Assignment 1 Solution: Math. Phys.

3 (PEU
455)

Prof. Ali Nassar

October 21, 2017

24.1
Let f (z) = u(x, y) + iv(x, y) be an analytic function whose imaginary part,
v(x, y), is given by

v(x, y) = (y cos y + x sin y)ex (1)

Analyticity of f (z) implies that it satisfies the Cauchy-Euler equations.
Namely,

u v
= (2)
x y
u v
= (3)
y x
Plugging Equation (1) into Equations (2) and (3) gives

u
= (cos y y sin y + x cos y)ex (4)
x
u
= ex y cos y sin yex x sin yex (5)
y
Integrating (4) and (5) with respect to x and y, respectively, yields

u(x, y) = ex cos y ex y sin y + cos yex (x 1) + g(x, y) (6)

x x x
u(x, y) = e (cos y + y sin y) + cos ye + x cos ye + h(x, y) (7)

where g(y) and h(x) are two generic functions of x and y, respectively. How-
ever, (6) and (7) are essentially equal. Therefore, matching their terms and
simplifying yields

u(x, y) = ex (x cos y y sin y) (8)

1
Hence, combining Equations (1) and (8) yields

f (z) = ex (x cos y y sin y) + i(y cos y + x sin y)ex (9)

2
24.2
Again, let f (z) = u(x, y) + iv(x, y) be an analytic function whose real part,
u(x, y) is given by
sin 2x
Re(f (z)) = u(x, y) = (10)
cosh 2y cos 2x
The fact that f (z) is analytic guarantees that u(x, y) and v(x, y) satisfy the
Cauchy-Riemann equations. Plugging Equation (10) in the Cauchy-Riemann
equations yields the two equations

u 2 sin 2x sinh 2y v
= 2
= (11)
x (cosh 2y cos 2x) x
u 2(1 + cos 2x cosh 2y) v
= 2
= (12)
x (cos 2x cosh 2y) y

Integrating the above two equations with respect to x and y, respectively,

yields exactly the same result

sinh(2y)
v(x, y) = +C
cosh(2y) cos(2x)

where C is an arbitrary constant.

Therefore, f (z) is given by
sin 2x i sinh 2y
f (z) = (13)
cosh 2y cos 2x
Some algebraic manipulation enables us to write Equation (13) in terms of
the variable z = x + iy as follows
cos z
f (z) = cot(z) = (14)
sin z
It is to be noted that f (z) has poles (essential singularities) whenever sin z is
equal to zero, namely at z = n, where n Z, where Z = {..., 2, 1, 0, 1, 2, ...}.

Note: Attached is a Mathematica notebook that you can use to reproduce

the results here very simply. Just run the sections of the notebook.

3
24.3
a)

X zn
n=2
ln n
Here,
1
an =
ln n
According to the results of Section 24.3, the radius of convergence of the
series, R, is given by

1 |an+1 | | ln(n + 1)
= lim = lim =1
R ninf |an | ninf ln n
Therefore, R = 1.

b)

X n!z n
n=1
nn
Similarly,
n!
an =
nn
To find the radius of convergence, we use Equation (24.13) from Hobson.
 n!  n1 1
1 1 n! n
= lim |an | n =  n  = lim (15)
 
R ninf n ninf n

However, for very large values of n, n! has an asymptotic behavior given by

the Stirling series

 n n  1 1 139 
n!
2n 1+ + + ...
e 12n 288n2 51840n3
n
Therefore, for very large values of n, n! n . Plugging this result into
Equation (15) yields
1 n
= lim =1
R ninf n
Therefore, R = 1.

4
24.5
a)
1
f (z) =
z2
Clearly, the function is analytic at z = 0. To examine the behavior at
z inf, we make the substitution
1
zw= (16)
z
Hence,
1 w
f (w) = 1 =
w 2 1 2w
Clearly,
w 1
lim =
winf 1 2w 2
Therefore, f (z) is analytic at z inf.

b)
1 + z3 1
f (z) = 2
= 2 +z
z z
While ,at z = 0, f (z) and zf (z) both diverge, z 2 f (z) = 1 + z 3 doesnt.
Therefore, f (z) has a second order pole at z = 0.
To examine the behavior at infinity, we do the same substitution as we have
done in part (a). Doing so, we obtain
1
f (w) = w2 +
w
Clearly, f (w) diverges at w = 0. However, wf (w) doesnt. Therefore, f (z)
has a simple pole at z inf.

c)
The power series expansion of the hyperbolic sine function is given by
inf
X z 2n+1
sinh z =
n=0
(2n + 1)!
Therefore,

1 inf
X 1
f (z) = sinh = 2n+1 (2n + 1)!
z n=0
z

5
 Since limz0 z n f (z) diverges for all n. Then f (z) has an essential singularity
at z = 0.
To examine the behavior at infinity, we do the substitution given by Equation
(16).
inf
X w2n+1
f (w) = sinh w =
n=0
(2n + 1)!
Clearly,

lim f (w) = 0
w0

This implies that f (z) is analytic at z inf.

d)
ez
f (z) =
z3
Clearly, f (z) diverges at z = 0. However, we know that ez is analytic for all
finite z. Therefore, z 3 f (z) is analytic at z = 0. Hence, f (z) has a third order
pole at z = 0.
To examine the behavior at infinity, we use Equation (16).
1
f (w) = w3 e w
Obviously,

1
 1 1 1 1 
lim w3 e w = lim w3 1 + + + + + ... = inf
w0 w0 w 2!w2 3!w3 4!w4
And since the sum in the above equation extends to infinity. Then there is
no n N such that wn f (w) is finite. Therefore, f(z) has an essential singularity
at z inf.

e)
 z  21
f (z) =
1 + z2
Obviously, f (z) does not diverge at z = 0. Hence, f (z) has no poles at z = 0.
However, if we make the substitution z = rei , we get
i
1
 rei  21 r2 2
1
f (z) = 2 2i
= (1 + r2 e2i ) 2
1+r e e
Obviously, under the transformation + 2, f(z) picks a minus sign.
Therefore, z = 0 is a branch point for f (z).
Again, using Equation (16), we can examine the behavior of f (z) at infinity.

6
 1  21

w
 w  12
f (w) =  2 =
1 1 + w2
1+ w

It is obvious that f (w) and f (z) have exactly the same form. Therefore,
f (w) must have a branch point at w = 0. Thus, f (z) has a branch point at
z = 0.

7
24.6
a)
sin z
f (z) = tan z =
cos z
Obviously, whenever sin z = 0, we have tan z = 0 as well. We know that
sin z = 0 whenever z = n, where n Z. Therefore, tan z has zeros at z = n,
where n Z.
On the other hand, tan z has essential singularities whenever cos z = 0. This
is because cos z is a transcendental function. Therefore, it cannot be expressed
as a finite sum of polynomials. Therefore, there is no finite n such that z n f (z)
is finite. Thus, f (z) = tan z has essential singularities at z = (n + 12 ), where
n Z.
To examine the behavior of f (z) at infinity, we use Equation (16), as always.
 
1 sin w1
tan z = tan =  
w cos 1 w

Taking the limit as w goes to zero we get

 
1
sin w
h 1  1  1 3 ih 1  1 2 i1
lim   = lim + ... 1 + ...
w0
cos 1 w0 w 3! w 2 w
w
h 1  1  1 3 ih 1  1 2 i
= lim + ... 1 + ...
w0 w 3! w 2 w
where in the last step, we have treated the series as two terms: 1 and the
rest of the terms, then Taylor-expanded the series according to

(1 + x)1 = 1 + x + x2 + x3 + ...
Obviously, the series expansion of f (w) has arbitrarily high powers of n.
Therefore, as w approaches zero, tan z approaches infinity with no n N such
that z n tan z is finite. Therefore, z = inf is an essential singularity of f (z) =
tan z.
Moreover, since there are no fractal powers of z involved in f (z). Then, f (z)
is single-valued, and there are no branch points.

b)
z2  1 
f (z) = sin
z2 1z
First, lets begin with identifying the zeros of f (z). Obviously, z = 2 is a
zero of the function. Also, plugging z = inf into f (z) yields f (inf) = sin 0
inf = 0.

8
  
1 1
Also, we know that sin 1z , whenever 1z = n, where n N. Therefore,
1
f (z) has zeros at z = 1 n ,where n Z.
On the other hand, at z = 0, f (z) diverges. However, z 2 f (z) doesnt. There-
fore, z = 0 is a second order pole of f (z). Also,
 according
 to the same argument
1
we followed in part (a), it is evident that sin 1z has an essential singularity
at z = inf.

c)
inf
1 X 1  1 n
f (z) = exp =
z n=0
n! z
The exponential function is non-zero everywhere in the complex plane. There-
fore, it has no zeros.
Apparently, f (z) diverges at z = 0. And since the exponential is a transcen-
dental function. Then, there is no n N such that z n f (z) is finite at z = 0.
Therefore, the point z = 0 is an essential singularity. The behavior at infinity
is regular, for f (inf) = e0 = 1.
Also, there are no branch points.

d)
1
f (z) = tan
z
 
1
The behavior of tan zcan be deduced very easily from tan z. Whenever
 
tan z has a zero at z = a, tan z1 has a singularity at the same point. Also,
both have no branch points.
1
Therefore, f (z) has zeros at z = inf and z = n , where n Z. Also, f (z)
1
has simple poles at z = (n+ 1 ) .
2  
It is to be noted that the function f (z) = tan z1 has a non-isolated singu-
larity at z = 0. That is for every neighborhood of radius  surrounding z = 0,
there exists at least on singularity other than the one at z = 0. The singularity
at z = 0 is an essential singularity.

e)
2
f (z) = z 3
Obviously, the function has a zero at z = 0. Also, if we rewrite z as rei ,
we can see that f (z) is not single-valued when makes a complete revolution
around z = 0. Therefore, z = 0 is a branch point.
To examine the behavior at infinity, we use Equation (16)
2
f (w) = w 3

9
 Obviously, w = 0 (i.e. z = inf) is a branch point. Also, it is a first order
pole.

10
Additional Question
i)
This is represented graphically in the Mathematica Notebook: Assignment 1 -
Supplementary Notebook 2. Refer to it for more details.

ii)
Although this is obvious from the graphs produced in part (i), we prove it here
using gradients.

u = i + j
v = i j

where i and j are unit vectors pointing in the x and t directions, respectively.
But the gradients of u and v are perpendicular to lines of constant u and lines
of constant v, respectively. Therefore, u v = 0 would imply that lines of
constant u and lines of constant v are perpendicular to each other. Evidently,

u v = (i + j) (i j) = 1 1 = 0
This establishes the desired result.

iii)
We know that
f f
df = dx + dt (17)
x t
We want to write the above equation in terms of u and v which are given by

u=x+t (18)
v =xt (19)

From Equations (17) and (18) we get

1
dx = (du + dv) (20)
2
1
dt = (du dv) (21)
2
Also,

11
 u
=1 (22)
x
u
=1 (23)
t
v
=1 (24)
x
u
= 1 (25)
t
Using the chain rule, we can rewrite Equation (17) as
h f u f v i h f u f v i
df = + dx + + dt (26)
u x v x u t v t
Putting it all together, we get
f f
df = du + dv (27)
u v

12