Modeling and Analysis of Structural Systems for

Efficient Design
Structural Dynamics

Week 11 Handout
1
2
 Week 11 Topics

◼ Topics for the week:

◼ Single Degree of Freedom Structural Dynamics
◼ Multi-Degree of Freedom Structural Dynamics
◼ Stiffness, Mass and Damping
◼ Linear Dynamic Analysis in Commercial Software

3
 Week 11 Learning Outcomes

◼ Upon completion of this week’s content you will be able to:

◼ Describe the fundamental components of the dynamic equilibrium equation
◼ Apply mass, damping and a dynamic forcing function to a statically loaded structural analysis
model to complete a dynamic analysis
◼ Complete a linear dynamic analysis using modal superposition and direct integration

4
 Week 11: Significance on the Job

◼ Importance to engineering practice

◼ Understanding the importance of structural dynamics is important for the design of
structures subjected to seismic, wind or other dynamic loads
◼ Correct construction and application of dynamic analysis models is critical for
designing structures

5
Next Scene: Single Degree of Freedom
Structural Dynamics

6
1
7
Single Degree of Freedom Structural Dynamics
9
 Idealized Single Degree of Freedom Structure

F(t)
Mass F ( t ), u( t )

Damping t
Stiffness
Y

X
u(t)

10
Equation of Dynamic Equilibrium
f I (t ) F (t )

f D (t )
0.5 f S ( t ) 0.5 f S ( t )

F (t ) − f I (t ) − f D (t ) − f S (t ) = 0 [1]

f I (t ) + f D (t ) + f S (t ) = F (t ) [2]
Inertia Damping Spring

( t ) + c u
mu ( t ) + k u( t ) = F ( t )
11
Properties of Structural (Seismic) MASS

INERTIAL FORCE
MASS

M
1.0

ACCELERATION
• Includes all dead weight of structure
• May include some live load
• Has units of FORCE/ACCELERATION

12
Properties of Structural DAMPING

DAMPING FORCE
DAMPING FORCE
DAMPING

C
1.0

VELOCITY DISPLACEMENT
AREA =
• In absence of dampers, is called Natural Damping ENERGY
• Usually represented by linear viscous dashpot DISSIPATED
• Has units of FORCE/VELOCITY
Damping vs Displacement
Response is Elliptical for
Linear Viscous Damping
13
Properties of Structural STIFFNESS

SPRING FORCE
STIFFNESS

K
1.0

DISPLACEMENT
• Includes all structural members
• May include some “nonstructural” members
• Requires careful mathematical modelling
• Has units of FORCE/DISPLACEMENT

14
 Undamped Free Vibration

Equation of Motion: mu( t ) + k u( t ) = 0

Initial Conditions: u0 u0
Assume : u( t ) = A sin(t ) + B cos(t )
u0 k
Solution: A=− B = u0 =
 m
u0
u( t ) = u0 cos(t ) + sin(t )

15
Undamped Free Vibration (2)
T = 0.5 seconds
u0

Displacement, inches
1.0
3
2
1
0
u0
-1
-2
-3
0.0 0.5 1.0 1.5 2.0
Time, seconds

Circular Frequency Period of Vibration Cyclic Frequency

(radians/sec) (seconds/cycle) (cycles/sec, Hertz)

k 2p 1 w
= T= f= =
m w T 2p
16
 Damped Free Vibration

Equation of Motion: mu ( t ) + c u( t ) + k u( t ) = 0

Initial Conditions: u0 u0
Assume: u( t ) = e st

Solution:
 u0 + u0 
u0 cos( D t ) + sin( D t )
−t
u( t ) = e
 D 
c c
= = D =  1 − 2
2 m cc
17
Damping in Structures
c c
= = Cc is the Critical Damping Constant
2 m cc
 is expressed as a ratio (0.0 <  < 1.0) in computations.
Sometimes  is expressed as a percent (0 <  < 100%)
Displacement, in

Time, sec

Response of Critically Damped System, =1.0 or 100% critical

18
Damped Free Vibration

Displacement, inches
3
2
1 0% Damping
0 10% Damping
-1 20% Damping
-2
-3
0.0 0.5 1.0 1.5 2.0
Time, seconds

19
 Undamped Harmonic Loading

Equation of Motion: (t ) + ku (t ) = p0 sin( t )

mu
 = Frequency of the Forcing Function
2
T =
 T = 0.25 Seconds
po=100 kips
150
100
Force, Kips

50
0
-50
-100
-150
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Time, Seconds

20
Undamped Harmonic Loading (2)

Equation of Motion: ( t ) + k u( t ) = p0 sin( t )

mu
Assume system is initially at rest:

:
Particular Solution u( t ) = C sin( t )

Complimentary Solution : u( t ) = A sin(t ) + B cos(t )

Solution:
p0 1   
u (t ) =  sin( t ) − sin(t ) 
k 1 − ( /  ) 2   

21
Undamped Harmonic Loading


=
LOADING FREQUENCY
Define
 Structure’s NATURAL FREQUENCY

Transient Response
Dynamic Magnifier (at Structure’s Frequency)

p0 1
2 (
u( t ) = sin( t ) −  sin(t ))
k 1− 
Steady State
Static Displacement Response
(At Loading Frequency)

22
 = 4 rad / sec  = 2 rad / sec  = 0.5 uS = 5.0 in.
200

Force, Kips
100
LOADING,kips 0
-100
-200
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00

Displacement, in. Displacement, in.

10
STEADY 5
STATE 0
-5
RESPONSE, in. -10
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00

10
5
TRANSIENT 0
RESPONSE, in. -5
-10
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Displacement, in.

10
5
TOTAL 0
RESPONSE, in. -5
-10
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Time, seconds
23
  4 rad / sec  = 4 rad / sec  = 0.99 uS = 5.0 in.
150
100

Force, Kips
50
LOADING, 0
-50
kips -100
-150
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00

500

Displacement, in.
STEADY 250
0
STATE
-250
RESPONSE, in
-500
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00

500
Displacement, in.

250
0
TRANSIENT
-250
RESPONSE, in
-500
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00

80
Displacement, in.

40
TOTAL 0
RESPONSE, in -40

-80
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Time, seconds
24
Undamped Resonant Response Curve

80

Displacement, in.
40
2 uS

-40

Linear Envelope

-80
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Time, seconds

25
Response Ratio: Steady State to Static
(Absolute Values)
12.00

Resonance
10.00
Magnification Factor 1/(1-  )
2

8.00

6.00

Slowly
4.00
Loaded
2.00
Rapidly
1.00
Loaded
0.00
0.00 0.50 1.00 1.50 2.00 2.50 3.00
Frequency Ratio 
26
Next Scene: Single Degree of Freedom
Dynamics II

27
28
Multi Degree of Freedom Structural
Dynamics
30
Damped Harmonic Loading

Equation of Motion:

(t ) + cu (t ) + k u(t ) = p0 sin( t )

mu
2
T = = 0.25 Seconds

150 po=100 kips
100
Force, Kips

50
0
-50
-100
-150
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
Time, Seconds

31
Damped Harmonic Loading
Equation of Motion:
(t ) + cu (t ) + k u(t ) = p0 sin( t )
mu
Assume system is initially at rest

Particular Solution: u(t ) = C sin( t ) + D cos( t )

Complimentary Solution:
u(t ) = e −t  A sin( Dt ) + B cos( Dt )
c
=
2m
Solution:
D =  1− 2
u(t ) = e  A sin( Dt ) + B cos( Dt )
− t

+ C sin( t ) + D cos( t )

32
Damped Harmonic Loading

Transient Response, at Structure’s Frequency

(Eventually Damps Out)

u(t ) = e − t
 A sin( Dt ) + B cos( Dt ) +

C sin( t ) + D cos( t )
Steady State Response,
at Loading Frequency

po 1−  2 po − 2
C= D=
k (1 −  2 ) 2 + (2 ) 2 k (1 −  2 ) 2 + (2 ) 2

33
Damped Harmonic Loading (5% Damping)

BETA=1 (Resonance)
Beta=0.5
Beta=2.0
50
40

Displacement Amplitude, Inches

30
20
10
0
-10
-20
-30
-40
-50
0.00 1.00 2.00 3.00 4.00 5.00
Time, Seconds

34
Damped Harmonic Loading (5% Damping)

50

40

Displacement Amplitude, Inches

1
 Static
30

20 2
10

-10

-20

-30

-40

-50
0.00 1.00 2.00 3.00 4.00 5.00
Time, Seconds

35
 14.00
Resonance
12.00 0.0% Damping
5.0 % Damping
10.0% Damping

Dynamic Response Amplifier

10.00 25.0 % Damping

8.00
1
RD =
6.00 (1 −  2 ) 2 + (2 ) 2
4.00
Slowly
Loaded
2.00 Rapidly
Loaded
0.00
0.00 0.50 1.00 1.50 2.00 2.50 3.00
Frequency Ratio, 

36
 General Dynamic Loading

◼ Examples of General Dynamic Loading

◼ Earthquake ground motion
◼ Wind velocity (or pressure) history
F(t) ◼ Blast profile
◼ Tsunami pressure history
◼ Machine vibration
◼ Vehicle crossing a bridge

Time, T

37
General Dynamic Loading Solution Techniques

• Fourier Transform
• Duhamel Integration
• Piecewise “Exact”
• Newmark Techniques
All techniques are carried out using a computer

38
Piecewise “Exact” Method

dF
F(t ) = Fo + t
dt
dF

Fo
dt
t
dt

39
Piecewise “Exact” Method

Advantages:

▪ Exact if load increment is linear

▪ Very computationally efficient

Disadvantages:

▪ Not generally applicable for inelastic behavior

40
Newmark Techniques
◼ Proposed by Nathan Newmark

◼ General method that encompasses a family of different integration schemes

◼ Derived by:
◼ Assuming a variation in acceleration over a time-step
◼ Formulating the equations of motion in incremental form to accommodate inelastic behavior (if needed)

◼ Parameters g and  are chosen by user to arrive at a numerical approximation

◼ Parameters control
◼ Stability of technique
◼ Accuracy of technique

◼ Two common Newmark approximations

◼ g=1/2 and =1/4 Constant average acceleration method
◼ g=1/2 and =1/6 Linear acceleration method

41
Newmark Method

Advantages:

• Works for both Elastic and Inelastic Response

• Very effective for Multiple Degree of Freedom Systems
Disadvantages:

• Potential Numerical Error if computation time step is

too long

42
Development of Effective Earthquake Force

0.40
GROUND ACC, g

0.20

0.00

-0.20

-0.40
0.00 1.00 2.00 3.00 4.00 5.00 6.00
TIME, SECONDS

43
Earthquake Ground Motion - 1940 El Centro
0.4

0.3 Ground Acceleration

Ground Acceleration (g's)
0.2

0.1
40
0
30 Ground Velocity

Ground Velocity (cm/sec)

-0.1
20
-0.2
10
-0.3
0 10 20 30 40 50 60 0
Time (sec)
-10
15
-20
Ground Displacement
Ground Displacement (cm)

10
-30
0 10 20 30 40 50 60
5
Time (sec)
0

-5
Many ground motions now
-10
available via the Internet
-15
0 10 20 30 40 50 60
Time (sec)

44
Development of Effective Earthquake Force
ut
ug ur
Ground ACCELERATION Time History
0.40

GROUND ACC, g
0.20

0.00

-0.20

-0.40
0.00 1.00 2.00 3.00 4.00 5.00 6.00
TIME, SECONDS

m [ug ( t ) + ur ( t )] + c ur ( t ) + k ur ( t ) = 0

mur ( t ) + c ur ( t ) + k ur ( t ) = − mug ( t )

45
“Simplified” form of Equation of Motion:

mur (t ) + cur (t ) + kur (t ) = −mug (t )

Divide through by m:

c k
ur (t ) + ur (t ) + ur (t ) = −ug (t )
m m
Make substitutions:

c k
= 2 =2
m m
Simplified form:

ur (t ) + 2ur (t ) +  2ur (t ) = −ug (t )

46
For a given ground motion, the response
history ur(t) is function of the structure’s
frequency  and damping ratio 

Structural frequency

ur (t ) + 2ur (t ) +  ur (t ) = −ug (t )

2

Damping ratio
Ground motion time history

47
Response to Ground Motion (1940 El Centro)

0.4
Excitation applied to
0.3
structure with given  and 

Ground Acceleration (g's)

0.2

0.1

-0.1
SOLVER
-0.2

-0.3
0 10 20 30 40 50 60
Time (sec)
6

Structural Displacement (in)

4 Computed Response
2
Change in ground motion
or structural parameters  0

and  requires re-calculation -2

of structural response -4
Peak Displacement
-6
0 10 20 30 40 50 60
Time (sec)
48
Next Scene: Multiple Degree of Freedom
Dynamics

49
50
Mass and Damping
52
 Multiple Degree of Freedom Systems

◼ Most structures have numerous (in some cases thousands) of degrees of freedom. These
systems are called Multiple Degree of Freedom (MDOF) systems.

◼ Solving such system is complex, but for systems that remain elastic the MDOF system can be
decomposed into a number of SDOF systems, the SDOF systems can be solved using the
techniques shown earlier, and the responses can be superimposed to produce the response of
the full system.

◼ The method can be further simplified by using a response spectrum as ground motion input. The
resulting analysis method is called Modal Response Spectrum Analysis.

◼ The Equivalent Lateral Force Method in ASCE 7 is a simplification of the Modal Response
Spectrum approach.

◼ For wind loading, the system dynamics are accounted for using the Gust Factor in static analysis,
certain classes of structures (dynamically sensitive) require more detailed studies (e.g. wind
tunnel testing)
53
Planar Frame with 36 Degrees of Freedom
uy
Majority of Mass
rz Typical Nodal
is in Floors
DOF
9 10 11 12 ux

5 6 7 8

1 2 3 4
Motion is
Predominantly
Lateral
13 14 15 16

54
Planar Frame with 36 Static Degrees of Freedom but with only THREE Dynamic
DOF

u1

u2  u1 
 
u3
U = u2 
u 
 3

55
Idealized Structural Property Matrices
m1 f1(t), u1(t)  k1 -k1 0 
K = -k1 k1+ k 2 -k 2 
k1  0 -k 2 k 2 + k 3
m2 f2(t), u2(t)
 m1 0 0
M =  0 m2 0 
k2
 0 0 m3
m3 f3(t), u3(t)

 f1 (t )   u1 (t ) 
   
k3 F (t ) =  f 2 (t )  U (t ) = u2 (t ) 
 f (t )  u (t ) 
 3   3 
Note: Damping to be shown later
56
Coupled Equations of Motion for Undamped Free Vibration

DOF 1

DOF 2

DOF 3

57
Solving Equations of Motion using Modal Decomposition

◼ This will be done by a transformation of coordinates from Normal Coordinates (displacements at

the nodes) to Modal Coordinates (amplitudes of the natural Mode shapes).

◼ Because of the Orthogonality Property of the natural mode shapes, the equations of motion
become uncoupled, allowing them to be solved as SDOF equations.

◼ After solving, we can transform back to the Normal Coordinates .

58
Solutions for Dynamic System in Undamped Free Vibration
(Natural mode shapes and frequencies)

Assume U(t) = f sin(wt) ˙˙ (t) = -w 2f sin(wt)

U
Then Kf - w 2 M f = 0{} has three (n) solutions:

1,1  1,2  1,3 

     
1 = 2,1  ,  1 2 = 2,2  ,  2 3 = 2,3  ,  3
     
 3,1   3,2   3,3 
Natural Frequency (radians/sec)

Natural Mode Shape (no units)

59
Solutions for System in Undamped Free Vibration

Kf = w 2Mf For a SINGLE Mode

KF = MFW 2
For ALL Modes
Where:  = 1 2 3 

ω12 
 
 =
2
ω2
2

 2
 ω 3
 1,i = 1.0
Note: Mode shape has arbitrary scale; usually or

F T MF = I
60
Mode Shapes for Idealized 3-Story Building
1,2
1,1 1,3

Node
2,1 2,2 2,3

Node Node

3,1 3,2 3,3

MODE 1 MODE 2 MODE 3

Decreasing Period of Vibration
Increasing Frequency of Vibration 61
Concept of Linear Combination of Mode Shapes
(Transformation of Coordinates)

U = Y
Mode Shape

1,1 1,2 1,3   y1  1,1  1,2  1,3 

        
U = 2,1 2,2 2,3   y2  U = 2,1  y1 + 2,2  y 2 + 2,3  y3
3,1 3,2 3,3   y3   
 3,1 
 
 3,2 
 
 3,3 
 

Modal Coordinate =
amplitude of mode
shape

62
Orthogonality Conditions

 = 1 2 3 
Generalized Mass
Generalized Stiffness
é m* ù é k* ù
ê 1 ú ê 1 ú
FT MF = ê m2* ú F T KF = ê k*2 ú
ê *
ú ê ú
ê m3 ú ê k3 ú
*
ë û ë û
Generalized Damping Generalized Force
é c* ù ì f * (t) ü
ê 1 ú ï 1 ï
ï ï
F TCF = ê ú F F(t) = í f2 (t) ý
T *
c*2
ê ú
ê c*3 ú
ï * ï
ë û ïî f3 (t) ïþ

63
Development of Uncoupled Equations of Motion

MDOF Equation of Motion:

Transformation of Coordinates: U = Y
Substitution:

Premultiply by T :

Using Orthogonality Conditions: Uncoupled Equations of Motion are:

é * ùì ü é c* ùì ü é k* ùì ü ì f * (t) ü
ê m1 úï y1 ï ê 1 úï y1 ï ê 1 úï y1 ï ï 1 ï
ê úï ï ê úï ï ê úï ï ï * ï
ý = í f 2 (t) ý
úí 2 ý +ê úí 2 ý +ê úí 2
m*2 y c*2 y k *2 y
ê
ê ï ï ê ï ï ê úï y ï ï * ï
êë m*3 úúïî
y3 ï
þ êë c*3 úúïî
y3 ï
þ êë k *3 úûï
î 3 ï
þ ï f (t) ï
û û î 3 þ

64
Development of Uncoupled Equations of Motion
(Explicit Form)

MODE 1 m1*y1 + c1*y1 + k1*y1 = f1* (t)

MODE 2 m*2 y 2 + c*2 y 2 + k*2 y 2 = f2* (t)

MODE 3 m*3y3 + c*3y3 + k*3y3 = f3* (t)

65
Development of Uncoupled Equations of Motion
(Explicit Form)

c*i
Simplify by Dividing Through by m* and defining modal damping ratio i = *
2mi i

MODE 1 y1 + 2x1w1y1 + w12 y1 = f1* (t) / m1*

MODE 2 y 2 + 2x 2w2 y 2 + w22 y 2 = f 2* (t) / m*2

MODE 3 y3 + 2x3w3y3 + w 2
y
3 3
= f3 (t) /
* *
m3

66
Earthquake “Loading” for MDOF Systems
ug ˙u˙ r,1
ì u (t) + u (t) ü
ï g r,1 ï
ï ï
FI (t) = M í ug (t) + u r,2 (t) ý=
ï ï
ïî ug (t) + u r,3 (t) ïþ

ì 1.0 ü ì u (t) ü
ï ï ï r,1 ï
ï ï
M í 1.0 ý ug (t) + M í u r,2 (t) ý
ï 1.0 ï ï ï
î þ ïî u r,3 (t) ïþ

Move to RHS as fFEFF (t) = - M Rug (t)

67
Modal Earthquake Loading F (t) = -F MR u g (t)
* T

m1=2
u1
m2=3 2  1
u2   
M= 3  R= 1
m3=1
1
u3  1 
ug (t )
m3=1
m1=2 u1
u3 2 + 1  1 
 
m2=3 u2 M =  3  R= 1 
 1 0 
 
ug (t )
68
Definition of Modal Participation Factor

For earthquakes: f (t) = -f MRu g (t)

i
* T
i

Typical Modal Equation:

f (t)*
f MR T
y i + 2xiwi y i + w y i = * = -
2
i
i
*
u g
i
(t)
m mi
i

Modal Participation Factor pi

69
Caution Regarding Modal Participation Factor

 MRT
pi = i
*
m i

 T

i M i

Its value is dependent on the (arbitrary) method

used to scale the mode shapes.

70
Concept of Effective Modal Mass

For each Mode i mi = pi mi 2 *

• The sum of the effective modal mass for all modes is

equal to the total structural mass.

• The value of effective modal mass is

independent of mode shape scaling.

• Use enough modes in the analysis to provide

a total effective mass not less than 90% of the
total structural mass.

71
Solution of MDOF Equations of Motion

◼ Explicit (Step by Step) Integration of Coupled Equations. This is referred to as Direct Integration
Response History Analysis.

◼ Explicit Integration of FULL SET or PARTIAL set of Uncoupled Equations. This is referred to as
Modal Response History Analysis.

◼ Modal Response Spectrum Analysis (Approximate)

◼ Equivalent Lateral Force Analysis (This is a vastly simplified Modal Response Spectrum
Analysis).

For a linear dynamic analysis, analysis of

Uncoupled Equations is much quicker and as
accurate as integration of Coupled Equations.

72
Next Scene: Stiffness, Mass and Damping

73
74
Modal Analysis
76
 Stiffness, Mass and Damping

◼ When creating a dynamic analysis model, three properties must be specified for
the numerical model:
◼ Stiffness
◼ Mass
◼ Damping

◼ Thetwo most important characteristics of a structural system for dynamic

analysis are:
◼ Period (or Frequency) of Vibration (based on the mass and stiffness)
◼ Damping Ratio

77
Period (or Frequency) of Vibration
◼ For a dynamic system, it is the
relationship between the mass and
stiffness that is a critical parameter
in determining the response

◼ For a single degree of freedom

system:

= km rad/sec

2
T= = 2 m sec/cycle
 k
◼ Adjusting the period of vibration by
changing the mass or stiffness (or
both) can have a significant effect
on the dynamic response
78
Critical Damping Ratio

◼ The critical damping ratio is another 14.00

important parameter of a dynamic Resonance
0.0% Damping
system 12.00
5.0 % Damping
10.0% Damping

Dynamic Response Amplifier

◼ It is important to note that when a 10.00 25.0 % Damping

structure is rapidly or slowly loaded the

damping ratio does not have a 8.00
significant effect
6.00
◼ When a structure is loaded at or near
the natural frequency, damping ratio is 4.00
very important Slowly
2.00 Loaded Rapidly
◼ There is a law of diminishing returns
Loaded
once a damping ratio exceeds
0.00
approximately 25% 0.00 0.50 1.00 1.50 2.00 2.50 3.00
Frequency Ratio, 

79
 Structural Stiffness

◼ Represents the resistance to deformation

response under applied loads
k26 k27 k28
◼ In a dynamics problem it is often referred
k13 k14 k15 k16
to as the restoring force k23 k24 k25

◼ Includes the beams, columns, k9 k10 k11 k12

k20 k21 k22
connections and foundations that provide
the load path for the loads that must be k5 k6 k7 k8
k17 k18 k19
resisted

◼ In a static equilibrium case, the only k1 k2 k3 k4

resistance is in the structural stiffness

◼ The stiffness matrix used in the analysis

is the same used for determining force
demands in a static analysis
80
 Mass

◼ Mass represents the resistance of the structure to

acceleration (inertia)
◼ The distribution of mass in the analysis model needs to
represent the distribution of mass in the structure m13 m14 m15 m16

◼ Can be modeled including either distributed mass

(mass/ft or mass/ft2) or point masses m9 m10 m11 m12

◼ Mass moment of inertia for rotational inertia can also be

applied (for point masses that are intended to represent m5 m6 m7 m8
rigid bodies such as large rectangular regions)
m1 m2 m3 m4
◼ Note that the applied forces are not the same as the
mass in the model (although some programs allow you to
specify mass based on specified load cases)
◼ The mass for the analysis model should represent all the
dead loads and could include some of the live loads
(such as partitions). Live load mass may be
different for wind and seismic.
81
Assigning Mass to a Structural Model (for Diaphragms)

Diaphragm
Element

Multiplier on Mass of Single Node with Mass and

Point Masses at Grid
Diaphragm (Shell Element) Rotational Inertia (Requires
Intersections
rigid diaphragm constraint)
If self-mass is included in the member
definitions, this must not be double counted. 82
 Damping Ratio

◼ The damping ratio for a linear dynamic analysis typically represents the natural or inherent
damping of the structural system

◼ For inherent damping, the damping matrix is typically a classical damping matrix which means
that it can be uncoupled using the mode shapes and frequencies

◼ Classical damping is also called proportional damping since the damping matrix is a
mathematical combination of the mass and stiffness matrices

◼ If supplemental damping is added to a dynamic system, the damping matrix is a non-classical

damping matrix which means it cannot be uncoupled by the mode shapes and frequencies

◼ This course focuses on analyses with classical damping matrices (only inherent damping)

83
Rayleigh Damping

◼ One of the more common forms of

proportional/classical damping is Rayleigh Damping

◼ The Rayleigh Damping Matrix is formed as a

mathematical combination of the mass and stiffness C = aM + K
matrices

◼ The formulation requires calculation of two where

constants C = Rayleigh Damping Matrix
M = Mass Matrix
◼ Each constant is based on the frequency of
K = Stiffness Matrix
vibration (or period of vibration) and the damping
a = Mass Proportional Damping Constant
ratio at that frequency
 = Stiffness Proportional Damping Constant
◼ Select two frequencies of vibration and the damping
ratio at each frequency to determine the constants
(a and )

84
Rayleigh Damping

◼ Rayleigh Damping from the two

specified frequencies and damping
ratios sets the curve for the damping
ratio for other frequencies of the system

◼ Any frequency between the two

specified frequency has a damping ratio
less specified ratio (if they are the same)

◼ Frequencies outside the two specified

frequencies have higher damping ratios

◼ Most programs will determine the

constants when you specify the periods
and damping ratios f1 = 13 rad/s f3 = 100 rad/s
z1 = 0.05 z3 = 0.05

85
 Modal Damping

◼ Another typical form of proportional damping is called Modal Damping

◼ It is also called Penzien-Wilson or Wilson Damping

◼ The damping ratio is specified for each mode that is used in the analysis

◼ In general, the full damping matrix is not formed and the equations are solved using modal
response history analysis

 2z 11 0 0 0 
 0 2z 22 0 0 
T C  = 
 0 0 ... ... 
 
 0 0 ... 2z nn 
86
 Modeling Damping in Dynamic Analysis

◼ Modeling damping in linear analysis is an

important part of the model
Welded Steel Frame  = 0.010
◼ Selecting a reasonable damping ratio for Bolted Steel Frame  = 0.020
the analysis purposes is an important part
of the process Uncracked Prestressed Concrete  = 0.015
Uncracked Reinforced Concrete  = 0.020
◼ The type of structure and the level of Cracked Reinforced Concrete  = 0.035
deformation (service, strength, extreme
Damaged Steel Structure  = 0.050
event) go into making this determination Damaged Concrete Structure  = 0.075

Glued Plywood Shear wall  = 0.100

Nailed Plywood Shear wall  = 0.150

Structure with Added Damping  = 0.250

87
 Modal Response History Analysis versus Direct
Integration Response History

◼ For linear elastic structural analysis with classical damping (Rayleigh Damping or Modal
Damping) it is more computationally efficient to do a Modal Response History Analysis

◼ Modal Response History Analysis is done by solving the series of single degree of freedom
equations representing the individual modes and then combining the individual modal responses
to determine the total system response

◼ Direct Integration Response History Analysis is done using the matrix equations in coupled form

◼ Direct Integration Response History Analysis would typically only be used when the system is not
classically damped or when running a nonlinear analysis

88
Next Scene: Solving for Modes and Mode
Shapes

89
90
Linear Dynamic Analysis I
92
 Solving for Modes and Mode Shapes

◼ Solving
for Modes and Mode Shapes is an important analysis case that should
be done when designing a structure for lateral loads

◼ ASCE 7 seismic design and wind design procedures allow for load determination
without calculating a period or frequency of vibration but the penalty for not
calculating the value with a model is that the approximate code equations result
in conservative estimates of forces

◼ Valuesfrom the analysis model can be compared to the approximate code

equations for wind (frequencies) and seismic (periods) to assess whether the
frequencies are in the range of expected values

◼ Mode shapes, especially where animated, are great tools for assessing whether
the structure has the correct connectivity and kinematics
93
Modal Analysis and Wind Design

◼ Frequency determination is required to determine

whether the structure is classified as flexible or rigid
which determines how to calculate the gust factor, Gf

◼ Low-rise buildings, as defined in ASCE 7, do not

require determining the frequency

◼ ASCE 7 approximate lower bound frequencies for

different lateral systems are provided

◼ Since these are lower bound estimates, an actual

calculation will provide a better estimate of the
frequency

◼ The approximate frequencies can also be used to

assess the model frequency of vibration as a
validation check

94
Modal Analysis and Seismic Design
Ta = Ct hn x T = CuTa *

◼ Determining a period of vibration is required for seismic [Eqn 12.8-7] [Section 12.8.2]
design but can be done using only approximate
equations (for the ELF method)

◼ For the ELF method, using a model to calculate a period

of vibration can result in a substantial reduction in your
equivalent static forces

◼ For drift-controlled structures, the period of vibration from

the analysis model can significantly benefit design forces
for drift

◼ The approximate periods can also be used to assess the

calculated period of vibration as a validation check

*Note: CuTa can only be used if a

fundamental period is determined
from an analysis model.

95
 Importance of Modes and Mode Shapes

◼ When subjected to dynamic loads, the most important

characteristic in the structural response is the ratio of the
frequency of the dynamically applied force or acceleration
and the structure’s natural frequency

◼ Modal periods (or frequencies) provide the dynamic

characteristics of the structural system

◼ The modes and mode shapes can also provide significant

insight into the behavior of the structural system and
identify aspects of building performance that may be
problematic
◼ If the 1st or 2nd period of vibration of a 3-D building model is
torsional, this is good indicator of a torsional irregularity
◼ If one of the first modes has a localized element mode shape,
this could indicate an improperly restrained or overly flexible
element 96
Determining Important Modes

◼ Identifying which modes are important in a dynamic analysis can be critical for analyzing 3-D
structures, especially if they have a unique shape

◼ Identifying important modes can be done by looking at the tabular output from the modal analysis
cases and looking at the Effective Modal Mass and the Cumulative Effective Modal Mass

◼ There are two different calculated ways to assess the participation of different modes in a
dynamic analysis:
i MR
T

◼ Modal Participation Factor pi = T

i Mi
◼ Effective Modal Mass mi = pi 2iT Mi
◼ The Modal Participation Factor is an indicator of mode importance but the values depend on the
way the mode shapes are scaled

◼ The Effective Modal Mass does not depend on scaling and is the best indicator of a mode’s
importance on the dynamic response
97
 Case Study Structure
A B C D E
N
1 2 3 4 5 6 7

14 ft
E
20 ft

6 @ 12 ft = 72 ft
D
20 ft

C
20 ft

B
20 ft

16 ft
A
Z
30 ft 30 ft 30 ft 30 ft 30 ft 30 ft 20 ft 20 ft 20 ft 20 ft

Plan View Elevation View

98
 Summary of Structure Loading

◼ Risk Category II Building (Office)

◼ Floor Dead Load: 75 psf + 13 psf = 88 psf

◼ Floor Live Load: 50 psf + 15 psf (partition)

◼ Roof Dead and Superimposed Dead Load: 60 psf + 18 psf = 78 psf

◼ Roof Live Load: 20 psf

◼ Seismic Criteria: Seismic Design Category D, SDS = 0.60, SD1 = 0.25

◼ Wind Criteria: V = 95 mph, Exposure Category C

◼ N-S: Buckling-Restrained Braced Frame, E-W: Steel Moment Frame

◼ Software – SAP2000 v23 (CSI)

99
 2-D Analysis Model for Determining Modal
Properties

◼ The 2-D model is built using the same principles as demonstrated in Week 8

◼ A planar model of the structural frame is constructed with all the elements and appropriate
modeling requirements (e.g. releases)

◼ Gravity and lateral loads were applied to those elements

◼ The mass representing ½ of the total structure is applied in the model

◼ The effect of the leaning columns are represented using a single leaning column that is
constrained at each of the floor levels and provides no lateral resistance

100
 Example Structure: 2-D Buckling-Restrained Braced Frame
Cumulative Effective
Effective Modal Mass Modal Mass
◼ ASCE 7 Frequency and Period Mode
Period of
UX UZ RY Sum UX Sum UZ Sum RY
Vibration (s)
Approximation: 1 1.484 0.789 8.91E-09 0.208 0.789 8.91E-09 0.208
75 75
◼ Wind
na = = = 0.73hz 2
3
0.537
0.316
0.158
0.033
3.17E-08
5.73E-10
0.558
0.109
0.946
0.979
4.06E-08
4.12E-08
0.766
0.875
h 102
4 0.231 0.013 2.12E-10 0.062 0.992 4.14E-08 0.936
◼ Seismic Ta = C h = 0.03(102)
x
t n
0.75
= 0.96s 5 0.188 0.005 4.34E-09 0.023 0.997 4.57E-08 0.960
6 0.150 0.001 2.15E-11 0.006 0.999 4.58E-08 0.966

Z Z Z
X X X
Mode 1: 1.484 s, 0.674 hz Mode 2: 0.537 s, 1.862 hz Mode 3: 0.316 s, 3.161 hz
101
 Modal Analysis and P-Delta Effects

◼ Modal properties can also be used to assess the importance of P-Delta Effects on the structure

◼ This can be done by running a modal analysis including P-Delta Effects

◼ In SAP2000, the modal analysis can start at the end of a nonlinear load case which includes the
gravity loads that would be used for the P-Delta analysis

◼ The magnitude of the change in the period or frequency of the fundamental mode, can be a good
indicator of how significant P-Delta effects will be for the structure

102
Example Structures – 2-D Buckling-Restrained Brace and Moment Frame

◼ Modal properties can also be used to assess the effects of P-Delta on a structure

◼ This is typically done by specifying a gravity load case that includes geometric stiffness which is
then included in calculating mode shapes and periods

103
Example Structure – 2-D Buckling-Restrained Brace and Moment Frame

Mode 1 without P-Delta: 1.484 s, 0.674 hz Mode 1 without P-Delta: 1.882 s, 0.531 hz
Mode 1 with P-Delta: 1.508 s, 0.663 hz Mode 1 with P-Delta: 1.931 s, 0.518 hz
(Ta = 0.96 s) (Ta = 1.13 s)
1.6% increase in 2.6% increase in
period of vibration period of vibration 104
Example Structure – 2-D Buckling-Restrained Brace and Moment Frame

Wind Displacement without P-Delta: 1.46 in Wind Displacement without P-Delta: 0.99 in
Wind Displacement with P-Delta: 1.51 in Wind Displacement with P-Delta: 1.04 in
3.4% increase in 5.2% increase in
roof displacement roof displacement 105
Next Scene: Case Study Dynamic Analysis for
3-D Structure

106
107
Linear Dynamic Analysis II
109
 Linear Dynamic Analysis Example

◼ A 3-D structural model will be constructed and analyzed for wind loads using a
linear dynamic analysis

◼ Modalproperties will be computed for the structure to identify the dynamic

properties of the structure

◼ Windloads will be analyzed separately in two directions (X and Y) as well as a

combined condition (0.707X + 0.707Y)

◼ Analyses will be run using Modal Response History Analysis and Direct
Integration Response History Analysis to compare the differences

110
Case Study Structure

N
A B C D E F G

Setback at
5th floor 5

20 ft
4

20 ft

3
20 ft

2
20 ft

Y
1

X
30 ft 30 ft 30 ft 30 ft 30 ft 30 ft

Plan View
111
Case Study Structure
B C D E F G 1 2 3 4 5

14 ft
8

7
A

6 @ 12 ft = 72 ft
6

16 ft
Z Z

X Y
30 ft 30 ft 30 ft 30 ft 30 ft 30 ft 20 ft 20 ft 20 ft 20 ft

Elevation View Elevation View

(Grid 1) (Grid B)
112
 Summary of Structure Loading

◼ Risk Category II Building (Office)

◼ Floor Dead Load: 75 psf + 13 psf = 88 psf

◼ Floor Live Load: 50 psf + 15 psf (partition)

◼ Roof Dead and Superimposed Dead Load: 60 psf + 18 psf = 78 psf

◼ Roof Live Load: 20 psf

◼ Seismic Criteria: Seismic Design Category D, SDS = 0.60, SD1 = 0.25

◼ Wind Criteria: V = 95 mph, Exposure Category C

◼ N-S: Buckling-Restrained Braced Frame, E-W: Steel Moment Frame

◼ Software – SAP2000 v23 (CSI)

113
 3-D Analysis Model for Determining Modal
Properties

◼ The 3-D model is built using the same principles as demonstrated in Week 8

◼ 3-D model was constructed using the preliminary sizes for the beams and columns

◼ A rigid diaphragm constraint and idealized boundary conditions were utilized

◼ Gravity and lateral loads were applied to the structural elements

◼ The mass was applied as joint masses at all beam-column intersections

◼ P-Delta effects were included using the geometric stiffness associated with the appropriate
gravity loads

114
Example Structure: 3-D Frame

◼ ASCE 7 Frequency and Period Approximation:

◼ Moment Frame
22.2 22.2
◼ Wind na = 0.8
= 0.8
= 0.55hz
h 102
◼ Seismic Ta = Ct hnx = 0.028(102)0.8 = 1.13s

Y
X

Mode 1: 1.777 s, 0.563 hz 115

Example Structure: 3-D Frame

◼ ASCE 7 Frequency and Period Approximation:

◼ Buckling Restrained Braced Frame

75 75
◼ Wind na = = = 0.73hz
h 102
◼ Seismic
Ta = Ct hnx = 0.03(102)0.75 = 0.96s

Note that lateral torsional

coupling is evident in this
mode shape.

Y
X

Mode 2: 1.362 s, 0.734 hz 116

Example Structure: 3-D Frame

◼ Torsional mode

Note that torsional coupling

with Y-translation is evident
in this mode shape.

Y
X

Mode 3: 0.999 s, 1.001 hz 117

Effective Modal Mass
Period of Frequency of Effective Modal Mass Cumulative Effective Modal Mass Effective Modal Mass Cumulative Effective Modal Mass
Mode
Vibration (s) Vibration (hz) UX UY UZ Sum UX Sum UY Sum UZ RX RY RZ Sum RX Sum RY Sum RZ
1 1.777 0.563 0.7885 0 7.23E-08 0.7885 0 7.23E-08 0 0.0504 0 0 0.0504 0
2 1.362 0.734 0 0.6517 1.56E-17 0.7885 0.6517 7.23E-08 0.1456 8.09E-20 0.1018 0.1456 0.0504 0.1018
3 0.999 1.001 3.00E-20 0.0979 3.15E-16 0.7885 0.7495 7.23E-08 0.0044 1.47E-18 0.6493 0.15 0.0504 0.7511
4 0.698 1.433 0.1333 0 2.29E-07 0.9218 0.7495 3.01E-07 2.21E-18 0.1196 1.79E-20 0.15 0.17 0.7511
5 0.502 1.992 1.13E-16 0.1647 4.80E-14 0.9218 0.9142 3.01E-07 0.255 1.15E-15 0.0113 0.405 0.17 0.7624
6 0.414 2.413 1.96E-15 0.0167 1.66E-14 0.9218 0.9309 3.01E-07 0.0313 2.12E-14 0.1673 0.4363 0.17 0.9297
7 0.392 2.551 0.042 2.67E-17 3.26E-10 0.9638 0.9309 3.02E-07 1.61E-16 0.0294 9.07E-19 0.4363 0.1994 0.9297
8 0.297 3.364 4.57E-15 0.035 6.89E-14 0.9638 0.9659 3.02E-07 0.0516 8.79E-15 0.0074 0.4879 0.1994 0.937
9 0.255 3.916 0.0204 1.26E-18 9.65E-08 0.9842 0.9659 3.98E-07 3.32E-15 0.0185 2.55E-19 0.4879 0.2179 0.937
10 0.232 4.311 1.36E-13 0.0088 1.89E-12 0.9842 0.9747 3.98E-07 0.0175 5.75E-13 0.0316 0.5054 0.2179 0.9686
11 0.212 4.723 3.07E-15 0.0127 8.60E-14 0.9842 0.9874 3.98E-07 0.0326 2.04E-13 0.0056 0.538 0.2179 0.9742
12 0.179 5.591 0.0094 1.00E-14 1.43E-09 0.9936 0.9874 3.99E-07 5.13E-14 0.0075 2.06E-16 0.538 0.2255 0.9742

• 1st mode is primarily UX only, very minimal coupling

• 2nd mode is coupled between UY and RZ due to center of mass and stiffness not
being coincident on every floor
• 3rd mode is coupled RZ and UY due to the same reason as previous
• To capture at least 90% of the effective modal mass, 6 modes would be needed
• To capture at least 95% of the effective modal mass, 10 modes would be required

118
Example Structure: 3-D Frame with Modal Analysis with P-Delta

2.3% increase in 1.3% increase in

period of vibration period of vibration

Mode 1: 1.777 s, 0.563 hz Mode 2: 1.362 s, 0.734 hz

Mode 1 w/ P-Delta: 1.818 s, 0.550 hz Mode 2 w/ P-Delta: 1.379 s, 0.725 hz
119
Example Structure: Static Wind Analysis including P-Delta Effects

Wind X-direction Wind Y-direction

DX = 0.967 in, DY = -0.005 in DX = 0.111 in, DY = 1.296 in

w/ P-Delta: DX = 1.032 in, DY = -0.005 in w/ P-Delta: DX = 0.137 in, DY = 1.331 in
6.7% increase in 2.7% increase in
roof displacement roof displacement 120
 Dynamic Wind Loading

◼ Static wind loads are applied to the structure based on the ASCE 7 Chapter 27 loads

◼ Dynamic Time History Functions are input to SAP2000 which provide the dynamic variation of the
applied loads

◼ These traces were developed in a wind tunnel test, the total response of the structure should also
include the torsional demands

◼ These traces represent the first 300 seconds of a record that is 1950 seconds long

121
 Defining Dynamic Load Cases (SAP2000)

◼ Dynamic analysis (other than typical periodic functions) requires definition of a Time History
Function
◼ Specification of the Load Name (static load case) and Function as well as the number of output
time steps and size is required

122
Modal versus Direct Integration Response History (or Time History)

◼ Modal response history is a much more efficient way to solve linear, classically damped structural dynamic
problems

◼ Direct integration response history is required for nonlinear response history analysis or nonclassically
damped systems but is much more computationally demanding

123
Damping in Dynamic Analysis

◼ For Modal Response History Analysis 2% modal damping was specified in all modes

◼ For Direct Integration Response History Analysis 2% damping was specified in the 1st and 10th
modes (Rayleigh Damping) to capture 95% of effective modal mass

124
 Modal versus Direct Integration Response
History

◼ Several different load cases were run using the same dynamic loading function

◼ Modal Response History is significantly faster than the Direct Integration Response History
analysis

Load Case Analysis Type Linear/Nonlinear Time (mm:ss)

Wind X Modal Response History Linear 0:01
Wind X Direct Integration Response History Linear 5:48
Wind X Direct Integration Response History Nonlinear (P-Delta) 8:10
Wind Y Modal Response History Linear 0:01
Wind Y Direct Integration Response History Linear 5:37
Wind Y Direct Integration Response History Nonlinear (P-Delta) 8:10
Wind XY Modal Response History Linear 0:01
Wind XY Direct Integration Response History Linear 5:35

125
Model Results

Joint 279
Joint 90

X
Y

126
Roof Displacement Response
Wind X-direction Wind Y-direction

Wind 0.707X-direction + 0.707 Y-direction

127
Base Shear Response
Wind Y-direction
Wind X-direction

Wind 0.707X-direction + 0.707 Y-direction

128
Modal versus Direct Integration Response History
Wind X-Direction Displacement
Wind Y-Direction Base Shear

Modal

Direct
Integration

129
Next Scene: Weekly Summary

130
131
Conclusion
133
1
 Recap of Week 11

◼ SDOF Structural Dynamics

◼ MDOF Structural Dynamics

◼ Stiffness, Mass and Damping

◼ Calculating and Using Modal Properties

◼ Case Study of 3-D Linear Dynamic Analysis

134
 What’s coming in Week 12?

◼ Modeling for Wind and Seismic Analysis

◼ Ductility Demands for Wind and Seismic

◼ Special Modeling Requirements for Wind and Seismic Design

135
136