MATH2023 Multivariable Calculus Problem Set 7

L3 (Fall 2019)

Solution to Extra Problem Set 7

1. (a) Let 𝑅 be the triangular region in ℝ2 bounded by the lines 𝑥 = 0, 𝑦 = 0 and 𝑥 + 𝑦 = 1. Then 𝑆 is the
graph of the function 𝑓(𝑥, 𝑦) = 1 − 𝑥 − 𝑦 defined on 𝑅. So

2
∬𝑥𝑦𝑧 𝑑𝑆 = ∬ 𝑥𝑦 𝑓(𝑥, 𝑦)√1 + [𝑓𝑥 (𝑥, 𝑦)]2 + [𝑓𝑦 (𝑥, 𝑦)] 𝑑𝐴
𝑆 𝑅

1 1−𝑦 1
1 − 𝑦 2 1 3 𝑥=1−𝑦
=∫ ∫ 𝑥𝑦(1 − 𝑥 − 𝑦)√1 + (−1)2 + (−1)2 𝑑𝑥 𝑑𝑦 = √3 ∫ 𝑦 [ 𝑥 − 𝑥 ] 𝑑𝑦
0 0 0 2 3 𝑥=0

1
√3 1 √3 1 1 √3
= ∫ 𝑦(1 − 𝑦)3 𝑑𝑦 = [ (1 − 𝑦)5 − (1 − 𝑦)4 ] = .
6 0 6 5 4 0 120

(b) 𝑆 can be parametrized by the vector-valued function 𝐫: [1, 4] × [0, 𝜋] → ℝ3 defined by

𝐫(𝑢, 𝑣) = 〈𝑢, 2 cos 𝑣 , 2 sin 𝑣〉,
so that (𝐫𝑢 × 𝐫𝑣 )(𝑢, 𝑣) = 〈0, −2 cos 𝑣 , −2 sin 𝑣〉 for every 𝑢 ∈ (1, 4) and 𝑣 ∈ (0, 𝜋). So

∬[(𝑦 + 1)2 + 𝑧 2 ]𝑑𝑆 = ∬ [(2 cos 𝑣 + 1)2 + (2 sin 𝑣)2 ]‖𝐫𝑢 × 𝐫𝑣 ‖𝑑𝑢 𝑑𝑣
𝑆 [1,4]×[0,𝜋]

=∬ [(2 cos 𝑣 + 1)2 + (2 sin 𝑣)2 ]√02 + (−2 cos 𝑣)2 + (−2 sin 𝑣)2 𝑑𝑢 𝑑𝑣
[1,4]×[0,𝜋]

𝜋 4 4 𝜋
= ∫ ∫ (10 + 8 cos 𝑣)𝑑𝑢 𝑑𝑣 = (∫ 𝑑𝑢) [∫ (10 + 8 cos 𝑣)𝑑𝑣 ]
0 1 1 0

= (3)(10𝜋) = 30𝜋.

Alternative solution: Since 𝑆 is symmetric about the plane 𝑦 = 0, we have ∬𝑆 𝑦 𝑑𝑆 = 0. Therefore

∬[(𝑦 + 1)2 + 𝑧 2 ]𝑑𝑆 = ∬[(𝑦 2 + 𝑧 2 ) + 2𝑦 + 1]𝑑𝑆 = ∬(4 + 2𝑦 + 1)𝑑𝑆

𝑆 𝑆 𝑆

1
= 5 ∬𝑑𝑆 + 2 ∬𝑦 𝑑𝑆 = 5(Area of 𝑆) + 0 = 5 [ 𝜋(2)2 (4 − 1)] = 30𝜋.
𝑆 𝑆 2

(c) Let 𝑅 be the disk in ℝ2 defined by the inequality 𝑥 2 + 𝑦 2 ≤ 2𝑥. Then 𝑆 is the graph of the function
𝑓(𝑥, 𝑦) = √𝑥 2 + 𝑦 2 defined on 𝑅. So

∬(𝑥 4 − 𝑦 4 + 𝑦 2 𝑧 2 − 𝑥 2 𝑧 2 + 1)𝑑𝑆
𝑆

2
= ∬ [𝑥 4 − 𝑦 4 + 𝑦 2 [𝑓(𝑥, 𝑦)]2 − 𝑥 2 [𝑓(𝑥, 𝑦)]2 + 1]√1 + [𝑓𝑥 (𝑥, 𝑦)]2 + [𝑓𝑦 (𝑥, 𝑦)] 𝑑𝐴
𝑅
2 2
𝑥 𝑦
= ∬ [𝑥 4 − 𝑦 4 + 𝑦 2 (𝑥 2 + 𝑦 2 ) − 𝑥 2 (𝑥 2 + 𝑦 2 ) + 1]√1 + ( ) +( ) 𝑑𝐴
𝑅 √𝑥 2 + 𝑦 2 √𝑥 2 + 𝑦 2

= √2 ∬ 𝑑𝐴 = √2(Area of 𝑅) = 𝜋√2.
𝑅

Page 1 of 11
MATH2023 Multivariable Calculus Problem Set 7
L3 (Fall 2019)

𝑥2
(d) 𝑆 is the graph of the function 𝑓(𝑥, 𝑦) = + 𝑦 defined on [0, 1] × [0, 1]. So
2

1 1 2
∬ 𝑑𝑆 = ∬ √1 + [𝑓𝑥 (𝑥, 𝑦)]2 + [𝑓𝑦 (𝑥, 𝑦)] 𝑑𝐴
𝑆 √𝑧 −𝑦+1 [0,1]×[0,1] √𝑓 (𝑥, 𝑦) −𝑦+1

1
=∬ √1 + (𝑥)2 + (1)2 𝑑𝐴
[0,1]×[0,1] 𝑥2
√( + 𝑦) − 𝑦 + 1
2

= √2 ∬ 𝑑𝐴 = √2.
[0,1]×[0,1]

2. (a) We may assume that the circular cone 𝑆 occupied by the thin sheet has vertex at the origin and base on the
plane 𝑧 = ℎ. Let 𝑅 be the closed unit disk in ℝ2 with radius 𝑎 centered at the origin. Then 𝑆 becomes
ℎ
the graph of the function 𝑓(𝑥, 𝑦) = √𝑥 2 + 𝑦 2 defined on 𝑅. Since the circular cone 𝑆 has surface area
𝑎
𝜋𝑎√𝑎2 + ℎ2 , the moment of inertia of the thin sheet about the 𝑧-axis is given by
𝑚 𝑚 2
𝐼𝑧 = ∬(𝑥 2 + 𝑦 2 )𝑑𝑆 = ∬ (𝑥 2 + 𝑦 2 )√1 + [𝑓𝑥 (𝑥, 𝑦)]2 + [𝑓𝑦 (𝑥, 𝑦)] 𝑑𝐴
𝜋𝑎√𝑎2 + ℎ2 𝑆 𝜋𝑎√𝑎2 + ℎ2 𝑅
2 2
𝑚 ℎ 𝑥 ℎ 𝑦
= ∬ (𝑥 2 + 𝑦 2 )√1 + ( ) +( ) 𝑑𝐴
𝜋𝑎√𝑎2 + ℎ2 𝑅 𝑎 √𝑥 2 + 𝑦 2 𝑎 √𝑥 2 + 𝑦 2

𝑚 ℎ2
= ∬ (𝑥 2 + 𝑦 2 )√1 + 𝑑𝐴
𝜋𝑎√𝑎2 + ℎ2 𝑅 𝑎2

𝑚 ℎ2 2𝜋 𝑎 2 𝑚 ℎ2 2𝜋 𝑎
= √1 + ∫ ∫ 𝑟 𝑟𝑑𝑟 𝑑𝜃 = √1 + (∫ 𝑑𝜃 ) (∫ 𝑟 3 𝑑𝑟)
2
𝑎 0 0 𝑎 2
𝜋𝑎√𝑎2 + ℎ2 𝜋𝑎√𝑎2 + ℎ2 0 0

𝑚 ℎ2 1 1
= √1 + (2𝜋) ( 𝑎4 ) = 𝑚𝑎2 .
𝜋𝑎√𝑎2 + ℎ2 𝑎2 4 2

(b) The surface 𝑆 occupied by the thin spherical shell can be parametrized by the vector-valued function
𝐫: [0, 𝜋] × [0, 2𝜋] → ℝ3 defined by
𝐫(𝑢, 𝑣) = 〈𝑎 sin 𝑢 cos 𝑣 , 𝑎 sin 𝑢 sin 𝑣 , 𝑎 cos 𝑢〉.
We have already seen before (e.g. in Q10 of Problem Set 5) that
‖𝐫𝑢 × 𝐫𝑣 ‖ = 𝑎2 sin 𝑢 .
Since 𝑆 has surface area 4𝜋𝑎2 , the moment of inertia of the thin spherical shell about the 𝑧-axis is given by
𝑚
𝐼𝑧 = ∬(𝑥 2 + 𝑦 2 )𝑑𝑆
4𝜋𝑎2 𝑆

𝑚
= ∬ [(𝑎 sin 𝑢 cos 𝑣)2 + (𝑎 sin 𝑢 sin 𝑣)2 ] ⏟
‖𝐫𝑢 × 𝐫𝑣 ‖ 𝑑𝑢𝑑𝑣
4𝜋𝑎2 [0,𝜋]×[0,2𝜋] 2 =𝑎 sin 𝑢

2𝜋 𝜋 2𝜋 𝜋
𝑚 𝑚
= 2
∫ ∫ 𝑎2 sin2 𝑢 𝑎2 sin 𝑢 𝑑𝑢 𝑑𝑣 = 2
(∫ 𝑎4 𝑑𝑣 ) (∫ sin3 𝑢 𝑑𝑢)
4𝜋𝑎 0 0 4𝜋𝑎 0 0

𝜋
𝑚 1 2
= 2
(2𝜋𝑎4 ) [ cos 3 𝑢 − cos 𝑢] = 𝑚𝑎2 .
4𝜋𝑎 3 0 3

Page 2 of 11
MATH2023 Multivariable Calculus Problem Set 7
L3 (Fall 2019)

3. With the given parametrization 𝐫: [0, 2𝜋] × [−1, 1] → ℝ3 of the Möbius strip defined by
𝑢 𝑢 𝑢
𝐫(𝑢, 𝑣) = 〈(2 + 𝑣 sin ) cos 𝑢 , (2 + 𝑣 sin ) sin 𝑢 , 𝑣 cos 〉 ,
2 2 2

we have
𝑣 𝑢 𝑢 𝑣 𝑢 𝑢 𝑣 𝑢
𝐫𝑢 (𝑢, 𝑣) = 〈( cos ) (cos 𝑢) + (2 + 𝑣 sin ) (− sin 𝑢), ( cos ) (sin 𝑢) + (2 + 𝑣 sin ) (cos 𝑢), − sin 〉
2 2 2 2 2 2 2 2

and
𝑢 𝑢 𝑢
𝐫𝑣 (𝑢, 𝑣) = 〈sin cos 𝑢 , sin sin 𝑢 , cos 〉
2 2 2

for every (𝑢, 𝑣) ∈ (0, 2𝜋) × (−1, 1), so

𝐢 𝐣 𝐤
(𝐫𝑢 × 𝐫𝑣 )(𝑢, 0) = | 𝑢 𝑢
−2 sin 2 cos 𝑢 0 | = 〈2 cos 𝑢 cos 𝑢 , 2 sin 𝑢 cos 𝑢 , −2 sin 𝑢〉
𝑢 𝑢 2 2 2
sin cos 𝑢 sin sin 𝑢 cos
2 2 2
for every (𝑢, 𝑣) ∈ (0, 2𝜋). Therefore the vectors
𝑢 𝑢 𝑢
lim (𝐫𝑢 × 𝐫𝑣 )(𝑢, 0) = lim+ 〈2 cos 𝑢 cos , 2 sin 𝑢 cos , −2 sin 〉 = 〈2, 0, 0〉
𝑢→0+ 𝑢→0 2 2 2

and
𝑢 𝑢 𝑢
lim −(𝐫𝑢 × 𝐫𝑣 )(𝑢, 0) = lim − 〈2 cos 𝑢 cos , 2 sin 𝑢 cos , −2 sin 〉 = 〈−2, 0, 0〉
𝑢→2𝜋 𝑢→2𝜋 2 2 2

point in opposite directions.

4. (a) Let 𝑅 be the triangular region in ℝ2 bounded by the lines 𝑥 = 0, 𝑦 = 0 and 𝑥 + 𝑦 = 1. Then 𝑆 is the
upward oriented graph of 𝑓: 𝑅 → ℝ defined by
𝑓(𝑥, 𝑦) = 1 − 𝑥 − 𝑦,
Therefore

∬𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅(𝑥, 𝑦, 𝑓(𝑥, 𝑦)) ⋅ 〈−𝑓𝑥 , −𝑓𝑦 , 1〉𝑑𝐴 = ∬ 〈𝑥 − 𝑦, 𝑦 2 , 1 − 𝑥 − 𝑦〉 ⋅ 〈1, 1, 1〉𝑑𝐴

𝑆 𝑅 𝑅

1 1−𝑦 1
=∫ ∫ (1 − 2𝑦 + 𝑦 2 )𝑑𝑥 𝑑𝑦 = ∫ (1 − 𝑦)3 𝑑𝑦
0 0 0

1
1 1
= [− (1 − 𝑦)4 ] = .
4 0 4

(b) Let 𝑅 be the annulus in ℝ2 defined by

𝑅 = {(𝑥, 𝑦) ∈ ℝ2 : 1 ≤ 𝑥 2 + 𝑦 2 ≤ 4}.
Then 𝑆 is the downward oriented graph of 𝑓: 𝑅 → ℝ defined by
𝑓(𝑥, 𝑦) = √𝑥 2 + 𝑦 2 .

Page 3 of 11
MATH2023 Multivariable Calculus Problem Set 7
L3 (Fall 2019)

Therefore
𝑥 𝑦
∬𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅(𝑥, 𝑦, 𝑓(𝑥, 𝑦)) ⋅ 〈𝑓𝑥 , 𝑓𝑦 , −1〉𝑑𝐴 = ∬ 〈𝑥 2 , 𝑦 2 , 𝑥 2 + 𝑦 2 〉 ⋅ 〈 , , −1〉 𝑑𝐴
𝑆 𝑅 𝑅 √𝑥 2 + 𝑦2 √𝑥 2 + 𝑦2

𝑥 3 + 𝑦3 2𝜋 2
𝑟 3 cos 3 𝜃 + 𝑟 3 sin3 𝜃
=∬( − 𝑥 2 − 𝑦 2 ) 𝑑𝐴 = ∫ ∫ ( − 𝑟 2 ) 𝑟𝑑𝑟 𝑑𝜃
𝑅 √𝑥 2 + 𝑦 2 0 1 𝑟
2 2𝜋 2 2𝜋
15𝜋
= (∫ 𝑟 3 𝑑𝑟 ) [∫ (cos 3 𝜃 + sin3 𝜃 − 1)𝑑𝜃 ] = (∫ 𝑟 3 𝑑𝑟 ) (∫ −1 𝑑𝜃 ) = − .
1 0 1 0 2

Alternative solution: As a surface of revolution, 𝑆 can be parametrized by the vector-valued function

3
𝐫: [1, 2] × [0, 2𝜋] → ℝ defined by
𝐫(𝑢, 𝑣) = 〈𝑢 cos 𝑣 , 𝑢 sin 𝑣 , 𝑢〉.
For every 𝑢 ∈ (1, 2) and 𝑣 ∈ (0, 2𝜋), we have

(𝐫𝑢 × 𝐫𝑣 )(𝑢, 𝑣) = 〈cos 𝑣 , sin 𝑣 , 1〉 × 〈−𝑢 sin 𝑣 , 𝑢 cos 𝑣 , 0〉 = 〈−𝑢 cos 𝑣 , −𝑢 sin 𝑣 , 𝑢
⏟〉,
>0

so the orientation of 𝑆 induced by the parametrization 𝐫 is opposite to the required downward orientation.
Therefore

∬𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅 ⋅ −(𝐫𝑢 × 𝐫𝑣 )𝑑𝑢 𝑑𝑣
𝑆 [1,2]×[0,2𝜋]

=∬ 〈𝑢2 cos 2 𝑣 , 𝑢2 sin2 𝑣 , 𝑢2 〉 ⋅ 〈𝑢 cos 𝑣 , 𝑢 sin 𝑣 , −𝑢〉𝑑𝑢 𝑑𝑣

[1,2]×[0,2𝜋]

2𝜋 2 2 2𝜋
=∫ ∫ (𝑢3 cos 3 𝑣 + 𝑢3 sin3 𝑣 − 𝑢3 )𝑑𝑢 𝑑𝑣 = (∫ 𝑢3 𝑑𝑢) [∫ (cos 3 𝑣 + sin3 𝑣 − 1)𝑑𝑣 ]
0 1 1 0

2 2𝜋
15𝜋
= (∫ 𝑢3 𝑑𝑢 ) (∫ −1 𝑑𝑣 ) = − .
1 0 2

(c) 𝑆 can be parametrized by the vector-valued function 𝐫: 𝑅 → ℝ3 defined by

𝐫(𝑢, 𝑣) = 〈1 − 𝑢2 , 𝑢, 𝑣〉
where 𝑅 = {(𝑢, 𝑣) ∈ ℝ2 : −1 ≤ 𝑢 ≤ 1 and 0 ≤ 𝑣 ≤ 1 − 𝑢2 }. For every (𝑢, 𝑣) in the interior of 𝑅, we have

(𝐫𝑢 × 𝐫𝑣 )(𝑢, 𝑣) = 〈−2𝑢, 1, 0〉 × 〈0, 0, 1〉 = 〈 ⏟

1 , 2𝑢, 0〉 ,
>0

so the orientation of 𝑆 induced by the parametrization 𝐫 agrees with the required orientation toward the
positive 𝑥-axis. Therefore
1 1−𝑢2
∬𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅 ⋅ (𝐫𝑢 × 𝐫𝑣 )𝑑𝑢 𝑑𝑣 = ∬ 〈𝑢2 , 𝑣, 𝑢2 − 1〉 ⋅ 〈1, 2𝑢, 0〉𝑑𝑢 𝑑𝑣 = ∫ ∫ (𝑢2 + 2𝑢𝑣)𝑑𝑣 𝑑𝑢
𝑆 𝑅 𝑅 −1 0

1 1 1
1 1 4
= ∫ [𝑢2 (1 − 𝑢2 ) + 𝑢(1 − 𝑢2 )2 ]𝑑𝑢 = ∫ 𝑢2 (1 − 𝑢2 )𝑑𝑢 = [ 𝑢3 − 𝑢5 ] = .
−1 −1 3 5 −1 15

Page 4 of 11
MATH2023 Multivariable Calculus Problem Set 7
L3 (Fall 2019)

(d) The surface 𝑆 consists of six parts. Denote the parts of 𝑆 lying on the planes 𝑥 = 0, 𝑥 = 𝑎, 𝑦 = 0, 𝑦 = 𝑎,
𝑧 = 0 and 𝑧 = 𝑎 by 𝑆1 , 𝑆2 , … , 𝑆6 respectively. Because of the positive orientation of 𝑆, the unit normal
vectors to 𝑆1 , 𝑆2 , … , 𝑆6 are −𝐢, 𝐢, −𝐣, 𝐣, −𝐤 and 𝐤 respectively. Now

∬ 𝐅 ⋅ 𝑑𝐒 = ∬ (2𝑥𝑦𝐢 + 2𝑦𝑧𝐣 + 2𝑥𝑧𝐤) ⋅ (−𝐢) 𝑑𝑆 = ∬ (−2𝑥𝑦)𝑑𝑆 = ∬ 0 𝑑𝑆 = 0,

𝑆1 𝑆1 𝑆1 𝑆1

and similarly ∬𝑆 𝐅 ⋅ 𝑑𝐒 = ∬𝑆 𝐅 ⋅ 𝑑𝐒 = 0. Also

3 5
𝑎 𝑎
∬ 𝐅 ⋅ 𝑑𝐒 = ∬ (2𝑥𝑦𝐢 + 2𝑦𝑧𝐣 + 2𝑥𝑧𝐤) ⋅ 𝐢 𝑑𝑆 = ∬ 2𝑥𝑦 𝑑𝑆 = 𝑎 ∬ 2𝑦 𝑑𝑆 = 𝑎 ∫ ∫ 2𝑦 𝑑𝑦 𝑑𝑧 = 𝑎4 ,
𝑆2 𝑆2 𝑆2 𝑆2 0 0
4
and similarly ∬𝑆 𝐅 ⋅ 𝑑𝐒 = ∬𝑆 𝐅 ⋅ 𝑑𝐒 = 𝑎 . Therefore
4 6

∯𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅 ⋅ 𝑑𝐒 + ∬ 𝐅 ⋅ 𝑑𝐒 + ⋯ + ∬ 𝐅 ⋅ 𝑑𝐒 = 0 + 𝑎4 + 0 + 𝑎4 + 0 + 𝑎4 = 3𝑎4 .
𝑆 𝑆1 𝑆2 𝑆6

Alternative solution: Let 𝐷 be the solid cube cut from the first octant by the three planes 𝑥 = 𝑎, 𝑦 = 𝑎
and 𝑧 = 𝑎. Then 𝐷 is the interior of the closed surface 𝑆. Applying Gauss’ Divergence Theorem, we have

∯𝐅 ⋅ 𝑑𝐒 = ∭ ∇ ⋅ 𝐅 𝑑𝑉 = ∭ (2𝑦 + 2𝑧 + 2𝑥)𝑑𝑉
𝑆 𝐷 𝐷

𝑎 𝑎 𝑎
= ∫ ∫ ∫ (2𝑦 + 2𝑧 + 2𝑥)𝑑𝑥 𝑑𝑦 𝑑𝑧 = 3𝑎4 .
0 0 0

(e) Let 𝑅 be the unit disk 𝑅 in ℝ2 centered at the origin. Then 𝑆 is the downward oriented graph of the
function 𝑓(𝑥, 𝑦) = 𝑥 2 + 𝑦 2 defined on 𝑅. So

∬𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅(𝑥, 𝑦, 𝑓(𝑥, 𝑦)) ⋅ 〈𝑓𝑥 , 𝑓𝑦 , −1〉𝑑𝐴 = ∬ 〈4𝑥, 4𝑦, 2〉 ⋅ 〈2𝑥, 2𝑦, −1〉𝑑𝐴
𝑆 𝑅 𝑅

2𝜋 1
= ∬ (8𝑥 2 + 8𝑦 2 − 2)𝑑𝐴 = ∫ ∫ (8𝑟 2 − 2)𝑟𝑑𝑟 𝑑𝜃 = (2𝜋)(1) = 2𝜋.
𝑅 0 0

Alternative solution: 𝑆 can be parametrized by the vector-valued function 𝐫: [0, 1] × [0, 2𝜋] → ℝ3 ,
𝐫(𝑢, 𝑣) = 〈𝑢 cos 𝑣 , 𝑢 sin 𝑣 , 𝑢2 〉.
For every 𝑢 ∈ (0, 1) and 𝑣 ∈ (0, 2𝜋), we have

(𝐫𝑢 × 𝐫𝑣 )(𝑢, 𝑣) = 〈cos 𝑣 , sin 𝑣 , 2𝑢〉 × 〈−𝑢 sin 𝑣 , 𝑢 cos 𝑣 , 0〉 = 〈−2𝑢2 cos 𝑣 , −2𝑢2 sin 𝑣 , 𝑢
⏟〉,
>0

so the orientation of 𝑆 induced by the parametrization 𝐫 is opposite to the required downward orientation.
Therefore

∬𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅 ⋅ −(𝐫𝑢 × 𝐫𝑣 )𝑑𝑢 𝑑𝑣 = ∬ 〈4𝑢 cos 𝑣 , 4𝑢 sin 𝑣 , 2〉 ⋅ 〈2𝑢2 cos 𝑣 , 2𝑢 2 sin 𝑣 , −𝑢〉𝑑𝑢 𝑑𝑣

𝑆 𝑅 𝑅

2𝜋 1 2𝜋 1
=∫ ∫ (8𝑢3 − 2𝑢)𝑑𝑢 𝑑𝑣 = (∫ 𝑑𝑣 ) [∫ (8𝑢3 − 2𝑢)𝑑𝑢] = (2𝜋)(1) = 2𝜋.
0 0 0 0

Page 5 of 11
MATH2023 Multivariable Calculus Problem Set 7
L3 (Fall 2019)

Alternative solution: Let 𝑆 ′ be the upward oriented unit disk in ℝ3 lying in the plane 𝑧 = 1, centered at
the point (0, 0, 1), and let 𝐷 be the portion of the solid paraboloid 𝑧 ≥ 𝑥 2 + 𝑦 2 below the plane 𝑧 = 1.
Then the surface consisting of 𝑆 together with 𝑆 ′ becomes a positively oriented closed surface, whose
interior is 𝐷. Applying Gauss’ Divergence Theorem, we have

∬𝐅 ⋅ 𝑑𝐒 + ∬ 𝐅 ⋅ 𝑑𝐒 = ∭ ∇ ⋅ 𝐅 𝑑𝑉 = ∭ (4 + 4 + 0)𝑑𝑉
𝑆 𝑆′ 𝐷 𝐷

2𝜋 1 1 1
=∫ ∫ ∫ 8 𝑑𝑧 𝑟𝑑𝑟 𝑑𝜃 = 2𝜋 ∫ 8(1 − 𝑟 2 )𝑟 𝑑𝑟 = 4𝜋.
0 0 𝑟2 0

On the other hand, because of the upward orientation of 𝑆 ′ , the unit normal vector to 𝑆 ′ is 𝐧
̂ = 〈0, 0, 1〉.
So

̂ 𝑑𝑆 = ∬ 〈4𝑥, 4𝑦, 2〉 ⋅ 〈0, 0, 1〉𝑑𝑆 = ∬ 2 𝑑𝑆 = 2(Area of 𝑆 ′ ) = 2𝜋.

∬ 𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅 ⋅ 𝐧
𝑆′ 𝑆′ 𝑆′ 𝑆′

Therefore

∬𝐅 ⋅ 𝑑𝐒 = 4𝜋 − 2𝜋 = 2𝜋.
𝑆

5. (a) Let
 𝑆1 be the downward oriented unit disk in ℝ3 lying on the 𝑥𝑦-plane centered at the origin,
 𝑆2 be the upward oriented portion of the plane 𝑧 = 𝑥 + 2 inside the solid circular cylinder 𝑥 2 + 𝑦 2 ≤ 1,
and
 𝑆3 be the portion of the circular cylinder 𝑥 2 + 𝑦 2 = 1 lying between the planes 𝑧 = 0 and 𝑧 = 𝑥 + 2,
oriented away from the 𝒛-axis.
Then the given oriented surface 𝑆 consists exactly of 𝑆1 , 𝑆2 and 𝑆3 . Now we have the following:
̂ = 〈0, 0, −1〉.
(i) Because of the downward orientation of 𝑆1 , the unit normal vector to 𝑆1 is 𝐧 So

̂ 𝑑𝑆 = ∬ 〈2𝑥, −3𝑦, 𝑧〉 ⋅ 〈0, 0, −1〉𝑑𝑆 = ∬ −𝑧 𝑑𝑆 = ∬ 0 𝑑𝑆 = 0.

∬ 𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅 ⋅ 𝐧
𝑆1 𝑆1 𝑆1 𝑆1 𝑆1

(ii) Let 𝑈 be the unit disk in ℝ2 centered at the origin. Then 𝑆2 is the upward oriented graph of the
function
𝑓(𝑥, 𝑦) = 𝑥 + 2
defined on 𝑈. So

∬ 𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅(𝑥, 𝑦, 𝑓(𝑥, 𝑦)) ⋅ 〈−𝑓𝑥 , −𝑓𝑦 , 1〉𝑑𝐴 = ∬ 〈2𝑥, −3𝑦, 𝑥 + 2〉 ⋅ 〈−1, 0, 1〉𝑑𝐴
𝑆2 𝑈 𝑈

= ∬ (2 − 𝑥)𝑑𝐴 = 2 ∬ 𝑑𝐴 = 2𝜋,
𝑈 𝑈

where we have observed that ∬𝑈 𝑥 𝑑𝐴 = 0 because the disk 𝑈 is symmetric about the line 𝑥 = 0.

Page 6 of 11
MATH2023 Multivariable Calculus Problem Set 7
L3 (Fall 2019)

(iii) 𝑆3 can be parametrized by 𝐫: 𝑅 → ℝ3 defined by

𝐫(𝑢, 𝑣) = 〈cos 𝑢 , sin 𝑢 , 𝑣〉,
where 𝑅 is the region 𝑅 = {(𝑢, 𝑣) ∈ ℝ2 : 0 ≤ 𝑢 ≤ 2𝜋 and 0 ≤ 𝑣 ≤ cos 𝑢 + 2}. For every (𝑢, 𝑣) in the
interior of 𝑅 we have
(𝐫𝑢 × 𝐫𝑣 )(𝑢, 𝑣) = 〈− sin 𝑢 , cos 𝑢 , 0〉 × 〈0, 0, 1〉 = 〈cos 𝑢 , sin 𝑢 , 0〉,
so the orientation of 𝑆3 induced by 𝐫 agrees with the chosen orientation away from the 𝑧-axis. So

∬ 𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅 ⋅ (𝐫𝑢 × 𝐫𝑣 )𝑑𝑢 𝑑𝑣 = ∬ 〈2 cos 𝑢 , −3 sin 𝑢 , 𝑣〉 ⋅ 〈cos 𝑢 , sin 𝑢 , 0〉𝑑𝑢 𝑑𝑣

𝑆3 𝑅 𝑅

2𝜋 cos 𝑢+2 2𝜋
=∫ ∫ (2 cos 2 𝑢 − 3 sin2 𝑢)𝑑𝑣 𝑑𝑢 = ∫ (2 cos 2 𝑢 − 3 sin2 𝑢)(cos 𝑢 + 2)𝑑𝑢
0 0 0

2𝜋 2𝜋
= ∫ (2 − 5 sin2 𝑢) cos 𝑢 𝑑𝑢 + 2 ∫ (2 cos 2 𝑢 − 3 sin2 𝑢)𝑑𝑢 = 0 + (2 − 3)(2𝜋) = −2𝜋.
0 0

Therefore by (i), (ii) and (iii) we have

∯𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅 ⋅ 𝑑𝐒 + ∬ 𝐅 ⋅ 𝑑𝐒 + ∬ 𝐅 ⋅ 𝑑𝐒 = 0 + 2𝜋 + (−2𝜋) = 0.
𝑆 𝑆1 𝑆2 𝑆3

(b) Let 𝐷 be the solid bounded by the closed surface 𝑆. Then by Gauss’ Divergence Theorem, we have
𝜕 𝜕 𝜕
∯𝐅 ⋅ 𝑑𝐒 = ∭ ∇ ⋅ 𝐅 𝑑𝑉 = ∭ [ (2𝑥) + (−3𝑦) + (𝑧)] 𝑑𝑉 = ∭ (2 − 3 + 1)𝑑𝑉 = 0.
𝑆 𝐷 𝐷 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝐷

6. [It is too difficult to compute the flux directly, so we aim to use Gauss’ Divergence Theorem. We need to create a
closed surface in order to use this theorem.]

Let 𝑆 ′ be the downward oriented unit disk in ℝ3 lying in the 𝑥𝑦-plane centered at the origin, and let 𝐷 be the
solid hemisphere 𝑥 2 + 𝑦 2 + 𝑧 2 ≤ 1 above the 𝑥𝑦-plane. Then the surface consisting of 𝑆 together with 𝑆 ′
becomes a positively oriented closed surface, whose interior is 𝐷. Applying Gauss’ Divergence Theorem, we have

∬𝐅 ⋅ 𝑑𝐒 + ∬ 𝐅 ⋅ 𝑑𝐒 = ∭ ∇ ⋅ 𝐅 𝑑𝑉 = ∭ (1 + 1 + 1)𝑑𝑉
𝑆 𝑆′ 𝐷 𝐷

1 4𝜋
= 3 ∭ 𝑑𝑉 = 3(Volume of 𝐷) = 3 ( ) = 2𝜋.
𝐷 2 3

On the other hand, because of the downward orientation of 𝑆 ′ , the unit normal vector to 𝑆 ′ is 𝐧
̂ = 〈0, 0, −1〉.
So we have

̂ 𝑑𝑆 = ∬ 〈𝑥 + 𝑦 2 + 𝑒 𝑧 , 𝑦 + sin(𝑧 − 𝑥) , 𝑧 + 2〉 ⋅ 〈0, 0, −1〉𝑑𝑆

∬ 𝐅 ⋅ 𝑑𝐒 = ∬ 𝐅 ⋅ 𝐧
𝑆′ 𝑆′ 𝑆′

= ∬ −(𝑧 + 2)𝑑𝑆 = ∬ −2 𝑑𝑆 = −2(Surface area of 𝑆 ′ ) = −2𝜋.

𝑆′ 𝑆′

Therefore

∬𝐅 ⋅ 𝑑𝐒 = 2𝜋 − (−2𝜋) = 4𝜋.
𝑆

Page 7 of 11
MATH2023 Multivariable Calculus Problem Set 7
L3 (Fall 2019)

7. Given the vector field 𝐅(𝑥, 𝑦, 𝑧) = 〈𝑥, 𝑦, 𝑧〉, by Gauss’ Divergence Theorem we have
1 1 1 𝜕 𝜕 𝜕 1
∯𝐅 ⋅ 𝑑𝐒 = ∭ ∇ ⋅ 𝐅 𝑑𝑉 = ∭ [ (𝑥) + (𝑦) + (𝑧)] 𝑑𝑉 = ∭ 3 𝑑𝑉 = ∭ 𝑑𝑉 = Volume of 𝐷.
3 𝑆 3 𝐷 3 𝐷 𝜕𝑥 𝜕𝑦 𝜕𝑧 3 𝐷 𝐷

8. (a) The divergence of the vector field 𝑓∇𝑔 = 〈𝑓𝑔𝑥 , 𝑓𝑔𝑦 , 𝑓𝑔𝑧 〉 is given by
𝜕 𝜕 𝜕
∇ ⋅ (𝑓∇𝑔) = (𝑓𝑔𝑥 ) + (𝑓𝑔𝑦 ) + (𝑓𝑔𝑧 )
𝜕𝑥 𝜕𝑦 𝜕𝑧

= (𝑓𝑥 𝑔𝑥 + 𝑓𝑔𝑥𝑥 ) + (𝑓𝑦 𝑔𝑦 + 𝑓𝑔𝑦𝑦 ) + (𝑓𝑧 𝑔𝑧 + 𝑓𝑔𝑧𝑧 )

= 𝑓(𝑔𝑥𝑥 + 𝑔𝑦𝑦 + 𝑔𝑧𝑧 ) + (𝑓𝑥 𝑔𝑥 + 𝑓𝑦 𝑔𝑦 + 𝑓𝑧 𝑔𝑧 )
= 𝑓(∇ ⋅ ∇𝑔) + ∇𝑓 ⋅ ∇𝑔.
Therefore applying Gauss’ Divergence Theorem to the vector field 𝑓∇𝑔, we have

∯(𝑓∇𝑔) ⋅ 𝑑𝐒 = ∭ ∇ ⋅ (𝑓∇𝑔) 𝑑𝑉 = ∭ [𝑓(∇ ⋅ ∇𝑔) + ∇𝑓 ⋅ ∇𝑔]𝑑𝑉 .

𝑆 𝐷 𝐷

(b) From the computation in (a), the divergence of the vector field 𝑓∇𝑔 − 𝑔∇𝑓 is given by
∇ ⋅ (𝑓∇𝑔 − 𝑔∇𝑓) = ∇ ⋅ (𝑓∇𝑔) − ∇ ⋅ (𝑔∇𝑓)
= [𝑓(∇ ⋅ ∇𝑔) + ∇𝑓 ⋅ ∇𝑔] − [𝑔(∇ ⋅ ∇𝑓) + ∇𝑔 ⋅ ∇𝑓]
= 𝑓(∇ ⋅ ∇𝑔) − 𝑔(∇ ⋅ ∇𝑓).
Therefore applying Gauss’ Divergence Theorem to the vector field 𝑓∇𝑔 − 𝑔∇𝑓, we have

∯(𝑓∇𝑔 − 𝑔∇𝑓) ⋅ 𝑑𝐒 = ∭ ∇ ⋅ (𝑓∇𝑔 − 𝑔∇𝑓) 𝑑𝑉 = ∭ [𝑓(∇ ⋅ ∇𝑔) − 𝑔(∇ ⋅ ∇𝑓)]𝑑𝑉 .

𝑆 𝐷 𝐷

9. [In order to apply Gauss’ Divergence Theorem, we need to convert the given surface integral of a function over 𝑆
into a surface integral of a vector field across 𝑆. ̂, we
In other words, denoting the unit normal vector of 𝑆 as 𝐧
̂ = 2𝑥 + 2𝑦 + 𝑧 2 .]
need to come up with a vector field 𝐅 such that 𝐅(𝑥, 𝑦, 𝑧) ⋅ 𝐧

Note that 𝑆 is a level surface of the function 𝑓(𝑥, 𝑦, 𝑧) = 𝑥 2 + 𝑦 2 + 𝑧 2 , so at every point (𝑥, 𝑦, 𝑧) on 𝑆, a unit
normal vector to 𝑆 is given by
1 1 1
̂=
𝐧 ∇𝑓 = 〈2𝑥, 2𝑦, 2𝑧〉 = 〈2𝑥, 2𝑦, 2𝑧〉 = 〈𝑥, 𝑦, 𝑧〉.
‖∇𝑓‖ √(2𝑥)2 + (2𝑦)2 + (2𝑧)2 2√𝑥 2 + 𝑦 2 + 𝑧 2
̂ gives the positive orientation of 𝑆.
This unit normal vector 𝐧 Now if we let
𝐅(𝑥, 𝑦, 𝑧) = 〈2, 2, 𝑧〉,
then

∯(2𝑥 + 2𝑦 + 𝑧 2 )𝑑𝑆 = ∯〈2, 2, 𝑧〉 ⋅ 〈𝑥, 𝑦, 𝑧〉𝑑𝑆 = ∯𝐅 ⋅ 𝐧

̂ 𝑑𝑆 ,
𝑆 𝑆 𝑆

so by Gauss’ Divergence Theorem we have

𝜕 𝜕 𝜕 4𝜋
∯(2𝑥 + 2𝑦 + 𝑧 2 )𝑑𝑆 = ∭ ∇ ⋅ 𝐅 𝑑𝑉 = ∭ [ (2) + (2) + (𝑧)] 𝑑𝑉 = ∭ 1 𝑑𝑉 = Volume of 𝐷 = .
𝑆 𝐷 𝐷 𝜕𝑥 𝜕𝑦 𝜕𝑧 𝐷 3

Page 8 of 11
MATH2023 Multivariable Calculus Problem Set 7
L3 (Fall 2019)

10. ℝ3 ∖ 𝑋 is simply connected in (a), (c), (d), (g) and (i); ℝ3 ∖ 𝑋 is not simply connected in (b), (e), (f), (h), (j), (k), (l).
ℝ3 ∖ 𝑋 is not even connected in (e), (h) and (k).

11. (a) The three edges of the triangle 𝐶 can be parametrized by

𝐫1 (𝑡) = 〈1 − 𝑡, 0, 𝑡〉, 𝐫2 (𝑡) = 〈0, 𝑡, 1 − 𝑡〉 and 𝐫3 (𝑡) = 〈𝑡, 1 − 𝑡, 0〉
respectively, where 𝑡 ∈ [0, 1] in each of these parametrizations. Since
𝐫1 ′ (𝑡) = 〈−1, 0, 1〉, 𝐫2 ′ (𝑡) = 〈0, 1, −1〉 and 𝐫3 ′ (𝑡) = 〈1, −1, 0〉
for 𝑡 ∈ (0, 1), we have
1 1
∮𝐅 ⋅ 𝑑𝐫 = ∫ 〈(1 − 𝑡)(0), (0)(𝑡), (1 − 𝑡)(𝑡)〉 ⋅ 〈−1, 0, 1〉𝑑𝑡 + ∫ 〈(0)(𝑡), (𝑡)(1 − 𝑡), (0)(1 − 𝑡)〉 ⋅ 〈0, 1, −1〉𝑑𝑡
𝐶 0 0

1
+ ∫ 〈(𝑡)(1 − 𝑡), (1 − 𝑡)(0), (𝑡)(0)〉 ⋅ 〈1, −1, 0〉𝑑𝑡
0

1
1
= ∫ 3𝑡(1 − 𝑡)𝑑𝑡 = .
0 2

(b) Let 𝑆 be the triangular portion of the plane 𝑥 + 𝑦 + 𝑧 = 1 lying in the first quadrant, oriented towards the
origin. Then the given curve 𝐶 is positively oriented with respect to 𝑆. Now the flux of 𝐅 is given by
𝐢 𝐣 𝐤
𝜕 𝜕 𝜕
∇×𝐅=| | = 〈−𝑦, −𝑧, −𝑥〉,
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝑥𝑦 𝑦𝑧 𝑥𝑧
and 𝑆 is the downward oriented graph of
𝑓(𝑥, 𝑦) = 1 − 𝑥 − 𝑦
defined on the triangular region 𝑅 in ℝ2 bounded by the lines 𝑥 = 0, 𝑦 = 0 and 𝑥 + 𝑦 = 1. So applying
Stokes’ Theorem, we have

∮𝐅 ⋅ 𝑑𝐫 = ∬(∇ × 𝐅) ⋅ 𝑑𝐒 = ∬ (∇ × 𝐅)(𝑥, 𝑦, 𝑓(𝑥, 𝑦)) ⋅ 〈𝑓𝑥 , 𝑓𝑦 , −1〉𝑑𝐴

𝐶 𝑆 𝑅

1
= ∬ 〈−𝑦, −(1 − 𝑥 − 𝑦), −𝑥〉 ⋅ 〈−1, −1, −1〉𝑑𝐴 = ∬ 𝑑𝐴 = Area of 𝑅 = .
𝑅 𝑅 2

12. (a) The curve 𝐶 can be parametrized by

𝐫(𝑡) = 〈cos 𝑡 , sin 𝑡 , 0〉
′ (𝑡)
for 𝑡 ∈ [0, 2𝜋]. Now 𝐫 = 〈− sin 𝑡 , cos 𝑡 , 0〉 for every 𝑡 ∈ (0, 2𝜋), so
2𝜋
∮𝐅 ⋅ 𝑑𝐫 = ∫ 𝐅(𝐫(𝑡)) ⋅ 𝐫 ′ (𝑡)𝑑𝑡
𝐶 0

2𝜋
= ∫ 〈sin 𝑡 , − cos 𝑡 , (cos 𝑡)2 (sin 𝑡)2 (0)〉 ⋅ 〈− sin 𝑡 , cos 𝑡 , 0〉𝑑𝑡
0

2𝜋
= ∫ −(sin2 𝑡 + cos 2 𝑡)𝑑𝑡 = −2𝜋.
0

Page 9 of 11
MATH2023 Multivariable Calculus Problem Set 7
L3 (Fall 2019)

(b) We note that 𝐶 is the boundary curve of the surface 𝑆, which is negatively oriented with respect to 𝑆. Now
𝐢 𝐣 𝐤
𝜕 𝜕 𝜕
∇ × 𝐅 = || || = 〈2𝑥 2 𝑦𝑧, −2𝑥𝑦 2 𝑧, −2〉 = 2〈𝑥 2 𝑦𝑧, −𝑥𝑦 2 𝑧, −1〉 = 2𝐆,
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝑦 −𝑥 𝑥 2𝑦 2𝑧
so applying Stokes’ Theorem, we have
1 1
∬𝐆 ⋅ 𝑑𝐒 = ∬2𝐆 ⋅ 𝑑𝐒 = ∬(∇ × 𝐅) ⋅ 𝑑𝐒
𝑆 2 𝑆 2 𝑆

1 1
= − ∮𝐅 ⋅ 𝑑𝐫 = − (−2𝜋) = 𝜋.
2 𝐶 2

Alternative solution: Although tedious, it is still possible to compute ∬𝑆 𝐆 ⋅ 𝑑𝐒 directly. 𝑆 can be

𝜋
parametrized by the vector-valued function 𝐫: [ , 𝜋] × [0, 2𝜋] → ℝ3 defined by
2

1
𝐫(𝑢, 𝑣) = 〈sin 𝑢 cos 𝑣 , sin 𝑢 sin 𝑣 , cos 𝑢〉 .
√3
𝜋
For every 𝑢 ∈ ( , 𝜋) and 𝑣 ∈ (0, 2𝜋) we have
2
1 1
(𝐫𝑢 × 𝐫𝑣 )(𝑢, 𝑣) = 〈 sin2 𝑢 cos 𝑣 , sin2 𝑢 sin 𝑣 , sin
⏟𝑢 ⏟cos 𝑢〉 ,
√3 √3 >0 <0

so the orientation of 𝑆 induced by the parametrization 𝐫 agrees with the required downward orientation.
Therefore

∬𝐆 ⋅ 𝑑𝐒 = ∬ 𝐆 ⋅ (𝐫𝑢 × 𝐫𝑣 )𝑑𝑢 𝑑𝑣
𝜋
𝑆 [ ,𝜋]×[0,2𝜋]
2

1 1
=∬ 〈(sin 𝑢 cos 𝑣)2 (sin 𝑢 sin 𝑣) ( cos 𝑢) , −(sin 𝑢 cos 𝑣)(sin 𝑢 sin 𝑣)2 ( cos 𝑢) , −1〉
𝜋
[ ,𝜋]×[0,2𝜋]
2
√3 √3

1 1
⋅〈 sin2 𝑢 cos 𝑣 , sin2 𝑢 sin 𝑣 , sin 𝑢 cos 𝑢〉 𝑑𝑢 𝑑𝑣
√3 √3
2𝜋 𝜋
1 1
=∫ ∫ ( sin5 𝑢 cos 𝑢 sin 𝑣 cos 3 𝑣 − sin5 𝑢 cos 𝑢 sin3 𝑣 cos 𝑣 − sin 𝑢 cos 𝑢) 𝑑𝑢 𝑑𝑣
0
𝜋 3 3
2

𝜋 2𝜋 2𝜋 𝜋
1 5
=∫ sin 𝑢 cos 𝑢 [∫ (sin 𝑣 cos 3 𝑣 − sin3 𝑣 cos 𝑣)𝑑𝑣 ] 𝑑𝑢 − (∫ 𝑑𝑣 ) (∫ sin 𝑢 cos 𝑢 𝑑𝑢 )
𝜋 3 ⏟0 0
𝜋
2 2
=0

𝜋
1
= 0 − (2𝜋) [ sin2 𝑢]𝜋 = 𝜋.
2
2

Page 10 of 11
MATH2023 Multivariable Calculus Problem Set 7
L3 (Fall 2019)

13. Let 𝑆 be the portion of the plane in ℝ3 enclosed by the curve 𝐶, oriented with the same unit normal vector 𝐧
̂,
so that 𝐶 is positively oriented with respect to 𝑆.
(a) Given the vector field 𝐅(𝑥, 𝑦, 𝑧) = 〈𝑥, 𝑦, 𝑧〉, if we write 𝐯 = 〈𝑎, 𝑏, 𝑐〉, then
𝐢 𝐣 𝐤
𝜕 𝜕 𝜕
∇ × (𝐯 × 𝐅) = | | = 〈2𝑎, 2𝑏, 2𝑐〉 = 2𝐯,
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝑏𝑧 − 𝑐𝑦 𝑐𝑥 − 𝑎𝑧 𝑎𝑦 − 𝑏𝑥
so by Stokes’ Theorem we have

∮(𝐯 × 𝐅) ⋅ 𝑑𝐫 = ∬(∇ × (𝐯 × 𝐅)) ⋅ 𝑑𝐒 = ∬2𝐯 ⋅ 𝑑𝐒 = 2 ∬𝐯 ⋅ 𝑑𝐒.

𝐶 𝑆 𝑆 𝑆
1
̂ is a constant vector since 𝑆 is a plane.
(b) Note that 𝐧 ̂ in the result obtained in (a), we have
Choosing 𝐯 = 𝐧
2

1
̂ × 𝐅) ⋅ 𝑑𝐫 = ∬𝐧
∮(𝐧 ̂ ⋅ 𝑑𝐒 = ∬𝐧 ̂⋅𝐧
̂ 𝑑𝑆 = ∬1 𝑑𝑆 = Area of 𝑆.
2 𝐶 𝑆 𝑆 𝑆

14. Consider the vector field 𝐇 ≔ 𝐅 − 𝐆, whose components also have continuous partial derivatives. Our aim is to
show that 𝐇 = 𝟎 on 𝐷.
Since 𝐷 is a simply connected region and since ∇ × 𝐇 = 𝟎 on 𝐷 by (iii), it follows that 𝐇 is a conservative
vector field in 𝐷, i.e. there exists a function 𝑓: 𝐷 → ℝ such that
𝐇 = ∇𝑓
on 𝐷. Now since ∇ ⋅ 𝐇 = 0 on 𝐷 by (ii), the divergence of the vector field 𝑓𝐇 is given by
∇ ⋅ (𝑓𝐇) = ∇𝑓 ⋅ 𝐇 + (𝑓)(∇ ⋅ 𝐇)
=𝐇⋅𝐇+0
=𝐇⋅𝐇
on 𝐷. ̂ = 0 on 𝑆 by (i), we have
So by Gauss’ Divergence Theorem together with that 𝐇 ⋅ 𝐧

∭ (𝐇 ⋅ 𝐇)𝑑𝑉 = ∭ ∇ ⋅ (𝑓𝐇)𝑑𝑉 = ∯𝑓𝐇 ⋅ 𝑑𝐒

𝐷 𝐷 𝑆

̂ 𝑑𝑆 = ∯0 𝑑𝑆 = 0.
= ∯𝑓𝐇 ⋅ 𝐧
𝑆 𝑆

Finally since 𝐇 ⋅ 𝐇 = ‖𝐇‖2 ≥ 0 on 𝐷 and ∭𝐷(𝐇 ⋅ 𝐇)𝑑𝑉 = 0, it follows that 𝐇 = 𝟎 on 𝐷. ∎

Page 11 of 11