FEM UNIT-V

III B Tech – II Semester

DepartmentofMechanicalEngineering

Vignan’sInstituteofInformationTechnology(A)

(Approved by AICTE and Permanently Affiliated to JNTU,Kakinada)

BesideVSEZ, Duvvada,Visakhapatnam – 530049

 Unit V
Concepts covered: Steady state heat transfer, element conductivity matrix derivation.

Dynamic analysis Modal analysis - Natural frequency and mode shapes

Harmonic analysis - Forced response ofsystem to a sinusoidal forcing

Transient analysis - Forced response for non-harmonic loads (impact, step or ramp forcing etc.)

DYNAMIC CONSIDERATIONS Static analysis holds when the loads are slowly applied.
When the loads are suddenly applied, or when the loads are of a variable nature, the mass and
acceleration effects come into the picture. If a solid body, such as an engineering structure, is
deformed elastically and suddenly released, it tends to vibrate about its equilibrium position.
This periodic motion due to the restoring strain energy is called free vibration.

The number of cycles per unit time is called frequency. The maximum displacement from the
equilibrium position is the amplitude.

FORMULATION: We define the Lagrangean by L = T -П where T is the kinetic energy and П is

the potential energy

Write a short notes on steady state heat transfer.

A steady state heat transfer analysis is used to determine the steady state temperature distribution
and heat flow. This type of analysis can be performed when the temperature at any point within
the model including the surfaces is independent of time. Because the analysis is steady state,
there must be a source of heat and a sink of heat. If there is only a source and no means of
removing heat the entire model should heat up to a constant temperature, but this leads to
unstable state.

Define conduction heat transfer.

Ans: Conduction heat transfer takes place in a body by the virtue of temperature difference.
Conduction is the only mode of heat transfer that takes place in a solid body which is influenced
by other mode of heat transfer at the boundary surfaces.

Write about one dimensional steady state conduction.

Conduction heat transfer takes place in a body by the virtue of temperature difference.
Conduction is the only mode of heat transfer that takes place in a solid body which is influenced
by other mode of heat transfer at the boundary surfaces.

Temperature distribution in a body is very important in many problems in engineering

practice. If a heated body is not permitted to expand or contract freely in all directions, some
stresses are developed within the body. The magnitude of this thermal stresses will influence the
design of various equipments like steam boilers, gas turbines, nuclear reactors, heat exchangers,
rocket motors etc.

The first step in the determination of thermal stresses is to determine the temperature
distribution with in the body.

Q – Internal heat generation per unit volume

Q – critical heat flux

A – area perpendicular to the direction of heat transfer

Fig 6.1 Composite wall through which heat transfer takes place
𝜕𝑞
q . A + Q . A dx = (q + 𝜕𝑥 . dx ) A

𝜕𝑞
q . A + Q . A dx = q . A + 𝜕𝑥 . dx A

𝜕𝑞
Q = 𝜕𝑥
 𝑑𝑇
We know that critical heat flux q = - K 𝑑𝑥

𝑑𝑇
𝜕(− K )
𝑑𝑥
Q= 𝜕𝑥

𝑑2𝑇
Q=-K 𝑑𝑥 2

𝑑2𝑇
Q+K =0
𝑑𝑥 2

Element conductivity matrix

Fig 6.2 Bar elemnt

T (x) = a0 + a1 x ----------- (1)

𝑎0
T (x) = [ 1 x ]
𝑎1
Applying boundary conditions

At x = 0 ; T(x) = T1

At x = L ; T(x) = T2

T1 = a0

T2 = a0 + a1 L
𝑇2−𝑇1
a1 = ( ) ----------(2)
𝐿

𝑇2−𝑇1
T(x) = T1 + ( )x
𝐿
 𝑥 𝑇2
T(x) = T1 (1 – 𝐿 ) + x
𝐿

T(x) = N1 T1 + N2 T2 ---------(3)

𝑥 𝑥 𝑇1
T(x) = [(1 – 𝐿 ) ]
𝐿 𝑇2
𝑑𝑇 𝑑𝑁1 𝑑𝑁2
= T1 + T2
𝑑𝑥 𝑑𝑥 𝑑𝑥

𝑑𝑇 −1 1 𝑇1
= { }
𝑑𝑥 𝐿 𝐿 𝑇2
𝑑𝑇
=[B]{q}
𝑑𝑥

Element conductivity matrix

𝐿
[ KT ] = 0
[ B ]T. K . [ B ] dx A

[ KT ] = [ B ]T. K . [ B ] L . A

[ KT ] = K A L [ B ]T [ B ]
−1
𝐿 −1 1
[ KT ] = K A L 1 𝐿 𝐿
𝐿

𝐾𝐴 1 −1
[ KT ] = 𝐿 −1 1
According to fourier law of conduction
𝐾𝐴
q= T
𝐿

𝑞 𝐾
Q=𝐴= T
𝐿

𝑄1 𝐾 1 −1 𝑇1
=𝐿
𝑄2 −1 1 𝑇2

Class work:

1. Define the word heat transfer. Explain the heat transfer in conduction.
2. Explain about steady state heat transfer.

3.Derivation of element conductivity matrix for stady state heat transfer.

Home work:

1. List out the modes of heat transfer. Explain

2. What is meant by steady state heat transfer.

Derive the temperature distribution in a fin.

Consider a fin of one dimensional heat flow

Fig. 6.3 A fin fixed at one end

Let,

A – Area of cross section of the fin

K – Thermal conductivity of fin material

h – convective heat transfer coefficient

Heat entering into the element at a distance x

From fourier law of conduction

 𝑑𝑇
Qx = - k A 𝑑𝑥

𝜕 𝑄𝑥
Qx+dx = Qx + . dx
𝜕𝑥

Heat transfer by convection Qc = h Af ( T – Ta )

Where,

T – surface temperature of fin

Ta – Ambient temperature

Af – Profile area

Qx – Qc = Qx+dx
𝜕 𝑄𝑥
Qx – Qc = Qx + . dx
𝜕𝑥

𝜕 𝑄𝑥
– Qc = . dx
𝜕𝑥

𝑑𝑇
𝜕(− K A )
- h Af ( T – Ta ) = 𝑑𝑥
. dx
𝜕𝑥

𝑑𝑇
𝜕( )
- h ( dx ) ( T – Ta ) = - K A 𝑑𝑥
dx
𝜕𝑥

h  ( T – Ta ) = K A (d2T/dx2)

K A (d2T/dx2) - h  ( T – Ta ) = 0

A generalized differential equation for minimum energy available within the element is given by
1 𝐿 𝑑𝑇 2 ℎ
I=2 0
[ (𝑑𝑥 ) + 𝐾𝐴 (T2 – 2 T Ta) ] dx

Let us assume temperature is varying linearly within the element

T(x) = a1 + a2x -----------(1)

𝑎0
T (x) = [ 1 x ]
𝑎1
Applying boundary conditions
At x = 0 ; T(x) = T1

At x = L ; T(x) = T2

T1 = a0

T2 = a0 + a1 L
𝑇2−𝑇1
a1 = ( ) ----------(2)
𝐿

𝑇2−𝑇1
T(x) = T1 + ( )x
𝐿

𝑥 𝑇2
T(x) = T1 (1 – 𝐿 ) + x
𝐿

T(x) = N1 T1 + N2 T2 ---------(3)
𝑑𝑇
Minimum energy available within if ; = 0;
𝑑𝑇𝑖

1 𝐿 𝑑𝑇 ℎ
I= [ ( 𝑑𝑥 )2 + (T2 – 2 T Ta) ] dx
2 0 𝐾𝐴

1 𝐿 ℎ
= [ (N1’T1 +N2’T2 )2 + [(N1T1+N2T2)2 – 2 (N1T1 + N2T2) Ta ] dx
2 0 𝐾𝐴

𝑑𝑇 1 𝐿 ℎ
= [ 2(N1’T1 +N2’T2 ) + [2(N1T1+N2T2) – 2 N1 Ta ] dx
𝑑𝑇1 2 0 𝐾𝐴

𝐿 −𝑇1 𝑇2 ℎ 𝑥 𝑥 𝑥
333= 0
[ ( + )+ 𝑇1 1 − 𝐿 + 𝑇2 − 1 − 𝐿 𝑇𝑎 ]𝑑𝑥
𝐿1 𝐿 𝐾𝐴 𝐿

On simplification we get
1 ℎ 𝐿 ℎ 𝑇𝑎 𝐿
𝑇1 − 𝑇2 + 2𝑇1 + 𝑇2 − =0 ----------(A)
𝐿 6𝐾𝐴 2𝐾𝐴

Similarly
𝑑𝑇 1 𝐿 ℎ
= [ 2(N1’T1 +N2’T2 ) + [2(N1T1+N2T2) – 2 N2 Ta ] dx
𝑑𝑇1 2 0 𝐾𝐴

1 ℎ 𝐿 ℎ 𝑇𝑎 𝐿
– 𝑇1 + 𝑇2 + 𝑇1 + 2𝑇2 − =0 -----------(B)
𝐿 6𝐾𝐴 2𝐾𝐴

Adding equations (1) & (2) we get

1 1 −1 𝑇1 ℎ 𝐿 2 1 𝑇1 ℎ 𝑇𝑎 𝐿 1
+ =
𝐿 −1 1 𝑇2 6𝐾𝐴 1 2 𝑇2 2𝐾𝐴 1

1 1 −1 ℎ 𝐿 2 1 𝑇1 ℎ 𝑇𝑎 𝐿 1
{𝐿 + } =
−1 1 6𝐾𝐴 1 2 𝑇2 2𝐾𝐴 1
[K][T]=[P]

{P}=[K][T]

2.Derive the expression of temperature distribution when the shaft is subjected to torsion .

Ans:
𝑇 𝐺𝜃
=
𝐽 𝐿

𝑇 𝐺𝐽
=
𝜃 𝐿

𝑇
Kt = 𝜃 = Torsional stiffness N-m/rad.

Fig 6.4 Shaft subjected to torsion

As the angle of twist is constant, the total angle of twist of an element can be expressed in terms
of nodal rotations and twisting moments

Let us assume 2 > 1

Net twist = 2 - 1
 𝑇𝐿
2 - 1 = 𝐺 𝐽

𝐺𝐽
T= (2 − 1 )
𝐿

T = Kt((2 − 1 )

For equilibrium condition T1+T2 = 0

T1 = - T2

T1 = - Kt (2 − 1 )

T2 = Kt ((2 − 1 )

𝑇1 𝐾𝑡 −𝐾𝑡 1
=
𝑇2 −𝐾𝑡 𝐾𝑡 2
𝑇1 1 −1 1
= 𝐾𝑡 Where 1, 2 are nodal rotations.
𝑇2 −1 1 2
Class work:

1.What is a shaft .

2.Explain the temperature distribution in a fin.

3.Derive the expression of temperature distribution when the shaft is subjected to torsion .

Home work:

1. Explain the temperature distribution in a fin.

2. Derive the expression of temperature distribution when the shaft is subjected to torsion .

Problems:

1.Estimate the temperature distribution in a fin whose cross section is 10*10mm and 500
mm long.Take K = 50w/mk and h= 75w/m2k at 25c. The base temperature is assumed to
be constant and its value may be taken as 900c . Also calculate the heat transfer rate.
 Fig 6.6 A bar element fixed at one end

Ans:

Given data:

K = 50w/mk

h= 75w/m2k

Ta = 25c

T = 900c

Area = 100mm

We know that

{P}=[K]{T} ---------(1)
ℎ 𝑇𝑎 𝐿 1 1 1 −1 ℎ 𝐿 2 1 𝑇1
={𝐿 + }
2𝐾𝐴 1 −1 1 6𝐾𝐴 1 2 𝑇2
1 1 −1 ℎ 𝐿 2 1
[K]=𝐿 +
−1 1 6𝐾𝐴 1 2
1 1 −1 75∗4∗0.01∗0.25 2 1
[ K1 ] = 0.25 +
−1 1 6∗50∗0.01∗0.01 1 2
4 −4 2 1
[ K1 ] = + 25 ∗
−4 4 1 2
54 21
[ K1 ] =
21 54
54 21
[ K2 ] =
21 54
[ K ] = [ K1 ] + [ k2 ]

54 21 0
[ K ] = 21 108 21
0 21 54
1
ℎ 𝑇𝑎 𝐿
[P]= 2𝐾𝐴
2
1
1 1875
75∗4∗0.01∗25∗0.25
[P]= 2∗50∗0.01∗0.01
2 = 3750
1 1875
1875 54 21 0 𝑇1
3750 = 21 108 21 𝑇2
1875 0 21 54 𝑇3
1875 = 54 T1 + 21 T2 ---------(1)

3750 = 21 T1 + 108 T2 + 21 T3 ---------(2)

1875 = 21 T2 + 54 T3 -----------(3)

Simplifying above equations we get

T2 = 98.94c
T3 = 37.32c

2. A composite wall consists of 3 materials as shown in fig.The outer temperature is at 20c

. Convection heat transfer takes place on the inner surface of the wall . Determine the
temperature distribution in the wall.Where K1 = 20w/mc , K2 = 30w/mc , K3 = 50w/mc
and h = 25w/m2c .

Fig 6.7

Ans:

Given data:

K1 = 20w/mc , K2 = 30w/mc , K3 = 50w/mc and h = 25w/m2c .

T = 800c T0 = 20c
Element connectivity table

Element 1 1 2
Element 2 2 3
Element 3 3 4
Conductivity matrix for element 1

𝐾1𝐴1 1 −1
[ KT ]1 = 𝐿1 −1 1
20∗1 1 −1 66.67 −66.67
[ KT ]1 = =
0.3 −1 1 −66.67 66.67
Conductivity matrix for element 2

𝐾2𝐴2 1 −1
[ KT ]2 = 𝐿2 −1 1
30∗1 1 −1 200 −200
[ KT ]2 = =
0.15 −1 1 −200 200

Conductivity matrix for element 3

𝐾3𝐴3 1 −1
[ KT ]3 = 𝐿3 −1 1
 50∗1 1 −1 333.33 −333.33
[ KT ]3 = =
0.15 −1 1 −333.33 333.33
Global conductivity matrix

[ KT ] = [ KT ]1 + [ KT ]2 + [ KT ]3

66.67 −66.67 0 0
−66.67 266.67 −200 0
[ KT ] =
0 −200 533.33 −333.33
0 0 −333.33 333.33
{ q } = [ KT ] [ T ]

25 ∗ 800 66.67 −66.67 0 0 T1

0 −66.67 266.67 −200 0 T2
=
0 0 −200 533.33 −333.33 T3
0 0 0 −333.33 333.33 20

Remove 4th row and 4th column C44*T4 is placed in q3

25 ∗ 800 66.67 −66.67 0 T1

0 = −66.67 266.67 −200 T2
333.33 ∗ 20 0 −200 533.33 T3
20000 = 66.67 T1 – 66.67 T2 ---------(1)

0 = -66.67T1+266.67T2-200T3 ----------(2)

333.33*20 = -200T2+533.33T3 ----------(3)

Solving the above equations we get

T1 = 180C

T2 = 480C

T3 = 80C

Class work:

1.Problems on steady state hear transfer through fin and shaft.

Home work

1.Find the temperature distribution in a fin of radius 1cm.Take h = 10W/cm2c, k = 70w/cmc,

T1 = 140c, L = 5cm.

Fig 6.7

Concepts covered : Dynamic analysis

EVALUATION OF EIGENVALUES AND EIGENVECTORS We observe here that K and M

are symmetric matrices. Further, K is positive definite for properly constrained problems.
Properties of Eigenvectors For a positive definite symmetric stiffness matrix of size n, there are n
real eigenvalues and corresponding eigenvectors.

The eigenvalues may be arranged in ascending order: If U1, U2…Un are the corresponding
eigenvectors, we have The eigenvectors possess the property of being orthogonal with respect to
both the stiffness and mass matrices The lengths of eigenvectors are generally normalized so that
The foregoing normalization of the eigenvectors leads to the relation

EIGENVALUE – EIGENVECTOR EVALUATION The eigenvalue-eigenvector evaluation

procedures fall into the following basic categories:

1. Characteristic polynomial technique

2. Vector iteration methods

3. Transformation methods Characteristic polynomial If the eigenvector is to be nontrivial, the

required condition is This represents the characteristic polynomial in λ

Explain about dynamic analysis in FEM .

If Generated stresses from external load are passing yield strength so your simulations are
categorized as dynamic FEA .Deformation or change in shape of the model because the exposed
loads are not neglected so continuously effects the model .This analysis is used to study the
behavior of the model under load conditions whether collisions occur or crack propagation etc.
To decide whether it is static or dynamic analysis we use stress strain diagram when the analysis
is performed. Dynamic analysis is performed by using software’s Ansys LS - DYNA etc.

Fig 6.9 model undergoing dynamic analysis

m x`` = mg – k(dt + x)

m x`` = mg – kdt – kx (since mg = 0 and kdt = 0)

m x`` = - kx
𝑑2𝑥
but x`` = 𝑑𝑡2

m x`` + kx = 0

x`` +kx/m = 0

x2 = k/m

x = 𝑘/𝑚 --- circular natural frequency

 Fig 6.10

m1x1`` = - k1 x1 + k2 ( x2 – x1 )

m2x2`` = - k2( x2 – x1)

m1x1`` + k1x1 – k2(x2 – x1 ) = 0

m2x2`` + k2(x2 – x1 ) = 0

Writing the above equations in matrix form we get

𝑚1 0 𝑥1`` 𝑘1 + 𝑘2 −𝑘2 𝑥1
+ =0
0 𝑚2 𝑥2`` −𝑘2 𝑘2 𝑥2
[ m ] { x`` } + [ K ] { x } = 0

2.Explain about lumped mass formulation and derive the expression of lumped mass
formulation.

Ans:

 Lumped mass technique is easy to handle. Therefore the natural frequencies obtained
from lumped mass systems are lower than the exact value.

 In this technique the total mass in either direction is distributed equally at the nodes of the
element and masses are associated with translation degree of freedom only.
Consider a truss bar element of length l, modulus of elasticity E, cross section area A and mass
density 

Fig 6.11 Truss bar element

We know that { F } = [ K ] { u }

𝐹1 𝐴𝐸 1 −1 𝑢1
=
𝐹2 𝐿 −1 1 𝑢2
AL
𝐹1 − ∗ 𝑢1`` 𝐴𝐸 1 −1 𝑢1
2
AL =
𝐹2 − ∗ 𝑢2`` 𝐿 −1 1 𝑢2
2

𝐹1 𝐴𝐸 1 −1 𝑢1 𝐴𝐿 1 0 𝑢1``
= +
𝐹2 𝐿 −1 1 𝑢2 2 0 1 𝑢2``
Lumped element mass matrix

𝐴𝐿 1 0 𝑢1``
2 0 1 𝑢2``
Lumped mass formulation

𝐹1 𝐴𝐸 1 −1 𝑢1 𝐴𝐿 1 0 𝑢1``
= + 2
𝐹2 𝐿 −1 1 𝑢2 0 1 𝑢2``
Derive the expression for consistent mass formulation.

We can formulate mass matrix by the use of Lagrange’s formula with the strain energy and
kinetic energy expressions written in terms of displacement functions. If the displacement
functions used in deriving the mass matrix are same as those used in deriving stiffness matrix.
The mass matrix is called as consistent mass matrix.

Lagrange’s formula
𝜕𝑇
𝑑 𝜕𝑈
𝜕𝑢𝑖
+ 𝜕𝑢𝑖 = Fi
𝑑𝑡

Where,

U – strain energy

T – kinetic energy

We know that

u(x) = N1u1 + N2u2

𝑥 𝑥
N1 = 1 - 𝐿 ; N2 = 𝐿

𝐸𝐴 𝐿
Strain energy U = 0
[ 𝑢` 𝑥 ]2 dx
2

𝐴 𝐿
Kinetic energy T = 2 0
[ 𝑢 𝑥 ]2 dx

u(x) = N1u1 + N2u2

u`(x) = N1`u1 + N2`u2

𝜕𝑇 𝜕 𝐴 𝐿
= 𝜕𝑢𝑖 ( 0
[ 𝑢′ ]2 dx )
𝜕𝑢𝑖 2

𝜕 𝐴 𝐿
= 𝜕𝑢𝑖 ( 0
[𝑁1𝑢′ + 𝑁2𝑢2′ ]2 dx )
2

𝐴 𝐿
= 0
2 𝑁1𝑢1′ + 𝑁2𝑢′ 𝑁1dx
2

𝜕𝑇 𝐿
= 𝐴 0
𝑁1𝑢1′ + 𝑁2𝑢2′ 𝑁1dx
𝜕𝑢𝑖
 𝜕𝑇
𝑑 𝐿
𝜕𝑢𝑖
= 𝐴 0
𝑁1𝑢1′′ + 𝑁2𝑢2′′ 𝑁1dx
𝑑𝑡

𝜕𝑈 𝜕 𝐸𝐴 𝐿
= 𝜕𝑢𝑖 ( 2 0
[ 𝑢′ 𝑥 ]2 dx )
𝜕𝑢𝑖

𝜕𝑈 𝜕 𝐸𝐴 𝐿
= 𝜕𝑢𝑖 ( 2 0
[ 𝑁1′𝑢1 + 𝑁2′𝑢2 ]2 dx )
𝜕𝑢𝑖

𝐸𝐴 𝐿
= 0
2 𝑁1′ 𝑢1 + 𝑁2′ 𝑢2 𝑁1′ dx
2

𝜕𝑈 𝐿
= 𝐸𝐴 0
[ 𝑁′𝑢1 + 𝑁2′𝑢2 ] N1’ dx
𝜕𝑢𝑖

𝜕𝑇
𝑑 𝜕𝑈
𝜕𝑢𝑖
+ 𝜕𝑢𝑖 = Fi
𝑑𝑡

𝐿 𝐿
{ 𝐴 0
𝑁1𝑢1′′ + 𝑁2𝑢2′′ 𝑁1 𝑑𝑥 } + { 𝐸𝐴 0
[ 𝑁′𝑢1 + 𝑁2′𝑢2 ] 𝑁1′ dx } = Fi

𝐿 𝐿 𝐿 𝐿
𝑢1′′ 𝑢1
𝐴 𝑁1. 𝑁1𝑑𝑥 𝐴 𝑁1. 𝑁2 𝑑𝑥 + 𝐸𝐴 𝑁1′. 𝑁1′𝑑𝑥 𝐸𝐴 𝑁1′. 𝑁2′𝑑𝑥
0 0 𝑢2′′ 0 0 𝑢2
= 𝐹𝑖

Where,
𝐿 𝐿 𝑥 𝑥
0
𝑁1. 𝑁1𝑑𝑥 = 0
1 − (1 − 𝐿 )𝑑𝑥
𝐿

𝐿 𝑥
= 0
( 1- )2
𝐿

𝐿 𝑥 𝑥 𝐿
= 0
( 1+(𝐿 )2 - 2𝐿 ) dx = 3

𝐿 𝐿
0
𝑁1. 𝑁1𝑑𝑥 =3

𝐿 𝐿
0
𝑁1. 𝑁2𝑑𝑥 =6

𝐿 1
0
𝑁1′. 𝑁1′𝑑𝑥 =𝐿

𝐿 −1
0
𝑁1′. 𝑁2′𝑑𝑥 = 𝐿

Substituting these values in above equation we get

 𝐹1 𝐴𝐸 1 −1 𝑢1 𝐴𝐿 2 1 𝑢1′′
= + 6
𝐹2 𝐿 −1 1 𝑢2 1 2 𝑢2′′

Let,

u1 = u1∗ sin(𝑡)

U1’’ = - 2 . u1sin(𝑡) = - 2u1

U2 = u2 *sin(𝑡)

U1’’ = - 2 . u2sin(𝑡) = - 2u2

𝐹1 𝐴𝐸 1 −1 𝑢1 𝐴𝐿 2 1 − 2u1
= + 6
𝐹2 𝐿 −1 1 𝑢2 1 2 − 2u2
𝐹1 𝐴𝐸 1 −1 𝐴𝐿 2 1 𝑢1
={ - 2 * }
𝐹2 𝐿 −1 1 6 1 2 𝑢2
Consistent mass formulation

𝐹1 𝐴𝐸 1 −1 𝐴𝐿 2 1 𝑢1
={ - 2 * }
𝐹2 𝐿 −1 1 6 1 2 𝑢2
2. Evaluate the equations for eigen values and eigen vectors .

Ans:

m1x1’’ = K2 ( x2 – x1) – K1x1

m2x2’’ = - K2 ( x2 – x1)

m1x1’’+ k1x1 – k2(x2 – x1) = 0 ;

m2x2’’ +k2( x2 – x1) = 0 ;

 Fig 6.12 A spring fixed at one end and load at other end

Rewrite the above equations in matrix form we get

𝑚1 0 𝑥1′′ (𝑘1 + 𝑘2) −𝑘2 𝑥1

+ =0
0 𝑚2 𝑥2′′ −𝑘2 𝑘2 𝑥2
[ M ] { x’’ } + [ K ] { x } = 0

Multiply with [ m ]-1

[ M ]-1 [ M ] { x’’ } + [ M ]-1 [ K ] { x’’ } = 0

I { x’’ } + D { x } = 0

x = x sin t

x’’ = - 2 sin t

- 2 I x sin t + D x sin t = 0

(D-I)x=0

Put  = 2
𝐷 − 𝐼 = 0

Class work

1. Explain about dynamic analysis.

2.Explain about lumped mass formulation and derive the expression for it

3.Derive the expression for eigen values and eigen vectors

Home work

1.Derive the expression for consistent mass matrix

Problems:

1.Determine the eigen values and eigen vectors for a stepped bar element shown in fig.
Take E = 30*105 N/m 2 and  = 0.283kg/m3

Fig 6.13

Ans:

Given data

A1 = 1m2

A2 = 0.5m2
L1 = 10m

L2 = 5m

E = 30*105 N/m 2

 = 0.283kg/m3

Stiffness matrix for element 1

𝐴1𝐸1 1 −1
[K1]= 𝐿1 −1 1
1∗30∗100000 1 −1
[K1]= 10 −1 1
300000 −300000
[K1]=
−300000 300000
Stiffness matrix for element 2

𝐴2𝐸2 1 −1
[K2]= 𝐿2 −1 1
0.5∗30∗100000 1 −1
[K2]= 5 −1 1
300000 −300000
[K2]=
−300000 300000
Global stiffness matrix

[ K ] = [ K1 ] + [ K2 ]

300000 −300000 0
[ k ] = −300000 600000 −300000
0 −300000 300000
Consistent mass matrix for element 1

1 𝐴1𝐿1 2 1
[M1]= 6 1 2
0.283∗1∗10 2 1
[M1]= 6 1 2
 0.9433 0.4716
[M1]=
0.4716 0.9433
Consistent mass matrix for element 2

2 𝐴2𝐿2 2 1
[M2]= 6 1 2
0.283∗0.5∗5 2 1
[M2]= 6 1 2
0.2358 0.1179
[M2]=
0.1179 0.2358
Global mass matrix

[ M ] = [ M1 ] + [M2 ]

0.9433 0.4716 0
[ M ] = 0.4716 1.1791 0.1179
0 0.1179 0.2358

[ F ] = { k – 2M } { u }

[ F ] = { k – M } { u }

𝐹1 300000 −300000 0 0.9433 0.4716 0 𝑢1

𝐹2 = −300000 600000 −300000 −  0.4716 1.1791 0.1179 𝑢2
𝐹3 0 −300000 300000 0 0.1179 0.2358 𝑢3
Since node 1 is fixed u1 = 0. Eliminate first row and first column

𝐹2 600000 −300000 1.1791 0.1179 𝑢2

= −
𝐹3 −300000 300000 0.1179 0.2358 𝑢3

𝐾 − 𝑀 = 0

600000 −300000 1.1791 0.1179.

− =0
−300000 300000 0.1179 0.2358
60000 − 1.179 −300000 − 0.117
=0
−300000 − 0.117 300000 − 0.236
{ [60000 − 1.179 ] [300000 − 0.236 ] } – { [−300000 − 0.117 ] [−300000 − 0.117 ]
=0
[ (18 * 1010 ) – 6*105*0.236 ] -  (1.179*3*105 ) + ( 1.179*0.2362 ) – 9*1010 – 3*0.117*105 -
3*0.117*105+0.117*0.117*2

0.2962 – 5.642*105+9*1010 = 0

Solving the above equation we get

1 = 1.75*105 ; 2 = 17.3*105

12 = 1 = 1.75*105 ; 22 = 2 = 17.3*105

1 = 418.33 rad/sec

2 = 1315.29 rad/sec
𝜔𝑛
F= 2𝜋

𝜔1 418.33
F1 = 2𝜋 = = 66.6 Hz
2𝜋

𝜔2 1315 .29
F2 = 2𝜋 = = 209.33 Hz
2𝜋

Eigen values ( 66.6 Hz , 209.33Hz )

Eigen vector corresponding to 1

( K - 1M ) = 0

600000 −300000 1.1791 0.1179 𝑢2

− 1.75 ∗ 100000 =0
−300000 300000 0.1179 0.2358 𝑢3
60000 − 1.75 ∗ 100000 ∗ 1.179 −300000 − 1.75 ∗ 100000 ∗ 1.179 𝑢2
=0
−30000 − 1.75 ∗ 100000 ∗ 0.117 300000 − 1.75 ∗ 100000 ∗ 0.230 𝑢3
393.675 ∗ 1000 −320.47 ∗ 1000 𝑢2
=0
−320.47 ∗ 1000 258.7 ∗ 1000 𝑢3
393.675*1000u2 = 320.47*1000u3
𝑢3 393.67
= 320.47
𝑢2

u3 = 1.222 u2
Normalization U1T. M . U1 = 1

1.179 0.128 𝑢2
𝑢2 1.222𝑢2 =1
0.128 0.236 1.222𝑢2
𝑢2
𝑢2 ∗ 1.179 + (1.222𝑢2 ∗ 0.118) 0.118𝑢2 + (1.222𝑢2 ∗ 0.236) =1
1.222𝑢2
𝑢2
1.323𝑢2 0.406𝑢2 =1
1.222𝑢2
1.323(u2)2 + 0.496 ( u2)2 = 1

u2 = 0.74

u3 = 0.910

Eigen vector corresponding to 2

( K - 2M ) = 0

600000 −300000 1.1791 0.1179 𝑢2

− 17.3 ∗ 100000 =0
−300000 300000 0.1179 0.2358 𝑢3
60000 − 17.3 ∗ 100000 ∗ 1.179 −300000 − 17.3 ∗ 100000 ∗ 1.179 𝑢2
=0
−30000 − 17.3 ∗ 100000 ∗ 0.117 300000 − 17.3 ∗ 100000 ∗ 0.230 𝑢3
−1.439 ∗ 1000000 −502.41 ∗ 1000 𝑢2
=0
−502.41 ∗ 1000 −108.28 ∗ 1000 𝑢3
- 1.439*106u2 = 502.41*1000u3
𝑢3 −1.439∗1000
=
𝑢2 502.41

u3 = - 2.864 u2

Normalization U2T. M . U2 = 1

1.179 0.1179 𝑢2
𝑢2 −2.864𝑢2 =1
0.1179 0.2358 −2.864𝑢2
𝑢2
𝑢2 ∗ 1.179 + (−2.864𝑢2 ∗ 0.1179) 0.1179𝑢2 + (−2.864𝑢2 ∗ 0.2358) =1
−2.864𝑢2
𝑢2
0.841𝑢2 −0.557𝑢2 =1
−2.864𝑢2
0.841(u2)2 + 1.59 ( u2)2 = 1

u2 = 0.41, u3 = -1.174

Class work:

1. Determine the eigen values and eigen vectors for a stepped bar element shown in fig.
Take E = 30*105 N/m 2 and  = 0.283kg/m3

Home work

1. Determine the eigen values and eigen vectors for a stepped bar element shown in fig.
Take E = 200*105 N/m 2 and  = 0.476kg/m3