Answers (Hw 14)

Math 321
By Lei Li

3.2
1. #4 Solution: 1 = 1x Replace x with x , and we get: 1 = 1 + x2
+ + 2 +

xn , |x| < 1
n=0

(x ) =
n=0 n=0

2 n

(1)n x2n

To nd the radius of convergence, we apply Ratio Test here:

n

lim |

an+1 | = lim | x2 | = x2 n an

We require |ratio| < 1 and thus |x| < 1, which implies the radius is 1. To understand this, we consider f (z ) =
1 z 2 +1

(1)n z 2n . You can see that it has

singularities at i. The series fails at these sigularity points. The radius must be less than or equal to the distance between the center and the closest sigularity. In our case, its just equal to this distance, which is true in most cases. #7: Solution: Apply the generalized Cauchys formula: f (z ) 2if (n) (a) dz = (z a)n+1 n!

n=0

if a is inside the contour here. In our case, you need to set a = 0 and replace n with n 1. The rst integral is thus: 2i dn1 I1 = cos z |z=0 (n 1)! dz n1

The second integral is I2 = Apply the rule

d cos z dz

2i dn1 sin z |z=0 (n 1)! dz n1

d sin z dz

= sin z and

= cos z : n even n = 4k + 1 n = 4k + 3 n odd n = 4k + 2 n = 4k

0
2i (n1)! i (n2 1)!

I1 =

and I2 = 0
2i (n1)! i (n2 1)!

1 2. Calculate |z|=4 (z+1)(z dz without calculating the partial fraction expressions. 1)(z +2i) Solution: To nd the singularities, we can set (z + 1)(z 1)(z + 2i) = 0. The singularities are 1, 1, 2i which are all inside our contour.

Denote the integral as I . We then deform our contour to the three small circles around these singular points (C1 : |z + 1| = 1 , C2 : |z 1| = 2 , C3 : |z + 2i| = 3 ) and we have: I = = + = = = = 1 dz |z |=4 (z + 1)(z 1)(z + 2i) 1 1 dz + dz C2 (z + 1)(z 1)(z + 2i) C1 (z + 1)(z 1)(z + 2i) 1 dz C3 (z + 1)(z 1)(z + 2i) 1/(z 1)(z + 2i) 1/(z + 1)(z + 2i) 1/(z + 1)(z 1) dz + dz + dz z+1 z1 z + 2i C2 C3 C1 1 1 1 2i( + + ) 2(1 + 2i) 2(1 + 2i) (2i + 1)(2i 1) (2 + i) (2 + i) 2i + 5 5 5 0

by Cauchys formula. This is not surprising, because all simple poles are inside our contour. (Sometimes, the innity is also a singularity and in those cases, the integral is not zero) 3. Calculate
1 dz |z |=2 z (z +3)2

and

1 dz |z |=2 z 2 (z +3)

Solution: Denote the rst as I1 and the second as I2 . By Cauchys formula: 1 dz = z (z + 3)2
1 (z +3)2

I1 =
|z |=2

|z |=2

1 1/(z + 3)2 dz = 2i 2 z 3

because the function f (z ) =

is quite good inside our contour.

I2 =
|z |=2

z 2 (z

1 dz = + 3)

|z |=2

1/(z + 3) 1 d 1 |z=0 dz = 2i 2 z 1! dz z + 3

2i 9

Also this is true because the numerator is good inside our contour.
e dz 4. Calculate |z|=3 (z 2)3 Solution: Apply Cauchys formula directly:
z

2i

1 d2 z e |z=2 = e2 i 2! dz 2

z 5. Calculate |z|=2 zsin 2 +1 dz Set z 2 + 1 = 0, and we get z = i. We get two small circles: C1 : |z i| = C2 : |z + i| = 2

and

I = = = = =

sin z dz 2 |z |=2 z + 1 sin z sin z dz + dz 2 2 C1 z + 1 C2 z + 1 sin z/(z + i) sin z/(z i) dz + dz zi z+i C1 C2 sin(i) sin i 2i + 2i 2i 2i sin i sin(i) = 2 sin i

6. (*)(Challenging problems) ez (This one could be really hard for you)Calculate |z|=1/2 sin(2 dz z) Solution: Lets look at the singular points. We set sin(2z ) = 0. First of all, we should know iw eiw where sin w = 0 in complex plane. sin w = e = 0 and we get ei2w = 1. Let 2i w = x + iy and we have ei2x e2y = 1. We take magnitue and we can see e2y = 1 we must have y = 0 since y is real. Then, we have e2ix = 1 for real x. This means 3

2x = 2k and x = k . We conclude that sin w = 0 only if w = k even if w can be complex. This means all of these zeros are real. Then 2z = k and z = k/2. Inside our contour, we only have z = 0 this singularity. Around this point, we can see sin(2z ) = 2z (2z )3 /3! + . . . and we can see, its a simple pole, since we can have z (2 8z 2 /6 + . . .) where 2 8z 2 /6 + . . . is really a good function. Then, we have the integrand as formula, we have:
g (z ) z

where g (z ) =

ez z sin(2z )

ez . 28z 2 /6+...

By Cauchys

2ig (0) = 2i

e0 = i 2 0 + ...

4
1 1. Calculate 0 2sin d(Hint: On unit circle, sin = z2zi ) Solution: On unit circle, z = ei , 0 z < 2 i ei by Eulers identity. Then, sin = e 2i dz dz i dz = ie d and thus d = ie i = iz . This integral becomes: 2
1

C 2

1 dz = 2 1 2 (z z )/(2i) iz

z2

1 dz 4iz 1

We want z 4iz 1 = 0 to nd the singularities. To do this, we use our quardratic formula: 4i (4i)2 4(1) 4i 12 = = (2 3)i z1,2 = 21 2 The denominator can be factored as (z (2 + 3)i)(z (2 3)i). Only one root is inside the circle that is (2 3)i. By Cauchys formula: 1 2(2i) = 2/ 3 (2 3)i (2 + 3)i 2. #1(Just consider the case a > b > 0) Solution: Method is the same. 1 dz I= =2 1 C a + b(z z )/(2i) iz The two roots: ai

1 dz bz 2 + 2aiz b

a2 + b 2 b = b ai a2 + b2 bi = a a2 b 2 4

The magnitudes of them are respectively.

b a+ a2 b2

<

b a

< 1 and

b a a2 b2

a+ a2 b2 b

>

a b

>1

So, there is only one root inside our contour and by Caucys formula, we have: 2 2i 1 1 = 2 2i 2 2 z1 z2 bi/(a + a b ) + bi/(a a2 b2 ) 1 2 = 2 2i = 2i a2 b2 a2 b 2
+

3. Redo the integral x41 dx to make sure you understand. +1 Solution: Please refer to you notes.(I just hope you could really understand.)
cos x 4. Calculate 0 x 2 +1 dx. iz Solution: Denote I = C ze 2 +1 dz . To nd the singularity, we set z 2 + 1 = 0 and get z = i. If we pick the contour as the C1 C2 where C1 = [R, R] and C2 is the upper half circle with radius R. We can see we only have one singularity i inside the contour provided R > 1. Since well let R +, we really only need to consider this case. The pole is simple. By Cauchys formula: +

I = 2i

eii = e1 i+i

We then prove that the integral on the semi-circle goes to zero. eiz dz | = | z2 + 1
0 0

|
C2

(eR sin eiR cos ) iRei d| R2 ei2 + 1 ReR sin d R2 1 R d 0 2 R 1

if we parametrize the upper half circle with z = Rei 0 < . We thus have:
+

eiz eix dx = lim dz = lim R C z 2 + 1 R x2 + 1 1 = I = e1

+

C1 +C2

eiz dz z2 + 1

cos x dx = Re(e1 ) = e1 x2 + 1

The original integral is e1 /2