Dynamics of SDOFcourse

Earthquake Engineering
EARTHQUAKE ENGINEERING systems

DYNAMICS OF SDOF SYSTEMS

Nicola Buratti Master course in Civil Enginering

University of Bologna
Earthquake Engineering Dynamics of SDOF systems

Single-Degree of Freedom Systems

1
 Earthquake Engineering Dynamics of SDOF systems

Examples of Single-Degree of Freedom

Systems

The concept of SDOF system will

be also used to reduce the
behaviour of a complex structures
to simpler ones.
1
 Earthquake Engineering Dynamics of SDOF systems

Forces in a SDOF system

a) Undamped SDOF;
SPRING (STIFFNESS)

MASS

b) SDOF with damping;

DAMPER

3
Earthquake Engineering Dynamics of SDOF systems

Different excitations of a SDOF system

GROUND MOTION
APPLIED FORCE p(t)
!!g (t )
ACCELERATION u

DISPLACEMENT total

DISPLACEMENTS
relative

4
 Earthquake Engineering Dynamics of SDOF systems

Parameters to describe a Single Degree of Freedom system:

•STIFFNESS : Linear – elastic relationship

STIFFNESS

Lateral force
f s = k × u(t )
Real case Rigid beam No stiffness beam

K depends on beam
stiffness EIb
 Earthquake Engineering Dynamics of SDOF systems

Parameters to describe a Single Degree of Freedom:

•DAMPER : Damping force

VISCOUS DAMPING COEFFICIENT

Damping force f D = c × u!(t )

c : viscous damping coefficient [F t / L]

It cannot be calculated from geometrical parameters of the structure because

it is not possible to identify all the mechanisms that dissipate vibrational
energy of actual structures (we will discuss later).
 Earthquake Engineering Dynamics of SDOF systems

Newton’s second law of Motion

APPLIED FORCE p(t) Free-body diagram at time t.

Inertia force f = m u!!
I

f D = c × u!(t ) f s = k × u(t )

Dynamic equilibrium, the excitation is p(t)

m u!! + cu! + ku = p (t )
 Earthquake Engineering Dynamics of SDOF systems

FREE VIBRATION - Undamped System

Differential equation of motion for undamped systems with no external force:

Initial conditions (at t=0):

m u!! + ku = 0
u=u(0) and u! = u! (0)

General solution: u(t ) = A cos wnt + B sin wnt

u! (0)
Enforcing the initial conditions: u (t ) = u (0) cos wnt + sin wnt
wn

Tn =
2Π wn
where w = k fn =
n m wn 2Π
Natural circular Natural period Natural frequency of
frequency of vibration of vibration vibration
 Earthquake Engineering Dynamics of SDOF systems
Different representation of the solution FREE VIBRATION - Undamped System

u! (0)
u(t ) = A cos wnt + B sin wnt u (t ) = u (0) cos wnt + sin wnt
wn

A
r
The solution can be also written as:
j
u(t ) = u0 sin(wn t + j)
B

PHASE
AMPLITUDE
Amplitude and phase
2
æ u! (0) ö
r= 2
A +B 2
= (u(0) ) 2
+ çç ÷÷
è wn ø
A
tgj =
B
 Earthquake Engineering Dynamics of SDOF systems
FREE VIBRATION

FREE VIBRATION - Undamped System

Solution: displacement as function of time

u! (0)
u (t ) = u (0) cos wnt + sin wnt
Initial conditions wn
period
2
2 é u! (0) ù
u0 = [u (0)] + ê ú
w
ë n û

wn = k m

2Π
Tn =
wn
 Earthquake Engineering Dynamics of SDOF systems
FREE VIBRATION

FREE VIBRATION - Undamped System

• Solution: displacement as function of time

u! (0)
u (t ) = u (0) cos wnt + sin wnt
wn

• Displacement: In more general form:

u(t ) = A cos wnt + B sin wnt

• Velocity:
Relation between displacement and
u!(t ) = -wn A cos wnt + wn B sin wnt
acceleration for undamped system

• Acceleration: !u!(t ) = -wn 2u(t )

u!!(t ) = -wn A cos wnt - wn B sin wnt
2 2
 Earthquake Engineering Dynamics of SDOF systems

FREE VIBRATION - SDOF with damping

Differential equation of motion of damped system in free

vibration:

m u!!(t ) + c u!(t ) + k u(t ) = 0

Dividing by m:
!u!(t ) + 2 zwn u!(t ) + wn 2u(t ) = 0
where
wn = k m
c c
z= =
2 m wn ccr Damping ratio
Natural circular
frequency of vibration
(of the undamped system)

2k
ccr = 2 m wn = 2 k m =
wn Critical damping
coefficient
 Earthquake Engineering Dynamics of SDOF systems

Free vibrations - SDOF with damping

c c
z= =
2 m wn ccr

Overdamped z >1
Critically damped z =1

Underdamped z < 1 (buildings, bridges, civil structures)

 Earthquake Engineering Dynamics of SDOF systems

Free vibrations - SDOF with damping

c c
u!!(t ) + 2 zwn u!(t ) + wn u(t ) = 0
2
z= =
2 m wn ccr
Form of the solution: u = e st
(s 2 + 2zwn s + wn ) est = 0
2
Substituting in the differential equation gives:

s 2 + 2 zwn s + wn = 0
2
It is satisfied for all values of time t, if :

In general:
(
s1,2 = wn -z ± z 2 - 1 )
Three cases are possible:
z = 1 Þ s1 = s2 = -zwn Critically damped

(
z > 1 Þ s1,2 = wn -z ± z 2 - 1 ) Overdamped

z < 1 Þ s1,2 = w ( -z ± i 1 - z )
n
2 Underdamped
Earthquake Engineering Dynamics of SDOF systems

Free vibrations - SDOF with damping

Critically damped z=1

(
s1,2 = wn -z ± z 2 - 1 ) z = 1 Þ s1 = s2 = -zwn

u (t ) = ( A1 + A2t ) e -wnt

Overdamped z>1
(
s1,2 = wn -z ± z - 1
2
) (
z > 1 Þ s1,2 = wn -z ± z 2 - 1 )
u (t ) = A e
( -z+ )
z 2 -1 wn t
+Ae
( -z- )
z 2 -1 wn t
1 2
 Earthquake Engineering Dynamics of SDOF systems

Free vibrations - SDOF with damping

Underdamped: z < 1

Form of the solution:

Substituting in the differential equation gives:

It is satisfied for all values of time t, if :

s1,2 = wn æç - z ± i 1 - z2 ö÷ if z <1
è ø

u(t ) = A1 e + A2 e
s1 t s2 t
 Earthquake Engineering Dynamics of SDOF systems
FREE VIBRATION

Free vibrations - SDOF with damping

Underdamped: z<1
Circular frequency
wD = wn 1 - z2
s1,2 = wn æç - z ± i 1 - z2 ö÷ of damped system
è ø
Solution: (
u(t ) = e-zwnt A1eiwD t + A2e-iwD t )
The second part can rewritten in terms of trigonometric functions (Euler Formula):

e ± iwDt = cos wDt ± i sin wDt Euler Formula

u(t ) = e-zwn t [A cos wDt + B sin wDt ] Solution

where A and B are constants to be determined using the initial conditions:

 Earthquake Engineering Dynamics of SDOF systems
FREE VIBRATION

Free vibrations - SDOF with damping

Underdamped: z<1
wD = wn 1 - z2
u(t ) = e-zwnt [A cos wDt + B sin wDt ]
Circular frequency of damped
system (related to the natural
Initial conditions – for t = 0 u(0), u! (0) frequency without damping)

The unknown coefficients

are: u! (0) + z wn u (0)
A = u (0) B=
wD
Solution:
é ù
-zw n t ê u! (0) + z wn u (0)
u (t ) = e u (0) cos w D t + sin w D t ú
ê wD ú
êë úû
Note: for z = 0 we obtain the same solution of undamped system
 Earthquake Engineering Dynamics of SDOF systems
FREE VIBRATION

Free vibrations - SDOF with damping

Underdamped: for z = 0.05 or 5% c c

z= =
2 m wn ccr

Tn
TD =
1 - z2
Natural period of
damped system
- zwn t é u! (0) + z wn u(0) ù
u (t ) = e ê u(0) cos wD t + sin wD tú (related to the
ë wn û natural period
without damping)
 Earthquake Engineering Dynamics of SDOF systems
FREE VIBRATION

Free vibrations - SDOF with damping

Effects of damping on the natural vibration frequency

Tn
wD = wn 1 - z2 TD =
1 - z2

For most structures, the

damped properties are
approximately equal to those
of undamped structure.

TD @ Tn wD @ wn
 Earthquake Engineering Dynamics of SDOF systems
FREE VIBRATION

Free vibrations - SDOF with damping

Effects of damping on the rate at which free vibrations decay

Systems subjected to the same initial displacement, different damping ratios:

 Earthquake Engineering Dynamics of SDOF systems
FREE VIBRATION

Free vibrations - SDOF with damping

Effect of damping on the rate of free vibration decay

u (t ) ui æ 2pz ö
u(t ) = e -zwn t
[ A cos wD t + B sin wD t ] = = exp( z wn T D) = expç ÷
u (t + TD ) ui +1 ç 1 - z2 ÷
è ø

TD
Logarithmic decrement
ui 2pz
d = ln = d @ 2p z
ui +1 1 - z2

If z is small, 1 - z2 @ 1
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF

We will analyze the following cases: p(t)

• UNDAMPED SYSTEM SUBJECTED TO HARMONIC FORCE

• DAMPED SYSTEM SUBJECTED TO HARMONIC FORCE

• ENERGY DISSIPATED IN VISCOUS DAMPING

• RESPONSE TO IMPULSES (for UNDAMPED & DAMPED SYSTEM)

• RESPONSE TO A Ground Motion)

• NUMERICAL SOLUTION METHODS

 Earthquake Engineering Dynamics of SDOF systems
FORCED VIBRATION
p(t) Forced Vibration of SDOF - UNDAMPED

Equation of motion: Applied force p(t)

mu!! + ku = p0 sin wt

p0
u!! + w2n u = w2n sin wt
k
Complete solution: w ¹ wn
p0 1
u(t ) = A cos wn t + B sin wn t + sin wt
k 1 - (w / wn )2
Frequency
of SDOF

Complementary solution Particular solution

(see the following) Frequency of
harmonic force
A, B are determined by imposing the initial condition for t=0
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - UNDAMPED

How to obtain the particular
solution?
unknown Applied force p(t)
Setting:

u p = C sin wt

u!!p = -w2C sin wt

Substituting in Equation of motion: w ¹ wn

2 2 p0
( -Cw sin wt ) + wn C sin wt = wn
2
sin wt
k
2 p0
( -Cw ) + wn C =
2

k
2
p0 wn p0 1
C= =
2
k wn - w 2
k 1 - (w / wn )2
 Earthquake Engineering Dynamics of SDOF systems
FORCED VIBRATION

Forced Vibration of SDOF - UNDAMPED

Complete solution
w ¹ wn
Initial condition: B
A
u = u(0) u! = u! (0)

é u% (0) p0 w / wn ù
u (t ) = u (0) cos wnt + ê - 2ú
sin wnt +
ë wn k 1 - (w / wn ) û
$!!!!!!!!#!!!!!!!!"
TRANSIENT

p0 1
+ sin wt
k 1 - (w / wn )2
$!!!#!!!"
STEADY - STATE

By imposing the initial condition we find the two unknown constants A and B
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - undamped

é u! (0) p0 w wn ù
u(t ) = u(0) cos wnt + ê - 2ú
sin wnt From equation of motion ignoring
ë wn k 1 - (w wn ) û the dynamic effect, we obtain the
TRANSIENT static deformation:
po 1 p p
+ sin w t u st = 0 sin wt max (ust )0 = 0
k 1 - (w wn )2 k k
STEADY STATE

Transient
component
continues forever for
undamped system,
but in real systems
because of it will
decrease decay with
t.
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - undamped

The steady-state can be also written as:

p(t) u(t)
é 1 ù
u (t ) = (ust )0 ê 2ú
sin w t where (ust )0 = p0
ë1 - (w wn ) û k
Static displacement

Sign of the factor in

brackets depends on
the ratio of forcing
frequency to the
natural one. When it is
positive it means that
when p is acting to the
rigth the system moves
to the rigth.
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - undamped

The steady-state can be also written as:

é 1 ù
u (t ) = (ust )0 ê 2ú
sin w t
ë1 - (w wn ) û w / wn = 1
Or :
u(t ) = u0 sin (w t - f ) = (ust )0 Rd sin (w t - f )
where
Deformation Response Factor:
u0 1
Rd = =
(ust )0 1 - (w / wn ) 2
phase
Out of phase
u and p have the
æ 0 w " wn
°
same sign
f = çç ° In phase

è180 w ! wn u and p have the

opposite sign
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - undamped

Deformation or Displacement
Response Factor:

Rd =
u0
=
1 w / wn = 1
(ust )0 1 - (w / wn ) 2
is the ratio between dynamic and static
deformation.

1) For small value of frequency ratio, 1) 3) 2)

Rd is similar to 1;

2) If w / wn ! 2 Rd ! 1
w/ wn
dynamic deformation < static deformation
3) If w/ wn similar to 1 Rd !! 1
Then, dynamic deformation due to a rapidly dynamic deformation >> static deformation
force is very small.
 Earthquake Engineering Dynamics of SDOF systems
FORCED VIBRATION

What happens at the resonant Forced Vibration of SDOF - undamped

frequency?
p0 1
u(t ) = A cos wn t + B sin wn t + sin wt
k 1 - (w / wn )2

Complementary solution
Particular solution
(see the following)

If w =the
wn particular solution defined previously is no longer valid because it is
contained in the complementary solution. The new particular solution is:

The new complete solution, considering at rest conditions at t =0 is:

 Earthquake Engineering Dynamics of SDOF systems
FORCED VIBRATION

What happens at the resonant

frequency?

Sec. 3.1 Harmonic Vibration of Undamped Systems 71

Figure 3.1.4 Response o f undamped system to sinusoidal force o f frequency w = a,,;

u(0) =
ii(0) = 0.
The displacement grows infinitely but it becomes infinite after an infinite time.
 Earthquake Engineering Dynamics of SDOF systems
FORCED VIBRATION

Damped Systems subjected to Harmonic Force

p0 sin wt
Equation of motion:
mu!! + cu! + ku = p0 sin wt

Initial condition:

u = u(0) u! = u! (0)

Applied force: p(t ) = p0 sin wt

w ¹ wn
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

Applied force
Equation of motion:

mu!! + cu! + ku = p0 sin wt

Complete solution: w ¹ wn

u(t ) = e -zwnt [ A cos wD t + B sin wD t ] + C sin w t + D cos w t

TRANSIENT STEADY STATE

wD = wn 1 - z2 p 1 - (w / wn )
2
C= 0
By imposing initial condition (at
[ n ]
k 1 - (w w )2 2 + [2x (w w )]2
n

t=0) we find the two unknown p0 - 2x w wn

D=
constants A and B. [ ]
k 1 - (w w )2 2 + [2x (w w )]2
n n
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

Solution:

u(t ) = e -zwnt [ A cos wD t + B sin wD t ] + C sin w t + D cos w t

TRANSIENT STEADY STATE

Transiet component
decay with t for
damped system, at a
rate depending on
frequency ratio and
damping ratio.
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

Response for: w = wn

u(t ) = e -zwnt
[ A cos wD t + B sin wD t ] + C sin w t + D cos w t
1 - (w / wn )
2
p p0 - 2x w wn
C= 0 D=
[ n ]
k 1 - (w w )2 2 + [2x (w w )]2
n [ ]
k 1 - (w w )2 2 + [2x (w w )]2
n n

In this case C=0 D= (ust )0 /2⇣

Assuming zero initial conditions we also obtain:
p
A = (ust )0 /2⇣ B = (ust )0 /2 1 ⇣2
1 é -zwnt æç z ö ù
u (t ) = (u st )o êe cos wDt + sin wDt ÷ - cos wnt ú
2z ê ç 1 - z 2 ÷ úû
ë è ø
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

Response for: w = wn
1 é -zwnt æç z ö ù
u (t ) = (u st )o êe cos wDt + ÷
sin wDt - cos wnt ú
2z ê ç 1 - z 2 ÷ úû
ë è ø
For lightly damped systems we can approximate the solution as: !D ' !n
1 ⇣!n t
u(t) ' (ust )0 (e 1)cos!n t
2⇣
Envelope curve (ust )o Bounded value
u (t ) =
2z

In comparison with
undamped system
damping lowers
each peak and
limits the
response.
for z = 0.05 or 5%
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

Response for: w = wn

• For damping ratio of 5% and 10 % -> the displacement is stable;

• For a ligthly damped system 1 % -> the maximum displacement increases

continuosly and it is similar to undamped systems
 Earthquake Engineering Dynamics of SDOF systems

Maximum Deformation and Phase Lag: Forced Vibration of SDOF - damped

The steady-state deformation of the system due

to harmonic force can be written as: w = wn
u(t ) = u0 sin (w t - f ) = (ust )0 Rd sin (w t - f )
1
Where: 𝑢" = 𝐶 % + 𝐷 % ; ϕ = 𝑡𝑎𝑛-. (− )
2

The deformation response factor and the phase

become:

𝑢" 1
𝑅5 = =
(𝑢67 )" [1 − 𝜔⁄𝜔; % ]% + [2ζ(𝜔⁄𝜔; )]%
2ζ(𝜔⁄𝜔; )
ϕ = 𝑡𝑎𝑛-.
1 − 𝜔⁄𝜔; %

If the system is undamped we have:

u0 1
z =0 Rd = =
(ust )0 1 - (w / wn ) 2
 Earthquake Engineering Dynamics of SDOF systems

Maximum Deformation and Phase Lag: Forced Vibration of SDOF - damped

The steady-state deformation of the system due

to harmonic force can be written as: w = wn
u(t ) = u0 sin (w t - f ) = (ust )0 Rd sin (w t - f )
1
Where: 𝑢" = 𝐶 % + 𝐷 % ; ϕ = 𝑡𝑎𝑛-. (− )
2

The deformation response factor and the phase

become:

𝑢" 1
𝑅5 = =
(𝑢67 )" [1 − 𝜔⁄𝜔; % ]% + [2ζ(𝜔⁄𝜔; )]%
2ζ(𝜔⁄𝜔; )
ϕ = 𝑡𝑎𝑛-.
1 − 𝜔⁄𝜔; %

Damping reduces
deformation response
factor Rd at all
excitation frequency.
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

Deformation or Displacement
Response Factor:

w / wn = 1
1) For small value of frequency ratio,Rd is
similar to 1. Then, dynamic and static
deformations are essentially the same
indipendent from damping.

2) If w / wn ! 2 Rd ! 1
dynamic deformation < static deformation, 1) 3) 2)
indipendent from damping.
Then, dynamic deformation due to a
rapidly force is very small.
w/ wn
3) If w/ wn similar to 1 Rd !! 1
In this area the
dynamic deformation >> static deformation response is controlled
by damping.
Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

Phase Lag: With damping,the phase is not necessarily only 0° or 180° as

in the undamped case.

2ζ(𝜔⁄𝜔; )
ϕ = 𝑡𝑎𝑛-.
1 − 𝜔⁄𝜔; % In phase
Out of phase

p(t) u(t) p(t) u(t)

In phase Out of phase

u and p have the u and p have the

same sign. opposite sign.
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

Considering the steady state vibration, the energy dissipated in viscous

damping in one cycle of armonic vibration is:
2p w
2p w
E D = ò f D du = ò (cu! )u! dt = ò0
0
cu! 2 dt

w 2
[wu0 cos(wt - f)]2 dt = pcwu02 = 2pz
2p w
= cò ku 0
0 wn
Proportional to the square of amplitude of motion u0

The energy input due to the applied force p(t) for

each cycle of vibration is:
2p w
Et = ò p(t ) du = ò p(t )u! dt Et = 2pz
w 2
ku 0
0
wn
[ p0 sin wt ][wu0 cos(wt - f)]dt = pp0u0 sin f
2p w
=ò
0 See Chopra for derivation
Proportional to the amplitude of motion u0.

The steady state amplitude u0 corresponds to Ed=Et

 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

The energy dissipated in viscous damping in one cycle of armonic vibration is:
f D = cu! (t ) = cwu0 cos(wt - f)
= cw u02 - u02 sin 2 (wt - f)
2 2
æ u ö æ fD ö
= cw u02 - [u (t )] çç ÷÷ + çç ÷÷ = 1
2
è u0 ø è cwu0 ø
Area of the ellipse represents
dissipated energy

Hysteresis loop for

Hysteresis loop for
spring and viscous
viscous damper.
damper in parallel.
Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

Specific Damping Capacity:

ED
ESo

Specific Damping Factor:

1 ED
x=
2p ESo

Equivalent viscous damping

from exp. test conducted at
can be evaluated as
1 ED
z eq =
4p ESo
 Earthquake Engineering Dynamics of SDOF systems

Forced Vibration of SDOF - damped

BASE GROUND-MOTION

The total displacement, velocity and acceleration of the mass are:

𝑢7 𝑡 = 𝑢 𝑡 + 𝑢@ (𝑡) 𝑢7 𝑡
𝑢 𝑡
𝑢̇ 7 𝑡 = 𝑢̇ 𝑡 + 𝑢̇ @ (𝑡)
𝑢̈ 7 𝑡 = 𝑢̈ 𝑡 + 𝑢̈ @ (𝑡)

Hence, the inertial force becomes:

𝑓D = 𝑚𝑢̈ 7 𝑡 = 𝑚𝑢̈ 𝑡 + 𝑚𝑢̈ @ (𝑡)
𝑢@ (𝑡)

The equation of motion becomes: Ground acceleration [m/s2] 2

1.5
1
𝑚𝑢̈ 7 𝑡 + 𝑐𝑢̇ 𝑡 + 𝑘𝑢 𝑡 = 0 0.5
𝑚[𝑢̈ 𝑡 + 𝑢̈ @ 𝑡 ] + 𝑐𝑢̇ 𝑡 + 𝑘𝑢 𝑡 = 0 0
0 10 20 30 40
-0.5
-1
𝑚𝑢̈ 𝑡 + 𝑐𝑢̇ 𝑡 + 𝑘𝑢 𝑡 = −𝑚𝑢̈ @ 𝑡 -1.5
Time [s]
 Earthquake Engineering Dynamics of SDOF systems

Methods of solution for systems under arbitrary forces

Equation of motion :
m u!! + cu! + ku = p(t )
.
Initial conditions must be specified in order to define the problem completely.
Typically, the structure is «at rest» bafore a dynamic aìexcitation, hence the initial
velocity and displacment are zero.

Initial conditions
Main Dynamic Analysis Methods :

1) Duhamel’s Integral;

2) Numerical Methods;
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Duhamel’s Integral

The response of a SDOF subjected to an arbitrary force p(t) can be

obtained decomposing the force in a sequence of infinitesimally short
impulses.

Hence, the response of the system at time t can be obtained by adding the
responses of all impulses up to that time. This method can be used ONLY
for LINEAR SYSTEMS

The impulse can be modeled as a Dirac delta function δ(t) for continuous-
time systems.

δ = 0 for t ≠ t
δ = 1 for t = t
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Duhamel’s Integral

According to Newton’s second law of motion:

𝑝 𝑡 = 𝑚𝑢̈
That, integrated with respect to t, gives:

7L
Impulse
J 𝑝 𝑡 𝑑𝑡 = 𝑚(𝑢̇ % − 𝑢̇ . ) = 𝑚∆𝑢̇
7M Theorem
Magnitude of
the impulse:

Hence, a unit impulse (magnitude equal 1) imparts to mass, m, the following 1

velocity:
u! (t) =
m
This formulation is also applicable to spring or dampers: in fact, if the
force acts for an infinitesimally short duration,the spring and the damper
has no time to respond.
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Duhamel’s Integral

Unit impulse causes free vibration of the SDOF due to initial velocity obtained before.
Assuming that the displacement of the system is zero up to the impulse, the response
of a damped system to a unit impulse can be written recalling the general solution for
damped systems:
é u! (0) + z wn u (0) ù
u (t ) = e -zw n t
ê u (0) cos w D t + sin w D t ú wD = wn 1 - z2
ë wD û

-zw n [ t -t ] é 1 ù
u (t ) = e ê sin[w D (t - t )]ú t≥t
ë mw D û
(response of Viscously damped system)

If z = 0 we obtain the same solution of undamped system

é 1 ù
u (t ) = ê sin[wn (t - t )]ú t≥t
ë mwn û
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Duhamel’s Integral

The unit impulse response function is commonly called h(.t - t )

The unit impulse response function for damped and undamped systems are shown in
the following graph:

é 1 ù
h(t - t ) h(t - t ) = u (t ) = ê sin[wn (t - t )]ú
ë mwn û

é 1 ù
h(t - t ) = u (t ) = e -zw n [t -t ] ê sin[w D (t - t )]ú
ë mw D û
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Duhamel’s Integral

Once the response of the dynamic system to a unit impulse is known, the solution can be
computed for a force varying arbitrarily with time.

The force p(t) can be represented as a sequence of infinitesimally short impulses.

The response of a linear dynamic system to the impulse at time t of magnitude p(t)dt, is
this magnitude times the unit impulse-response function.

𝑑𝑢 𝑡 = 𝑝 𝜏 𝑑𝜏 ℎ(𝑡 − 𝜏) 𝑡>𝜏

The response of the system at time t is the sum of the responses to all impulses up to
that time:
7 Convolution
𝑢 𝑡 = J 𝑝 𝜏 ℎ 𝑡 − 𝜏 𝑑𝜏 Integral
"
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Duhamel’s Integral

Substituting the unit impulse response for the

damped system, we obtain Duhamel’s
Integral:
7
1
𝑢 𝑡 = J 𝑝 𝜏 𝑒 -STU (7-V) 𝑠𝑖𝑛 𝜔1 (𝑡 − 𝜏) 𝑑𝜏
𝑚𝜔1 "
assuming «at rest» initial conditions:
𝑢 0 =0 𝑢̇ 0 = 0

Duhamel’s Integral is based on the principle of

superposition of effects, valid only for linear elastic
systems.
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

For more general solution of the equation motion of SDOF systems

subjected to an arbitrary force it is possible to use TIME-STEPPING
METHODS.

In the present section we will consider only two numerical methods

for solving II order differential equation:

1. CENTRAL DIFFERENCE

2. NEWMARK’s METHOD

The can be used for linear and non linear systems.

 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

The considered methods are based on the discretization of the applied force p, the
acceleration and velocity with respect to time.
In fact, the continuous function p(t) is substituted with discrete values pi, computed at
the discrete time instant ti.

A constant time interval is commonly adopted: Dti = ti +1 - ti

As a result, the response of
the system (displacements,
accelerations or velocities) are
also obtained for the discrete
time instant ti.
54
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

At each time i the following equation has to be satisfied:

𝑚𝑢̈ Z + 𝑐𝑢̇ Z + (𝑓6 )Z = 𝑝Z

Where (𝑓6 )Z represents the resisting force at time i (𝑘𝑢Z ) if the system is linear
elastic, but depends on the previous history of displacement if the system is
nonlinear.
The time-stepping methods allow to solve the previous equation for each instant
i, giving the desired response at each ti.
Approximate numerical procedures are adopted to step from time i to i+1. The
three important requirements for a numerical procedure are:
1. convergence - as the time step decreases, the numerical solution should
approach the exact solution;
2. stability - the numerical solution should be stable in the presence of
numerical round-off errors;
3. accuracy - the numerical procedure should provide results that are close
enough to the exact solution.
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

P(t)
CENTRAL DIFFERENCE

Pi-1
Pi Pi+1 Taking constant time steps, Dti = Dt, the
central difference expressions for velocity gives:
𝑢Z[. − 𝑢Z-.
𝑢̇ Z =
t 2∆𝑡
ti-1 ti ti+1 Similarly:
u Dt Dt ] -] ] -]
𝑢̇ Z-./% = ^ ^_M 𝑢̇ Z[./% = ^`M ^
∆7 ∆7
ui
ui-1 ui+1
The central difference expressions for acceleration
gives:
t 𝑢̇ Z[./% − 𝑢̇ Z-./%
ti-1 ti ti+1 𝑢̈ Z =
Dt Dt
∆𝑡
𝑢̇ Substituting the expressions of velocity in the latter
equation:
𝑢̇ i-1/2
𝑢̇ i 𝑢Z[. − 2𝑢Z + 𝑢Z-.
𝑢̈ Z =
𝑢̇ i+1/2 ∆𝑡 %
t
ti-1/2 ti ti+1/2
Dt/2 Dt/2
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

P(t)
CENTRAL DIFFERENCE

Pi Pi+1
Pi-1
In the case of a inear elastic system the previous
expressions can be substituted in the equation of
t motion as follows:
ti-1 ti ti+1
u Dt Dt
𝑚𝑢̈ Z + 𝑐𝑢̇ Z + 𝑘𝑢Z = 𝑝Z
ui
ui-1 ui+1
𝑢Z[. − 2𝑢Z + 𝑢Z-. 𝑢Z[. − 𝑢Z-.
𝑢̈ Z = 𝑢̇ Z =
∆𝑡 % 2∆𝑡
t
ti-1 ti ti+1

𝑢̇
Dt Dt 𝑢Z[. − 2𝑢Z + 𝑢Z-. 𝑢Z[. − 𝑢Z-.
𝑚 + 𝑐 + 𝑘𝑢Z = 𝑝Z
∆𝑡 % 2∆𝑡
𝑢̇ i-1/2
𝑢̇ i

𝑢̇ i+1/2
t
ti-1/2 ti ti+1/2
Dt/2 Dt/2
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

CENTRAL DIFFERENCE

Rearranging the previous equation, for linear system we obtain a linear equation with
only one unknown parameter, ui+1

é m c ù é m c ù é 2m ù
ê + úui +1 = pi - ê + úui -1 - êk - u
2ú i
ë (Dt ) 2Dt û ë (Dt ) 2Dt û ë (Dt ) û
2 2

The displacement at the time step i+1 is only function

of the displacement at previous time steps (i and i-1).
For this reason the central difference method is an
explicit method.

Requirement for stability of numerical solution: Dt 1

<
Tn p
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

NEWMARK’s METHODS

The Newmark’s method is a stepping numerical procedure that allows to evaluate the
response of the structure. It is an implicit method: the expressions that allow to evaluate
the response at a certain time step depend on quantities (i.e. accelerations) related to the
same time step. Hence, various iterations might be needed to find the response.

According to this method, the applied force p is discretized and an assumption must be
made about the variation of the acceleration within the time step. In fact, Newmark
developed a family of methods (not just one) depending on the selected fuction of
acceleration vs time.

At first, two particular cases of Newmark’s method will be presented and then a general
formulation for the family of methods will be given.
Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

NEWMARK’s METHODS:
Constant acceleration method
𝑢̈
𝑢̈ Z[.
Constant acceleration:
𝑢̈ Z (𝑢̈ Z[. + 𝑢̈ Z )
𝑢̈ =
2
t
ti ti+1
𝑢̇ Integrating :
𝑢̇ Z[. 𝜏(𝑢̈ Z[. + 𝑢̈ Z )
𝑢̇ 𝜏 = 𝑢̇ Z +
2
𝑢̇ Z hence
∆𝑡(𝑢̈ Z[. + 𝑢̈ Z )
𝑢̇ Z[. = 𝑢̇ Z +
t 2
ti ti+1
𝑢 Integrating again:
𝜏 % (𝑢̈ Z[. + 𝑢̈ Z )
𝑢 𝜏 = 𝑢Z + 𝑢̇ Z 𝜏 +
𝑢Z[. 4
𝑢Z hence
∆𝑡 % (𝑢̈ Z[. + 𝑢̈ Z )
t 𝑢Z[. = 𝑢Z + 𝑢̇ Z ∆𝑡 +
ti Dt ti+1 4
Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

NEWMARK’s METHODS:
The method is implicit: the
Constant acceleration method velocity and displacement at
the instant i+1 depend on the
𝑢̈ acceleration at the same time
𝑢̈ Z[. step
Constant acceleration:
𝑢̈ Z (𝑢̈ Z[. + 𝑢̈ Z )
𝑢̈ =
2
t
ti ti+1
𝑢̇ Integrating :
𝑢̇ Z[. 𝜏(𝑢̈ Z[. + 𝑢̈ Z )
𝑢̇ 𝜏 = 𝑢̇ Z +
2
𝑢̇ Z hence
∆𝑡(𝑢̈ Z[. + 𝑢̈ Z )
𝑢̇ Z[. = 𝑢̇ Z +
t 2
ti ti+1
𝑢 Integrating again:
𝜏 % (𝑢̈ Z[. + 𝑢̈ Z )
𝑢 𝜏 = 𝑢Z + 𝑢̇ Z 𝜏 +
𝑢Z[. 4
𝑢Z hence
∆𝑡 % (𝑢̈ Z[. + 𝑢̈ Z )
t 𝑢Z[. = 𝑢Z + 𝑢̇ Z ∆𝑡 +
ti Dt ti+1 4
Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

NEWMARK’s METHODS:
Linear acceleration method

𝑢̈
𝑢̈ Z[.
Constant acceleration:
𝑢̈ Z (𝑢̈ Z[. − 𝑢̈ Z )
𝑢̈ 𝜏 = 𝑢̈ Z + 𝜏
∆𝑡
t
ti ti+1
𝑢̇ Integrating :
𝑢̇ Z[. 𝜏 % (𝑢̈ Z[. − 𝑢̈ Z )
𝑢̇ 𝜏 = 𝑢̇ Z + 𝑢̈ Z 𝜏 +
2 ∆𝑡
𝑢̇ Z hence
∆𝑡
𝑢̇ Z[. = 𝑢̇ Z + (𝑢̈ Z[. + 𝑢̈ Z )
t 2
ti ti+1 Integrating again:
𝑢 𝜏 % 𝜏 b (𝑢̈ Z[. − 𝑢̈ Z )
𝑢 𝜏 = 𝑢Z + 𝑢̇ Z 𝜏 + 𝑢̈ Z +
𝑢Z[. 2 6 ∆𝑡
hence
𝑢Z
∆𝑡 %
𝑢Z[. = 𝑢Z + 𝑢̇ Z ∆𝑡 + (𝑢̈ Z[. + 2𝑢̈ Z )
6
t
ti Dt ti+1
Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

NEWMARK’s METHODS:
The method is implicit: the
Linear acceleration method velocity and displacement at
the instant i+1 depend on the
𝑢̈ acceleration at the same time
𝑢̈ Z[. step
Constant acceleration:
𝑢̈ Z (𝑢̈ Z[. − 𝑢̈ Z )
𝑢̈ 𝜏 = 𝑢̈ Z + 𝜏
∆𝑡
t
ti ti+1
𝑢̇ Integrating :
𝑢̇ Z[. 𝜏 % (𝑢̈ Z[. − 𝑢̈ Z )
𝑢̇ 𝜏 = 𝑢̇ Z + 𝑢̈ Z 𝜏 +
2 ∆𝑡
𝑢̇ Z hence
∆𝑡
𝑢̇ Z[. = 𝑢̇ Z + (𝑢̈ Z[. + 𝑢̈ Z )
t 2
ti ti+1 Integrating again:
𝑢 𝜏 % 𝜏 b (𝑢̈ Z[. − 𝑢̈ Z )
𝑢 𝜏 = 𝑢Z + 𝑢̇ Z 𝜏 + 𝑢̈ Z +
𝑢Z[. 2 6 ∆𝑡
hence
𝑢Z
∆𝑡 %
𝑢Z[. = 𝑢Z + 𝑢̇ Z ∆𝑡 + (𝑢̈ Z[. + 2𝑢̈ Z )
6
t
ti Dt ti+1
 Earthquake Engineering Dynamics of SDOF systems

Systems under arbitrary forces: Numerical Methods

NEWMARK’s METHODS
Hence, in general the following equations have been proposed by Newmark to define
velocoties and displacements at the time i+1:

𝑢̇ Z[. = 𝑢̇ Z + 1 − 𝛾 ∆𝑡 𝑢̈ Z + 𝛾∆𝑡 𝑢̈ Z[.

𝑢Z[. = 𝑢Z + ∆𝑡 𝑢̇ Z + 0.5 − 𝛽 ∆𝑡 % 𝑢̈ Z + 𝛽∆𝑡 % 𝑢̈ Z[.

It can be observed that if 𝛾 = 1/2 and 𝛽 = 1/4 these expressions become those of the
constant acceleration method, while if 𝛾 = 1/2 and 𝛽 = 1/6 they are those of the linear
acceleration method.

For linear systems Newmark’s method becomes expicit: these expression can be substituted in
the equation of motion giving:

𝑚𝑢̈ Z[. + 𝑐 𝑢̇ Z + 1 − 𝛾 ∆𝑡 𝑢̈ Z + 𝛾∆𝑡 𝑢̈ Z[. + 𝑘 𝑢Z + ∆𝑡 𝑢̇ Z + 0.5 − 𝛽 ∆𝑡 % 𝑢̈ Z + 𝛽∆𝑡 % 𝑢̈ Z[. = 𝑝Z[.

[𝑚 + 𝑐 𝛾∆𝑡 + 𝑘(𝛽∆𝑡 % )]𝑢̈ Z[. = 𝑝Z[. − 𝑐[𝑢̇ Z + 1 − 𝛾 ∆𝑡 𝑢̈ Z ] − 𝑘 𝑢Z + ∆𝑡 𝑢̇ Z + 0.5 − 𝛽 ∆𝑡 % 𝑢̈ Z

The response at the time step i+1 is only function of the response at previous time steps (i and i-1).
The equation of motion can be solved without iterations.