Vibration Analysis
Flexibility Formulation
Earlier the discussion was based on stiffness
matrix. But sometimes, it is more convenient to
express the elastic properties of the structural
system by means of the flexibility matrix rather
than the stiffness matrix.
[a] = [k]-1
Referring again to equation (4): k 2 mv~ 0
Multiplying by (1/ω2)[a] results into
1
2 I a m vˆ 0 (16)
2
Vibration Analysis
As before, this set of homogeneous equations
can have a nonzero solution only if the
determinant of the square matrix vanishes;
1
I a m 0 (17)
2
The equation is solved similarly to that of
equation (5), and the mode shapes are
evaluated in the similar fashion. The basic
difference is that the roots of the equation (17)
represent the reciprocals of the frequency
squared values rather than the frequency
squared.
3
Orthogonality Conditions
Basic Conditions:
Certain properties of the free-vibration mode shapes ϕn
are very useful in structural-dynamics analyses. These
properties, known as orthogonality conditions, can be
shown by the use of Betti’s law.
Considering two different modes of vibration of a
structural system, as shown in the figure:
Vibration mode shapes and resulting inertial forces. 4
Orthogonality Conditions
In the example, considering the lumped-mass system, the
equations of motion for a system in free- vibration is given by
equation (4), viz.,
k m v~ 0
2
n n
k vˆ n n2 m vˆ n (18)
elastic-resisting applied inertia load
Force vector {fs} vector - {fI}
Thus the free vibration motion may be considered to involve
deflections produced by inertia forces acting as applied loads,
as shown in the figure.
5
Orthogonality Conditions
The two vibration modes then could represent
two different applied-load systems and their
resulting dispacements:
Applying Betti’s law,
f Im vˆ n f In vˆ m
T T
(19)
Substituting from (18) for the inertia force,
m vˆ vˆ m vˆ vˆ
2
m m
T
n
2
n n
T
m
vˆ m m vˆ n vˆ n m vˆ m (20)
2
m
T 2
n
T
6
Orthogonality Conditions
Note that the matrix products in equation (20)
are scalars and can be transposed arbitrarily.
Then, the equation becomes,
2
m 2
n vˆ mvˆ 0
m
T
n (21)
If the two frequencies are not the same, it gives
the first orthogonality condition:
vˆ m mvˆ n 0
T
ωm n (22)
7
Orthogonality Conditions
Premultiplying equation (18) by vˆ m T ,
v m k v n v m n mvˆ n
ˆ T
ˆ ˆ 2T
vˆ m m vˆ n
2 T
n
When equation (22) is applied to the right hand
side, it becomes,
vˆ m k vˆ n 0
T
m n (23)
-second orthogonality condition.
8
Orthogonality Conditions
Hence, vibrating shapes are orthogonal with
respect to the stiffness matrix as well with
respect to the mass.
Dividing equations (22) and (23) by any
reference displacement value, then the
conditions will be in terms of dimensionless
mode-shape vectors ϕn.
m m n 0
T
mn (24a)
m k n 0
T
mn (24b)
9
Orthogonality Conditions
Equation (24a) shows that the different mode
shapes are orthogonal with respect to the
mass matrix.
Equation (24b) shows that the different mode
shapes are orthogonal with respect to the
stiffness matrix.
Additional relationships of the mode shapes:
10
Normalizing
The vibration mode amplitudes obtained from
the eigen-problem solution are arbitrary, any
amplitude will satisfy the basic frequency
equation: k 2 mv~ 0 , and only the
resulting shapes are uniquely defined. In the
analysis process discussed earlier, the
amplitude of one degree of freedom (the first,
actually) has been set to unity, and the other
displacements have been determined relative to
this reference value. This is called normalizing
the mode shapes with respect to the specified
coordinate.
11
