AST1100 Lecture Notes

1–2 Celestial Mechanics

1 Kepler’s Laws
Kepler used Tycho Brahe’s detailed observations of the planets to deduce
three laws concerning their motion:
1. The orbit of a planet is an ellipse with the Sun in one of the foci.
2. A line connecting the Sun and the planet sweeps out equal areas in
equal time intervals.
3. The orbital period around the Sun and the semimajor axis (see figure
4 on page 8 for the definition) of the ellipse are related through:

P 2 = a3 , (1)

where P is the period in years and a is the semimajor axis in AU

(astronomical units, 1 AU = the distance between the Earth and
the Sun).
Whereas the first law describes the shape of the orbit, the second law
is basically a statement about the orbital velocity: When the planet is
closer to the Sun it needs to have a higher velocity than when far away
in order to sweep out the same area in equal intervals. The third law is a
mathematical relation between the size of the orbit and the orbital period.
As an example we see that when√ the semimajor axis doubles, the orbital
period increases by a factor 2 2 (do you agree?).
The first information that we can extract from Kepler’s laws is a relation
between the velocity of a planet and the distance from the Sun. When
the distance from the Sun increases, does the orbital velocity increase or
decrease? If we consider a nearly circular orbit, the distance traveled by
the planet in one orbit is 2πa, proportional to the semimajor axis. The
mean velocity can thus be expressed as vm√= 2πa/P which using Kepler’s
third law simply gives vm ∝ a/(a3/2 ) ∝ 1/ a (check that you understood
this!). Thus, the mean orbital velocity of a planet decreases the further
away it is from the Sun.

1
When Newton discovered his law of gravitation,
Gm1 m2
F~ = ~er ,
r2

he was able to deduce Kepler’s laws from basic principles. Here F~ is the
gravitational force between two bodies of mass m1 and m2 at a distance
r and G is the gravitational constant. The unit vector in the direction of
the force is denoted by ~er .

2 General solution to the two-body prob-

lem
Kepler’s laws is a solution to the two-body problem: Given two bodies with
mass m1 and m2 at a positions ~r1 and ~r2 moving with speeds ~v1 and ~v2
(see figure 1). The only force acting on these two masses is their mutual
gravitational attraction. How can we describe their future motion as a
function of time? The rest of this lecture will be devoted to this problem.
In order to solve the problem we will now describe the motion from the
rest frame of mass 1: We will sit on m1 and describe the observed motion
of m2 , i.e. the motion of m2 with respect to m1 . (As an example this
could be the Sun-Earth system, from the Earth you view the motion of
the Sun). The only force acting on m2 (denoted F~2 ) is the gravitational
pull from m1 . Using Newton’s second law for m2 we get
m1 m2
F~2 = −G ~r = m2~r¨2 , (2)
|~r |3

where ~r = ~r2 − ~r1 the vector pointing from m1 to m2 (or from the Earth
to the Sun in our example). Overdots describe derivatives with respect to
time,
d~r
~r˙ =
dt
2
d ~r
~r¨ = 2
dt

Sitting on m1 , we need to find the vector ~r(t) as a function of time (in

our example this would be the position vector of the Sun as seen from the
Earth). This function would completely describe the motion of m2 and be
a solution to the two-body problem (do you see this?).
Using Newton’s third law, we have a similar equation for the force acting
on m1
m1 m2
F~1 = −F~2 = G ~r = m1~r¨1 . (3)
|~r |3

2
 Figure 1: The two-body problem.

Subtracting equation (3) from (2), we can eliminate ~r1 and ~r2 and obtain
an equation only in ~r which is the variable we want to solve for,
m1 + m2 ~r
~r¨ = ~r¨2 − ~r¨1 = −G 3
~r ≡ −m 3 , (4)
|~r | r
where r = |~r | and m = G(m1 + m2 ). This is the equation of motion of
the two-body problem,
~r
~r¨ + m 3 = 0. (5)
r
We are looking for a solution of this equation with respect to ~r(t), this
would be the solution to the two-body problem predicting the movement
of m2 with respect to m1 .
To get further, we need to look at the geometry of the problem. We
introduce a coordinate system with m1 at the origin and with ~er and ~eθ
as unit vectors. The unit vector ~er points in the direction of m2 such that
~r = r~er and ~eθ is perpendicular to ~er (see figure 2). At a given moment,
the unit vector ~er (which is time dependent) makes an angle θ with a given
fixed (in time) coordinate system defined by unit vectors ~ex and ~ey . From
figure 2 we see that (do you really see this? Draw some figures to convince
yourself!)
~er = cos θ~ex + sin θ~ey
~eθ = − sin θ~ex + cos θ~ey
The next step is to substitute ~r = r~er into the equation of motion (equa-
tion 5). In this process we will need the time derivatives of the unit
vectors,
~e˙ r = −θ̇ sin θ~ex + θ̇ cos θ~ey
= θ̇~eθ
~e˙ θ = −θ̇ cos θ~ex − θ̇ sin θ~ey
= −θ̇~er

3
 Figure 2: Geometry of the two-body problem.

Using this, we can now take the derivative of ~r = r~er twice,

~r˙ = ṙ~er + r~e˙ r

= ṙ~er + rθ̇~eθ
~r¨ = r̈~er + ṙ~e˙ r + (ṙθ̇ + rθ̈)~eθ + rθ̇~e˙ θ
1d 2
= (r̈ − rθ̇2 )~er + (r θ̇)~eθ .
r dt
Substituting ~r = r~er into the equation of motion (equation 5), we thus
obtain
1d 2 m
(r̈ − rθ̇2 )~er + (r θ̇)~eθ = − 2 ~er .
r dt r
Equating left and right hand sides, we have
m
r̈ − rθ̇2 = − (6)
r2
d 2
(r θ̇) = 0 (7)
dt
The vector equation (equation 5) has thus been reduced to these two scalar
equations. Go back and check that you understood the transition.
The last of these equations indicates a constant of motion, something
which does not change with time (why?). What constant of motion enters
in this situation? Certainly the angular momentum of the system should
be a constant of motion so let’s check the expression for the angular mo-
mentum vector ~h (note that h is defined as angular momentum per mass,
(~r ×~p)/m2 (remember that m1 is at rest in our current coordinate frame)):

|~h| = |~r × ~r˙ | = |(r~er ) × (ṙ~er + rθ̇~eθ )| = r2 θ̇.

So equation (7) just tells us that the magnitude of the angular momentum
h = r2 θ̇ is conserved, just as expected.
To solve the equation of motion, we are left with solving equation (6). In
order to find a solution we will

4
 1. solve for r as a function of angle θ instead of time t. This will give us
the distance of the planet as a function of angle and thus the orbit.
2. Make the substitution u(θ) = 1/r(θ) and solve for u(θ) instead of
r(θ). This will transform the equation into a form which can be
easily solved.
In order to substitute u in equation (6), we need its derivatives. We start
by finding the derivatives of u with respect to θ,
du(θ) dt ṙ 1 ṙ
= u̇ = − 2 = −
dθ dθ r θ̇ h
2
d u(θ) 1 d 1 1
2
=− ṙ = − r̈ .
dθ h dθ h θ̇
In the last equation, we substitute r̈ from the equation of motion (6),
d2 u(θ) 1 m m 1 m
2
= ( 2 − rθ̇2 ) = 2 − = 2 − u,
dθ hθ̇ r h r h
where the relation h = r2 θ̇ was used twice. We thus need to solve the
following equation
d2 u(θ) m
+ u =
dθ2 h2
This is just the equation for a harmonic oscillator (if you have not encoun-
tered the harmonic oscillator in other courses yet, it will soon come, it is
simply the equation of motion for an object which is attached to a spring
in motion) with known solution:
m
u(θ) = 2 + A cos (θ − ω),
h
where A and ω are constants depending on the initial conditions of the
problem. Try now to insert this solution into the previous equation to
see that this is indeed the solution. Substituting back we now find the
following expression for r:
The general solution to the two-body problem
p
r= (8)
1 + e cos f

where p = h2 /m, e = (Ah2 /m) and f = θ − ω.

We recognize this expression as the general expression for a conic section.

3 Conic sections
Conic sections are curves defined by the intersection of a cone with a
plane as shown in figure 3. Depending on the inclination of the plane,
conic sections can be divided into three categories with different values of
p and e in the general solution to the two-body problem (equation 8),

5
Figure 3: Conic sections: Circle: e=0,p=a, Ellipse: 0 ≤ e < 1, p =
a(1 − e2 ), Parabola: e = 1, p = 2a, Hyperbola: e > 1 and p = a(e2 − 1)

1. the ellipse, 0 ≤ e < 1 and p = a(1 − e2 ) (of which the circle, e = 0,

is a subgroup),
2. the parabola, e = 1 and p = 2a,
3. the hyperbola, e > 1 and p = a(e2 − 1).
In all these cases, a is defined as a positive constant a ≥ 0. Of these curves,
only the ellipse represents a bound orbit, in all other cases the planet just
passes the star and leaves. We will discuss the details of an elliptical orbit
later. First, we will check which conditions decides which trajectory an
object will follow, an ellipse, parabola or hyperbola. Our question is thus:
If we observe a planet or other object close to a star, is it in orbit around
the star or just passing by? For two masses to be gravitationally bound,
we expect that their total energy, kinetic plus potential, would be less
than zero, E < 0. Clearly the total energy of the system is an important
initial condition deciding the shape of the trajectory.
We will now investigate how the trajectory r(θ) depends on the total
energy. In the exercises you will show that the total energy of the system
can be written:
Total energy of a two-body system
1 µ̂m
E = µ̂v 2 − , (9)
2 r

where v = |~r˙ |, the velocity of m2 observed from m1 (or vice versa) and
µ̂ = m1 m2 /(m1 + m2 ).

We will now try to rewrite the expression for the energy E in a way which
will help us to decide the relation between the energy of the system and
the shape of the orbit. We will start by rewriting the velocity in terms of

6
its radial and tangential components using the fact that ~v = ~r˙ = ṙ~er + r~e˙ r

v 2 = vr2 + vθ2 = ṙ2 + (rθ̇)2 , (10)

decomposed into velocity along ~er and ~eθ (check that you got this!). We
need the time derivative of r. Taking the derivative of equation (8),
pe sin f
ṙ = θ̇,
(1 + e cos f )2
we get from equation (10) for the velocity

2 2p2 e2 sin2 f
v = θ̇ 4
+ r2 θ̇2 .
(1 + e cos f )

Next step is in both terms to substitute θ̇ = h/r2 (where did this come
from?) and then using equation (8) for r giving

2 h2 e2 sin2 f h2 (1 + e cos f )2
v = + .
p2 p2
Collecting terms and remembering that cos2 f + sin2 f = 1 we obtain
h2
v2 = (1 + e2 + 2e cos f ).
p2
We will now get back to the expression for E. Substituting this expression
for v as well as r from equation (8) into the energy expression (equation
9), we obtain
1 h2 1 + e cos f
E = µ̂ 2 (1 + e2 + 2e cos f ) − µ̂m (11)
2 p p
Total energy is conserved and should therefore be equal at any point in
the orbit, i.e. for any angle f . We may therefore choose an angle f which
is such that this expression for the energy will be easy to evaluate. We
will consider the energy at the point for which cos f = 0,
1 h2 µ̂m
E = µ̂ 2 (1 + e2 ) −
2 p p
We learned above (below equation 8) that p = h2 /m and thus that h =
√
mp. Using this to eliminate h from the expression for the total energy
we get
µ̂m 2
E= (e − 1).
2p
If the total energy E = 0 then we immediately get e = 1. Looking back at
the properties of conic sections we see that this gives a parabolic trajectory.
Thus, masses which have just too much kinetic energy to be bound will
follow a parabolic trajectory. If the total energy is different from zero, we
may rewrite this as
µ̂m 2
p= (e − 1).
2E

7
We now see that a negative energy E (i.e. a bound system) gives an ex-
pression for p following the expression for an ellipse in the above list of
properties for conic sections (by defining a = µ̂m/(2|E|). Similarly a pos-
itive energy gives the expression for a hyperbola. We have shown that
the total energy of a system determines whether the trajectory will be
an ellipse (bound systems E < 0), hyperbola (unbound system E > 0)
or parabola (E = 0). We have just shown Kepler’s first law of motion,
stating that a bound planet follows an elliptical orbit. In the exercises
you will also show Kepler’s second and third law using Newton’s law of
gravitation.

4 The elliptical orbit

We have seen that the elliptical orbit may be written in terms of the
distance r as
a(1 − e2 )
r= .
1 + e cos f

In figure (4) we show the meaning of the different variables involved in

this equation:
• a is the semimajor axis
• b is the semiminor axis
p
• e is the eccentricity defined as e = 1 − (b/a)2
• m1 is located in the principal focus
• the point on the ellipse closest to the principal focus is called peri-
helion
• the point on the ellipse farthest from the principal focus is called
aphelion
• the angle f is called the true anomaly

Figure 4: The ellipse.

8
The eccentricity is defined using the ratio b/a. When the semimajor and
semiminor axis are equal, e = 0 and the orbit is a circle. When the
semimajor axis is much larger than the semiminor axis, e → 1.

5 Center of mass system

In the previous section we showed that seen from the rest frame of one
of the masses in a two-body system, the other mass follows an elliptical
/ parabolic / hyperbolic trajectory. How does this look from a frame
of reference which is not at rest with respect to one of the masses? We
know that both masses m1 and m2 are moving due to the gravitational
attraction from the other. If we observe a distant star-planet system, how
does the planet and the star move with respect to each other? We have
only shown that sitting on either the planet or the star, the other body
will follow an elliptical orbit.
An elegant way to describe the full motion of the two-body system (or in
fact an N-body system) is to introduce center of mass coordinates. The
center of mass position R ~ is located at a point on the line between the
two masses m1 and m2 . If the two masses are equal, the center of mass
position is located exactly halfway between the two masses. If one mass
is larger than the other, the center of mass is located closer to the more
massive body. The center of mass is a weighted mean of the position of
the two masses:
~ = m1 ~r1 + m2 ~r2 ,
R (12)
M M
where M = m1 + m2 . We can similarly define the center of mass for an
N-body system as
N
~
X mi
R= ~ri , (13)
i=1
M
P
where M = i mi and the sum is over all N masses in the system. New-
ton’s second law for one object in the system is

f~i = mi~r¨i

where f~i is the total force on object i. Summing over all bodies in the
system, we obtain Newton’s second law for the full N-body system
N
X
F~ = mi~r¨i , (14)
i=1

where F~ is the total force on all masses in the system. We may divide the
total force on all masses into one contribution from internal forces between
masses and one contribution from external forces,
XX
F~ = f~ij + F~ext ,
i j6=i

9
Figure 5: Info-figure: A diagram of the trajectory that enabled NASA’s
Voyager 2 spacecraft to tour all the four gas giants and achieve a large
enough velocity to escape our solar system. Celestial mechanics obviously
played an integral part in the extremely careful planning that was needed
in order to carry out the probe’s ambitious tour of the outer solar system.
The planetary flybys not only allowed for close-up observations of the
planets and their moons, but also accelerated the probe so that it could
reach the next object. In 2012 Voyager 2 was at a distance of roughly
100 AU from the Sun, traveling outward at around 3.3 AU per year. It is
expected to keep transmitting weak radio messages until at least 2025.

10
Figure 6: The center of mass system: The center of mass (CM) is indicated
by a small point. The two masses m1 and m2 orbit the center of mass in
elliptical orbits with the center of mass in one focus of both ellipses. The
center of mass vectors ~r1CM and ~r2CM start at the center of mass and point
to the masses.

where f~ij is the gravitational force on mass i from mass j. Newton’s third
law implies that the sum over all internal forces vanish (f~ij = −f~ji ). The
right side of equation (14) can be written in terms of the center of mass
coordinate using equation (13) as
N
~¨
X
mi~r¨i = M R,
i=1

giving
~¨ = F~ext .
MR
(Check that you followed this deduction!). If there are no external forces
on the system of masses (F~ext = 0), this equation tells us that the center
of mass position does not accelerate, i.e. if the center of mass position is
at rest it will remain at rest, if the center of mass position moves with a
given velocity it will keep moving with this velocity. We may thus divide
the motion of a system of masses into the motion of the center of mass and
the motion of the individual masses with respect to the center of mass.
We now return to the two-body system assuming that no external forces
act on the system. The center of mass moves with constant velocity and
we decide to deduce the motion of the masses with respect to the center
of mass system, i.e. the rest frame of the center of mass. We will thus be
sitting at the center of mass which we define as the origin of our coordinate
system, looking at the motion of the two masses. When we know the

11
motion of the two masses with respect to the center of mass, we know the
full motion of the system since we already know the motion of the center
of mass position.
~ = 0.
Since we take the origin at the center of mass location, we have R
Using equation (12) we get
m1 CM m2 CM
0= ~r + ~r ,
M 1 M 2
where CM denotes position in the center of mass frame (see figure 6).
Combining this equation with the fact that ~r = ~r2 − ~r1 = ~r2CM − ~r1CM we
obtain
µ̂
~r1CM = − ~r, (15)
m1
µ̂
~r2CM = ~r, (16)
m2

The reduced mass µ̂ is defined as

m1 m2
µ̂ = .
m1 + m2

The relative motion of the masses with respect to the center of mass can
be expressed in terms of ~r1CM and ~r2CM as a function of time, or as we have
seen before, as a function of angle f . We already know the motion of one
mass with respect to the other,
p
|~r | = .
1 + e cos f

Inserting this into equations (15) and (16) we obtain

µ̂ µ̂p
|~r1CM | = |~r | =
m1 m1 (1 + e cos f )
µ̂ µ̂p
|~r2CM | = |~r | =
m2 m2 (1 + e cos f )

For a bound system we thus have

µ̂
m1
a(1 − e2 ) a1 (1 − e2 )
|~r1CM | = ≡
1 + e cos f 1 + e cos f
µ̂
m2
a(1 − e2 ) a2 (1 − e2 )
|~r2CM | = ≡
1 + e cos f 1 + e cos f
We see from these equations that for a gravitationally bound system, both
masses move in elliptical orbits with the center of mass in one of the foci
(how do you see this?). The semimajor axis of these two masses are given

12
by
µ̂a
a1 = ,
m1
µ̂a
a2 = ,
m2
a = a1 + a2

(check that you understand how these equations come about) where a1
and a2 are the semimajor axis of m1 and m2 respectively and a is the
semimajor axis of the elliptical orbit of one of the masses seen from the
rest frame of the other. Note that the larger the mass of a given body
with respect to the other, the smaller the ellipse. This is consistent with
our intuition: The more massive body is less affected by the same force
than is the less massive body. The Sun moves in an ellipse around the
center of mass which is much smaller than the elliptical orbit of the Earth.
Figure (6) shows the situation: the planet and the star orbit the common
center of mass situated in one common focus of both ellipses.

6 Problems
Problem 1 (20–45 min.)
The scope of this problem is to deduce Kepler’s second law. Kepler’s
second law can be written mathematically as
dA
= constant,
dt
i.e. that the area A swept out by the vector ~r per time interval is constant.
We will now show this step by step:
1. Show that the infinitesimal area dA swept out by the radius vector
~r for an infinitesimal movement dr and dθ is dA = 21 r2 dθ.
2. Divide this expression by dt and you obtain an expression for dA/dt
in terms of the radius r and the tangential velocity vθ .
3. By looking back at the above derivations, you will see that the tan-
gential velocity can be expressed as vθ = h/r.
4. Show Kepler’s second law.

Problem 2 (20–45 min.)

The scope of this problem is to deduce Kepler’s third law. Again we will
solve this problem step by step:
1. In the previous problem we found an expression for dA/dt in terms
of a constant. Integrate this equation over a full period P and show

13
 that
2πab
P =
h
(Hint: the area of an ellipse is given by πab).
2. Use expressions for h and b found in the text to show that

4π 2
P2 = a3 (17)
G(m1 + m2 )

3. This expression obtained from Newtonian dynamics differs in an

important way from the original expression obtain empirically by
Kepler (equation 1). How? Why didn’t Kepler discover it?

Problem 3 (15–30 min.)

1. How can you measure the mass of a planet in the solar system by
observing the motion of one of its satellites? Assume that we know
only the semimajor axis and orbital period for the elliptical orbit
of the satellite around the planet. Hint 1: Kepler’s third law (the
exact version). Hint 2: You are allowed to make reasonable ap-
proximations.
2. Look up (using Internet or other sources) the semimajor axis and
orbital period of Jupiter’s moon Ganymede.
(a) Use these numbers to estimate the mass of Jupiter.
(b) Then look up the mass of Jupiter. How well did your esti-
mate fit? Is this an accurate method for computing planetary
masses?
(c) Which effects could cause discrepancies from the real value and
your estimated value?

Problem 4 (70–90 min.)

1. Show that the total energy of the two-body system in the center of
mass frame can be written as
1 GM µ̂
E = µ̂v 2 − ,
2 r
where v = |d~r/dt| is the relative velocity between the two objects,
r = |~r | is their relative distance, µ̂ is the reduced mass and M ≡
m1 + m2 is the total mass. Hint: make the calculation in the center
of mass frame and use equation (15) and (16).
2. Show that the total angular momentum of the system in the center
of mass frame can be written

P~ = µ̂~r × ~v ,

14
 3. Looking at the two expressions you have found for energy and an-
gular momentum of the system seen from the center of mass frame:
Can you find an equivalent two-body problem with two masses m01
and m02 where the equations for energy and momentum will be of
the same form as the two equations which you have just derived?
What are m01 and m02 ? If you didn’t understand the question, here is
a rephrasing: If you were given these two equations without know-
ing anything else, which physical system would you say that the
describe?

Problem 5 (optional 30–45 min.)

1. At which points in the elliptical orbit (for which angels f ) is the
velocity of a planet at maximum or minimum?
2. Using only the mass of the Sun, the semimajor axis and eccentricity
of Earth’s orbit (which you look up in Internet or elsewhere), can
you find an estimate of Earth’s velocity at aphelion and perihelion?
3. Look up the real maximum and minimum velocities of the Earth’s
velocity. How well do they compare to your estimate? What could
cause discrepancies between your estimated values and the real val-
ues?
4. Use Python (or Matlab or any other programming language) to plot
the variation in Earth’s velocity during one year.
Hint 1: Use one or some of the expressions for velocity found in section
(3) as well as expressions for p and h found in later sections (including the
above problems). Hint 2: You are allowed to make reasonable approxi-
mations.

Problem 6 (optional 10–30 min.)

1. Find our maximum and minimum distance to the center of mass of
the Earth-Sun system.
2. Find Sun’s maximum and minimum distance to the center of mass
of the Earth-Sun system.
3. How large are the latter distances compared to the radius of the
Sun?

Problem 7: Numerical solution to the 2-body/3-body problem

In this problem you are first going to solve the 2-body problem numerically
by a well-known numerical method. We will start by considering the ESA
satellite Mars Express which entered an orbit around Mars in December
2003 (http://www.esa.int/esaMI/Mars\_Express/index.html). The
goal of Mars Express is to map the surface of Mars with high resolution

15
images. When Mars Express is at an altitude of 10107 km above the
surface of Mars with a velocity 1166 m/s (with respect to Mars), the
engines are turned off and the satellite has entered the orbit.
In this exercise we will use that the radius of Mars is 3400 km and the
mass of Mars is 6.4 × 1023 kg. Assume the weight of the Mars Express
spacecraft to be 1 ton.
1. (4–5 hours) A distance of 10107 km is far too large in order to
obtain high resolution images of the surface. Thus, the orbit of
Mars Express need to be very eccentric such that it is very close to
the surface of Mars each time it reaches perihelion. We will now
check this by calculating the orbit of Mars Express numerically. We
will introduce a fixed Cartesian coordinate system to describe the
motion of Mars and the satellite. Assume that at time t = 0 Mars
has position [x1 = 0, y1 = 0, z1 = 0] and Mars express has position
[x2 = 10107 + 3400 km, y2 = 0, z2 = 0] in this fixed coordinate
system (see Figure 8).
The velocity of Mars express is only in the positive y-direction at
this moment. In our fixed coordinate system, the initial velocity
vectors are therefore ~v1 = 0 (for Mars) and ~v2 = 1166 ms ~j (for Mars
Express), where ~j is the unitvector along the y-axis. There is no
velocity component in the z-direction so we can consider the system
as a 2-dimensional system with movement in the (x, y)-plane.
Use Newton’s second law,
d2~r
m 2 = F~ ,
dt
to solve the 2-body problem numerically. Use the Euler-Cromer
method for differential equations. Plot the trajectory of Mars Ex-
press. Do 105 calculations with timestep dt = 1 second. Is the result
what you would expect?
Hints - Write Newton’s second law in terms of the velocity vector.
 
d~v ~ dv~x~ dv~y ~
m =F ⇒m i+ j = Fx~i + Fy~j
dt dt dt
Then we have the following relation between the change in the com-
ponents of the velocity vector and the components of the force vector;
dvx Fx
=
dt m
dvy Fy
=
dt m
These equations can be solved directly by the Euler-Cromer method
and the given initial conditions. For each timestep (use a for- or
while-loop), calculate the velocity vx/y (t + dt) (Euler’s method) and
the position x(t + dt), y(t + dt) (standard kinematics) for Mars and

16
Figure 7: Info-figure: 433 Eros was the target of the first long-term, close-
up study of an asteroid. After a four year journey the NEAR-Shoemaker
space probe was inserted into orbit around the 33 km long, potato-shaped
asteroid in February 2000 and encircled it 230 times from various distances
before touching down on its surface. The primary scientific objective was
to return data on the composition, shape, internal mass distribution, and
magnetic field of Eros. Asteroids are a class of rocky small solar system
bodies that orbit the Sun, mostly in the asteroid belt between Mars and
Jupiter. They are of great interest to astronomers as they are leftover
material from when the solar system formed some 4.6 billion years ago..

Figure 8: Mars and Mars Express at time t = 0.

17
 the space craft (or even easier: use vectors). A good approach is to
make a function that calculates the gravitational force. From the
loop you send the previous timestep positiondata (4 values) to the
function which calculates and returns the components of the force (4
values, two for Mars and two for the space craft) with correct nega-
tive/positive sign. Collect the values in arrays and use the standard
python/scitools plot-command.
2. (60–90 min) Mars Express contained a small lander unit called
Beagle 2. Unfortunately contact was lost with Beagle 2 just after
it should have reached the surface. Here we will calculate the path
that Beagle 2 takes down to the surface (this is not the real path
that was taken). We will assume that the lander does not have any
engines and is thus moving under the influence of only two forces: the
force of gravity from Mars and the force of friction from the Martian
atmosphere. The friction will continously lower the altitude of the
orbit until the lander hits the surface of the planet. We will assume
the weight of the lander to be 100 kg.
We will now assume that Mars Express launches Beagle 2 when Mars
Express is at perihelion. We will assume that it adjusts the velocity
of Beagle 2 such that it has a velocity of 4000 m/s (with respect
to Mars) at this point. Thus we have 2-body problem as in the
previous exercise. At t = 0, the position of Mars and the lander is
[x1 = 0, y1 = 0, z1 = 0] and [x2 = −298 − 3400 km, y2 = 0, z2 = 0]
respectively (Figure 9). The initial velocity vector of the lander is
v~1 = −4000 ms~j with respect to Mars. Due to Mars’atmosphere a
force of friction acts on the lander which is always in the direction
opposite to the velocity vector. A simple model of this force is given
by
f~ = −k~v ,
where k = 0.00016 kg/s is the friction constant due to Mars’ at-
mosphere. We will assume this to have the same value for the full
orbit.
Plot the trajectory that the lander takes down to the surface of Mars.
Set dt = 1 second.
Hints - You can use most of the code from the previous exercise.
First, we write Newton’s second law in terms of the cartesian com-
ponents;
dvx Fx + fx
=
e m
dvy Fy + fy
=
dt m
(or use the vector from direcly if you prefer). The best approach is
to make one more function that calculates the force of friction with
the lander’s velocity components as arguments. In this problem you

18
 Figure 9: Mars and Beagle 2 at time t = 0.

should use a while-loop. For each evaluation, first call the gravita-
tional function (as before) and then the friction function. Remem-
ber to send the space craft’s velocity components from the previous
timestep. In the friction function, first calculate the total force F ,
and then (if you do not use vectors) the components Fx and Fy (use
simple trigonometry) with correct positive/negative-sign by check-
ing the sign of the velocity components. Then return the force com-
ponents to the loop. For each evaluation (in the while-loop) check
whether the spaceship has landed or not.
3. Use the trajectory of the previous exercise to check the landing site:
Was the lander supposed to study the ice of the Martian poles or the
rocks at the Martian equator? Use figure 9 to identify the position
of the poles with respect to the geometry of the problem (the result
does not have any relation with the objectives or landing site of the
real Beagle 2 space craft)
4. (90–120 min.) Finally, we will use our code to study the 3-body
problem. There is no analytical solution to the 3-body problem, so
in this case we are forced to use numerical calculations. The fact that
most problems in astrophysics consider systems with a huge number
of objects strongly underlines the fact that numerical solutions are
of great importance.
About half of all the stars are binary stars, two stars orbiting a
common center of mass. Binary star systems may also have planets
orbiting the two stars. Here we will look at one of many possible
shapes of orbits of such planets. We will consider a planet with the
mass identical to the mass of Mars. One of the stars has a mass
identical to the mass of the Sun (2 × 1030 kg), the other has a mass
4 times that of the Sun.
The initial positions are [x1 = −1.5 AU, y1 = 0, z1 = 0] (for the
planet), [x2 = 0, y2 = 0, z2 = 0] (for the small star) and [x3 = 3 AU,
y3 = 0, z3 = 0] (for the large star) (Figure 3). The initial velocity
vectors are v~1 = −1 km ~j (for the planet), v~2 = 30 km~j (for the small
s s

19
 Figure 10: The binary star system with the planet at time t = 0.

km~
star) and v~3 = −7.5 s
j (for the large star).
Plot the orbit of the planet and the two stars in the same figure.
Use timestep dt = 400 seconds and make 106 calculations. It should
now be clear why it is impossible to find an analytical solution to
the 3-body problem. Note that the solution is an approximation. If
you try to change the size and number of time steps you will get
slighly different orbits, small time steps cause numerical problems
and large time steps is too inaccurate. The given time step is a
good trade-off between the two problems but does not give a very
accurate solution. Accurate methods to solve this problem is outside
the scope of this course. Play around and try some other starting
positions and/or velocities.
Hints: There is really not much more code you need to add to the
previous code to solve this problem. Declare arrays and constants
for the three objects. In your for-/while-loop, calculate the total
force components for each object. Since we have a 3-body problem
we get two contributions to the total force for each object. In other
words, you will have to call the function of gravitation three times
for each time-evaluation. For each time step, first calculate the force
components between the planet and the small star, then the force
components between the planet and the large star, and finally the
force components between the small and the large star. Then you
sum up the contributions that belong to each object.
5. Look at the trajectory and try to imagine how the sky will look like
at different epochs. If we assume that the planet has chemical condi-
tions for life equal to those on earth, do you think it is probable that
life will evolve on this planet? Use your tracetory to give arguments.

20