2019F MAS102 Final Sol
2019F MAS102 Final Sol
1 Let F (x, y, z) be a C 1 function defined on R3 , and let S be the level surface F (x, y, z) =
3+4 5. Using the implicit function theorem, we may show that S is the graph of a function
points z = f (x, y) locally at a point (x0 , y0 , z0 ) ∈ S.
∂F
Solution. (a) The implicit function theorem requires the condition ∂z 6= 0 at (x0 , y0 , z0 ). (+3
points)
(b) By substituting z = f (x, y) into F (x, y, z) = 5, we have F (x, y, f (x, y)) = 5. By the chain
∂F ∂F ∂f ∂F ∂F ∂f
rule, it gives ∂x + ∂z ∂x = 0 and ∂y + ∂z ∂y = 0. (+2 points) Therefore the gradient is
∂F
!
∂f ∂f ∂F
∂x ∂y
∇f = , = − ∂F
,− ∂F
. (+2 points)
∂x ∂y ∂z ∂z
R1R1R1 5
2 Evaluate the integral 0 x
√
y
ez dzdydx.
7 points
R1 5
Solution. Since √ ez dz does not seem to be doable, we may change the order of this iterated
y
√
integral. Starting from witnessing that the boundaries are given by z = y, y = x, the domain is
{(x, y, z) ∈ R3 |0 ≤ x ≤ y, 0 ≤ y ≤ z 2 , 0 ≤ z ≤ 1}. (+3 points) One can modify the integral to
R 1 R z2 R y z5 R 1 R z2 5 R1 4 5
0 0 0
e dxdydz = 0 0 yez dydz = 0 z2 ez dz = e−1 10 . (+4 points)
• If you have set the right integral domain, you get (+3 points).
Solution 1. Notice that the curve x belongs to the surface S at t = 0, as x(0) = (1, 0, 0) satisfies
the equation x3 + y 3 + z 3 − xz = 1. (+1 point) In addition, S is the level surface of the function
F (x, y, z) = x3 + y 3 + z 3 − xz and the normal vector to S at (1, 0, 0) is given by the gradient
∇F = (3x2 − z, 3y 2 , 3z 2 − x) = (3, 0, −1). On the other hand, the tangent vector x0 (0) is computed
to be x0 (0) = (1, 1, 3). (+3 points) Since the inner product (3, 0, −1) · (1, 1, 3) = 0, the path x(t)
is tangent to the surface S at t = 0. (+2 points)
Solution 2. Let G(t) = F (x(t)) = (et cos t)3 + (et sin t)3 + (3t)3 − 3et cos t. By computation, we
have G(0) = 1, G0 (0) = 0. (+4 points) Then the curve x(t) meets tangentially the level surface
S at t = 0. (+2 points)
• If you have presented the tangent criterion clearly, you get (+3 points).
H
4 Let D be the triangle with vertices (0, 0), (0, 1), (1, 1). Evaluate ∂D
F · ds for
9 points
2 1 2 x+1
F(x, y) = ey + , x(2y + 1)ey −
y+1 (y + 1)2
∂N ∂M
H RR
• However, if you exactly wrote down ∂D
F · ds = D ∂x
− ∂y dxdy , only (−2 points).
Solution 2. Let x1 (t) = (t, t), x2 (t) = (1 − t, 1), and x3 (t) = (0, 1 − t) for −1 ≤ t ≤ 1, then
I Z 1 Z 1 Z 1
F · ds = F(x1 (t)) · x01 (t)dt + F(x2 (t)) · x02 (t)dt + F(x3 (t)) · x03 (t)dt
∂D 0 0 0
So, we get
e−1
I
F · ds = .
∂D 2
F · ds, for F(x, y, z) = (ex sin y)i + (ex cos y)j + (3z 2 + z + 2)k
R
5 Evaluate the integral C
2 √
9 points and C: the curve parametrized by x(t) = (et − e, 1 − t2 , t), −1 ≤ t ≤ 1.
Since the domain R3 is simply connected, F is consevative and path independent. (+4 points)
z2
Solution 2. For f (x, y, z) = ex sin y + z 3 + 2 + 2z, we get
∇f = (ex sin y)i + (ex cos y)j + (3z 2 + z + 2)k = F(x, y, z). (+4 points)
Moreover,
So, we get
Z Z
F · ds = ∇f · ds = f ((0, 0, 1)) − f ((0, 0, −1)) = 6. (+3 points)
C C
Z Z 2π
F · ds = F(x(t)) · x0 (t) dt
x 0
Z 2π
= (a cos t − ab sin t, ab cos t, a sin t) · (−a sin t, a cos t, 0) dt (+1 point)
0
Z 2π
= −a2 cos t sin t + a2 b sin2 t + a2 b cos2 t + 0 dt
0
2π
a2
Z
= − sin 2t + a2 b dt
0 2
= 0 + 2πa2 b
= 2πa2 b (+3 points)
• If you have paramatrization of the circle with wrong value a or b, you will get (−2 points).
Also, you will get (−1 point), if the range of t is incorrect. No points will be given if this
is not a circle at all.
• If you have correct answer without intermediate compuation step, you will get full points
out of 4 points. Otherwise, no points will be given.
• You will get full points if your orientation, parametrization and answer are all matched.
= 2b × a2 π
= 2πa2 b (+3 points)
• If you have the disk with wrong value a or b, you will get (−2 points). If the boundary of
the disk is not a circle at all, no points will be given to that point.
• If you have correct answer without intermediate compuation step, you will get full points
out of 4 points. Otherwise, no points will be given.
• You will get full points if your normal vector and answer are all matched.
where C1 is the bottom boundary circle of S on the plane z = 0 centered at the z-axis, radius 2
with same orientation of the circle in problem 6, and C2 is the top boundary circle of S on the
plane z = 1 centered at the z-axis, radius 5 with opposite orientation of the problem 6’s. (+3
points)
By problem 6, this is the circulation of F along C2,0 subtracted by the circulation of F along
C5,1 .
So, ZZ
(∇ × F) · dS = 2π · 22 · 0 − 2π · 52 · 1 = −50π. (+2 points)
S
• For each correct circle as the integral path, you will get (+1 point) each (ignoring orienta-
tions).
• Only the (+3 points) will be given if both of the boundary circles are correctly oriented;
no partial points on this.
• (+1 point) for having correct value of circulation of F along C1 , C2 each (with sign). It is
OK to direcrly compute these circulations without using problem 6.
Consider a solid V with its boundary is given by ∂V = S ∪ S1 ∪ S2 , where S1 is the disk in the
plane z = 0 centered at the z-axis with radius 2, and S2 is the disk in the plane z = 1 centered at
the z-axis with radius 5. (+2 points)
By Gauss’s theorem,
ZZZ ZZ
∇ · (∇ × F)dV = (∇ × F) · dS
V ∂V
ZZ ZZ ZZ
= (∇ × F) · dS + (∇ × F) · dS + (∇ × F) · dS. (+1 point)
S S1 S2
and
ZZ Z 2π Z 1 Z 2π Z 1
(∇ × F) · dS = (1 − 5s cos t, −5s sin t, 2) · (0, 0, 1) · 25s dsdt = 50s dsdt
S2 0 0 0 0
So, ZZ
(∇ × F) · dS = −50π. (+1 point)
S
• You get (−1 point) for the wrong or omitted boundary disk, with either one disk is correct.
We parametrize S by X(s, t) = ((2s + 3s ) cos t, (2s + 3s ) sin t, s) for s ∈ [0, 1], t ∈ [0, 2π]. (+2
points)
Note that ∇ × F = (1 − x, −y, 2z).
To compute the outward normal vector, we compute Xs ×Xt , which is either inward or outward
normal vector.
Xs × Xt = ((2s log 2 + 3s log 3) cos t, (2s log 2 + 3s log 3) sin t, 1) × (−(2s + 3s ) sin t, (2s + 3s ) cos t, 0)
So,
ZZ
(∇ × F) · dS
S
Z 2π Z 1
= (1 − (2s + 3s ) cos t, −(2s + 3s ) sin t, 2s) · (2s + 3s ) · (cos t, sin t, −(2s log 2 + 3s log 3)) dsdt
0 0
Z 2π Z 1
= (2s + 3s ) · (cos t − (2s + 3s ) − 2s(2s log 2 + 3s log 3)) dsdt
0 0
Z 2π Z 1 Z 2π Z 1
s s
= (2 + 3 ) cos t dsdt − (2s + 3s )2 + 2s(2s log 2 + 3s log 3)(2s + 3s ) dsdt
0 0 0 0
Z 2π Z 2π
=0− [s(2s + 3s )2 ]s=1
s=0 dt = 25 dt
0 0
• If you compute the integral with inward normal vector, you get (−2 points). Also, if you
get 50π as the final answer by using this method, you can have at most 7 points.
• No partial points for the computations and answer part except the above case.
• If you take the wrong orientation, then you get (−2 points).
• If you take the wrong boundary ∂S, then you get no point in the calculation part.
• The minor mistakes (−1 point) and the critical mistakes (−2 points).
since the z-coordinate of F is -1 on D and so F · ~n = 1, where ~n = (0, 0, −1) is the unit normal
vector of D. Also, note that
√
ZZZ ZZZ Z 2π Z 2 Z 1−r 2
∇ · FdV = 1dV = rdzdrdθ
W W 0 0 −1
√
Z 2π Z 2
= (2r − r3 )drdθ
0 0
Z 2π
= 1dθ = 2π. (+2 points)
0
Hence, ZZ ZZZ ZZ
F · dS = ∇ · FdV − F · dS = 2π − 2π = 0.
S W D
• If you take the wrong orientation, then you get (−2 points) of 4 points in the second and
third parts.
• If you take the wrong region D, then you get no point in the calculation part.
• The minor mistakes (−1 point) and the critical mistakes (−2 points).
Solution. (a) We use the Extreme Value Theorem. Since the domain E is compact (or closed and
bounded) and the function f is continuous, it satisfies the condition of the theorem. Thus, f
has a global minimum and a global maximum on E. (+3 points)
The Hessian of f is
" 2 2 2 2 2
#
−x2
−4x2 ey −x + 2ey −x + 2 4xyey
Hf (x, y) = 2
−x2 2
−x2 2
−x2
,
4xyey −4y 2 ey − 2ey +2
and
2
F 0 (t) = −3 sin 2t − 5 sin 2t e1−5 cos t .
We see that F 0 (t) = 0 when sin 2t = 0, i.e., t is π/2, π, or 3π/2. Thus, we need to consider
the following five points: (0, 0), (±2, 0), and (0, ±1). (+3 points)
(x, y) f (x, y)
(0, 0) −1
(±2, 0) 4 − e−4
(0, ±1) 1−e
Thus, the global maximum is 4 − e−4 (+1 point) and the global minimum is 1 − e (+1
point).
Solution 1. Let x = r cos θ and y = 21 r sin θ, then by the change of variables, we get
π
where 1 ≤ r ≤ 2 and 0 ≤ θ ≤ 2. (+2 points)
By the simple computation,
π
ZZ Z Z 2
2 1 1
y(x2 + 4y 2 )3/2 dA = r sin θ · r3 · r drdθ
0 1 2 2
D
π
Z Z 2
2 1 5
= r sin θ drdθ
0 1 4
Z π
2 21
= sin θ dθ
0 8
21
= . (+4 points)
8
Solution 2. We directly calculate the double integral by defining two regions:
Clearly D = D1 − D2 , so
ZZ ZZ ZZ
y(x2 + 4y 2 )3/2 dA = y(x2 + 4y 2 )3/2 dA − y(x2 + 4y 2 )3/2 dA (+2 points)
D D1 D2
q q
4−x2 1−x2
Z 2Z 4
Z 1 Z 4
2 2 3/2
= y(x + 4y ) dydx − y(x2 + 4y 2 )3/2 dydx
0 0 0 0
r
c − x2
2 2
∵ x + 4y = c, y ≥ 0 ⇔ y= (+2 points)
4
q q
2 2
Z 2 y= 4−x Z 1 y= 1−x
1 2 1 2
4 4
= (x + 4y 2 )5/2 dx − (x + 4y 2 )5/2 dx
0 20 0 0 20 0
Z 2 Z 1
1 1
= (32 − x5 ) dx − (1 − x5 ) dx
20 0 20 0
1 160 5
= −
20 3 6
21
= . (+4 points)
8
where W is the region bounded above by the paraboloid z = 4 − (x2 + y 2 ) and below
p
by the cone z = 3 x2 + y 2 .
Solution. Let x = r cos θ, y = r sin θ, and z = z for the cylindrical coordinates, then by the change
of variables, we get
cos θ −r sin θ 0
∂(x, y, z)
= sin θ r cos θ 0 =r (+2 points)
∂(r, θ, z)
0 0 1
• No partial point for integration domain and Jacobian on the change of variables.
x
12 Let V be the finite solid bounded by x2 + y2 = 4, z = x + 2, z = 2 − 1. Evaluate
RR
16 points ∂V
(y − z)dS.
Solution. Easy to see that ∂V is union of S1 , S2 , S3 , respectively top, side and bottom face of
the truncated cylinder, with their edges overlapped; each being subset of graph of z = x + 2,
x
x2 + y 2 = 4, z = 2 − 1 each.
WARNING.
• This problem asks you to compute scalar surface integral, so NO CREDIT UNLESS you
actually START CALCULATION
• NO CREDIT for triple integral, vector integral, divergence theorem, Stoke’s theorem, or any
such irrelevant work
RR
Solution 1. By symmetry, easy to see that ∂V
ydS = 0. (+2 points)
S1 and S3 are inclined ellipses and have symmetry so denote z¯1 and z¯3 be the z coordinate of
center of them.
ZZ
zdS = z¯1 · Area(S1 ) (+3 points)
S1
√
= 2 · Area(S1 ) = 8 2π (+2 points)
and
ZZ
zdS = z¯3 · Area(S3 ) (+3 points)
S3
√
= −1 · Area(S3 ) = −2 5π. (+2 points)
Solution 2. On the other hand, one can enjoy computing every surface integral by calculating
everything using definitions only.
For the calculation, 4 points for the side face S2 , and 5 points for each top and bottom, S1
and S3 , and 2 points bonus to final answer for those without symmetry argument at all and thus
16 points total for correct answers.
Then one has to choose parameterization for each face. For S1 and S3 , for example, parametriza-
tion as graph from D, where D is projection of the cylinder onto xy plane, so disk with radius
2.
ZZ ZZ
(y − z)dS = (y − x − 2)||N||dxdy
∂S1 D
ZZ √
= (y − x − 2) 2 dxdy (+3 points)
D
√
= −2 2 · Area(D)
√
= −8 2π (+2 points)
ZZ ZZ
(y − z)dS = (y − x/2 + 1)||N||dxdy
∂S3 D
ZZ √
5
= dxdy (+3 points)
D 2
√
= 2 5π (+2 points)
Then
I Z 2π
1
F · ds = (a cos t + a sin t, a sin t − a cos t) · (−a sin t, a cos t) dt
Ca 0 a2
Z 2π
= −(sin2 t + cos2 t) dt = −2π. (+5 points)
0
• If you made any minor mistake, you will get (−1 point) for each.
(b) Let simple closed path C turning counter-clockwise around the origin be given. For sufficiently
small a > 0, the circle Ca defined in (a) could be placed inside C. Let R be the region bounded
by C and Ca . Notice that, on R,
x2 − y 2 − 2xy x2 − y 2 − 2xy
∂ y−x ∂ x+y
− = − = 0. (+3 points)
∂x x2 + y 2 ∂y x2 + y 2 (x2 + y 2 )2 (x2 + y 2 )2
By Green’s theorem,
y−x
I ZZ
∂ ∂ x+y
F · ds = − dxdy = 0. (+4 points)
∂R R ∂x x2 + y 2 ∂y x2 + y 2
Therefore, by (a),
I I I
F · ds = F · ds + F · ds = 0 − 2π = −2π. (+4 points)
C ∂R Ca
• If you misunderstood the region R, you will get (−2 points), at most (−4 points).
14 Let S be the boundary of the solid bounded by the graphs f (x, y) = 2 − x2 − y2 and
16 points g(x, y) = x2 + y 2 . Orient S with outward-pointing normal. Find the flux of the vector
field F = xi + yj + z 2 k across S, using the standard parametrizations given by the
graphs.
Solution 1. Let D = (s, t)|s2 + t2 ≤ 1 .
Define surface and parametrization for f (x, y) and g(x, y). For (s, t) ∈ D,
Then, we can find normal vector of each surface with given parametrization.
Now,
‹ ‹
F · dS = 2s2 + 2t2 + (2 − s2 − t2 )2 dS (+2 points)
S1 D
ˆ 1 ˆ √
1−s2
= √ 2s2 + 2t2 + (2 − s2 − t2 )2 dtds (+1 point)
−1 − 1−s2
ˆ 2π ˆ 1
= 2
(2r + (2 − r2 )2 )r drdθ
0 0
ˆ 1
1 1 10
= 2π (r5 − 2r3 + 4r)dr = 2π( − + 2) = π, (+2 points)
0 6 2 3
‹ ‹
F · dS = 2s2 + 2t2 − (s2 + t2 )2 dS (+2 points)
S2 D
ˆ 1 ˆ √
1−s2
= √ 2s2 + 2t2 − (s2 + t2 )2 dtds (+1 point)
−1 − 1−s2
ˆ 2π ˆ 1
= 2
(2r − r4 )r drdθ
0 0
ˆ 1
1 1 2
= 2π (2r3 − r5 )dr = 2π( − ) = π. (+2 points)
0 2 6 3
Thus, we have
‹ ‹ ‹
10 2
F · dS = F · dS + F · dS = π + π = 4π.
3 3
S S1 S2
• You can use other parametrization(ex. (s cos t, s sin t, f (s cos t, s sin t))). It will follow the
above criteria.
• In problem, we said ’using the standard parametrizations’. If you use Gauss’s theorem, the
maximum score is 10pts.
Solution 2. Let D = {(x, y, z)|g(x, y) ≤ z ≤ f (x, y)}. F is class C 1 in domain D. Using Gauss’s
theorem, we can find the flux of F as follows:
" ˚
F · dS = ∇ · FdV (+2 points)
S D
˚
= (2 + 2z) dV. (+3 points)
D
15 Let
6+10
x 4y 5z
points F(x, y, z) = + 3y, 2 + 3x, 2 −1 .
x2 + 4y 2 + 5z 2 x + 4y 2 + 5z 2 x + 4y 2 + 5z 2
Solution 1 of (a). We can find that ∇ × F = 0 (+2 points). Note that the domain of F is
R3 − {(0, 0, 0)}, which is simply connected (+4 points). Thus, F is conservative.
Solution 2 of (a). We attempt to find f satisfies F = ∇f (+2 points). Indeed one of such f is
given by
1
f (x, y, z) = ln(x2 + 4y 2 + 5z 2 ) + 3xy − z. (+4 points)
2
16 Let
16 points S = {(x, y, z) ∈ R3 : x2 + (z − 1)2 = 1 and 4 − x2 − z 2 ≤ y ≤ 4}
be the oriented surface with the unit normal vector n = −k at point (0, 2, 2), and
let C be the boundary curve of S that is oriented consistently with S. Let x(t) be
an orientation-preserving parametrization of the portion of C from point (0, 0, 2) to
(−1, 2, 1). For each (x, y, z) ∈ R3 , let
M (x, y, z) = −y 2 ex+z
2
N (x, y, z) = 2(zey − yex )ez + yey
P (x, y, z) = (2z + 2)ey+z − y 2 ex+z .
Solution 1. Let us find a scalar potential function of F such that F = ∇f . (+2 points)
Since
∂f
= M = −y 2 ex+z , (+2 points)
∂x
∂f
we integrate ∂x by x to get f (x, y, z) = −y 2 ex+z + g(y, z). Thus, we obtain
∂f ∂g 2
= −2yex+z + = N = 2(zey − yex )ez + yey . (+2 points)
∂y ∂y
∂g 2
Now integrate ∂y by y to get g(y, z) = 2zey+z + 12 ey + h(z). We have
∂f
= −y 2 ex+z + 2ey+z + 2zey+z + h0 (z) = P = (2z + 2)ey+z − y 2 ex+z . (+2 points)
∂z
and h0 (z) = 0 implies that h is constant. We then obtain the scalar potential function,
1 2
f (x, y, z) = −y 2 ex+z + 2zey+z + ey + C. (+2 points)
2
Therefore, F is conservative, and the line integral of F along the path x satisfies
Z
F · ds = f (−1, 2, 1) − f (0, 0, 2) (+4 points)
C
1 4 9
= e + 2e3 − 4e2 − . (+2 points)
2 2
• No partial score if you just write down an incorrect potential function without explanation.
17 Let D = [0, π] × [0, 1] and for each (u, v) ∈ D, let Φ(u, v) = (v cos u, v sin u, u). Let
√ √
16 points S = Φ(D) be oriented with the normal field n such that n = (0, 1/ 2, 1/ 2) at
A(−1, 0, π). Let C be the simple curve from B(1, 0, 0) to A along the boundary curve
of S that is oriented consistently with S. For each (x, y, z) ∈ R3 , let
and ZZ Z Z Z
curl F · dS = F · ds = − F · ds + F · ds, (+2 points)
S ∂S C C1
where C1 is a curve given by a union of the following three paths: