You are on page 1of 7

Project One

Due October 3rd, 2014

A Note on the Solutions: The solutions are interwoven with the original project. If you were to remove
the project parts, the solution would read correctly. It is in no way dependent on the assigned steps of the project
and they are only included to see how the particular questions were addressed.

Purpose: This project is designed for the student to develop a deeper understanding of mathematical con-
cepts connected to the course and to learn to communicate effectively about mathematics and technical content.
In particular, this project is designed to allow the student to investigate mathematical concepts by first under-
standing an argument thoroughly enough to apply the argument to a different situation and then to be able to
think critically enough to fill in and make rigorous the steps of another proof. Through explaining the steps taken
and reasoning behind them, the student will develop not only the ability to communicate mathematics effectively
but also the necessary critical thinking and problem solving skills that are important in life.

Details: This project is to be worked in groups of two or three. All answers are to be typed (or hand written
very clearly), correct, and fully explained. While some answers will require equations, steps that are not obvious
should be explained and if needed, motivated. That being said, steps should not be numbered as if you are
turning in a homework assignment. Any person (with enough mathematical background to know the terms) should
be able to pick up the paper and follow along with the argument.

Project

Main Goal: If we take our frame of reference so that the sun is at the origin and the earth is at the point (1, 0)
(thus the units are what is called astronomical units, approximately 1.496108 km), then there are five locations
L1 , L2 , L3 , L4 , and L5 in the plane of rotation of the earth about the sun where a satellite remains motionless
with respect to the earth (due to forces acting on the satellite from the earth and the sun). These locations are
called libration points. If m1 is the mass of the sun and m2 is the mass of the earth, and r = m1m+m 2
2
then the
x-coordinate of L1 is the unique root of the fifth-degree equation

p(x) = x5 (2 + r)x4 + (1 + 2r)x3 (1 r)x2 + 2(1 r)x + r 1 = 0. (1)

and the x-coordinate of L2 is the root of the equation

p(x) 2rx2 = 0. (2)

If we use the value r 3.04042 106 , find the locations of the libration points L1 and L2 .

1
Solution:
When placing a satellite in orbit around the earth (or any planet) it is desirable to place them in a location
that will minimize the energy required to keep them in place. These particular locations are known as the libration
points. If we consider plane of our solar system as a coordinate system with the Sun as the origin and the Earth
at the point (1, 0) (and thus we are working in astronomical units) then the first two libration points, L1 and L2 ,
would lie on the same axis as the Earth (i.e. the x-axis) which allows their position to be described with only one
variable (the other three libration points are not as easy to describe). In particular, L1 is the unique root of the
equation
p(x) = x5 (2 + r)x4 + (1 + 2r)x3 (1 r)x2 + 2(1 r)x + r 1 = 0 (3)
and L2 is the root of the equation
p(x) 2rx2 = 0 (4)
where r 3.04042 106 for the Sun and the Earth (in general it is r = m1m+m
2
2
where m1 is the mass of the larger
object -the Sun- and m2 is the mass of the smaller object - the Earth).
Unfortunately, these equations are both fifth-order equation and thus we have no formula for calculating the
roots (or zeros) of the equations. Since there seems to be no obvious way to find the roots precisely, we instead are
left attempting to approximate the roots through a numerical method, in particular we will use something called
Newtons Method.

Necessary Background Information


Since there is no general formula for a fifth degree polynomial, unless there is some clever trick to use, we may
need to hope for an approximation. To do so we develop something called Newtons Method. First we will want
to recall that the tangent line could be used to approximate the value of the function and at the point (x0 , f (x0 ))
was given by the equation y = f 0 (x0 )(x x0 ) + f (x0 ).

1. We want to use this to approximate our first guess of a zero x1 . We do this by finding where the linear
equation equals zero. Assuming that x1 is a zero of the linear equation, show that x1 = x0 ff0(x 0)
(x0 ) .

2. Now we use this new x-value, x1 and find the equation of the tangent line at x = x1 . Find the equation of
the tangent line.

3. Use the tangent line to approximate the next zero x2 .

4. Explain why the generalization of this process yields the formula xn+1 = xn ff0(x n)
(xn ) . This is called Newtons
Method of approximation.

Solution:
Newtons Method is actually based on using the tangent line to keep getting a closer and closer approximation
of the actual zero. For example, lets suppose that we have a function f (x) and we take an initial guess of

2
the zero to be x0 . Then we can find the equation of the tangent line at x0 which is

y = f 0 (x0 )(x x0 ) + f (x0 ).

While it is difficult to find the root of the original equation, it is not difficult to find the zero for a linear
equation (assuming that the slope is not zero). Moreover, if we assume that the linear equation is not a
horizontal line, we actually know that there is a zero, so we will call it x1 . If we let x = x1 then we get

f (x0 )
0 = f 0 (x0 )(x1 x0 ) + f (x0 ) x1 = x0 .
f 0 (x0 )

If the process worked correctly, we know have a better approximation for our zero than we did before. We
repeat the process by finding the tangent line at x1 . If the slope is again not zero, then we have a zero and
we can call it x2 . Plugging this in gives

f (x1 )
0 = f 0 (x1 )(x2 x1 ) + f (x1 ) x2 = x1 .
f 0 (x1 )

Notice that the formula looks the same as it did before. If we continued this process for x3 we would have
x3 = x2 ff0(x 2)
(x2 ) since each time we are really just solving the equation of the tangent line for x. Thus if we
continued this process to find the nth approximation we would have

f (xn1 )
xn = xn1
f 0 (xn1 )

which is Newtons Method of approximation.

Before we continue, we should practice a little with Newtons Method. There are a couple of important
things we need to be careful about though. The first is that we can always use the Intermediate Value
Theorem to guess our x0 (remember if the function is continuous and changes sign, then there was a zero
in between). The other deals with the shape of the graph. Since Newtons Method uses the derivative and
thus the slope of the tangent line, when looking to approximate a particular zero we may have to choose our
first guess carefully.

5. Use Newtons Method to find an approximation of 5 3 with a margin of error of less than 104 . (Hint: It is
the only solution to x5 3 = 0.)

Solution:
Before we return to the main goal and thus the harder task, we will practice
some with Newtons Method on
an easier example. For example, lets saywe wanted to approximate 5 3. Of course in order to use Newtons
Method we need an equation such that 5 3 is the root. For this we notice that it would be the only solution

3
to the equation x5 3 = 0. Thus we will let f (x) = x5 3 and thus f 0 (x) = 5x4 . Thus Newtons Method
looks like
x5 3
xn = xn1 n14 .
5xn1

While there are several initial guesses for a zero, we will try x0 = 1. Plugging in gives that x2 = 1.4.
Continuing this until the fifth decimal place no longer changes (i.e. accurate to 104 ). This gives that
x 1.2457.

6. Explain why Newtons Method will fail if the initial guess is somewhere where the derivative is zero.
It turns out that the derivative being zero is not the only trouble that we can encounter with Newtons
Method. The following theorem will tell us when Newtons Method is valid to use:

Solution:
Of course we should notice that in the previous example if we had encountered where the first derivative
was zero, then Newtons Method would have failed. In such a case, the equation of the tangent line would
have been y = f (x0 ) and thus as long as the function was not zero at x0 then there would be no zero for the
tangent line.

Thm: Let f : [a, b] < be a differentiable function.

(a) If f (a) < 0 < f (b), then Newtons Method will generate a sequence of values approaching a unique root
of f in [a, b] if one of the following conditions is satisfied (i.e. we will get a unique solution to f (x) = 0):
i. f is concave up on [a, b] and x0 := b.
ii. f is concave down on [a, b] and x0 := a.
(b) If f (a) > 0 > f (b), then Newtons Method will generate a sequence of values approaching a unique root
of f in [a, b] if one of the following conditions is satisfied (i.e. we will get a unique solution to f (x) = 0):
i. f is concave up on [a, b] and x0 := a.
ii. f is concave down on [a, b] and x0 := b.

4
Now we will practice with this before we return to the main goal.

7. Use Newtons Method to find the solutions of x2 + 4 = 0. Be sure to fully explain your steps.

Solution:
Unfortunately, the derivative being zero is not the only issue that can arise when trying to apply Newtons
Method to approximate a zero. Fortunately there is a theorem that aids in choosing an initial guess in a
way that Newtons Method will yield the correct approximate:
Thm: Let f : [a, b] < be a differentiable function.

(a) If f (a) < 0 < f (b), then Newtons Method will generate a sequence of values approaching a unique root
of f in [a, b] if one of the following conditions is satisfied (i.e. we will get a unique solution to f (x) = 0):
i. f is concave up on [a, b] and x0 := b.
ii. f is concave down on [a, b] and x0 := a.
(b) If f (a) > 0 > f (b), then Newtons Method will generate a sequence of values approaching a unique root
of f in [a, b] if one of the following conditions is satisfied (i.e. we will get a unique solution to f (x) = 0):
i. f is concave up on [a, b] and x0 := a.
ii. f is concave down on [a, b] and x0 := b.

Take for example the function f (x) = x2 + 4. Of course the derivatives are f 0 (x) = 2x and the second
derivative is f 00 (x) = 2. To locate the first positive zero (of course this is x = 2) we notice that f (1.9) > 0
and f (2.1) < 0. We also note that the function is concave down. Newtons Method takes the form of

x2n1 4
xn = xn1 .
2xn

By the theorem we know to take the right end-point as the initial guess of x0 = 2.1. We find within two
steps we are accurate to 5 decimal places. If we look for x = 2 (the other zero) we have that f (2.1) < 0
and f (1.9) > 0. Since the function is also concave down we use the initial guess of x0 = 2.1. Again after
two steps we are accurate to 5 decimal places.

Back to the Main Goal:


We are finally ready to return to the original problem.
First we find L1 (you may use a computer or calculator):

8. First find the derivative p0 (x) and write the general formula for Newtons Method.

5
9. Use the r-value provided to approximate the formula.

10. Show that the function changes sign between 0.95 and 1.

11. Using the concavity of the function, use Newtons Method to find the location accurate to 5 decimal places.

Solution: Now we wish to return to the original problem of finding the first libration point. We will need
the derivative, which is

p0 (x) = 5x4 4(2 + r)x3 + 3(1 + 2r)x2 2(1 r)x + 2(1 r).

Using this, we have that Newtons Method will look like


x5n1 (2 + r)x4n1 + (1 + 2r)x3n1 (1 r)x2n1 + 2(1 r)xn1 + r 1
xn = xn1 .
5x4n1 4(2 + r)x3n1 + 3(1 + 2r)x2n1 2(1 r)xn1 + 2(1 r)

Of course we can plug the particular value of r 3.04042 106 to approximate. Before we choose the
initial guess for the zero, we need to know if the function is concave up or down and where we should take
the initial guess. We first notice that f (0.95) < 0 and f (1) > 0. Next we can see that the function is
concave down on the interval by graphing the second derivative and thus by the theorem we take the initial
guess as x0 = 0.95. We repeat using Newtons Method until we are accurate to 5 decimal places, giving
L1 0.989989. Thus the libration point is at 0.989989 astronomical units from the Sun.

Next we find L2 (you may use a computer or calculator):

12. First find the derivative p0 (x) and write the general formula for Newtons Method.

13. Use the r-value provided to approximate the formula.

14. Show that the function changes sign between 1 and 1.02.

15. Using the concavity of the function, use Newtons Method to find the location accurate to 5 decimal places.

Solution: Now we wish to find the second libration point. First we find the function for the libration point,
which is
p(x) = x5 (2 + r)x4 + (1 + 2r)x3 (1 + r)x2 + 2(1 r)x + r 1 = 0. (5)

6
We will need the derivative, which is

p0 (x) = 5x4 4(2 + r)x3 + 3(1 + 2r)x2 2(1 + r)x + 2(1 r).

Using this, we have that Newtons Method will look like

x5n1 (2 + r)x4n1 + (1 + 2r)x3n1 (1 r)x2n1 + 2(1 r)xn1 + r 1


xn = xn1 .
5x4n1 4(2 + r)x3n1 + 3(1 + 2r)x2n1 2(1 + r)xn1 + 2(1 r)

Of course we can plug the particular value of r 3.04042 106 to approximate. Before we choose the initial
guess for the zero, we need to know if the function is concave up or down and what the initial guess should be.
We first notice that f (1) < 0 and f (1.02) > 0. Next we can see that the function is concave up on the interval (by
looking at the graph of the second derivative) and thus by the theorem we take the initial guess as x0 = 1.02. We
repeat using Newtons Method until we are accurate to 5 decimal places, giving L2 1.010078. Thus the libration
point is at 1.010078 astronomical units from the Sun.

This was adapted from Stewart Caculus 6E.

You might also like