You are on page 1of 16

27

Convolution
Convolution is an operation involving two functions that turns out to be rather useful in many
applications. We have two reasons for introducing it here. First of all, convolution will give
us a way to deal with inverse transforms of fairly arbitrary products of functions. Secondly, it
will be a major element in some relatively simple formulas for solving a number of differential
equations.
Let us start with just seeing what convolution is. After that, well discuss using it with the
Laplace transform and in solving differential equations.
27.1 Convolution, the Basics
Denition and Notation
Let f (t ) and g(t ) be two functions. The convolution of f and g , denoted by f g , is the
function on t 0 given by
f g(t ) =
_
t
x=0
f (x)g(t x) dx .
!

Example 27.1: Let


f (t ) = e
3t
and g(t ) = e
7t
.
Since we will use f (x) and g(t x) in computing the convolution, let us note that
f (x) = e
3x
and g(t x) = e
7(t x)
.
So,
f g(t ) =
_
t
x=0
f (x)g(t x) dx
=
_
t
x=0
e
3x
e
7(t x)
dx
=
_
t
x=0
e
3x
e
7t
e
7x
dx
539
540 Convolution and Laplace Transforms
= e
7t
_
t
x=0
e
4x
dx
= e
7t

1
4
e
4x

t
x=0
=
1
4
e
7t
e
4t

1
4
e
7t
e
40
=
1
4
e
3t
+
1
4
e
7t
.
Simplifying this slightly, we have
f g(t ) =
1
4
_
e
7t
e
3t
_
when f (t ) = e
3t
and g(t ) = e
7t
.
It is common practice to also denote the convolution f g(t ) by f (t ) g(t ) where, here,
f (t ) and g(t ) denote the formulas for f and g . Thus, instead of writing
f g(t ) =
1
4
_
e
7t
e
3t
_
when f (t ) = e
3t
and g(t ) = e
7t
,
we may just write
e
3t
e
7t
=
1
4
_
e
7t
e
3t
_
.
This simplies notation a little, but be careful t is being used for two different things in this
equation: On the left side, t is used to describe f and g ; on the right side, t is the variable in
the formula for the convolution. By convention, if we assign t a value, say, t = 2 , then we are
setting t = 2 in the nal formula for the convolution. That is,
e
3t
e
7t
with t = 2
means compute the convolution and replace the t in the resulting formula with 2 , which, by the
above computations, is
1
4
_
e
72
e
32
_
=
1
4
_
e
14
e
6
_
.
It does NOT mean to compute
e
32
e
72
,
which would give you a completely different result, namely,
e
6
e
14
=
_
t
x=0
e
6
e
14
dt = e
20
t .
!

Example 27.2: Let us nd


1

t
t
2
when t = 4 .
Here,
f (t ) =
1

t
and g(t ) = t
2
.
So
f (x) =
1

x
and g(t x) = (t x)
2
,
and
1

t
t
2
= f g(t ) =
_
t
x=0
1

x
(t x)
2
=
_
t
x=0
x

1
/
2
_
t
2
2t x + x
2
_
dx
Convolution, the Basics 541
=
_
t
x=0
_
t
2
x

1
/
2
2t x
1
/
2
+ x
3
/
2
_
dx
= t
2
2x
1
/
2
2t
2
3
x
3
/
2
+
2
5
x
5
/
2

t
x=0
= 2t
2
t
1
/
2

4
3
t t
3
/
2
+
2
5
t
5
/
2
.
After a little algebra and arithmetic, this reduces to
1

t
t
2
=
16
15
t
5
/
2
. (27.1)
Thus, to compute
1

t
t
2
when t = 4 ,
we actually compute
16
15
t
5
/
2
with t = 4 ,
obtaining
16
15
4
5
/
2
=
16
15
2
5
=
512
15
.
Basic Identities
Let us quickly note a few easily veried identities that can simplify the computation of some
convolutions.
The rst identity is trivial to derive. Let be a constant, and let f and g be two functions.
Then, of course,
_
t
x=0
[f (x)]g(t x) dx =
_
t
x=0
f (x)[g(t x)] dx =
_
t
x=0
f (x)g(t x) dx ,
which we can rewrite as
[f ] g = f [g] = [ f g] .
In other words, we can factor out constants.
A more substantial identity comes from looking at how switching the roles of f and g
changes the convolution. That is, how does the result of computing
g f (t ) =
_
t
x=0
g(x) f (t x) dx
compare to what we get by computing
f g(t ) =
_
t
x=0
f (x)g(t x) dx ?
Well, in the last integral, lets use the substitution y = t x . Then x = t y , dx = dy and
f g(t ) =
_
t
x=0
f (x)g(t x) dx
=
_
t t
y=t 0
f (t y)g(y)(1) dy
=
_
0
y=t
g(y) f (t y) dy =
_
t
y=0
g(y) f (t y) dy .
542 Convolution and Laplace Transforms
The last integral is exactly the same as the integral for computing g f (t ) , except for the cosmetic
change of denoting the variable of integration by y instead of x . So that integral is the formula
for formula for g f (t ) , and our computations just above reduce to
f g(t ) = g f (t ) . (27.2)
Thus we see that convolution is commutative.
!

Example 27.3: Lets consider the convolution


t
2

t
.
Since we just showed that convolution is commutative, we know that
t
2

t
=
1

t
t
2
.
What an incredible stroke of luck! Weve already computed the convolution on the right in
example 27.2. Checking back to equation (27.1), we nd
1

t
t
2
=
16
15
t
5
/
2
.
Hence,
t
2

t
=
1

t
t
2
=
16
15
t
5
/
2
.
In addition to being commutative, convolution is distributive and associative. That is,
given three functions f , g and h ,
[ f + g] h = [ f h] + [g h] , (27.3)
f [g +h] = [ f g] + [ f h] (27.4)
and
f [g h] = [ f g] h . (27.5)
The rst and second equations are that addition distributes over convolution. They are easily
conrmed using the basic denition of convolution. For the rst:
[ f + g] h(t ) =
_
t
x=0
[ f (x) + g(x)]h(t x) dx
=
_
t
x=0
[ f (x)h(t x) + g(x)h(t x)] dx
=
_
t
x=0
f (x)h(t x) dx +
_
t
x=0
g(x)h(t x)] dx
= [ f g] + [g h] .
The second, equation (27.4) follows in a similar manner or by combining (27.3) with the commu-
tativity of the convolution. The last equation in the list, equation (27.5), states that convolution
Convolution and Products of Transforms 543
is associative, that is, when convolving three functions together, it does not matter which two
you convolve rst. Its verication requires showing that the two double integrals dening
f [g h] and [ f g] h
are equivalent. This is a relatively straightforward exercise in substitution, and will be left as a
challenge for the interested student (exercise 27.3 on page 552).
Finally, just for fun, lets make a few more simple observations:
0 g(t ) = g 0(t ) =
_
t
x=0
0 g(t x) dx = 0 .
f 1(t ) = 1 f (t ) =
_
t
x=0
f (s) 1 dx =
_
t
x=0
f (s) dx .
f g(0) =
_
0
x=0
f (x)g(0 x) dx = 0 .
Observations on the Existence of the Convolution
The observant reader will have noted that, if f and g are at least piecewise continuous on
(0, ) , then, for any positive value t , the product f (x)g(t x) is a piecewise continuous
function of x on (0, t ) . It then follows that the integral in
f g(t ) =
_
t
x=0
f (x)g(t x) dx
is well dened and nite for every positive value of t . In other words, f g is a well-dened
function on (0, ) , at least whenever f and g are both piecewise continuous on (0, ) . (In
fact, it can then even be shown that f g(t ) is a continuous function on [0, ) .)
But now observe that one of the functions in example 27.2, namely t
1/2
, blows up at
t = 0 and, thus, is not piecewise continuous on (0, ) . So that example also demonstrates that,
sometimes, f g is well dened on (0, ) even though f or g is not piecewise continuous.
27.2 Convolution and Products of Transforms
To see one reason convolution is important in the study of Laplace transforms, let us examine the
Laplace transformof the convolution of two functions f (t ) and g(t ) . Our goal is a surprisingly
simple formula of the corresponding transforms,
F(s) = L[ f (t )]|
s
=
_

0
f (t )e
st
dt
and
G(s) = L[g(t )]|
s
=
_

0
g(t )e
st
dt .
544 Convolution and Laplace Transforms
(The impatient can turn to theorem 27.1 on page 545 for that formula.)
Keep in mind that we can rename the variable of integration in each of the above integrals.
In particular, note (for future reference) that
F(s) =
_

0
e
sx
f (x) dx and G(s) =
_

0
e
sy
g(y) dy .
Now, simply writing out the integral formulas for the Laplace transform and for the convo-
lution yields
L[ f g(t )]|
s
=
_

t =0
e
st
f g(t ) dt
=
_

t =0
e
st
_
t
x=0
f (x)g(t x) dx dt
=
_

t =0
_
t
x=0
e
st
f (x)g(t x) dx dt .
Combined with the observation that
e
st
= e
st +sxsx
= e
s(t x)
e
sx
,
the above sequence becomes
L[ f g(t )]|
s
=
_

t =0
_
t
x=0
e
sx
f (x) e
s(t x)
g(t x) dx dt
=
_

t =0
_
t
x=0
K(x, t ) dx dt
(27.6)
where, simply to simplify expressions in the next few lines, weve set
K(x, t ) = e
sx
f (x) e
s(t x)
g(t x) .
It is now convenient to switch the order of integration in the last double integral. According
to the limits in that double integral, we are integrating over the region R in the XTplane
consisting of all (x, t ) for which
0 < t <
and, for each of these values of t ,
0 < x < t .
As illustrated in gure 27.1, region R is the portion of the rst quadrant of the XTplane to
the left of the line t = x . Equivalently, as can also be seen in this gure, R is the portion of
the rst quadrant above the line t = x . So R can also be described as the set of all (x, t ) for
which
0 < x <
and, for each of these values of x ,
x < t < .
Thus,
_

t =0
_
t
x=0
K(x, t ) dx dt =
_ _
R
K(x, t ) d A =
_

x=0
_

t =x
K(x, t ) dt dx .
Convolution and Products of Transforms 545
T
X
t = t
0
x = x
0
t = x
0
x = t
0
(x
0
, t
0
)
(x
0
, x
0
)
(t
0
, t
0
)
t
=
x
R
Figure 27.1: The region R for the transform of the convolution. Note that the coordinates
of any point (x
0
, t
0
) in R must satisfy 0 < x
0
< t
0
< .
Combining this with equation (27.6), and then continuing, we have
L[ f g(t )]|
s
=
_

x=0
_

t =x
K(x, t ) dt dx
=
_

x=0
_

t =x
e
sx
f (x) e
s(t x)
g(t x) dt dx
=
_

x=0
e
sx
f (x)
__

t =x
e
s(t x)
g(t x) dt
_
dx .
Let us simplify the inner integral with the substitution y = t x (remembering that t is the
variable of integration in this integral):
_

t =x
e
s(t x)
g(t x) dt =
_
x
y=xx
e
sy
g(y) dy =
_

y=0
e
sy
g(y) dy = G(s) !
Combining this with our last formula for L[ f g] then yields
L[ f g(t )]|
s
=
_

x=0
f (x)e
sx
G(s) dx
=
_

x=0
e
sx
f (x) dx G(s) = F(s) G(s) !
Thus,
L[ f g(t )]|
s
= F(s)G(s) .
Equivalently,
f g(t ) = L
1
[F(s)G(s)]|
t
.
If we had been a little more complete in our computations, we would have kept track of
the exponential order of all the functions involved (see exercise 27.9), and obtained all of the
following theorem.
Theorem 27.1 (Laplace convolution identities)
Assume f (t ) and g(t ) are two functions of exponential order s
0
, and with Laplace transforms
F(s) = L[ f (t )]|
s
and G(s) = L[g(t )]|
s
.
546 Convolution and Laplace Transforms
Then the convolution f g(t ) is of exponential order s
1
for any s
1
> s
0
. Moreover,
L[ f g(t )]|
s
= F(s)G(s) for s > s
0
(27.7)
and
L
1
[F(s)G(s)]|
t
= f g(t ) . (27.8)
Do remember that identities (27.7) and (27.8) are equivalent. It is also worthwhile to rewrite
these identities as
L[ f g(t )]|
s
= L[ f ]|
s
L[g(t )]|
s
(27.7

)
and
L
1
[F(s)G(s)]|
t
= L
1
[F(s)]|
t
L
1
[G(s)]|
t
, (27.8

)
respectively. These forms, especially the latter, are sometimes a little more convenient in practice.
!

Example 27.4: Consider nding the inverse Laplace transform of


1
s
2
10s +21
.
Factoring the denominator and applying the above, we get
L
1
_
1
s
2
10s +21
_

t
= L
1
_
1
(s 3)(s 7)
_

t
= L
1
_
1
s 3

1
s 7
_

t
= L
1
_
1
s 3
_

t
L
1
_
1
s 7
_

t
= e
3t
e
7t
.
As luck would have it, this convolution was computed in example 27.1 on page 539),
e
3t
e
7t
=
1
4
_
e
7t
e
3t
_
.
Thus,
L
1
_
1
s
2
10s +21
_

t
= e
3t
e
7t
=
1
4
_
e
7t
e
3t
_
.
The inverse transform in the last example could also have been computed using partial
fractions. Indeed, many of the inverse transforms we computed using partial fractions can also
be computed using convolution. Whether one approach or the other is preferred depends on the
opinion of the person doing the computing. However, as the next example shows, there are cases
where convolution can be applied, but not partial fractions. We will also use this example to
demonstrate how convolution naturally arises when solving differential equations.
!

Example 27.5: Consider solving the initial-value problem


y

+ 9y =
1

t
with y(0) = 0 and y

(0) = 0 .
Convolution and Products of Transforms 547
Taking the Laplace transform of both sides:
L
_
y

+ 9y
_

s
= L
_
1

t
_

s
L
_
y

s
+ 9L[y]|
s
=

s
s
2
Y(s) s y(0)
.,,.
0
y

(0)
.,,.
0
+ 9Y(s) =

_
s
2
+9
_
Y(s) =

s
Y(s) =

s
_
s
2
+9
_ .
Thus, y(t ) is the inverse Laplace transform of

s
_
s
2
+9
_ .
Because the denominator does not factor into two polynomials, we cannot use partial fractions
we must use convolution,
L
1
_

s
_
s
2
+9
_
_

t
= L
1
_

1
s
2
+9
_

t
= L
1
_

s
_

t
L
1
_
1
s
2
+9
_

t
.
Reversing the transform made on the right side of the above equations, we have
L
1
_

s
_

t
=
1

t
.
Using our tables, we nd that
L
1
_
1
s
2
+9
_

t
=
1
3
L
1
_
3
s
2
+3
2
_

t
=
1
3
sin(3t ) .
Combining the above and recalling that constants factor out, we then obtain
y(t ) = L
1
_

s
_
s
2
+9
_
_

t
= L
1
_

s
_

t
L
1
_
1
s
2
+9
_

t
=
_
1

t
_

_
1
3
sin(3t )
_
=
1
3
1

t
sin(3t ) .
That is,
y(t ) =
1
3
_
t
x=0
1

x
sin(3[t x]) dx .
Admittedly, this last integral is not easily evaluated by hand. But it is something that can
be accurately approximated for any specic (nonnegative) value of t using routines found in
many computer math packages. So it is still a usable formula.
548 Convolution and Laplace Transforms
27.3 Convolution and Differential Equations
(Duhamels Principle)
As illustrated in our last example, convolution has a natural role in solving differential equations
when using the Laplace transform. However, if we look a little more carefully at the process
of solving differential equations using the Laplace transform, we will nd that convolution can
play an even more signicant role.
!

Example 27.6: Lets consider solving the nonhomogeneous initial-value problem


y

10y

+ 21y = f (t ) with y(0) = 0 and y

(0) = 0
where f = f (t ) is any Laplace transformable function. Naturally, we will use the Laplace
transform. So let
Y(s) = L[y(t )]|
s
and F(s) = L[ f (t )]|
s
.
Because of our initial conditions, the transform of the derivatives identities simplify con-
siderably:
L
_
y

s
= s
2
Y(s) s y(0)
.,,.
0
y

(0)
.,,.
0
= s
2
Y(s)
and
L
_
y

s
= sY(s) y(0)
.,,.
0
= sY(s) .
Consequently,
L
_
y

10y

+21y
_

s
= L[ f (t )]|
s
L
_
y

s
10L
_
y

s
+ 21L[y]|
s
= F(s)
s
2
Y(s) 10sY(s) + 21Y(s) = F(s)

_
s
2
10s +21
_
Y(s) = F(s) .
Dividing through by the polynomial, we get
Y(s) = H(s)F(s) where H(s) =
1
s
2
10s +21
.
Thus,
y(t ) = L
1
[Y(s)]|
t
= L
1
[H(s)F(s)]|
t
.
Applying the convolution identity then yields
y(t ) = h f (t ) (27.9a)
where
h(x) = L
1
[H(s)]|
x
= L
1
_
1
s
2
10s +21
_

t
. (27.9b)
Convolution and Differential Equations (Duhamels Principle) 549
The convolution h f can be computed using either
_
t
0
h(x) f (t x) dx or
_
t
0
h(t x) f (x) dx .
For no particular reason, we will choose the rst integral formula.
To compute h(x) , we can use partial fractions or convolution. Or we can glance at
example 27.4 a few pages ago, discover that weve already computed h(t ) , and just replace
the t in that formula with x ,
h(x) =
1
4
_
e
7x
e
3x
_
.
With this and our chosen integral formula for h f , formula (27.9a), the solution to our
initial-value problem, becomes
y(t ) =
_
t
0
1
4
_
e
7x
e
3x
_
f (t x) dx . (27.10)
Formula (27.10) is a convenient way to describe the solutions to our initial-value problem,
especially if we want to solve this problem for a number of different choices of f (t ) . Using
it, we can quickly write out a relatively simple integral formula for the solution corresponding
to each f (t ) . For example:
If f (t ) = e
4t
, then f (t x) = e
4(t x)
and formula (27.10) yields
y(t ) =
_
t
0
1
4
_
e
7x
e
3x
_
e
4(t x)
dx .
If f (t ) = 1 , then f (t x) = 1 and and formula (27.10) yields
y(t ) =
1
4
_
t
0
_
e
7x
e
3x
_
1 dx .
And nally, if f (t ) =
3

t , then f (t x) =
3

t x and formula (27.10) yields


y(t ) =
1
4
_
t
0
_
e
7x
e
3x
_
3

t x dx .
The rst two integrals are easily evaluated, giving us
y(t ) =
1
2
e
7t
+
1
4
e
3t

1
3
e
4t
and y(t ) =
1
28
e
7t

1
12
e
3t

1
21
,
respectively. The last integral,
y(t ) =
1
4
_
t
0
_
e
7x
e
3x
_
3

t x dx ,
is not easily evaluated by hand, but can be accurately approximated for any value of t using
routines found in our favorite computer math package.
?

Exercise 27.1: Using the formula (27.10), nd the solution to


y

10y

+ 21y = e
3t
with y(0) = 0 and y

(0) = 0 .
550 Convolution and Laplace Transforms
Generalizing what we just derived in the last example is easy. Suppose we have any second-
order initial-value problem of the form
ay

+ by

+ cy = f (t ) with y(0) = 0 and y

(0) = 0
where a , b and c are constants, and f is any Laplace transformable function. Then, taking
the Laplace transform of both sides of the differential equation, letting
Y(s) = L[y(t )]|
s
and F(s) = L[ f (t )]|
s
,
and noting that, because of our initial conditions, the transform of the derivatives identities
simplify to
L
_
y

s
= s
2
Y(s) s y(0)
.,,.
0
y

(0)
.,,.
0
= s
2
Y(s)
and
L
_
y

s
= sY(s) y(0)
.,,.
0
= sY(s) ,
we see that
L
_
ay

+by

+cy
_

s
= L[ f (t )]|
s
aL
_
y

s
+ bL
_
y

s
+ cL[y]|
s
= F(s)
as
2
Y(s) + bsY(s) + cY(s) = F(s)

_
as
2
+bs +c
_
Y(s) = F(s) .
Dividing through by the polynomial, we get
Y(s) = H(s)F(s) where H(s) =
1
as
2
+bs +c
.
So,
y(t ) = L
1
[Y(s)]|
t
= L
1
[H(s)F(s)]|
t
,
and the convolution identity tells us that
y(t ) = h f (t ) (27.11a)
where
h(x) = L
1
[H(s)]|
x
= L
1
_
1
as
2
+bs +c
_

t
. (27.11b)
The fact that the formula for y in equation set (27.11) is the solution to
ay

+ by

+ cy = f (t ) with y(0) = 0 and y

(0) = 0
is often called Duhamels principle. The function H(s) is usually referred to as the transfer
function, and its inverse transform, h(t ) , is usually called the impulse response function.
1
Keep
in mind that a , b and c are constants, and that we assumed f is Laplace transformable.
1
The reason why h is called the impulse response function will be revealed in chapter 29. A few authors also
refer to h as a weight function.
Convolution and Differential Equations (Duhamels Principle) 551
As illustrated in our example, Duhamels principle makes it easy to write down solutions to
the given initial-value problem once we have found h . This is especially useful if we need to
nd solutions to
ay

+ by

+ cy = f (t ) with y(0) = 0 and y

(0) = 0
for a number of different choices of f .
Bur whystopat second-order problems? It shouldbe clear that the above differential equation
did not have to be second order. A completely analogous derivation can be done starting with
any nonhomogeneous linear differential equation with constant coefcients, provided all the
appropriate initial values are zero. Doing so, leads to the following theorem:
Theorem 27.2 (Duhamels principle)
Let N be any positive integer, let a
0
, a
1
, . . . and a
N
be any collection of constants, and let
f (t ) be any Laplace transformable function. Then, the solution to
a
0
y
(N)
+ a
1
y
(N1)
+ + a
N2
y

+ a
N1
y

+ a
N
y = f (t )
satisfying the N
th
-order zero initial conditions,
y(0) = 0 , y

(0) = 0 , y

(0) = 0 , . . . and y
N1
(0) = 0 ,
is given by
y(t ) = h f (t )
where
h(x) = L
1
[H(s)]|
x
and H(s) =
1
a
0
s
n
+a
1
s
n1
+ +a
n
.
Three quick notes:
1. As noted a few pages ago, the convolution h f can be computed using either
_
t
0
h(x) f (t x) dx or
_
t
0
h(t x) f (x) dx .
In practice, use whichever appears easier to compute given the h and f involved. In the
examples here, we used the rst. Later, when we re-examine resonance in mass/spring
systems (section 28.7), we will use the other integral formula.
2. It turns out that the f (t ) in Duhamels principle (as described above) does not have to
be Laplace transformable. By applying the above theorem with
f
T
(t ) =
_
f (t ) if t < T
0 if t T
,
and then letting T , you can show that y = h f is well dened and satises the
initial-value problem even when f is merely piecewise continuous on (0, ) .
3. It is not hard to show that, if you want the solution to
a
0
y
(N)
+ a
1
y
(N1)
+ + a
N2
y

+ a
N1
y

+ a
N
y = f (t ) ,
552 Convolution and Laplace Transforms
but satisfying nonzero initial conditions, then you simply need to add the solution obtained
by Duhamels principle to the solution to the corresponding homogeneous differential
equation
a
0
y
(N)
+ a
1
y
(N1)
+ + a
N2
y

+ a
N1
y

+ a
N
y = 0
that satises the desired initial conditions.
Additional Exercises
27.2. Compute the convolution f g(t ) of each of the following pairs of functions:
a. f (t ) = e
3t
and g(t ) = e
5t
b. f (t ) =
1

t
and g(t ) = t
2
c. f (t ) =

t and g(t ) = 6 d. f (t ) = t and g(t ) = e


3t
e. f (t ) = t
2
and g(t ) = t
2
f. f (t ) = sin(t ) and g(t ) = t
g. f (t ) = sin(t ) and g(t ) = sin(t ) h. f (t ) = step
3
(t ) and g(t ) = e
2t
i. f (t ) = step
3
(t ) and g(t ) = t
2
27.3. Verify the associative property of convolution. That is, verify equation (27.5) on page
542.
27.4. Usingconvolution, compute the inverse Laplace transformof eachof the following:
a.
1
(s 4)(s 3)
b.
1
s(s 3)
c.
1
s(s
2
+4)
d.
1
(s 3)(s
2
+1)
e.
1
(s
2
+9)
2
f.
s
2
(s
2
+4)
2
g.
e
4s
s(s
2
+1)
h.
1

s(s 3)
(leave in integral form)
27.5. For each of the following initial-value problems, nd the corresponding transfer function
H and the impulse response function h , and write down the corresponding convolution
integral formula for the solution:
a. y

+ 4y = f (t ) with y(0) = 0 and y

(0) = 0
b. y

4y = f (t ) with y(0) = 0 and y

(0) = 0
c. y

6y

+ 9y = f (t ) with y(0) = 0 and y

(0) = 0
d. y

6y

+ 18y = f (t ) with y(0) = 0 and y

(0) = 0
e. y

+ 16y

= f (t ) with y(0) = 0 and y

(0) = 0
Additional Exercises 553
27.6. Using the results from exercise 27.5 a, nd the solution to
y

+ 4y = f (t ) with y(0) = 0 and y

(0) = 0
for each of the following choices of f :
a. f (t ) = 1 b. f (t ) = t c. f (t ) = e
3t
d. f (t ) = sin(2t ) e. f (t ) = sin(t ) where = 3
27.7. Using the results from exercise 27.5 c, nd the solution to
y

6y

+ 9y = f (t ) with y(0) = 0 and y

(0) = 0
for each of the following choices of f :
a. f (t ) = 1 b. f (t ) = t c. f (t ) = e
3t
d. f (t ) = e
3t
e. f (t ) = e
t
where = 3
27.8. Using the results from exercise 27.5 e, nd the solution to
y

+ 16y

= f (t ) with y(0) = 0 and y

(0) = 0
for each of the following choices of f :
a. f (t ) = 1 b. f (t ) = t c. f (t ) = e
3t
d. f (t ) = sin(4t ) e. f (t ) = sin(at ) where = 4
27.9. Let f and g be two piecewise continuous functions on the positive real line satisfying,
for all t > 0 ,
| f (t )| < M
f
e
s
0
t
and |g(t )| < M
g
e
s
0
t
for some constants M
f
, M
g
and s
0
.
a. Show that | f g(t )| < M
f
M
g
e
s
0
t
t whenever t > 0 .
b. Why does this tell us that f g is of exponential order s
1
for any s
1
> s
0
?

You might also like