Scribd
Upload
239 views7 pages

November2007 Duhamel Integral

The document derives and evaluates Duhamel's integral, which represents the displacement response of a damped single degree of freedom system subjected to arbitrary loading. It first derives Duhamel's integral by taking the Laplace transform of the equation of motion. It then describes numerically evaluating the integral when the loading is given at discrete time points, by approximating the loading as linear between points and analytically evaluating the integral term.

Uploaded by

Alak Majumder
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
239 views7 pages

November2007 Duhamel Integral

The document derives and evaluates Duhamel's integral, which represents the displacement response of a damped single degree of freedom system subjected to arbitrary loading. It first derives Duhamel's integral by taking the Laplace transform of the equation of motion. It then describes numerically evaluating the integral when the loading is given at discrete time points, by approximating the loading as linear between points and analytically evaluating the integral term.

Uploaded by

Alak Majumder
Copyright
© Attribution Non-Commercial (BY-NC)
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOC, PDF, TXT or read online on Scribd
You are on page 1/ 7

Derivation and numerical evaluation of Duhamel's integral.

1. Derivation of Duhamel's integral:


Let's start by considering the equation of motion of a damped single degree of freedom system
subjected to arbitrary loading.
fig.1 Schematic representation
fig.2 Forces acting on mass M along the X direction
Mass equilibrium:
f t () F
el
F
in
F
c
0
1 ( )

m
2
t
x t ( )
d
d
2
c
t
x t ( )
d
d
+ k x t ( ) + f t ( )
2 ( )
where:
f t ( )

is a general real-valued function of time t
The solution to equation (2) can be obtained by taking the Laplace transform of equation (2)
L m
2
t
x t ( )
d
d
2
c
t
x t ( )
d
d
+ k x t ( ) +

,
L f t ( ) ( )
3 ( )
m L
2
t
x t ( )
d
d
2

,
c L
t
x t ( )
d
d

,
+ k L x t ( ) ( ) + L f t ( ) ( )
4 ( )
m s
2
X s ( ) s x 0 ( )
t
x 0 ( )
d
d

,
c s X s ( ) x 0 ( ) ( ) + k X s ( ) +F s ( )
5 ( )
in equation (5)
X s ( )

is the Laplace transform of
x t ( )

and
Fs ( )

is the Laplace transform of
f t ( )
By rearranging the terms in equation (5) and factoring
X s ( )
we obtain the following equation:
m s
2
c s +k +
( )
X s ( ) F s ( ) m
t
x 0 ( )
d
d
+m s x 0 ( ) +c x 0 ( ) +
6 ( )
or
X s ( )
F s ( )
m s
2
c s +k +
m
t
x 0 ( )
d
d

m s
2
c s +k +
+
m s c + ( ) x 0 ( )
m s
2
c s +k +
+
7 ( )
Rewrite
c
as
c 2 m
where

is a constant, and
k
2
m


ms
2
c s +k +ms
2
2 s +
2
+
( )
8 ( )

in equation (8) represents the natural frequency of the undamped SDF system:
If we solve equation
s
2
2 s +
2
+0
for
s
, we obtain the following roots:
S
1
i
2

2
+ i

+
9 ( )
S
2
i
2

2
i


in equations (9),


represents the natural frequency of the damped SDF system.
Now equation (8) can be rewritten as:
ms
2
2 s +
2
+
( )
ms S
1

()
s S
2

()
ms i

+
( )
s +i

+
( )

or
m s
2
2 s +k +
( )
m s + ()
2
i

()
2

1
]
m s + ()
2

()
2
+

1
]

10 ( )
Hence equation (7) becomes:
X s ( )
F s ( )

m s + ( )
2

( )
2
+

1
]

t
x 0 ( )
d
d

s + ( )
2

( )
2
+

1
]

+
s 2 + ( )x 0 ( )
s + ( )
2

( )
2
+
+
11 ( )

X s ( )
F s ( )

m
L e
t
sin

t
( )

( )

t
x 0 ( )
d
d

L e
t
sin

t
( )

( )
+
s + ( ) x 0 ( ) x 0 ( ) +
s + ( )
2

( )
2
+
+
...
12 ( )

X s ( )
F s ( )

m
L e
t
sin

t
( )

( )

t
x 0 ( )
d
d

L e
t
sin

t
( )

( )
+
x 0 ( ) L e
t
cos

t
( )

( )

x 0 ( )

L e
t
sin

t
( )

( )
+ +
...
13 ( )

X s ( )
1

m
L
0
t
f
( )
e
t
( )

sin

t ( )

1
]

1
1
1
]

t
x 0 ( )
d
d

Le
t
sin

t
( )

( )
x 0 ( ) Le
t
cos

t
( )

( )
+ +
...
x 0 ( )

Le
t
sin

t
( )

( )
+
...
14 ( )

X s ( ) L
1

m
0
t
f
()
e
t
( )

sin

t ( )

1
]

d
t
x 0 ( )
d
d

e
t
sin

t
( )

( )
x 0 ( ) e
t
cos

t
( )

sin

t
( )
+

,
+ +
...

1
1
1
1
1
1
1
]
15 ( )
Now the general solution to equation (2) is finally:
x t ( )
1

m
0
t
f
()
e
t
( )

sin

t ( )

1
]

d
t
x 0 ( )
d
d

e
t
sin

t
( )

( )
+
x 0 ( ) e
t
cos

t
( )

sin

t
( )
+

,
+
...
16 ( )
D t ( )
1

m
0
t
f
()
e
t
( )

sin

t ( )

1
]

d
is the Convolution integral
(Duhamel's integral) and represents the displacement due to the forcing function only

2. Numerical evaluation of the Convolution integral
A

the time span of the forcing function is [0 ; T]
B

the forcing function is given at discrete time points,
t
0
0t
1
,t
2
,............ , t
i
,.................. , t
n
T ,
f t
0
()
f
0
f t
1
()
f
1
, f t
2
()
f
2
, .............. ,f t
n
()
f
n
,
we define a linear approximation of
f t ( )
at time interval [ti ; ti+1] :
for t

[ti ; ti+1]
f
l
t ( ) f t
i
()
f t
i 1 +
( )
f t
i
()

t
i 1 +
t
i

t t
i

( )
+
17 ( )
or
f
l
t ( ) f
i
f
i 1 +
f
i

t
i 1 +
t
i

t t
i

( )
+
18 ( )
let
t t
i


f
l
t ( )


f
l

( )
f
i
f
i 1 +
f
i

t
i
+

[
0
;
t
i
]
assume we are interested in the solution within time interval [ti ; ti+1]. The solution is:
for
t

[
0
;
t
i
]
x t ( )
1

m
0
t
f
l

()
e
t
( )

sin

t ( )

1
]

d
t
X t
i
()
d
d

e
t
sin

t
( )

( )
+
X t
i
()
e
t
cos

t
( )

sin

t
( )
+

,
+
...
19 ( )
t
X t
i
( )
d
d
and
Xt
i
( )
in equation (19) are the velocity and displacement respectively (the initial
conditions), at time point
t
i
Since
f
l
t ( )
is linear in
t
within time interval [
0
;
t
i
], the first term on the right hand side of
equation (19) can be evaluated analytically. Such analytic evaluation yields the following
expression for the first term: (it is a particularly long expression, therefore analytic evaluation
details have been omitted.)
1

m
0
t
f
l

()
e
t
( )

sin

t ( )

1
]

d . .

( )
2

f
l
t ( ) 2

t
f
l
t ( )
d
d

( )
2

1
]
2

( )
2

f
i
e
t
cos

t
( )
+
t
f
l
0 ( )
d
d
f
l
0 ( ) +

1
1
]

( )
2

1
]
e
t
sin

t
( )
2

t
f
l
0 ( ) e
t

cos

t
( )

( )
2

1
]
2

d
d
+ +
...
2
2

t
f
l
0 ( ) e
t
sin

t
( )

d
d

( )
2

1
]
2
+
...
20 ( )
with
t
f
l
0 ( )
d
d
f
i 1 +
f
i

t
i
t
f
l
t ( )
d
d

f
l
t ( ) f
i
f
i 1 +
f
i

t
i
t +

and

f
l
0 ( ) f
i
Equation (20) represents the solution due to the forcing function only, within time interval
[
0
;
t
i
]
By rearranging the terms in equation (20) and adding the initial condition terms to equation (20)
the solution
x t ( )
within time interval [
0
;
t
i
] or [
t
i
;
t
i 1 +
] can be rewritten as follows:
x t ( )
1

( )
2

f
i

e
t
cos

t
( )
e
t
sin

t
( )

( )

f
i 1 +
f
i

t
i

t 2

( )
2

e
t
sin

t
( )

2

( )
2

e
t
cos

t
( )
+
...
2
2
e
t
sin

t
( )

( )
2

+
...

1
1
1
1
1
1
1
1
1
1
]
+
...

1
1
1
1
1
1
1
1
1
1
1
1
]

t
X t
i
( )
d
d

e
t
sin

t
( )

( )
X t
i
( )
e
t
cos

t
( )

sin

t
( )
+

,
+ +
...
21 ( )
By taking the derivative of
x t ( )
with respect to time we obtain the following expression for the
velocity
t
x t ( )
d
d
:
t
x t ( )
d
d
1

( )
2

f
i

( )
2
e
t
sin

t
( )

e
t
cos

t
( )
+

e
t
cos

t
( )

2
e
t
sin

t
( )
+ +
...

1
1
1
]

f
i 1 +
f
i

t
i

e
t
cos

t
( )
e
t
sin

t
( )
+
2

( )
2

e
t
sin

t
( )

e
t
cos

t
( )
+
...

,
+
...
2
2

( )
2

e
t
cos

t
( )

e
t
sin

t
( )
+
...

,
+
...

1
1
1
1
1
1
1
1
]
+
...

1
1
1
1
1
1
1
1
1
1
1
]

t
X t
i
( )
d
d

e
t
cos

t
( )
e
t
sin

t
( )

( )
+
...
X t
i
( )
e
t

sin

t
( )
cos

t
( )
+
( )
e
t
cos

t
( )

sin

t
( )
+
...

1
1
1
1
]
+
...
22 ( )
Expressions (21) and (22) can be used to compute the displacement and velocity at time point
t
i 1 +
by replacing
t
with
t
i
. The displacement and velocity at time point
t
i 1 +
are then used as
initial conditions for the solution in the next time interval. This process is repeated for the next
time interval and so on.

You might also like