EST56 - Tópico 5a

EST - 56
DINÂMICA DE ESTRUTURAS E
AEROELASTICIDADE
Prof. Airton Nabarrete
Tópico 5-a: Sistemas com 2 GDL
1
Instituto Tecnológico da Aeronáutica
Contents
o Two-Degree-of-Freedom Model (Undamped)

o Eigenvalues and Natural Frequencies
o Eigenvector Normalization for Mode Shapes
o Matlab Commands
2
Introduction
The Millennium bridge required many degrees of freedom
to model and design with.
London
Video: https://www.youtube.com/watch?v=gQK21572oSU&t=11s
3
Introduction
The dynamic structural model of a missile:
4
Introduction
The dynamic structural model of an airplane:
5
Introduction
Two DOF (2 DOF) Systems
Referência (Brandão, 1996)
6
2 DOF
(Two-Degree-of-Freedom
Model)
7
Two DOF (2 DOF) Systems
The first step in analyzing multiple degrees of freedom (DOF)
is to look at 2 DOF
 DOF: Minimum number of coordinates to specify the
position of a system
 Many systems have more than 1 DOF
 Example of 2 DOF systems
 car with sprung and unsprung mass (both heave)
Fig 4.1 (Inman)
8
2 DOF Model (Undamped)
4.1
A two degree of freedom (2 DOF) system used to base

much of the analysis and conceptual development of
MDOF systems on.
9
Free-body diagram of each mass:
Figure 4.2 (Inman)
k1 x1 k2(x2 -x1)
m1 m2
x1 x2
10
Summing forces yields the equations of motion:
x1 (t ) =
m1  − k1 x1 (t ) + k2 ( x2 (t ) − x1 (t ) )
(4.1)
x2 (t ) =
m2  − k2 ( x2 (t ) − x1 (t ) )
Rearranging terms:
x1 (t ) + (k1 + k2 ) x1 (t ) − k2 x2 (t ) =
m1  0
(4.2)
x2 (t ) − k2 x1 (t ) + k2 x2 (t ) =
m2  0
11
Note that it is always the case that…
 A 2 DOF system has

 Two equations of motion!
 Two natural frequencies (as we shall see)!
12
The dynamics of a 2 DOF system consists of 2 homogeneous
and coupled equations:
 Free vibrations, so homogeneous eqs.
 Equations are coupled:
 Both have x1 and x2.
 If only one mass moves, the other follows
 Example: pitch and heave of a car model
 In this case the coupling is due to k2.
 Mathematically and Physically
 If k2 = 0, no coupling occurs and can be solved as two
independent SDOF systems
13
Initial Conditions:
 Two coupled, second -order, ordinary differential
equations with constant coefficients
 Needs 4 constants of integration to solve
 Thus 4 initial conditions on positions and
velocities
=
x1 (0) x= 
10 , x1 (0) =
x10 =
, x2 (0) x20 , x2 (0) x20
14
Solution by Matrix Methods
The two equations can be written in the form of a single
matrix equation (see pages 272-275 if matrices and vectors are a
struggle for you) :
 x1 (t )   x1 (t )   
x1 (t ) 
=x(t ) =  , x (t ) =  , 
x(t )  (4.4), (4.5)
 x2 (t )   x2 (t )  x2 (t ) 
 
 m1 0   k1 + k2 − k2 
M = , K 
k2 
(4.6), (4.9)
 0 m2  − k2
m1 x1 (t ) + (k1 + k 2 ) x1 (t ) − k 2 x2 (t ) = 0
x + Kx =
M 0
m2 x2 (t ) − k 2 x1 (t ) + k 2 x2 (t ) = 0
15
Initial Conditions:
IC’s can also be written in vector form:
 x10   x10 
=x(0) =  , and x (0)  
 x20   x20 
16
The approach to a Solution:
For 1DOF we assumed the scalar solution aeλt Similarly,

now we assume the vector form:
Let x(t ) = ue jωt (4.15)
/ 0, ω , u unknown
j =−1, u =
⇒ ( -ω 2 M + K ) ue jωt =
0 (4.16)
⇒ ( -ω M + K ) u =
2
0 (4.17)
17
This changes the differential equation of motion into algebraic
vector equation:
( -ω 2
M + K )u =
0 (4.17)
This is two algebraic equation in 3 uknowns
( 1 vector of two elements and 1 scalar):
 u1 
u =   , and ω
u2 
18
The condition for solution of this matrix equation requires that
the matrix inverse does not exist:
If the inv ( -ω 2 M + K ) exists ⇒ u =

0 : which is the
static equilibrium position. For motion to occur
/ 0 ⇒ ( -ω M + K )
−1
u= 2
does not exist
or det ( -ω 2 M + K ) =
0 (4.19)
The determinant results in 1 equation

in one unknown ω (called the characteristic equation)
19
Back to our specific system: the characteristic equation is
defined as:
det(-ω 2 M + K ) = 0 ⇒
−ω 2 m1 + k1 + k2 − k2 
det =0 ⇒ (4.20)
 − k2 −ω m2 + k2 
2
m1m2ω 4 − ( m1k2 + m2 k1 + m2 k2 )ω 2 + k1 k2 = 0 (4.21)
Eq. (4.21) is quadratic in ω 2 so four solutions result:

ω12 and ω22 ⇒ ±ω1 and ± ω2
20
Once ω is known, use equation (4.17) again to calculate the
corresponding vectors u1 and u2
This yields vector equation for each squared frequency:
(−ω12 M + K )u1 =
0 (4.22)
and
(−ω22 M + K )u 2 =
0 (4.23)
Each of these matrix equations represents 2 equations in the 2

unknowns components of the vector, but the coefficient matrix is
singular so each matrix equation results in only 1 independent
equation. The following examples clarify this.
21
Examples 4.1.5 & 4.1.6
Calculating u and ω :
 m1=9 kg,m2=1kg, k1=24 N/m and k2=3 N/m

 The characteristic equation becomes
ω4-6ω2+8=(ω2-2)(ω2-4)=0
ω2 = 2 and ω2 =4 or
ω 1,3 = ± 2 rad/s, ω 2,4 = ± 2 rad/s
Each value of ω2 yields an expression or u:
22
Examples 4.1.5 & 4.1.6
Computing the vectors u:
 u11 
For ω =2, denote u1 =   then we have
2
1
u12 
(-ω12 M + K )u1 =
0⇒
 27 − 9(2) −3   u11  0 
 −3   =   ⇒
 3 − (2)  u12  0 
9u11 − 3u
=12 0 and − 3u11 + u
=12 0
2 equations, 2 unknowns but DEPENDENT!
(the 2nd equation is -3 times the first)
23
Examples 4.1.5 & 4.1.6
Only the direction of vectors u can be determined, not the
magnitude as it remains arbitrary
u11 1 1
= ⇒ u11 = u12 results from both equations:
u12 3 3
only the direction, not the magnitude can be determined!
This is because: det(−ω12 M + K ) =
0.
The magnitude of the vector is arbitrary. To see this suppose
that u1 satisfies
(−ω12 M + K )u1 =
0, so does au1 , a arbitrary. So
(−ω12 M + K )au1 =
0 ⇔ (−ω12 M + K )u1 =
0
24
Examples 4.1.5 & 4.1.6
Likewise for the second value of ω2:
 u21 
For ω = 4, let u 2 =   then we have
2
2
u22 
(-ω12 M + K )u =
0⇒
 27 − 9(4) −3   u21  0 
 −3   =   ⇒
 3 − (4)  u22  0 
1
−9u21 − 3u22 = 0 or u21 = − u22
3
Note that the other equation is the same
25
Examples 4.1.5 & 4.1.6
What to do about the magnitude?
Several possibilities, here we just fix one element:
 13
Choose: u12 = 1 ⇒ u 1 =  
1
 −1 3 
Choose: u22 = 1 ⇒ u2 =  
1
26
Examples 4.1.5 & 4.1.6
Thus the solution to the algebraic matrix equation is
 13
ω 1, 3 = ± 2, has mode shape u 1 =  
1
 −1 3 
ω 2, 4 = ± 2, has mode shape u 2 =  
1
Here we have introduce the name mode shape

to describe the vectors u1 and u2.
The origin of this name comes later.
27
Return now to the time response:
We have computed four solutions:
x(t ) u1e − jω1t , u1e jω1t , u 2 e − jω2t , u 2 e jω2t ⇒ (4.24)
Since linear, we can combine as:
x(t ) = au1e − jω1t + bu1e jω1t + cu 2 e − jω2t + du 2 e jω2t
⇒ x(t=
) ( ae − jω1t
+ be jω1t
) 1 (
u + ce − jω2t
+ de jω2t
) u2
= A1 sin(ω1t + φ1 )u1 + A2 sin(ω2t + φ2 )u 2 (4.26)
where A1 , A2 , φ1 , and φ2 are constants of integration

Note that to go from the exponential
determined by initial conditions. form to to sine requires Euler’s formula
for trig functions and uses up the
+/- sign on omega
28
Physical interpretation of all that math!
 Each of the TWO masses is oscillating at TWO natural

frequencies ω1 and ω2
 The relative magnitude of each sine term, and hence of
the magnitude of oscillation of m1 and m2 is determined by
the value of A1u1 and A2u2
 The vectors u1 and u2 are called mode shapes because
the describe the relative magnitude of oscillation between
the two masses
29
What is a mode shape?
 First note that A1, A2, φ1 and φ2 are determined by the

initial conditions
 Choose them so that A2 = φ1 = φ2 =0
 Then:
 x1 (t )   u11 
x(t ) =   = A1   sin ω 1t = A1u 1 sin ω 1t
 x2 (t )  u12 
 Thus each mass oscillates at (one) frequency ω1 with
magnitudes proportional to u1 the 1st mode shape
30
Graphic Look at Mode Shapes
If IC’s correspond to mode 1 or 2, then the response is purely
in mode 1 or mode 2.
x1 x2
k1 k2
 13
Mode 1: m1 m2 u1 =  
1
x1= A/3 x2= A
x1 x2
k1 k2
 −1 3 
Mode 2: m1 m2 u2 =  
1
x1= -A/3 x2= A
31
Example 4.1.7
Given the initial conditions, compute the time response
1  0 
consider x(0) =   mm, x (0) =  
0 0 
 
sin ( 2t + φ1 ) − sin (2t + φ2 )
A1 A2
 1 
x (t )
 x (t ) = 3 3
 2   A1 sin ( 2t + φ1 ) + A2 sin (2t + φ2 ) 
 
2 cos( 2t + φ1 ) − 2 cos(2t + φ2 )
A A
 x1 (t )  1 2

 x (t ) = 3 3
 2   A1 2 cos( 2t + φ1 ) + A2 2 cos(2t + φ2 ) 
32
Example 4.1.7
At t = 0, we have
 
sin (φ1 ) − sin (φ2 )
A A
1 mm 1 2
 0 = 3 3
   A1 sin (φ1 ) + A2 sin (φ2 ) 
 
2 cos(φ1 ) − 2 cos(φ2 )
A1 A2
0  
0  =  3 3

 A1 2 cos(φ1 ) + 2 A2 cos(φ2 ) 
 
33
Example 4.1.7
4 equations in 4 unknowns:
3 = A1 sin (φ1 ) − A2 sin (φ2 )

0 = A1 sin (φ1 ) + A2 sin (φ2 )
0 = A1 2 cos(φ1 ) − A2 2 cos(φ2 )
0 = A1 2 cos(φ1 ) + A2 2 cos(φ2 )
π
Yields: A1 = 1.5 mm, A2 = −1.5 mm, φ1 = φ2 = rad
2
34
Example 4.1.7
The final solution is: x1 (t ) = 0.5 cos 2t + 0.5 cos 2t
(4.34)
x2 (t ) = 1.5 cos 2t − 1.5 cos 2t
These initial conditions gives a response that is a combination
of modes. Both harmonic, but their summation is not.
Figure 4.3
35
Sum of Modes
Solution as a sum of modes
x( t ) = a1u1 cosω1t + a 2u 2 cosω 2 t
Determines how the second

Determines how the first frequency contributes to the
frequency contributes to the response
response
36
2 DOF Model
Things to note:
 Two degrees of freedom implies two natural frequencies

 Each mass oscillates at with these two frequencies
present in the response and beats could result
 Frequencies are not those of two component systems
k1 k2
ω1 = 2≠ = 1.63, ω2 = 2 ≠ = 1.732
m1 m2
 The above is not the most efficient way to calculate
frequencies as the following describes
37
Some Matrix and Vector Reminders
a b −1 1 d −b 
=
A   ⇒ A=  −c
 c c  ad − cb  a 
xT =
x x12 + x22
 m1 0 
M =  ⇒ x T
Mx =m x
1 1
2
+ m x
2 2
2
 0 m 2
M > 0 ⇒ xT Mx > 0 for every value of x except 0
Then M is said to be positive definite
38
EIGENVALUE PROBLEM
(NATURAL FREQUENCIES
AND MODE SHAPES)
39
Eigenvalues and Natural Frequencies
 Can connect the vibration problem with the

algebraic eigenvalue problem developed in math
 This will give us some powerful computational
skills
 And some powerful theory
 All the codes have eigen-solvers so these painful
calculations can be automated
40
Some matrix results to help us use available computational
tools:
A matrix M is defined to be symmetric if M=M T
A symmetric matrix M is positive definite if
xT Mx > 0 for all nonzero vectors x
A symmetric positive definite matrix M can be factored
M = LL T
Here L is upper triangular, called a Cholesky matrix

41
If the matrix L is diagonal, it defines the matrix square root:
The matrix square root is the matrix M 1/ 2 such that

M 1/ 2 M 1/ 2 = M
If M is diagonal, then the matrix square root is just the root
of the diagonal elements:
 m1 0 
L = M 1/ 2 =  (4.35)
 0 m2 
42
A Change of Coordinates
For a symmetric, positive definite matrix M :
 m1 0  −1  1 0  −1/ 2
 1
m1
0 
M =  0=
m1
 , M 1 
, M  
 0 m2   m2 
 m2 
0 1

Let x(t ) = M −1/ 2q(t ) and multiply by M −1/ 2 :
−1/ 2 −1/ 2 −1/ 2 −1/ 2
M
  MM  
q (t ) + M
  KM  q(t ) =
0 (4.38)
I identity K symmetric
or (t ) + K=
q =
q(t ) 0 where K M −1/ 2 KM −1/ 2
 k
K is called the mass normalized stiffness and is similar to the scalar
m
used extensively in single degree of freedom analysis. The key here is that
K is a SYMMETRIC matrix allowing the use of many nice properties and
computational tools
43
Eigenvalue Problem
How the vibration problem relates to the real symmetric
eigenvalue problem:
Assu me= q(t ) ve jωt ni q(t ) + K=

q(t ) 0
−ω 2 ve jωt + K ve jωt= 0, v ≠ 0 or
Kv = ω 2 v ⇔ Kv = λ v v≠0
   
vibration problem real symmetric
eigenvalue problem
(4.40) (4.41)
Note that the martrix K contains the same type of information

as does ωn2 in the single degree of freedom case.
44
Eigenvalue Problem
Important properties of the n x n real symmetric eigenvalue
problem:
 There are n eigenvalues and they are all real valued

 There are n eigenvectors and they are all real valued
 The set of eigenvectors are orthogonal
 The set of eigenvectors are linearly independent
 The matrix is similar to a diagonal matrix
Window 4.1 page 285 (Inman)
45
Eigenvalue Problem
Square matrix review:
 Let aik be the ikth element of A then A is symmetric if

aik = aki denoted AT=A
 A is positive definite if xTAx > 0 for all nonzero x (also
implies each λi > 0)
 The stiffness matrix is usually symmetric and positive
semi definite (could have a zero eigenvalue)
 The mass matrix is positive definite and symmetric (and
so far, its diagonal)
46
Normalization of Vectors
Normal and orthogonal vectors:
 x1   y1 
    n
x =   , y =   , inner product is xT y = ∑ xi yi
x  y  i =1
 n  n
x is orthogonal to y if xT y = 0


x is normal if x x = 1
T


if the set of vectors is both orthogonal and normal, it is
called an orthonormal set
The norm of x is x = xT x (4.43)

47
Normalization of Vectors
Normalizing any vector can be done by dividing it by its norm:
x (4.44)
has norm of 1
T
x x
To see this compute
x xT x xT x
= = T
=1
xT x xT x xT x x x
48
Example 4.2.2
 −1/ 2  1 0 
−1/ 2 27 −3   1 0
=
3 3
K M KM  0 1   −3 3   0 1 
  
  3 −1
so K =   which is symmetric.
 −1 3 
3-λ -1 
det(K − λ I ) = det 

 = λ 2
− 6λ + 8 = 0
 -1 3-λ 
which has roots: λ1= 2= ω12 and λ2= 4= ω22
49
Example 4.2.2
( K − λ1 I ) v1 =
0⇒
3 − 2 −1   v11   0 
 −1 3 − 2   v =  0 ⇒
   12   
1
v11 − v12 =0 ⇒ v1 =α  
1
v1 = α 2 (1 + 1) =1 ⇒ α =1 2
1 1
v1 = 1 The first normalized eigenvector
2  
50
Example 4.2.2
Likewise the second normalized eigenvector is computed
and shown to be orthogonal to the first, so that the set is
orthonormal
1 1 1
v2 =  −1 , v1 v 2 = 2 (1 − 1) = 0
T
2 
1
v1 v1 = (1 + 1) = 1
T
2
1
v 2 v 2 = (1 + (−1)(−1)) = 1
T
2
⇒ v i are orthonormal
51
Example 4.2.2
Modes u and Eigenvectors v are different but related,
u 1 ≠ v1 and u 2 ≠ v 2
x = M −1/ 2 q ⇒ u = M −1/ 2 v (4.37)
Note
 3 0   1  1
M 1/ 2 u 1 =    3 = 1 = v1
 0 1   1   
52
Orthonormal Vectors
This orthonormal set of vectors is used to form an orthogonal
matrix,
P = [ v1 v2 ] called a matrix of eigenvectors (normalized)
 v T
1 v1 v1T v 2  1 0 
=T
P P  T = T  =  I
 v 2 v1 v 2 v 2  0 1 
P is called an orthogonal matrix

PT KP P=
T
 K v1 K v 2  PT [ λ1 v1 λ2 v 2 ]
λ1 v1T v1 λ2 v1T v 2  λ1 0 
=  T  =  = diag(ω 2
, ω 2) = Λ
2
(4.47)
λ1 v 2 v1 λ2 v 2 v 2   0 λ2 
T 1
P is also called a modal matrix

53
Example 4.2.3
Compute P and show that it is an orthogonal matrix,
From the previous example 4.2.2:
1 1 
P = [v1 v1 ] =
1
1 −1 ⇒
 2 
1 1 1 1  1 1 
P P=
T
  
2 2 1 −1 1 −1
1 1 + 1 1 − 1 1  2 0 
=   =   =I
2 1 − 1 1 + 1 2  0 2 
54
Example 4.2.4
Compute the square of the frequencies by matrix manipulation,
 1 1 1   3 −1 1 1 1 
P KP =
T
    
2 1 −1  −1 3  2 1 −1
1 1 1   2 4 
= 
2 1 −1  2 −4 
1 4 0 2 0 ω12 0 
=   =  =Λ= 2
2 0 8 0 4  0 ω 2 
⇒ ω1 = 2 rad/s and ω 2 = 2 rad/s
=
In general: T 
Λ P= ( λi ) diag(ωi2 )
KP diag= (4.48)
55
Example 4.2.5
The equations of motion:
x1 + (k1 + k2 ) x1 − k2 x2 =
m1  0
(4.49)
x2 − k2 x1 + (k2 + k3 ) x2 =
m2  0
Figure 4.4
In matrix form these become:

 m1 0  k1 + k2 −k2 
0  x+
  x=
0 (4.50)
 m2   − k2 k 2 + k3 
56
Example 4.2.5
Next substitute numerical values and compute P and λ
m1 = 1 kg, m2 = 4 kg, k1 = k3 = 10 N/m and k2 =2 N/m
1 0  12 −2 
=
⇒M  =  , K 
0 4   −2 12 
 −1/ 2 −1/ 2 12 −1

=
⇒K M =  −1 12 
KM

12 − λ −1 
⇒ det ( K − λ I ) = det 

 = λ 2
− 15λ + 35 = 0
 −1 12 − λ 
= ⇒ λ1 2.8902 = and λ2 12.1098
=⇒ ω1 1.7 rad/s
= and ω2 12.1098 rad/s
57
Example 4.2.5
Next compute the eigenvectors
For λ1 equation (4.41 ) becomes:
12 - 2.8902 −1   v11 
    =0
 −1 3 - 2.8902   v21 
⇒ 9.1089v11 = v21
Normalizing v1 yields
1= v1 = v112 + v21
2
= v112 + (9.1089) 2 v112
⇒ v11 =
0.1091, and v21 = 0.9940
 0.1091  −0.9940 
v1 =  , likewise v 2  
 0.9940   0.1091 
58
Example 4.2.5
Next check the value of P to see if it behaves as its suppose to
 0.1091 −0.9940 
P [=
v1 v 2 ]  
 0.9940 0.1091 
  0.1091 0.9940  12 −1  0.1091 −0.9940   2.8402 0 

=
T
P KP
−   −1 3   0.9940 0.1091   0 
 0.9940 0.1091    12.1098 
 0.1091 0.9940   0.1091 −0.9940  1 0 

P P=
T
   =  Yes!
 −0.9940 0.1091 0.9940 0.109 1   0 1 
59
Example 4.2.5
A note on eigenvectors
In the previous section, we could have chosed v 2 to be

 0.9940   -0.9940 
v2 =   instead of v 2 =  
 −0.1091   0.1091 
because one can always multiple an eigenvector by a constant
and if the constant is -1 the result is still a normalized vector.
Does this make any difference?
No! Try it in the previous example
60
“Solve by Hand”
All of the previous examples can and should be solved by “hand”
to learn the methods
However, they can also be solved on calculators with

matrix functions and with the codes listed in the last section
In fact, for more then two DOF one must use a code to
solve for the natural frequencies and mode shapes.
Next we examine 3 other formulations for solving for

modal data
61
MATLAB COMMANDS
62
Matlab Commands
 To compute the inverse of the square matrix A: inv(A) or

use A\eye(n) where n is the size of the matrix
 [P,D]=eig(A) computes the eigenvalues and

normalized eigenvectors (watch the order). Stores them in
the eigenvector matrix P and the diagonal matrix D (D=Λ)
63
More Commands
 To compute the matrix square root use sqrtm(A)

 To compute the Cholesky factor: L= chol(M)
 To compute the norm: norm(x)
 To compute the determinant det(A)
 To enter a matrix: K=[27 -3;-3 3]; M=[9 0;0 1];
 To multiply: K*inv(chol(M))
64
Normalization of Mode Shapes
An alternate approach to normalizing mode shapes
From equation (4.17) (− Mω 2

)
+ K u = 0, u≠0
Now scale the mode shapes by computing α such that
(α i u i ) M (α i u i ) = 1 ⇒ α i =
T 1
u Ti u i
w i = α i ui is called mass normalized and it satisfies:

−ωi2 Mw i + Kw i =0 ⇒ ωi2 =w Ti Kw i , i =1, 2 (4.53)
65
Computing the Eigenvalue Problem
There are 3 approaches to computing mode shapes and
frequencies
−1 −1
(i) ω Mu = Ku (ii) ω u = M Ku
2 2 −1
(iii) ω v = M
2 2
KM 2
v
(i) Is the Generalized Symmetric Eigenvalue Problem

easy for hand computations, inefficient for computers
(ii) Is the Asymmetric Eigenvalue Problem

very expensive computationally
(iii) Is the Symmetric Eigenvalue Problem

the cheapest computationally
66
Reading Recommendation
Topic 7.7 – Standard Eigenvalue Problem (Rao’s book);

Examples: 7.7, 7.8 and 7.11
67
Exercises
Problems 7.40 through 7.51 (Rao’s book).
68

EST56 - Tópico 5a

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

EST56 - Tópico 5a

Uploaded by

Copyright:

Available Formats

EST - 56

Prof. Airton Nabarrete

Tópico 5-a: Sistemas com 2 GDL

o Two-Degree-of-Freedom Model (Undamped)

Two DOF (2 DOF) Systems

Referência (Brandão, 1996)

Fig 4.1 (Inman)

A two degree of freedom (2 DOF) system used to base

Figure 4.2 (Inman)

 A 2 DOF system has

IC’s can also be written in vector form:

For 1DOF we assumed the scalar solution aeλt Similarly,

Let x(t ) = ue jωt (4.15)

If the inv ( -ω 2 M + K ) exists ⇒ u =

The determinant results in 1 equation

m1m2ω 4 − ( m1k2 + m2 k1 + m2 k2 )ω 2 + k1 k2 = 0 (4.21)

Eq. (4.21) is quadratic in ω 2 so four solutions result:

Each of these matrix equations represents 2 equations in the 2

 m1=9 kg,m2=1kg, k1=24 N/m and k2=3 N/m

ω 1,3 = ± 2 rad/s, ω 2,4 = ± 2 rad/s

Each value of ω2 yields an expression or u:

Several possibilities, here we just fix one element:

Here we have introduce the name mode shape

where A1 , A2 , φ1 , and φ2 are constants of integration

 Each of the TWO masses is oscillating at TWO natural

 First note that A1, A2, φ1 and φ2 are determined by the

3 = A1 sin (φ1 ) − A2 sin (φ2 )

x( t ) = a1u1 cosω1t + a 2u 2 cosω 2 t

Determines how the second

 Two degrees of freedom implies two natural frequencies

M > 0 ⇒ xT Mx > 0 for every value of x except 0

Then M is said to be positive definite

 Can connect the vibration problem with the

A symmetric matrix M is positive definite if

xT Mx > 0 for all nonzero vectors x

A symmetric positive definite matrix M can be factored

Here L is upper triangular, called a Cholesky matrix

The matrix square root is the matrix M 1/ 2 such that

Assu me= q(t ) ve jωt ni q(t ) + K=

Note that the martrix K contains the same type of information

 There are n eigenvalues and they are all real valued

Window 4.1 page 285 (Inman)

 Let aik be the ikth element of A then A is symmetric if

The norm of x is x = xT x (4.43)

P is also called a modal matrix

From the previous example 4.2.2:

⇒ ω1 = 2 rad/s and ω 2 = 2 rad/s

In matrix form these become:

 −1/ 2 −1/ 2 12 −1

  0.1091 0.9940  12 −1  0.1091 −0.9940   2.8402 0 

 0.1091 0.9940   0.1091 −0.9940  1 0 

In the previous section, we could have chosed v 2 to be

Does this make any difference?

No! Try it in the previous example

However, they can also be solved on calculators with

Next we examine 3 other formulations for solving for

 To compute the inverse of the square matrix A: inv(A) or

 [P,D]=eig(A) computes the eigenvalues and

 To compute the matrix square root use sqrtm(A)

 To enter a matrix: K=[27 -3;-3 3]; M=[9 0;0 1];

From equation (4.17) (− Mω 2

w i = α i ui is called mass normalized and it satisfies:

(i) Is the Generalized Symmetric Eigenvalue Problem

(ii) Is the Asymmetric Eigenvalue Problem