Anda di halaman 1dari 48

[1] (8 points) Find the equation of the tangent line to the path c(t) = (cos t, sin t, t2 ) at t = .

Solution. The derivative of this path is given by


c! (t) = (sin (t) , cos (t) , 2t)
The tangent line to c (t) at the point t = is given by
L (t) = c () + t c! ()
! "
= cos () , sin () , 2 + t (sin () , cos () , 2)
! "
= 1, 0, 2 + t (0, 1, 2)
! "
= 1, t, 2 + 2t

(Or you may use L (t) = c () + (t ) c! ().)

[2] (8 points) Let f : R3 R be defined by:


2 y 2
f (x, y, z) = 3z + ex

Let C be the set of the heads of unit vectors v in R3 such that f increases at 1/3 of its maximum
rate of change in the direction v starting from (0, 0, 1) . Find the equation(s) which determine(s)
the set C. (Hint : C is a circle in R3 .)

Solution. The gradient of f is given by


! ! " ! " "
f = 2x exp x2 y 2 , 2y exp x2 y 2 , 3
f (0, 0, 1) is a vector in the direction of greatest increase for the function f at the point (0, 0, 1).
So
f (0, 0, 1) = (0, 0, 3)
= 3 (0, 0, 1)

Calculate the maximimum rate of change of f by evaluating the directional derivative of f .


Df(0,0,1) = f (0, 0, 1)
= (0, 0, 3) (0, 0, 1)
= 3

1/3 of this rate is just 1, so the goal of this problem is to find and equation for the unit vectors
v = (x, y, z) such that Dfv = 1 at the point (0, 0, 1).

1 = Dfv
= f v
= (0, 0, 3) (x, y, z)
= 3z
1/3 = z

2
Now applying the constraint that v be a unit vector
vv = 1
(x, y, 1/3) (x, y, 1/3) = 1
x2 + y 2 + 1/9 = 1
x2 + y 2 = 8/9

The final equations are : x2 + y 2 = 8/9 and z = 1/3.

[3] (8 points) Find the absolute minimum and maximum for the function f (x, y) = x + 2y 2 + 1 on
the unit disk D = {(x, y)| x2 + y 2 1}.

Solution. On the interior of D, which is {(x, y)| x2 + y 2 < 1}, if (x, y) is a critical point of f ,
f (x, y) = (0, 0). But f (x, y) = (1, 4y) %= (0, 0). Hence f has its extrema on D.
On D = {(x, y)| x2 + y 2 = 1}, we use Lagarange Multiplier. Let g(x, y) = x2 + y 2 , and solve
f (x, y) = g(x, y) and x2 + y 2 = 1 simultaneously. Then we have 1 = 2x, 4y = 2y, and
x2 + y 2 = 1.
If y = 0, x = 1, hence we have (x, y) = (1, 0), (1, 0). If y %= 0, then = 2 and x = 14 , hence

y= 4 .
15
Thus (x, y) = ( 14 , 4
15
), ( 1
4 , 15
).
4
Now f (1, 0) = 2, f (1, 0) = 0, f ( 4 , 4 ) = f ( 14 , 415 )
1 15
= 8 .
25
Therefore f has the absolute

maxima ( 14 , 415 ) and ( 14 , 415 ), and the absolute minimum (1, 0).

[4] (8 points) Let C be a simple closed curve in R3 , and c : [a, b] R3 be a parametrization of C


such that ||c! (t)|| = 1 for all t [a, b]. The curvature of C at c(t) is defined to be
(t) = ||c!! (t)||
and the total curvature of C is defined by
#
ds.
c

Compute the total curvature of a circle C = {(x, y, z) R3 | x2 + y 2 = R2 , z = 0}.

Solution. Let c(s) = (Rcos(s/R), Rsin(s/R), 0), s [0, 2]. Note that
c! (s) = (sin(s/R), cos(s/R), 0)

is of unit length for all s. We have c!! (s) = ( cos(s/R)


R , sin(s/R)
R , 0) and (s) = ||c!! (s)|| = 1/R for
all s. Hence the total curvature is
# # 2
1 1 1
ds = ds = length(C) = 2R = 2.
c 0 R R R

[5] (8 points) Let C be the ellipse


x2 y 2
+ 2 = 1.
a2 b
3
#
Evaluate the line integral F ds where F(x, y) = (x/a2 , y/b2 ).
C

Solution. Method I: Consider the region

x2 y 2
S = {(x, y, ) | + 2 1}.
a2 b
By Stokes theorem, we have
# ##
F ds = F dS.
C S

But F = 0, hence the integral is 0.

Method II: The ellipse can be parametrized by

c(t) = (a cost, b sint), t [0, 2].

Since we have
cost sint
F(c(t)) c! (t) = (a sint) + b cost = 0,
a b
the line integral is zero.

#
2
[6] (8 points) Evaluate the line integral ex dx xy dy + y 2 dz, where c(t) = (1, t, t2 ), 0 t 1.
c
# # 1 # 1
2
Solution. e dx xy dy + y dz =
x 2
(e 0 1 t 1 + t 2t) dt =
2
(t + 2t3 ) dt =
$ %c 0 0
1 2 2 4 1
t + t = 0.
2 4 0

[7] (8 points) Find the area of the portion of the unit sphere inside the cylinder x2 + y 2 = 12 and
z > 0.
Solution. The intersection of the unit sphere and the cylinder is a circle, and the angle between
the z-axis and a line from the origin to the circle is 4 . Denote the surface by S. We parametrize
S by (, ) = (sin cos , sin sin , cos ) where 0 4 , 0 2. Then || || = sin .
## # 2#
4 /4
Hence dS = sin d d = [cos]0 2 = (2 2).
S 0 0

! 3 (8 points)
[8] W is the "volume defined by x2 + y 2 + z 2 1 and y x. Find the flux of
x 3x, y + xy, z 3 xz out of W .
3

Solution. Calculate that

F = 3x2 3 + 3y 2 + x + 3z 2 x
! "
= 3 x2 + y 2 + z 2 1

4
By Gauss Theorem the flux out of W is given by
# # # # #
F dS = F dW
W W
# # #
= 3 x2 + y 2 + z 2 1dxdydz

In spherical coordinates, the defining equations of the volume W are 2 1 and sin cos.
The latter inequality implies 3/4 /4.
The Jacobian determinant for spherical coordinates is 2 sin, so the integral is just:

# # # # 1 # /4 # ! 2 "
3 x + y + z 1dxdydz = 3
2 2 2
1 2 sinddd
0 3/4 0
= 4/5

[9] (9 points) Let S be the


# # surface of the tetrahedron whose vertices are (1, 0, 0), (0, 1, 0), (0, 0, 2)
and the origin. Evaluate f dS where f (x, y, z) = xz.
S

Solution. Let S!, S !! , S !!! ,


T denote the faces of the tetrahedron in the xz-plane, yz-plane, xy-
plane, and the plane 2x + 2y + z = 2 (the frontal face), respectively. Since x = 0 on S !! and z = 0
on S !!! , it follows that the integral vanishes on those faces. Hence we have
## ## ##
xz dS = xz dS + xz dS.
S S! T
(1)
## # 1 # 1x
xz dS = xz dz dx
S! #0 1 0

= x(1 x)2 /2 dx = 1/24.


0
(2) T is the graph of z = 2 2x 2y over the triangle with vertices (1, 0, 0), (0, 1, 0), and
the origin. Let g(x, y) = 2 2x 2y. Then we have
## ## &
xz dS = xg(x, y) 1 + gx2 + gy2 dA
T # 1# 1x
= x(2 2x 2y)3 dy dx
#0 1 0
= 3x(1 x)2 dx = 1/4.
0
##
(3) xz dS = 1/24 + 1/4 = 7/24.
S

##
[10] (9 points) Evaluate F dS where
S
' ! "2 (
F (x, y, z) = 1, 1, z x2 + y 2

5
and S is the surface of the cylinder x2 + y 2 1, 0 z 1 (oriented with the outward unit normal).

Solution. To use Gauss Theorem, write W = S and so


# # # #
F dS = F dS
S W
# # #
= div (F ) dW
W

! "2
div (F ) = 0 + 0 + x2 + y 2
! "2
= x2 + y 2

The Jacobian determinant of cylindrical coordinates is r, so


# # # # # #
! 2 "2
div (F ) dW = x + y 2 dW
W W
# 1 # 2 # 1
! 2 "2
= r rdzddr
0 0 0
= /3

[11] (9 points) Let S be the portion


## of the unit sphere centered at the origin that is cut out by the
)
cone z x2 + y 2 . Evaluate F dS where
S

F(x, y, z) = (xy + cos z, yx + x2 + z 3 , 2z 2 + x).


Solution. The boundary C of S is the circle of radius 1/ 2 on z = 1/ 2 and with its center on
the z-axis. Let D be the disc in the plane z = 1/ 2 bounded by C, and T = S D. Then we have
## ## ##
F dS = F dS F dS.
S T D

(1) T
** is a surface without boundary and we can apply Gauss theorem to the first integral
T F dS. Let W be the (3-dimensional) region bounded by T .
** ***
T F dS = ***W F dV
= W y x + 4z dV.
*** ***
But W y dV = W x dV = 0 because of the apparent symmetry.
*** * 2 * 1 * 1z 2
W 4z dV = 0

1/ 2 0 4z r dr dz d
*1 1z 2
= 8 1/2 z dz = .
2 4
(2) Since n = k on D, we have
** **
D F dS = D **F (k) dS
= D 2z 2 + x dA.
6
**
But D x dA = 0 because of symmetry.
** **
D 2z 2 dA = **D 2 12 dA
= D dA = /2.
**
(3) S F dS = /4 (/2) = 3/4.

##
[12] (9 points) Evaluate (F) dS, where F(x, y, z) = (xy, y, z) and S is the surface described
S
by x2 + y 2 + z 2 = 2 and z 1. (S is oriented with the upward unit normal.)
Solution.
## S is a circle
# with the equations x + y = 1 and z = 1. Use Stokes Theorem.
2 2

( F) dS = F ds. We parametrize S by c(t) = (cos t, sin t, 1), 0 t 2.


# S # 2 S # 2
1
F ds = (cos tsin t(sin t) + sin tcos t + 1 0) dt = (sin2 tcos t + sin 2t) dt =
2
$ S 0%
2
0
1 3 1
sin t cos 2t = 0.
3 4 0

7
Problem 1) True/False questions (20 points). No justifications are needed.

1) T F The surface x2 + y 2 z 2 = 1 is a one-sheeted hyperboloid.

Solution:
Look at the xy-trace.

~ w~ (~v ) of a vector ~v onto a non-zero vector w


The vector projection proj ~ is
2) T F always non-zero.

Solution:
The two vectors can be perpendicular

The linearization of f (x, y) = 5 + 7x + 3y at any point (a, b) is the function


3) T F L(x, y) = 5 + 7x + 3y.

Solution:
A linear function has the function as a linearization.

For any function f (x, y, z), for any unit vector ~u and any point (x0 , y0 , z0 )
4) T F we have D~u f (x0 , y0 , z0 ) = D~u f (x0 , y0 , z0 ).

Solution:
Changing the direction of ~u changes the sign by definition Du f = f ~u.

5) T F There is a vector field F~ = hP, Q, Ri such that curl(F~ ) = div(F~ ).

Solution:
The right hand side is a scalar while the left hand side is a vector.

The formula |~v w|


~ = |~v ||w|
~ sin() holds if ~v , w
~ are vectors in space and
6) T F is the angle between them.

Solution:
Yes, this is an important formula for the length of the cross product.
If F~ = h2y, 2xi and C is the circle x2 + y 2 = 4 oriented counterclockwise,
7) T F
then C F~ d~r = 16.
R

Solution:
Just write out the integrand F~ (~r(t))~r 0 (t) = h4 cos(t), 4 sin(t)ih2 sin(t), 2 cos(t)i = 8.
When integrated from 0 to 2, we get 16.


The parametrization ~r(u, v) = hu, 1 u2 v 2 , vi, u2 + v 2 1 describes a
8) T F half sphere.

Solution:
Indeed, the entries x, y, z satisfy x2 + y 2 + z 2 = 1

9) T F The vectors ~v = h1, 0, 1i and w


~ = h1, 1, 1i are perpendicular.

Solution:
Their dot product zero.

If div(F~ )(x, y, z) = 0 for all (x, y, z) then C F~ d~r = 0 for any closed curve
R
10) T F
C.

Solution:
It would be true if div were replaced by curl.

11) T F The vector field F~ = hex , ey , ez i satisfies grad(div(F~ )) = F~ .

Solution:
Indeed. Just compute it and use that (ex )0 = ex .

12) T F If F~ (x, y, z) has zero curl everywhere in space, then F~ is a gradient field.

Solution:
This is a result which follows from the Stokes theorem.
3
If ~r(u, v) is a parametrization of the surface g(x, y, z) = x2 + ey + z 2 = 5
13) T F then for any u and v we have g(~r(u, v)) ~ru (u, v) = 0.

Solution:
This is true for any surface g(x, y, z) = c. The reason is that the gradient of g is perpen-
dicular to the level surface and that ~ru is tangent to the surface.

14) T F The equation grad(div(grad(f ))) = h0, 0, 0i always holds.

Solution:
A simple example is f (x, y, z) = x3 .

There is a non-constant function f (x, y, z) such that grad(f ) =


15) T F curl(grad(f )) everywhere.

Solution:
The right hand side is 0. So, f has to be constant.

If the vector field F~ has constant divergence 1 everywhere, then the flux
16) T F of F~ through any closed surface S oriented outwards is the volume of the
enclosed solid.

Solution:
By the divergence theorem.

If f (x, y) is maximized at (a, b) under the constraint g(x, y) = c, then


17) T F f (a, b) and g(a, b) are parallel.

Solution:
They indeed are by the Lagrange equations

The distance between a point P and the line L through two different points
18) T F A, B is given by the formula |P~A AB|/|
~ P~A|.

Solution:
It would be true if P~A were replaced by AB.
~
19) T F The unit tangent vector T~ (t) is always perpendicular to the vector T~ 0 (t).

Solution:
We have shown that in class. It follows from differentiating T T = 1.

The vector field F~ (x, y, z) = hx5 , x6 , x7 i can not be the curl of another vector
20) T F
field.

Solution:
Its divergence is not zero. So that it can not happen
Problem 2) (10 points) No justifications are necessary.

a) (2 points) Match the following surfaces. There is an exact match.


A B C D
Surface ~r(u, v) = A-D
hu, u2 v, v 2 i
hu2 cos(v), u, u2 sin(v)i
hcos(v), sin(u), sin(v)i
hv cos(u), v sin(u), sin(v)i

b) (2 points) Match the following 2D region plots. There is an exact match.


A B C D
Region A-D
1 x2 4, 1 y 2 4
1 x2 + y 2 4
1 x4 + y 4 4
1 |x| + |y| 2

c) (2 points) Match the following 3D regions. There is an exact match.

Solid A-D A B C D
|x| + |y| + |z| 2
|x| < |y| < |z|
1 < |xyz| < 2
x2 y 2 < z 2

d) (2 points) The following figures display vector fields. There is an exact match.

Field A-D A B C D
~
F (x, y) = h0, yi
F~ (x, y) = hx, y 3 i
F~ (x, y) = hy 3 , xi
F~ (x, y) = hx, y 3 i

e) (2 points) The following figures display polar regions. There is an exact match.

Polar region A-D A B C D


r 1 + cos(3)
r 2 + sin(3)
r 3 sin(3)
r | sin(3)|
Solution:
a) BDCA
b) DCBA
c) BCAD
d) BADC
e) DCAB

Problem 3) (10 points)

a) (6 points)

The concept of boundary plays an important role in integral theorems. In each of the following
six rows, check exactly one entry which best describes the boundary.

The boundary of solid surface curves points empty

x2 + y 2 + z 2 = 1

x2 + y 2 = 1, z = 0

x2 + y 2 + z 2 1, x = y = 0

x2 + y 2 + z 2 1

x2 + y 2 1, z = 0

x2 + y 2 = 1, z 2 1

b) (4 points) Match the following partial differential equations (PDEs) by picking 4 from the 5
given choices A-E.
PDE Enter A-E uxx = ut A

heat equation uxx = ut + uux B

wave equation ux = ut C

transport equation uxx = utt D

Burgers equation uxx = utt E

Solution:
a) empty, empty, points, surface, curves, curves.
Here is some more explanation:
The first is a closed surface which has no boundary.
The second is a closed curves which has no boundary.
The third is a line segment which has two points as a boundary.
The fourth is the unit ball which has the unit sphere, a surface, as a boundary.
The fifth is a closed disk, which has a circle, a curve as a boundary.
The sixth is a finite cylinder, which has two circles, two curves, as a boundary.
b) A,E,C,B.

Problem 4) (10 points)


a) (3 points) A surface S is parameterized by
~r(u, v) = hu, v, uvi ,
where u2 + v 2 1. Find its surface area.

b) (3 points) Parametrize the boundary curve


C matching the orientation ~rRu ~rv of S,
then compute the line integral C F~ d~r with
F~ (x, y, z) = hy, x, 1i.

c) (2 points) The coordinates of the surface S


satisfy xy z = 0. Find the tangent plane to S at
(2, 1, 2).

d) (2 points) Find the linearization L(x, y) of


f (x, y) = xy at the point (2, 1).

Solution: R 2 R 1
2 1 2
a) Since |~ru ~rv | = u + v + 1 we have in polar coordinates 0 0 1 + r rdrd =
2/3( 8 1).
b) The boundary curve is ~r(t) = hcos(t), sin(t), cos(t) sin(t). The line integral can be
computed directly as ~r 0 (t) = h sin(t), cos(t), cos2 (t) sin2 (t)i. which gives F~ (~r(t))
~r 0 (t) = 1 and the integral is 2 .
P.S. Some have computed it with Stokes using curl(F~ ) = h0, 0, 2i and ~ru ~rv = hv, u, 1i
as computed in a). We have to integrate the function 2 over the disk of radius 1. This is
twice the area of the disk, which is again 2 .
c) The gradient vector is hy, x, 1i which is h1, 2, 1i so that the equation of the plane
is x + 2y z = d. The constant can be obtained by plugging in the point. This is
x + 2y z = 2 .
d) The gradient of f is hy, xi. At the point (2, 1) this is h1, 2i. The linearization is
L(x, y) = f (2, 1) + 1(x 2) + 2(y 1) = 2 + (x 2) + 2(y 1)

Problem 5) (10 points)

On November 17 2016, the NASA Eagleworks paper appeared, making the EM drive more
probable. It might in future be used for deep space missions. The drive produces a thrust,
apparently violating momentum conservation.
a) (5 points) Assume the drive flies in the gravita-
tional field
F~ (x, y, z) = hx7 + xy 2 z 2 , x2 yz 2 , x2 y 2 zi
along the path

C : ~r(t) = ht cos(t), t sin(t), t(5 t)i


with 0 t 5. Compute the work
Z 5
F~ d~r.
0
R 5
b) (3 points) Compute d = | 0 ~r 0 (t) dt|.

c) (2 points) If L is the arc length of C, circle the one box below which applies:

d=L d>L d<L

Solution:
a) The field is conservative. There is a potential f (x, y, z) = x2 y 2 z 2 + x8 /8. We can by
the fundamental theorem of line integral, evaluate
R 0
f (5, 0, 0) f (0, 0, 0) = (5)8 /8.
b) By the fundamental theorem of calculus, ~r (t) dt = ~r(5) ~r(0) = h5, 0, 0i. Its
length is 5 .
c) The arc length of the curve is longer than the shortest connection, the line which has
been computed in b). Note that the arc length has the length | | inside the integral. We
have
d=L d>L d<L

Problem 6) (10 points)


a) (2 points) Find the equation ax + by + cz = d of
the plane through A = (1, 1, 1), B = (3, 4, 5), C =
(4, 4, 2).

b) (3 points) Compute the area of the parallelo-


~ and AC.
gram spanned by AB ~

c) (3 points) Determine the volume of the


~ AC,
parallelepiped spanned by AB, ~ AP
~ where
P = (1, 3, 4).

d) (2 points) Find the distance |P~Q|, where Q


is the mirror image of P opposite of the plane.
It is determined by the fact that the middle
point (P + Q)/2 is on the plane and that P~Q is
perpendicular to the plane.

Solution:
a) The normal vector is ~n = AB~ AC ~ = h2, 3, 4i h3, 3, 1i = h9, 10, 3i. The equa-
tion is 9x 10y + 3z = d. The constant is obtained by plugging in a point. It is
9x 10y + 3z = 2 .

b) The area of the parallelogram is the length of the vector ~n = h9, 10, 3i = 190 .
~ ~n = 11 .
c) The volume of the parallel epiped is AP
d) The distance is volume divided by area which is 11/ 190. The distance between P

and Q is 22/ 190 .

Problem 7) (10 points)


The triple scalar product is also written as

~ = ~u (~v w).
[~u, ~v , w] ~
The torsion of a space curve is defined as
[~r 0 , ~r 00 , ~r 000 ]
.
|~r 0 ~r 00 |2
a) (3 points) Compute ~r 0 (0), ~r 00 (0), ~r 000 (0) for
~r(t) = hcos(t), sin(t), ti.
b) (4 points) Compute the torsion of the curve at
the point ~r(0).

c) (3 points) Assume you have an arbitrary curve


~r(t) which is contained in the xy-plane. What is
its torsion?

Solution:
a) ~r 0 (0) = h0, 1, 1i.
~r 00 (0) = h1, 0, 0i.
~r 000 (0) = h0, 1, 0i.
b) The triple scalar product is 1. The cross product |~r 0 ~r 00 |2 = |h0, 1, 1i|2 = 2. The
torsion is 1/2 .
c) If ~r(t) = hx(t), y(t), 0i, then ~r 0 (t) ~r 0 (t) = h0, 0, x0 y 00 x00 y 0 i. On the other hand,
~r 000 (t) = hx000 (t), y 000 (t), 0i. The dot product is zero .
P.S. Also more intuitive explanations worked: the three derivatives are all in the plane.
They span a parallelepiped of zero volume. Therefore, the torsion, which has the volume
in the nominator, is zero.

Problem 8) (10 points)


Let E be the solid
x2 + y 2 z 2 , x2 + y 2 + z 2 9, y |x|.
a) (7 points) Integrate
ZZZ
x2 + y 2 + z 2 dxdydz.
E

b) (3 points) Let F~ be a vector field

F~ = hx3 , y 3 , z 3 i.

Find the flux of F~ through the boundary surface


of E, oriented outwards.

Solution:
a) The region is best described in spherical coordinates. The angle goes from /4 to
3/4. The angle goes from /4 to 3/4. The radius goes from zero to 3. The integral
is Z 3/4 Z 3/4 Z 3
2 2 sin() ddd .
/4 /4 0

The answer is (/2)(35 /5) 2 = 243 2/10 .
b) Since the divergence is 3x2 + 3y 2 + 2z 2 the result is just three times the result found

in a). It is 729 2/10 .

Problem 9) (10 points)

The vector field

~ y, z) = hy, x, 0i
A(x,
(x2 + y 2 + z 2 )3/2
is called the vector potential of the magnetic
field
~ = curl(A).
B ~
The picture shows some flow lines of this mag-
~ Find the flux of B
netic dipole field B. ~ through
2 2 2
the lower half sphere x + y + z = 1, z 0 ori-
ented downwards.
Solution:
Since we have an integral of the curl of the vector field A, ~ we use Stokes theorem and
~
integrate A(~r(t)) along the boundary curve ~r(t) = hcos(t), sin(t), 0i. First of all, we
have A(~~ r(t)) = hsin(t), cos(t), 0i. The velocity is ~r 0 (t) = h sin(t), cos(t), 0i. The integral
R 2
is 0 1 dt = 2. The answer is 2 .

Problem 10) (10 points)

a) (8 points) Classify the critical points of the area


51 function
f (x, y) = x51 51x y 51 + 51y

using the second derivative test. The reason why


this function was chosen is classified.

b) (2 points) Does the function have a global max-


imum or global minimum on the region x2 +y 2 1
including the boundary? Write yes or no with
a brief explanation. There is no need to find the
global extrema.

Solution:
a) Setting the gradient f = h51x50 51, 51y 50 51i to h0, 0i gives the solutions x =
1, y = 1 and so the solutions (1, 1), (1, 1), (1, 1), (1, 1). We have fxx = 50 51x49
and fyy = 50 51y 49 and fxy = 0 so that D = fxx fyy fxy 2
= (50 51)2 x49 y 49 . This is
negative if the two coordinates have the same sign and positive else. Furthermore fxx > 0 if
x is positive. Therefore (1, 1), (1, 1) are saddle points and (1, 1) is the maximum
and (1, 1) is the minimum .
b) Yes by Bolzano , the continuous function f has both a maximum and minimum on
the closed disc x2 + y 2 1 as this region is both bounded and closed.

Problem 11) (10 points)


Using the Lagrange method, find the maximum
and minimum of the elliptic curve function

f (x, y) = y 2 x3 x2 x
on the circle g(x, y) = x2 + y 2 = 1.
This problem is motivated from a real life application. To encrypt communication in
WhatsApp, the elliptic curve 25519 given by y 2 = x3 + 486662x2 + x over the
prime p = 2255 19 =

57896044618658097711785492504343953926634992332820282019728792003956564819949

is used.

Solution:
The Lagrange equations are
3x2 2x 1 = (2x)
2y = 2y
x + y2 = 1
2

Eliminating gives 2y(3x2 2x 1) = 2x2y. The first possibility is that y = 0. In that


case, x = 1. If y is not zero, then we can divide it out and get 3x2 2x 1 = 2x which
is equivalent to 3x2 + 4x + 1 = (3x + 1)(x + 1). We had already x = 1 so that the only
other solution is x = 1/3 which gives y = 8/3. Now we have all the critical points.
(1, 0), (1, 0), (1/3, 8/3), (1/3, 8/3). To find the maximum and minimum, we

evaluate the function f at the critical points. The maximum is at (1/3, 8/3) . The
minimum is at (1, 0) .

Problem 12) (10 points)

Given the scalar function f (x, y) = x5 + xy 4 ,


compute the line integral of
F~ (x, y) = h5y + 3y 2 , 6xy + y 4 i + f

along the boundary of the Monster region


given in the picture. There are four boundary
curves, oriented as shown in the picture: a large
ellipse of area 16, two circles of area 1 and 2
as well as a small ellipse (the mouth) of area
3. The picture describes the orientations of the
boundary curves perfectly and they are as they
are! We warn you not to ask about this, or else
we will bring in Mike from Monsters, Inc.
Solution:
The curl is 5. By Greens theorem we just have to compute the right areas. If the eye
of Mike were oriented differently, we would get the area itself which is 16(21)3 = 12
and the answer would be 60. However, since the inner eye is oriented wrong, that area is
counted negatively. So instead of adding 1 we subtract 1 for that line integral contribution.
The answer is (5)[16 2 3 + 1] = 50 .

Problem 13) (10 points)

ProtEgg is a defense spell. It produces an egg


shaped solid E enclosed by the surfaces
S : z = 2 2x2 2y 2 , z 0 ,
where S is oriented upwards and

T : z = x2 + y 2 1, z 0 ,
where T is oriented downwards.
RRR
a) (4 points) Find the volume E 1 dxdydz of E.

b) (4 points) The spell uses a force field

F~ (x, y, z) = h0, 0, 2x2 + 2y 2 + zi .


S is parametrized by ~r(u, v) = hu, v, 2 2u2 2v 2 i
withRRu2 + v 2 1 oriented upwards. Compute the
flux S F~ dS~ without an integral theorem.

c) (2 points) The flux T F~ dS~ can be determined using an integral theorem. What is the
RR

value of the flux? Check all that apply:


A B C D

F~ dS
~ F~ dS
~ F~ dS
~ 1 dV F~ dS
~
RR RR RRR RR RRR RR
S S E 1 dV + S E S
Solution:
a) The region E is described by x2 + y 2 1, x2 + y 2 1 z 2 2x2 2y 2 which is
in polar coordinates:

0 2, 0 r 1, r2 1 z = 2 2r2 .

ZZZ Z 2 Z 1 Z 22r2
1 dx dy dz = r dz dr d
E 0 0 r2 1
Z 2 Z 1
3
= (3 3r2 )r dr d = .
0 0 2

The answer is 3 2
.
b) Since ~ru = h1, 0, 4ui and ~rv = h0, 1, 4vi, we have ~ru ~rv = h4u, 4v, 1i and
ZZ ZZ
F~ dS
~ = F~ (~r(u, v)) (~ru = ~rv ) du dv
S u2 +v 2 1
ZZ
= 2 du dv = 2.
u2 +v 2 1

The answer is 2 .
c) Since S and T bound the solid E, by the divergence theorem we have
ZZ ZZ ZZZ ZZZ
F~ dS
~+ F~ dS
~= divF~ dV = 1 dV.
S T E E

This implies that


ZZ ZZZ ZZZ ZZ
F~ dS
~= divF~ dV = 1 dV = F~ dS.
~
T E E S

Only answer D is correct.


Problem 1) True/False questions (20 points), no justifications needed

The function f (x, y) = x3 y/(x6 + y 5 ) can be filled in at the origin with a


1) T F value f (0, 0) = a so that f is continuous everywhere.

Solution:
Go to polar coordinates. The function diverges at (0, 0).

R1
2) T F The chain rule assures that 0 (f (~r(t)) ~r 0 (t)) dt = f (~r(1)) f (~r(0)).

Solution:
Just integrate the chain rule and use the fundamental theorem of calculus.

R1R1 R1R1
3) T F The formula 0 0 f (x, y) dy dx = 0 0 f (y, x) dy dx holds.

Solution:
Just change the variables x, y first on the left hand side to get 01 01 f (y, x) dx dy, then
R R

use Fubini to get the right hand side.

If u(x, t) solves the partial differential equation ut = ux , then so does the


4) T F function ux .

Solution:
We can use Clairaut to see that.

There is a surface S containing the curve ~r(t) = ht, t2 , t3 i for which the
5) T F tangent plane to S at (0, 0, 0) is x + 2y + 3z = 0.

Solution:
The velocity vector is perpendicular to the surface, not parallel to the surface.

6) T F For any two unit vectors ~u and ~v , and any f , we have D~u D~v f = D~v D~u f .
Solution:
Write down the definitions, we have in both cases fxx u1 v1 + fyy u2 v2 + fxy u1 v2 + fyx u2 v1 .

If the tangent plane to z = f (x, y) at (0, 0, f (0, 0)) is 4 + 3x + 2y + z = 0,


7) T F then L(x, y) = 4 + 3x + 2y is the linearization of f (x, y) at (0, 0).

Solution:
This is fast because L(x, y) = 4 3x 2y

2
8) T F For f (x, y) = x3 ey cos y
x4 cos y the function fxyxyxyxyxy is zero everywhere.

Solution:
Use Clairaut. We differentiate 5 times with respect to x.

9) T F The point (0, 0) is a critical point of f (x, y) = x3 y 2 .

Solution:
There are many critical points but (0, 0) belongs there.

The gradient of f (x, y) = x2 + y 2 is a vector perpendicular to the surface


10) T F z = f (x, y).

Solution:
It is not a 3-vector.

If the function f (x, y) attains an absolute maximum on the region x2 +y 2 4


11) T F at the point (2, 0), then we must have fxx (2, 0) 0.

Solution:
The maximum does not need to be in the interior.

R 2 R 7
If f (x, y) 5 for all values of (x, y), then 0 0 f (r cos , r sin )r dr d
12) T F 5(72 ).
Solution:
The integral is smaller or equal to 02 07 5r dr d which is 5 times the area of the disk.
R R

R a R a  x2 
13) T F For any constant a, we have a 0 e sin y dx dy = 0.

Solution:
Use the symmetry

2 +y
The linearization of the function f (x, y) = ex at the point (0, 0) is the
14) T F 2 2
function L(x, y) = 1 + 2x2 ex +y + yex +y .

Solution:
The linearization is linear


15) T F Let ~u be the unit vector in the direction h1, 1i/ 2. Then D~u f = fxy .

Solution:
The directional derivative only invokes the first derivatives of f .

R 2 R 1
16) T F The integral of f (x, y) = x2 + y 2 over the unit disk is 0 0 r drd.

Solution:
We have forgotten the factor r.

17) T F There is a function f (x, y) for which D~v f (0, 0) = 1 for all directions ~v .

Solution:
Switching the direction gives a negative value.

dy
18) T F Given f (x, y(x)) = 0, then fx + fy dx = 0.
Solution:
This is implicit differentiation

19) T F Any function on a closed and bounded region must have a critical point.

Solution:
Take a linear non-constant function like x + y.

The integral x2 +y2 1 |f (x, y)| dxdy computes the surface area of the surface
RR
20) T F z = f (x, y), x2 + y 2 1.

Solution: q
The surface area is 1 + fx2 + fy2 . In general the areas are different. Take f = 2 for
example, then the integral under consideration is 2 while the surface area is .
Problem 2) (10 points) No justifications needed

RR RR
a) (6 points) Double integrals like R 1 dxdy or R r drd can be interpreted both as the area
of the region R as well as the volume of the solid under the graph of the constant function
f (x, y) = 1 or f (, r) = 1. Match the regions with the integrals:

A B C

D E F
Enter A-F Integral
R 10 R x
0 0 1 dydx
R 10 R /4
0 0 r ddr
R /2 R 20/
0 0 r drd
R /2 R 10
r drd
R010 R y20/
1 dxdy
R010 R010
0 10x 1 dydx

b) (4 points)

You know the Transport, Wave, Heat, or Burgers equation.


Given in a possibly different order, these differential equations are
ut = uxx , ut = ux , utt = uxx , ut + uux = uxx . Check all the boxes
where the given function solves the given PDE.

f (x, t) = x/(1 + t) solves Name f (x, t) = xt solves Name

Burgers Heat

Transport Wave
Solution:
a) The order is DBACF E.
b) x/(1 + t) solves the Burgers equation and xt solves the wave equation.

Problem 3) (10 points)

3a) (7 points) All the parts of this problem refer to the labeled points and the differentiable
function f (x, y) whose level curves are shown in the following plot:

a) At the point , the gradient f has maximal length

b) At the point , fx > 0 and fy = 0

c) At the point , fx < 0 and fy < 0

d) At the point , DD Ef = 0 and fx 6= 0


1 , 1
2 2

e) At the point , f achieves a global min on 4 x 4 and 4 y 4

f) At the point , f = ~0 and fxx < 0

g) At the point , f points straight toward the top of the page.

y
-4 -3 -2 -1 0 1 2 3 4
4 2 2 1 4

7
3
3 U 4 3

9 7 3
2 10 2
V P 8
8 9
10
1 1
T
5
2
0 3
R
6 S 0 x

-4 Y
-1 -1
1

-2 -1 -8 -2
Q
-9
-7 -2
-6
-3 -3
-3 -5
W X
-4 0 -4
-4 -3 -2 -1 0 1 2 3 4
3b) (3 points) Check the cases where
the maximum, minimum or saddle
Critical Point x4 + y 2 xy x2 y 4
point of the function can be established
Maximum
conclusively using the second deriva-
Minimum
tive test. Dont check the box if the test
Saddle point
does not apply, (even if it might be a
sort of minimum, maximum or saddle).

Solution:
a) The answer is Y V W SQP R.
b) Only check Saddle xy. The other cases are critical points but situations where D = 0
and the second derivative test is inconclusive.

Problem 4) (10 points)

a) (4 points) A math candy of the form


f (x, y, z) = 3x2 y 2 + 3y 2 z 2 + 3x2 z 2 + x2 + y 2 + z 2 = 12
is leaning at (1, 1, 1) at the plane tangent to it. Find
that plane.

b) (3 points) Estimate f (1.1, 1.01, 0.98) using lineariza-


tion.

c) (3 points) A fruit fly just dipped some sugar from


the candy at (1, 1, 1) and moves along a path ~r(t) with
constant speed 1 perpendicularly away from the candy.
What is dtd f (~r(t)) at the moment of take-off?
Solution:
a) f (1, 1, 1) gives a normal vector to the tangent plane. As f = h6xy 2 + 6xz 2 +
2x, 6x2 y + 6yz 2 + 2y, 6y 2 z + 6x2 z + 2zi, we have f (1, 1, 1) = h14, 14, 14i. The plane is
14x + 14y + 14z = 42 , where we got the constant on the right hand side by plugging in
the point (1, 1, 1).
b) The linearization L(x, y, z) of f (x, y, z) at (1, 1, 1) is given by L(x, y, z) = f (1, 1, 1) +
f (1, 1, 1)hx1, y1, z1i = 12+14(x1)+14(y1)+14(z1). So f (1.1, 1.01, 0.98)
L(1.1, 1.01, 0.98) = 12 + 1.4 + 0.14 0.28 = 13.26.
c) This is the length of the gradient as we know that this is the directional derivative
in the direction of gradient. The length is 14 3. If you should have forgotten the
interpretation of the length of the gradient, here is the computation: by the chain rule,
we have dtd f (~r(t)) = f (~r(t)) ~r 0 (t). Since the fly moves perpendicularly to the candy
at the unit speed, ~r 0 (t) at the moment of take -off is the unit vector in the direction of
f (1, 1, 1) = h14, 14, 14i, and thus it is 13 h1, 1, 1i. Note that f (1, 1, 1) points outward,
so this is the direction of the flys path.) Hence dtd f (~r(t)) at the moment of take-off is

f (1, 1, 1) 13 h1, 1, 1i = 14 3.

Problem 5) (10 points)

In order to figure out the Egos x and y of the US presi-


dential candidates, we want to minimize the sum of the
perimeter of the letters H and T written in units x and
y if the total area is fixed. The letter H has area 7x2 and
perimeter 16x, the letter T has area 5y 2 and perimeter
12y. Minimize

f (x, y) = 16x + 12y .


under the constraint
g(x, y) = 7x2 + 5y 2 = 2016 .

We dont actually need to know x and y. As political


pundits, we are only interested in the ratio y/x
at the minimum. Find this ratio!

Solution:
Write down the Lagrange equations 16 = 14x, 12 = 10y, then eliminate to get
y/x = 21/20. There is slightly more Ego for T.

Problem 6) (10 points)


With F (x, y, z) = 2x2 + y 2 + z 2 and the surface S
parametrized by ~r(x, y) = h2x, y, 2x2 + y 2 1i, the func-
tion f (x, y) = F (~r(x, y)) giving the value F on S is
f (x, y) = 4x4 + 4x2 y 2 + 4x2 + y 4 y 2 + 1 .

a) (8 points) Find all the critical points of f and classify


them with the second derivative test. Please organize
your work carefully so that we can see your method
and your conclusions easily.

b) (2 points) The minimum could be obtained by mini-


mizing F (x, y, z) on the surface G(x, y, z) = x2 /2 + y 2
1 z = 0. We would then use a method found by some
mathematician. Which one? Just check the name. No
additional work is needed in b).
Fubini Burgers Laplace Lagrange Bolzano Clairaut

Solution:
a) This is a standard problem for exrema without constraint. The gradient of f is
h8x (2x2 + y 2 1) + 16x, 4y (2x2 + 2
y 1) + 2yi.
2 2
We have fxx = 8 + 48x + 8y . There
are three critical points (0, 0), (0, 2/2), (0, 2/2). The first is a saddle with D = 16,
the other two are minima with D = 48 and fxx = 12.
x y D fxx Type
0 0 -16 8 saddle
0 1/
2 48 12 minimum
0 1/ 2 48 12 minimum
b) The problem has been reformulated as a Lagrange problem. Some were crossing
Bolzano but since the surface is not bounded, one can not conclude the existence of
a minimum from Bolzano.

Problem 7) (10 points)


Integrate Z 1Z 1
sin(x5 ) dxdy .
0 (1y)1/4

The figure just shows a fancy plot of the func-

tion sin(x5 ).

Solution:
Change the order of integration to get
Z 1Z 1
sin(x5 ) dy dx .
0 1x4

This simplifies to Z 1
x4 sin(x5 ) dx = (1 cos(1))/5 .
0

Problem 8) (10 points)

Integrate the double integral


ZZ
x2 dxdy ,
R

where R is the region

1 x2 + y 2 4
and
x 0, y 0, y x .
Solution: R /4
Use polar coordinates. We integrate 0 (15/4) cos2 () d. Now use the double angle
formula. We end up with (15 + 30)/32 .

Problem 9) (10 points)

a) (7 points) Compute A = |~r ~r | for the half cylinder


parametrized by
~r(, ) = hcos(), sin(), cos()i .
with 0 /2 and 0 and use this to find
the surface area of the half cylinder

b) (3 points) Compute B = |~r ~r | for the quarter


sphere parametrized by

~r(, ) = hsin() cos(), sin() sin(), cos()i


with 0 /2 and 0 to show that (remark-
ably!) it is the same factor than in part a).
Remark: The fact that the surface area elements A and B are the same has been realized by

Archimedes already. It allowed him to compute the surface area of the sphere in terms of the

surface area of the cylinder.

Solution: R R /2
a) The integration factor is A = sin(). The integral is 0 0 sin() = .
b) The integration factor is again sin(). We did not ask for the integral as the integral
is the same.
P.S. It is remarkable that one can project the sphere onto the cylinder to see that the
areas are the same. You can do that also in a way how Archimedes might have derived it:
compare a thin slice of size dz on each surface. The surface area of the cylinder is 2dz
which integrates to 2, the area of the full cylinder of height 1 and radius 1. At height
z = cos() the radius is r sin() and since the surface is slanted, the length of the cross
section of the piece is dz/ sin() Now the surface area piece is is 2r sin()dz/ sin() =
2dz again.
18.02 Final Exam Solutions
Problem 1.
a) Line L has direction vector v = 1, 2, 3 which lies in P.
To get a point P0 on L take t = 0 P0 = (1, 1, 2).

P0 Q = 1, 1, 2 1, 1, 2 = 2, 0, 0 also lies in P.
A normal to P is


i j k


n = v
P0 Q = 1 2 3 = i(0) j(6) + k(4) = 0, 6, 4.


2 0 0

So, the equation of P is

0(x 1) + 6(y 1) + 4(z 2) = 0 or 6y = 4z = 14 or 3y + 2z = 7.

b) nQ = 2, 1, 1 n
= 1 2, 1, 1, v = 1, 2, 3
v
= 1 1, 2, 3
6 14
Component of n
on v
is
1 3
n =

v (2 + 2 3) =
6
14 84

Problem 2.
a) Direction vector for L: v = 1, 2, 0.
P0 = (0, 0, 1) equation for L:

r = x, y, z = 0, 0, 1 + t1, 2, 0

or
x = t, y = 2t, z = 1.

b) n = normal vector for P = 1, 2, 0 since L P.

P0 = (0, 0, 1) 1(x 0) + 2(y 0) + 0(z 1) or x + 2y = 0.

c) P on L P = (t, 2t, 1) for some t = 0 (part (a))



P = (t, 2t, 1) since then dist(P0 , P ) = dist(P0 , P ) = |t| 5.

Problem
3.
1 0 3
1 1 2 1
2 1
a)
2 1 1
= 1


1
1 1

= 3 3 = 0.

+ 3

1 1
2

1
2
2

b) To get a non-zero solution take the cross-product of any two rows of A2 ; for example

i j k

1 0
3 = 3, 5, 1


2 1 1



3
This implies all solutions to Ax = 0 are x = t 5 .
1
c)
1 0 0
A 1
1 A1 = I3 = 0 1 0
0 0 1

1 0 3
3 p 5 2 1 1 = 3 2p 5 82p = 0 p = 4.
1 1 1

Problem 4.
a) r (t) = sin(et )et , cos(et )et , et |r (t)| = et 1 + 1 = et 2
r (t) 1
T(t) = = sin(et ), cos(et ), 1.
|r (t)| 2
b)
1 et
T (t) = cos(et ), sin(et ), 0 = cos(et ), sin(et ), 0.
2 2

Problem 5.
F xz
a) Fx = = 2
Fx (1, 3, 23) = 2/2 = 1.
x (x + y)1/2
F z 2

Fy = = 2 1/2
+ Fy (1, 3, 2) = 3/2.
y 2(x + y) z
F 2y 1
Fz = = (x2 + y)1/2 2 Fz (1, 3, 2) = .
z z 2
3 1
n = F (1, 3, 2) = 1, , , P0 = (1, 3, 2)
2 2

tangent plane equation

3 1
1(x 1) + (y 3) + (z 2) or 2x + 3y + z = 13.
2 2

b) At P0 = (1, 3, 2) we have |Fy | = 3/2 > |Fx |, |Fz |. So, a change in y produces the
largest change in F .
3
F = Fy y = (0.1) = 0.15.
2

df 1 1 1
c) =u F (P0 ) = 2, 2, 1 1, 3/2, 1/2 = (2 + 3 1/2) = .
ds P0 ,u
3 3 6

dF 1
F s 0.1 = s s = 0.6
ds
P0 ,u
6

Problem 6. a)

fx = 1 2/(x2 y) = 0 x2 y = 2

x = 4y
fy = 4 2/(xy 2 ) = 0 xy 2 = 1/2

1 1 1
4y 3 = y3 = y= x = 2.
2 8 2
There is one critical point at (x, y) = (2, 1/2).

b) fxx = 4/(x2 y), fyy = 4/(xy 3 ), fxy = fyx = 2/(x2 y 2 )

A = fxx (2, 1/2) = 1, C = fyy (2, 1/2) = 16, B = fxy (2, 1/2) = 2

AC B 2 = 12 > 0, A > 0 f has a relative minimum at (2, 1/2).

Problem 7.
f (x, y, z) = dist2 = (x x0 )2 + (y y0 )2 + (z z0 )2
subject to g(x, y, z) = Ax + By + Cz = D.
f = 2x x0 , y y0 , z z0 , g = A, B, C.

2(x x0 ) = A
2(y y0 ) = B
f = g, and g = D
2(z z0 ) = C
Ax + By + Cz = D.

Problem 8. a)
F F x F y F z
= + +
x y z

= cos cos x (2, /4, /4) = 2 cos(/4) cos(/4) = 1.



y
= cos sin y (2, /4, /4) = 2 cos(/4) sin(/4) = 1.

z
= sin z (2, /4, /4) = 2 sin(/4) = 2.

b) NOT Possible

y, x is not a gradient eld. (Test: y, x = P, Q: Py = 1 = Qx = 1.)

Problem 9.

x2 y 2 x

a) R =

0 x 2

b)

4

y
y 2 /8 x

y

R =
f dA = f (x, y) dx dy.

0 y 4
R 0 y 2 /8

Problem 10.

4u9

Ru,v :
1v2

xu xv u2/3 v 1/3 /3 u1/3 v 4/3 /3


2 1 1

Jacobian J
=

2/3 2/3
1/3 1/3
=
+
u1/3 v 2/3 =
u1/3 v 2/3 .

yu yv
u v /3 2u v /3
9 9 3


9
1
f (x, y) dA = int21 f (u1/3 v 1/3 , u1/3 v 2/3 )( u1/3 v 2/3 ) du dv.
R 4 3

Problem 11.
a) Net ux out of R will be positive (more ow out than into R)

b)

F

n
ds = N dx + M dy = x dx + x dy

C C C
C1 : x = 1 t, y = t dx = dt, dy = dt
1
1

F

n
ds = (1 t)(1) + (1 t)(1) dt =
(1 t)2
0 = 1.

C1 0


C2 : x = 0 F n ds = 0.
C2
1
x2

1
C3 : y = 0, dy = 0 F n ds = x dx = = .
C3 0 2 2
Thus,
1
Fn
ds = = 1 + 0 + (1/2) = .
C C1 +C2 +C3 2
c)

1 1
div(F) = Mx + Ny = 1 div(F) dA = dA = area(R) = 11=
R R 2 2

Problem 12.
a) Limits on G are 2r z 2; 0 r 1; 0 2.

In cylindrical coordinates dV = dz r dr d.
Thus
2 1 2 2 1
1
M= z dV = z dz r dr d = 2(1r2 ) r dr d = 4 = .
G 0 0 2r 0 0 4

b) 2 1 2
1 1
z = z dV = z 2 dz r dr d.
M G 0 0 2r

c) In spherical coordinates: z = 2 cos = 2 = 2 sec .


Limits on G: 0 2 sec ; 0 tan1 (1/2); 0 2.
In spherical coordinates: dV = 2 sin d d d and z = cos , so
1
1 2 tan (1/2) 2 sec

1
z = z dV = ( cos )2 2 sin d d d.
M G 0 0 0

Problem 13.
a) We have F = P, Q, R, where P = y + y 2 z, Q = x z + 2xyz, R = y + xy 2 .

P R Q R P Q
= y2 = ; = 1 + 2xy = ; = 1 + 2yz = .
z x z y y x
b)

x1 y1 z1
f (x1 , y1 , z1 ) = P (x, 0, 0) dx + Q(x1 , y, 0) dy + Q(x1 , y1 , z) dz.
0 0 0
P (x, 0, 0) = 0; Q(x1 , y, 0) = x1 ; R(x1 , y1 , z) = y1 + x1 y12 .
y1 z1
f (x1 , y1 , z1 ) = 0 + x1 dy + (y1 + x1 y12 ) dz
0 0
f (x1 , y1 , z1 ) = x1 y1 y1 z1 + x1 y12 z1 f (x, y, z) = xy yz + xy 2 z + C.

c) F dr = f (1, 1, 2) f (2, 2, 1) = 10 + 3 = 7.
C

Problem 14. a)

= 12 x, y, 0, F = x, 0, 0.
b) n
1
Thus, F n = x2 and in cylindrical coordinates dS = 2dz d.
2

6
On the surface x = 2 cos and the limits of integration are 0 z 4, and 0 /2

/2 4 4 /2
1 2 1
2

n
dS = x
dS =
(2 cos ) dz 2d = 4 dz cos2 () d = 16 .

S 2 S 2
0 0 0 0 4

1
(We used the half angle formula cos2 = (1 + cos ).

1
c) div(F) = F = 1 F dV = 1 dV = Vol(G) = 22 4 = 4.
G G 4
d) Flux of F across all four at faces of G is zero.
Check:
= j F

n
face on xz-plane: n = 0.
= i F

n
face on yz-plane: n = x = 0 on yz plane.
faces on xy-plane and plane z = 4: n = k and k respectively, in either case F
n.

Problem 15. a)


i j k

F =
x y

z = i(x) j(y) + k(2z) = x, y, 2z.


yz xy 1

dS = zx , zy , 1 dA = 2x, 2y, 1 dA.


n

dS = (2x2 +2y 2 2z) dA = (subst. for z ) = (2x2 +2y 2 2(4x2 y 2 )) dA = 4(x2 +y 2 2) dA.

(F)
n
Limits of integration on R are 0 r 2; 0 /2.
/2 2 2
r4

2 2 2 2

(F)
n dS = 4 (x
+y
2) dA = 4 (r
2)r dr d = 4 ( r

= 2(44) = 0.

S R 0 0 2 4 0

b) F = yz, xz, 1 F dr = (yz) dx (xz) dy + 1 dz.
C C

7
C1 is in the yz-plane: x = 0, dx = 0, y = t, z = 4 t2 , dz = (2t) dt t goes from 2 to 0.
0
F dr = 1 dz = (2t) dt = 4.
C1 C1 2

C2 is in the xz-plane: y = 0, dy = 0, x = t, z = 4 t2 , dz = (2t) dt t goes from 0 to 2.


2
F dr = 1 dz = (2t) dt = 4.
C2 C2 0

C3 is in the xy-plane: z = 0, dz = 0

F dr = 0.
C3

Thus,
F dr = F dr = 4 + (4) = 0.
C C1 +C2 +C2

18.02 Final Exam


No books, notes or calculators.
15 problems, 250 points.

Useful formula: cos2 () = 12 (1 + cos(2))

Problem 1. (20 points)


a) (15 pts.) Find the equation in the form Ax + By + Cz = D of the plane P
which contains the line L given by x = 1 t, y = 1 + 2t, z = 2 3t and the point
(- 1, 1, 2).
b) (5 pts.) Let Q be the plane 2x + y + z = 4. Find the component of a unit normal
vector for Q projected on a unit direction vector for the line L of part(a).

Problem 2. (15 points)


Let L denote the line which passes through (0,0,1) and is parallel to the line in the

xy-plane given by y = 2x.

a) (5 pts.) Sketch L and give its equation in vector-parametric form.

b) (5 pts.) Let P be the plane which passes through (0,0,1) and is perpendicular to

the line L of part(a). Sketch in P (above) and give its equation in point-normal form.

(0, 0, 1) lies on L. Write down the method or

c) (5 pts.) Suppose that the point P =


formula you would use to nd the point P which is: (i) on L; (ii) the same distance
away from the point (0,0,1) as P ; and is (iii) on the other side of P from P .

Problem 3. (20 points)


1 0 3
Given the 3 3 matrix: A a = 2 1 1 :
1 1 a
a) (5 pts.) Let a = 2: show that | A2 | = 0


x 0
b) (7 pts.) Find the line of solutions to A2 y = 0
z 0


c) (8 pts.) Suppose now that a = 1, and that A1 1 = 3 p 5 . Find p.

Problem 4. (10 points)


Let r(t) = cos(et ), sin(et ), et .
r (t)
a) (5 pts.) Compute and simplify the unit tangent vector T(t) = .
| r (t) |
b) (5 pts.) Compute T (t)

Problem 5. (20 points)


y
Consider the function F (x, y, z) = z x2 + y + 2 :
z
a) (10 pts.) The point P0 : (1, 3, 2) lies on the surface F (x, y, z) = 7. Find the
equation of the tangent plane to the surface F = 7 at P0 .
b) (5 pts.) If starting at P0 a small change were to be made in only one of the
variables, which one would produce the largest change (in absolute value) in F ? If
the change is this variable was of size 0.1, approximately how large would the change
in F be ?
c) (5 pts.) What distance from P0 in the direction 2, 2, 1 will produce an
approximate change in F of size 0.1 units, according to the (already computed) lin
earization of F ?

Problem 6. (15 points)


2
Let f (x, y) = x + 4y + .
xy
a) (10 pts.) Find the critical point(s) of f (x, y)
b) (5 pts.) Use the second-derivative test to test the critical point(s) found in part(a).

Problem 7. (10 points)


Let P be the plane with equation Ax + By + Cz = D and P0 = (x0 , y0 , z0 ) be a point

which is not on P.

Use the Lagrange multiplier method to set up the equations satised by the point

(x, y, z) on P which is closest to P0 . (Do not solve.)

Problem 8. (15 points)


a) (10 pts.) Let F (x, y, z) be a smooth function of three variables for which F (1, 1, 2) =
1, 2, 2.
F
Use the Chain Rule to evaluate at (, , ) = (2, 4 , 4 ).

(Use x = sin cos , y = sin sin , z = cos .)
b) (5 pts.) Suppose f (x, y) is a smooth, non-constant function. Is it possible that, for
all points (x, y), the gradient of f at the point (x, y) is equal to the vector y, x ?
Justify (briey).

Problem 9. (10 points)


2 22x
f dA = f (x, y) dy dx .
R 0 x2
a) (5 pts.) Sketch the region R.
b) (5 pts.) Rewrite the double integral as an iterated integral with the order inter
changed.

2
Problem 10. (15 points)

Set up the integral f (x, y) dA where R is the region bounded by the four curves
R
x2 y = 4, x2 y = 9, xy = 1, and xy = 2 as a double integral in the variables u = x2 y
and v = xy . (Your answer should be completely ready to integrate, once the function
f is given.)
1 1 1 2
Note: the inverse transformation is given by x = u
3 v 3 ,
y = u
3 v
3 .

Problem 11. (15 points)

F(x, y) = x (i + j), and let C be the closed curve in the xy-plane formed by the

triangle with vertices at the origin and the points (1,0) and (0,1).

a) (5 pts.) Give a rough sketch of the eld F in the rst quadrant, and use it to

predict whether the net ux out of the region R = the interior of C will be positive

or negative.


b) (5 pts.) Compute the ux integral Fn ds directly.
C
(Specify which orientation you are using for C.)

c) (5 pts.) Compute the ux integral Fn ds using the appropriate double
C
integral. (Set up, then using short-cut is ok.)

Problem 12. (20 points)


Let G be the solid 3-D cone bounded by the lateral surface given by z = 2 x2 + y 2
and by the plane z = 2. The problem is to compute
z = the z-coordinate of the center of mass of G, in the case where the density is equal
to the height above the xy-plane.
a) (5 pts.) Find the mass of G using cylindrical coordinates
(b) (5 pts.) Set up the calculation for z using cylindrical coordinates
(Answers should be ready to integrate out but do not evaluate.)
(c) (10 pts.) Set up the calculation for z using spherical coordinates.
(Answers should be ready to integrate out but do not evaluate.)

Problem 13. (15 points)


F(x, y, z) = y + y 2 z i + (x z + 2xyz) j + y + xy 2 k

a) (3 pts.) Show that F(x, y, z) is a gradient eld using the derivative conditions.
b) (10 pts.) Find a potential function f (x, y, z) for F(x, y, z), using any systematic
method. Show the method used and all work clearly.

c) (2 pts.) Find F dr, where C is the straight line joining the points (2, 2, 1) and
C
(1, -1, 2) (in that order), using as little computation as possible.

3
Problem 14. (25 points)
In this problem S is the surface given by the quarter of the right-circular cylinder

centered on the z-axis, of radius 2 and height 4, which lies in the rst octant. The

eld F(x, y, z) = x i.

a) (5 pts.) Sketch the surface S and the eld F.

(Suggestion: use a coordinate system with y pointing out of the paper.)


b) (10 pts.) Compute the ux integral Fn dS.
S
(Use the normal which points outward from S, i.e. on the side away from the z-axis.)

c) (5 pts.) G be the 3D solid in the rst octant given by the interior of the quarter
cylinder dened above. Use the divergence theorem to compute the ux of the eld
F = x i out of the region G.
d) (5 pts.) The boundary surface of G is comprised of S together with four other
faces. What is the ux outward through these four faces, and why ? Use the answers
to parts(b) and (c), and also verify using the sketch of part(a).

Problem 15. (25 points)


F(x, y, z) = (yz) i + (xz) j + k. Let S be the portion of surface of the paraboloid
z = 4 x2 y 2 which lies above the rst octant; and let C be the closed curve C
= C1 + C2 + C3 , where the curves C1 , C2 and C3 are the three curves formed by
intersecting S with the xy, yz and xz planes respectively (so that C is the boundary
of S). Orient C so that it is traversed CCW when seen from above in the rst octant.

a) (15 pts.) Use Stokes Theorem to compute the loop integral F dr by using
C
the surface integral over the capping surface S.


b) (10 pts.) Set up and evaluate the loop integral F dr directly by parametriz
C
ing each piece of the curve C and then adding up the three line integrals.

1. Find the circulation


- given: vector fieldregion
- apply: 1) Stokes Theorem 2) Direct calculation of line integrals
- interpret the results

2. Derive the mass continuity equation and differential forms


- Given:
a) rate of increase of mass within finite control volume
=-net mass flow rate from control volume across surface S
b) equation for flux(=mas per area):
mass flow rate crossing area A:

,

mass flux:mass flow rate per unit area =

- apply: 1)Control Volume method
a) mass flow rate out of one surface:

b) all the surface?
c) rate of decrease of mass of volume element?
2) Gauss Divergence Theorem
- interpret the results

3. calculate total kinetic energy of a cylingder


- given: radius R, height H, density , rotating about axis angular velocity , physical eq.
- apply: 1) Volume integrals. See if the result gives vector or scalar
- interpret the results


4. Calculate surface integral (



) for given vector field for each surface

- given: vector field,


,
- apply: 1) Surface Integrals. check for the region. 2) Check for divergence of the
field.
- interpret the results
5. Show that field
is conservative and find its scalar potential
- given:
at given point.
- apply: 1) Conservative field theory. check for
- interpret the results

6. Evaluate double integral over an area bounded by certain region to get a volume
of something.
- given: region, function
- apply: 1) double integrals
- interpret the results

7. Find the scalar potential of the vector field


. Sketch the field lines and
equipotentials.
- given: vector field
- apply: 1) conservative field and obtaining scalar. 2) sketch field lines. 3) Sketch
equipotential lines, constant
- interpret the results

8. Derive equation for rotation of fluids


- given: Velocity field in 2-D, curl operator.
applying :( cos , sin ), components of : , effective
components of : use drawing,

- apply: 1) =? 2) mean angular velocity = ?

- interpret the results: curl V = vorticity, field lines are vortex lines.

9. Derive equation for heat conduction. (Using Laplaces eq and potential theory)
- given: Temperature field, law of conduction.

Rate of increase of energy in small volume = - Net efflux of heat from small volume
- apply: 1) divergence and laplace equation.
- interpret the results
10. Apply gradient to obtain physical equation of heat flow / conduction / diffusion.
Obtain scalar potential.
- given: Some physical law
- apply: 1) Gradient operator
- interpret the results

11. Find the equation of field lines and sketch field lines for velocity field
.
- given: Vector field.
- apply: find the field line
- interpret the results

12. Derive an expression for rate of chance of pressure with a distance at arbitrary
angle . (Force vortex flow)
- given: function for pressure. v(x,y) = kx
- apply:
- interpret the results

13. Prove triple product formula.


- given: pv = RT
- apply: chain rule and partial derivative.

14. Prove Laplace equation. (Using chain Rule)


- apply: chain rule

15. Draw a curve in 3-D.


- given: x( ), y( ), z( )
- interpret the results

16. Give definition and geometrical meaning for.


- gradient, divergence, curl, gauss theorem, stokes theorem
- partial derivative of scalar functions.