
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
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
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
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)
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)
0
1
3
x=0
x = 2L
x = --- L
N 1 = 1 , N 2 = dx = Ld
2
(a)
(b)
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
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
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
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
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)
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
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)
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
D3
D2
k1 = k2 =
L 2
Eq.
system:
3L 2
D5
D4
D6
Distributed load contribution
to the nodal force vector:
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
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
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
y/L
(0,1)
(1,1)
2
Kx
1
(1,0)
L
Q
2L
Q2
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)
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 =
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
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
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
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
R3
D4
R4
D7
D8
D5
R5
D6
R6
D7
Q 2
Q
D8
-------------, 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
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
5.4 (6)
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
R y = P,
M y = P2L
5.4
D2
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
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 )
Ne = 4
Ne = 1
Ne = 2
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)
N1 = 1
y/L
1
N 2 = c
3
1
N3 =
= x L, = y L
x/L
1
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
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
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
y/L
6.2 (24)
x
L
12
1
1
a
(b)
19
15
11
b+l
(a)
1 x/L
2
1
1
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
A
1
y
b
thickness h
x
a
6.3 (24)
6
5
4
3
2
1
Body force
Kx = 0
Ky = g
(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
(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
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)
y
1 0
t = ------2 0
L
L
s
1
Shape functions:
x y
x
N 1 = 1 --- ---, N 2 = ---,
L L
L
Element stiffness matrix:
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
(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
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)
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)
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)
(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
6.8 (24)
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)
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
(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
6.5
19
N i = 1 N 20 = 1
i=1
i=1
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)
= ------------------- , 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
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 + )
= N 1 ( 0, 0 ) ( g )hL ( 2 2 ) = h L g
2
= 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)
Se
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
=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)
6.11 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
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
f 3x =
0 hL
(----------------1 + )
hLd = ------------2
3
R1x + p 0 hL 3
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)
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
(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
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
(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
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
6.18 (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
f 2x =
N 2 tx h dy =
1
1
N 3 tx h dy =
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
t = 0 y L
0
x
fe =
1
f 2x =
T
=1
thLd
0 hL
(---------------1 )hLd = ------------2
3
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)
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
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
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
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)
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)
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
x=0
x = 2L
2.2 %
22.2 %
Exact
xx xx
100%
Relative eroor = ---------------------------------Exact
xx
x=L
8.3 %
D 1x = D 1y = D 2x = D 3x = D 3y = D 4y = D 7y = 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)
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
0
t =
0
fs =
T
=1
thbd
1
1
f2y =
=1
thbd = 0 hb
1 )(---------------d = 0 hb
2
f 3y =
=1
( 1 + )
thbd = 0 hb ----------------- d = 0 hb
2
according to:
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
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)
L1
L2
7.1 (3)
Solutions
7.1
(a) Temperature distribution
Boundary conditions:
x = 0: T = 80 oC
T
x = L: kA ------ = hA ( T T )
x
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
FEM-analysis: 2 linear elements
D1
D2
D3
Boundary conditions:
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
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
FEM-analysis: 4 linear elements
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
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
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