The Heaviside Function or Step Function.

We define the Heaviside function or step function as:

0, if t < a
Ha (t) =
1, if t > a
(See figure 1)
Note: Ha (t) has a jump at t = a and therefore is piecewise continuous!
(because it takes the constant values 0 or 1 and constants are always continuous)
Furthermore we do not really care of defining it at t = a.
Let us compute its Laplace transform:
L(Ha (t)) =

R
0

est Ha (t)dt,

you know this definition by now, dont you? ;o)

est dt, by definition of Ha (t)

h st it=
sa
= es
= es
t=a
=

Let us do it a numbered equation:

L(Ha (t)) =

esa
s

(1)

Remark: Suppose f (t) = 0 for all t < 0. Then f (t a)Ha (t) with a > 0 is
f (t) shifted (translated) to the right by amount a. Please study the differences
between the figures 2, 3 and 4.

-2

10

-1

-2

-3

Figure 1: Heaviside funtion with a = 4.

0.5

-10

-8

-6

-4

-2

-0.5

-1

Figure 2: f (t) = sin t, t > 0.

10

0.5

-10

-8

-6

-4

-2

10

10

-0.5

-1

Figure 3: g(t) = H2 (t) sin t.

0.5

-10

-8

-6

-4

-2

-0.5

-1

Figure 4: h(t) = H2 (t) sin(t 2).

Example 1 How does it look like H1 (t) 2H4 (t) + 3H6 (t)?
By definition we have:

0,
H1 (t) =
1,

if t < 1
if t > 1,

0, if t < 4
2H4 (t) =
2, if t > 4
and finally:

0,
3H6 (t) =
3,

if t < 6
if t > 6

In the interval (0, 1), H1 (t), H4 (t), H6 (t) vanish so H1 (t), 2H4 (t), +3H6 (t) =
0, t (0, 1) . In the interval (1, 4) the only function that does not vanish is H1 (t),
so H1 (t) 2H4 (t) + 3H6 (t) = 1 if t (1, 4). In the interval (4, 6), H1 (t) = 1 and
2H4 (t) = 2 and H6 = 0, therefore H1 (t) 2H4 (t) + 3H6 (t) = 1, t (4, 6)
and in the interval (6, ) H1 (t) = 1 and 2H4 (t) = 2 and 3H6 = 3 so
H1 (t) 2H4 (t) + 3H6 (t) = 2, t (6, ). Hence we have the following piecewise
function:

0, if t < 1

1, if 1 < t < 4
H1 (t) 2H4 (t) + 3H6 (t) =

1, if 4 < t < 6

2 otherwise

Note: Define

1,
f (t) =
0,

if 0 < t < C, C constant

otherwise

Then:
f (t) = 1 HC (t).
(Check it!. I am sure it will be useful for some exercises...)
4

-2

10

-2

-4

Figure 5: Sum of Heaviside functions of Example 1

Theorem 1 Second shifting theorem. Time shifting.
If f (t) has the transform F (s), then the shifted function

0, if t < a

f (t) = f (t a)Ha (t) =

f (t a), if t > a
has transform eas F (s). That is:
L(f(t)) = eas F (s),
and
L1 (eas F (s)) = f(t).
Proof: if you feel like checking the proof go to Erwin Kreyszigs book! But it is
true, believe me...

Example 2 Suppose you are given the Laplace transform
5

e2s .
s + 2
2

Calculate the inverse Laplace transform.

e2s indicates that there is a time delay of 2 units. We identify the inverse
Laplace transform of
1
s + 2
2

via the tables and we conclude:

1
sin t

Therefore:
L1 (

e2s
1
sin (t 2)H2 (t),
2
2) =

s +

using the Second Shifting theorem.

Example 3 Remainder of the use of the First Shifting Theorem:
Let us use Laplace transfom method to solve the following ODE:
y 00 + 2y 0 + 2y = 0,

y(0) = 0, y 0 (0) = 1

Taking Laplace Transforms and by the theorem of laplace transforms of derivatives we get:
s2 Y (s) sy(0) y 0 (0) + 2(sY (s) y(0)) + 2Y (s) = 0
Therefore:
Y (s) =

1
1
=
= F (s (1)),
s + 2s + 2
(s + 1)2 + 1
2

1 , and we know that the inverse Laplace transform of F (s) is

s2 + 1
sin t. Hence by the first shifting theorem we obtain the solution of our ODE:
where F (s) =

y(t) = L1 (F (s + 1)) = et sin t.

Note: The First Shifting theorem is about s shifting. The second Shifting
theorem is about t shifting.
6

Example 4 Application of both Shifting Theorems.

Compute the inverse transform of

e2s + e3s .
s + 2 (s + 1)2
2

We have:
2s
e3s ) = L1 (e2s 1 ) + L1 (e3s
1
L1 ( 2e
),
2 +
s +
(s + 1)2
s2 + 2
(s + 1)2

by linearity of the inverse transform. Now, first observe that e2s , e3s indicate
that we have time delays of 2 and 3 units resp. We need to identify then inverse
laplace transform of
1
s + 2
2

and
1
.
(s + 1)2
Looking in the table we see that
L1 (

1
1
sin t.
2) =

s +
2

Using the Second Shifting Theorem we deduce then:

L1 (e2s

1
1
sin (t 2)H2 (t)
2) =

s +
2

Moreover we observe that

1
= F (s + 1),
(s (1))2
where
F (s) =

1
s2

and we know, having a look at the table of course ;o) that

L1 (

1
) = t,
s2

so using the First Shifting Theorem we conclude that

L1 (

1
) = tet .
(s (1))2
7

1
0.8
0.6
0.4
0.2

-2

10

-0.2
-0.4
-0.6
-0.8
-1

Therefore using the Second Shifting Theorem we get:

L1 (e3s

1
) = (t 3)e(t3) H3 (t)
(s + 1)2

Hence:
L1 (

e2s
e3s
1
+
sin (t 2)H2 (t) + (t 3)e(t3) H3 (t).
2
2
2) =

s +
(s + 1)

This can be rewritten as follows:

0, if 0 < t < 2

1
1 sin (t 2), if 2 < t < 3
sin (t2)H2 (t)+(t3)e(t3) H3 (t) =

1
(t3)
,
sin (t 2) + (t 3)e
See the figure above!.

Example 5 Consider the differential equation:

0 t<3

y 00 + y 0 6y =
2 3t<5

0 5t
8

otherwise,

with y(0) = y 0 (0) = 0. The first thing to do is to rewrite the right hand side
in terms of the Heaviside function. The key point is that the Heaviside function
Ha (t) is zero until you get to a and then it is one after that. Now the function on
the right hand side is zero until we get to three and then it is two, so it behaves
like 2H3 (t), however, at five it goes back down to zero, so we have to take away
2H5 (t), in short:

0 t<3

2H3 (t) 2H5 (t) =

2 3t<5

0 5t
The first Heaviside function switches on at t = 3 and brings you up to two,
the second switches on at five and brings you back down to zero.
Now, the differential equation is:
y 00 + y 0 6y = 2H3 (t) 2H5 (t)
with y(0) = y 0 (0) = 0 and we take the Laplace transform of both sides:
s2 Y + sY 6Y =

2 3s
e e5s
s

or
Y (s) =

2
e3s e5s
s(s + 3)(s 2)

Now, partial fractions has

2
1
2 1
1 1
= +
+
s(s + 3)(s 2)
3s 15 s + 3 5 s 2
Via the tables


1
2 1
1 1
1
2 3t 1 2t
+
+
=L + e + e
3s 15 s + 3 5 s 2
3 15
5
Next, in the expression for Y this gets multiplied by various exponential factors:
Y (s) =

1
2 1
1 1
+
+
e3s e5s
3s 15 s + 3 5 s 2
9

the effect of this is to delay the answer:


y(t) =




1
2 3t+9 1 2t6
1
2 3t+15 1 2t10
+ e
+ e
H3 (t) + + e
+ e
H5 (t)
3 15
5
3 15
5
F TRY THEM F

More examples:
1. Compute the inverse Laplace transform of:
4s
i) e 2 ,
s
s
ii) 2se 2 ,
s +w
s
iii) 2 e
,
s + 2s + 2
2s
e8s .
iv) e
2
s +1

2. Compute the solution of the following differential equations:

i) y 00 + 3y 0 + 2y = r(t), where r(t) = 1 if 0 < t < 1 and r(t) = 0, t > 1,
with initial conditions: y(0) = 0 and y 0 (0) = 0.
ii) y 00 + y 0 = r(t), where r(t) = t if 0 < t < 1 and r(t) = 0, t > 1, with
initial conditions: y(0) = 0 and y 0 (0) = 0.
iii) y 00 + 9y = r(t), where r(t) = 8 sin t if 0 < t < and r(t) = 0, t > ,
with initial conditions: y(0) = 0 and y 0 (0) = 4.
Solutions:
1.

4s
i) Notice that the numerator of e 2 indicates a shifting of 4 in t. Then
s
we need to identify the inverse Laplace transform of 12 and apply the
s
Second Shifting Theorem. Via the tables

L1 (
10

1
2
2) = t ,
s

and then:
L1 (

e4s
) = (t 4)H4 (t).
s2

ii) Notice that we know that:

L1 (

ses
) = cos wt
s2 + w 2

and using The Second Shifting Theorem we get:

L1 (

ses
) = cos w(t 1)H1 (t).
s2 + w 2

iii) This example is very interesting because it uses the First and Second
Shifting Theorems at the same time: (wow!!) First we complete the
square in the denominator as follows:
s2 + 2s + 2 = s2 + 2s + 1 + 1 = (s + 1)2 + 1,
then we can rewrite our fraction:
es
es
=
.
s2 + 2s + 2
(s + 1)2 + 1
es in the numerator indicates that we have to use the Second Shifting
theorem. But this implies that we need to compute the inverse Laplace
transform of
1
,
(s + 1)2 + 1
By comparison with the tables we notice that
L1 (

1
)
(s + 1)2 + 1

is very similar to the inverse Laplace of sin t but s-shifted 1 unit,

or in another words, applying the First Shifting theorem we get:
L1 (

1
) = et sin t.
(s + 1)2 + 1
11

Therefore the inverse Laplace Transform of

es
,
(s + 1)2 + 1
by the Second Shifting Theorem we have a t shifted of units:
sin(t )H (t)
iv) This example is really similar to the two first examples but we need
to use linearity first:
L1 (

2s
8s
e2s e8s
1 e
1 e
)
=
L
(
)

L
(
).
s2 + 1
s2 + 1
s2 + 1

We know that L1 (

1 ) = sin t, so using the Second Shifting Thes2 + 1

orem:
L1 (
2.

8s
e2s
1 e
)

L
(
) = sin(t 2)H2 (t) sin(t 8)H8 (t).
s2 + 1
s2 + 1

i) First notice that we can rewriten r(t) in terms of the Heaviside function
as follows:

1,
0,

if 0 < t < 1

= 1 H1 (t)

if t > 0

Then taking Laplace Transform y 00 + 3y + 2y = 1 H1 (t) and using the

theorem to compute Laplace transform of derivatives, linearity and (1)
we get:
s2 Y (s) + 3Y (s) + 2Y (s) =

1 es

s
s

so
Y (s) =

1
es

,
s(s + 2)(s + 1) s(s + 2)(s + 1)

since s2 + 3s + 2 = (s + 2)(s + 1). Now we observe that if we are able to

identify the inverse Laplace Transform of

12

1
then we will
s(s + 2)(s + 1)

be able to identify the inverse Laplace transform of

es
s(s + 2)(s + 1)

using the Second Shifting Theorem.

Let us do partial fractions then:
1
A
B
C
= +
+
,
s(s + 2)(s + 1)
s
s+2 s+3
setting s = 0 we obtain A = 21 , setting s = 2 we obtain B = 12 and
setting s = 1 we have that C = 1. Therefore:
1 1
1
1
1
= ( +
)
,
s(s + 2)(s + 1)
2 s s+2
s+3
and via the tables:
L1 (

1 )) L1 ( 1 )
1
) = 12 (L1 ( 1s ) + L1 ( s +
2
s+3
s(s + 2)(s + 1)
= 12 + 12 e2t et .

Using the Second Shifting Theorem and the above inverse Laplace
transform we have:
L1 (

s
s
s
es
) = 21 (L1 ( es ) + L1 ( se+ 2 )) L1 ( se+ 3 )
s(s + 2)(s + 1)
= 21 + 12 e2(t1) H1 (t) e(t1) H1 (t).

Hence the solution for our ODEs is:

y(t) =

1 1 2t
1 1
+ e et e2(t1) H1 (t) + et1 H1 (t)
2 2
2 2

Dont worry to much about how the solution looks like, what I would
like you to do is to train yourself using the Shifting Theorems and
tables of Laplace Transform. (And of course partial fraction decomposition).
ii) We rewrite r(t) in terms of the Heaviside function as follows:

t, if 0 < t < 1
= (1 H1 (t))t
0, if t > 1
13

Then taking Laplace Transform y 00 + y 0 = (1 H1 (t))t and using the

theorem to compute laplace transform of derivatives and linearity we
get:
s2 Y (s) + sY (s) =

1
L(tH1 (t)).
s2

We compute the Laplace transform of tH1 (t) by definition:

L(tH1 (t)) =

R
0

est tH1 (t)dt

est tdt, because the heaviside function has value 0 in (0, 1)

h st i
R
= es
+ 1s 1 est dt, using integration by parts
t=1
s
s
e
= s e2
s
=

so
(s2 + s)Y (s) =

1
es es
+ 2 .

s
s2
s

Since (s + 1)s = s2 + s, we get:

1
es
es
Y (s) = 3

+
.
s (s + 1) s2 (s + 1) s3 (s + 1)
Therefore if we are able to identify the inverse Laplace Transform
1
and 2 1
then we will be able to identify the inverse
s3 (s + 1)
s (s + 1)
s
s
Laplace transform of 3 e
and 2 e
using the Second Shifting
s (s + 1)
s (s + 1)
Theorem.
of

Let us do partial fractions then:

A B
C
D
1
= + 2+ 3+
,
s
s+1
s (s + 1)
s
s
3

1 = As2 (s+1)+Bs(s+1)+C(s+1)+Ds3 = As3 +As2 +Bs2 +Bs+Cs+C+D3 ,

we conclude that C = 1 = A and B = 1 = D :
1
1
1
1
1
= 2+ 3
,
s s
s+1
s (s + 1)
s
3

14

and via the tables:

L1 (

1
1 )
) = L1 ( 1s ) L1 ( 12 ) + L1 ( 13 ) L1 ( s +
1
s3 (s + 1)
s
s
= 1 t + 12 t2 et

Using the Second Shifting Theorem and the above inverse Laplace
transform we have:
L1 (

es ) = L1 ( es ) L1 ( es ) + L1 ( es ) L1 ( es )
s
s+1
s (s + 1)
s2
s3
3

= 1 (t 1)H1 (t) + 21 t(1)2 H1 (t) e(t1) H1 (t)

Similarly we obtain the partial fraction decomposition of the second
fraction:
A B
C
1
,
= + 2+
s
s+1
s (s + 1)
s
2

1 = As(s + 1) + B(s + 1) + Cs2 = As2 + As + Bs + B + Cs2 ,

giving:
1
1
1
1
,
= + 2 +
s s
s+1
s (s + 1)
2

with inverse Laplace transform:

L1 (

1
1 )
) = L1 ( 1s ) + L1 ( 12 ) + L1 ( s +
1
s (s + 1)
s
2

= 1 + t + et
By the Second shifting theorem we get:
L1 (

es ) = L1 ( es ) + L1 ( es ) + L1 ( es )
s
s+1
s2 (s + 1)
s2
= 1 + (t 1)H1 (t) + e(t1)H1 (t)

Hence the solution for our ODEs is:

y(t) = L1 (

2
s
1
) L1 ( 2 e
) + L1 ( 3 e
)
s (s + 1)
s (s + 1)
s (s + 1)
3

= 1 t + 21 t2 et
+1 (t 1)H1 (t) e(t1) H1 (t)
+1 (t 1)H1 (t) + 21 (t 1)2 H1 (t) e(t1) H1 (t)
= 3 t + 21 t2 et 2(t 1)H1 (t) 2e(t1) H1 (t) + 21 (t 1)2 H1 (t)
15

iii) Rewrite r(t) in terms of the Heaviside function:

8 sin t, if 0 < t <

= 8(1 H (t)) sin t
0, if t >
Then taking Laplace Transform y 00 + 9y = 8(1 H (t)) sin t and using
the theorem to compute laplace transform of derivatives and linearity
we get:
s2 Y (s) + 9Y (s) = 8

1
8L1 (sin tH1 (t)).
s +1
2

We compute the Laplace transform of tH1 (t) by definition:

R

est sin tH1 (t)dt

R
= I = est sin tdt,

L(H1 (t) sin t) =

=
=
=
=

because the heaviside function has value 0 in (0, )

h
st i
R
sin t es
+ 1s est cos tdt, using integration by parts
t=
1 R est cos tdt, because sin = 0
hs st i
R
cos t e 2
12 est sin tdt, using integration by parts
s t= s
1
s
12 I
2e
s
s

so
1
(1 + 2 )
s

est sin tdt = I(1 +

1
1 s
,
2) = 2e
s
s

then:
Z
I=

st

es
sin tdt = 2
s +1

giving:
1
es
Y (s) = 8 2
8 2
.
(s + 1)(s2 + 9)
(s + 1)(s2 + 9)
Once again, if we are able to identify the inverse Laplace Transform of
of

1
then deduce the inverse Laplace transform
(s2 + 1)(s2 + 9)

es
just using the Second Shifting Theorem.
(s + 1)(s+ 9)
2

16

Let us do partial fractions decomposition noticing that neither of the

polynomials in the denominator have real roots:
1
As + B Cs + D
= 2
+ 2
,
2
(s + 1)(s + 9)
s +1
s +9
2

1 = As(s2 +9)+B(s2 +1) = As3 +9As+Bs2 +9B+Cs3 +Cs+Ds2 +D,

we conclude that B = 1 = D and A = 98 = C. Then:
9
9
s1 s+1
1
= 82
+ 82
(s2 + 1)(s2 + 9)
s +1
s +9
= 98 2 s 2 1 98 2 s + 2 1
s +1 s +1
s +9 s +9
and via the tables and using linearity:
L1 (

1
) = 89 L1 ( 2 s ) L1 ( 2 1 )
(s2 + 1)(s2 + 9)
s +1
s +1
98 L1 ( 2 s ) + L1 ( 2 1 )
s +9
s +9
9
9
= 8 cos t sin t 8 cos 3t + 13 sin 3t

Using the Second Shifting Theorem, we have:

L1 (

es
) = ( 89 cos(t ) sin(t )
(s2 + 1)(s2 + 9)
98 cos 3(t ) + 31 sin 3(t ))H (t)

Hence our solution is:

s
1
) + 8L1 ( 2 e 2
)
2
(s + 1)(s + 9)
(s + 1)(s + 9)
= 9 cos t 8 sin t 9 cos 3t + 38 sin 3t
(9 cos(t ) 8 sin(t ) 9 cos 3(t ) + 83 sin 3(t ))H (t)

y(t) = 8L1 (

17