MATH2011 SEVERAL VARIABLE CALCULUS

S1 2004
TEST 1 SOLUTIONS

2. [B]
1
,
1 + x + 2y
2
2g
=
,
x2
(1 + x + 2y)3

g(x, y) =

The versions in this test were colour coded. In these solutions, they are indicated by codes
B = BLUE,

G = GREEN,

P = PINK,

Y = YELLOW .

g = 1,

1.
x
y
y
x
= cos ,
= r sin ,
= sin ,
= r cos .
r

b) [ALL] For z = f (x, y),

x
y
x
y

f
= 2xy 2 ,
x

[Not required in test: The Taylor series in full can be found as:


1
=
(1)n (x + 2y)n
1 + (x + 2y) n=0
n  



n k nk k
2 x
=
(1)n
y .
]
k
n=0

f
= 2x2 y
y

2. [G]
1
,
1 + x 2y
2
2
g
=
,
x2
(1 + x 2y)3

g(x, y) =

z
= r sin (2x) + r cos 2y = 2r 2 sin cos + 2r 2 cos sin

= 4r 2 sin cos = 2r 2 sin 2 .

f
= 2y
y

f
= 2x,
x

2
g
=
,
y
(1 + x 2y)2
2
g
8
=
,
y 2
(1 + x 2y)3

g
= 1,
x

2g
= 2,
x2

g
= 2,
y

2g
= 4,
g x

2g
=8.
y 2

so the Taylor series of g about (0, 0) up to and including the 2nd order terms is:

c) [Y] For f (x, y) = x2 + y 2,

g
1
=
,
x
(1 + x 2y)2
2
4
g
=
,
g x
(1 + x 2y)3

hence at (0, 0),

z
= r sin 2x + r cos (2y) = 2r 2 sin cos 2r 2 cos sin

= 4r 2 sin cos = 2r 2 sin 2 .

if |x + 2y| < 1

k=0

f
= 2y
y

f
= 2x,
x

f
= 2x,
x

2g
=8.
y 2

 1


1

1(x 0) 2(y 0) +
2(x 0)2 + 2 4(x 0)(y 0) + 8(y 0)2 +
1!
2!
= 1 x 2y + x2 + 4xy + 4y 2 +

c) [P] For f (x, y) = x2 y 2,

2g
= 4,
g x

1+

c) [G] For f (x, y) = y 2 x2 ,

2g
= 2,
x2

g
= 2,
y

g(x, y) =

z
= r sin 2xy 2 + r cos 2x2 y = 2r 4 cos sin3 + 2r 4 cos3 sin

1
= 2r 4 cos sin (cos2 sin2 ) = r 4 sin 2 cos 2 = r 4 sin 4 .
2

g
= 1,
x

so the Taylor series of g about (0, 0) up to and including the 2nd order terms is:

c) [B] For f (x, y) = x2 y 2,

2
g
=
,
y
(1 + x + 2y)2
2g
8
=
,
y 2
(1 + x + 2y)3

hence at (0, 0),

Also the code ALL indicates a common question to all versions.

g
1
=
,
x
(1 + x + 2y)2
4
2g
=
,
g x
(1 + x + 2y)3

 1


1

1(x 0) + 2(y 0) +
2(x 0)2 + 2 (4)(x 0)(y 0) + 8(y 0)2 +
1!
2!
= 1 x + 2y + x2 4xy + 4y 2 +

1+

[Not required in test: The Taylor series in full can be found as:
f
= 2y
y


1
=
(1)n (x 2y)n
1 + (x 2y) n=0
n  



n
(2)k xnk y k .
=
(1)n
k
n=0

if |x 2y| < 1

g(x, y) =

z
= r sin 2x + r cos 2y = 2r 2 sin cos + 2r 2 cos sin

= 0.

k=0

2. [P]

g
1
=
,
x
(1 x 2y)2
4
2g
=
,
g x
(1 x 2y)3

1
,
1 x 2y
2
2g
=
,
x2
(1 x 2y)3

g(x, y) =

2
g
=
,
y
(1 x 2y)2
2g
8
=
,
y 2
(1 x 2y)3

hence at (0, 0),

g
= 2,
y

g
= 1,
x

g
= 2,
x2

g
= 4,
g x

3. [B] Maximise and minimise

g
=8.
y 2


 1

1

1+
1(x 0) + 2(y 0) +
2(x 0)2 + 2 4(x 0)(y 0) + 8(y 0)2 +
1!
2!
= 1 + x + 2y + x2 + 4xy + 4y 2 +


1
=
(x + 2y)n
1 (x + 2y) n=0
n  


n k nk k
2 x
=
y .
]
k
n=0

1
g(x, y) =
,
1 x + 2y
2
2g
=
,
x2
(1 x + 2y)3

if |x + 2y| < 1

(2)
(3)

The simplest solution is to eliminate by taking y (1) and x (2), obtaining two expressions
for 2xy,
4x
y(2x 6) = x(2y 8) so 2xy 6y = 2xy 8x, 6y = 8x or y =
3
 2
25x2
4x
16x2
2
2
2
=
= 25, x = 9,
= 25 from (3), i.e. x +
x +
3
9
9

and
and

g
1
=
,
x
(1 x + 2y)2
4
2g
=
,
g x
(1 x + 2y)3

(1)

2y 8 = 2y
x2 + y 2 = 25

k=0

2. [Y]

2x 6 = 2x

(* equivalently, the Lagrange Multiplier Thm. states local extrema for h on = 0 occur at
critical points of h for some constant , on = 0, i.e. h = and = 0.),

[Not required in test: The Taylor series in full can be found as:
g(x, y) =

(x, y) = x2 + y 2 25 = 0 .

With F (x, y, ) = h(x, y) (x, y), then F = 0 gives (*)

so the Taylor series of g about (0, 0) up to and including the 2nd order terms is:

subject to

2
g
=
,
y
(1 x + 2y)2
2g
8
=
,
y 2
(1 x + 2y)3

x = 3,

y = 4

(same signs)

h(3, 4) = 32 + 42 6 3 8 4 50 = 25 18 32 50 = 75

h(3, 4) = (3)2 + (4)2 6 (3) 8 (4) 50 = 25 + 18 + 32 50 = 25

Hence on = 0, h has a maximum value of 25 (at (3, 4)) and a minimum value of 75 (at
(3, 4))
A slower solution is to use (1) and (2) to express x and y in terms of , substitute into (3) to
find an equation for , solve for and hence find x and y. This method gives
3
4
,
y=
1
1
2 
2

3
4
+
= 25
1
1

x=

hence at (0, 0),

g
= 1,
x

g
= 2,
y

2g
= 2,
x2

2g
= 4,
g x

2g
=8.
y 2

so the Taylor series of g about (0, 0) up to and including the 2nd order terms is:
 1


1

1(x 0) 2(y 0) +
2(x 0)2 + 2 (4)(x 0)(y 0) + 8(y 0)2 +
1!
2!
= 1 + x 2y + x2 4xy + 4y 2 +

1+

or 25 = 25(1 )2


1
=
(x 2y)n
1 (x 2y) n=0

n  

n
(2)k xnk y k .
=
k
n=0

=0

(x, y) = (3, 4)

=2

(x, y) = (3, 4)

if |x 2y| < 1
]

k=0

( 2) = 0

and

[Not required in test: The Taylor series in full can be found as:
g(x, y) =

or

etc.
etc.

3. [P] Maximise and minimise

3. [G] Maximise and minimise

(x, y) = x2 + y 2 8x 6y 75 = 0 .

subject to

h(x, y) = x2 + y 2 ,

With F (x, y, ) = h(x, y) (x, y), then F = 0 gives (*)

subject to

(x, y) = x2 + y 2 + 6x + 8y 75 = 0 .

With F (x, y, ) = h(x, y) (x, y), then F = 0 gives (*)

2x = (2x 8)

(1)

2x = (2x + 6)

(1)

2y = (2y 6)
2
2
x + y 8x 6y = 75

(2)
(3)

2y = (2y + 8)
2
2
x + y + 6x + 8y = 75

(2)
(3)

(* equivalently, the Lagrange Multiplier Thm. states local extrema for h on = 0 occur at
critical points of h for some constant , on = 0, i.e. h = and = 0.),

(* equivalently, the Lagrange Multiplier Thm. states local extrema for h on = 0 occur at
critical points of h for some constant , on = 0, i.e. h = and = 0.),

The simplest solution is to eliminate by taking (2y 6) (1) and (2x 8) (2), obtaining
two expressions for (2x 8)(2y 6),

The simplest solution is to eliminate by taking (2y + 8) (1) and (2x + 6) (2), obtaining
two expressions for (2x + 6)(2y + 8),

2x(2y 6) = 2y(2x 8)

x2 +

so 4xy 12x = 4xy 16y,

12x = 16y

2

3x
= 75 from (3),
4
50x
x2
2x
25x2

= 75,

=3
i.e.
16
4
16
4
2
i.e. x 8x 24 = 0, i.e. (x 12)(x + 4) = 0
(x, y) = (9, 12) or (4, 3)

3x
4

8x 6

and
and

x2 +

h(4, 3) = (4) + (3) = 25

so 4xy + 16x = 4xy + 12y,

4x
= 75 from (3),
3
50x
x2
2x
25x2
+
= 75,
+
=3
i.e.
9
3
9
3
2
i.e. x + 6x 27 = 0, i.e. (x + 9)(x 3) = 0
(x, y) = (3, 4) or (9, 12)

4x
3

+ 6x + 8

and

16x = 12y

2

h(9, 12) = 92 + 122 = 81 + 144 = 225

2x(2y + 8) = 2y(2x + 6)

h(3, 4) = 32 + 42 = 25

h(9, 12) = (9)2 + (12)2 = 81 + 144 = 225

and

Hence on = 0, h has a maximum value of 225 (at (9, 12)) and a minimum value of 25 (at
(4, 3))

Hence on = 0, h has a maximum value of 225 (at (9, 12)) and a minimum value of 25
(at (3, 4))

A slower solution is to use (1) and (2) to express x and y in terms of , substitute into (3) to
find an equation for , solve for and hence find x and y. This method gives

A slower solution is to use (1) and (2) to express x and y in terms of , substitute into (3) to
find an equation for , solve for and hence find x and y. This method gives

4
3
,
y=
1
1
2 
2





3
4
3
4
6
= 75
+
8
1
1
1
1
x=

or

25 + 50(1 ) 75(1 ) = 0

2 + 2 22 3 + 6 32 = 0

3
4
,
y=
1
1
2 
2





4
3
4
3
+8
= 75
+
+6
1
1
1
1
x=

or 252 + 50(1 ) 75(1 )2 = 0

42 8 + 3 = 0
i.e (2 1)(2 3) = 0

4 8 + 3 = 0
i.e (2 1)(2 3) = 0

and

1
2
3
=
2
=

(x, y) = (4, 3)
(x, y) = (12, 9)

2 + 2 22 3 + 6 32 = 0

etc.

etc.

and

1
2
3
=
2
=

(x, y) = (3, 4)

etc.

(x, y) = (9, 12)

etc.

3. [Y] Maximise and minimise

subject to

(x, y) = x2 + y 2 25 = 0 .

With F (x, y, ) = h(x, y) (x, y), then F = 0 gives (*)

2x + 6 = 2x

(1)

2y 8 = 2y
x2 + y 2 = 25

(2)
(3)

(* equivalently, the Lagrange Multiplier Thm. states local extrema for h on = 0 occur at
critical points of h for some constant , on = 0, i.e. h = and = 0.),
The simplest solution is to eliminate by taking y (1) and x (2), obtaining two expressions
for 2xy,
4x
y(2x + 6) = x(2y 8) so 2xy + 6y = 2xy 8x, 6y = 8x or y =
3
2

25x2
4x
16x2
2
2
2
=
= 25, x = 9,
= 25 from (3), i.e. x +
x +
3
9
9

x = 3,

y = 4

(opp. signs)

and

h(3, 4) = 32 + (4)2 + 6 3 8 (4) 50 = 25 + 18 + 32 50 = 25

and

h(3, 4) = (3)2 + 42 + 6 (3) 8 4 50 = 25 18 32 50 = 75

Hence on = 0, h has a maximum value of 25 (at (3, 4)) and a minimum value of 75 (at
(3, 4))
A slower solution is to use (1) and (2) to express x and y in terms of , substitute into (3) to
find an equation for , solve for and hence find x and y. This method gives
3
4
,
y=
1
1
2 
2

3
4
+
= 25
1
1

x=

or

25 = 25(1 )2

or ( 2) = 0

=0

(x, y) = (3, 4)

etc.

and

=2

(x, y) = (3, 4)

etc.