The Finite Element Method for Dynamic Analysis

Swiss Federal Institute of Technology Page 1
The Finite Element Method

for the Analysis
y of Non-Linear and Dynamic
y Systems
y
Prof. Dr. Michael Havbro Faber

Swiss Federal Institute of Technology
ETH Zurich, Switzerland
Method of Finite Elements II

C t t off T
Contents Today's
d ' Lecture
L t
• S l i
Solution off Equilibrium
E ilib i Equations
E i in
i Dynamic
D i Analysis
A l i
Direct Integration methods
- The central difference method
- The Houbolt method
- The Wilson θ method
- The Newmark method
- Coupling of integration operators

I t d ti to
Introduction t Dynamic
D i Analysis
A l i
• I
Introduction
d i
The very basics ☺ k Unstretched p

position
Newtons 2‘nd law of motion: kΔ
mx = w − k (Δ + x) Δ
m m
⇓ Stretched position
w = mg (static equilibrium)
mx − w + k (Δ + x) = 0
⇓ x x ,
x
mx − k Δ + k (Δ + x) = mx + kx = 0
k
x + x =
x + ωn2 x = 0 ωnτ = 2π ⇒ τ =
2π
= 2π
m
m ωn k
1 1 k
⇓ fn =
τ
=
2π m
x = A sin ωnt + B cos ωn t ⇒ t x = 0 x sin ωnt + 0 x cos ωnt

I t d ti to
D i Analysis
A l i
• I
Introduction
d i
We have previously considered the equilibrium equations

governing the linear dynamic response of a system of finite
elements:
+ CU
MU + KU = R FI (t ) + FD (t ) + FE (t ) = R (t )
M : Mass matrix
C : Damping matrix
FI (t ) = MU
K : Stiffness matrix
F (t ) = CU
D
U : Displacements
: Velocities
FE (t ) = KU
U
: Accelerations
U
I t d ti to
D i Analysis
A l i
• I
Introduction
d i
Whether a dynamic analysis is needed or not is generally up to

engineering judgment –
requires
q understanding
g of the interaction between loading
g and
structural response!
IIn generall – if th
the lloading
di varies
i over time
ti with
ith frequencies
f i
higher than the Eigen-frequencies of the structure –
then dynamic analysis will be required

I t d ti to
D i Analysis
A l i
• I
Introduction
d i
+ CU
MU + KU = R
In principle the equilibrium equations may be solved by any

standard numerical integration scheme – BUT !
Efficiency – numerical efforts – must be considered and it is

worthwhile to look at special techniques of integration which are
especially suited for the analysis of finite element assemblies.

Di t IIntegration
Direct t ti Methods
M th d
+ CU
MU + KU = R
• Di
Direct Integration
I i Methods
M h d
Direct means: The equations are solved in their original form!
Two ideas are utilized
1) The equilibrium equations are satisfied only at time steps,

i.e. at discrete times with intervals Δt
2) A particular variation of displacements, velocities and

accelerations within each time interval is assumed
The accuracy depends on these assumptions as well as the

choice of time intervals!

Di t IIntegration
Direct t ti Methods
M th d
+ CU
MU + KU = R
• Di
Direct Integration
I i Methods
M h d
The displacements, velocities and accelerations

0
U : Displacement vector at time t = 0
0
U : Velocity vector at time t = 0
0
U : Acceleration vector at time t = 0
are assumed to known and we aim to establish the solution of
the equilibrium equations for the period 0 - T.
So we sub-divide
S b di id T into
i t n intervals
i t l off length
l th Δt = T/n
T/ and
d
establish solutions for the times Δt , 2 Δt , 3 Δt , …,T.

Di t IIntegration
Direct t ti Methods
M th d
+ CU
MU + KU = R
• Di
Direct Integration
I i Methods
M h d
We distinguish principally between
Explicit methods:
Solution is based on the equilibrium equations at time t
Implicit methods:
S l ti is
Solution i based
b d on the
th equilibrium
ilib i equations
ti att time
ti t + Δt

Di t IIntegration
Direct t ti Methods
M th d
+ CU
MU + KU = R c
• Th Central
The C l difference
diff method
h d
t
U = 1 ( t −Δt U − 2 t U + t +Δt U) a
Δt 2
t 1
U= (− t −Δt U + t +Δt U) b
2Δt
a and b inserted in c
⎛ 1 1 ⎞ t +Δt ⎛ 2 ⎞t ⎛ 1 1 ⎞ t −ΔΔt
⎜ 2 M + C ⎟ U = t
R − ⎜ K − M ⎟ U − ⎜ 2 M − C⎟ U
⎝ Δt 2Δt ⎠ Δt ⎝ Δt 2Δt ⎠
2
⎝ ⎠
⇓
( a0M + a1C ) t +Δt U = t R − ( K − a2M ) t U − ( a0M − a1C ) t −Δt U

Di t IIntegration
Direct t ti Methods
M th d
( a0M + a1C ) t +Δt U =
t
R − ( K − a2 M ) t U − ( a0 M − a1C ) t −Δt U
• Th Central
The C l difference
diff method
h d
We see that we do not need to factorize the stiffness matrix ☺
We also see that in order to calculate the displacements

p at time
Δt we need to know the displacements at time 0 and –Δt
In general 0 U, 0 U and 0 U
are known and we may use a and b to
obtain −Δt Ui
−Δt + Δt 2
0
U i = U i − Δt U
0
i
0
Ui
2
⇓
−Δ
Δt + a 0U
U i = 0 U i − Δt 0 U
i 3 i

Di t IIntegration
Direct t ti Methods
M th d
• Th Central
The C l difference
diff method:
h d Solution
S l i procedure:
d
A: Initial calculations
1) Form stiffness matrix, mass matrix and damping matrix

2)) Initialize 0 U, 0 U and 0 U

3) Select time step Δt, Δt ≤ Δtcr and calculate integration constants
1 1 1
a0 = , a1 = , a2 = 2a0 , a3 =
Δt 2 2Δt a2
4) Calculate −Δt + a 0U
U = 0 U − Δt 0 U
3
5) Form effective mass matrix M̂ = ( a0 M + a1C )
6) Triangularize M M̂ = LDLT

Di t IIntegration
Direct t ti Methods
M th d
• Th Central
The C l difference
diff method:
h d Solution
S l i procedure:
d
B: For each time step
1) Calculate effective loads at time t:

t ˆ = t R − ( K − a M ) t U − ( a M − a C ) t −Δt U
R 2 0 1
2) Solve for the displacements U at time t + Δt

ˆ
LDLT t +Δt U = t R
3) If required, solve for the corresponding velocities and

accelerations
t = a ( t −Δt U − 2 t U + t +Δt U)
U 0
t = a (− t −Δt U + t +Δt U) It is not required to factorize the
U 1 stiffness matrix – explicit method
Di t IIntegration
Direct t ti Methods
M th d
• The Central difference method
The effectiveness of the central difference method depends on the

efficiency of the time step solution – because we need a lot of
them
F this
For thi reason the th method
th d isi usually ll only
l applied li d when
h a lumped l d
mass matrix can be assumed and when the velocity dependent
damping can be neglected, i.e.:
1 t +Δt t ˆ
M U = R
t ˆ Δt
2
Δt 2
t +Δt
U i = Ri ( ), mii > 0
⎛ 2 ⎞t ⎛ 1 ⎞ t −ΔΔt mii
t
R̂ = R − ⎜ K − 2 M ⎟ U − ⎜ 2 M ⎟ U
R t
⎝ Δt ⎠ ⎝ Δt ⎠
K t U = ∑ K (i ) t U = ∑ t F
(i )
1
R = t R − ∑ t F − 2 M ( t −ΔΔt U − 2 t U)
t ˆ (i ) i i
i Δt
Di t IIntegration
Direct t ti Methods
M th d
• The Central difference method: Example:
With natural periods:

T1 = 4.45
k1 = 4 T2 = 2.8
m1 = 2 R1 = 0
U1 ,U1 ,U1
⎡ 2 0 ⎤ ⎡U1 ⎤ ⎡ 6 −2 ⎤ ⎡U1 ⎤ ⎡ 0 ⎤
k2 = 2 ⎢ 0 1 ⎥ ⎢ ⎥ + ⎢ −2 4 ⎥ ⎢U ⎥ = ⎢10 ⎥
⎣ ⎦ ⎣U 2 ⎦ ⎣ ⎦⎣ 2⎦ ⎣ ⎦
m2 = 1 R2 = 10
U 2 ,U 2 ,U2
k3 = 2

Di t IIntegration
Direct t ti Methods
M th d + Δt
2
−Δt
Ui = Ui − Δt U
0 0 0
U
i i
2
⇓
−Δ
Δt + a 0U
Ui = 0 Ui − Δt 0 U
• The Central difference method: Example: i 3 i
⎡ 2 0 ⎤ ⎡U1 ⎤ ⎡ 6 −2 ⎤ ⎡U1 ⎤ ⎡ 0 ⎤ With Eigen periods:

⎢ 0 1 ⎥ ⎢ ⎥ + ⎢ −2 4 ⎥ ⎢U ⎥ = ⎢10 ⎥ T1 = 4.45
⎣ ⎦ ⎣U 2 ⎦ ⎣ ⎦⎣ 2⎦ ⎣ ⎦
T2 = 2.8
We will calculate the the response
p of the system
y for
Δt=T2 /10 and for Δt=10 T2 over 12 time steps
First we calculate
0
U
⎡ 2 0 ⎤ ⎡ 0U1 ⎤ ⎡ 6 −2 ⎤ ⎡0 ⎤ ⎡ 0 ⎤ ⎡ 0U1 ⎤ ⎡ 0 ⎤
⎢ 0 1 ⎥ ⎢ 0 ⎥ + ⎢ −2 4 ⎥ ⎢0 ⎥ = ⎢10 ⎥ ⇒ ⎢ 0 ⎥ = ⎢10 ⎥
⎣ ⎦ ⎣ U2 ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ ⎦ ⎣ U2 ⎦ ⎣ ⎦
Then we continue with the steps ☺

Di t IIntegration
Direct t ti Methods
M th d
⎡ 2 0 ⎤ ⎡U1 ⎤ ⎡ 6 −2 ⎤ ⎡U1 ⎤ ⎡ 0 ⎤
⎢ 0 1 ⎥ ⎢ ⎥ + ⎢ −2 4 ⎥ ⎢U ⎥ = ⎢10 ⎥
⎣ ⎦ ⎣U 2 ⎦ ⎣ ⎦⎣ 2⎦ ⎣ ⎦
1 1 1
a0 = , a1 = , a2 = 2a0 , a3 =
Δt 2 2Δt a2
1 1
a0 = = 12.8, a1 = = 1.79,
(0.28) 2
2 ⋅ 0.28 For Δt = 0.28
1 1
a2 = 2 ⋅ = 25.5, a3 = = 0.0392
(0.28) 2 25.5
−Δt ⎡0 ⎤ ⎡0 ⎤ ⎡0⎤ ⎡ 0 ⎤
Ui = ⎢ ⎥ − 0.28
0 28 ⎢ ⎥ + 00.0392
0392 ⎢ ⎥ = ⎢ ⎥
0
⎣ ⎦ 0
⎣ ⎦ 10
⎣ ⎦ ⎣ 0.0392 ⎦
ˆ ⎡2 0⎤ ⎡0 0 ⎤ ⎡ 25.5 0 ⎤
M = 12.8
12 8 ⎢ ⎥ + 1.79
1 79 ⎢ ⎥ =⎢ ⎥
⎣ 0 1 ⎦ ⎣ 0 0 ⎦ ⎣ 0 12.8 ⎦
ˆ ⎡ 0 ⎤ ⎡ 45.0 2 ⎤t ⎡ 25.5 0 ⎤ t −Δt
t
R =⎢ ⎥+⎢ ⎥ U−⎢ ⎥ U
⎣10 ⎦ ⎣ 2 21.5⎦ ⎣ 0 12.8⎦
Di t IIntegration
Direct t ti Methods
M th d
⎡ 2 0 ⎤ ⎡U1 ⎤ ⎡ 6 −2 ⎤ ⎡U1 ⎤ ⎡ 0 ⎤
⎢ 0 1 ⎥ ⎢ ⎥ + ⎢ −2 4 ⎥ ⎢U ⎥ = ⎢10 ⎥
⎣ ⎦ ⎣U 2 ⎦ ⎣ ⎦⎣ 2⎦ ⎣ ⎦
The equation which must be solved for each time step is:
⎡ 25.5 0 ⎤ ⎡ U1 ⎤ t ˆ
t +Δt
⎢ 0 12.8⎥ ⎢ t +Δt ⎥ = R
⎣ ⎦ ⎢⎣ U 2 ⎥⎦
5 0 ⎤ ⎡ U1 ⎤
t −Δt
t ˆ ⎡ 0 ⎤ ⎡ 45
45.0
0 2 ⎤ ⎡ t
U1 ⎤ ⎡ 25
25.5
R =⎢ ⎥+⎢ ⎥ ⎢t ⎥ − ⎢ ⎥ ⎢ t −Δt ⎥
10
⎣ ⎦ ⎣ 2 21.5 ⎦ ⎣ U2 ⎦ ⎣ 0 12.8 ⎦ ⎢⎣ U 2 ⎥⎦

k1 = 4
Di t IIntegration
Direct t ti Methods
M th d m1 = 2 R1 = 0
U 1 ,U1 ,U1 k2 = 2

m2 = 1 R2 = 10
U 2 ,U 2 ,U2
The results are: k3 = 2
CentralDifference_U1 For Δt=28

5
CentralDifference_U2 the solution increase
4
steadily
3
0
0 0.5 1 1.5 2 2.5 3 3.5 4

Di t IIntegration
Direct t ti Methods
M th d
+ CU
MU + KU = R c
• Th Houbolt
The H b l method
h d
t +Δt = 1 (2 t +Δt U − 5 t U + 4 t −Δt U − t − 2 Δt U)
U a
Δt 2
t +Δt 1
U= (11t +Δt U − 18 t U + 9 t −Δt U − 2 t − 2 Δt U) b
6Δt
a and b inserted in c
⎛ 2 12 ⎞ t +Δt
⎜ 2 M + C + K ⎟ U=
⎝ Δt 6Δt ⎠
t +Δt ⎛ 5 3 ⎞ ⎛ 4 3 ⎞ t −Δt ⎛ 1 1 ⎞ t − 2 Δt
R + ⎜ 2 M + C⎟ t U − ⎜ 2 M + C⎟ U + ⎜ 2 M + C⎟ U
⎝ Δt Δt ⎠ ⎝ Δt 2Δt ⎠ ⎝ Δt 3Δt ⎠

Di t IIntegration
Direct t ti Methods
M th d
• Th Houbolt
The H b l method
h d
We will not consider the Houbolt in more detail – however it is

noted that it is necessary to factorize the stiffness matrix (implicit
method)
Furthermore, if the mass and damping terms are neglected, the

Houbolt method results in the static analysis equations
⎛ 2 12 ⎞ t +Δt
⎜ 2 M + C + K ⎟ U=
⎝ Δ t 6 Δt ⎠
t +Δt ⎛ 5 3 ⎞ ⎛ 4 3 ⎞ t −Δt ⎛ 1 1 ⎞ t − 2 Δt
R + ⎜ 2 M + C⎟ t U − ⎜ 2 M + C⎟ U + ⎜ 2 M + C⎟ U
⎝ Δt Δt ⎠ ⎝ Δt 2Δt ⎠ ⎝ Δt 3Δt ⎠

Di t IIntegration
Direct t ti Methods
M th d
• Th Wilson
The Wil θ method
h d
In this method the acceleration is assumed to vary linearly from

time t to t + Δt
t +τ
U + τ ( t +θΔt U
= t U )
− t U
θΔt
By integration we obtain
τ 2
t +τ = U
U + U
t τ + t
2θΔt
( t +θΔt )
U− tU
1 t 2 1 3 t +θΔt t
t +τ
U = U + Uτ + Uτ +
t t
τ ( U − U)
2 6θΔt

Di t IIntegration
Direct t ti Methods
M th d
• Th Wilson
The Wil θ method
h d
Setting τ = θ Δt we get
t +θΔt
U + θΔt ( t +θΔt U
= tU )
+ t U
2
t +θΔt θΔt + 1 (θΔt )2 ( t +θΔt U

U = tU+ tU )
+ 2 t U
6
from which we can solve
6 6 t
t +θΔt =
U ( t +θΔt
U − t U) −
U − 2t U
(θΔt ) θΔt
2
t +θΔt
U − θΔt t U
= 3 ( t +θΔt U − t U ) − 2 t U
θΔt 2
Di t IIntegration
Direct t ti Methods
M th d
6 6 t
t +θΔt =
U ( t +θΔt
U − t U) −
U − 2t U
(θΔt ) θΔt
2
• Th Wilson
The Wil θ method
h d t +θΔt
U − θΔt t U
= 3 ( t +θΔt U − t U ) − 2 t U
θΔt 2
We now solve for the displacements, velocities and accelerations

by inserting into the dynamic equilibrium equation
+ C t +θΔt U
M t +θΔt U + K t +θΔt U = t +θΔt
R
t +θΔt
R = t R + θ ( t +Δt R − t R )

Di t IIntegration
Direct t ti Methods
M th d
• The Wilson θ method : Solution procedure:

2) Initialize 0 U, 0 U and 0 U

3) Select time step Δt and calculate integration constants θ = 1.4
6 3 θΔt a0
a0 = , a1 = , a2 = 2a1 , a3 = , a4 =
(θΔt ) θΔt θ
2
2
−a2 3 Δt Δt 2
a5 = , a6 = 1 − , a7 = , a8 =
θ θ 2 6
4) Form effective stiffness matrix ˆ =K+a M+aC
K 0 1
5) Triangularize K ˆ = LDLT
Implicit procedure !

Di t IIntegration
Direct t ti Methods
M th d
• Th Wilson
The Wil θ method
h d : Solution
S l i procedure:
d
1) Calculate effective loads at time t+∆t:

t +θΔt + 2t U
ˆ = t R + θ ( t +θΔt R − t R ) + M (a t U + a t U
R )
0 2
+C(a t U + 2 t U +a tU )
1 3

LDLT t +θΔt U = t +θΔt ˆ
R
3) Solve for the corresponding velocities and accelerations
t +Δt = a ( t +θΔt U − t U) + a t U
U +a tU
4 5 6
t +Δt + a ( t +Δt U + t U
)
U = tU 7
t +Δt
U = t U + Δt t U + a ( t +Δt U + 2 t U
)
8
Di t IIntegration
Direct t ti Methods
M th d
• Th Newmark
The N k method
h d
This method may be seen as an extension of the Wilson θ method
t +Δt = tU
U + ⎡(1 − δ ) t U
+ δ t +Δt U
⎤ Δt
⎣ ⎦
t +Δt Δt + ⎡( 1 − α ) t U
U = tU+ tU ⎤ Δt 2
+ α t +Δt U
⎢⎣ 2 ⎥⎦
δ and α are parameters which may be adjusted to achieve

accuracy and stability
δ =0.5, α = 1/6 corresponds to the linear acceleration method which

also correspond to the Wilson θ method with θ = 1
Newmark
Ne ma k originally
o iginall proposed
p oposed δ =0.5,
0 5 α = 1/4 which
hich results
es lts in an
unconditionally stable scheme (the trapetzoidal rule)
Di t IIntegration
Direct t ti Methods
M th d
• Th Newmark
The N k method
h d
We now solve for the displacements, velocities and accelerations

by inserting into the dynamic equilibrium equation
t +Δt + ⎡(1 − δ ) t U
+ δ t +Δt U
⎤ Δt
U = tU ⎣ ⎦
+ C t +Δt U
M t +Δt U + K t +Δt U = t +Δt R
t ⎡ 1 ⎤ Δt 2
+ α t +Δt U
t +Δt
U = U + UΔt + ⎢( − α ) t U
t
⎥⎦
⎣ 2

Di t IIntegration
Direct t ti Methods
M th d
• The Newmark method : Solution procedure:

2) Initialize 0 U, 0 U and 0 U

3) Select time step Δt and parameters α and δ
1 δ 1 1 δ
a0 = , a1 = , a2 = , a3 = − 1, a4 = −1
αΔt 2 αΔt αΔt 2α 2α
δ Δt δ
a5 = − 1,
1 a6 = 2) a7 = Δt (1 − δ ),
( − 2), ) a8 = δΔt
α 2 α
4) Form effective stiffness matrix ˆ =K+a M+aC
K 0 1
5) Triangularize K ˆ = LDLT
Implicit procedure !

Di t IIntegration
Direct t ti Methods
M th d
• The Newmark method : Solution procedure:
1) Calculate effective loads at time t:

t +Δt ˆ +a tU
R = t +Δt R + M (a0 t U + a2 t U )
3
+C(a t U + a t U +a tU )
1 4 5

ˆ
LDLT t +Δt U = t +Δt R
3) Solve for the corresponding velocities and accelerations

t +Δt
U = a ( t +Δt U − t U) − a t U −a tU

0 2 3
Δt
t +Δ + a t +ΔΔt U + a t U

U = tU 7 6

k1 = 4
Di t IIntegration
Direct t ti Methods
M th d m1 = 2 R1 = 0
U 1 ,U1 ,U1 k2 = 2
m2 = 1 R2 = 10
• The Newmark method : Example: U 2 ,U 2 ,U2

k3 = 2
6 Newmark_U1
Newmark_U2
5
CentralDifference_U1
4 CentralDifference_U2
0
0 0.5 1 1.5 2 2.5 3 3.5 4

k1 = 4
Di t IIntegration
Direct t ti Methods
M th d m1 = 2 R1 = 0
U 1 ,U1 ,U1 k2 = 2
m2 = 1 R2 = 10
• The Newmark method : Example: U 2 ,U 2 ,U2

k3 = 2
7
6 Newmark_U1
_ Δt = 28
Newmark_U2
= ⎡ 0 ⎤
5
0
4
U ⎢10 ⎥
⎣ ⎦
3
0
0 05
0.5 1 15
1.5 2 25
2.5 3 35
3.5 4
3.5
3 Newmark_U1
2.5
Newmark_U2 Δt = 28
= ⎡0 ⎤
2
0
U ⎢0 ⎥
1.5
⎣ ⎦
1
0.5
0
Method of Finite
0 Elements
0.5 II 1 1.5 2 2.5 3 3.5 4
Di t IIntegration
Direct t ti Methods
M th d
• C
Coupling
li off integration
i i operators
For some problems it may be an advantage to combine the

different types of integration schemes – e.g. if a structure is
subjected to dynamic load effect from hydrodynamic loading then
the analysis
y of the hydrodynamic
y y forces mayy be assessed usingg an
explicit scheme and the structural response by using an implicit
scheme.
The best choice of strategy will depend on the problem in

regard stability and accuracy!

The Finite Element Method for Dynamic Analysis

Загружено:

Сведения о документе

Исходное описание:

Оригинальное название

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

The Finite Element Method for Dynamic Analysis

Загружено:

Авторское право:

Доступные форматы

Swiss Federal Institute of Technology Page 1

The Finite Element Method

Prof. Dr. Michael Havbro Faber

Method of Finite Elements II

Direct Integration methods

- The central difference method

- The Houbolt method

- The Wilson θ method

- The Newmark method

- Coupling of integration operators

Method of Finite Elements II

The very basics ☺ k Unstretched p

x = A sin ωnt + B cos ωn t ⇒ t x = 0 x sin ωnt + 0 x cos ωnt

We have previously considered the equilibrium equations

Whether a dynamic analysis is needed or not is generally up to

Method of Finite Elements II

In principle the equilibrium equations may be solved by any

Efficiency – numerical efforts – must be considered and it is

Method of Finite Elements II

Direct means: The equations are solved in their original form!

Two ideas are utilized

1) The equilibrium equations are satisfied only at time steps,

2) A particular variation of displacements, velocities and

The accuracy depends on these assumptions as well as the

Method of Finite Elements II

The displacements, velocities and accelerations

Method of Finite Elements II

We distinguish principally between

Solution is based on the equilibrium equations at time t

Method of Finite Elements II

Method of Finite Elements II

We see that we do not need to factorize the stiffness matrix ☺

We also see that in order to calculate the displacements

Method of Finite Elements II

1) Form stiffness matrix, mass matrix and damping matrix

Method of Finite Elements II

B: For each time step

1) Calculate effective loads at time t:

2) Solve for the displacements U at time t + Δt

3) If required, solve for the corresponding velocities and

• The Central difference method

The effectiveness of the central difference method depends on the

• The Central difference method: Example:

With natural periods:

Method of Finite Elements II

⎡ 2 0 ⎤ ⎡U1 ⎤ ⎡ 6 −2 ⎤ ⎡U1 ⎤ ⎡ 0 ⎤ With Eigen periods:

Then we continue with the steps ☺

Method of Finite Elements II

Method of Finite Elements II

• The Central difference method: Example:

CentralDifference_U1 For Δt=28

Method of Finite Elements II

Method of Finite Elements II

We will not consider the Houbolt in more detail – however it is

Furthermore, if the mass and damping terms are neglected, the

Method of Finite Elements II

In this method the acceleration is assumed to vary linearly from

Method of Finite Elements II

t +θΔt  θΔt + 1 (θΔt )2 ( t +θΔt U

We now solve for the displacements, velocities and accelerations

Method of Finite Elements II

• The Wilson θ method : Solution procedure:

x = A sin ωnt + B cos ωn t ⇒ t x = 0 x sin ωnt + 0 x cos ωnt

⎡ 2 0 ⎤ ⎡U1 ⎤ ⎡ 6 −2 ⎤ ⎡U1 ⎤ ⎡ 0 ⎤ With Eigen periods:

t +θΔt θΔt + 1 (θΔt )2 ( t +θΔt U

• The Newmark method : Example: U 2 ,U 2 ,U2

• The Newmark method : Example: U 2 ,U 2 ,U2