Modeling Flight over a Spherical Earth

David Zitter

May 9, 2008

1 Introduction
Deriving the equations of motion of a point-mass projectile moving in a vacuum solely under the
influence gravity is one of the first things we learn regarding the physics of flight. This model of flight is
of course greatly simplified and is not an accurate enough description of the motion of objects for
almost all practical modern-day settings. We wish to expand the model of flight to be able to more
accurately model a ballistic missile whose trajectory is flown in the immediate neighborhood of the
Earth. Beyond simply modeling a rocket that has variable thrust and mass, we will take into account
that the surface over which the rocket is traveling, the earth, is spherical in shape, rotates around its
axis, and that the strength of gravity varies depending on the altitude of the missile. We will examine a
more specific case of flight in that we will study a class of paths which are flown in a great-circle plane.
A great circle is a circle on the surface of a sphere, in this case the Earth, that has the same
circumference as the sphere. The corresponding great-circle plane is the plane which contains this great
circle. The great circle we will be choosing will be perpendicular to the equator, or along a fundamental
parallel, which will be defined in more detail later.

2 The Model
If we assume that the rocket is a solid body (i.e. there is no fluid within the rocket that could affect
its motion) and take into account the Coriolis acceleration acting on the rocket that results from the
rotation of the Earth, the general dynamical equation that describes the absolute motion of the rocket is



+  +  =   +2 × 

(1)

where
is the thrust produced by the rocket,  is the total aerodynamic force,  is the mass of the
rocket, is angular velocity of the Earth with respect to the Fixed Stars, and  is the acceleration of
gravity. Also, is the velocity of the rocket with respect to the Earth and is given by


=

(2)

where  is the vector joining point  on the surface of the Earth with point  on the rocket, as shown
in Figure 1.
 We wish to find the scalar equations associated with this vectorial relationship so that we can
find the location of the rocket on the sphere at a specific time under given initial conditions. To do this
we must define several relevant coordinate systems and then establish the relationships between
position and motion in one system with respect to another. We do this because many of the forces
involved are produced by the rocket and are easiest to implement in the rocket’s frame of reference,
but ultimately we want to find the rocket’s position in the Earth’s frame of reference, so we have to be
able to transform between the two systems.

2.1 Coordinate Systems

   , the curvilinear ground system , the local horizon system    , and the body axes
We employ four different coordinate systems in our investigation; the Earth axes system

system    .

origin E is a point on the Earth’s surface; the  -axis is vertical and positive downward; the  -axis and
The Earth axes system is a Cartesian reference frame which is rigidly attached to the Earth. Its

the  -axis are tangent to the Earth’s surface and are directed such that the system is right-handed. The
the great circle tangent to the  -axis is called the fundamental meridian, while the great circle tangent
to the  -axis, our desired path, is called the fundamental parallel.

origin  is a point on the Earth’s surface; the -coordinate is measured from  on the fundamental
The curvilinear ground system is an orthogonal reference frame which is fixed to the Earth. Its

parallel; the -coordinate is measured from  on the fundamental meridian; and the -coordinate is
measured radially from . Furthermore, the positive senses for , ,  are consistent with the positive
sense for the Earth axes. These three coordinate are sufficient to determine the position of the rocket
with respect to the Earth.

position of the aircraft ; the  -axis is vertical and positive downward; the  -axis and the  -axis are
The local horizon system is a Cartesian reference frame that has its origin at the instantaneous

   is right-handed. This sphere surface, which has the same center as the Earth sphere, is called
contained in the plane tangent to the spherical surface passing through the aircraft and are such that

the local horizon.

The body axes system also has its origin at ; the  -axis is tangent to the flight path and is
positive forward; the  -axis is perpendicular to the  -axis, contained in the plane of symmetry of the
rocket, and positive downward for the normal flight attitude of the aircraft; the  -axis is directed such
that    is right-handed.
 Figure 1. Depicts the Earth and local horizon axes, as well as the intermediate axes used to transform
between the two. Also shows the curvilinear ground system and labels all points and angles as defined in the
previous and next sections. The rocket travels above the labeled fundamental parallel, which is constructed
based on the chosen location of E, the origin of the Earth system.

2.2 Angular Relationships

We wish to be able to convert from the body axes system to the Earth axes system and vice
versa since most of the forces act on either the rocket or the Earth. We will therefore investigate the
angular relationships between the local horizon-Earth axes and body axes-local horizon pairs.

two angular parameters, the longitude  and the latitude . Looking at Figure 1 we can see that the
The orientation of the local horizon with respect to the Earth axes can be described in terms of

intermediate coordinate systems are introduced. The system    is obtained from the Earth axes
necessary rotations to transform from one system to the other are easily understood if two

system by means of a rotation  around the  -axis plus a translation; the system !" " " is obtained
from    by means of a rotation  around the  -axis plus a translation; finally, the local horizon
system is such that its axes and the corresponding axes of the system !" " " are parallel and have the

coordinates and the angles of rotation are related by X = r%  and Y = r% λ, where r% is the radius of the
same positive sense. Using the arc length formula, from Figure 1 it is clear that the curvilinear

Earth. Using matrix-vector notation we can therefore represent the total rotation in the form
 i cos1X⁄23 5 0 sin1X⁄23 5 i
( j , = - −sin1X⁄23 5 sin1Y⁄23 5 cos1Y⁄23 5 cos1X⁄r% 5 sin1X⁄23 5 9 ( j ,.
k −sin1X⁄23 5 cos1Y⁄23 5 −sin1Y⁄23 5 cos1X⁄r% 5 cos1Y⁄23 5 k
(3)

two angular parameters, the velocity yaw ; and the velocity pitch <. Using a similar process as for the
The orientation of the body axes with respect to the local horizon can be described in terms of

system = = = is obtained from the local horizon by means of a rotation ; around the  -axis and the
local horizon-Earth axes transformation, we introduce an intermediate coordinate system such that the

wind axes system is obtained from = = = by means of a rotation < around the = -axis. The velocity
yaw essentially measures angular deviation of the  body axis from the  local horizon axis, which is

the  body axis from the  local horizon axis in the plane containing the  and  axes. Therefore
contained in the plane normal to the local horizon. The velocity pitch measures the angular deviation of

the total rotation takes the form

i> cos γ cos χ cos γ sin χ − sin γ i

( j> , = ( − sin χ cos χ 0 , ( j ,.
k> sin γ cos χ sin γ sin χ cos γ k 
(4)

2.3 Angular Velocity

Because we are dealing with a rotating Earth, we must take into account the forces acting on the
rocket that come from the angular velocity of the Earth. We therefore need to determine how the
angular velocity is transformed when going from the Earth axes system to the local horizon system.

When going from the Earth axes to the local horizon system, we rotated along the  and 
axes. The infinitesimal angular displacement of the local horizon with respect to the Earth is therefore
given by

A = λB − τD .

Consequently, the angular velocity of the local horizon with respect to the Earth becomes

A
E = = λFB − τF D


which, in consideration of Eq. (3) and the transformation matrices to the intermediate coordinate
systems, can be rewritten in the form

F F F
(5)
E = B − cos1Y⁄r% 5D + sin1Y⁄r% 5 G  .
23 23 23

We also need to know the evolutory velocity, that is, the angular velocity of the body axes with
respect to the Earth axes, which we will represent as

 = H> B + q  D + r G . (6)

2.4 Kinematic Relationships

Because the velocity of the rocket is collinear with the  -axis, we can write the velocity as
We must now derive the scalar relationships corresponding to the vectorial equation (2).

= JB = J1cos γ cos χ B + cos γ sin χ D − sin γG  5

(7)

after using the transformation matrix in Eq. (4). We can then rewrite the vector joining the origin of the
Earth axes system with the aircraft as

 = K − 123 + ℎ5G 

where EQ is a vector rigidly attached to the Earth and ℎ is the altitude of the rocket above sea level as
shown in Figure 1. If we then take the time derivate of this equation as is required for Eq. (2) we get

 G 
= −ℎFG  − 123 + ℎ5
 
(8)

where , because of Poisson’s formulas, the time derivative of the unit vector perpendicular to the local
horizon is given by

G  F F
= E × G  = − cos1Y⁄r% 5B − D .
 23 23
(9)

As a final step, we combine Eqs. (2), (7), (8), and (9) to get the following kinematic relationships:
23 cos < cos ;
F = J
23 + ℎ cos1⁄23 5
(10)

23
F = J cos < cos ;
23 + ℎ

ℎF = J sin <

2.5 Dynamic Relationships

We will now derive the scalar form of the vectorial equation (1) using a process similar to that of
the previous section, in which we determined the components of each vector on the body axes.

We first introduce the thrust angle of attack M, which determines the angle of rotation in the 
and  plane to which the body axes system must be subjected in order to turn the  -axis in a direction
parallel to the thrust. Consequently, the thrust becomes

= NOcos M B − sin M G  P. (11)

We then write the aerodynamic force in terms of its components on the body axes as

 = −1!B + QD + LG  5
(12)

the same direction as the  -axis and can write it as

where D is the drag, Q the side force, and L the lift. We then note that the acceleration of gravity has

(13)
  = SO− sin < B + cos < G  P

We then take the time derivative of Eq. (7) to find the acceleration of the aircraft relative to the Earth
and get

 B
(14)
= JF B + J = JF B + J1 × B 5 = JF B + Vr D − Vq G 
  

after we use Poisson’s formulas to find the time rate of change of the unit vector tangent to the flight
path in conjunction with Eq. (6) and substitute the result into the equation.

We must now examine the Coriolis acceleration term, which can be written as follows

UV = 2 × = 2J1r  D − q  G  5 (15)

where p  , r  , and q  denote the components of the angular velocity of the Earth on the body axes.

We then apply the law of variation of the acceleration of gravity with the altitude to get

23 Z
S = S3 X Y .
23 + ℎ
(16)

If we then combine Eqs. (1) and (11)-(16), the following scalar equations are derived:

23 Z
N cos M − ! − S3 X Y sin < − JF = 0
23 + ℎ
(17)

23 Z
N sin M + [ − S3 X Y cos < − J1\ + 2\  5 = 0
23 + ℎ

−] − J12 + 22  5 = 0

2.6 Governing Equations

We have derived general equations for the motion of the rocket over the Earth, but we wish to
examine the case where the rocket travels solely in a great-circle plane. The great circle we will take

mathematically by setting  = 0 and ; = 0 at all time instants. Under this condition, the components
under consideration is the fundamental parallel, the particular motion of which is expressed

of the evolutory velocity on the body axes simplify to

F
H = 0, \ = <F − , 2 = 0.
23
(18)

As a final result, the kinematic and dynamic relationships describing the motion of an aircraft in
a great-circle plane are given by the reduced forms of Eqs. (10) and (17) with Eq. (18) to get
23
F = J cos <
23 + ℎ
 ℎF = J sin <

23 Z
N cos M − ! − S3 X Y sin < = JF
23 + ℎ

23 Z J cos <
N sin M + [ − S3 X Y cos < = JO<F − + 2\  P
23 + ℎ 23 + ℎ

, ℎ, J, and <, which we have to solve for. The values for 23 and \  can be found in tables, leaving the
In this system of equations there is one independent variable, the time, and four dependent variables,

user to specify values for N, !, [, , and M, as well as the initial values for the four dependent variables.
Figure 2 shows the directions of the relevant parameters in the great-circle plane in which the rocket is
traveling.

Figure 2. Shows the great-circle plane in which the rocket travels as a cross-section of the Earth.
The instantaneous direction the rocket is traveling in is labeled as (t). All other angles and forces are
labeled as described in the previous sections.

3 Results
We will study the trajectory of the rocket for various values of initial velocity pitch angle (<3 ), thrust
angle of attack (M), drag (!5, lift ([5, mass (), initial velocity (J3 ),and thrust (N5. We will use a value of
6367442.5m for 23 and a value of 7.2921159 × 10f" 2gih for \  , both values of which can be easily
looked up.
 We will track the motion for 100s and then plot the parametric curve with horizontal distance (5

oriented, we will plot the negative of the vertical distance since the positive direction for ℎ is directed
on the x-axis and vertical distance (−ℎ) on the y-axis. Because of the way our coordinate system is

initial values:  = 1000jS, 3 = 0, ℎ3 = 0, M = 0, [ = 0, ! = 0, N = 2000k, <3 = li4, and

towards the Earth, and we wish to see the height away from the Earth. We will begin with the following

J3 = 1 ⁄h. We obtain the following graph.

4
x 10
6

3
-h

0
-40 -20 0 20 40 60 80 100
X

We can see that with the thrust always pointing in the direction of motion of the rocket and no drag
or lift to influence the path, the rocket will travel in the direction first oriented and then will loop back

the initial settings are very impractical for real world use. We will now modify only the value of M to see
on itself. It travels far higher than it does longitudinally and doesn’t continue in the direction desired so

set M = li100. This is essentially pointing the thrust away from the Earth by a very small amount, as
the effects of having the thrust point in a different direction than the orientation of the rocket. We will

depicted in Figure 2. The following graph is obtained.

 4
x 10
6

3
-h

0
0 50 100 150 200 250 300 350
X

We can see that having the thrust point towards the Earth corrected the problem of having the
rocket loop back on its path for this time interval and also resulted in the rocket traveling farther in the

back on itself, though further increasing the value of M corrects this.

longitudinal direction. Increasing the time interval however shows that the rocket does indeed loop

We will now vary the amount of drag placed on the rocket. Leaving M = 0, we set ! = 500k and
obtain the following graph.
 4
x 10
6

3
-h

0
-100 -80 -60 -40 -20 0 20 40 60
X

We can see that increasing the drag merely decreases the speed of the rocket, as expected.
Compared to there being no drag, the rocket simply doesn’t go as far before reversing its course.

[ = 100k, we get the following graph.

Increasing the amount of drag amplifies this effect. If we keep this amount of drag and then set
 4
x 10
6

3
-h

0
0 50 100 150 200 250 300 350 400 450 500
X

the rocket to go farther. We next increase the initial velocity J3 to 100 ⁄h and obtain the following
We can see that adding even a small amount of lift greatly straightens out the trajectory and allows

graph.
 4
x 10
4.5

3.5

2.5

2
-h

1.5

0.5

-0.5
0 2000 4000 6000 8000 10000 12000
X

Increasing the initial velocity greatly increases the range of the rocket, increasing it by a factor of

equally large factors. The last parameter to test is the initial velocity pitch angle <3 , which we first
about 24 over having no initial velocity. Increasing the initial velocity even more increases the range by

increase to 89li180 in order to see the effects of velocity pitch that is nearly straight up vertically. The
result is intuitive; the rocket doesn’t travel as far longitudinally.
 4
x 10
4.5

3.5

2.5

2
-h

1.5

0.5

-0.5
0 100 200 300 400 500 600 700
X

Correspondingly, lower the velocity pitch angle from 45° causes the rocket to go a bit farther.

4 Conclusion
We have only studied the most general possibilities for the various parameters involved. We have
kept their values as constants when in fact they should vary highly based on the position and speed of
the rocket. More accurate models for drag and lift are in themselves very complicated problems and
beyond the scope of this project. We also used a very simple model for thrust and mass. In a typical

of the rocket would decrease. Also, the thrust angle of attack M is a parameter that can be controlled
rocket, thrust would only be available for so long as there is fuel, and as the fuel is consumed, the mass

during flight and is changed to direct the rocket towards a particular target. We have also not observed
any attempts to land the rocket. This is due to our particular combinations of parameters and mostly
because we use a constant thrust. In a real rocket once the thrust stopped a combination of drag and
gravity would cause the rocket to fall back to the Earth.

Despite these limitations however, we have verified the general behavior of the rocket in response
to major changes in parameter values. A more detailed verification however would need to include
more detailed models of lift, drag, thrust, mass, and thrust angle of attack if we wish to model a real
rocket’s trajectory.
5 References
Miele, A., Flight Mechanics: Theory of Flight Paths, Addison-Wesley, 1962; pp. 35-68.