Solutions to Exam 3

Note Solutions to both exams are given (since the two versions are identical except for the order). The problem numbers are ordered for the white (version A) exam with a note at the beginning of the solution for the corresponding problem for the blue (version B) exam.

Part I: Multiple Choice


White 1c Blue 4c
 
 Since we have an indeterminate quotient, 0 / 0, we may apply
 l'Hospital's rule. Then use a limit fact from Section 3.4 (or use
 l'Hospital's rule again if you prefer).
 


                      cos(x) - 1                 sin(x)
               lim    ---------- =  lim    - 1/2 ------
              x -> 0       2       x -> 0          x
                          x


                          /        sin(x)\
                    - 1/2 | lim    ------| = -1/2
                          \x -> 0    x   /

 White 2a  Blue 2a

 Plug and chug; don't forget the Chain Rule!


                                        2
                           f := arctan(x )


                                      x
                            Df := 2 ------
                                         4
                                    1 + x

  White 3d Blue 1d
 
 To find the absolute maximum and minimum values of f(x)=x^3 - 3*x + 1
 on the interval [-1, 3], first find the critical values of f which lie
 in the stated interval. (These are where the derivative of f is zero
 or undefined. In our case, since f is a polynomial, this amounts to
 where f ' is zero.) To obtain the largest and smallest function
 values, evaluate f at its critical numbers and at the endpoints of the
 interval. For efficiency, we'll evaluate all the function values at
 once by mapping f onto a list of x-values! (This parallel list
 processing is also available on an HP48G/GX via DOLIST.)
 


                                   3
                        f := x -> x  - 3 x + 1


                                       2
                         Df := x -> 3 x  - 3


                     eq := 3 (x - 1) (x + 1) = 0


                critical_numbers := {x = 1}, {x = -1}


                        x_values := [-1, 1, 3]


                    function_values := [3, -1, 19]

 White 4b Blue 3b

 STRATEGY: Use the IVT to show there's at least one solution to the
 equation f(x)=0, then draw a rough sketch (using some of your Chapter
 5 curve analysis skills) to conclude that there is exactly one root.


                                   3
                        f := x -> x  - 3 x + 3


                          x_vals := [-3, -1]


                        func_vals := [-15, 5]

 Since f(-3) = -15 and f(-1) = 5, we have by the IVT that there is a
 root of f in the interval (-3, -1).

                                       2
                         Df := x -> 3 x  - 3


                     eq := 3 (x - 1) (x + 1) = 0


                critical_numbers := {x = 1}, {x = -1}


                   interior_x_values := [-3, 0, 3]


                    Df_sign_analysis := [1, -1, 1]


                         crit_nums := [-1, 1]


                       local_max_min := [5, 1]

 We now know that:
      1. the graph of f crosses the x-axis between x=-3 and x=-1;
      2. f(-1)=5 is a local max of f;
      3. f(1)=1 is a local min of f;
      4. f is increasing on (-infinity, -1] and on [1, infinity);
      5. f is decreasing on [-1, 1].
 This enables us to draw a rough sketch like the one below.
> plot(f(x), x=-3..2, xtickmarks=6, ytickmarks=4);

 We conclude that f(x)=0 has exaclty one real solution.
 
White 5d Blue 7d
 
 By the MVT, there is a number c in (0, 2) such that `f '`(c) = (f(2) -
 f(0)) / (`2` - `0`). Solving this equation for f(2) and applying the
 other facts of the problem yields f(2) = 2*`f '`(c) + f(0)``<=2*`(3)`
 + 2``=8. That is, f(2)<=8.
 
 White 6c Blue 5c
 
 From the graph of f ' on the exam, we conclude via the First
  Derivative Test that there is a local minimum at x=2, since at this
  value f ' changes sign from -`` to +.
  
 White 7b Blue 6b
  
  From the graph of f ' on the exam, x=1 is a point of inflection, as
  follows. At a point of inflection, f '' changes sign. Equivalently, (f
  ') ' changes sign. In other words, "the slope of the depicted curve, f
  ', changes sign." Clearly this occurs at x=1 (and NOT at x=3).
  
  White 8d Blue 8d
  
  The choices all speak of the increase and concavity of the function
  f(x)=x*exp(x). Accordingly, we perform sign analyses on f ' and f ''.
  The factored form of f ' shows that it changes sign from -`` to + at
  x=-1. Thus f is increasing on (-1, infinity). The factored form of f
  '' indicates it changes sign from -`` to + at x=-2. Hence f is concave
  down on (-infinity, -2).


                            f := x exp(x)


                       Df := exp(x) + x exp(x)


                    Df_factored := exp(x) (x + 1)


                      D2f := 2 exp(x) + x exp(x)


                    D2f_factored := exp(x) (x + 2)

 White 9e Blue 9e
  
  Determine the list of critical values, then compute f, f ' and f ''
  thereat. Here are conclusions via the Second Derivative Test for Local
  Extrema (you could also use the First Derivative Test):


                                 4        3        2
                  f := x -> 1/4 x  - 4/3 x  + 3/2 x


                                  3      2
                      Df := x -> x  - 4 x  + 3 x


                                     2
                      D2f := x -> 3 x  - 8 x + 3


           Df_factored_set_to_zero := x (x - 1) (x - 3) = 0


                    critical_numbers := [0, 1, 3]


                  function_values := [0, 5/12, -9/4]


                    derivative_values := [0, 0, 0]


                second_derivative_values := [3, -2, 6]

> plot(f(x), x=-1..4, xtickmarks=5, ytickmarks=5);

 White 10a Blue 11a
  
  Antidifferentiate (i.e., indefinitely integrate); don't forget the +C!
  Then use the numerical data to resolve the constant.


                      Df := 3 cos(x) + 5 sin(x)


                                  /
                                 |
           antidifferentiate :=  |  3 cos(x) + 5 sin(x) dx
                                 |
                                /


                  f := x -> 3 sin(x) - 5 cos(x) + C


                           eq := -5 + C = 4


                            sol := {C = 9}


               answer := f(x) = 3 sin(x) - 5 cos(x) + 9

 White 11e Blue 10e
  
  since the integral from 1 to 5 equals 7 and the integral from
  3 to 5 equals 2, we subtract and find that the integral
  from 1 to 3 is 7-2 = 5. Adding this value to the integral
  from 3 to 8, which is 6, we obtain that the entire
  integral from 1 to 8 is 5+6=11.
  
 

Part II: Work Out

  
  White 12 Blue 12
  
  Here's the equation for exponential growth or decay, XGD. Substitute
  the numerical data into this equation and solve for the decay constant
  k. Then use this to define the amount of radon as a function of time.
  (Here the initial amount is a symbolic constant.)


                        XGD := y = y0 exp(k t)


                         3/5 y0 = y0 exp(5 k)


                         ksol := 1/5 ln(3/5)


                   f := t -> y0 exp(1/5 ln(3/5) t)

  Now find out how much time must elapse until half the radon remains;
  this is called the half-life. The exact answer is 5*ln(`1/2`) /
  ln(`3/5`), which we subsequently float to give a feel for how many
  days this really is.


                    1/2 y0 = y0 exp(1/5 ln(3/5) t)


                                  ln(2)
                              -5 -------
                                 ln(3/5)


                    half_life := 6.784577245 days

  NOTE: You can do this problem automatically on an HP48G/GX. Simply
  choose any numerical value for y[0] (say 100 g) and use the SOLVR.
  
  White 13 Blue 14
  
  For the limit Limit((1+2*sin(x))^`1/x`, x=0), we have an indeterminate
  power, 1^infinity. Accordingly, first compute the limit L of the
  natural logarithm of y=(1+sin(2*x))^`1/x`. The final answer is then
  exp(L).


                                       ln(1 + 2 sin(x))
                lim    ln(y) =  lim    ----------------
               x -> 0          x -> 0         x


                       Apply l'Hospital's rule.


                                           cos(x)
             L = lim ln(y) :=  lim    2 ------------ = 2
                              x -> 0    1 + 2 sin(x)


                   Thus y equals := exp(L) = exp(2)

  Maple agrees with our findings.


                                       (1/x)
                  lim    (1 + 2 sin(x))      = exp(2)
                 x -> 0

  White 14 Blue 13
  
  STRATEGY: By implicitly computing the derivative dy / dx, we will
  actually have computed Diff(arccos(x), x).


                           eq := cos(y) = x


                            y = arccos(x)


                         d      d
                         -- y = -- arccos(x)
                         dx     dx

  Here are the steps involved in the implicit differentiation.


      cos(y(x)) = x,      Regard y as an implicit funciton of x.


              /d      \
   -sin(y(x)) |-- y(x)| = 1,      Differentiate with respect to x.
              \dx     /


             d               1
            {-- y(x) = - ---------},      Solve for dy/dx.
             dx          sin(y(x))


                                       1
                          dy_dx := - ------
                                     sin(y)

  Now express sin(y) in terms of x. While Maple can do this via brute
  force, we would draw a right triangle and use trigonometry.


                                          2 1/2
                       sine_of_y := (1 - x )


                         d             1
                         -- y = - -----------
                         dx             2 1/2
                                  (1 - x )


                           d                     1
               finished := -- arccos(x) = - -----------
                           dx                     2 1/2
                                            (1 - x )

 White 15 Blue 15
  
  (a) Diagram: Draw  a simple sketch of a rectangular box, labeling it
  as follows.
  LEGEND:   x=width,   2*x=length,   h=height.
  (b) Surface area function
  Recall the top is open.
  The surface area is (area of base)  +  (area of left side  +  area of
  right side  +  area of front side  + area of back side).


                                    2
                         area := 2 x  + 6 h x

  (c) Volume constraint


                                       2
                           volume = 2 x  h


                                     2
                             36 = 2 x  h


                                        18
                          hsol := {h = ----}
                                         2
                                        x

  (d) Area as a function of width


                                      2   108
                         S := x -> 2 x  + ---
                                           x


  (e) Find the dimensions of the box which give the (absolute) minimum
  area.
  Note that for 03, `S'`(x)>0.
  Therefore, S has an absolute minimum at x=3.


                                          108
                         DS := x -> 4 x - ---
                                           2
                                          x


                                     108
                         eq := 4 x - --- = 0
                                      2
                                     x


                                      1/2                       1/2
 xsol := {x = 3}, {x = - 3/2 + 3/2 I 3   }, {x = - 3/2 - 3/2 I 3   }


                     optimal_width_in_inches := 3


                    optimal_length_in_inches := 6


                    optimal_height_in_inches := 2


                                                  2
                   minimum_surface_area := 54, in.

  "MOAPF" Supplement for Problems 16 and 17 (or, "Vectors Revisited")
  Yes, friends, that "Mother of All Physics Formulas" for position under
  constant acceleration. Here we'll derive it once and for all (why
  repeat history?), in an n-dimensional vector setting. Recognize that
  Problems 16 and 17 are applications of this same formula; only n
  differs! In  16, n=2, whereas in  17, n=1 (regarding scalars as
  vectors with one component). Not only that, you can use the same
  formula in Calc 2 and Calc 3 for motion in space (n=3) or in higher
  dimensional theaters of the mind...
  
   dv / dt = a     (The derivative of velocity in this case is constant
  acceleration. Here dv / dt is a vector derivative.)
   v = a t  +  C     (Antidifferentiate. The a t denotes scalar
  multiplication; C is a vector constant.)
   v(0) = C     (Resolve the vector constant C by plugging in
  the initial velocity v``[0] at time t=0.)
   dr / dt = v = a t  +  v(0)     (We now know the velocity, the
  derivative of position, as a function of time.)
   r = 1/2 a t^2  +  v(0) t  +  K     (Antidifferentiate again; the
  first rule of comedy: if it works once, it will work twice!)
   r(0) = K     (Resolve the vector constant K by plugging in
  the initial position r(0) at time t=0.)
   r = 1/2 a t^2  +  v(0) t  +  r(0)     (Finally, we have position
  as a function of time, the "MOAPF.")
  
  Now you can use this formula for any problem involving constant
  acceleration (as you will repeatedly in the first few weeks of your
  physics classes, perhaps next term). All you need do is change the
  constants from problem to problem (the constant acceleration, initial
  velocity, and initial position).
  
 White 16 Blue 18
  
  After having derived the MOAPF in the preceding supplement, we simply
  fill in the (vector) constants for this particular problem. Here the
  dimension is n=2, signifying motion in the plane. (NOTE: You did have
  to derive something for this problem to get full credit. The point of
  the supplement was that we don't need to derive the formula every time
  we do a problem like this: we derived a general formula.)
  General formula for position under constant acceleration.

                                     2
                    r := t -> 1/2 a t  + v0 t + r0

  Plug `em in!


                            a := [0, -9.8]


                                    1/2
                        v0 := [250 3   , 250]


                             r0 := [0, 0]


                          [       1/2                2        ]
       Voila!, position = [250 t 3   , -4.900000000 t  + 250 t]

 White 17 Blue 16
  
  Same drill as  16, but with a different dimension: n=1, or motion
  along a straight line.


                                     2
                    r := t -> 1/2 a t  + v0 t + r0


                               a := -20


                               r0 := 0


                                      2
                       r := t -> -10 t  + v0 t

  When the car has stopped, r(t)=160 and v(t)=r '(t)=0. This gives
  us two equations and two unknowns (final time and velocity), which we
  then solve.


                                        2
                      eq1 := 160 = -10 t  + v0 t


                        eq2 := 0 = -20 t + v0


                 {v0 = 80, t = 4}, {t = -4, v0 = -80}

  We discard the solution in which time is negative. Thus the initial
  velocity is +80 ft/s in the direction of motion of the car, and the
  speed (the magnitude of the velocity) is 80 ft/s, or 54.55 mph.
  
  White 18 Blue 17
  
  Hand approach: multiply the uniform stepsize
  h=Delta*x``=(b-a)/n``=(3-1)/4 ``=1/2 by the sum of the function values
  at the midpoints of the subintervals.


                                    2     1/2
                        f := x -> (x  + 1)


                               h := 1/2


                  midpoints := [5/4, 7/4, 9/4, 11/4]


                          1/2        1/2        1/2         1/2
      func_vals := [1/4 41   , 1/4 65   , 1/4 97   , 1/4 137   ]


                          1/2         1/2         1/2          1/2
   midpoint_rule := 1/8 41    + 1/8 65    + 1/8 97    + 1/8 137


                 numerical_equivalent := 4.502367463

  Here's an automated midpoint rule approximation using Maple's
  middlesum (in the student package) and its numerical equivalent.


                     /  3                          \
                     |-----                        |
                     | \                  2     1/2|
                 1/2 |  )   ((5/4 + 1/2 i)  + 1)   |
                     | /                           |
                     |-----                        |
                     \i = 0                        /


                 numerical_equivalent := 4.502367463

  
  DONE!