2.

IMPROPER INTEGRALS
We shall extend the concept of definite integral to include
 integrals over infinite intervals;
 integrals in which the integrand become infinite within
the interval of integration.

These are called improper integrals.

b
So far we have worked with integrals  f ( x) dx where [a,b] is finite/bounded.
a
Knew that:
b

For b>a,  f ( x) dx
a
represents
b

the area under the curve y=f(x).  f ( x) dx

1
 2.1 Integrals over infinite intervals (unbounded intervals)

If f is continuous on unbounded interval

the improper integral  f ( x) dx

 b

 f ( x) dx  lim  f ( x) dx
a
b 
a

If the limit exists, let say equal to L, the improper integral is said to
converge to L. We write


 f ( x) dx  L
a

If the limit does not exists, the improper integral is said to diverge, in
which case it is not assigned a value.

2


 f ( x) dx
a
can be regarded as the area under y  f ( x )

over the interval  a,  .

 f ( x) dx
a

Examples
Evaluate:

1. dx

1
x2
 b b
Solution dx dx  1
1 x 2  blim
 
1
x 2  lim
b  


x 
1
 1 
 lim    1  1
b 
 b  3
 
2. dx

1
x b

 

dx  lim dx
Solution
 x b  
1
x
 lim
b 
ln x b
1
1
 lim ln b   This integral diverges.
b 

dx
Generally,  p converges if p  1 and diverges if p  1.
 1
x

2 x
3. e dx
0  b

 
Solution 2 x 2 x
e dx  lim e dx
b 
0 0
b
 e  
2 x
1 1 1
 lim     lim   2b
  
b 
 2 0 b 
 2e 2  2
4
 If f is continuous on the interval  - , b, then we define
b
the improper integral  f ( x)dx as the limit :

b b

 f ( x) dx  lim  f ( x) dx

a  
a

If the limit exists, the improper integral is said to converge. Otherwise it

 dx
x
1. Evaluate e


 
1 1

e
x
dx lim  e dx a  
 lim e x
e a
e
a  
 a 5
 0 
If the two improper integrals  f ( x)dx and  f ( x)dx
-  0

are both converge, then we say that  f ( x)dx


 f ( x)dx   f ( x)dx   f ( x)dx
 0

Example

1
1. Evaluate


1 x 2
dx
Solution

 
 b
dx dx b
 lim 
1
 lim tan x
 1 x2 b 
0
1  x 2 b 
0


0
1
 lim(tan b) 
b 
2 6
0 0


dx
1 x 2
 lim
a   
dx
1 x 2  lim
a  
tan 1
x   0
a
 a
1
 lim (  tan a ) 

a  
2
 
1 0
dx dx


1 x 2
dx
 1 x 2
  1 x 2
 0


Remark
The split of the integral at x=0 is arbitrary.
The split can be made at any other point x=c without
affecting the convergence, the divergence, or the value of
integral. 7 7
Test for convergence and divergence
When an improper integral cannot be evaluated directly,
need to do these:
 establishing the fact of convergence, and then
 approximating the integral numerically.
Cannot
Example  evaluate
Investigate the convergence of 
2
x
e dx directly-
nonelementar
1 y
Solution
 b

e dx lim  e
2
x  x2
By definition, dx
b 
1 1
 x2
But for every value of x  1, have e  e  x , so
 
b 1 1
 dx  e dx  lim[e e ]e
2
x x
e
1 b 
 1

e
Hence,  x2
dx converges for some finite value.
1 8
  x2
The comparison of e and e  x is the special case of the following test :
Theorem 1 Direct Comparison Test
If f and g be continuous on  a,   and suppose that
0  f ( x)  g ( x) for all x  a (as in picture),

then
 
1.  f ( x)dx converges if  g ( x)dx converges
a a
 
2.  g ( x)dx diverges if  f ( x)dx diverges
a a
9
 cannot evaluate by any method
Examples
we’ve learnt

dx
1. The improper integral  converges since
1 1  x3

1 1 dx
 for x  1,   and  3 / 2 converges.
1 x 3 x3/ 2 1
x

sin x
1 x 2
2. converges since
dx

sin x 1 dx
0  2  2 for x  1,   and 1 x 2 converges.
x x

dx
3.

1 x 2  0.1
diverges since


1 1 dx
 for x  1,   and 1 x diverges.
x 2  0.1 x 10
 Theorem 2 Limit Comparison Test
If f and g be continuous on  a,   and suppose that
f ( x)
lim L (0  L  ), then
x  g ( x )
 

 f ( x)dx and  g ( x)dx

Examples

 
dx dx
1. Compare
1 x 2 and 1 1  x 2 with the LCT.

Solution
1 1
Let f ( x)  2 and g(x) 
x 1 x2

11
 f ( x) 1  x 2
 1 
lim
x  g ( x )
 lim 2
 lim  2  1 1
x  x x  
x 
positive finite limit
 
dx dx
Therefore
1 1  x 2 converges since 1 x 2 converges.
Note: However, the integrals converge to different values.
 
3dx dx
2.
1 e x  5 converges since 1 e x converges.
This is because
1/ ex ex  5 1 5 
lim  lim  lim    1/ 3
x   3 /( e  5)
x x x  3 x
3e 
x   3e 

12