Chapter 4- Periodic loading

Chapter 4: Response to periodic loading

4.1 Fourier series expressions of periodic loading

Periodic loading is the load varying on time t where p x + T p = p ( x )

Periodic loading of period T p can be expressed in a Fourier series
series form
with harmonic loading components at discrete values of frequency.
frequency.
p( x)

Trigonometric form
x

Tp

Tp

Tp

n =1

n =1

p ( t ) = a0 + an cos nt + bn sin nt

(4.1)

Tp

(4.2)
n = n 1 = n
in which
Tp

1 Tp
a0 = 0 p ( t ) dt
Tp

2 Tp
& harmonicharmonic-amplitude coefficients an = 0 p ( t ) cos t dt n = 1, 2,3,...
Tp

can be evaluated by
(4.3)

2 Tp
bn = p ( t ) sin t dt n = 1, 2,3,...
Tp 0

Response is obtained by adding up the responses to the individual

individual
1
harmonic loadings

Chapter 4- Periodic loading

Undamped systems:
systems: (omitting the transient response term) response by

b 1
a sinesine-wave loading term bn sin 2 n : v ( t ) = n
sin nt (4.4)

Tp
k 1 n 2

2 n
a 1
(4.5)
a cosinecosine-wave term an cos
: v (t ) = n
cos nt
with

Tp

n n

k 1 n 2

(4.6)

v0 = a0 k

(4.7)

1
1
a
a cos nt + bn sin n t )
+
0
2 ( n
k
n =1 1 n

(4.8)

the constant load a0 is the static deflection

Total periodic response:

v (t ) =

For damping system, using (3.17) the total response is given by

v (t ) =

1
1
2
2 an n + bn (1 n ) sin n t
(a0 +
2
2

k
n =1 (1 n ) + ( 2 n )

2
+ an (1 n ) 2 bn n cos n t} )

(4.9)
2

Chapter 4- Periodic loading

m
k
2

Example 4.1

p (t )

k
2

One-story building system is subjected the periodic load as

shown. This load consists of the positive portion of a simple
sine function having amplitude p0 = 54 kN & the period
Tp = 4 3 T = 1.733 s . Plot the response of system.

p0 sin 2T t

p( t )

Solution:

Determine Fourier coefficients by (4.3)

p0

t
Tp 2

Tp 2

Tp 2

1
an =
Tp

Tp 2

Tp 2

1
bn =
Tp

1
Tp

Tp 2

p0 sin

p
2 t
dt = 0
Tp

n odd
0
2 t
2 n t

p0 sin
cos
dt = p0 1
Tp
Tp
1 n 2 n even

Tp 2

a0 =

p0
2 t
2 n t

p0 sin
dt = 2
sin
Tp
Tp
0

n =1
n >1

Chapter 4- Periodic loading

Using (4.1)

p (t ) =

p0
2
2
2

1 + sin 1t cos 2 1t cos 4 1t cos 6 1t + ...

2
3
15
35

rad
where 1 = 2 Tp = 2 1.733 = 3.625 s

Using (4.8), total response :

p
v (t ) = 0
k

( k = 180 kN

(a)

cm )

8
1
8

sin 1t + cos 2 1t + cos 4 1t + ...

1 +
7
15
60

(b)

The load (a) & response (b) are plotted as shown

p (t )

The period load (a) (kN)

Response (b) of the building

v ( t ) (cm )

54 kN

Tp = 1.733 s
4

Chapter 4- Periodic loading

When p ( t ) is of arbitrary periodic form, the integrals in (4.3)

(4.3) must be
evaluated numerically by applying approximate integration

Tp into N equal intervals t = Tp N

Divide the period

Evaluate ordinates of the integrands at discrete values of t = tm = m t

( m = 1, 2,...N ) and denote them by q0 , q1 , q2 .., qN
Then apply the trapezoidal rule of integration

Tp

q
q N 1
p ( t ) dt  t 0 + qm + N
2
2 m =1

In practical solutions,

q0 & q N

Tp

can be set to equal to zero

zero
N 1

p ( t ) dt  t qm

(4.10)

m =1

The harmonic amplitude coefficients of (4.3) then may be expressed

expressed

a0
2t N
an =
qm
Tp m =1

bn

Chapter 4- Periodic loading

where

2 p ( tm )

qm = p ( t m ) cos n ( m t ) (4.11)

p ( t m ) sin n ( m t ) 5

Exponential form (Complex form)

Exponential form of (4.1) is obtained by substituting the inverse

inverse Euler
relations

cos n t =

1 int
e + e int
2

into (4.1) & (4.3), we have

and

p(t ) =

sin n t =

Pe

m =

in which

Pn =

1
Tp

It should be noted:
their sum is real.

Tp

p ( t )e in t dt

i n t

i int
e e int
2

n = 0, 1, 2,...

)
(4.12)

(4.13)

Pm & P m are complex conjugate pairs, therefore,

If the loading is expressed in terms of individual harmonics of (4.12),

total response of a damped SDOM can be found by superposition:
- Equation of motion due to unit complex forcing function:
function:

mv( t ) + cv ( t ) + kv ( t ) = 1.e i t

- SteadySteady-state solution has the following form

v ( t ) = H ( ) e i t

Chapter 4- Periodic loading

where H ( ) , the complexcomplex-frequencyfrequency-response function, can be

obtained by differentiating and substituting into EoM:
EoM:
2
H ( ) ( i ) m + ( i ) c + k e i t = e i t

H ( ) =

Complex-frequencyresponse-function

k + 2i + 1
2

(4.14)

Thus, H ( n ) correspondingly to
1
H ( n1 ) =
the n-th term of loading (4.12)
2
k n + 2i n + 1
with n = n1 can be written as
where

1 =

2
Tp

1 =

H n H ( n1 ) =

n =

n1

= n1

k n2 12 + 2in1 + 1

(4.15)

(4.16)

H n expresses the harmonic response due to a unit harmonic loading

at the frequency n
From principle of superposition, the
total steadysteady-state response to any
periodic function can be written as

Chapter 5- Impulsive loading

p (t )

v (t ) =

1
H n Pne in1t (4.17)

7
2 n=

Chapter 5: Response to impulsive loading

5.1 General nature of impulsive loading

Impulsive load consists of

- a single principal impulse that is typically large in magnitude
- generally is of relatively short duration t1
t Maximum response, including displacements, velocity,
t1
acceleration, will be reached in a very short time.
So, damping forces absorb less energy from structure & have much
much less
importance. In order to simplify, only undamped response are consider
The response is divided into two phases corresponding to
- first, the forcedforced-vibration phase in the loading interval ( 0 t t1 )
- second, the freefree-vibration phase ( t t1 )
Maximum response by impulse loading depends only on the ratio of
impulse duration t 1 to natural period T of structure, i.e on t1 T
Maximum response vmax is not strongly
influenced by shape of loading but rather
depends mainly on magnitude of the impulse:
impulse:

t1

I = p ( t ) dt
0

Impulsive or shock loads are of importance in

design of vehicles such as trucks, traveling cranes

Chapter 5- Impulsive loading

5.2 SineSine-wave impulse

The response is divided into two phases corresponding to

p0 sin t

p (t )

Phase I ( 0 t t1 ) : Structure is subjected to harmonic

loading starting from rest. Response consisting of
steady & transient terms is (from (3.19))

p0

v (t ) =

t
t

t1

Phase 1

Phase 2

p0 1
( sin t sin t )
k 1 2

(5.1)

Phase II ( t = t t1 0 ) : Free vibration depends on v ( t1 ) & v ( t1 ) at the

end of Phase I.

v ( t = 0) = v ( t1 )
Like (2.19) :

v(t ) =

&

v ( t1 )

v ( t = 0) = v ( t1 )

(5.2)

sin t + v ( t1 ) cos t

(5.3)

Chapter 5- Impulsive loading

Depend on the ratio t1 T the maximum value of response occurs

occurs in
Phase I or Phase II.
- If

vmax in Phase I t < t1 = , from (5.1):

dv ( t ) p0 1
=
( cos t cos t ) = 0
dt
k 1 2

cos t = cos t t = 2 n t ( n = 0, 1, 2,...)

Substitute (5.4) into (5.1), vmax is determined.
, i.e = < 1
n = 1 & sign ((-) in (5.4), we have: t =

Special case,
taking

(5.4)

2
2
=
1+ 1+1

Substituting it into (5.1) gets the peak value.

If

> 1 , i.e > : vmax

will fall just in Phase II,

II, freefree-vibration

response with the initial conditions

(5.1) at t = t1 =

v ( t1 ) & v ( t1 )

determined
determined from
10

Chapter 5- Impulsive loading

v ( 0) =

p0 1

0 sin
2
k 1

& v ( 0) =

This freefree-vibration response,

response, (2.20) :

p0 1

1 cos
2
k 1

v ( t ) = cos ( t )

(5.5)

v ( t1 )
2
p0

2 + 2 cos
with amplitude:
amplitude: =
+ v ( t1 ) =
2

k (1 )

2

(5.6)

Dynamic magnification factor:

D=

vmax

2 1 + cos
=
=
2
p0 k p0 k 1

by using the trigonometric identity

D=

2 (1 + cos ) = 2 cos

cos
2
1
2

(5.7)

11

Chapter 5- Impulsive loading

p (t )

5.3 Rectangular impulse

t1
Phase 1

Phase I 0 t t1 : The load applies suddenly &

remains constant during this phase (step
(step loading)
loading) .
-Particular solution is simply the static deflection:
t

p0

t
Phase 2

v p = p0 k

(5.8)

p0
(5.9)
(1 cos t )
k T
p
From dv ( t ) dt = 0 sin t = 0 t = At t = = : vmax = 2 0
k
2
T
t
(5.10)
1
, we have D = 2 with
Thus, if t t1 , i. e t1
0.5
2
T
Phase II ( t = t t1 0 ) : Like (5.3), freefree-vibration response:
v ( t1 )
v ( t ) = cos ( t )
or
v(t ) =
sin t + v ( t1 ) cos t

12
2 12
v ( t ) 2

2
p0 2 2
p
t
2

1
t1 + 1 cos t1 = 0 2sin 1
= vmax =
+ v ( t1 ) = sin
k
T
T
k
T

(5.10)
-General solution at the rest initial conditions: v ( t ) =

Thus,

D=

vmax
t
= 2sin 1
po k
T

with

t1
0.5
T

(5.11)
12

Chapter 5- Impulsive loading

p (t )

p0

5.4 Triangular impulse

Triangular impulse p ( t ) = p0 1 causes response

t1
consisting 2 phases
Phase I ( 0 t t1 ) :
p t
t
v p ( t ) = 0 1
Particular
solution
p ( t ) = p0 1
k t1
t1 t
- At the restrest-initial conditions, general solution:

t1

Phase 1

v (t ) =

Phase 2

dv ( t ) dt = 0 vmax
at

t=

p0 sin t
t
cos t + 1

k t1
t1
p0 arctg t
= 2 1

k
t1

2arctg t1

t1

(5.12)

t1
0.371
T

From (5.12): The condition to occur

maximum response in Phase I:

(5.13)

arctg t1
D = 2 1

t1

(5.14)
13

Chapter 5- Impulsive loading

Phase II

(t

Free-vibration response determined from the

= t t1 0 ) : Free-

initial conditions corresponding at the end of Phase I, i.e at

v ( t = 0 ) = v ( t1 ) & v ( t = 0 ) = v ( t1 )

v ( t1 ) =
v ( t1 )

v(t ) =
sin t + v ( t1 ) cos t with

v t =
( 1)

v t = cos t
Like (2.19):

( )

t = t1 :

p0 sin t1
cos t1

k t1

p0 cos t1
t1

k t1

From (5.5) & (5.6), the amplitude

v ( t )
p
+ 1 = 0
k

2

= v ( t1 )

v
D = max =
p0 k

(1 cos t1 ) + ( t1 sin t1 )
2

t1

(1 cos t1 ) + ( t1 sin t1 )
2

(5.15)

(5.16)

t1

The condition to occur the maximum response in Phase II:

t1
< 0.371
14
T

Chapter 5- Impulsive loading

5.5 Shock or Response Spectra

Dynamic magnification factor, D

Maximum response produced by each type of impulsive loading

depends only on the ratio of impulse duration to natural period
period of
structure, i.e on t1 T
2.5

Rectangular

Half sine wave

2

1.5

Triangular

Useful to plot to magnification factor D

(maximum value of response ratio Rmax )
as a function of t1 T for various forms
of impulsive loading.

0.5

Such plots are called

displacementdisplacement-response spectra
or simply response spectra

0
0

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

t1 T = Impulse duration / Period

Note:

D <

Structure is stiffer, the dynamic effect is greater

It
Its used to predict the maximum effect from a given type of loads.
It
Its used to indicate response of structure to a base acceleration pulse.
15

Chapter 5- Impulsive loading

Response spectra also serve to indicate the response of structure to

an acceleration impulse applied to its base
Effective impulsive loading due to applied
peff ( t ) = mvg ( t )
(5.17)
based acceleration is given by

(5.18)
p0,eff ( t ) = mvg 0
Maximum effective impulsive load is
applied
base
acceleration

vg (t ) =
vg 0 = maximum base acceleration & measured by accelerometers

Dynamic magnification factor D = vmax = vmax = kvmax

for this case is
p0,eff / k
mvg 0 / k
mvg 0
in which only the absolute magnitude of response generally is of interest
Since in undamped system,

D=

t
f s = f I kv max = mvmax

t
kvmax
mvmax
vt
=
= max
mvg 0
mvg 0
vg 0

(5.19)

t
(vmax
: maximum total acceleration of the mass)

Accordingly, response spectrum plots may therefore be used to predict

predict
maximum acceleration response of mass m to an impulsive base
16
acceleration. So the plots are referred to as shock spectra

Chapter 5- Impulsive loading

Example 5.1

A single-story building to triangular blast

load. By using the above response spectrum,
p(t )
determine the maximum response.
Solution: The natural period of structure

Total weight W = 2670 kN

Total lateral stiffness:

k = 17500 kN/cm

Blast load p ( t )

4450 kN
t

t1 = 0.05 s

W
2670
= 2
= 0.078 s
kg
17500 981
t1 0.05
Ratio of impulse duration to
=
= 0.64
natural period becomes:
T 0.078
T=

= 2

From the above response spectrum for triangular

impulsive load, dynamic magnification factor D = 1.33

Maximum displacement will be:

p
1450
v max = D v st = D 0 = 1.33
= 0.038 cm
17500
k

The maximum total elastic force developed in the structure is:

f S ,max = k vmax = 17500 0.332 = 5810 kN

t1 = 0.005 s t1 T = 0.064 D = 0.2

If
For impulse of very short duration, a large part of the load is
17
resisted by the inertia of structure.

Chapter 5- Impulsive loading

5.6 Approximate analysis of impulsiveimpulsive-load response

From study response spectra, 2 general conclusions:
1- For longlong-duration loads,
loads, t1 T > 1 , D depends on the rate of increase
of the load to its maximum value.
For step loads: D = 2 ; for a very gradual increase: D = 1
2- For shortshort-duration,
duration, t1 T < 0.25, vmax depends on the magnitude of impulse
t1
& is not strongly influenced by the form of the load.
I = p ( t ) dt
However, D is quite dependent on the form of impulse.
0
p (t )
From the equilibrium,
t

I = p ( t ) dt
1

f I + f S = p ( t ) mv = p ( t ) kv ( t )

t1

t1
dv
= p ( t ) kv ( t ) m v = p ( t ) kv ( t ) dt
0
dt
2
For small t1 , v ( t1 ) is of the order ( t1 ) v ( t1 ) is neglected.

Thus, approximately

m v p ( t ) dt
t1

1 t1
v = p ( t ) dt = v ( t1 ) v ( 0 )
m 0

18

Chapter 5- Impulsive loading

1 t1
p ( t ) dt = v ( t1 ) v ( 0 )
m 0
1 t1
Since v ( 0 ) = 0,
v ( t1 ) =
p ( t ) dt
v =

(5.20)

The response after termination of loading is the free vibration

v(t ) =

v ( t1 )

v(t ) =

sin t + v ( t1 ) cos t

1
m

t1

(5.21)

p ( t ) dt sin t

Example E5.2
Weight W=8900 kN
Lateral Stiffness
k=89.5 kN/cm

Using the approximate formula, consider

response of the SDF one-store building to
impulsive load as shown.
Its physical properties:
total weight W = 222 kN
p0 = 222.5 kN total lateral stiffness k = 89.5 kN cm

p (t )

p0

t , sec
0.1

0.1

19

t1

0.1

Chapter 5- Impulsive loading

Solution:
The natural frequency & period of the system

Weight W=8900
kN
Lateral Stiffness
k=89.5
kN/cm

p (t )

kg
89.5 981
=
= 3.14
W
8900

&

T=

2
= 2s
3.14

Equivalent instantaneous pulse:

p0

I = p ( t ) dt = 44.5 kN
t1

p0 = 222.5kN
t , sec
0.1

0.1

0.1

t1

v(t ) =

Since t1 T = 0.3 2 = 0.15 < 0.25, shortshort-duration,

approximate response is (5.18)

g
981
I sin t =
44.5sin t = 1.56sin t
W
8900 3.14

Maximum response results when

sin t = 1, that is v max = 1.56 cm

Maximum lateral elastic force in the building is

f S ,max = k vmax = 89.5 1.56 = 139.6 kN

The exact v max determined by direct integration of the equation of
motion: 1.53 cm
The error in the approximate result is less than 2%.
20
Thus, the approximate analysis in this case is quite accurate.

10