Date post: | 15-Jan-2016 |
Category: |
Documents |
Upload: | prateekkumarjaysinghani |
View: | 7 times |
Download: | 0 times |
1
Residue Theory
Definition.1. (Residue). Let f (z) has a non-removable isolated singularity at the point z0. Then f(z) has the Laurent series representation for all z in some punctured disk DR
*(z0) given
by f ( z )= ∑n=−∞
∞
an ( z−z0 )n
The coefficient a-1 of 1
z−z0 is called the residue of f(z) at z0 .
It is denoted by Res[f,z0] = a-1
Example .1.1.
Consider f (z) = e2/z
Then the Laurent series of f about the point z0 = 0is given by
¿1+ 21! z
+ 22
2 ! z2 +23
3! z3 +…,
The co-efficient of 1
z−z0 =
1z−0 =
1z is 2
Hence by definition of residue, residue of f (z) = e2/z at z0 = 0 is given by Res [f, z0] = 2
Example .1.2. Find residue of f (z) = 3
2 z+z2−z3 at z0 = 0
f(z) = 3
z (2+z−z2 ) = 3
z ( z+1 ) (2−z )
Now 3
z ( z+1 ) (2−z )= Az
+Bz+1
+ C
2−z
⇒A(z+1)(2-z) + Bz(2-z) + Cz (z+1) = 3
⇒A(-z2 +2 + z) +B(2z –z2) +C(z2 +z) = 3
⇒-A –B +C = 0---- (a)
A +2B +C = 0 ---- (b)
2A =3 ---- ( c) ⇒A = 3/2
2
(a) ⇒-B + C = A= 3/2
(b) ⇒2B + C = -A = -3/2
-----------------------------
Adding B + 2C =0⇒ B = - 2C
Put B = - 2C in -B + C = 3/2
⇒3C = 3/2⇒C= 1/2
Put C= 1/2, B = -2C = -1⇒ B = -1
Hence f(z) = = Az
+Bz+1
+ C
2−z =
32 z
- 1
z+1 +
12(2−z )
= 3
2 z - (1+z) -1 +
12
(2-z)-1 = 3
2 z - (1- z+ z2 -….) +
12
2-1(1− z2 )
−1
= 3
2 z - (1- z+ z2 -….) +
14
(1 + z2
+ ( z2 )2
+ ….)
= 3
2 z -
34
+9 z8
- …..
The residue of f at 0 is given by Res [f,0] = coefficient of 1z
= 32
Example 1.3. Find residue of f (z) = ez
z3 at z0 = 0
Laurent expansion of f(z) = 1z3 {1+z+ z2
2 !+ z3
3!…}
= 1
z3+ 1
z2+ 1z 2!
+ 13 !
…
The residue of f at 0 is given by Res [f,0] = coefficient of 1z
= 12
Contour integration
Contour integration is the process of calculating the values of a contour integral around a given
3
contour in the complex plane.
The Cauchy integral formulae are useful in evaluating contour integrals over a simple closed
contour C where the integrand has the form f ( z )
( z−z0 )k and f is an analytic function
Example 1.
Evaluate ∫C
ydzalong the curve C : x = t-1, y = et-1, 2< t < 3
Solution.
Let z = x+iy ⇒ dz = dx + i dy
Given curve x = t-1 ⇒ dx = dt
And y = et-1 ⇒ dy = et-1dt
∫C
ydz = ∫2
3
et−1(dx+ idy) = ∫2
3
et−1(dt+i e t−1dt) = ∫2
3
et−1dt + ∫2
3
ie2 (t−1)dt
= e t−1 ] 32+ie2 t−2
2 ]32= e2 – e + ie4−e2
2 =
Example 2.
If C is the curve y = x3-3x2+4x-1 joining the points (1,1) and (2,3) then
find the value of ∫C
(12 z2−4 iz)dz
∫C
(12 z2−4 iz)dz = ∫1+i
2+3 i
(12 z2−4 iz)dz = 12 z3
3 - 4 i z2
2 ]2+3i1+i
= 4(2+3i)3 – 2i(2+3i)2 - 4(1+i)3 + 2i(1+i)2 = -156 +38i
Recall (i).(Cauchy’s integral Theorem)
Let D be any simply connected domain. Let C be any closed contour contained in D and f(z)
4
analytic in D, then ∮C
f (z)dz = 0
Recall (ii).
For a function f(z) analytic in DR*(z0)and for any r with 0 < r < R, the Laurent series
coefficients of f(z) are given by
an=1
2πi∮C
f ( ξ )(ξ−z0 )n+1 dξfor n =0,±1 ,±2, ….. ------(I)
WhereC denotes the circle {z: |z-z0| = r} with positive orientation.
Put n = -1 in Equation (I) and replace C with any positively oriented simple closed contour C
containingz0, provided z0 is the still only singularity of f(z) that lies inside C,
then we obtain a-1 = 1
2πi∮C
f (z)dz.
We know that a-1 is the Res[f,z0]⇒ Res[f,z0] = 1
2πi∮C
f (z)dz
⇒∮C
f (z)dz = 2πi Res[f,z0]
If we know the Laurent series expansion for f(z), then using above equation we can evaluate contour integrals.
Example ii.1. Evaluate ∮C
e2z dzwhere C denotes the circle C = {z: |z| =1}with positive
orientation.
5
Solution.
Let f(z) = e2z
From Example.1, we have Res [f, 0] = 2
Recall 2. gives us ∮C
f (z)dz = 2πi Res[f,z0]
Hence ∮C
e2z dz = 2πi Res[f,0] = 2πi (2) = 4πi
Theorem 1 (Cauchy's Residue Theorem).
Let D be a simply connected domain, and let C ⊂D be a closed positively oriented contour within and on the functionf(z) is analytic, except finite number of singular z1,z2,….,zn , then
∮C
f (z)dz = 2πi∑k=1
n
Res [ f , zk ]
Proof.
Let Ci be the neighborhood of zi, (i=1,2…n) lies inside C such that all Ciare disjoint.
Since each zi is a singular point of f and each Ci is a neighborhood of corresponding zi(i=1,2,..n), f is analytic in and on C except these neighborhoods Ci (i=1,2…n).
Then by Cauchy’s Theorem,(Recall 1)
∮C
f (z)dz - ∮C1
f (z)dz - …… - ∮Cn
f (z)dz = 0
6
⇒∮C
f (z)dz = ∮C1
f (z)dz + …… + ∮Cn
f (z)dz
⇒∮C
f (z)dz = 2πiRes[f,z1] +….. + 2πi Res [f,zn] ( by Recall 2)
⇒ ∮C
f (z)dz = 2πi∑k=0
n
Res [ f , zk ]
Note 1.1: The residue at z0depend only the coefficient a-1in the Laurent expansion, if f(z) has a
removable singularity at z0, then the Laurent expansion has no negative power term and
hencea-1=0 ⇒ Res[f,z0] = 0.
Theorem 2.(Residues at Poles).
(i) If f(z) has a simple pole at z0, then Res[f,z0] = lim ¿ z→ z0 ( z−z0 ) f (z)
(ii) If f(z) has a pole of order 2 at z0, then Res[f,z0] = lim ¿ z→ z0ddz
(( z−z0 )¿¿2¿¿ f (z ))¿¿
(iii) If f(z) has a pole of order 3 at z0, then Res[f,z0] = 12!
lim ¿ z→z0d2
d z2 (( z−z0 )¿¿3¿¿ f (z ))¿¿
(v) If f(z) has a pole of order k at z0,
then Res[f,z0] = 1
(k−1)!lim ¿ z→z0
dk−1
d zk−1 ( ( z−z0 )¿¿k¿¿ f (z))¿¿
Proof.
i) Suppose f(z) has a simple pole at z = z0, then the Laurent series expansion
f(z) = ∑n=0
∞
an ( z−z0 )n+a−1 ( z−z0 )−1
⇒ (z -z0)f(z) = (z-z0) ∑n=0
∞
an ( z−z0 )n+ a-1
7
⇒ (z -z0)f(z) = (z-z0) ∑n=0
∞
an ( z−z0 )n + Res [f,z0]
Taking lim z→z0, both sides
¿ lim ¿ z→z 0 ( z−z0¿ f (z) = lim ¿ z→ z0 (z-z0) ∑n=0
∞
an ( z−z0 )n +lim ¿ z→ z0Res[f,z0]
= 0 + Res[f,z0]
Hence Res[f,z0] = ¿ lim ¿ z→z 0 ( z−z0¿ f (z)
v) Suppose f(z) has a pole of order k at z = z0, then the Laurent series expansion
f(z) = ∑n=0
∞
an ( z−z0 )n+a−1 ( z−z0 )−1+ a−2 ( z−z0)−2 + …… + a−k ( z−z0 )−k
Multiply both sides by (z-z0)k
(z-z0)kf(z) = a−k + ….. + a−1 ( z−z0 )k−1 + ∑n=0
∞
an ( z−z0 )n ( z−zo )k
Differentiate both sides k-1 times with respect to z,
dk−1
d zk−1 ( z−z0 )k f (z ) = 0+0+…..+ a−1 ( k−1 ) ! + dk−1
d zk−1 ¿
= a−1 ( k−1 ) ! + a0(z−z0)k ! + a1 ( z−z0 )2( (k+1 )!2! ) + ….
Taking lim z→z0, both sides
lim ¿ z→ z0dk−1
d zk−1 ( z− z0 )k f (z ) = lim ¿ z→ z0a−1 ( k−1 ) ! + 0 +…
= a−1 ( k−1 ) ! = Res [f,z0] (k-1)!
Hence Res [f,z0] = 1
(k−1 )!lim ¿ z→z0
dk−1
d zk−1 ( z−z0 )k f (z)
ii) and iii) are the particular case of v) (take k = 2 and k= 3)
8
Example 2.1.
Find residue of f (z) = ez
z2−1 at z0 = 1
Solution.
Given f (z) = ez
z2−1 =
ez
(z−1)(z+1)
The poles of f(z) are z = 1 , z =-1 (simple poles)
Res [f,1] = lim ¿ z→1 ( z−1 ) f (z) = lim ¿ z→1( z−1 ) ez
( z−1 ) (z+1 ) = lim ¿ z→1
ez
( z+1 ) =
e2
Example 2.2 .
Find the residue of f(z) = (z8- ω8)-1, where ω is any complex constant.
Solution.
Given f(z) = (z8- ω8)-1 = 1
(z8−ω8)
The poles of f(z) are the zeros of z8-ω8 ⇒ zeros are given by z8-ω8 = 0
⇒ z8 = ω8 ⇒ z8 = ω8(cos2nπ +isin2nπ), n = 0,1,2,…7 ⇒ z8 = ω8 e2niπ
⇒ z = ω e2nπi/8 ⇒ z = ω enπi/4 , n = 0,1,2,…7
Hence z = ω enπi/4 , n = 0,1,2,…7 are the simple poles of f(z)
Let an = ω enπi/4, n = 0,1,2,…7
The residue of f (z) at z = an , is given by Res [f, an] = lim ¿ z→an ( z−an ) f (z )
=lim ¿ z→an ( z−an ) 1
( z8−an8)
Since it is not easy to factories (z8- an8) into eight factors , so we have to use L’Hospital’s rule
(that is differentiating Nr and Dr separately w.r.to z)
9
= lim ¿ z→an1
8 z7 = 1
8an8 , n = 0,1,….7.
Example 2.3
Find the residue of 1
sinhπz
Solution.
Given f(z) = 1
sinhπz
The poles of f(z) are the zeros of sinhπz,
Also the zeros of sinhπz are z = ni, for all integer n (since sinhnπi = 0 for all n)
Hence Res [f,ni] = lim ¿ z→∋¿ ( z−¿ ) 1sinhπz
¿
By using L’Hospital’s rule
Res [f,ni] = lim ¿ z→∋¿1
πcoshπz¿
= 1
πcoshπni =
1πco sπn
(Since coshix= cosx)
= (-1)n/π (Since
Example 2.3. Find the residue of f(z) = π cot (πz )
z2 at z0 = 0
Solution.
Given f(z) = π cot (πz )
z2 = π cos (πz )z2 sin (πz )
Since z2 has a zero of order 2 at z0 = 0 and sin(πz) has a simple zero 1 at z0 = 0, we have
z2sin(πz) has a zero of order 3at z0 = 0 and πcos(πz) ≠ 0.
Hence f(z) has a pole of order 3 at z0 = 0 .
By part (iii) of Theorem .2, we have
10
Res [f,0] = 12!
lim ¿ z→0d2
d z2 (( z−0 )¿¿3¿¿ f (z ))¿¿
= 12!
lim ¿ z→0d2
d z2 (z¿¿3¿¿π cot (πz )
z2 )¿¿
= 12!
lim ¿ z→0d2
d z2 (z¿¿ ¿¿ π cot (πz))¿¿ =
12!
lim ¿ z→0dd z
(¿¿¿ π cot (πz )−πzcsc2(πz) π)¿¿
= 12!
lim ¿ z→0dd z
(¿¿¿ π cot (πz )−π2 zcs c2(πz ))¿¿
=12!
lim ¿ z→0(¿−¿π csc2 (πz )π−π 2cs c2 (πz )+π2 z2csc (πz )csc (πz)cot (πz )π )¿¿
= 12!
lim ¿ z→0(¿−¿2π2 csc2 (πz )+2 π3 z csc2 (πz ) cot (πz ))¿¿
= 2π2
2 !lim ¿ z→0csc2(πz)(¿−¿1+π zcot (πz))¿¿
= π2 lim ¿ z→0
1
sin2 (πz )(¿−¿1+π z
cos (πz )sin (πz )
)¿¿
Res [f,0] = π2 lim ¿ z→0
(πzcos (πz )−sin (πz ))sin3 (πz )
As lim z → 0, LHS is indeterminate, so we have to use L’Hospital’s rule to evaluate the limit
(that is differentiating Nr and Dr separately w.r.to z)Res [f,0] = π
2 lim ¿ z→0(πcos (πz )−πzsin(πz)π−πcos (πz ))
3 sin2(πz )cos (πz )π
= π2 lim ¿ z→0
−π2 zsin(πz)3πsin2 (πz ) cos (πz )
= π2 lim ¿ z→0
−π z3sin (πz ) cos (πz )
11
= π 2
3lim ¿ z→0
−πzsin (πz ) cos (πz )
= −π 2
3lim ¿ z→0 ( πz
sin (πz ) ) lim ¿ z→0( 1cos (πz ) )
= −π 2
31
lim ¿ z→0( sin (πz )πz )
lim ¿ z→0( 1cos (πz ) ) = −π 2
3(1)(1) = −π 2
3
Example 2.4.
Find ∫C
dzz4+z3−2 z2 where C denotes the circle {z: |z| =3} with positive orientation.
Solution.
Let f(z) = 1
z4+z3−2 z2 = 1
z2(z¿¿2+z−2)¿ =
1
z2(z¿¿ +2)(z−1)¿
The singularities of f(z) that lie inside C are simple poles at the points z =1 and z= -2, and a pole of order 2 at z =0.
To find the Residue at z = 0 :
Res [f,0] =lim ¿ z→0ddz
(( z−0 )¿¿2¿¿ f (z ))¿¿ = lim ¿ z→0ddz
¿
= lim ¿ z→0ddz
¿= lim ¿ z→0ddz
¿
= lim ¿ z→0−2 z−1
(z2+z−2)2 = - 14
To find the Residue at z = 1 :
12
Res [f,1] = lim ¿ z→1 ( z−1 ) f (z ) = lim ¿ z→1 ( z−1 ) 1
z2(z¿¿+2)(z−1)¿
= lim ¿ z→11
z2( z¿¿ +2)¿= ¿
13
To find the Residue at z = -2 :
Res [f,-2] = lim ¿ z→−2 ( z+2 ) f (z ) = lim ¿ z→−2 ( z+2 ) 1
z2(z¿¿+2)(z−1)¿
= lim ¿ z→−21
z2(z−1) =
−112
By Cauchy’s residue theorem ∮C
f (z)dz = 2πi∑k=0
n
Res [ f , zk ]
∫C
dzz4+z3−2 z2 = 2πi(Res[f,0] +Res[f,1]+Res[f,-2]) = 2πi(
−14
+ 13− 1
12¿ = 0
Example 2.5. Find ∫C
dzz4+4
where C denotes the circle {z: |z-1| =2 } with positive orientation
Solution.
Let f(z) = 1
z4+4
To find the poles of f(z) , we know that poles of f(z) is nothing but the zeros of z4 +4
Now we have to find the zeros of z4 +4
Put z4 +4 =0 ⇒z4 = -4 = 4i2 = (2i)2⇒ z2 = ±2i
Let z = a+ib ⇒ z2 = (a+ib)2 = a2 + 2iab – b2
13
Supposez2= 2i ⇒a2–b2+ 2iab = 2i ⇒ a2 –b2=0 and ab = 1
⇒ a2 = b2and b = 1/a
⇒ a = ±b and b = 1/a
If a = b , then b = 1 ⇒ a,b=1
If a = -b , then b = -1⇒a = 1 , b = -1
The zeros are z = a+ib , 1+i, 1-i
Suppose z2= -2i ⇒a2 –b2+ 2iab = -2i ⇒ a2 –b2 =0 and ab = -1
⇒ a2 = b2 and b = -1/a
⇒ a = ±b and b = -1/a
If a = b , then b = -1 ⇒ a,b=-1
If a = -b , then b = 1⇒a = -1 , b =1
The zeros are z = a+ib ,-1-i, -1+i
Hence the poles of f(z) are 1±i, -1±i (simple poles)
The poles lie inside the circle {z: |z-1| =2} with positive orientation are 1±i
Res[f,1+i] = lim ¿ z→1+ i ( z−(1+i)) f (z) =lim ¿ z→1+ i (z−(1+i ) )( 1
z4+4)
As lim z →1+i, LHS is indeterminate, so we have to use L’Hospital’s rule to evaluate the limit
(that is differentiating Nr and Dr separately w.r.to z) = lim ¿ z→1+ i( 1
4 z3) =lim ¿ z→1+ i ( z
4 z4) =
1+ i4 (1+i )4
= 1+i
4 (−4 ) = 1+i−16
Similarly
Res[f,1-i] = 1−i−16
By Cauchy’s residue theorem ∮C
f (z)dz = 2πi∑k=0
n
Res [ f , zk ]
14
∫C
dzz4+4
= 2πi(Res[f,1+i] +Res[f,1-i]) = 2πi(1+i−16
+ 1−i−16
¿ = - πi4
Result 3.
Let P(z) be a polynomial of degree at most 2. If a ,b and c are distinct complex numbers, then
f(z) = P ( z )
(z−a)(z−b)(z−c ) =
A(z−a) +
B(z−b) +
C(z−c)
Where A = Res [f,a] = P (a )
(a−b)(a−c)
B = Res [f,b] = P (b)
(b−a)(b−c )
C = Res [f,c] = P (c)
(c−a)(c−b)
Example 3.1.
Find the residue of f(z) = 3 z+2
z ( z−1 )(z−2) and express f(z) in partial fractions.
Solution.
In Result I, take a= 0, b = 1, c= 2 and P(z) = 3z+2.
The residues are
A = Res[f,0] = P(0)
(0−1)(0−2) = 1
B = Res [f,1] =P(1)
(1−0)(1−2) = -5
C = Res [f,2] = P(2)
(2−0)(2−1) = 4
The partial fraction expression of f(z)is given by
f(z) = A
(z−a) + B
(z−b) + C
(z−c)
= 1
(z−0) + −5
(z−1) + 4
(z−2)
15
= 1z
- 5
(z−1) + 4
(z−2)
Example 3.2. Find the residue of f (z) = 1
z4−1 and express in partial fractions.
f (z) = 1
z4−1 =
1
(z¿¿2−1)(z2+1)¿ =
1( z−1 )(z+1)(z+i)(z−i) =
A
z−1 +
B
z+1 +
C
z−i +
D
z+i
Where A = Res[f,1] = lim ¿ z→1(z−1) f (z) = lim ¿ z→1(z−1) 1( z−1 )(z+1)(z+i)(z−i)
= 1
2(1+i)(1−i) = 1
2(1−(−1)) = 14
B= Res[f,-1] = lim ¿ z→−1(z+1) f (z ) = lim ¿ z→−1(z+1) 1( z−1 )(z+1)(z+ i)(z−i)
= 1
−2(−1+i)(−1−i) = 1
2(−1−1)¿¿ =
−14
C= Res [f,i] = lim ¿ z→i(z−i) f (z ) = lim ¿ z→i(z−i) 1( z−1 )(z+1)(z+i)(z−i)
= 1
2i(i+1)(i−1) = 1
2i(−1−1) = −14 i
D=Res[f,-i] = = lim ¿ z→−i(z+i) f (z) = lim ¿ z→−i(z+i) 1( z−1 )(z+1)(z+i)(z−i)
= 1
−2i(−i+1)(−i−1) = 1
2i(1−(−1 )) = 14 i
f(z) = A
z−1 +
B
z+1 +
C
z−i +
D
z+i =
1
4 (z−1) -
14 (z¿¿+1)¿ -
14 i(z¿¿−i)¿ +
14 i(z¿¿ +i)¿
Result 4.
If a repeated root occurs in partial fraction, and P(z) has degree of at most 2, then f(z) = P ( z)
( z−a )2(z−b) =
A
( z−a )2 +
B(z−a)
+ C
(z−b)
16
Where A = Res [(z-a)f(z),a]
B = Res [f, a]
C = Res [f, b]
Example 4.1. Find the residue of f(z) = z2+3 z+2z2(z−1)
and express in partial fraction.
Solution.
In Result II, take a = 0, b = 1 and P(z) = z2+3 z+2 , we have
f(z) = P (z)
(z−0)2(z−1) =
A
( z−0 )2 +
B(z−0) +
C(z−1)
Where A = Res [(z-0)f(z),0] = Res [ z z2+3 z+2z2(z−1)
,0] = Res [z2+3 z+2z (z−1)
,0]
= lim ¿ z→0(z−0)( z2+3 z+2z (z−1)
)= lim ¿ z→0( z2+3 z+2(z−1)
) = -2
B = Res [f, 0] = lim ¿ z→0ddz
(z−0 )2 z2+3 z+2z2(z−1) = lim ¿ z→0
ddz
z2 z2+3 z+2z2(z−1) =
lim ¿ z→0ddz
z2+3 z+2(z−1)
= lim ¿ z→0[ (2 z+3 ) ( z−1 )−(1 ) ( z2+3 z+2 ) ]
( z−1 )2
¿
¿ = lim ¿ z→0[ ( z2−2 z−5 ) ]
( z−1 )2
¿
¿= -5
C = Res [f, 1] = Res [z2+3 z+2z2(z−1)
,1] = lim ¿ z→1(z−1)( z2+3 z+2z2(z−1)
)
= lim ¿ z→1( z2+3 z+2
z2 ) = 6
f(z) = −2
( z−0 )2 +
−5(z−0) +
6(z−1) =
−2
z2 + −5z
+ 6
(z−1)
17
Example 4.2. Find the residue of f(z) = 1
( z−1 )2(z−3)
Take P(z) = 1, a= 1, b=3
f(z) = P(z )
( z−1 )2(z−3) =
A
( z−1 )2 +
B
( z−1 ) +
C(z−3)
A = Res [(z-1)f(z),1] = Res [ (z-1) 1
( z−1 )2(z−3),1¿ = lim ¿ z→1(z−1)( 1
(z−1)( z−3))
= 1
(1−3) = 1
−2
B = Res [f,1] = lim ¿ z→1ddz
( z−1 )2 1
(z−1)2( z−3) = lim ¿ z→1 −1(1)( z−3 )2
= −14
C = Res [f,3] = lim ¿ z→3(z−3)( 1
( z−1 )2(z−3)) = lim ¿ z→3
1
( z−1 )2 =
14
EVALUATION OF REAL DEFINITE INTEGRALS
Cases of poles are not on the real axis.
Type I
Evaluation of the integral ∫0
2π
f (cosθsinθ )dθ where f(cosθ sinθ) is a real rational function of
sinθ,cosθ.
First we use the transformation z = eiθ = cosθ + i sinθ ------ ( a)
And 1z
= 1
eiθ = e-iθ = = cosθ - i sinθ -------- (b)
From (a) and (b) , we have cosθ = 12(z+ 1
z) , sinθ =
12i
(z−1z)
Now z = eiθ ⇒ dz = ieiθdθ ⇒ dθ = dziz
18
Hence ∫0
2π
f (cosθsinθ )dθ = ∫C
f [¿ 12 (z+ 1
z ) , 12 i
(z−1z)] dziz
¿
Where C, is the positively oriented unit circle |z| = 1
The LHS integral can be evaluated by the residue theorem and
∫C
f [¿ 12 (z+ 1
z ) , 12 i
(z−1z)] dziz
¿ = 2πi ∑ Res(zi) , where zi is any pole in the interior of the
circle |z| =1
Example I.1.
Evaluate ∫0
2π
e−cosθ cos (nθ+sinθ )dθ ,where nis a positive integer .
Solution.
Let I = ∫0
2π
e−cosθ¿¿
= ∫0
2π
e−cosθ e−i (nθ+sinθ ) dθ = ∫0
2π
e−cosθ−isinθe−i (nθ ) dθ
= ∫0
2π
e−(cosθ+isinθ)e−i (nθ) dθ = ∫0
2π
e−e iθ
e−i (nθ )dθ
Let z = eiθ, dz = ieiθ dθ ⇒ dθ = dz
ie iθ = dziz
and C denotes the unit circle |z| = 1
Therefore I = ∫C
e− z
zn ( dzzi ) = 1i∫C
e−zdzzn+1 = ∫
C
f ( z )dz where f(z) = e− z
i zn+1
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z)
To find the poles of f(z) :
Since poles of f(z) = to the zeros of izn+1 ,and the only zero of izn+1 is z = 0 of order n+1
19
Hence the pole of f(z) is z =0 of order n+1
There are no poles on the real axis
To find the residue of f(z):
Res [f,0] = 1n !
lim ¿ z→0dn
d zn( z−0 )n+1
f ( z) =1n !
lim ¿ z→0dn
d zn( z−0 )n+1 e−z
i zn+1
= 1n !
lim ¿ z→0dn
d znzn+1 e−z
i zn+1 = 1n !
lℑz→0
dn
d zne− z
i =
1n !
lim ¿ z→0 (−1 )n e−z
i
= (−1 )n
n ! i
Hence ∑ Res[f,zk] = Res[f,0]= (−1 )n
n ! i
Therefore ∫C
f ( z )dz = 2πi∑ Res[f,zk] = 2πi (−1 )n
n ! i = 2π
(−1 )n
n!
⇒ I = 2π (−1 )n
n!
⇒ ∫0
2π
e−cosθ¿¿ = 2π (−1 )n
n!
⇒ ∫0
2π
e−cosθ cos (nθ+sinθ )dθ – i∫0
2π
e−cosθ sin(nθ+si nθ)¿ dθ ¿ = 2π (−1 )n
n!
Equating real and imaginary parts,
∫0
2π
e−cosθ cos (nθ+sinθ )dθ = 2π (−1 )n
n!
And ∫0
2π
e−cosθ sin(nθ+sinθ)¿dθ ¿ = 0
Example I.2.
Prove that ∫0
2πdθ
a+bcosθ =
2π
√a2−b2 , a >b >0.
20
Solution.
Let I = ∫0
2πdθ
a+bcosθ
Put z = eiθ ⇒ dθ = dziz
and let C denotes the unit circle |z| = 1
Since z = eiθ = cosθ + isinθ and 1z
= cosθ – isinθ, we have cosθ = 12
(z + 1z¿
I = ∫0
2πdθ
a+bcosθ =
1i∫C
dz
z (a+ b2 (z+ 1
z )) = 1i∫C
dz
z a+ z2b2
+ b2
= 2bi∫C
dz
z2ab
+z2+1
= ∫C
f ( z )dz where f(z) = 2
bi(z2+2azb
+1)
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
To find the poles of f(z):
Poles of f(z) = to the zeros of bi(z2+2azb
+1) and the zeros are given by bi(z2+2azb
+1) = 0
⇒ (z2+2azb
+1) =0 -------- (a)
⇒ z = −2ab
±√ 4 a2
b2−4
2
= −2ab
±√ 4 a2−4b2
b2
2
= −2ab
±2√a2−b2
2b =
−a±√a2−b2
b
⇒ z = −a+√a2−b2
b or
−a−√a2−b2
b are the simple poles of f(z)
Let α = −a+√a2−b2
b and β = −a−√a2−b2
b , these the roots of the equation (a)
Now the product of the roots αβ = 11
= 1
21
Now | αβ| = 1 ⇒ |α||β| = 1
Since a > b > 0, |β| = |−a−√a2−b2
b | = |a+√a2−b2
b | Here a > b ⇒ a2 > b2 ⇒ a2- b2>0⇒ √a2−b2 > 0 ⇒ a + √a2−b2 > a >b
⇒ a + √a2−b2 >b ⇒ a+√a2−b2
b > 1 ⇒ |a+√a2−b2
b | > 1
Hence z = α = −a+√a2−b2
b < 1 is the only simple pole lie inside the circle |z| = 1
There are no poles on the real axis
To find the residue of f(z):
Res [f,α] = lim ¿ z→α (z−α) f (z )
Here f(z) = 2
bi(z2+2azb
+1) and α,β are the factors of z2+2azb
+1
⇒ f(z) = 2
bi(z2+2azb
+1) = 2
bi(z−α)(z−β)
∴ Res [f,α] = lim ¿ z→α (z−α) f (z ) = lim ¿ z→α (z−α) 2bi(z−α )(z−β ) =
lim ¿ z→α2
bi(z−β) = 2
bi(α−β)
=
2
bi((−a+√a2−b2
b )−(−a−√a2−b2
b )) = 2
bib
(−a+√a2−b2+a+√a2−b2 ) = 2
2i √a2−b2 =
1
i√a2−b2
Hence ∑ Res[f,zk] = Res[f,α]= 1
i√a2−b2
22
∴∫C
f ( z )dz = 2πi∑ Res[f,zk] = 2 πi
i√a2−b2 = 2π
√a2−b2
∴ ∫0
2πdθ
a+bcosθ =
2π
√a2−b2
Example I.3.
Prove that ∫0
2πdθ
1+a2−2acosθ= 2π
1−a2 , 0 ≤ a < 1.
Solution
Let I = ∫0
2πdθ
1+a2−2acosθ
Let z = eiθ ⇒ dθ = dziz
and let C denotes the unit circle |z| = 1
Since z = eiθ = cosθ + isinθ and 1z
= cosθ – isinθ, we have cosθ = 12
(z + 1z¿
∴ I = ∫0
2πdθ
1+a2−2acosθ = ∫
C
dziz
1+a2−2a( 12 )(z+ 1
z ) =
1i∫C
dzz
(1+a2 )−a ( z2+1 )z
=
1i∫C
dzz
(1+a2 ) z−a ( z2+1 )z
= 1i∫C
dz
(1+a2 ) z−a ( z2+1 ) =
1i∫C
dzz+a2 z−az2−a
= −1ai
∫C
dz−za
−az+z2+1 =
−1ai
∫C
dz
(−za
+z2)+(−az+1)
23
= −1ai
∫C
dz
z (−1a
+z)−a(z−1a) =
−1ai
∫C
dz
(−1a
+z )(z−a) = ∫
C
f ( z )dz
Where f(z) =
−1
ai(−1a
+z )(z−a)
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
To find the poles of f(z):
Poles of f(z) = zeros of ai(−1a
+ z)(z−a) and
these zeros are given by ai(−1a
+ z)(z−a) = 0
⇒ (−1a
+z ) = 0 or (z−a) = 0
⇒ z = 1a
or z = a which are the simple poles of f(z)
Since 0≤ a < 1, 1a
> 1
Hence a<1 is the only pole lie inside the unit circle |z| =1
There are no poles on the real axis
To find the residue of f(z):
Res[f,a] = lim ¿ z→a(z−a) f (z) = lim ¿ z→a(z−a) −1
ai(−1a
+z )(z−a)
= lim ¿ z→a
−1
ai(−1a
+z) =
−1
ai(−1a
+a) = −1
i (−1+a2 ) = 1
i (a2−1 )
24
Hence ∑ Res[f,zk] = Res[f,a] = 1
i (a2−1 )
∴∫C
f ( z )dz = 2πi∑ Res[f,zk] = 2πi
i (a2−1 ) = 2 π
(a2−1 )
⇒ I = 2 π
(a2−1 )
⇒ ∫0
2πdθ
1+a2−2acosθ =
2 π
(a2−1 )
Example I.4.
Evaluate ∫0
πadθ
a2+sin2θ
Solution.
First we have to change the limits to 0 to 2π from 0 to π for the given integral.Let I = ∫
0
πadθ
a2+sin2θ = ∫
0
πadθ
a2+ 1−cos 2θ2
= ∫0
π2adθ
2a2+1−cos 2θ
To change the limit , take 2θ = ϕ ⇒ 2dθ = dϕ If θ =0 , then ϕ = 0If θ= π, then ϕ = 2πHence I = ∫
0
2πadϕ
2a2+1−cosϕ
Put z = eiϕ ⇒ dϕ = dziz
and let C denotes the unit circle |z| = 1
Since z = eiϕ = cosϕ + isinϕ and 1z
= cosϕ – isinϕ, we have cosϕ = 12
(z + 1z¿
∴ I = ∫0
2πadϕ
2a2+1−cosϕ = ∫
C
adziz
2a2+1−12(z+
1z) = ∫
C
adziz
4 za2+2 z−( z2+1 )2 z
25
= 2ai∫C
dz4 z a2+2 z−z2−1
= −2ai
∫C
dz
z2−2 z (2a2+1 )+1 = ∫
C
f ( z )dz
Where f(z) = −2a
i( z¿¿2−2 z (2a2+1 )+1)¿
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
To find the poles of f(z):
Poles of f (z) = zeros of i(z2-2z(2a2+1)+1) , these zeros are given by i(z2-2z(2a2+1)+1) = 0
⇒ z2-2z(2a2+1)+1 = 0
⇒ z = 2(2a2+1)±√(2 (2a2+1 ))2−4 (1)(1)
2 (1) = 2(2a2+1)±2√( (2a2+1 ))2−1
2
= 2a2+1±√(2a2+1 )2−1 = 2a2+1±√4a4+1+4 a2−1
= 2a2+1±√4a4+4 a2 = 2a2+1±2a√a2+1
∴ z = 2a2+1+2a√a2+1 = α (say)
Or z = 2a2+1−2a √a2+1 = β (say)
Hence the poles of f(z) are α,β which are simple poles.
Now α,β are the roots of the equation z2-2z(2a2+1)+1 = 0
Product of the roots αβ = 1 ⇒ | αβ| = 1⇒|α||β| = 1
Clearly |α| = |2a2+1+2a√a2+1 | > 1 ⇒ |β| < 1
∴ the only pole lie inside the unit circle |z| =1 is β = 2a2+1−2a √a2+1
There are no poles on the real axis
To find the residue of f(z) :
Res[f,β] = lim ¿ z→ β(z−β) f (z ) = lim ¿ z→ β(z−β) −2a
i(z¿¿2−2 z ( 2a2+1 )+1)¿
26
= lim ¿ z→ β(z−β) −2a
i(z−α)(z−β)= lim ¿ z→ β
−2a
i(z−α) =
−2a
i( β−α ) =
−2a
i(2a2+1−2a√a2+1−(2a2+1+2a√a2+1 ))
= −2a
i(2a2+1−2a√a2+1−2a2−1−2a√a2+1) =
−2a
i(−4 a√a2+1) =
1
2i √a2+1¿¿
Hence ∑ Res[f,zk] = Res[f,β] = 1
2i √a2+1¿¿
∴∫C
f ( z )dz = 2πi∑ Res[f,zk] = 2πi
2i √a2+1¿¿ =
π
√a2+1¿¿
⇒ I = ∫0
πadθ
a2+sin2θ = π
√a2+1¿¿
Example I.5.
Evaluate ∫0
2πdθ
(a+bcosθ )2 ( a > 0, b > 0 ; a > b)
Solution
Let I = ∫0
2πdθ
(a+bcosθ )2
Take z = eiθ ⇒ dθ = dziz
and let C denotes the unit circle |z| = 1
Since z = eiθ = cosθ + isinθ and 1z
= cosθ – isinθ, we have cosθ = 12
(z + 1z¿
⇒ I = ∫C
dziz
(a+ b2(z+1
z))
2 = ∫C
dziz
( 2az+b ( z2+1 )2 z )
2 = 4i∫C
zdz
(2az+b(z2+1))2 =
4i∫C
zdz
( 2az+bz2+b¿)2
27
= ∫C
f ( z )dz where f(z) = 4 z
i (2az+b z2+b¿)2
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
To find the poles of f(z):
Poles of f(z) = zeros of i(2az+bz2+b)2 , these zeros are given by i(2az+bz2+b)2 = 0
⇒ bz2+2az +b = 0 ⇒ z2+2azb
+1 = 0
⇒ z = −2ab
±√(2ab
)2
−4
2 = −2a±√4 a2−4b2
2b = −a±√a2−b2
b
⇒ z = −a+√a2−b2
b = β (say)
Or z = −a−√a2−b2
b = α (say)
Hence the poles of f(z) are α,β both order 2
Since α,β are the roots of the equation bz2+2az +b = 0
Product of the root αβ = b/b = 1 ⇒ | αβ| = 1 ⇒ |α||β| = 1
Given a > b ⇒ a2 > b2 ⇒ a2- b2>0⇒ √a2−b2 > 0 ⇒ a + √a2−b2 > a >b
⇒ a + √a2−b2 >b ⇒ a+√a2−b2
b > 1 ⇒ |a+√a2−b2
b | > 1
Hence z = β = −a+√a2−b2
b < 1 is the only pole lie inside the circle |z| = 1
There are no poles on the real axis
To fine the residue of f(z):
28
Res [f,β] = lim ¿ z→ βddz
( z−β )2 f (z ) = lim ¿ z→ β
ddz
( z−β )2 4 z
ib2( 2azb
+z2+1¿)2
= lim ¿ z→ βddz
( z−β )2 4 z
ib2¿¿¿¿ = lim ¿ z→ β
ddz
( z−β )2 4 z
ib2 ( z−α )2 ( z−β )2
= lim ¿ z→ βddz
4 z
i b2 ( z−α )2 = lim ¿ z→ β
4
ib2
ddz
z
( z−α )2 = lim ¿ z→ β
4
ib2 [ ( z−α )2−z 2 ( z−α )( z−α )4 ] =
= lim ¿ z→ β4
ib2(z−α )[ ( z−α )−z 2
( z−α )4 ] = lim ¿ z→ β4
ib2 [ ( z−α )−z2
( z−α )3 ] = 4
ib2 [ (β−α )−2β
(β−α )3 ]
= −4
i b2 [ α+β(β−α )3 ] =
−4
i b2 [ α+β(β−α )3 ] =
−4i b2 [ −a−√a2−b2
b+−a+√a2−b2
b
(−a+√a2−b2
b−−a−√a2−b2
b )3 ] =
−4i b2 [ −2a
b
(−a+√a2−b2
b+ a+√a2−b2
b )3 ]
=−4i b2 [ −2a
b
( 2√a2−b2
b )3 ] = [ 8ab3
8b3i (√a2−b2 )3 ] = a
i (√a2−b2 )3
Hence ∑ Res[f,zk] = Res[f,β] = a
i (√a2−b2 )3
∴∫C
f ( z )dz = 2πi∑ Res[f,zk] = 2πia
i (√a2−b2 )3 = 2πa
(√a2−b2 )3
⇒ I = ∫0
2πdθ
(a+bcosθ )2 =
2πa
(√a2−b2 )3
Type II.
29
Evaluation of the integral ∫−∞
∞
f ( x )dx where f(x) is a real rational function of the real
variable x.
If the rational function f(x) = g (x)h(x )
, then degree of h(x) exceeds that of g(x) and g(x) ≠ 0.To
find the value of the integral, by inventing a closed contour in the complex plane which includes the required integral. For this we have to close the contour by a very large semi-circle in the upper half-plane. Suppose we use the symbol “R” for the radius. The entire contour integral comprises the integral along the real axis from −Rto +Rtogether with the integral along the semi-circular arc. In the limit as R→∞the contribution from the straight line part approaches the required integral, while the curved section may in some cases vanish in the limit.
The poles z1,z2,….,zk of g (x)h(x )
, that lie in the upper half-plane
∫−∞
∞
f ( x )dx = ∫−∞
∞g(x )h (x)
dx = 2πi∑ Res[f,zk]
Example II.1
Using the residue of calculus compute ∫−∞
∞dx
(x2+1 ) ( x2+4 )
Solution
Consider the integral ∫C
f ( z )dz where f(z) = 1
( z2+1 ) ( z2+4 )
To find the poles of f(z) :
The poles of f(z) = zeros of (z2+1)(z2+4) , these zeros are given by (z2+1)(z2+4) = 0
⇒ z2+1 = 0 or z2+4 = 0
30
⇒ z2 = -1 = i2 ⇒ z = ± i
Or z2 = -4 = (2i)2⇒ z = ± 2i
Hence the poles of f(z) are ± i, ± 2i (all are simple poles)
And the poles z = i and z = 2i are the only poles lie inside the upper half of semi-circle.
There are no poles on the real axis
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
Now ∫C
f ( z )dz = ∫−R
R
f ( x )dx + ∫C R
f ( z )dz ------------- (a) ( on the real line –R to R (LR)+ the upper
half of the semi circle CR)
To find the residue of f(z):
Res[f,i] = lim ¿ z→i(z−i) f (z ) = lim ¿ z→i(z−i) 1
( z2+1 ) ( z2+4 ) = = lim ¿ z→i(z−i) 1
( z+i )¿¿¿
= lim ¿ z→i1
( z+i ) (z+2 i)( z−2i) =
1
(i+ i) (i+2i)(i−2i) =
1(2 i)(3 i)(−i) =
16 i
= -i6
Res[f,2i] = lim ¿ z→2i(z−2 i) f (z) = lim ¿ z→2i(z−2 i) 1
( z2+1 ) ( z2+4 ) = =
lim ¿ z→2i(z−2 i) 1
( z+ i)¿¿¿
= lim ¿ z→2i1
(z−i) ( z+i )(z+2i) =
1(2i−i )(i+2 i)(2 i+2i) =
1(i)(3i)(4 i) =
−112i
= i
12
31
Consider |∫CR
f ( z )dz| = |∫CR
1
( z2+1 ) ( z2+4 )dz|≤ ∫
C R
¿ dz
( z2+1 ) ( z2+4 )∨¿≤
∫C R
¿dz∨ ¿¿ ( z2+1 ) ( z2+4 )∨¿
¿¿
≤ ∫C R
¿dz∨ ¿¿¿ ¿¿ ---------(b)
Let z = Reiθ, dz = iReiθdθ⇒ |dz| = |iReiθdθ| = R dθ (∵ |i| =1 =|eiθ|)If z = -R, then Reiθ = -R ⇒ eiθ = -1⇒ θ = πIf z = R, then Reiθ = R ⇒ eiθ = 1⇒ θ = 0Hence (b) ⇒ |∫CR
f ( z )dz|≤ ∫0
πRdθ
(R2−1 ) (R2−4 ) =
R
(R2−1 ) (R2−4 ) ∫0π
dθ = Rπ
(R2−1 ) (R2−4 )
As R → ∞, Rπ
(R2−1 ) (R2−4 ) → 0 ⇒ ∫C R
f ( z )dz → 0
Now as R→ ∞, (a) ⇒ ∫C
f ( z )dz = ∫−∞
∞
f ( x )dx + 0 =∫−∞
∞
f ( x )dx
Where f(x) = 1
(x2+1 ) (x2+4 )
Hence ∫−∞
∞1
(x2+1 ) ( x2+4 )dx = ∫
−∞
∞
f ( x )dx = ∫C
f ( z )dz = 2πi∑ Res[f,zk]
= 2πi{Res[f,i] + Res[f,2i]} = 2πi{−i6
+ i
12 }=2πi(
−i12
¿= π6
Example II.2
Using the residue of calculus compute ∫−∞
∞dx
(x2+4 )3
Solution
32
Consider the integral ∫C
f ( z )dz where f(z) = 1
( z2+4 )3
To find the poles of f(z):
Poles of f(z) = zeros of (z2+4)3, these zeros are given by (z2+4)3 = 0
⇒ z2 +4 = 0 ⇒ z2 = -4 ⇒ z2 = (2i)2⇒ z = ±2i ⇒ z = 2i or z = -2i Hence the poles of f(z) are z = 2i , z = -2i , both of order 3
The only pole lie inside the upper half of the semi-circle is z = 2i of order 3
There are no poles on the real axis
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
Now ∫C
f ( z )dz = ∫−R
R
f ( x )dx + ∫C R
f ( z )dz ( on the real line –R to R (LR)+ the upper half of the semi
circle CR)
To find the residue of f(z):
Res[f,2i] = 1
(2 )!lim ¿ z→2 i
d2
d z2 ( z−2 i )3 f (z ) = 12
lim ¿ z→2 id2
d z2 ( z−2 i )3 1
( z2+4 )3
= 12
lim ¿ z→2 id2
d z2 ( z−2 i )3 1¿¿¿ ¿ =
12
lim ¿ z→2 id2
d z2 ( z−2 i )3 1
( z−2 i )3 ( z+2 i )3
= 12
lim ¿ z→2 id2
d z2
1
( z+2 i )3 =
12
lim ¿ z→2 idd z
dd z ( 1
( z+2 i )3 ) = 12
lim ¿ z→2 idd z (−3 ( z+2i )2
( z+2i )6 )
33
= 12
lim ¿ z→2 idd z ( −3
( z+2 i )4 ) = 12
lim ¿ z→2 i( 3 (4 ) ( z+2 i )3
( z+2i )8 ) = 122
lim ¿ z→2 i( 1
( z+2i )5 )= 6( 1
(2 i+2 i )5 ) = 6
(4 i )5 = 6
1024 i =
3
512i
Consider |∫CR
f ( z )dz| = |∫CR
1
( z2+4 )3dz|≤ ∫
C R
¿ dz
( z2+4 )3∨¿≤ ∫
C R
¿dz∨ ¿¿ ( z2+4 )3∨¿
¿¿
≤ ∫C R
¿dz∨ ¿¿¿¿ ¿¿ ---------(b)
Let z = Reiθ, dz = iReiθdθ⇒ |dz| = |iReiθdθ| = R dθ (∵ |i| =1 =|eiθ|)If z = -R, then Reiθ = -R ⇒ eiθ = -1⇒ θ = πIf z = R, then Reiθ = R ⇒ eiθ = 1⇒ θ = 0Hence (b) ⇒ |∫CR
f ( z )dz|≤ ∫0
πRdθ
(R2−4 )3 =
R
(R2−4 )3 ∫0
π
dθ = Rπ
(R2−4 )3
As R → ∞, Rπ
(R2−4 )3 → 0 ⇒ ∫C R
f ( z )dz → 0
Now as R→ ∞, (a) ⇒ ∫C
f ( z )dz = ∫−∞
∞
f ( x )dx + 0 =∫−∞
∞
f ( x )dx
Where f(x) = 1
(x2+4 )3
Hence ∫−∞
∞1
(x2+4 )3dx = ∫
−∞
∞
f ( x )dx = ∫C
f ( z )dz = 2πi∑ Res[f,zk]
= 2πi{Res[f,2i]} = 2πi{ 3
i512 }=
3π256
Example II.3
34
Prove that ∫−∞
∞ (x¿¿2−x+2)dx(x4+10 x2+9 )
¿ = 5π12
Solution
Consider the integral ∫C
f ( z )dz where f(z) = (z¿¿2−z+2)( z4+10 z2+9 )
¿
To find the poles of f(z):
Poles of f(z) = zeros of z4+10z2+9 , these zeros are given by z4+10z2+9 =0
⇒ z4+z2+9z2+9=0 ⇒ (z2+1)(z2+9) = 0
⇒ z2 = -1 = i2 or z2 = -9 = (3i)2
⇒ z = ±i or z = ±3iHence the poles of f(z) are i,-i,3i,-3i (all are simple poles)
The poles that are lying the upper half of the semi-circle are i,3i
There are no poles on the real axis
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
Now ∫C
f ( z )dz = ∫−R
R
f ( x )dx + ∫C R
f ( z )dz ( on the real line –R to R (LR)+ the upper half of the
semi-circle CR)
To find the residue of f(z):
35
Res[f,i] = lim ¿ z→i(z−i) f (z ) = lim ¿ z→i(z−i)(z¿¿2−z+2)( z4+10 z2+9 )
¿ =
lim ¿ z→i(z−i)(z¿¿2−z+2)
( z−i¿(z+ i)(z+3 i)(z−3 i)¿
= lim ¿ z→i(z¿¿2−z+2)
(z+ i)(z+3 i)(z−3 i)¿ =
(i¿¿2−i+2)(i+i)(i+3 i)(i−3 i)
¿ = 1−i
(2 i)(4 i)(−2 i) = 1−i1 6 i
Res[f,3i] = lim ¿ z→3i(z−3i) f (z ) = lim ¿ z→3i(z−3i)(z¿¿2−z+2)( z4+10 z2+9 )
¿
= lim ¿ z→3i(z−3i)(z¿¿2− z+2)
( z−i¿(z+i)(z+3i)(z−3 i)¿ = lim ¿ z→3i
(z¿¿2−z+2)( z−i¿(z+i)(z+3 i)
¿ =
((3 i)¿¿2−(3 i)+2)(3 i−i¿(3 i+ i)(3 i+3i)
¿
= −7−3 i
(2i¿(4 i)(6 i) = −(7+3i)−48 i
= 7+3 i48i
Consider |∫CR
f ( z )dz| = ¿≤ ∫C R
¿( z¿¿2−z+2)dz
( z4+10 z2+9 )∨¿¿≤
∫C R
¿(z¿¿2−z+2)∨¿dz∨ ¿¿ ( z4+10 z2+9 )∨¿
¿¿¿
≤ ∫C R
¿(z¿¿2−z+2)∨¿dz∨ ¿¿¿¿ ¿¿¿ ≤ ∫
C R
(¿ z∨¿2−¿ z∨+2)∨dz∨ ¿¿¿¿ ¿¿¿ ---------(b)
Let z = Reiθ, dz = iReiθdθ⇒ |dz| = |iReiθdθ| = R dθ (∵ |i| =1 =|eiθ|)If z = -R, then Reiθ = -R ⇒ eiθ = -1⇒ θ = πIf z = R, then Reiθ = R ⇒ eiθ = 1⇒ θ = 0Hence (b) ⇒ |∫CR
f ( z )dz|≤ ∫0
πR2dθ
(R2−1 )(R2−9) - ∫
0
πRdθ
(R2−1 )(R2−9) + ∫
0
π2dθ
(R2−1 )(R2−9) =
R2
(R2−1 )(R2−9)∫0
π
dθ - R
(R2−1 )(R2−9)∫0
π
dθ + 2
(R2−1 )(R2−9)∫0
π
dθ
36
= R2π
(R2−1 )(R2−9) -
Rπ
(R2−1 )(R2−9) +
2π
(R2−1 )(R2−9)
As R → ∞, R2π
(R2−1 )(R2−9) → 0 ,
Rπ
(R2−1 )(R2−9) → 0 and
2π
(R2−1 )(R2−9) → 0 ⇒ ∫
C R
f ( z )dz
→ 0
Now as R→ ∞, (a) ⇒ ∫C
f ( z )dz = ∫−∞
∞
f ( x )dx + 0 =∫−∞
∞
f ( x )dx
Where f(x) = (x¿¿2−x+2)(x4+10 x2+9 )
¿
Hence ∫−∞
∞ (x¿¿2−x+2)(x4+10x2+9 )
dx ¿ = ∫−∞
∞
f ( x )dx = ∫C
f ( z )dz = 2πi∑ Res[f,zk]
= 2πi {1−i1 6 i
+ 7+3 i48 i } = 2πi {3−3i+7+3i
48i } =10π24
= 5π12
Example II.4
Evaluate ∫0
∞dx
x4+a4
Solution
Let us take ∫−∞
∞dx
x4+a4
Consider the integral ∫C
f ( z )dz where f(z) = 1
z4+a4
To find the poles of f(z):
Poles of f(z) = zeros of z4+a4 , these zeros are given by z4+a4 =0
⇒ z4 = - a4 ⇒ z4 = a4eiπ (∵ eiπ = -1)
⇒ z4 = a4eiπ ei2nπ (∵ ei2nπ = 1)
⇒ z4 = a4eiπ+2nπi = a4ei(2n+1)π⇒ z = a ei(2n+1)π/4 , n = 0,1,2,3
37
If n=0, z = a eiπ/4 = a( cosπ4+i sin
π4
¿ = a( 1
√2+ i
√2¿ = α (say)
If n = 1, z = a ei3π/4 = a( cos3π4
+i sin3 π4
¿ = a( −1
√2+ i
√2¿ = β (say)
If n = 2, z = a ei5π/4 = a( cos5π4
+i sin5 π4
¿ = -a( 1
√2+ i
√2¿ = γ(say)
If n=3, z = a ei7π/4 = a( cos7 π4
+i sin7 π4
¿ = a( 1
√2− i
√2¿ = δ(say)
The poles lying inside the upper hemi circle are aeiπ/4 =α , aei3π/4 = β (both are simple poles)There are no poles on the real axis
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
Now ∫C
f ( z )dz = ∫−R
R
f ( x )dx + ∫C R
f ( z )dz ( on the real line –R to R (LR)+ the upper half of the
semi-circle CR)
To find the residue of f(z):
Res[f,α] =lim ¿ z→α (z−α) f (z ) = lim ¿ z→α (z−α) 1
z4+a4 It is difficult to solve while factoring 1
z4+a4 and taking limit, so we will use L’Hospital rule (that is differentiating Nr and Dr separately w.r.to z)= lim ¿ z→α
1
4 z3 = 1
4 α3 = α
4 α4 Now α = a eiπ/4 ⇒ α4 = a4eiπ ⇒ α4 =- a4 (∵eiπ = -1)
38
∴ Res [f,α] = −α
4 a4 = - ae iπ4
4a4 = - e iπ
4
4 a3
Now β = a ei3π/4 ⇒ β4 = a4ei3π ⇒ β4 =- a4 (∵ei3π = -1)Similarly , Res [f,β] = 1
4 β3 = β
4 β4 = - ae 3iπ4
4a4 = - e 3iπ
4
4 a3
Consider |∫CR
f ( z )dz| = |∫CR
1
z4+a4 dz|≤ ∫
C R
¿ dzz4+a4 ∨¿≤ ∫
C R
¿dz∨ ¿¿ z 4+a4∨¿
¿¿
≤ ∫C R
¿dz∨ ¿¿ z∨¿4−a4 ¿¿ ---------(b)
Let z = Reiθ, dz = iReiθdθ⇒ |dz| = |iReiθdθ| = R dθ (∵ |i| =1 =|eiθ|)If z = -R, then Reiθ = -R ⇒ eiθ = -1⇒ θ = πIf z = R, then Reiθ = R ⇒ eiθ = 1⇒ θ = 0Hence (b) ⇒ |∫CR
f ( z )dz|≤ ∫0
πRdθ
(R4−a4 ) ≤
R
(R4−a4 )∫0
π
dθ ≤ Rπ
(R4−a4 )
As R → ∞, Rπ
(R4−a4 )→ 0 ⇒ ∫C R
f ( z )dz → 0
Now as R→ ∞, (a) ⇒ ∫C
f ( z )dz = ∫−∞
∞
f ( x )dx + 0 =∫−∞
∞
f ( x )dx
Where f(x) = 1
(x4+a4 )
Hence ∫−∞
∞1
(x4+a4 )dx = ∫
−∞
∞
f ( x )dx = ∫C
f ( z )dz = 2πi∑ Res[f,zk]
= 2πi { - e iπ4
4 a3 - e i3π
4
4a3 } = - 2πi
4 a3 { e iπ4 +¿ e i3π
4 } = - πi
2a3 {( 1
√2+ i
√2¿ +( −1
√2+ i
√2¿ }
39
= 2π
2√2a3 = π
√2a3
We know that ∫−∞
∞1
(x4+a4 )dx = 2∫
0
∞1
(x4+a4 )d x
⇒ ∫0
∞1
(x4+a4 )dx =
12∫−∞
∞1
(x4+a4 )dx =
π
2√2a3
Type III.
Evaluation of the integral ∫−∞
∞
f ( x ) sinmxdx , ∫−∞
∞
f ( x ) cosmxdx where m > 0 and f(x) is a real
rational function of the real variable x.
If the rational function f(x) = g (x)h(x )
, then degree of h(x) exceeds that of g(x) and g(x) ≠ 0.
Let g(x) and h(x) be polynomials with real coefficients, of degree p and q, respectively, where q ≥ p+1.
If h(x) ≠0 for all real x, and m is a real number satisfying m > 0, then
∫−∞
∞g(x )h (x)
cos mxdx= limR→∞
∫−R
Rg (x)h(x )
cosmxdx and ∫−∞
∞g(x )h (x)
sinmx dx= limR→∞
∫−R
Rg (x)h(x )
sinmxdx
We know that Euler’s formula e imx = cos mx + i sin mx , where cos mx = Re[e imx]
and sin mx = Im[e imx] , m is a positive real.
We have ∫−∞
∞g(x )h (x)
eimx dx = ∫−∞
∞g(x )h (x)
cos mxdx + i ∫−∞
∞g(x )h (x)
sinmx dx
Here we are going to use the complex function f(z) = g (z)h(z )
eimz to evaluate the given integral.
∫−∞
∞g(x )h (x)
cos mxdx = Re {2πi ∑ Res[f,zk]} and
40
∫−∞
∞g(x )h (x)
sinmx dx = Im {2πi ∑ Res[f,zk]}, where z1,z2,…..zk are the poles lies on the upper half
of the semi-circle.
Lemma III.1.(Jordan’s Lemma)
If f (z) → 0 uniformly as z →∞, then limR→∞∫C1
e imz f (z )dz = 0, (m > 0) where C1 denotes the semi-
circle |z| = R, I(z) > 0.
Proof.
Given f (z) → 0 uniformly as z →∞⇒ given 𝜀 > 0 , ∃ a R0 > 0 such that | f(z) – 0| < 𝜀 , ∀ R ≥ R0 That is | f(z)| < 𝜀 , ∀ R ≥ R0 --------(a)Let |z| = R which is the semi-circle Put z = Reiθ ⇒ dz = R eiθ i dθ ⇒ dz = izdθ, 0≤ θ ≤ πNow e imz = eℑ eiθ = e imR(cosθ+isinθ) = e imRcosθ−mRsinθ = e imRcosθ e−mRsinθ
⇒ ¿e imR(cosθ+isinθ) | = ¿e imRcosθ∨¿ ¿e−mRsinθ∨¿ = e−mRsinθ ( ∵¿e imRcosθ∨¿ = 1) ----- (b)
We know that sinθθ
is monotonically decreases as θ increases from 0 to π2
.
If 0 ≤ θ ≤ π2
, then sin ( π
2)
π2
≤ sinθθ
41
⇒ 1π2
≤ sinθθ
⇒ 2π
≤ sinθθ
⇒ sinθ ≥ 2θπ
⇒ - 2θπ
≥ - sinθ⇒ -
mR2θπ
≥ - mRsinθ⇒ e−mR2θ
π ≥ e−mRsinθ ------ ( c)
From (a), (b) and (c),
|∫C1
e imz f ( z )dz| ≤ ∫C1
¿ eimz f ( z )dz∨¿
≤∫C1
¿ eimz∨|f ( z )|∨dz∨¿
≤ ε∫0
π
e−mRsinθ Rdθ ( ∵ |dz| = |izdθ| = |z|dθ = Rdθ) ≤ 2𝜀R ∫
0
π2
e−m Rsinθ
dθ
≤ 2𝜀R ∫0
π2
e−mR2θ
π dθ
≤ 2𝜀R [ e−mR2θ
π
−2mRπ
] π20
≤ 2𝜀R{(e¿¿−mR ) π−2mR
¿ - (e0) π−2mR
} ≤ 2 εRπ
2mR{¿¿
≤ επm
{−e−mR+1 } ≤ επm
(∵{−e−mR+1 }<1¿
42
|∫C1
e imz f ( z )dz| ≤ επm
= 𝜀’ (say)As lim R → ∞ , |∫C1
e imz f ( z )dz−0| ≤ 𝜀’
⇒ ∫C1
eimz f ( z )dz → 0 as R → ∞Example III.1.Use the method of contour integration to prove that ∫
0
∞cosmx
(x2+a2 )dx =
π e−ma
2a and
∫0
∞sinmx
(x2+a2 )dx = 0
Solution.
Consider the integral ∫C
f ( z )dz where f(z) = eimz
( z2+a2 )
To find the poles of f(z):
Poles of f(z) = zeros of z2+a2 , these zeros are given by z2+a2 =0
⇒ z2 = - a2 ⇒ z = (ai)2 ⇒ z = ± ai ⇒ z = ai or z = -ai
Poles of f (z) are ai, -ai (both are simple poles)There are no poles on the real axis
The only pole lie inside the upper half of semi-circle is z = ai
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
43
Now ∫C
f ( z )dz = ∫−R
R
f ( x )dx + ∫C R
f ( z )dz ( on the real line –R to R (LR)+ the upper half of the
semi-circle CR).
To find the residue of f(z):
Res[f,ai] = lim ¿ z→ai(z−ai) f (z ) = lim ¿ z→ai(z−ai) eimz
( z2+a2 ) =
lim ¿ z→ai(z−ai) eimz
(z−ai¿(z+ai)
= lim ¿ z→aie imz
(z+ai) =
e imai
(ai+ai) =e−ma
2ai
Now lim ¿ z→∞1
( z2+a2 ) = 0
∴ by Jordan’s lemma , lim ¿R→∞∫CR
eimz
( z2+a2 )dz = 0
As R → ∞ ,
∫C
f ( z )dz = ∫−∞
∞
f ( x )dx + lim ¿R→∞∫CR
eimz
( z2+a2 )dz = ∫
−∞
∞eimx
(x2+a2 )dx +0
= ∫−∞
∞e imx
(x2+a2 )dx
Hence ∫−∞
∞e imx
(x2+a2 )dx = ∫
C
f ( z )dz = 2πi∑ Res[f,zk] = 2πi e−ma
2ai = π e
−ma
a
⇒ ∫−∞
∞cosmx+isinmx
(x2+a2 )dx =
π e−ma
a
⇒ ∫−∞
∞cosmx
(x2+a2 )dx + i∫
−∞
∞sinmx
(x2+a2 )dx =
π e−ma
a
Equating real and imaginary parts,
∫−∞
∞cosmx
(x2+a2 )dx = π e
−ma
a and ∫
−∞
∞sinmx
(x2+a2 )dx = 0
44
⇒ ∫0
∞cosmx
(x2+a2 )dx =
12∫−∞
∞cosmx
(x2+a2 )dx =
π e−ma
2a
Example III.2.Apply the calculus of residue to evaluate ∫
−∞
∞cosx
(x2+a2 )(x2+b2)dx , (a > b > 0)
Solution.Consider the integral ∫
C
f ( z )dz where f(z) = e iz
( z2+a2 )(z2+b2)
To find the poles of f(z):
Poles of f(z) = zeros of (z2+a2) (z2+b2), these zeros are given by (z2+a2) (z2+b2) =0
⇒ z2 = - a2 or z2
= - b2 ⇒ z = (ai)2 or z = (bi)2 ⇒ z = ± ai or z = ± bi ⇒ z = ai or z = -ai or z =bi or z = -biPoles of f (z) are ai,-ai, bi,-bi (all are simple poles)There are no poles on the real axis.
The poles lie inside the upper half of semi-circle are z = ai , z = bi
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
Now ∫C
f ( z )dz = ∫−R
R
f ( x )dx + ∫C R
f ( z )dz ( on the real line –R to R (LR)+ the upper half of the
semi-circle CR).
To find the residue of f(z):
45
Res[f,ai] = lim ¿ z→ai(z−ai) f (z ) = lim ¿ z→ai (z−ai ) e iz
( z2+a2 )(z2+b2)
=lim ¿ z→ai(z−ai) e iz
( z+ai )(z−ai)(z+bi)(z−bi) = lim ¿ z→ai
e iz
( z+ai )(z+bi)(z−bi)
= e iai
(ai+ai )(ai+bi)(ai−bi) =
−e−a
2ai(a2−b2) =
e−a
2ai(b2−a2)
Similarly,
Res[f,bi] = lim ¿ z→bi(z−bi) f (z) = e−b
2bi(a2−b2)
Now lim ¿ z→∞ 1
( z2+a2 )(z2+b2) = 0
By Jordan’s Lemma,
lim ¿R→∞∫CR
eiz
( z2+a2 )(z2+b2)dz = 0
As R → ∞ ,
∫C
f ( z )dz = ∫−∞
∞
f ( x )dx + lim ¿R→∞∫CR
eiz
( z2+a2 ) ( z2+b2 )dz = ∫
−∞
∞eix
(x2+a2 ) (x2+b2 )dx +0
= ∫−∞
∞eix
(x2+a2 ) (x2+b2 )dx
Hence ∫−∞
∞eix
(x2+a2 ) (x2+b2 )dx = ∫
C
f ( z )dz = 2πi∑ Res[f,zk] = 2πi ¿ +e−b
2bi(a2−b2)¿
= 2πi(be−a−ae−b)
2abi(b2−a2) =
π (be−a−ae−b)ab(b2−a2)
⇒ ∫−∞
∞cosx+isinx
(x2+a2 ) (x2+b2 )dx =
π (be−a−ae−b)ab(b2−a2)
Equating real and imaginary parts,
∫−∞
∞cosx
(x2+a2 ) (x2+b2 )dx =
π (be−a−ae−b)ab(b2−a2)
and ∫−∞
∞sinx
(x2+a2 ) (x2+b2 )dx = 0
46
Example III.3
Evaluate ∫−∞
∞xcosxx2+4
dx and ∫−∞
∞xsinxx2+4
dx
Solution
Consider the integral ∫C
f ( z )dz where f(z) = z eiz
z2+4
To find the poles of f(z):
Poles of f(z) = zeros of (z2+4), these zeros are given by (z2+ 4) =0
⇒ z2 = - 22 ⇒ z = (2i)2 ⇒ z = ± 2i
⇒ z = 2i or z = -2i Poles of f (z) are 2i,-2i (both are simple poles)There are no poles on the real axis.
The only pole lie inside the upper half of semi-circle is z = 2i
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
Now ∫C
f ( z )dz = ∫−R
R
f ( x )dx + ∫C R
f ( z )dz ( on the real line –R to R (LR)+ the upper half of the
semi-circle CR).
To find the residue of f(z):
Res[f,2i] = lim ¿ z→2i(z−2 i) f (z) = lim ¿ z→2i(z−2 i) zeiz
z2+4 =
lim ¿ z→2i(z−2 i) zeiz
(z+2 i)(z−2 i)
47
=lim ¿ z→2izeiz
(z+2i) =
2 ie i2i
(2 i+2i) = e
−2
2
Now lim ¿ z→∞ z
( z2+4 ) = 0
By Jordan’s Lemma,
lim ¿R→∞∫CR
zeiz
( z2+4 )dz = 0
As R → ∞ ,
∫C
f ( z )dz = ∫−∞
∞
f ( x )dx + lim ¿R→∞∫CR
z eiz
( z2+4 )dz = ∫
−∞
∞x eix
(x2+4 )dx +0
Hence ∫−∞
∞xe ix
(x2+4 )dx = ∫
C
f ( z )dz = 2πi∑ Res[f,zk] = 2πi ¿ ¿ = e-2πi ⇒ ∫
−∞
∞xcosx+ixsinx
(x2+4 )dx = e-2πi
Equating real and imaginary parts,
∫−∞
∞xcosx
(x2+4 )dx = 0and ∫
−∞
∞sinx
(x2+4 )dx = e-2π
Example III.4
Evaluate ∫0
∞cosmx
x4+x2+1dx (m > 0)
Solution
Consider the integral ∫C
f ( z )dz where f(z) = e imz
z4+z2+1
To find the poles of f(z):
Poles of f(z) = zeros of (z4+z2+1), these zeros are given by z4+z2+1=0
z4+- z2 + z2+ z2 +1=0 ⇒z2 (z2 +1) + z2 +1 = z2
⇒ (z2 +1)2 - z2 = 0 ⇒ (z2 +1-z)( z2 +1+z) = 0
48
⇒ z2 +1-z = 0 or z2 +1+z = 0
⇒ z = 1+√3 i2
,or z = 1−√ 3i
2, or z =
−1+√3 i2
,or z = −1−√ 3i
2
Poles of f (z) are1+√3 i2
,1−√ 3i2
, −1+√3 i
2 , −1−√ 3i
2 (all are simple poles)
There are no poles on the real axis.
The poles lie inside the upper half of semi-circle is z = 1+√3 i
2 = α (say) and z = −1+√3 i
2
= β(say)
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
Now ∫C
f ( z )dz = ∫−R
R
f ( x )dx + ∫C R
f ( z )dz (on the real line –R to R (LR) + the upper half of the
semi-circle CR).
To find the residue of f(z):
Res[f,α] = lim ¿ z→α (z−α) f (z ) = lim ¿ z→α (z−α) e imz
(z−α )(z−β)(z+α )(z+β ) =
lim ¿ z→αe imz
(z−β )(z+α)(z+β)
= eimα
(α−β )(α+α )(α+β) =
e imα
(α−β )(2α)(α+β) =
e imα
(α−β )(2α)(α+β) =
eℑ (1+√3 i)
2
[ 1+√3 i2
−−1+√3 i2
]¿¿
49
=e
ℑ (1+√3i )2
(2 ( 1+√3 i )
2)( 2√3 i
2 ) = e
ℑ (1+√3 i )2
(√3 i−3) = e
ℑ2 e
−√ 3m2
(√3 i−3)
Similary,
Res[f,β] = e
ℑ (−1+√3 i)2
[(−1+√3 i )
2−1+√3 i
2] (
2 (−1+√3 i )2
)[(1+√3 i )
2+
(−1+√3 i)2
] =
eℑ (−1+√3 i)
2
(−1 ) (2 (−1+√3 i) )2
(2√3i )
eℑ (−1+√3i )
2
(−1)(2 (−1+√3 i )
2)( 2√3 i
2 ) = e
ℑ(−1+√3 i )2
(√3 i+3) = e
−ℑ2 e
−√ 3m2
(√3 i+3)
Now lim ¿ z→∞ 1
( z4+z2+1 ) = 0
By Jordan’s Lemma,
lim ¿R→∞∫CR
e imz
( z4+z2+1 )dz = 0
As R → ∞ ,
∫C
f ( z )dz = ∫−∞
∞
f ( x )dx + lim ¿R→∞∫CR
e imz
( z4+z2+1 )dz = ∫
−∞
∞e imx
(x4+x2+1 )dx +0
Hence ∫−∞
∞e imx
(x4+x2+1 )dx = ∫
C
f ( z )dz = 2πi∑ Res[f,zk] = 2πi [ eℑ2 e
−√3m2
(√3 i−3 )+ e
−ℑ2 e
−√3m2
(√3 i+3 ) ] = 2πi [ e−√3m
2 {(3+√3 i)eℑ2 +(√3i−3 ) e
−ℑ2 }
(√3i−3 ) (√3 i+3 ) ] =
2πi [ e−√3m2 {(3+√3 i )(cos (m2 )+ isin(m2 ))+(√3 i−3 ) (cos (m2 )−isin(m2 ))}
(−3−9 ) ]
50
= 2πi [ e−√3m
2
−12
{(3+√3 i )(cos (m2 )+ isin(m2 ))+(√3 i−3 ) (cos (m2 )−isin(m2 ))}]=
2πi [ e−√3m2
−12
{2√3 i cos(m2 )+6 isin (m2 )} ] = −4πe
−√ 3m2 ¿¿ = πe
−√3m2 ¿¿
⇒ ∫−∞
∞cosmx+isinmx
(x4+x2+1 )dx = πe
−√3m2 ¿¿
Equating real and imaginary parts,
∫−∞
∞cosmx
(x4+x2+1 )dx = πe
−√3m2 ¿¿
and ∫−∞
∞sinmx
(x4+x2+1 )dx = 0
Hence ∫0
∞cosmx
(x4+x2+1 )dx =
12∫−∞
∞cosmx
(x4+x2+1 )dx = πe
−√3m2 ¿¿
Example III.5
Prove that ∫0
∞cosmx
(a2+x2 )2dx= π
2a3 (1+ma) e−ma (m > 0, a > 0)
Solution
Consider the integral ∫C
f ( z )dz where f(z) = e imz
(a2+z2 )2
To find the poles of f(z):
Poles of f(z) = zeros of (z2+a2 )2, these zeros are given by (z2+a2 )2=0
⇒ z2 = - a2 ⇒ z = (ai)2 ⇒ z = ± ai ⇒ z = ai (twice)or z = -ai(twice)
Poles of f (z) are ai, -ai (both are order 2)
51
There are no poles on the real axis
The only pole lie inside the upper half of semi-circle is z = ai (order 2)
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
Now ∫C
f ( z )dz = ∫−R
R
f ( x )dx + ∫C R
f ( z )dz ( on the real line –R to R (LR)+ the upper half of the
semi-circle CR).
To find the residue of f(z):
Res[f,ai] = lim ¿ z→aiddz
( z−ai )2 f (z) = lim ¿ z→aiddz
( z−ai )2 e imz
(a2+z2 )2
= lim ¿ z→aiddz
( z−ai )2 eimz
( z+ai )2 ( z−ai )2= lim ¿ z→ai
ddz
eimz
( z+ai )2 =
lim ¿ z→ai[ ( z+ai )2 ℑe imz−eimz 2 ( z+ai ) ]
( z+ai )4
= lim ¿ z→ai[ ( z+ai ) ℑe imz−eimz 2 ]
( z+ai )3 =
[ (ai+ai ) ℑe imai−eimai2 ](ai+ai )3
= [−2ame−ma−2e−ma ]
(2ai )3 =
−2e−ma (ma+1 )−8a3i
= e−ma (ma+1 )
4a3 i
Now lim ¿ z→∞1
( z2+a2 )2 = 0
52
∴ by Jordan’s lemma , lim ¿R→∞∫CR
e imz
( z2+a2 )2dz = 0
As R → ∞ ,
∫C
f ( z )dz = ∫−∞
∞
f ( x )dx + lim ¿R→∞∫CR
e imz
( z2+a2 )2dz = ∫
−∞
∞eimx
(x2+a2 )2dx +0
= ∫−∞
∞eimx
(x2+a2 )2dx
Hence ∫−∞
∞eimx
(x2+a2 )2dx = ∫
C
f ( z )dz = 2πi∑ Res[f,zk] = 2πi e−ma (ma+1 )
4a3 i= π e−ma(ma+1)
2a3
⇒ ∫−∞
∞cosmx+isinmx
( x2+a2 )2dx =
π e−ma(ma+1)2a3
⇒ ∫−∞
∞cosmx
(x2+a2 )2dx + i∫
−∞
∞sinmx
(x2+a2 )2dx =
π e−ma(ma+1)2a3
Equating real and imaginary parts,
∫−∞
∞cosmx
(x2+a2 )2dx =
π e−ma(ma+1)2a3 and ∫
−∞
∞sinmx
(x2+a2 )2dx = 0
⇒ ∫0
∞cosmx
(x2+a2 )2dx =
12∫−∞
∞cosmx
(x2+a2 )2 dx = π e−ma(ma+1)
4 a3
Note:III.1
z = reiθ ⇒ r = |z| and θ= arg (z)
logz = Log r+i arg(z)
If z = x+iy , r = (x2+y2)1/2 , θ= arg (z) = arg (x+iy) = tan-1(y/x)
log(x+i) = log (x2+1)1/2 + iarg(x) = log (x2+1)1/2 +0 = log (x2+1)1/2
Example III.6
Prove that ∫0
∞log (1+x2)
1+ x2 dx=πlog 2
53
Solution
Consider the integral ∫C
f ( z )dz where f(z) = log (z+i)z2+1
To find the poles of f(z):
Poles of f(z) = zeros of (z2+1 ), these zeros are given by (z2+1 )=0
⇒ z2 = - 1 ⇒ z = (i)2 ⇒ z = ± i ⇒ z = i or z = -i
Poles of f (z) are i, -i (both are simple poles)There are no poles on the real axis.
The only pole lie inside the upper half of semi-circle is z = i
By Cauchy’s residue theorem, ∫C
f ( z )dz = 2πi∑ Res[f,zk] where zk are the singularities(poles) of
f(z).
Now ∫C
f ( z )dz = ∫−R
R
f ( x )dx + ∫C R
f ( z )dz ( on the real line –R to R (LR)+ the upper half of the
semi-circle CR).
To find the residue of f(z):
Res[f,i] = lim ¿ z→i(z−i) f (z ) =lim ¿ z→i(z−i)log (z+i)z2+1
= lim ¿ z→i(z−i)log (z+i)
(z−i)(z+i)
= lim ¿ z→ilog (z+i)
(z+i) = log (i+i)
(i+i) = log 2 i
2i =
log (22 )12+ i tan−1( 2
0)
2 i =
log 2+itan−1(∞)2 i
(using Note III.1)
= log 2+ iπ
22i
54
Now lim ¿ z→∞log(z+i)z2+1
= lim ¿ z→∞log(z+i)
(z+i)(z−i) = lim ¿ z→∞1
(z−i)
lim ¿ z→∞log(z+i)(z+i)
Consider lim ¿ z→∞1
(z−i) = 0
Consider lim ¿ z→∞log (z+i)
( z+i) it is undetermined, so we have to use L’Hospital’s rule
lim ¿ z→∞
1z+i1
= lim ¿ z→∞1z+ i
= 0
Hence lim ¿ z→∞log(z+i)z2+1
= 0
⇒ lim ¿ z→∞∫C R
log(z+i)z2+1
= 0 ⇒ lim ¿R→∞∫CR
log (z+i)z2+1
= 0 (∵ |z| = R)
As R → ∞ ,
∫C
f ( z )dz = ∫−∞
∞
f ( x )dx + lim ¿R→∞∫CR
log (z+i)z2+1
dz = ∫−∞
∞log (x+i)x2+1
dx +0
= ∫−∞
∞log (x2+1 )
12
x2+1dx = ∫
−∞
∞12
log (x2+1)
x2+1dx (By using the Note III.1)
= 12∫−∞
∞log (x2+1)
x2+1dx = ∫
0
∞log (x2+1)
x2+1dx
Hence ∫0
∞log (x2+1)
x2+1dx = ∫
C
f ( z )dz = 2πi∑ Res[f,zk] = 2πi log 2+ iπ
22i
=πlog2 +iπ 2
2
Equating real part ∫0
∞log (x2+1)
x2+1dx = πlog2
Case of poles are on the real axis.
55
Type IV
If the rational function f(z) = g (z)h(z ) , then degree of h(z) exceeds that of g(z) and g(z) ≠ 0.
Suppose h(z) has simple zeros on the real axis ( that is simple poles of f(z) on the real axis) , let it be a1,a2,…ak
and h(z) has zeros inside the upper half of semi-circle ( that is poles of f(z) inside the upper half of semi-circle), let it be b1,b2,…bs,
then ∫−∞
∞
f ( x )dx = πi∑ Res[f,ak] + 2πi∑ Res[f,bs] , where k = 1,2,….k and s = 1,2,…s
Where C1,C2,….Ck are the semi circles and b1,b2,…bs are lie upper half of these semi circles.
Example IV.1.
Evaluate ∫−∞
∞x
x3−8dx
Solution.
Consider the integral ∫C
f ( z )dz where f(z) = z
z3−8
To find the poles of f(z):
Poles of f(z) = zeros of (z3- 8 ), these zeros are given by (z3- 8)=0
⇒ z3 = 8 ⇒ z3 = (2)3 ⇒ z = 2
Since z-2 is a factor of z3- 8 , z3- 8 = (z-2)(z2+2z+4) = 0
⇒ z2+2z+4 = 0 ⇒ z = −2±√4−16
2 =
−2±√−4∗32
= −2±2 i√3
2 = -1±i√3
Poles of f (z) are 2 , -1+i√3 and -1-i√3 (all are simple poles)
56
Pole lie on the real axis z =2
Pole lie inside the upper half of semi-circle z = -1+i√3
To find the residue of f(z):
Res[f,2] = lim ¿ z→2(z−2) f (z) = lim ¿ z→2(z−2) z
z3−8 =
lim ¿ z→2(z−2) z(z−2)(z+1−i √3)(z+1+i √3)
= lim ¿ z→2z
(z+1−i √3)( z+1+i √3) =
2(2+1−i√3)(2+1+ i√3)
= 2
(3−i√3)(3+i √3)
=2
(9+3) = 16
Res[f, -1+i√3] = lim ¿ z→−1+i √3 (z−(−1+ i√3)) f (z )
= lim ¿ z→−1+i √3 (z+1−i √3) z
z3−8 =
lim ¿ z→−1+i √3 (z+1−i √3) z(z−2)(z+1−i√3)(z+1+i √3)
=lim ¿ z→−1+i √3z
(z−2)(z+1+i √3) =
−1+i √3(−1+i √3−2)(−1+ i√ 3+1+i√3)
=
−1+ i√ 3(−3+i √3)(2 i √3) =
−1+i √3(−6 i √3−6)
=−1+i √3
−6(i √3+1) =
(−1+i √3)(1−i√3)−6(i √3+1)(1−i√3)
= −(−1+i √3)(−1+i √3)
−6(1+3) =
(1−3−2 i√3)24
=(−2−2i √3)
24
= −2(1+i √3)
24 =
−(1+i √3)12
57
We know that ∫−∞
∞
f ( x )dx = πi∑ Res[f,ak] + 2πi∑ Res[f,bs] where ak ‘s are the poles lie on real
axis and bs ‘s are the poles lie inside the upper half of semi-circle.
∫−∞
∞x
x3−8dx = πi∑ Res[f,ak] + 2πi∑ Res[f,bs] = πi(
16¿ + 2πi(
−(1+i √3)12
¿
= πi6
- πi6
-πi √3 i
6 =
√3 π6
Type V
If the rational function f(z) = g (z)h(z )
, then degree of h(z) exceeds that of g(z) and g(z) ≠ 0.
Suppose h(z) has simple zeros on the real axis ( that is simple poles of f(z) on the real axis) , let it be a1,a2,…ak
and h(z) has zeros inside the upper half of semi-circle ( that is poles of f(z) inside the upper half of semi-circle), let it be b1,b2,…bs,
Let m be a positive real number and if f(z) = eimz g (z)h(z)
, then
∫−∞
∞
cosmxg ( x )h ( x )
dx = Re ∫−∞
∞
cosmx f ( x )dx
= ℜ[2 πi∑i=1
s
Res [ f , b i ]] + ℜ[πi∑j=1
k
Res [ f , aj ]]And
∫−∞
∞
sinmxg ( x )h (x )
dx = Img ∫−∞
∞
sinmx f ( x )dx
= Img[2πi∑i=1
s
Res [ f , bi ]] + Img[ πi∑j=1
k
Res [ f , aj ] ] Where b1,b2,…bs, are the poles of f(z) that lie in the upper half of the semi-circles
C1,C2,….Ck .
Example V.1.
58
Prove that ∫−∞
∞cos x
(x−1)(x2+4 )dx= π
10(−1e2 −2 sin 1) and
∫−∞
∞sinx
(x−1)(x2+4 )dx=π
5(−1e2 +cos1)
Solution.
Consider the integral ∫C
f ( z )dz where f(z) = e iz
(z−1)(z2+4)
To find the poles of f(z):
Poles of f(z) = zeros of (z-1)(z2+4), these zeros are given by (z-1)(z2+4) =0
⇒ z-1 = 0 or z2+4 =0 ⇒ z = 1 or z2 = -4 = (2i)2
⇒ z = 1 or z = 2i or z = -2iPoles of f(z) are z = 1 , z = 2i , z = -2i (all are simple poles)
The only pole lie on the real axis is z = 1
The only pole lie inside the semi-circle is z = 2i
To find the residue of f(z):
Res[f,1] = lim ¿ z→1(z−1) f (z) = lim ¿ z→1(z−1) e iz
(z−1)(z2+4) = lim ¿ z→1
eiz
(z2+4 )
= e i
(1+4) = e
i
5
Res[f,2i] = lim ¿ z→2i(z−2 i) f (z) = lim ¿ z→2i(z−2 i) eiz
(z−1)(z+2i)(z−2 i)
59
= lim ¿ z→2ie iz
(z−1)(z+2 i) =
e i2 i
(2 i−1)(2i+2 i) =
e−2
(2 i−1)(4 i)
We know that
∫−∞
∞
cosmxg ( x )h ( x )
dx =
ℜ∫−∞
∞
cosmx f ( x )dx = ℜ[2 πi∑i=1
s
Res [ f , b i ]] + ℜ[πi∑j=1
k
Res [ f , aj ]]And ∫
−∞
∞
sinmxg ( x )h ( x )
dx =
Img∫−∞
∞
sinmx f ( x )dx = Img[2πi∑i=1
s
Res [ f , bi ]] + Img[ πi∑j=1
k
Res [ f , aj ] ]Where ak’s are the poles lie on the real axis and bs’s are the poles lie inside the upper half of the semi-circle.
∫−∞
∞cosx
(x−1)(x2+4 )dx = Re ∫
−∞
∞
cos x f ( x )dx = Re ∫−∞
∞
cos xe i x
(x−1)(x2+4)dx = ℜ [2πi Res [ f ,2i ] ]
+ ℜ [ πi Res [ f ,1 ] ]
= ℜ[2 πie−2
(2 i−1)(4 i) ] + ℜ[πi e i
5 ]= ℜ[π e−2
(2 i−1)(2) ] + ℜ[πi( cos1+isin15
)]= ℜ[π e−2(2 i+1)
2(2i−1)(2 i+1) ] + ℜ[ πicos1−π sin 15 ] = ℜ[π e−2(2i+1)
2(−4−1) ] - πsin15
= ℜ[π e−2(2i+1)−10 ] -
πsin15
= [ π e−2
−10 ] - πsin15
= −π10
[e−2+2 sin1 ]
∫−∞
∞sin x
(x−1)(x2+4 )dx = Img ∫
−∞
∞
sinx f ( x )dx = Img ∫−∞
∞
cosxe ix
(x−1)(x2+4 )dx
60
= Img[2πi∑i=1
s
Res [ f , bi ]] + Img[ πi∑j=1
k
Res [ f , aj ] ] = Img[π e−2(2 i+1)
2(2 i−1)(2i+1) ] + Img[ πicos1−πsin15 ]
= Img[ π e−2(2 i+1)−10 ] +
π cos15
= πe−2 2−10
+ πcos 1
5 = π
e−2
−5 +
πcos 15
= π5
[−e−2+cos1 ]
Example V.2.
Prove that ∫0
∞sinm x
xdx=π
2
Solution.
Consider the integral ∫C
f ( z )dz where f(z) = eim z
z
To find the poles of f(z):
Poles of f(z) = zeros of (z), this zero is given by z =0
The only pole of f(z) is z = 0 simple and lie on real axis
To find the residue of f(z):
Res[f,0] = lim ¿ z→0(z−0) f (z) = lim ¿ z→0 zeiz m
z = e0 = 1
We know that ∫−∞
∞
cosmxg ( x )h ( x )
dx =
ℜ∫−∞
∞
cosmx f ( x )dx = ℜ[2 πi∑i=1
s
Res [ f , b i ]] + ℜ[πi∑j=1
k
Res [ f , aj ]]
61
And ∫−∞
∞
sinmxg ( x )h ( x )
dx =
Img∫−∞
∞
sinmx f ( x )dx = Img[2πi∑i=1
s
Res [ f , bi ]] + Img[ πi∑j=1
k
Res [ f , aj ] ]Where ak’s are the poles lie on the real axis and bs’s are the poles lie inside the upper half of the semi-circle.
∴ ∫−∞
∞sinmx
xdx = Img∫
−∞
∞
sinmx f ( x )dx = Img∫−∞
∞
sinmxe imx
xdx
= Img[2πi∑i=1
s
Res [ f , bi ]] + Img[ πi∑j=1
k
Res [ f , aj ] ]= Img [πi Res [ f ,0 ] ] = Img [πi(1)] = πHence ∫
0
∞sinmx
xdx =
12∫−∞
∞sinmxx
dx = π2