CHAPTER 8

MULTI-DEGREE-OF-FREEDOM SYSTEMS
Equations of Motion, Problem Statement,
and Solution Methods

Expected outcome:
1. Students can formulate equations of motion of MDF systems.
2. Students can use method of influence coefficient to formulate mass matrix and
stiffness matrix.
3. Students can reduce number of degrees of freedom by static condensation.

When the system consists of more than one mass, or the mass can displace and/or
rotate in more than one independent directions, more than one degree of freedom are required
to completely describe positions of all mass. The system is considered as a multi-degree-of-
freedom (MDF) system. This chapter will discuss how to derive equations of motion and
related dynamic properties, i.e., mass and stiffness matrices of the system and define the
problem statement.

8.1 Two-story shear building

A shear building is the building whose beams or floor systems are rigid in flexure and
axial deformation of beams and columns are assumed to be very small and neglected, so the
beam-column joints have no rotation. Primary deformation of the structure is double-
curvature flexural deflection of columns and the lateral stiffness of each column is 12EI h3 .
We will use this two-story shear building with a mass lumped at each floor as an example of
2-DOF system to demonstrate formulation of equations of motion.

The equations of motion are formulated by considering equilibrium of forces acting

on each mass. Any of the two approaches can be used

(1) Newton’s second law of motion

(2) D’Alembert’s principle of dynamic equilibrium

8-1
 m2 u2
p2(t) p2(t)
f S2
c2 f D2
a
u1 a f D1
m1 f S1
p1(t) f S1= + p1(t) + = f D1
b
f S1 f b
c1 D1

(a) (b)

Figure 8.1 (a) Two-story shear frame; (b) forces acting on the two masses

8.2 Equations of motion

Newton’s second law of motion

Newton’s second law of motion stated that vector summation of all forces acting on a
body is equal to product of mass and acceleration vector. If we consider only horizontal
direction, then the equation involves scalar magnitude of force and acceleration in horizontal
component:

 F  mu (8.1)

Applying Eq. (8.1) to each floor mass, we obtain two equations for j =1 and 2:

p j  f Sj  f Dj  m j uj or m j uj  f Sj  f Dj  p j  t  (8.2)

The two equations ( j =1 and 2) can be written in matrix form

 m1 0   u1   f D1   f S 1   p1  t  
        (8.3)
0 m2  u2   f D 2   f S 2   p2  t  

or   f D  f S  p  t 
mu (8.4)

where
u  m 0 f  f  p 
u   1 m 1 f D   D1  fS   S1  p   1.
u2  0 m2   fD2   fS 2   p2 

8-2
 u1 and u2 are displacement of mass at story 1 and 2, respectively.

u1 and u2 are acceleration of mass at story 1 and 2, respectively.

m1 and m2 are lumped mass at story 1 and 2, respectively.

f D1 and f D 2 are damping force acting on mass at story 1 and 2, respectively.

f S 1 and f S 2 are structural resisting force acting on mass at story 1 and 2, respectively.

p1 and p2 are external force acting on mass at story 1 and 2, respectively.

Because all beams are assumed to be rigid, the story shear force can be directly related to the
relative displacement between two adjacent stories, called “story drift,”  j .

Vj  k j j (8.5)

12 EI c
where story drift  j  u j  u j 1 and story lateral stiffness k j  
columns h3
.

In addition, E is modulus of elasticity of material of column. I c is second moment of cross-

sectional area of each column (often known as moment of inertia of cross section). h is the
height of each column.

The structural resisting force acting on the first story mass comes from columns
below ( f Sb1 ) and above ( f Sa1 ) the floor (see Fig. 8.1b).

f S 1  f Sb1  f Sa1 (8.6)

f S 1  k1u1  k2  u1  u2  (8.7)

f S 2  k2  u2  u1  (8.8)

 f S 1   k1  k2 k2   u1 
    or f S  ku (8.9)
 f S 2   k2 k2  u2 

The structural resisting force vector f S is related to displacement vector u through

the stiffness matrix k .

8-3
 The damping forces f D1 and f D 2 are derived from story shear V j due to viscous

damping as Eq. (8.10) where story shear V j is equal to product of damping coefficient c j at

story j and relative velocity between two adjacent floors  j .

V j  c j  j (8.10)

where relative velocity, or rate of story drift,  j  u j  u j 1 . From Fig. 8.1b, we have

f D1  c1u1  c2  u1  u2  (8.11)

f D 2  c2  u2  u1  (8.12)

We can write the two damping forces f D1 and f D 2 together in a damping force vector f D as

 f D1  c1  c2 c2   u1  (8.13)

    or f D  cu
 f D 2   c2 c2  u2 

The damping force vector f D and velocity vector u are related by the damping matrix c .

Substituting Eqs.(8.9) and (8.13) into Eq. (8.4), we then have the equations of motion as

  cu  ku  p  t 
mu (8.14)

The two rows of this vector equation represents two ordinary differential equations (ODEs)
governing the displacements u1  t  and u2  t  of the two-story shear frame subjected external

forces p1  t  and p2  t  .

Each equation contains both unknowns u1  t  and u2  t  , so two equations are said to

be “coupled” and they must be solved simultaneously.

Dynamic equilibrium (D’Alembert’s principle)

Alternative to Newton’s law, the equations of motion can be derived by D’Alembert’s

principle of dynamic equilibrium. Consider each mass in the structure. The external force
must be in balance with

8-4
 (1) inertia force (resisting acceleration) acting in the opposite direction to acceleration

(2) damping force (resisting velocity) acting in the opposite direction to velocity and

(3) structural force resisting deformation

fI  fD  fS  p t  (8.15)

where the inertia force vector f I  mu

 . This is just the same as Eq. (8.4).

f I2
p2(t)
f S2
f D2
f I1
p1(t) f S1 f D1

Figure 8.2 Free-body diagrams for D’Alembert’s principle of dynamic equilibrium

Example 8.1a

Formulate the equations of motion for the two-story shear frame shown in Fig. E8.1a

m
p2(t) u2
EI b=
8

h EI c EI c
2m
p1(t) u1
EI b=
8

h 2EI c 2EI c

Figure E8.1a
Solution

The mass and stiffness of each story are

m1  2m m2  m

12  2 EI c  48EI c 12  EI c  24 EI c
k1  2  k2  2 
h3 h3 h3 h3

8-5
The mass and stiffness matrices are

2 0 24 EI c  3 1
m  m  k  1 1 
0 1 h3  

The governing equations for this system without damping is

  ku  p  t 
mu

 2 0  u1  24 EI c  3 1 u1   p1 (t ) 

m     h3  1 1  u    p (t ) 
 0 1  u2    2  2 

Observe that the stiffness matrix is nondiagonal, implying that the two equations are coupled,
and in their present form must be solved simultaneously.

Example 8.1b

Formulate the equations of motion for the two-story shear frame in Fig. E8.1a using
influence coefficients.

Solution
T
The two DOFs of this system are u  u1 u2 .

1. Determine the stiffness matrix. To obtain the first column of the stiffness matrix, we
impose u1  1 and u2  0 . The stiffness influence coefficients are kij (Fig. E8.1b). The forces

necessary at the top and bottom of each story to maintain the deflected shape are expressed in
terms of story stiffnesses k1 and k2 [part (b) of the figure], determined in Example 8.1a:

48EI c 24 EI c
k1  k2 
h3 h3

The two sets of forces in parts (a) and (b) of the figure are one and the same. Thus,

72 EI c 24 EI c
k11  k1  k2  k21  k2  
h3 h3

The second column of the stiffness matrix is obtained in a similar manner by imposing u2  1

with u1  0 . The stiffness influence coefficients are ki2 [part (c) of the figure] and the forces

8-6
necessary to maintain the deflected shape are shown in part (d) of the figure. The two sets of
forces in parts (c) and (d) of the figure are one and the same. Thus,

24 EI c 24 EI c
k12  k2   k22  k2 
h3 h3

With the stiffness influence coefficients determined, the stiffness matrix is

24 EI c  3 1
k
h3  1 1 

u1=1, u2=0
k 21 k2

u1=1

k11 k1 k2

k1
(a) (b)

u2=1, u1=0 u2=1

k22 k2

k12 k2

(c) (d)

Figure E8.1b

2. Determine the mass matrix. With the DOFs defined at the locations of the lumped masses,
the diagonal mass matrix is:
2 0
m  m 
0 1
3. Determine the equations of motion
 2 0  u1  24 EI c  3 1 u1   p1 (t ) 
  ku  p  t 
mu or m     h3  1 1  u    p (t ) 
 0 1  u2    2  2 

8-7
8.3 General approach for linear systems

Discretization

A frame structure can be idealized by an assemblage of elements—beams, columns,

walls—interconnected at nodal points or nodes. Displacements of nodes are degrees of
freedom. A node in a planar two-dimension frame has 3 DOFs—two translations and one
rotation.

u6 u7 u8
u2
Node
u3 u4 u5
u1
Structural
elements

(a) (b)
Figure 8.3 Degrees of freedom: (a) axial deformation included, 18 DOFs; (b) axial
deformation neglected, 8 DOFs

If axial deformations are neglected, the number of DOFs can be reduced because some
translational DOF are equal. The external forces are applied at the nodes which correspond
to the DOFs.

Elastic forces

The elastic forces are related to displacement through stiffness matrix. The stiffness
matrix can be obtained from stiffness influence coefficient kij , which is the force required

along DOF i due to a unit displacement at DOF j and zero displacement at all other DOFs.

For example, the force ki1 ( i  1, 2,...,8 ) are required to maintain the deflected shape

associated with u1  1 and all other u j  0 .

8-8
 f S6 f S7 f S8
f S2

f S3 f S4 f S5
f S1

(a)
k61 k71 k81 k64 k74 k84
k21 k24
u1=1
u4=1
k31 k41 k51 k34 k 44 k54
k11 k14

(b) (c)

Figure 8.4 (a) Stiffness component of frame; (b) stiffness influence coefficients for u1=1; (c)
stiffness influence coefficients for u4=1

The force f Si at DOF i associated with displacement u j ( j  1 to N ) is obtained by

superposition:

f Si  ki1u1  ki 2u2  ...  kiN u N (8.16)

Such equation applies to each of f Si where i  1 to N , so

 f S 1   k11 k12 .. k1N   u1 

 f  k k22 .. k2 N   u2 
 S 2   21
    or f S  ku (8.17)
 :   : : .. :   : 
 f SN   k N 1 kN 2

.. k NN  u N 

where k is the stiffness matrix of the structure.

This approach can be cumbersome for complex structures in order to visualize a

deflected shape with a unit displacement at DOF j and zero displacement at all other DOFs.

8-9
 The direct stiffness method must be used instead. It involves assembling of stiffness
matrices of structural members into the stiffness matrix of the whole system. The appropriate
method should be used for a given problem.

Damping forces

Damping forces are related to velocities of nodes through damping matrix. The
method of damping influence coefficient cij can be used to derive the damping matrix in a

similar manner as stiffness matrix relating elastic forces to displacements.

However, it is impractical to compute the coefficient cij of damping matrix directly

from the size of the structural elements. Instead, damping of a MDF system is usually
specified in term of damping ratio and the corresponding damping matrix can be constructed
accordingly.

Inertia forces

Inertia forces are forces related to acceleration of the mass. An approach to consider
inertia forces acting at nodes is to lump the mass of structural components to nodes.

Inertial forces are related to acceleration at nodes through the mass matrix m . Mass
matrix can be derived using mass influence coefficient mij which is the external force in DOF

i due to unit acceleration along DOF j . For example, the force mi1 ( i  1, 2,...8 ) are required

in various DOF to equilibrate the inertia forces associated with u1  1 and all other uj  0 .

The force at DOF i due to acceleration at various nodes can be obtained by

superposition

f Ii  mi1u1  mi 2u2  ...  miN uN (8.18)

Such inertia forces at all DOFs are written together in the inertia force vector f I , which is

equal to
 f I 1   m11 m12 .. m1N   u1 
 f  m m22 .. m2 N   u2 
 I 2   21
    or f I  mu
 (8.19)
 :   : : .. :  : 
 f IN   mN 1 mN 2

.. mNN  uN 

8 - 10
When lumped-mass model is used, the mass matrix will be diagonal. Rotational inertia
forces at the nodes are neglected, so the mass associated with rotational DOFs are zero.

mij  0 if i  j m jj  m j or 0 (8.20)

Equations of motion

fI  fD  fS  p t  (8.21)

 + cu + ku = p  t 
mu (8.22)

The off-diagonal terms in the coefficient matrices m, c, and k are known as coupling
terms. The coupling in a system also depends on the choice of DOFs.

Example 8.2

A uniform rigid bar of total mass m is supported on two springs k1 and k2 at the two
ends and subjected to dynamic forces shown in Fig. E8.2a. The bar is constrained so that it
can move only vertically in the plane of the paper; with this constraint the system has two
DOFs. Formulate the equations of motion with respect to displacements u1 and u2 of the two
ends as the two DOFs.

Solution

1. Determine the applied forces. The external forces do not act along the DOFs and should
therefore be converted to equivalent forces p1 and p2 along the DOFs (Fig. E8.2b) using
equilibrium equations. This can also be achieved by the principle of virtual displacements.
Thus if we introduce a virtual displacement δu1 along DOF 1, the work done by the applied
forces is

 u1  u1
 W  pt  p (a)
2 L

Similarly, the work done by the equivalent forces is

 W  p1 u1  p2  0  (b)

8 - 11
Because the work done by the two sets of forces should be the same, we equate Eqs. (a) and
(b) and obtain

pt p
p1   (c)
2 L

In a similar manner, by introducing a virtual displacement δu2, we obtain

pt p
p2   (d)
2 L

Figure E8.2

2. Determine the stiffness matrix. Apply a unit displacement u1  1 with u2  0 and identify

the resulting elastic forces and the stiffness influence coefficients k11 and k21 (Fig. E8.2c). By
statics, k11  k1 and k21  0 . Now apply a unit displacement u2  1 with u1  0 and identify

the resulting elastic forces and the stiffness influence coefficients (Fig. E8.2d). By statics,
k12  0 and k22  k2 . Thus the stiffness matrix is

k 0
k 1 (e)
0 k2 

In this case the stiffness matrix is diagonal (i.e., there are no coupling terms) because the two
DOFs are defined at the locations of the springs.

8 - 12
3. Determine the mass matrix. Impart a unit acceleration u1  1 with u2  0 , determine the

distribution of accelerations of (Fig. E8.2e) and the associated inertia forces, and identify
mass influence coefficients (Fig. E8.2f). By statics, m11  m / 3 and m21  m / 6 . Similarly,

imparting a unit acceleration u2  1 with u1  0 , defining the inertia forces and mass

influence coefficients, and applying statics gives m12  m / 6 and m22  m / 3 . Thus the mass

matrix is

m 2 1
m (f)
6 1 2 

The mass matrix is coupled, as indicated by the off-diagonal terms, because the mass is
distributed and not lumped at the locations where the DOFs are defined.

4. Determine the equations of motion. Substituting Eqs. (c)-(f) in Eq. (8.14) with c  0 gives

m  2 1  u1   k1 0  u1   pt / 2    p / L  

    
6 1 2  u2   0 k2  u2   pt / 2    p / L   (g)

The two differential equations are coupled because of mass coupling due to the off-diagonal
terms in the mass matrix.

Example 8.3

Formulate the equations of motion of the system of Fig. E8.2a with the two DOFs
defined at the center of mass O of the rigid bar: translation ut and rotation uθ (Fig. E8.3a).

Solution

1. Determine the stiffness matrix. Apply a unit displacement ut  1 with u  0 and identify

the resulting elastic forces and ktt and k t (Fig. E8.3b). By statics, ktt  k1  k2 and

k t  (k2  k1 ) L / 2 . Now, apply a unit rotation u  1 with ut  0 and identify the resulting

elastic forces and kt and k (Fig. E8.3c). By statics, kt  (k2  k1 ) L / 2 and

k  (k1  k2 ) L2 / 4 . Thus the stiffness matrix is

8 - 13
  k1  k2  k2  k1  L / 2 
k (a)
 k2  k1  L / 2  k1  k2  L2 / 4

Observe that now the stiffness matrix has coupling terms because the DOFs chosen are not
the displacements at the locations of the springs.

Figure E8.3

2. Determine the mass matrix. Impart a unit acceleration ut  1 with u  0 , determine the

acceleration distribution (Fig. E8.3d) and the associated inertia forces, and identify mtt and

m t (Fig. E8.3e). By statics, mtt  m and m t  0 . Now impart a unit rotational acceleration

u  1 with ut  0 , determine the resulting accelerations (Fig. E8.3f) and the associated

inertia forces, and identify mt and m (Fig. E8.3g). By statics, mt  0 and m  mL2 /12 .

8 - 14
Note that m  I O , the moment of inertia of the bar about an axis that passes through O and

is perpendicular to the plane of rotation. Thus the mass

matrix is
m 0 
m 2  (b)
 0 mL /12 

Now the mass matrix is diagonal (i.e., it has no coupling terms) because the DOFs of this
rigid bar are defined at the mass center.
T T
3. Determine the equations of motion. Substituting u  ut u p  pt p , and Eqs. (a)

and (b) in Eq. (8.14) gives

m 0  ut   k1  k2  k2  k1  L / 2  ut   pt 
    (c)
 0 mL2 /12 
  u   k2  k1  L / 2  k1  k2  L2 / 4 u   p 

The two differential equations are now coupled through the stiffness matrix.

We should note that if the equations of motion for a system are available in one set of
DOFs, they can be transformed to a different choice of DOF. This concept is illustrated for
the system of Fig. E8.2a. Suppose that the mass and stiffness matrices and the applied force
T
vector for the system are available for the first choice of DOF, u  u1 u2 . These
T
displacements are related to the second set of DOF u  ut u , by

u1  1  L / 2  ut 
    or u = au (d)
u2  1 L / 2  u 

where a denotes the coordinate transformation matrix. The stiffness and mass matrices and
the applied force vector for the u DOFs are given by

k  aT ka m = aT ma p = aT p (e)

Substituting for a from Eq. (d) and for k, m, and p from Example 8.2 into Eq. (e) leads to k
and m , which are identical to Eqs. (a) and (b) and to the p in Eq. (c).

8 - 15
Example 8.4

A massless cantilever beam of length L supports two lumped masses mL/2 and mL/4
at the midpoint and free end as shown in Fig. E8.4a. The flexural rigidity of the uniform
beam is EI. With the four DOFs chosen as shown in Fig. E8.4b and the applied forces p1(t)
and p2(t), formulate the equations of motion of the system. Axial and shear deformations in
the beam are neglected.

p2(t) p1(t) u2 u1
u4 u3
(a) EI EI (b)
mL/2 mL/4 Element (1) Element (2)

L/2 L/2 Node (2) Node (1)

u1=1, u2=u3=u4=0 k11 u3=1, u1=u2=u4=0

k 21 k31 k23 k13
k 41 u1=1 k43 k33 u3=1
(c) (g)

96EI / L3
24EI / L2
24EI / L2 24EI / L 2 8EI / L
(d) (h)
4EI / L
96EI / L3 24EI / L2

u2=1, u1=u3=u4=0 u4=1, u1=u2=u3=0

k 22
k 42 k12 k24 k14
k 32 k44 u4=1
(e) u2=1 k 34 (i)

96EI / L3
24EI / L 2
24EI / L2 24EI / L 2 8EI / L
(f) (j)
4EI / L
96EI / L3 24EI / L2
3
96EI / L
24EI / L2
24EI / L 2 24EI / L2 4EI / L

8EI / L
96EI / L3 24EI / L2

Figure E8.4

8 - 16
Solution

The beam consists of two beam elements and three nodes. The left node is constrained and
each of the other two nodes has two DOFs (Fig. E8.4b). Thus, the displacement vector
T
u  u1 u2 u3 u4 .

1. Determine the mass matrix. With the DOFs defined at the locations of the lumped masses,
the diagonal mass matrix is given by Eq. (8.19):

 mL / 4 
 mL / 2 
m 
 0  (a)
 
 0

2. Determine the stiffness matrix. Several methods are available to determine the stiffness
matrix. Here we use the direct equilibrium method based on the definition of stiffness
influence coefficients.
To obtain the first column of the stiffness matrix, we impose u1  1 and

u2  u3  u4  0 . The stiffness influence coefficients are ki1 (Fig. E8.4c). The forces necessary

at the nodes of each beam element to maintain the deflected shape are determined from the
beam stiffness coefficients (Fig. E8.4d). The two sets of forces in figures (c) and (d) are one
and the same. Thus k11  96 EI / L3 , k21  96 EI / L3 , k31  24 EI / L2 , and k41  24 EI / L2 .

The second column of the stiffness matrix is obtained in a similar manner by

imposing u2  1 with u1  u3  u4  0 . The stiffness influence coefficients are ki2 (Fig. E8.4e)

and the forces on each beam element necessary to maintain the imposed displacements are
shown in Fig. E8.4f. The two sets of forces in figures (e) and (f) are one and the same. Thus
k12  96 EI / L3 , k32  24 EI / L2 , k22  192 EI / L3 , and k42  0 .

The third column of the stiffness matrix is obtained in a similar manner by imposing
u3  1 with u1  u2  u4  0 . The stiffness influence coefficients ki3 are shown in Fig. E8.4g

and the nodal forces in Fig. E8.4h. Thus k13  24 EI / L2 , k23  24 EI / L2 , k33  8EI / L , and

k43  4 EI / L .

The fourth column of the stiffness matrix is obtained in a similar manner by imposing
u4  1 with u1  u2  u3  0 . The stiffness influence coefficients ki4 are shown in Fig. E8.4i,

8 - 17
and the nodal forces in Fig. E8.4j. Thus k14  24 EI / L2 , k34  4 EI / L , k24  0 , and

k44  16 EI / L .

With all the stiffness influence coefficients determined, the stiffness matrix is

 12 12 3L 3L 

 0 
8EI 12 24 3L
k 3  (b)
L  3L 3L L2 L2 / 2 
 2 
 3L 0 L / 2 2 L 
2

3. Determine the equations of motion. The governing equations are

  ku  p  t 
mu

where m and k are given by Eqs. (a) and (b), and p  t   p1  t  p2  t  0 0 .

T

Example 8.5

Derive the equations of motion of the beam of Example 8.4 (also shown in Fig. E8.5a)
expressed in terms of the displacements u1 and u2 of the masses (Fig. E8.5b).

f S1=1, f S2=0 f S1=1

p2(t) p1(t) f 21
EI EI f 11
(a) (c)
mL/2 mL/4

L/2 L/2
f S2=1, f S1=0
u2 u1 f S2=1
f 12
(b) (d)
Element (1) Element (2)
Node (2) Node (1)
f 22

Figure E8.5

8 - 18
Solution

This system is the same as that in Example 8.4, but its equations of motion will be formulated
considering only the translational DOFs u1 and u2 (i.e., the rotational DOFs u3 and u4 will be
excluded).

1. Determine the stiffness matrix. In a statically determinate structure such as the one in Fig.
E8.5a, it is usually easier to calculate first the flexibility matrix and invert it to obtain the
stiffness matrix. The flexibility influence coefficient fˆíj is the displacement in DOF i due to

unit force applied in DOF j (Fig. E8.4c and d). The deflections are computed by standard
procedures of structural analysis to obtain the flexibility matrix:

ˆf  L
3
16 5 
48 EI  5 2
 

The off-diagonal elements fˆ12 and fˆ21 are equal, as expected, because of Maxwell’s theorem

of reciprocal deflections. By inverting f̂ , the stiffness matrix is obtained:

48EI  2 5 (a)

k  5 16 
7 L3  

2. Determine the mass matrix. This is a diagonal matrix because the lumped masses are
located where the DOFs are defined:

 mL / 4 
m (b)
 mL / 2 

3. Determine the equations of motion. Substituting m, k, and p  t   p1  t  p2  t 

T
in Eq.

(8.14) with c = 0 gives

 mL / 4  u1  48EI  2 5 u1   p1  t  

      (c)
 mL / 2  u2  7 L3  5 16  u2   p2  t  

8 - 19
Example 8.6

Formulate the free vibration equations for the two-element frame of Fig. E8.6a. For
both elements, the flexural stiffness is EI, and axial deformations are to be neglected. The
frame is massless with lumped masses at the two nodes as shown.

f 11 f 12
2m m f S1=1
u1
EI f 21 f S2=1 f 22
L EI u2

L
(a) (b) (c)
Figure E8.6
Solution

The two degrees of freedom of the frame are shown. The mass matrix is

3m 
m (a)
 m 

Note that the mass corresponding to u1  1 is 2m  m  3m because both masses will undergo

the same acceleration since the beam connecting the two masses is axially inextensible.
The stiffness matrix is formulated by first evaluating the flexibility matrix and then
inverting it. The flexibility influence coefficients are identified in Fig. E8.6b and c, and the
deflections are computed by standard procedures of structural analysis to obtain the flexibility
matrix:

L3  2 3
fˆ 
6 EI  3 8

This matrix is inverted to determine the stiffness matrix:

6 EI  8 3
kˆ  3  3 2 
7L  

8 - 20
Thus the equations in free vibration of the system (without damping) are

3m  u1  6 EI  8 3 u1  0 

      
 m  u2  7 L3 
 3 2  u2  0 

Example 8.7

Formulate the equations of motion for the two-story frame in Fig. E8.7a. The flexural
rigidity of the beams and columns and the lumped masses at the floor levels are as noted. The
dynamic excitation consists of lateral forces p1(t) and p2(t) at the two floor levels. The story
height is h and the bay width 2h. Neglect axial deformations in the beams and the columns.

u5 m u6 k51 k61 k53 k63

p2(t) u2 k21 k 23
EI
EI u1=1
h EI u3=1
u3 2m u4 k 31 k41 k 43
p1(t) u1 k11 k13
2EI k 33
h 2EI 2EI

L=2h
(a) (b) (c)

Figure E8.7
Solution

The system has six degrees of freedom shown in Fig. E8.7a: lateral displacements u1 and u2
of the floors and joint rotations u3, u4, u5, and u6. The displacement vector is

T
u  u1 u2 u3 u4 u5 u6 (a)

The mass matrix is given by Eq. (8.19):

2 
 1 
 
 0 
m  m 
 0  (b)
 0 
 
 0

8 - 21
The stiffness influence coefficients are evaluated following the procedure of Example 8.4. A
unit displacement is imposed, one at a time, in each DOF while constraining the other five
DOFs, and the stiffness influence coefficients (e.g., shown in Fig. E8.7b and c for u1  1 and

u3  1 , respectively) are calculated by statics from the nodal forces for individual structural

elements associated with the displacements imposed. These nodal forces are determined from
the beam stiffness coefficients (Appendix 1). The result is

 72 24 6h 6h 6h 6h 

 24 24 6h 6h 6h 6h 

EI  6h 6h 16h 0 
2
2h 2 2h 2
k 3   (c)
h  6h 6h 2h 2 16h 2 0 2h 2 
 6h 6h 2h 2 0 6h 2 h 2 
 
 6h 6h 0 2h 2 h 2 6h 2 

The dynamic forces applied are lateral forces p1(t) and p2(t) at the two floors without
any moments at the nodes. Thus the applied force vector is

p(t )  p1  t  p2  t  0 0 0 0
T
(d)

The equations of motion are

  ku  p  t 
mu (e)

where u, m, k, and p(t) are given by Eqs. (a), (b), (c), and (d), respectively.

8.4 Static condensation

Static condensation is a method to exclude the DOFs with no force from dynamic
analysis. Typically, the formulation of stiffness matrix in static analysis considers all
unrestrained DOFs at joints between structural members. Some of DOFs may not be
associated with any mass in dynamic analysis, for example, rotation DOFs in a lumped-mass
model, so they should be excluded to simplify the dynamic analysis.

8 - 22
 u6 u7 u8
u2 u2
Node
u3 u4 u5
u1 u1

(a) (b)
Figure 8.5 (a) Degrees of freedom (DOFs) for elastic forces-axial deformation neglected;
(b) DOFs for inertia forces.

The equations of motion for a building shown above is

m tt 0 ut   k tt k to   ut  pt  t  
      (8.23)
 0

 o  k ot
0  u k oo  u o   0 

It is partitioned into translation ( ut ) and rotation ( u o ) DOFs. Each part involves vectors and

sub-matrices. Each group of partitioned equations are

t  k tt u t  k tou o  pt  t 
m tt u and k ot ut  k oou o  0 (8.24)

Because no inertia terms and external forces are associated with the rotations, u o can be

solved.

u o  k oo 1k ot ut (8.25)

Then, we can substitute u o into the equation for translational DOFs and obtain equations of

motion which are simpler as they involve only translation DOFs.

t  kˆ tt ut  pt  t 
mtt u (8.26)

where the condensed stiffness matrix is

kˆ tt  k tt  k toT k oo
1
k ot (8.27)

Note that k to  k Tot because k is a symmetric matrix.

8 - 23
Example 8.8

Examples 8.4 and 8.5 were concerned with formulating the equations of motion for a
cantilever beam with two lumped masses. The degrees of freedom chosen in Example 8.5
were the translational displacements u1 and u2 at the lumped masses; in Example 8.4 the four
DOFs were u1, u2, and node rotations u3 and u4. Starting with the equations governing these
four DOFs, derive the equations of motion in the two translational DOFs.

Solution
T T
The vector of four DOFs is partitioned in two parts: ut  u1 u2 and u 0  u3 u4 . The

equations of motion governing ut are given by Eq. (9.3.4), where

 mL / 4 
pt  t   p1  t  p2  t 
T
mtt  
mL / 2 
(a)


To determine kˆ tt , the 4 × 4 stiffness matrix determined in Example 8.4 is partitioned:

 12 12  3L 3L 
k k t0  8 EI  12 24 3L 0 
k =  tt   
k 0t k 00  L3  3L 3L L2 L2 / 2  (b)
 
 3L 0 L2 / 2 2 L2 

Substituting these submatrices in Eq. (9.3.5) gives the condensed stiffness matrix:

48EI  2 5
kˆ tt   5 16  (c)
7 L3  

This stiffness matrix of Eq. (c) is the same as that obtained in Example 8.5 by inverting the
flexibility matrix corresponding to the two translational DOFs.
Substituting the stiffness submatrices in Eq. (8.24) gives the relation between the
condensed DOF u0 and the dynamic DOF ut :

1  2.57 3.43
u 0 = Tu t T (d)
L  0.857 0.857 

The equations of motion are given by Eq. (8.25), where mtt and pt(t) are defined in
Eq. (a) and kˆ tt in Eq. (c). These are the same as Eq. (c) of Example 8.5.

8 - 24
Example 8.9

Apply static condensation to Examples 8.7 to reduce number of DOFs for dynamic
analysis.

 72 24 6h 6h 6h 6h 

 24 24 6h 6h 6h 6h 

k k t0  EI  6h 6h 16h 2
2h 2 2h 2 0 
k =  tt   
k 0t k 00  h3  6h 6h 2h 2 16h 2 0 2h 2 
 6h 6h 2h 2 0 6h 2 h 2 
 
 6h 6h 0 2h 2 h 2 6h 2 
From Eq. (8.27),

EI  54.885 17.508
kˆ tt  k tt  k toT k oo
1
k ot  3 
h  17.508 11.607  (a)

From Eq. (8.25), rotation DOFs can be determined after translation DOFs are obtained.

u o  k oo 1k ot ut

1  2.57 3.43
u 0 = Tu t T (b)
L  0.857 0.857 

8.5 Planar systems subjected to translational ground motion

Equations of motion

At each instant of time, displacement of each mass is

u tj  t   u j  t   u g  t  (8.28)

For N masses, the displacements can be written in compact form as a vector.

ut  t   u  t   u g  t  1 (8.29)

where 1 is a vector of order N with each element equal to unity.

8 - 25
 Rigid-body utj utj
motion uj uj
N N

j j
Rigid-body
motion
1 1

ug ug ug
(a) (b)

Figure 8.6 (a) Building frame; (b) simplified tower

The equations of motion previously derived for a MDF system subjected to external
for p  t  is still valid except that the external force for this case (ground excitation) is zero.

fI  fD  fS  0 (8.30)

Only relative displacements u between masses and the base produce deformation and elastic
and damping forces.

t .
The inertia forces f I are related to the total acceleration u

f I  mu
t (8.31)

Substituting f I  mu
t in the equilibrium equation, we get

  cu  ku  m1ug  t 
mu (8.32)

The right-hand side is the effective earthquake forces due to ground motion excitation.

p eff  t   m1ug  t  (8.33)

8 - 26
 mN mN
..
-m N ug (t)

mj mj
..
-m j ug (t)

=
m1 m1
..
-m1 ug (t)

..
ug(t) Stationary base

Figure 8.7 Effective earthquake forces

This is valid when a unit ground displacement results in a unit total displacement of
all DOFs. In general, this is not always the case. We introduce the influence vector ι to
represent the influence of ground displacement on total displacement at DOFs.

u t  t   ιu g  t   u  t  (8.34)

The equations of motion are

  cu  ku  mιug  t 
mu (8.35)

For example, vertical DOF u3 is not displaced when the ground moves horizontally. The

influence vector is

1 
 
ι  1  (8.36)
 
0 

8 - 27
 Figure 8.8 (a) L-shaped frame; (b) influence vector  : static displacements due to u g  1 ;

(c) effective earthquake forces

The effective earthquake force is

 m1  1   m1 
p eff  t   mιug  t   ug  t    1  u t m  m 
  
g   2
m2  m3 (8.37)
  3

  
m3  0   
 0 

Note that the mass corresponding to u2  1 is m2  m3 because both masses will undergo the

same acceleration since the connecting beam is axially rigid. The effective earthquake force
is zero in the vertical DOF because the ground motion is horizontal.

8.6 Inelastic systems

For inelastic systems, the force resisting deformation is no longer linear relationship and is
described by a nonlinear function

f s  f s  u, u  (8.38)

The equation of motion becomes

  cu  f s  u, u   mιug  t 
mu (8.39)

Such equation has to be solved by numerical methods as presented in Chapter 5.

8 - 28
8.7 Problem statement

Given a system with known, mass matrix m, damping matrix c, stiffness matrix k, and
excitation p(t) or ug  t  , we want to determine the response of the system.

Response can be any response quantity such as displacement u  t  , velocity,

acceleration of masses or internal forces, which is closely related to the relative displacement.

By the concept of equivalent static force, internal forces can be obtained by static
analysis of structure subjected to a set of equivalent static forces

f S  t   ku  t  (8.40)

8 - 29
Exercises

1. Formulate the equations of motion with respect to the two DOFs defined at the left end of
the rigid bar in Fig. P8.1. The bar has total mass m supported by tow springs k1 and k2
subjected to dynamic forces pt and p . The bar is constrained so that it can move only in

vertical direction.

Figure P8.1

2. Consider tow-story shear frame with lumped masses shown in Fig. P8.2. The beams are
assumed to be rigid and the flexural rigidity of all columns is EI . Neglect axial
deformation in all elements. Formulate the equations of motion for the system if
a) the columns have fixed support,
b) the columns have pinned support.

m/2 m/2
p2(t) p2(t)

h EI EI EI EI
m m
p1(t) p1(t)

h EI EI EI EI

2h 2h
(a) (b)
Figure P8.2

3. Consider a uniform slab supported on four fixed-base columns as shown in Fig. P8.3. The
slab has a total mass m and is rigid in plane and out of plane. The cross section of each

8 - 30
column is circular with moment of inertia as noted in the Fig. P8.3. With DOFs selected
as u x , u y , and u at the center of the slab,

a) formulate the mass and stiffness matrices in terms of m and lateral stiffness of the
smaller column, k  12 EI / h3 ,
b) formulate the equation of motion for ground motion in the x-direction, the y-direction,
and the direction A-C.

Figure P8.3

8 - 31