HD 7 Modal Analysis Undamped MDOF

ME617 - Handout 7
(Undamped) Modal Analysis of

MDOF Systems
The governing equations of motion for a n-DOF linear
mechanical system with viscous damping are:
+ D U
MU +K U =F (1)
(t ) (t )
where U, U, and U are the vectors of generalized displacement,

velocity and acceleration, respectively; and F( t ) is the vector of
generalized (external forces) acting on the system.
M, D,K represent the matrices of inertia, viscous damping and
stiffness coefficients, respectively1.
The solution of Eq. (1) is uniquely determined once initial

conditions are specified. That is,
=U
at t = 0 → U (0) = U o , U (2)
(0) o
In most cases, i.e. conservative systems, the inertia and stiffness

matrices are SYMMETRIC, i.e. M = M , K = K . The kinetic
T T
energy (T) and potential energy (V) in a conservative system are
1 T , V = 1 UT K U
T= U MU (3)
2 2
1
The matrices are square with n-rows = n columns, while the vectors are n-
rows.
MEEN 617 – HD#7 Undamped Modal Analysis of MDOF systems. L. San Andrés © 2008 1
In addition, since T > 0, then M is a positive definite matrix2. If V
>0, then K is a positive definite matrix. V=0 denotes the existence
of a rigid body mode, and makes K a semi-positive matrix.
In MDOF systems, a natural state implies a certain

configuration of shape taken by the system during motion.
Moreover a MDOF system does not possess only ONE natural
state but a finite number of states known as natural modes of
vibration. Depending on the initial conditions or external forcing
excitation, the system can vibrate in any of these modes or a
combination of them. To each mode corresponds a unique
frequency knows as a natural frequency. There are as many
natural frequencies as natural modes.
The modeling of a n-DOF mechanical system leads to a set of n-

coupled 2nd order ODEs, Hence the motion in the direction of one
DOF, say k, depends on or it is coupled to the motion in the other
degrees of freedom, j=1,2…n.
In the analysis below, for a proper choice of generalized

coordinates, known as principal or natural coordinates, the
system of n-ODE describing the system motion is independent of
each other, i.e. uncoupled. The natural coordinates are linear
combinations of the (actual) physical coordinates, and conversely.
Hence, the motion in physical coordinates can be construed or
interpreted as the superposition or combination of the motions in
each natural coordinate.
2
Positive definite means that the determinant of the matrix is greater than
zero. More importantly, it also means that all the matrix eigenvalues will be
positive. A semi-positive matrix has a zero determinant, with at least an
eigenvalues equaling zero.
For simplicity, begin the analysis of the system by neglecting
damping, D=0. Hence, Eq.(1) reduces to
+ K U = F
MU (4)
(t ) (t )
=U
and at t = 0 → U (0) = U o , U
(0) o
Presently, set the external force F=0, and let’s find the free
vibrations response of the system.
+ K U = 0
MU (5)
The solution to the homogenous Eq. (5) is simply
U = φ cos(ω t −θ ) (6)
which denotes a periodic response with a typical frequency ω .

From Eq. (6),
= − φ ω 2 cos(ω t −θ )
U (7)
Note that Eq. (6) is a simplification of the more general solution
U = φ e s t with s = i ω and where i = −1 (8)
Substitution of Eqs. (6) and (7) into the EOM (5) gives:
+ K U = 0 →
MU
→ − M φ ω 2 cos(ω t −θ ) + K φ cos(ω t −θ ) = 0
→ ⎡⎣ − M ω 2 + K ⎤⎦ φ cos(ω t −θ ) = 0
and since cos(ω t −θ ) ≠ 0 for most times, then
⎡⎣ − M ω 2 + K ⎤⎦ φ = 0 (9)
or
ω 2M φ = K φ (10)
Eq. (10) is usually referred as the standard eigenvalue

problem (mathematical jargon):
Aφ = λ φ
(11)
where Α =M K and λ = ω −1 2
Eq.(9) is a set of n-homogenous algebraic equations. A nontrivial

solution, φ ≠ 0 exists if and only if the determinant Δ of the
system of equations is zero, i.e.
Δ = −M ω2 + K =0 (12)
Eq. (12) is known as the characteristic equation of the system.

It is a polynomial in ω = λ , i.e.
2
Δ = 0 = a0 + a1 ω 2 + a2 ω 4 + a3 ω 6 + .... an ω n
n (13)
Δ = 0 = a0 + ∑ ( ai λ i
)
i =1
This polynomial or characteristic equations has n-roots, i.e. the
set {λk }k =1,2,....n or {± ωk }k =1,2,....n since ω = ± λ .
The ω’s are known as the natural frequencies of the system. In

the MATH jargon, the λ’s are known as the eigenvalues (of matrix
A)
Knowledge summary
a) A n-DOF system has n-natural frequencies.
b) If M and K are positive definite, then

0 < ω1 ≤ ω2 .......ωn −1 ≤ ωn .
c) If K is semi-positive definite, then

0 = ω1 ≤ ω2 .......ωn −1 ≤ ωn , i.e. at least one natural frequency
is zero, i.e. motion with infinite period. This is known as
rigid body mode.
Note that each of the natural frequencies satisfies Eq. (9).

Hence, associated to each natural frequency (or eigenvalues)
there is a corresponding natural mode vector (eigenvector) such
that
[ − M λi + K ] φ(i ) = 0, i =1,...n (14)
The n-elements of an eigenvector are real numbers (for

undamped system), with all entries defined except for a constant.
The eigenvectors are unique in the sense that the ratio between two
elements is constant, i.e.
⎛ ϕ( k ) j ⎞
⎜ ϕ ⎟ = constant for any j , i = 1,....n
⎝ ( k )i ⎠
The actual value of the elements in the vector is entirely

arbitrary. Since Eq. (14) is homogenous, if φ is a solution, so it is
α φ for any arbitrary constant α. Hence, one can say that the
SHAPE of a natural mode is UNIQUE but not its amplitude.
For MDOF systems with a large number of degrees of freedom,

n>>3, the eigenvalue problem, Eq. (11), is solved numerically.
Nowadays, PCs and mathematical computation software allow,
with a single (simple) command, the evaluation of all (or some)
eigenvalues and its corresponding eigenvectors in real time, even
for systems with thousands of DOFs.
Long gone are the days when the graduate student or practicing
engineer had to develop his/her own efficient computational
routines to calculate eigenvalues. Handout # 9 discusses briefly
some of the most popular numerical methods to solve the
eigenvalue problem.
A this time, however, let’s assume the set of eigenpairs

{ }
ωi , φ ( i ) is known.
i =1,2...n
Properties of natural modes

The natural modes (or eigenvectors) satisfy important
orthogonality properties. Recall that each eigenpair
{ }
ωi , φ ( i ) satisfies the equation
i =1,2...n
⎡⎣ − M ωi2 + K ⎤⎦ φ ( i ) = 0, i =1,...n . (15)
Consider two different modes, say mode-j and mode-k, each

satisfying
ω 2j M φ ( j ) = K φ ( j ) and ωk2 M φ ( k ) = K φ ( k ) (16)
T T
Pre-multiply the equations above by φ ( k ) and φ ( j ) to obtain
ω 2j φT( k ) M φ ( j ) = φT( k ) K φ ( j )
and (17)
ωk2 φT( j ) M φ ( k ) = φT( j ) K φ ( k )
Now, perform some matrix manipulations. The products

φ M φ and φ TK φ are scalars, i.e. not a matrix nor a vector. The
T
transpose of a scalar is the number itself. Hence,
( φT( j )K φ( k ) ) = ( K φ( k ) ) ( φT( j ) )
T T T
= φT( k ) K T φ ( j )
= φT( k ) K φ ( j ) since K = K T
and
(φ M φ ( k ) ) = φT( k ) M φ ( j ) since M = MT
T
T
( j)
for symmetric systems. Thus, Eqs. (17) are rewritten as
ω 2j φT( j ) M φ ( k ) = φT( j ) K φ ( k ) (a )
and (18)
ωk2 φT( j ) M φ ( k ) = φT( j ) K φ ( k ) (b)
Subtract (b) from (a) above to obtain
(ω 2
j − ωk2 ) φT( j ) M φ ( k ) = 0 (19)
if ω j ≠ ωk , i.e. for TWO different natural frequencies; then it

follows that
for j ≠ k φT( j ) M φ ( k ) = 0 and φT( j ) K φ ( k ) = 0 (20)
for j = k φT( j ) M φ ( j ) = M j and φT( j ) K φ ( j ) = K j = ω 2j M j (20)
where Kj and Mj are known as the j-modal stiffness and j-modal

mass, respectively.
Define a modal matrix Φ has as its columns each of the

eigenvectors, i.e.
Φ = [φ1 φ 2 .. φ n ] (21)
and the modal properties are written as
ΦT M Φ = [ M ] ; ΦT K Φ = [ K ] (22)
where [M] and [K] are diagonal matrices containing the modal
mass and stiffnesses, respectively.
The eigenvector set φ k=1,…n is linearly independent. Hence,

any vector (v) in n-dimensional space can be described as a linear
combination of the natural modes, i.e.
n
v = ∑ a j φ ( j ) = Φa (23)
j =1
⎡ a1 ⎤
⎢a ⎥
v = φ1 a1 + φ 2 a2 + .. + φ n an = [φ1 φ 2 .. φ n ] ⎢ 2 ⎥ = Φa
⎢ .. ⎥
⎢ ⎥
⎣ an ⎦
System Response in Modal Coordinates
The orthogonality property of the natural modes (eigenvectors)
permits the simplification of the analysis for prediction of system
response. Recall that the equations of motion for the undamped
system are
MU + K U = F (4)
(t ) (t )
=U
and at t = 0 → U (0) = U 0 , U
(0) 0
Consider the modal transformation U(t ) = Φ q(t ) (24)3

= Φ q
And with U ( t ) , then EOM (4) becomes:
(t )
+ K Φq = F( t )
M Φq
which offers no advantage in the analysis. However, premultiply

T
the equation above by Φ to obtain
( Φ M Φ ) q + ( Φ K Φ ) q = Φ F
T T T
(t ) (25)
and using the properties of the natural modes,

ΦT M Φ = [ M ]; ΦT K Φ = [ K ] , then Eq. (25) becomes
[ ] [ ]
M
q + K q = Q = Φ T
F( t ) (26)
3
Eq. (24) sets the physical displacements U as a function of the
modal coordinates q. This transformation merely uses the property
of linear independence of the natural modes.
And since [M] and [K] are diagonal matrices. Eq. (26) is just a set
of n-uncoupled ODEs. That is,
M 1 q1 + K1 q1 = Q1
M 2 q2 + K 2 q2 = Q2
(27)
.....
M n qn + K n qn = Qn
Or M j qj + K j q j = Q j with ωn j = Kj
Mj , j =1,2...n (28)
The set of q’s are known as modal or natural

coordinates (canonical or principal, too). The vector
Q = ΦT F( t ) is known as the modal force vector.
Thus, the major advantage of the modal transformation (24)

is that in modal space the EOMS are uncoupled. Each equation
describes a mode as a SDOF system.
The unique solution of Eqs. (28) needs of initial conditions

specified in modal space, i.e. {q o , q o } .
Using the modal transformation, = Φq

U o = Φq o ; U , it
o o
follows

q o = Φ −1U o ; q o = Φ −1 U (28)
o
However, Eq. (28) requires of the inverse of modal matrix Φ ,

−1
i.e. Φ Φ = I . For systems with a large number of DOF, n>> 1,
−1
finding the matrix Φ is computationally expensive.
A more efficient to determine the initial state {q o , q o } in modal
coordinates follows. Start with the fundamental transformation,
U o = Φq o , and premultiply this relationship by ΦT M to
obtain,
ΦT M U o = ΦT M Φq o
since [ M ] = Φ M Φ , hence
T
= [ M ]qo ,
q o = [ M ] ΦT M U o ,
−1
(29a)
q o = [ M ]
−1 T
Φ MU o
or
1 T 1 T
qok = φ( k ) ( M U o ) , qok = φ( k ) ( M U
)
o (29b)
Mk Mk
Eqs. (29) are much easier to calculate efficiently when n-DOF is
large. Note that finding the inverse of the modal mass matrix [M]-1
is trivial, since this matrix is diagonal.
Comparing eqs. (28) and (29a) it follows that
Φ −1= [ M ] ΦT M
−1
(30)
The solution of ODEs M j qj + K j q j = Q j with initial
{ }
conditions qo j , qo j follows an identical procedure as in the
solution of the SDOF response. That is, each modal response adds
the homogeneous solution and the particular solution. The
particular solution clearly depends on the time form of the modal
force Q(t), i.e step-load, ramp-load, pulse-load, periodic load, or
arbitrary time form.
Free response in modal coordinates

Without modal forces, Q=0, the modal equations are
M j q H j + K j qH j = 0 = Q j (31a)
with solutions, for an elastic mode

qo j
( )
qH j = qo j cos ωn j t +
ωn
( )
sin ωn j t if ωn j ≠ 0 (31b)
j
; and for a rigid body mode

qH j = qo j + qo j t if ωn j = 0 (31c)
j=1,2,….n
Forced response in modal coordinates
For step-loads, QS j , the modal equations are

M j qj + K j q j = QS j (32a)
and; for an elastic mode, ωn ≠ 0 , j
qo j QS j
( )
q j = qo j cos ωn j t +
ωn
( )
sin ωn j t +
Kj ⎣
nj ( )
⎡1 − cos ω t ⎤
⎦
(32a)
j
; and for a rigid body mode, ωn j = 0 ,
1 QS j 2
q j = qo j + qo j t + t (32c)
2Mj
j=1,2,….n
For periodic loads,, the modal equations are
M j qj + K j q j = QPj cos(Ωt ) (33a)
with solutions
for an elastic mode, ωn ≠ 0 , and Ω ≠ ωn
j j
⎡ ⎤
QPj ⎢
( )
q j = C j cos ωn j t + S j sin ωn j t +( ) 1 ⎥ cos ( Ω t )
K j ⎢1 − Ω ω
( )
2⎥
⎢⎣ nj ⎥⎦
(33b)
Note that if Ω = ωn j , a resonance appears that will lead to system

destruction.
For a rigid body mode, ωn j = 0 ,
QPj
q j = qo j + qo j t − cos(Ωt ) (33c)
M jΩ2
For arbitrary-loads Q j , the modal response is

q jo 1 t
qj = qjo cos(ωnj t) +
ωn
sin(ωnj t) +
M jωnj ∫Qτ
0 j( ) sin ⎡⎣ωnj (t −τ )⎤⎦ dτ
j
(34)
for an elastic mode, ωn ≠ 0 . j
System Response in Physical Coordinates
Once the response in modal coordinates is fully determined, the
system response in physical coordinates follows using the modal
transformation
U(t ) = Φ q(t ) =
⎡ q1( t ) ⎤
⎢ ⎥
q
U ( t ) = [φ1 φ 2 .. φ n ] ⎢ 2 ⎥ = φ1 q1 + φ 2 q2 + .. + φ n qn
⎢ .. ⎥
⎢ ⎥
⎣⎢ qn ⎥⎦
n
U ( t ) = ∑ φ j q j( t ) (35)
j =1
One important question follows: are all the modal

responses important and need be accounted for to
obtain the response in physical coordinates? If not,
savings in computation time are evident. Hence, the physical
response becomes
m
U ( t ) ≈ ∑ φ j q j( t ) , m < n (36)
j =1
If m<n, then how many modes are to be included to ensure the

physical response is accurate? That is, which modes are important
and which others are not?
Example: Consider the case of force excitation with frequency
Ω ≠ ωn j and acting for very long times. The EOMs in physical
space are
+ K U = F cos ( Ωt )
MU P
Let’s assume there is a little damping; hence, the steady state

periodic response in modal coordinates is (see eq. (33b)):
⎡ ⎤
QPj ⎢ 1 ⎥ cos ( Ω t )
qj ≈ (37a)
K j ⎢1 − Ω ω
( )
2⎥
⎢⎣ nj ⎥⎦
And thus,
⎛ ⎡ ⎤⎞
⎜ QPj n
1
U = U P cos(Ωt ) = Φq = ∑ φ j ⎢ ⎥ ⎟ cos ( Ω t )
⎜ Kj
( )
2 ⎟
j =1 ⎜
⎢1 − Ω ω ⎥⎟
⎝ ⎢⎣ nj ⎥⎦ ⎠
(38)
The physical response is also periodic with same frequency as the
force excitation.
Recall that K j = ωn j M j = φ ( j ) K φ ( j ) and QPj = φ ( j ) FP

2 T T
However, nowadays the engineer in a hurry prefers to dump the

problem into a super computer; and for U = U P cos(Ωt ) , finds the
solution
−1
U P = ⎡⎣ K − Ω M ⎤⎦ FP 2
(39)
at a fixed excitation frequency Ω. Brute force substitutes beauty

and elegance, time savings in lieu of understanding!
Example: Find natural frequencies and natural ORIGIN := 1
mode shapes of UNDAMPED system.
Given EOMs for a 2DOF - undamped- system:
⎛⎜ M2 0 ⎟⎞ d2 ⎛⎜ X2 ⎟⎞ ⎛⎜ 2 ⋅K2 −2K2 ⎞ ⎛ X2 ⎞
⎟ ⋅⎜ ⎟ = ⎛ 0 ⎞
⋅ +
⎜ 0 M1 ⎟ dt2 ⎜ X1 ⎟ ⎜ −2 ⋅K2 2 ⋅K2 + K1 ⎟ ⎜ X1 ⎟ ⎜ K ⋅Z ⎟ (1)
⎝ ⎠ ⎝ ⎠ ⎝ ⎠⎝ ⎠ ⎝ 1 ⎠
where M2 =mo, M1 =5 m o , K2=k o; K1 =5 ko
⎛⎜ mo 2 ⋅ ko −2ko
⎟⎞ ⋅ d2 ⎛⎜ 2 ⎟⎞ + ⎛⎜ ⎞⎟ ⎛⎜ X2 ⎟⎞ ⎛ 0 ⎞
0 X
⎜ 0
⋅ = ⎜ K ⋅Z ⎟
5 ⋅mo ⎟ dt2 ⎜ X1 ⎟ ⎜ −2 ⋅ko 2 ⋅ ko + 5 ⋅ ko ⎟ ⎜ X1 ⎟ ⎝ 1 ⎠
⎝ ⎠ ⎝ ⎠ ⎝ ⎠⎝ ⎠
(a) PROCEDURE TO FIND NATURAL FREQUENCIES AND NATURAL MODES: Assume the
motions are periodic with frequency ω, ie
X2 = a1 ⋅ cos ( ω ⋅t) X1 = a2 ⋅ cos ( ω ⋅t) (2)
Set the RHS of Eq. (1) equal to 0. Substitution of (2) into (1) gives
⎛ 2 ⋅ko − mo ⋅ω2 −2ko ⎞ ⎛ a1 ⎞

⎜ ⎟ ⋅⎜ ⎟ ⋅cos ( ω ⋅t) = ⎛ 0 ⎞
⎜ 2⎟ ⎜ a ⎟
⎜ ⎟
⎝ 0⎠
−2 ⋅ ko 7 ⋅ko − 5 ⋅mo ⋅ ω
⎠⎝ ⎠
2
⎝ cancel cos(ωt) since it
is NOT zero for all times
The homogeneous system of eqns
⎛ 2 ⋅ko − mo ⋅ω2 −2ko ⎞ ⎛ a1 ⎞

⎜ ⎟ ⋅⎜ ⎟ = ⎛ 0 ⎞
⎜ 2⎟ ⎜ a ⎟
⎜ ⎟ (3)
−2 ⋅ko 7 ⋅ko − 5 ⋅mo ⋅ ω ⎝ 2⎠ ⎝ 0⎠
⎝ ⎠
has a non-trivial solution if the determinant of the system of equations equals zero, i.e. if
( 2
)(
Δ ( ω ) = 7 ⋅ ko − 5 ⋅ mo ⋅ω ⋅ 2 ⋅ko − mo ⋅ ω
2
) − 4 ⋅ ko2
=0
2
Let λ=ω , and expanding the products in the determinant
2 2
(
0 = λ ⋅5mo − λ ⋅ 7 ⋅ko ⋅mo + 10 ⋅ ko ⋅ mo + 14 ⋅ ko − 4 ⋅ko) 2 2
Let ⎯ ⎛ mo ⎞ Leads to: ⎡ ( ⎯) 2 + b ⋅⎯ ⎤

λ = λ ⋅⎜ ⎟ 0 = ⎣a ⋅ λ λ + c⎦ (4)
⎝ ko ⎠ with: a := 5 b := −17 c := 10
The roots (eigenvalues) of the characteristic equation are

(
− b − b2 − 4 ⋅ a ⋅ c ) 0.5 (
− b + b2 − 4 ⋅ a ⋅ c ) 0.5 ⎛ 0.757 ⎞ ⎛ ko ⎞
λ 1 := λ 2 := λ=⎜
2 ⋅a 2 ⋅a ⎟ ⎜m ⎟
⎝ 2.643 ⎠ ⎝ o⎠
and the natural frequencies are:
ω 1 := ( λ 1) ω 2 := ( λ 2)
0.5 0.5 0.5
⎛ ⎞ ⎛ ko ⎞
0.87
ω=⎜ ⎟ ⎜ ⎟
⎝ 1.626 ⎠ ⎝ mo ⎠
Find the eigenvectors:
The two equations in (3) are linearly dependent. Thus, one cannot solve for a1 and a2. Set
φ 1 := 1
arbitrarily; and from the first equation
for ω 1
φ2 =
( 2 ⋅ko − mo ⋅ ω 1 )
=
(2 ⋅ko − 0.757 ⋅ko)
2
φ 2 :=
( 2 − 0.757)
2
2 ⋅ ko 2 ⋅ ko
φ 2 = 0.621
φ 1 := φ
φ1 = ⎜
⎛ 1 ⎞ is the first eigenvector (natural mode)
⎟
⎝ 0.621 ⎠
(b) Explanation: DOF1 (X2) and DOF2 (X1) move in phase, with X2>X1
ω2 φ 1 := 1
( 2 ⋅ko − mo ⋅ ω 2 )
for
(2 ⋅ko − 2.643 ⋅ko)
2
φ2 = = ( 2 − 2.643)
2 ⋅ ko 2 ⋅ ko φ 2 :=
2
φ 2 := φ ⎛ 1 ⎞
φ2 = ⎜ ⎟ is the 2nd eigenvector (natural mode)
⎝ −0.321 ⎠
(b) Explanation:DOF1 (X2) and DOF2 (X1) move 180 deg OUT of phase, with |X2|>|X1|
(c) find the numerical value for each natural frequency:

lb
0.5 mo :=
1000lb ko := 105 ⋅
Since
⎛ 0.87 ⎞ ⎛ ko ⎞ g in
ω := ⎜ ⎟ ⋅⎜ ⎟
⎝ 1.626 ⎠ ⎝ mo ⎠ 0
Note that mass must be
ω=⎜
⎛ 170.947 ⎞ rad expressed in physical units
⎟
⎝ 319.495 ⎠ sec consistent with the problem, i.e.
ω
fn := ⎛ 27.207 ⎞ Hz lb ⋅ sec2
2 ⋅π fn = ⎜ ⎟ mo = 2.59
⎝ 50.849 ⎠ in
Perform same work using a calculator
Use BUILT IN functions
-
M := ⎜
⎛1 0⎞ ⎛ 2 −2 ⎞ ⋅ k Not much learning
⎟ ⋅mo K := ⎜ ⎟ o
⎝0 5⎠ ⎝ −2 7 ⎠
−1
Let Z := M ⋅K
λ := sort ( eigenvals ( Z) ) ⎛ 2.921 × 104 ⎞ 1

λ=⎜ ⎟ 2
⎝ × ⎠ sec
5
1.021 10
ω 1 := ( λ 1)
.5
⎛ 170.914 ⎞ rad ω ⎛ 27.202 ⎞ Hz

ω 2 := ( λ 2) ω=⎜ ⎟ =⎜ ⎟
.5
⎝ 319.495 ⎠ sec 2 ⋅π ⎝ 50.849 ⎠
φ 1 := eigenvec ( Z , λ 1)
natural modes:
⎛ 0.849 ⎞
φ1 = ⎜ ⎟ ( φ1) 2
⎝ 0.528 ⎠ = 0.622
( φ1) 1
φ 2 := eigenvec ( Z , λ 2) ⎛ 0.952 ⎞
φ2 = ⎜ ⎟
⎝ −0.306 ⎠ ( φ2) 2
= −0.322
( φ2) 1
which are the same ratios as for
the vectors found earlier
1000lb 5 lb ft
Example: Undamped Modal Analysis mo := k o := 10 ⋅ g = 32.174 ORIGIN := 1
g in 2
sec
Equations of motion: natural frequencies, modal matrix (eigenvectors)
⎛⎜ mo 0 ⎞⎟ d 2 ⎛⎜ X2 ⎞⎟ ⎛⎜ 2 ⋅ ko −2k o ⎞⎟ ⎛⎜ X2 ⎞⎟ ⎛ 0 ⎞ ⎛ 170.95 ⎞ ⋅ rad ⎛ 1 1 ⎞

ω n := ⎜ ⎟ Φ := ⎜ ⎟
⎜⎝ 0 5⋅ mo ⎟⎠ dt2 ⎜⎝ X1 ⎟⎠ ⎜⎝ −2 ⋅ ko 7⋅ k o ⎟⎠ ⎜⎝ X1 ⎟⎠ ⎜⎝ k o⋅ Z ⎟⎠
⋅ + ⋅ = ⎝ 319.5 ⎠ sec ⎝ 0.621 −0.321 ⎠
given: Zo := 0.01⋅ in provides a Fo := k o ⋅ Zo constant force

Define matrices:
⎛⎜ mo 0 ⎞⎟ ⎛⎜ 2 ⋅ ko −2k o ⎞⎟ ⎛ 0⋅ lb ⎞ at t=0s, Initial conditions:
M := K := F := ⎜F ⎟
⎜⎝ 0 5 ⋅ mo ⎟⎠ ⎜⎝ −2 ⋅ ko 7⋅ k o ⎟⎠ ⎝ o⎠
system is at REST
(a) FIND modal masses and stiffnesses MM := ΦT ⋅ M⋅ Φ KM := Φ ⋅ K⋅ Φ

T
⎛ 7.584 8.534 × 10 ⎞⎟ lb⋅ sec2

−3
MM = ⎜ non-diagonal elements are very small= non zero b/c of
⎜ −3 ⎟ in roundoff in numerical calculator
⎝ 8.534 × 10 3.925 ⎠
2
modal masses and stiffnesses: Mode 1 M m := M M Km := ⎛ ω n ⎞ ⋅ M M
1 1, 1 1 ⎝ 1⎠ 1, 1
M m := M M
Mode 2 2, 2 2
2 Km := ⎛ ω n ⎞ ⋅ M M
2 ⎝ 2⎠ 2, 2
⎛ 2.216 × 105 ⎞ lb
2
⎛ 7.584 ⎞ lb⋅ sec
Mm = ⎜ ⎟ Km = ⎜ ⎟
⎝ 3.925 ⎠ in ⎜ 5 ⎟ in
⎝ 4.006 × 10 ⎠
(b) Find initial moddal displacements and velocities and modal force vector (Q)
At time t=0s, the system is at REST at its static equilibrium position, hence the initial conditions are null
displacements and null velocities. Of course, the same applies to modal space, i.e. null initial displacements
and velocities
⎛0⎞ X1 ⎛ 0 ⎞ ft
for generality, define: Xo := ⎜ ⎟ ⋅ ft Vo := ⎜ ⎟ ⋅ Calculate inverse of A matrix −1
⎝0⎠ X 2
⎝ 0 ⎠ sec Φ inv := Φ
and in modal coordinates q o := Φ inv⋅ Xo q o_dot := Φ inv ⋅ Vo velocity
(disp & velocities)
⎛ 0 ⎞ ft ⎛ 0 ⎞ ft No need for actual calculation
qo = ⎜ ⎟ q o_dot = ⎜ ⎟
⎝0⎠ ⎝ 0 ⎠ sec - a knowledge statement suffices
Define modal force
T
Q := Φ ⋅ F ⎛ 621 ⎞ lb
Q= ⎜ ⎟ Both natural modes will be excited
⎝ −321 ⎠
(c) Modal EOMs and modal responses
⎛⎜ d2 ⎞⎟
The EOMs in modal space are uncoupled Mm q + Km ⋅ q = Q
i⎜ 2 i⎟ i i i i = 1,2
and equal to ⎝ dt ⎠
Using the cheat sheet, and since the Initial conditions Q Q
1 2
are null, the response in modal coordinates are δ m := δ m :=
1 Km 2 Km
1 2
⎛ 170.95 ⎞ rad
q 1 ( t) := δ m ⋅ ⎛ 1 − cos⎛ ω n ⋅ t⎞ ⎞ q 2 ( t) := δ m ⋅ ⎛ 1 − cos⎛ ω n ⋅ t⎞ ⎞ ωn = ⎜ ⎟
1⎝ ⎝ 1 ⎠⎠ 2 ⎝ 2 ⎝ ⎠⎠ where: ⎝ 319.5 ⎠ sec
⎛ 2.802 × 10− 3 ⎞
where δm = ⎜ ⎟ in are the "static" deflections in modal space. δ2 << δ1, thus first
⎜ − 4⎟ modal response is MORE important
⎝ −8.013 × 10 ⎠
(d) The response in physical coordinates, X1 and X2, equals (from transformation x=Aq)
with ⎛ 1 1 ⎞
X1 ( t) := q 1 ( t) + q 2 ( t)
X1 ( t) = δ m ⋅ ⎛ 1 − cos⎛ ω n ⋅ t⎞ ⎞ + δ m ⋅ ⎛ 1 − cos⎛ ω n ⋅ t⎞ ⎞
Φ= ⎜ ⎟
⎝ 0.621 −0.321 ⎠
1 ⎝ ⎝
1 ⎠⎠ 2 ⎝ 2 ⎝ ⎠⎠
X2 ( t) := 0.621 ⋅ q 1 ( t) − 0.321 ⋅ q 2 ( t) for graph below:
1
X2 ( t) = δ m ⋅ 0.621 ⋅ ⎛ 1 − cos⎛ ω n ⋅ t⎞ ⎞ + δ m ⋅ ( −0.321 ) ⋅ ⎛ 1 − cos⎛ ω n ⋅ t⎞ ⎞
⎛ 2 ⋅ π ⎟⎞
Tlarge := 10⋅ ⎜
1 ⎝ ⎝ 1 ⎠⎠ 2 ⎝ 2 ⎝ ⎠⎠ ⎜ ω n1 ⎟
−4
⎝ ⎠
−3 δ m ⋅ ( −0.321 ) = 2.572 × 10 in
δ m ⋅ 0.621 = 1.74 × 10 in 2
1
Explanation: Since q1 and q2 are non-zero, then physical motion, X1 &X2, shows excitation of the
TWO fundamental modes of vibration - BUT response for second mode is much less
GRAPHs not needed for exam:
Note that there is no damping or
0.006
attenuation of motions.
0.005
displacements (inch)
0.004 Not too complicated physical response. It

0.003 shows dominance of first mode (lowest
0.002
natural freq or largest period)
0.001
2⋅ π
0 = 0.037 sec
ωn
0.001 1
2⋅ π
0.002
0 0.046 0.092 0.14 0.18 0.23 0.28 0.32 0.37 = 0.02 sec
ωn
time (sec) 2
X1
X2
terminal value
Terminal condition:
If damping is present and since the applied force is a constant, the system will achieve a new steady state
condition.
In the limit as t approaches very, very large values ⎛⎜ X1 ⎞⎟ ⎛ 0 ⎞
2 ⎛ X1end ⎟⎞
K⋅ ⎜
d
=⎜ ⎟ ; hence ===> =F
dt ⎜⎝ X2 ⎟⎠ ⎝ 0 ⎠ ⎜ X2end ⎟
2
⎝ ⎠
And equations of motion reduce to:
⎛⎜ 2 ⋅ ko −2k o ⎞⎟ ⎛⎜ X1end ⎞⎟ ⎛ 0 ⎞
⎜⎝ −2 ⋅ ko 7⋅ k o ⎟⎠ ⎜⎝ X2end ⎟⎠ ⎜⎝ Fo ⎟⎠
⋅ = 3
Fo = 1 × 10 lb
And solving this system of equations using Cramer's rule 2 2

Δ := 14⋅ k o − 4 ⋅ k o determinant of system of eqns.
Fo ⋅ 2 ⋅ k o 2 ⋅ k o ⋅ Fo
X1end := X2end :=
Δ Δ
−3 −3
X1end = 2 × 10 in X2end = 2 × 10 in
Note that the graph of undamped periodic motions Z(t) and X(t)
shows oscillatory motions abut these terminal or end values. recall
Zo = 0.01 in
⎛ 2 × 10− 3 ⎞
K ⋅F = ⎜ ⎟ in
OR −1
⎜ − 3⎟ Zo
⎝ 2 × 10 ⎠ =5
X1end
0.006
COMPARE actual 0.005
response with a
displacements (inch)
0.004
response
0.003
neglecting q2.
0.002
Indeed mode 2 does
0.001
not afffect the physical
0
response, except for
motion X2 sligthly 0.001
0.002
0 0.12 0.25 0.37
time (sec)
X1
X2
terminal
Normalization of eigenvectors (natural modes)
Recall that the components of an eigenvector φ j are
ARBITRARY but for a multiplicative constant. If one of the
elements of the eigenvector is assigned a certain value, then this
vector becomes unique, since then n-1 remaining elements are
automatically adjusted to keep constant the ratio between any two
elements in the vector.
In practice, the eigenvectors are normalized. The resulting

vectors are called NORMAL MODES.
Some typical NORMS are
L1 norm: q ( j ) =1= max q jk ( ) (39a)
L2 norm: q ( j ) =1= q j1 + q j2 + .... + q jn

2 2 2
(39b)
Or making the mass modal matrix equal to the identity matrix,

[M]=I, i.e.
φT( j ) M φ ( j ) = M j = 1 (39c)
hence
φT( j )K φ ( j ) = K j = ω 2j M j = ωn 2j (39d)
This normalization has obvious advantages since it will reduce

the number of operations when conducting the modal analysis.
However, the physical significance of the modal equations is lost.
Note that the modal Eqs. (26) become:
qj + ωn2j q j = Q j
Your lecturer recommends this normalization procedure be

conducted only for systems with large number of degrees of
freedom, n>>>1.
Note that the normalization process is a mere convenience,

devoid of any physical significance.
Rayleigh’s Energy Method

The method is a procedure to determine an approximate value
(from above) for the fundamental natural frequency of a MDOF
system. At times, the full solution of the eigenvalue problem is of
NO particular interest and an estimate of the system lowest natural
frequency suffices.
{
Recall that the pairs ωi , φ ( i ) }
i =1,2...n
satisfy Kφ ( i ) = ωi2M φ ( i )
with properties ΦT M Φ = [ M ] ; ΦT K Φ = [ K ]
i.e. with modal stiffness and masses calculated from:
Ki
K i = φT( i ) Kφ ( i ) ; M i = φT( i ) Mφ ( i ) , and ωi2 = (41)
Mi
Ki 1
2 φT( i ) Kφ ( i )
That is, ωi2 = Mi
= T
(42)
2 φ ( i ) Mφ ( i )
1
Above, the numerator relates to the potential or strain energy of

the system for the i-mode, and the denominator to the kinetic
energy for the same mode.
Consider an arbitrary vector u and define Rayleigh’s
quotient R(u) as
T
u Ku
1
R (u) = 2 T (43)
1 u Mu
2
R (u) is a scalar whose value depends not only on the matrices M

& K, but also on the choice of the vector u.
Clearly, if the arbitrary vector u coincides with (or is a multiple

of) one of the natural mode vectors, then Rayleigh’s quotient will
deliver the exact natural frequency for that particular mode. It can
also be shown that the quotient has a stationary value, i.e. a
minimum, in the neighborhood of the system natural modes
(eigenvectors). To show this, since u is an arbitrary vector and the
natural modes are a set of linearly independent vectors, then one
can represent
n
u = ∑ φ j c j = Φc (44)
j =1
Where c = {c1 c2 .. cn } is the vector of coefficients in the
T
expansion. Substitution of the expression above into Rayleigh’s

quotient gives
2 ( Φc ) K ( Φc ) cT ( ΦT KΦ ) c
T
1
R (u) = = T T
2( ) M ( Φc ) c ( Φ MΦ ) c
T
1 Φc
cT [ K ] c
R (u) = T (45)
c [ M ]c
Assume the modes have been normalized with respect to the
mass matrix, i.e.
n
c ⎡⎣ω ⎤⎦ c T 2
n
∑c ω 2
i
2
ni
R (u) = = i =1
n
(46a)
cT Ic
∑c i =1
2
i
Next, consider that the arbitrary vector u (which at this time can be
regarded as an assumed mode vector) differs very little from the
natural mode (eigenvector) φ ( r ) . This means that in the expansion
of vector u, the coefficients ci << cr ; for i = 1,2,...n and i ≠ r
Or
ci = ς i cr ; ς i <<1 for i = 1,2,...n and i ≠ r
Then, Rayleigh’s quotient is expressed as
n
c ω +c
2
r
2
nr
2
r ∑ςω
i =1, i ≠ r
i
2 2
ni
R (u) = n
cr2 + cr2 ∑ς
i =1, i ≠ r
i
2
( )
n n 2
ω +2
nr ∑ςω
i =1, i ≠ r
i
2 2
ni 1+ ∑
i =1, i ≠ r
ς iωni
ωnr
R (u) = n
=ω 2
nr n
(46b)
1+ ∑ς
i =1, i ≠ r
i
2
1+ ∑ς
i =1, i ≠ r
i
2
{ }
The quantities ς i are small, of second order, hence R(u) differs
2
from the natural frequency by a small quantity of second order.
This implies that R(u)has a stationary value in the vicinity of the
modal vector φ ( r ) .
The most important property of Rayleigh’s quotient is that it

shows a minimum value in the neighborhood of the
fundamental mode, i.e. when r=1.
( )
n 2
1+ ∑
ς iωni
( ) > 1 (47)
ωn1
ωni
R (u) = ω = ω 2 2
n1
i =2
n
, since ωn1
1 + ∑ς i
2
i=2
Then each term in the numerator is greater than the corresponding
one in the denominator. Hence, it follows that
R(u) = ω 2 ≥ ωn21 (48)
i.e., Rayleigh’s quotient provides an upper bound to the first

(lowest) natural frequency of the undamped MDOF system.
Clearly, the equality holds above if one selects u = c1φ (1) ; c1 ≠ 0 .
Closure
Rayleigh’s energy method is generally used when one is
interested in a quick (but particularly accurate) estimate of the
fundamental natural frequency of a continuous system, and for
which a solution to the whole eigenvalue problem cannot be
readily obtained. The method is based on the fact that the natural
frequencies have stationary values in the neighborhood of the
natural modes.
In addition, Rayleigh’s quotient provides an upper bound to the
first (lowest) natural frequency. The engineering value of this
approximation can hardly be overstated. Rayleigh’s energy
method is the basis for the numerical computing of eigenvectors
and eigenvalues as will be seen later.
3 ORIGIN  1
N  5 number of DOFS mass  1 stiff  10
Example for estimation of first natural frequency 2
d
M  identity ( N) mass M U  K U = 0
2
dt
 2 1 0 0 0  U3
 1 
U1 U2 U4 U5
2 1 0 0
 
K  stiff  0 1 2 1 0  K K K K K K
 0 0 1 2 1 
  M M M M M
 0 0 0 1 2 
Prepared by
Lecturer Luis San
Andres for ME617
course
Estimate for first natural frequency - Use Rayleigh-Ritz method:
ASSUME T
u  ( .5 .75 1 .75 .5 ) 1
calculate Rayleigh's quotient

uT K u1  1
R ( u) 
uT Mu1  1
 n_ap  R ( u)  n_ap  16.903 Estimate
Actual  n  16.369 rad/s

1
  
% difference =  1  n_ap  100  3.262
 n 
 1 
1.5
0.5
0
0
0 1 2 3 4 5 6
M1
u (approx)
Estimate for second natural frequency - Use Rayleigh-Ritz method:
ASSUME T
u  ( .5 .5 0 .5 .5 ) 1

uT K u1  1
R ( u) 
uT Mu1  1
 n_ap  R ( u)
 n_ap  31.623 Estimate

2
  
% difference =  1  n_ap  100  1.11  10 14
 n 
 2 
1.5
0
0
1
0 1 2 3 4 5 6
M2
u (approx)
3 ORIGIN  1
N  5 number of DOFS mass  1 stiff  10
Example for estimation of first natural frequency 2
d
M  identity ( N) mass M U  K U = 0
2
dt
 2 1 0 0 0 
 1  U1 U2 U3 U4 U5
2 1 0 0
 
K  stiff  0 1 2 1 0  K K K K K
 0 0 1 2 1  M
  M M M M
 0 0 0 1 1 
Estimate for first natural frequency - Use Rayleigh-Ritz method:
ASSUME T
u  ( .1 .3 .5 .7 1 ) 1

uT K u1  1
R ( u) 
uT Mu1  1
 n_ap  R ( u)  n_ap  10.935 Estimate

1
  
% difference =  1  n_ap  100  21.485
 n 
 1 
1.5
0.5
0
0
0 1 2 3 4 5 6
M1
u (approx)
Mode Acceleration Method
Recall that the response in physical coordinates is
m
U ( t ) ≈ ∑ φ j q j( t ) , m < n (36)
j =1
where m<n. The procedure is known as the mode displacement

method.
This method, however, fails to give an accurate solution even

when a static load is applied (See Structural Dynamics, by R.
Craig, J. Wiley Pubs, NY, 1981.).
The difficulty is overcome by using the procedure detailed

below. Recall that the system motion is governed by the set of
equations
+ K U = F
MU (4)
(t ) (t )
And, if there are no rigid body modes, i.e. all natural frequencies
are greater than zero, then
U ( t ) = K −1 ( F( t ) − M U
) (51)
−1
where K is a flexibility matrix. From Eq. (36),
m
≈ ∑ φ qj , m < n
U (52)
( j ) (t )
j =1
Hence, Eq, (51) can be written as
m
U ( t ) ≈ K −1F( t ) − K −1M ∑ φ( j ) qj( t ) (53)
j =1
Using the fundamental identity,

1
Kφ ( i ) = ωi2 Mφ ( i ) ⇒ φ = K −1
Mφ ( i )
ωi 2 (i )
Write Eq. (53) as

⎛ φ( j ) ⎞
m
U ( t ) ≈ K F( t ) − ∑ ⎜ 2 ⎟ qj( t )
−1
⎜
j =1 ⎝ ω j ⎠
⎟ (54)
Note that U S = K −1F( t ) (55)
is the displacement response vector due to a “pseudo-static” force

F(t), i.e. without the system inertia accounted for. Hence write Eq.
(54), as
m⎛ φ( j ) ⎞
U ( t ) ≈ U s ( t ) − ∑ ⎜ 2 ⎟ qj( t )
⎜
j =1 ⎝ ω j ⎠
⎟ ; m<n (56)
The second term above can be thought as the “inertia induced

response.”
Example: Consider the case of force excitation with frequency

Ω ≠ ωn j and acting for very long times. The EOMs in physical
space are:
+ K U = F cos ( Ωt )
MU P
With a little damping, the steady state periodic response in modal
coordinates is
⎡ ⎤
QPj ⎢ 1 ⎥ cos ( Ω t ) (37a)
qj ≈
(
K j ⎢1 − Ω ω
)
2⎥
⎢⎣ nj ⎥⎦
Recall that, using the mode displacement method, the
response in physical coordinates is:
⎛ ⎡ ⎤⎞
m
⎜ QPj 1
U≈∑ φ j ⎢ ⎥ ⎟ cos ( Ω t ) (38)
⎜ Kj
( )
2 ⎟
j =1 ⎜
⎢1 − Ω ω ⎥⎟
⎝ ⎢⎣ nj ⎥⎦ ⎠
From each of the modal responses,

QPj ⎡ ⎤
1
qj ≈
Kj
( −Ω ) ⎢
2
⎢1 − ( Ω ω ) ⎥
2
⎥ cos ( Ω t ) ;
⎣ j ⎦
− qjQPj ⎛ Ω 2 ⎞ ⎡ 1
⎤
≈ ⎜⎜ 2 ⎟⎟ ⎢ ⎥ cos ( Ω t ) (57)
ωj
2
K j ⎝ ω j ⎠ ⎢1 − ( Ω ω ) ⎥
2
⎣ j ⎦
Since K j = ω j M j ; then using the mode acceleration method, the

2
response is
⎧ QPj ⎛ Ω2 ⎞ ⎡ ⎤ ⎫⎪
⎪ m
1
U ≈ ⎨ U SP + ∑ φ j ⎜⎜ 2 ⎟⎟ ⎢ ⎥ cos ( Ωt ) (58
2 ⎬
⎪⎩ j =1 Kj ⎝ ω j ⎠ ⎢⎣1 − ( Ω ω j ) ⎥⎦ ⎪⎭
)
−1
where the pseudo-static response is U SP = K FP . Now, in the
limit, as the excitation frequency decreases, i.e., as Ω → 0 , the
second term in Eq. (58) above disappears, and hence the physical
response becomes:
U = USP = K −1 FP (59)
which is the exact response, regardless of the number of modes

chosen. Hence, the mode acceleration method is more accurate
than the mode displacement method. Known disadvantages
include more operations.
Finding the flexibility matrix is, in actuality, desirable. In

particular, if derived from measurements, the flexibility matrix is
easier to determine than the stiffness matrix.
STEP FORCED RESPONSE of Undamped 2-DOF ORIGIN := 1
mechanical system
Dr. Luis San Andres (c) MEEN 363, 617 February 2008
The undamped equations of motion are: d2 (1)

M⋅ X + K ⋅ X = Fo
dt 2
where M,K are matrices of inertia and stiffness coefficients, and X, V=dX/dt, d2X/dt2 are the
vectors of physical displacement, velocity and acceleration, respectively.
The FORCED undamped response to the initial conditions, at t=0, Xo,Vo=dX/dt, follows:
========================================================================
The equations of motion are:
⎛ M11 M12 ⎞ d2 ⎛ x1 ⎞ ⎛ K11 K12 ⎞ ⎛ x1 ⎞ ⎛ F1o ⎞

⎜ ⎟ ⋅ 2⎜ ⎟ +⎜ ⎟ ⋅⎜ ⎟ = ⎜ ⎟ (2)
⎝ 21
M M22 ⎠ dt ⎝ x2 ⎠ ⎝ 21 22 ⎠ ⎝ x2 ⎠ ⎝ F2o ⎠
K K
1. Set elements of inertia and stiffness matrices DATA FOR problem
⎛ 100 0 ⎞ ⋅ kg K := ⎛ 2 ⋅ 10 6 −1 ⋅ 10 6 ⎞ N
M := ⎜ ⎟ ⎜ ⎟⋅ m
⎝ 0 50 ⎠
⎝ −1 ⋅ 10
6
2 ⋅ 10 ⎠
6 n := 2 # of DOF
Note M and K are symmetric matrices
⎛0⎞
Xo := ⎜ ⎟ ⋅ m Vo := ⎜
⎛ 0.0 ⎞ ⋅ m
initial conditions ⎟
⎝0⎠ ⎝ 0 ⎠ sec
Applied force vector:

Fo := ⎜
⎛ 10000 ⎞ ⋅ N
⎟
⎝ −5000 ⎠
2. Find eigenvalues (undamped natural frequencies) and eigenvectors

Set determinant of system of eqns = 0
(
Δ = ⎡⎣ K11 − M11 ⋅ ω
2
) ⋅ (K 22 − M22 ⋅ ω ) − (K
2
12 − M12 ⋅ ω ) ⋅ (K
2
21 − M21 ⋅ ω
2
) ⎤⎦ = (2a)
0
4 2
( 2
Δ = a ⋅ ω + b ⋅ ω + c = a ⋅ λ + b ⋅ λ + c = 0with λ = ω ) 2 (2b)
where the a := M1 , 1 ⋅ M2 , 2 − M1 , 2 ⋅ M2 , 1
coefficients
are: b := K1 , 2 ⋅ M2 , 1 − K1 , 1 ⋅ M2 , 2 − K2 , 2 ⋅ M1 , 1 + K2 , 1 ⋅ M1 , 2 (2c)
c := K1 , 1 ⋅ K2 , 2 − K1 , 2 ⋅ K2 , 1
The roots of equation (2b) are:
⎡ −b − ( b 2 − 4 ⋅ a ⋅ c) .5 ⎤ ⎡ −b + ( b 2 − 4 ⋅ a ⋅ c) .5 ⎤
λ 1 :=
⎣ λ 2 :=
⎦⎣ ⎦ (3)
2⋅a 2⋅a
also known as eigenvalues. The natural frequencies follow as:

j := 1 .. n ⎛ 112.6 ⎞ rad
( ) .5
ω j := λ j
ω
ω=⎜ ⎟ (4)
f := ⎝ 217.53 ⎠ sec
2⋅π
f=⎜
⎛ 17.92 ⎞ Hz
⎟
⎝ 34.62 ⎠
( ) ( )
Note that: Δ ω 1 = Δ ω 2 = 0
For each eigenvalue, the eigenvectors (natural modes) are

j := 1 .. n
Set arbitrarily first element of vector = 1
⎡ 1 ⎤
⎢ ⎥
a j := ⎢ K1 , 1 − M1 , 1 ⋅ λ j ⎥
⎢ −( K1 , 2 − M1 , 2 ⋅ λ j) ⎥
⎣ ⎦
a1 = ⎜
⎛ 1 ⎞ a =⎛ 1 ⎞
⎟ 2 ⎜ ⎟ (5)
MODAL matrix 〈 j〉
⎝ 0.73 ⎠ ⎝ −2.73 ⎠
A := a j
A is the matrix of eigenvectors (undamped ⎛ 1 1 ⎞
modal matrix): each column corresponds to an A = ⎜ ⎟
eigenvector ⎝ 0.73 −2.73 ⎠
Plot the mode shapes:
Aj , 1
0.37
Aj , 2
2.73
j
DOF
mode 1
mode 2
3. Modal transformation of physical equations to (natural) modal coordinates
Using transformation: X = A⋅q (6)
EOMs (1) become uncoupled in modal space:
d2 (7)
Mm ⋅ q + K m ⋅ q = Qm
dt 2
Qm = A ⋅ Fo
T (8)
with modal force vector:
and initial conditions (modal displacement=q and modal velocity dq/dt=s)
qo = Mm
−1
( T
⋅ A ⋅ M ⋅ Xo ) so = Mm
−1
( T
⋅ A ⋅ M ⋅ Vo ) (9)
The modal responses are of the form: k=1....n

so Qm
(
qk = qo ⋅ cos ω k ⋅ t +
k
) ωk
k
(
⋅ sin ω k ⋅ t +
Km
) k
( ( ))
⋅ 1 − cos ω k ⋅ t ω ≠ 0
k
(10a)
k, k
for an elastic mode
OR
Qm
qk = qo + so ⋅ t +
1
⋅
k
⋅t
2 for ω k = 0 (10b)
k k 2 Mm for a rigid body mode
k, k
And, the response in the physical coordinates is given

X ( t) = A ⋅ q ( t) (5)
by the superposition of the modal responses, i.e.
=== CHECK ========================================================
Verify the orthogonality properties of the natural mode shapes
T ⎛ 126.79 −2.24 × 10 − 14 ⎞
Mm := A ⋅ M ⋅ A Mm = ⎜ ⎟ kg
−
⎝ −1.58 × 10 ⎠
14
473.21
T
Km := A ⋅ K ⋅ A ⎛ 1.61 × 10 6 3.18 × 10 − 10 ⎞ N
Km = ⎜ ⎟m
− 10
⎝ 3.51 × 10 2.24 × 10 7 ⎠
ω=⎜
⎛ 112.6 ⎞ s -1
⎟
⎝ 217.53 ⎠
=========================================================================================
4. Find Modal and Physical Response for given initial condition and
Constant Force vector
Recall the vectors of initial conditions ⎛0⎞

Xo = ⎜ ⎟ m
⎛0⎞ m
Vo = ⎜ ⎟
⎝0⎠ ⎝0⎠ s
and Constant forces:

⎛ 1 × 10 4 ⎞ NDATA FOR problem being analyzed:
Fo = ⎜ ⎟m m
⎝ −5 × 10 ⎠
3
4.a Find initial conditions in modal coordinates (displacement = q, velocity = s)
Set inverse of modal mass matrix Ainv := Mm

−1 ( T
⋅ A ⋅M )
qo := Ainv ⋅ Xo so := Ainv ⋅ Vo
⎛0⎞
qo = ⎜ ⎟ m
⎛0⎞
so = ⎜ ⎟ m s
-1
⎝0⎠ ⎝0⎠
4.b Find Modal forces:
T
Qm := A ⋅ Fo ⎛ 6.34 × 10 3 ⎞
Qm = ⎜ ⎟N
⎝ 2.37 × 10 ⎠
4
4.c Build Modal responses:
so Qm
q1 ( t) := qo ⋅ cos ( ω 1 ⋅ t) + ⋅ sin ( ω 1 ⋅ t) + ⋅ ( 1 − cos ( ω 1 ⋅ t) )
1 1
1 ω1 Km
1 , 1
so Qm
2 2
2 ω2 Km
2 , 2
for plots:
4.d Build Physical responses: X ( t) := a1 ⋅ q1 ( t) + a2 ⋅ q2 ( t) 6
Tplot :=
f1
4.e Graphs of Modal and Physical responses:
Response in modal coordinates

0.01
0.005
0
0 0.056 0.11 0.17 0.22 0.28 0.33
time (s)
q1
q2
Response in physical coordinates

0.01
0.005
0.005
0.01
0 0.056 0.11 0.17 0.22 0.28 0.33
time (s)
x1
x2
5. Interpret response: analyze results, provide recommendations

S-S displacement
−1 ⎛ 5 × 10 − 3 ⎞
K ⋅ Fo = ⎜ ⎟m
Recall natural frequencies & periods ⎝ 0 ⎠
f=⎜
⎛ 17.92 ⎞ Hz 1
=
⎛ 0.056 ⎞
⎟ f ⎜⎝ 0.029 ⎟⎠
s
⎝ 34.62 ⎠ A=⎜
⎛ 1 1 ⎞
⎟
⎝ 0.73 −2.73 ⎠
ω=⎜
⎛ 112.6 ⎞ s -1
⎟
⎝ 217.53 ⎠
RIGID BODY MODE
STEP FORCED RESPONSE of Undamped 2-DOF ORIGIN := 1
mechanical system
Dr. Luis San Andres (c) MEEN 363, 617 February 2008
The undamped equations of motion are: d2 (1)

M⋅ X + K ⋅ X = Fo
dt 2
where M,K are matrices of inertia and stiffness coefficients, and X, V=dX/dt, d2X/dt2 are the
vectors of physical displacement, velocity and acceleration, respectively.
The FORCED undamped response to the initial conditions, at t=0, Xo,Vo=dX/dt, follows:
========================================================================
The equations of motion are: WITH RIGID BODY
MODE
⎛ M11 M12 ⎞ d2 ⎛ x1 ⎞ ⎛ K11 K12 ⎞ ⎛ x1 ⎞ ⎛ F1o ⎞
⎜ ⎟ ⋅ 2⎜ ⎟ +⎜ ⎟ ⋅⎜ ⎟ = ⎜ ⎟ (2)
⎝ 21
M M22 ⎠ dt ⎝ x2 ⎠ ⎝ 21 22 ⎠ ⎝ x2 ⎠ ⎝ F2o ⎠
K K
1. Set elements of inertia and stiffness matrices DATA FOR problem
⎛ 100 0 ⎞ ⋅ kg K := ⎛ 1 ⋅ 10 6 −1 ⋅ 10 6 ⎞ N
M := ⎜ ⎟ ⎜ ⎟⋅ m
⎝ 0 50 ⎠
⎝ −1 ⋅ 10
6
1 ⋅ 10 ⎠
6 n := 2 # of DOF
Note M and K are symmetric matrices
⎛0⎞
Xo := ⎜ ⎟ ⋅ m Vo := ⎜
⎛ 0.0 ⎞ ⋅ m
initial conditions ⎟
⎝0⎠ ⎝ 0 ⎠ sec
Applied force vector:

Fo := ⎜
⎛ 1000 ⎞ ⋅ N
⎟
⎝ −980 ⎠
2. Find eigenvalues (undamped natural frequencies) and eigenvectors

Set determinant of system of eqns = 0
(
Δ = ⎡⎣ K11 − M11 ⋅ ω
2
) ⋅ (K 22 − M22 ⋅ ω ) − (K
2
12 − M12 ⋅ ω ) ⋅ (K
2
21 − M21 ⋅ ω
2
) ⎤⎦ = (2a)
0
4 2
( 2
Δ = a ⋅ ω + b ⋅ ω + c = a ⋅ λ + b ⋅ λ + c = 0with λ = ω ) 2 (2b)
where the a := M1 , 1 ⋅ M2 , 2 − M1 , 2 ⋅ M2 , 1
coefficients
are: b := K1 , 2 ⋅ M2 , 1 − K1 , 1 ⋅ M2 , 2 − K2 , 2 ⋅ M1 , 1 + K2 , 1 ⋅ M1 , 2 (2c)
c := K1 , 1 ⋅ K2 , 2 − K1 , 2 ⋅ K2 , 1
, , , ,
The roots of equation (2b) are:

⎡ −b − ( b 2 − 4 ⋅ a ⋅ c) .5 ⎤ ⎡ −b + ( b 2 − 4 ⋅ a ⋅ c) .5 ⎤
λ 1 :=
⎣ λ 2 :=
⎦⎣ ⎦ (3)
2⋅a 2⋅a
also known as eigenvalues. The natural frequencies follow as:

j := 1 .. n
( ) .5 ω = ⎛⎜ ⎞ rad
0
ω j := λ j ⎟
ω ⎝ 173.21 ⎠ sec
(4)
f :=
2⋅π
f = ⎛⎜
0 ⎞
⎟ Hz
⎝ 27.57 ⎠
( ) ( )
Note that: Δ ω 1 = Δ ω 2 = 0
For each eigenvalue, the eigenvectors (natural modes) are

j := 1 .. n
Set arbitrarily first element of vector = 1
⎡ 1 ⎤
⎢ ⎥
a j := ⎢ K1 , 1 − M1 , 1 ⋅ λ j ⎥
⎢ −( K1 , 2 − M1 , 2 ⋅ λ j) ⎥
⎣ ⎦
a1 = ⎛⎜ ⎟⎞ a2 = ⎛⎜ ⎞
1 1
⎟ (5)
MODAL matrix 〈〉
⎝1⎠ ⎝ −2 ⎠
A j := a j
A = ⎛⎜ ⎞
A is the matrix of eigenvectors (undamped 1 1
⎟
modal matrix): each column corresponds to an ⎝1 −2 ⎠
eigenvector
Plot the mode shapes:
Aj , 1
0
Aj , 2
2
j
DOF
mode 1
mode 2
3. Modal transformation of physical equations to (natural) modal coordinates
Using transformation: X A (6)

Using transformation: X = A⋅q (6)
EOMs (1) become uncoupled in modal space:
d2 (7)
Mm ⋅ q + K m ⋅ q = Qm
2
dt
Qm = A ⋅ Fo
T (8)
with modal force vector:
and initial conditions (modal displacement=q and modal velocity dq/dt=s)
(
qo = Mm − 1 ⋅ A ⋅ M ⋅ Xo
T
) so = Mm − 1 ⋅ A ⋅ M ⋅ Vo ( T
) (9)
The modal responses are of the form: k=1....n

so Qm
(
qk = qo ⋅ cos ω k ⋅ t +
k
) ωk
k
(
⋅ sin ω k ⋅ t +) Km
k
( ( ))
⋅ 1 − cos ω k ⋅ t ω ≠ 0
k
(10a)
k, k
for an elastic mode
OR
Qm
qk = qo + so ⋅ t + ⋅
1 k
⋅ t2 for ω k = 0 (10b)
k k 2 Mm for a rigid body mode
k, k
And, the response in the physical coordinates is given

X ( t) = A ⋅ q ( t) (5)
by the superposition of the modal responses, i.e.
=== CHECK ========================================================
Verify the orthogonality properties of the natural mode shapes
Mm = ⎛⎜ ⎞ kg
T 150 0
Mm := A ⋅ M ⋅ A ⎟
⎝ 0 300 ⎠
T
Km := A ⋅ K ⋅ A
⎛0 ⎞N
ω = ⎛⎜ ⎞ -1
0 0
Km = ⎜ ⎟ ⎟s
⎝0 9 × 10 6 ⎠ m ⎝ 173.21 ⎠
=========================================================================================
4. Find Modal and Physical Response for given initial condition and
Constant Force vector
Xo = ⎛⎜ ⎟⎞ m Vo = ⎛⎜ ⎟⎞
Recall the vectors of initial conditions 0 0 m
⎝0⎠ ⎝0⎠ s
dC t tf
and Constant forces:
⎛ 1 × 10 3 ⎞ N
DATA FOR problem being analyzed:
Fo = ⎜ ⎟m m
⎝ −980 ⎠
4.a Find initial conditions in modal coordinates (displacement = q, velocity = s)
Set inverse of modal mass matrix Ainv := Mm − 1 ⋅ A ⋅ M ( T )

Ainv = ⎜
⎛ 0.67 0.33 ⎞
⎟
⎝ 0.33 −0.33 ⎠
qo := Ainv ⋅ Xo so := Ainv ⋅ Vo
qo = ⎛⎜ ⎞⎟ m so = ⎛⎜ ⎟⎞ m s -1
0 0
⎝0⎠ ⎝0⎠
4.b Find Modal forces:
T
Qm := A ⋅ Fo ⎛ 20 ⎞
Qm = ⎜ ⎟ N
⎝ 2.96 × 10 3
⎠
4.c Build Modal responses:
Qm
1 t2 response for rigid body mode
q1 ( t) := qo + so ⋅ t + ⋅
1 1 Mm 2
1 , 1
so Qm
2 2
2 ω2 Km
2 , 2
for plots:
4.d Build Physical responses: X ( t) := a1 ⋅ q1 ( t) + a2 ⋅ q2 ( t) 6
Tplot :=
f2
4.e Graphs of Modal and Physical responses:
Response in modal coordinates

0.004
0.003
0.002
0.001
0
0 0.022 0.044 0.065 0.087 0.11 0.13 0.15 0.17 0.2 0.22
time (s)
time (s)
q1
q2
Response in physical coordinates
0.002
0.002
0 0.022 0.044 0.065 0.087 0.11 0.13 0.15 0.17 0.2 0.22
time (s)
x1
x2
5. Interpret response: analyze results, provide recommendations

S-S displacement - NONE
Recall natural frequencies & periods
f = ⎛⎜ ⎞ Hz A = ⎛⎜ ⎞ ω = ⎛⎜ ⎞ s -1
0 1 1 0
⎟ ⎟ ⎟
⎝ 27.57 ⎠ ⎝1 −2 ⎠ ⎝ 173.21 ⎠

HD 7 Modal Analysis Undamped MDOF

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

HD 7 Modal Analysis Undamped MDOF

Transféré par

Droits d'auteur :

Formats disponibles

ME617 - Handout 7

(Undamped) Modal Analysis of

where U, U, and U  are the vectors of generalized displacement,

The solution of Eq. (1) is uniquely determined once initial

In most cases, i.e. conservative systems, the inertia and stiffness

energy (T) and potential energy (V) in a conservative system are

In MDOF systems, a natural state implies a certain

The modeling of a n-DOF mechanical system leads to a set of n-

In the analysis below, for a proper choice of generalized

The solution to the homogenous Eq. (5) is simply

which denotes a periodic response with a typical frequency ω .

Note that Eq. (6) is a simplification of the more general solution

U = φ e s t with s = i ω and where i = −1 (8)

and since cos(ω t −θ ) ≠ 0 for most times, then

Eq. (10) is usually referred as the standard eigenvalue

Eq.(9) is a set of n-homogenous algebraic equations. A nontrivial

Eq. (12) is known as the characteristic equation of the system.

The ω’s are known as the natural frequencies of the system. In

b) If M and K are positive definite, then

c) If K is semi-positive definite, then

Note that each of the natural frequencies satisfies Eq. (9).

[ − M λi + K ] φ(i ) = 0, i =1,...n (14)

The n-elements of an eigenvector are real numbers (for

The actual value of the elements in the vector is entirely

For MDOF systems with a large number of degrees of freedom,

A this time, however, let’s assume the set of eigenpairs

Properties of natural modes

⎡⎣ − M ωi2 + K ⎤⎦ φ ( i ) = 0, i =1,...n . (15)

Consider two different modes, say mode-j and mode-k, each

ω 2j M φ ( j ) = K φ ( j ) and ωk2 M φ ( k ) = K φ ( k ) (16)

Now, perform some matrix manipulations. The products

transpose of a scalar is the number itself. Hence,

for symmetric systems. Thus, Eqs. (17) are rewritten as

Subtract (b) from (a) above to obtain

if ω j ≠ ωk , i.e. for TWO different natural frequencies; then it

for j = k φT( j ) M φ ( j ) = M j and φT( j ) K φ ( j ) = K j = ω 2j M j (20)

where Kj and Mj are known as the j-modal stiffness and j-modal

Define a modal matrix Φ has as its columns each of the

and the modal properties are written as

The eigenvector set φ k=1,…n is linearly independent. Hence,

Consider the modal transformation U(t ) = Φ q(t ) (24)3

which offers no advantage in the analysis. However, premultiply

and using the properties of the natural modes,

The set of q’s are known as modal or natural

Thus, the major advantage of the modal transformation (24)

The unique solution of Eqs. (28) needs of initial conditions

Using the modal transformation,  = Φq

However, Eq. (28) requires of the inverse of modal matrix Φ ,

Comparing eqs. (28) and (29a) it follows that

The solution of ODEs M j qj + K j q j = Q j with initial

Free response in modal coordinates

with solutions, for an elastic mode

; and for a rigid body mode

Forced response in modal coordinates

For step-loads, QS j , the modal equations are

and; for an elastic mode, ωn ≠ 0 , j

Note that if Ω = ωn j , a resonance appears that will lead to system

For arbitrary-loads Q j , the modal response is

One important question follows: are all the modal

If m<n, then how many modes are to be included to ensure the

Let’s assume there is a little damping; hence, the steady state

Recall that K j = ωn j M j = φ ( j ) K φ ( j ) and QPj = φ ( j ) FP

However, nowadays the engineer in a hurry prefers to dump the

at a fixed excitation frequency Ω. Brute force substitutes beauty

where U, U, and U are the vectors of generalized displacement,

Using the modal transformation, = Φq

The solution of ODEs M j qj + K j q j = Q j with initial