M337 1998 Exam Solutions Issue 1.

0
1998 Question 1
(a) 5 marks
(a)(i)
2
i 1
i 1
i 1
i 1
1
i 1
1
w

= |
.
|

\
|

+
=
+
=
ThereIore Arg w - t / 4. (Unit A1, Section 2, Para. 8)
(a)(ii)
2
1
2
2
w = =
The principal Iourth root oI w is (Unit A1, Section 3, Para. 3)
z
0

|
|
.
|

\
|
|
|
.
|

\
|
|
.
|

\
| t
+
|
|
.
|

\
|
|
.
|

\
| t

4 4
1
sin i
4 4
1
cos 2
8 / 1
|
|
.
|

\
|
|
.
|

\
| t
+ |
.
|

\
| t
=

16
sin i
16
cos 2
8 / 1
(Unit A1, Section 2, Para. 6)
(b) 3 marks

i 3
1 1 Log i 3 exp = (Unit A2, Section 5, Para. 3)
1 iArg 1 log i 3 exp
e
+ = (Unit A2, Section 5, Para. 1)
. 3 exp i 0 i 3 exp t = t + =
M337 1998 Exam Solutions Issue 1.0
1998 Question 2
(a) 3 marks
A
B
B-A
(b) 5 marks
(b)(i) (Unit A3, Section 4, Paras. 6 and 7)
A is not a region since it is not open.
B is a region.
B-A is not a region as it is not connected.
(b)(ii) (Unit A3, Section 5, Para. 5)
A is compact.
B is not compact as it is not closed or bounded.
B - A is not compact as it is not closed or bounded.
M337 1998 Exam Solutions Issue 1.0
1998 Question 3
(a) 4 marks
The standard parametrization (Unit A2, Section 2, Para. 3) Ior I
1
is

1
(t) (1 t) ti, t e |0, 1|
and
1
/
(t) -1 i.
As
1
is diIIerentiable on |0, 1|,
1
/
is continuous on |0, 1|, and
1
/
is non-zero on |0, 1| then
1
is a
smooth path (Unit A4, Section 4, Para. 3).
Since
1
is a smooth path then (Unit B1, Section 2, Para. 1)
< dt t t Re dz z Re
1
1
0
1
1
'
=
} }
I

}
+ =
1
0
dt i 1 t 1


1
0
2
2
t 1
i 1
(

+ =
2
i 1+
=
(b) 3 marks
Let I(z) 1/z, E(z) Log z and the region # - x : x s 0}.
I is continuous on # and E is a primitive oI I on # (Unit A4, Section 3, Para. 4). Thus, by the
Eundamental Theorem oI Calculus (Unit B1, Section 3, Para. 2), since I
1
is a contour in #

}
I
=
1
1 E i E dz
z
1
1 Log i Log =
< 0 i Arg i i log
e
+ = (Unit A2, Section 5, Para. 1)
2
it
=
(c) 1 mark
Since I
2
is also a contour in # with the same start and end points as I
1
then by the Contour
Independence Theorem (Unit B1, Section 3, Para. 4)
2
i
dz
z
1
dz
z
1
1 2
t
= =
} }
I I
M337 1998 Exam Solutions Issue 1.0
1998 Question 4
(a) 2 marks
The zeros oI z
2
2 are at z 2
1/2
i. These are outside the contour C
1
.
# z : ,z, 2} is a simply-connected region (Unit B2, Section 1, Para. 3) and
2 z
z
2
3
+
is analytic
(quotient rule) on #. Since C
1
is a closed contour in # then by Cauchy`s Theorem (Unit B2,
Section 1, Para. 4)
0 dz
2 z
z
1
C
2
3
=
+
}
(b) 3 marks
| Eollow similar strategy to that given in Unit B2, Section 4, Para. 3|
Both the zeros oI z
2
2 are inside the contour C
2
.

)
`

=
+
=
+ 2 i z
1
2 i z
1
2 i 2
1
2 i z 2 i z
1
2 z
1
2
ThereIore
)
`

=
+
} } }
2 2 2
C
3
C
3
C
2
3
dz
2 i z
z
dz
2 i z
z
2 i 2
1
dz
2 z
z
As is a simply-connected region which contains the simple-closed contour C
2
(Unit B2, Section
1, Para. 1) and I(z) z
3
is analytic on then using Cauchy`s Integral Eormula (Unit B2, Section
2, Para. 1) gives
< 2 i I i 2 2 i I i 2
2 i 2
1
dz
2 z
z
2
C
2
3
t t =
+
}
< i 4 i 2 2 i 2 2
2
t =
t
=
(c) 3 marks
C
2
is a simply-closed contour in the simply-connected region and I(z) z
3
is analytic on .
Since the point 2 lies in C
2
then by Cauchy`s n`th Derivative Iormula (Unit B2, Section 3, Para.
1), with n 1 and o -2, we have

i 24 2 3 * i 2 2 I
! 1
i 2
dz
2 z
z
2
C
2
3
2
t = t = '
t
=
+
}
M337 1998 Exam Solutions Issue 1.0
1998 Question 5
Identical to 2001 Question 5.
1998 Question 6
2,78
I and g are not direct analytic continuations oI each other since D
0
D
1
|.

=
|
.
|

\
|
=
0 n
2
2
z
z I is a geometric series with sum
z 2
2
1
1
2
z

on D
0
.

=
|
.
|

\
|
0 n
2
z
2
is a geometric series with sum
2 z
z
1
1
z
2

on D
1
.
ThereIore
z 2
2
2 z
z
z
2
z
2
z
2
z
2
z g
0 n
2
1 n
2

= |
.
|

\
|

= |
.
|

\
|
= |
.
|

\
|
=

=
on D
1
.
Let
z 2
2
z h

= on - 2}.
I and h are analytic Iunctions whose domains are regions and I(z) h(z)
Ior z e D
0
_ D
0
( 2}) then h(z) is an analytic continuation oI I(z) to 2} (Unit C3,
Section 1, Para. 1).
Since the domains oI h and g are regions and h(z) g(z) Ior z e D
1
_ ( 2}) D
1
then g(z) is
an analytic extension oI h(z) to D
1
.
As the Iunctions (I, D
0
), (h, 2}), (g, D
1
) Iorm a chain then I and g are indirect analytic
continuations oI each other (Unit C3, Section 2, Para. 3).
M337 1998 Exam Solutions Issue 1.0
1998 Question 7
(a) 1 mark
The conjugate velocity Iunction
2
z / 1 ) z ( q = .
Since q is a steady continuous 2-dimensional velocity Iunction on the region 0} and q is
analytic on 0} then q is a model Iluid Ilow (Unit D2, Section 1, Para. 14).
(b) 5 marks
On 0},
z
1
z = O is a primitive oI q . ThereIore O is a complex potential Iunction Ior the
Ilow (Unit D2, Section 2, Para. 1).
The stream Iunction z Im y , x O = + (Unit D2, Section 2, Para. 4)
|
|
.
|

\
|
+
=
iy x
1
Im , where z x iy, (x,y) = (0, 0)
2 2 2 2
y x
y
y x
iy x
Im
+
=
|
|
.
|

\
|
+

=
A streamline through the point i satisIies the equation
1 1 , 0
y x
y
2 2
= + =
+
(Unit D2, Section 2, Para. 4)
ThereIore the streamline through i has the equation x
2
y
2
y 0 or

4
1
2
2
1
2
y x = +
Since q(i) -1 (-x direction) then the direction oI Ilow is as shown.
(c) 2 marks
The Ilux oI q across the unit circle C z : ,z, 1}is (Unit D2, Section 1, Para. 10)
0 dz
z
1
Im dz z q Im
C
2
C
= |
.
|

\
|
=
} }
by Cauchy`s Residue Theorem.
M337 1998 Exam Solutions Issue 1.0
1998 Question 8
(a) 3 marks
Same as 2002 Qu. 8(a).
(b) 5 marks
(b)(i) | Unit D3, Exercise 4.1(c) |
P
c
(0) -1 i.
. i 1 i 1 i 2 i 1 i 1 0 P
2 2
c
= + + = + + + =
. i 3 1 i 1 i 2 i 1 i 1 0 P
2 3
c
+ = + + = + + =
As 2 0 P
3
c
> then c does not lie in the Mandelbrot set (Unit D3, Section 4, Para. 5).
(b)(ii)
Since
4
1
8
1
8
1
i 1 c < = = + then P
c
has an attracting 2-cycle (Unit D3, Section 4, Para. 9).
ThereIore c belongs to the Mandelbrot set (Unit D3, Section 4, Para. 8).
M337 1998 Exam Solutions Issue 1.0
1998 Question 9
(a) 8 marks
Putting z x iy we have
y , x iv y , x u y x iy x z z z I
2 2
2
+ = + = = ,
where u(x,y) x x
2
y
2
, and v(x,y) -y.
x 2 1 y , x
x
u
=
c
c
, y 2 y , x
y
u
=
c
c
0 y , x
x
v
=
c
c
, 1 y , x
y
v
=
c
c
II I is diIIerentiable the Cauchy-Riemann equations (Unit A4, Section 2, Para. 1) hold.
They will hold at (a, b) iI
b , a
y
v
1 a 2 1 b , a
x
u
c
c
= = =
c
c
, and
b , a
y
u
b 2 0 b , a
x
v
c
c
= = =
c
c
Since the Cauchy-Riemann equations only hold at (1, 0) then
I is not diIIerentiable on 1}.
As I is deIined on the region , and the partial derivatives
y
v
,
x
v
,
y
u
,
x
u
c
c
c
c
c
c
c
c
1. exist on
2. are continuous at (1, 0).
3. satisIy the Cauchy-Riemann equations at (1, 0)
then, by the Cauchy-Riemann Converse Theorem (Unit A4, Section 2, Para. 3), I is diIIerentiable
at (1, 0).
ThereIore I is only diIIerentiable at (1, 0).
M337 1998 Exam Solutions Issue 1.0
(b) 10 marks
(i) g(z) is analytic on the region 0} (Unit A4, Section 3, Para. 4),
and
3
z
2
z g = ' on 0}.
As 0 i 2
i
2
i g
3
= = = ' and g is analytic at i , then g is conIormal at z i. (Unit A4,
Section 4, Para. 6)
(ii) t/2 is in the domain oI
1
so i e 2
2 / i
1
= = t
t
.
0 is in the domain oI
2
so i 0
2
= . ThereIore I
1
and I
2
meet at the point i.
(iii) As g is analytic on - 0} and g
/
(i) = 0 then a small disc centred at i is mapped
approximately (Unit A4, Section 1, Para. 11) to a small disc centred at g(i) -1. The disc
is rotated by Arg (g
/
(i)) Arg -2i -t/2, and scaled by a Iactor , g
/
(i), 2.
In the diagram below g(I
1
) is the vertical line.
M337 1998 Exam Solutions Issue 1.0
1998 Question 10
(a) 9 marks
(i) There are poles at z 0, and z 3. As
3
1
z I 0 z
0 z
lim
=

and
3
1
z I 3 z
3 z
lim
=

then these are simple poles.

+
+

=
+
=
h 1
1
h 2
1
3
1
2 h h 1
1
z I

)
`

+
+

=
h 1 1 h
1
2 h 1 2
1
3
1
As 1 , z 1 , , h , 2 then , h/2 , 1 and , 1/h , 1. ThereIore

+ |
.
|

\
|
+ |
.
|

\
|
+

+ |
.
|

\
|
+ |
.
|

\
|
+ = ...
h
1
h
1
1
h 3
1
...
2
h
2
h
1
6
1
z I
2 2
when 1 ,h, 2. (Unit B3, Section 3, Para. 5)



...
24
1 z
12
1 z
6
1
1 z 3
1
1 z 3
1
...
2
2
+

+ =
when 1 ,z - 1, 2.
M337 1998 Exam Solutions Issue 1.0
(b) 9 marks
(i)
...
! 4
z
! 2
z
1 z cos
4 2
+ = , Ior z e . (Unit B3, Section 3, Para. 5)
...
! 2
z
z 1 e
2
z
+ + + = , Ior z e . (Unit B3, Section 3, Para. 5)
By the Composition Rule (Unit B3, Section 4, Para. 3)
... ...
! 4
z
! 2
z
! 2
1
...
! 4
z
! 2
z
1 z g
2
4 2 4 2
+
|
|
.
|

\
|
+ +
|
|
.
|

\
|
+ + =
...
6
z
2
z
1 ...
8
1
24
1
z
2
z
1
4 2
4
2
+ + = + |
.
|

\
|
+ + = Ior z e .
Since g is analytic on then by Taylor`s Theorem (Unit B3, Section 3, Para. 1) this is the unique
power series Ior g on .
(ii)
z
3
g(1/z) is analytic on the punctured disc - 0}.
The Laurent series about 0 Ior z
3
g(1/z) on this disc is

=
= + + = |
.
|

\
|
n
n
n
3 3
z a ...
z 6
1
2
z
z
z
1
g z
ThereIore as C is a circle with centre 0 (Unit B4, Section 4, Para. 2)
}
t
= |
.
|

\
|
t = t = |
.
|

\
|

C
1
3
3
i
6
1
i 2 ia 2 dz
z
1
g z
M337 1998 Exam Solutions Issue 1.0
1998 Question 11
(a) 8 marks
(a)(i)
Let g(z) 3z
2
and I(z) z
3
3z
2
1.
Using the Triangle Inequality (Unit A1, Section 5, Para. 2) then on I z : ,z, 1} we have
, I(z) g(z) , , z
3
1 , s , z
3
, 1 2
3 , 3z
2
, , g(z) ,
As I and g are analytic (Unit A4, Section 1, Para. 7) on the simply-connected region , and I is a
simple-closed contour in then by Rouch`s Theorem (Unit C2, Section 2, Para. 4) I has the
same number oI zeros inside I as g.
ThereIore I(z) 0 has 2 solutions in the disc z : ,z, 1}.
Let g(z) z
3
and I z : ,z, 4}.
Using the Triangle Inequality then on I we have
, I(z) g(z) , , 3z
2
1 , s , 3z
2
, 1 49
64 , z
3
, , g(z) ,
As I and g are analytic on the simply-connected region , and I is a simple-closed contour in
then by Rouch`s Theorem I has the same number oI zeros inside I as g. ThereIore I(z) 0 has 3
solutions in the disc z : ,z, 4}.
On z : ,z, 1} then (Unit A1, Section 5, Para. 3e)
,I(z), , 3z
2
z
3
1 , > ,3z
2
, - , z
3
, 1 3 1 - 1 ~ 0.
As there are no zeros on z : ,z, 1} then there is a single solution on the annulus
z : 1 ,z, 4}.
(a)(ii) Taking the conjugate oI I(z) 0 gives
0 1 z 3 z
2 3
= + +
ThereIore iI z is a solution so is its conjugate. As there is only one solution in the annulus then it it
must be real. II z is real then 3z
2
1 ~ 0 so iI I(z) 0 then z
3
is negative.
ThereIore the root in the annulus is real and negative.
M337 1998 Exam Solutions Issue 1.0
(b) 10 marks
(b)(i)
Let I(z) exp(z
3
) and R z : ,z, 3}.
II we write z r e
iu
then z
3
r
3
e
i3u
and
, exp(z
3
), exp(Re z
3
) exp(r
3
cos 3u).
Since I is analytic on the bounded region R and continuous on R then, by the Maximum Principle,
there exists an o e cR such that , I(z) , s , I(o) , Ior all z e R (Unit C2, Section 4, Para. 4).
On c# each point can be written in the Iorm z 3e
iu
Ior u e |0, 2t|.
ThereIore max , exp(z
3
), : z e c# }
max ,exp(27 cos 3u), : u e |0, 2t|}
e
27
.
The maximum occurs when cos 3u 1. ThereIore the maximum is attained when
z 3, 3 exp(2t/3), and 3 exp(4t/3).
Eor points in R we have r 3, so the maximum value cannot be attained elsewhere in the disc ,z, s
3.
(b)(ii)
II ,z, 3 then using the Triangle Inequalities (Unit A1, Section 5, Para. 2)
4 1 z 1 z = + s + , and
2 1 3 1 z 1 z = = > .
Hence on the contour C we have

27 27 3
e 2 e *
2
4
z exp
1 z
1 z
= s

+
.
The length oI the contour C is 2t * 3 6t.
Since
3
z exp
1 z
1 z

+
is continuous on C then, by the Estimation Theorem (Unit B1, Section 4, Para.
3), we have

27 27
C
3
e 12 e 2 * 6 dz z exp
1 z
1 z
t = t s

+
}
.
M337 1998 Exam Solutions Issue 1.0
1998 Question 12
(a) 5 marks
Using the Implicit Eormula (Unit D1, Section 2, Para. 11) then we have







1 w
i w
1 i 2
1 z
i
1
1 w
i w
1 i 2
i 2
z
1 z
+

+
+

Hence wz w z 1 2iw 2 w i.
Rearranging gives w(z 2i) -z (3 i ).
ThereIore the required extended Mbius transIormation is I
`
, where


i 2 z
i 3 z
z I

+ +
=
(b) 13 marks
(b)(i)
II z x iy then
Im z ~ 1 - Re z y ~ 1 x x y ~ 1.
M337 1998 Exam Solutions Issue 1.0
(b)(ii)
As I
1
is a Mbius transIormation and R is a region in the domain oI I
1
then I
1
(R) is a region and
the boundary oI R maps onto the boundary oI I(R) (Unit D1, Section 4, Para. 3).
The boundary oI R D
1
D
2
, where D
1
z : Re z Im z 1, 0 s Re z s 1 } and D
2
z : ,z,
1, 0 s Arg z s t/2}
As I
1
(i) 0 and I
1
(1) then both D
1
and D
2
are mapped to extended lines which start at the
origin. Since the angle between D
1
and D
2
at i is t/4 and the transIormation I
1
is conIormal, then
the angle between the lines at the origin is also t/4.
(1 i)/2 e D
1
and 1
i 1
i 1
1
i
2
i 1
I
2
i 1
2
i 1
1
=
+

= |
.
|

\
| +
+
+
.
ThereIore D
1
is mapped to the extended negative real axis x : x s 0} }. As the interior oI
the region R is on the leIt as we travel along D
1
Irom i to (i 1)/2, then this is also the case when
we travel Irom I(i) to I( (i 1)/2 ) in the transIormed region. ThereIore the transIormed region is
as shown below.
I(R) z : -t Arg z -3t/4}
(b)(iii)
w g(z) z
4
is a one-one conIormal mapping (Unit D1, Section 4, Para. 5) that maps the image
oI R to the upper halI plane. ThereIore a one-one conIormal mapping Irom R to the upper-halI
plane is
g
o
I
1
(z)
4
1 z
i z
|
.
|

\
|