MAT 303 Spring 2013 Calculus IV with Applications

Final Exam May 21, 2013, 8:00 to 10:45 AM

Name: Solution Key

Circle your recitation:

R01 (Claudio Fri) R02 (Xuan Wed) R03 (Claudio Mon)

You have a maximum of 2 12 hours. This is a closed-book, closed-notes exam. No

calculators or other electronic aids are allowed.

Read each question carefully. Show your work and justify your answers for full
credit. You do not need to simplify your answers unless instructed to do so.

If you need extra room, use the back sides of each page. If you must use extra paper,
make sure to write your name on it and attach it to this exam. Do not unstaple or
detach pages from this exam.

Grading

1 /15 6 /10

2 /15 7 /20

3 /20 8 /10

4 /10 9 /20

5 /15 10 /15

Total /150
MAT 303 Spring 2013 Calculus IV with Applications
MAT 303 Spring 2013 Calculus IV with Applications

dy 1 2x
1. (15 points) Consider the differential equation = .
dx y
(a) (10 points) Find the general solution to this DE.
dy
Solution: This DE is separable, so we write it as y dx = 1 2x. Integrating, we obtain

1 2 p
y = x x2 + C y2 = 2x 2x2 + C y= 2x 2x2 + C.
2
Note that because the square-root function always returns non-negative values, we
require a sign to capture all of the solutions.

(b) (5 points) Find the solution matching the condition y(1) = 2. On what interval is
this solution defined?
Solution: We apply x = 1 and y = 2 to the implicit form of the solution to solve for
C. Then 4 = 2(1) 2(1)2 + C, so C = 4. Since y(1) < 0, we must take the negative
branch of the solution, so p
y( x ) = 2x 2x2 + 4.
We note that this function is defined where 2x 2x2 + 4 0. Equivalently, x2 x
2 0, so since x2 x 2 = ( x 2)( x + 1), this is where 1 x 2. From the
original DE, though, we should exclude points where y = 0, so the solution is defined
on the open interval (1, 2).

1
MAT 303 Spring 2013 Calculus IV with Applications

2. (15 points) Find the general solution to the linear system

0
x (t) 1 0 0 x (t)
y ( t ) = 1 2 0 y ( t )
z(t) 3 2 1 z(t)

Solution: Letting the matrix above be denoted A, we find its eigenvalues. As a shortcut,
we note that it it lower triangular, so its eigenvalues are the values along the diagonal: 1,
with multiplicity 2, and 2. To check, we compute det( A I ) by row expansion:
 
1 0 0   
  2 0 
det( A I ) =  1
 2 0  = (1 ) 
  = (1 )(2 )(1 ).
2 1 
 3 2 1 
 T
We find eigenvectors for these eigenvalues. First, we find solutions v = a b c to
( A I )v = 0 by row reducing A I:

0 0 0 1 1 0 1 1 0 0 1 0 1 0 0
A I = 1 1 0 3 2 0 1 0 0 1 0 0 0 1 0
3 2 0 0 0 0 0 0 0 0 0 0 0 0 0
Translating this back into equations, the two non-trivial rows give that a = 0 and b = 0,
 T
so v = 0 0 c , with c still a free variable. Letting c = 1, we obtain a single linearly
 T
independent eigenvector, v1 = 0 0 1 . Since = 1 has multiplicity 2, we expect a
 T
generalized eigenvector v2 = a b c with ( A I )v2 = v1 , which we solve by row
reducing [ A I | v1 ]:

0 0 0 0 1 1 0 0 1 1 0 0
1 1 0 0 3 2 0 1 1 0 0 1
3 2 0 1 0 0 0 0 0 0 0 0

0 1 0 1 1 0 0 1
1 0 0 1 0 1 0 1
0 0 0 0 0 0 0 0
 T  T
Then a = 1 and b = 1, so v2 = 1 1 c . Choosing c = 0 this time, v2 = 1 1 0
(although any choice of c is valid).
 T
Next, we find an eigenvector v3 = a b c for = 2 by row reduction of A 2I:

1 0 0 0 0 0 1 0 0
A 21 = 1 0 0 1 0 0 0 2 1
3 2 1 0 2 1 0 0 0
 T
Then a = 0 and 2b c = 0, so 2b = c. Taking b = 1, c = 2, and v3 = 0 1 2 . Therefore,
the general solution is

0 0 1 0
t t 2t
x ( t ) = c1 e 0 + c2 e t 0 + 1 + c3 e 1 .
1 1 0 2

2
MAT 303 Spring 2013 Calculus IV with Applications

3. (20 points)
An 8 kg mass m is attached to a spring of constant k = 2 N/m,
k
and allowed to return to its equilibrium position. The system is m
naturally damped due to friction with a constant of c = 8 N/m-s.
(a) (5 points) Find the general solution to the motion x (t) of the mass.
Solution: The DE describing the motion is 8x 00 + 8x 0 + 2x = 0, which we normalize to
x 00 + x 0 + 14 x = 0. Its characteristic equation is
 2
2 1 1
r +r+ = 0 r+ = 0,
4 2

so it has a double root at r = 12 , and therefore the general solution is

1 1
x (t) = c1 e 2 t + c2 te 2 t .

(b) (5 points) Suppose that the mass is moved 10 cm to the right of the equilibrium posi-
tion and released at a speed of 7 cm/s to the left at time t = 0. Find the displacement
x (t) of the mass, in cm.
Solution: From the statement, the initial conditions are x (0) = 10 and x 0 (0) = 7,
1 1 1
where x is measured in cm. Since x 0 (t) = 12 c1 e 2 t + c2 ( 21 te 2 t + e 2 t ), x 0 (0) =
12 c1 + c2 . Then x (0) = c1 , so c1 = 10, and c2 = 7 + 12 (10) = 7 + 5 = 2. Hence,
the displacement function is
1
x (t) = (10 2t)e 2 t .

(continued on next page)

3
MAT 303 Spring 2013 Calculus IV with Applications

(c) (5 points) Find the first time t > 0 at which the mass crosses the equilibrium position.
Solution: The mass crosses the equilibrium whenever its displacement is 0. Using
1 1
that x (t) = (10 2t)e 2 t , we solve (10 2t)e 2 t = 0: dividing by the never-zero
exponential, 10 2t = 0, so 2t = 10, and t = 5 seconds.

(d) (5 points) An external force f (t) = 20 cos t is then applied to the mass. Find the
amplitude of the steady-state motion of the mass.
Solution: Since the forcing term contains cos t, and since neither cos t nor sin t appears
in the general solution of the unforced equation, we guess x p (t) = A cos t + B sin t as
a particular solution. We plug its derivatives

x 0p = A sin t + B cos t, x 00p = A cos t B sin t = x p ,

into the forced DE, 8x 00 + 8x 0 + 2x = 20 cos t:

8x 00p + 8x 0p + 2x p = 8x 0p 6x p = 8A sin t + 8B cos t 6A cos t 6B sin t = 20 cos t.

Isolating the coefficients of the cos t and sin t components, we have the linear system
6A + 8B = 20 and 8A 6B = 0. Then B = 34 A, so 6A 32 3 A = 20, and
6 8
50A = 60. Hence, A = 5 , so B = 5 . Then the amplitude is
s
 2  2 r 2
p 6 8 10
C= A2 + B2 = + = = 2.
5 5 52

4
MAT 303 Spring 2013 Calculus IV with Applications

4. (10 points) Fresh water flows at a constant rate of r = 6 liters/minute into a tank
contaning V1 = 20 liters of salt solution. The well-mixed solution then flows at the same
rate into a second tank containing V2 = 30 liters of solution, and then drains out of that
tank at the same rate.
(a) (5 points) Write a linear system of DEs describing the amount of salt x1 (t) and x2 (t) in
each of the two tanks.
Solution: For each tank, the rate of change of the amount of salt is rin cin rout cout . For
each tank, rin = rout = 6. For tank 1, cin = 0 and cout = Vx11 = 20
1
x1 , so

1 3
x10 = 0 (6) x1 = x1 .
20 10
x1 1 x2 1
For tank 2, cin = V1 = 20 x1 , while cout = V2 = 30 x2 , so

1 1 3 1
x20 = (6) x1 (6) x2 = x1 x2 .
20 30 10 5

(b) (5 points) The general solution to this system is x1 (t) = c1 e3/10t , x2 (t) = c2 e1/5t
3c1 e3/10t . At t = 0, tank 1 contains salt at a concentration of 0.5 kg/liter and tank 2
at a concentration of 0.2 kg/liter. Find x1 (t) and x2 (t) matching this initial condition.
Solution: Since we are given the initial concentrations, not the amounts, we multiply
by the tank volumes to obtain x1 (0) = (0.5)(20) = 10 and x2 (0) = (0.2)(30) = 6.
Then x1 (0) = c1 = 10 and x2 (0) = c2 3c1 = 6, so c2 = 6 + 3c1 = 36. Hence, the
solution with this IC is

x1 = 10e3/10t , x2 (t) = 36e1/5t 30e3/10t .

5
MAT 303 Spring 2013 Calculus IV with Applications

5. (15 points) Consider the linear system x 0 = 3x 5y, y0 = x y.

(a) (10 points) Find the general solution to this system.
 
3 5
Solution: Writing the system as x0 = Ax, with A = , we find the eigenvalues
1 1
of A:
 
3 5  = (3 )(1 ) + 5 = 2 2 + 2.

det( A I ) = 
1 1 

The roots of this characteristic polynomial are = 2 248 = 1 i. We find a complex
eigenvector for = 1 i by row reduction of A (1 i ) I:
     
2+i 5 0 0 1 2 + i
A (1 i ) I = .
1 2 + i 1 2 + i 0 0
 T
Then v1 = 2 i 1 is a reasonable complex eigenvector, giving the complex-
valued solution
 
(1 i ) t t 2i
x1 ( t ) = e v1 = e (cos t i sin t)
1
   
t 2 cos t + sin t 2 sin t 2 cos t
=e +i .
cos t sin t

Taking linear combinations of the real and imaginary parts of this solution (and mak-
ing the choice to multiply the imaginary part by 1, to remove the negative signs)
gives the general real-valued solution:
   
t 2 cos t + sin t t 2 sin t + 2 cos t
x ( t ) = c1 e + c2 e .
cos t sin t

(b) (5 points) Characterize the behavior of the system around its only critical point, (0, 0).
Solution: Since A has complex eigenvalues, = 1 i, with positive real part, we
expect an unstable spiral point at the origin.

6
MAT 303 Spring 2013 Calculus IV with Applications

dy
6. (10 points) Find the general solution to the differential equation x = 2y + x3 cos x.
dx
dy dy
Solution: We normalize the DE to isolate dx : dx = 2x y + x2 cos x. We then recognize the DE
as linear and rearrange it into the normal form

dy 2
y = x2 cos x.
dx x
2x dx
R
The corresponding integrating factor is then ( x ) = e = e2 ln |x| = x 2 . Multiply-
ing the DE by ( x ), it becomes

( x 2 y)0 = x 2 x2 cos x = cos x,

which we integrate to obtain x 2 y = sin x + C. Isolating y by multiplying by x2 , the

general solution is
y = x2 sin x + Cx2 .

7
MAT 303 Spring 2013 Calculus IV with Applications

7. (20 points)
Consider a damped pendulum of length L = 5/8 m. Assuming that
g = 10 m/s2 , the angle (t) it makes with the vertical is controlled by
L
the nonlinear differential equation 00 + 6 0 + 16 sin = 0. Introducing
the new variable = 0 , we obtain the nonlinear autonomous system

0 = , 0 = 16 sin 6.
(a) (5 points) Find all of the critical points (, ) of this system with 0 < 2.
Solution: To find the critical points, we must find the values in this range where 0 =
0 and 0 = 0 simultaneously. From the first equation, = 0. From the second,
16 sin 6 = 0, so sin = 0 as well. This occurs where = n for all integers n,
so in this range the values of are 0 and . Hence, the critical points (, ) are (0, 0)
and (, 0).

(b) (5 points) Find the Jacobian matrix J (, ) of this system.

Solution: Letting 0 = F (, ) = and 0 = G (, ) = 16 sin 6, the Jacobian
matrix is    
F (, ) F (, ) 0 1
J (, ) = = .
G (, ) G (, ) 16 cos 6

(continued on next page)

8
MAT 303 Spring 2013 Calculus IV with Applications

(c) (10 points) Characterize the behavior of the system at the critical points you found in
part (a). Interpret this behavior in terms of the motion of the pendulum.
Solution: We evaluate the Jacobian matrix at these two critical points:
 
0 1
At (0, 0), cos = 1, so J (0, 0) = . Computing its eigenvalues,
16 6
 
 1  = 2 + 6 + 16.

det J (0, 0) I = 
16 6 

This polynomial has roots = 6 23664 = 3 i 7, which are complex with
negative real part. Hence, the system has an asymptotically stable spiral point at
(0, 0).
 
0 1
At (, 0), cos = 1, so J (, 0) = . Computing its eigenvalues,
16 6
 
 1  = 2 + 6 16 = ( + 8)( 2).

det J (, 0) I = 
16 6 

Therefore, the eigenvalues are 2 and 8, so the system has a saddle point at
(, 0), which is always unstable.
In terms of the motion of the pendulum, we see that if the pendulum starts close to
= 0 (pointing straight downward) with a small angular velocity , it will get closer
and closer to resting at = 0. On the other hand, if it is close to the straight-upward
position = , it will not settle there, and will instead settle at another equilibrium.
There is one trajectory that does converge to this upwards equilibrium, but in practice
small forces on the pendulum will perturb it out of this trajectory.

9
MAT 303 Spring 2013 Calculus IV with Applications

e e2t
   t 
5 3
8. (10 points) Let B = . A fundamental matrix for x = Bx is (t) =
0 .
6 4 2et e2t
 2t 
3e
Find a particular solution to the system x0 = Bx + f(t), where f(t) = .
2e2t
Solution: From the fundamental matrix given in the problem statement, we see that the
eigenvalues of B are 1 and 2, which do not overlap with the eigenvalue 2 represented in
the forcing functions. Hence, we may guess a particular solution of the form x p (t) = e2t a,
T
where a = a1 a2 . Then x0p = 2e2t a. Rearranging the DE to be x0 Bx = f(t) and


plugging in this guess, we obtain

 2t   
2t 2t 3e 3
2e a e Ba = 2t (2I B)a = ,
2e 2
2t
where we factor
  out the a on the left and divide by the nonzero factor e . Then a =
3 7 3
(2I B)1 . Since 2I B = , and using the positional formula for the inverse
2 6 2
of a 2 2 matrix,
        
1 1 2 3 1 3 1 2 3 3 0
(2I B) = (2I B) = =
4 6 7 2 4 6 7 2 1
 
0
Therefore, a particular solution is x p = .
e2t

R also obtain a particular solution with the variation of parameters formula x(t) =
We can
(t) (t)1 f(t) dt:
  t
et
 2t
e2t

1 1 e e
(t) = =
et 2et et 2e2t e2t
 t
et 3e2t et
    
1 e
(t) f(t) = =
2e2t e2t 2e2t 4e4t
et e2t et
     
0
Z
1
(t) (t) f(t) dt = = .
2et e2t e4t e2t

This agrees with the answer calculated with undetermined coefficients. This computation
could also be done using e Bt as the fundamental matrix, but the additional work required
to compute and to work with e Bt is probably not justified, given that it was relatively easy
to compute (t)1 directly.

10
MAT 303 Spring 2013 Calculus IV with Applications

9. (20 points) Ninjas board a ship full of pirates, and they start to fight, thereby reducing
each others populations. Because of the ninjas superior training and skills, they are four
times more effective at fighting than the pirates are. A model governing the evolution of
the populations x (t) of pirates and y(t) of ninjas is

dx dy 1
= 2y, = x.
dt dt 2
(a) (5 points) Find the general solution to this system of DEs.
   
0 x 0 2
Solution: We write this system as x = Ax, with x = and A = . We
y 21 0
compute the eigenvalues of A:
 
2
det( A I ) = = 2 1 = ( 1)( + 1).
21

Hence, the eigenvalues are 1 and 1. We find eigenvectors for both:

     
1 2 1 2 1 2
AI = ,
21 1 0 0 0 0


2
so 1 = 1 has an eigenvector v1 = . Similarly,
1
   
1 2 1 2
A+I =
21 1 0 0
 
2
so 2 = 1 has an eigenvector v2 = . The general solution is then
1
   
t 2 t 2
x ( t ) = c1 e + c2 e .
1 1

(continued on next page)

11
MAT 303 Spring 2013 Calculus IV with Applications

(b) (5 points) Plot some trajectories of these populations in the xy-plane on the axes below.
Pay close attention to the behavior along any eigendirections.
y

(c) (5 points) Explain what happens in the first quadrant. How can you tell from the
starting populations which side will be victorious?
Solution: In the first quadrant, we see that the line y = 21 x separates two regions with
different behaviors: below this line, trajectories intersect the positive x-axis, indicat-
ing that pirates are left but not ninjas (x > 0, but y = 0). Conversely, above this line,
trajectories intersect the positive y-axis, indicating that ninjas are left but not pirates.
Hence, this line is a separatrix between these two outcomes, and the location of the
initial condition above or below the separatrix determines the outcome entirely.

(d) (5 points) If the pirate ship initially contains 50 pirates, how many ninjas are required
to defeat all of them (that is, to reduce their population to 0)?
Solution: In order for all the pirates to be destroyed, at least 25 ninjas are required, so
that the trajectory does not intersect the positive x-axis. This differs from the naive
estimate that 13 ninjas would be required (rounding up from 50/4 = 12.5), which
one might expect from the 4-to-1 skill advantage. (More generally, an n2 -to-1 skill
advantage is required in this model of combat to overcome being outnumbered n-to-
1.)

12
MAT 303 Spring 2013 Calculus IV with Applications
 
3 2
10. (15 points) Let A = .
0 4
(a) (10 points) Find e At .
Solution:
 Although
  A=
we would like to write  D + B, the sum of a diagonal matrix
3 0 0 2
D= and a nilpotent matrix B = , this is not useful for computing e At ,
0 4 0 0
since D and B do not commute (BD 6= DB). Likewise, although 3I and A 3I com-
mute, A 3I is not nilpotent and therefore does not have an easily computed matrix
exponential. Instead, we construct a fundamental matrix (t) from the eigendata of
A. Since A is upper triangular, its eigenvalues are 3 and 4 by inspection. We compute
eigenvectors by row reducing A 3I and A 4I:
     
0 2 0 1 1
A 3I = v1 =
0 1 0 0 0
     
1 2 1 2 2
A 4I = v2 =
0 0 0 0 1

Then a fundamental matrix (t) for the system x0 = Ax is

 3t
2e4t
    
e 1 2 1 1 2
(t) = , (0) = , (0) = .
0 e4t 0 1 0 1

Finally,

e3t 2e4t
  3t
2e4t 2e3t
  
At 1 1 2 e
e = ( t ) (0) = = .
0 e4t 0 1 0 e4t

d At
(b) (5 points) Evaluate e at t = 0.
dt
d d
Solution: We first compute e At : by the properties of the matrix exponential, e At =
dt dt
Ae At . At t = 0, this is Ae A(0) = Ae0 = AI = A, so the answer is
  
d At  3 2
e =A= .
dt t=0 0 4

Although we have the answer, we check that it is consistent with our result from
part (a). Differentiating e At termwise and evaluating at t = 0,
 3t
8e4t 6e3t
     
d At 3e d At  3 86 3 2
e = , e = = = A.
dt 0 4e4t dt t=0 0 4 0 4

13