AEE-400 Summer Practice III

Ömer Uğur ZAYIFOĞLU – ID: 2300945

Company: Roketsan A.Ş.

Dates:09 Agust-06 September 2021

Introduction ....................................................................................................... 2
Company ............................................................................................................ 3
Description ....................................................................................................... 3
Brief Histor and Main areas of Business........................................................... 3

Organziation Chart ............................................................................................ 3

Facilities ............................................................................................................ 3

Employee Profile ............................................................................................... 4

Summer Practice .............................................................................................. 5

Week-1 ................................................................................................................. 5
Week-2 ................................................................................................................ 6

Week-3 ................................................................................................................ 7

Week-4 ................................................................................................................. 8

Conclusion .......................................................................................................... 9

Appendix .......................................................................................................... 10

1
INTRODUCTION
I have conducted my second summer practice at ROKETSAN A.Ş. at Lalahan Ankara for 20
work days on 9 Agust – 6 September 2021. My internship was on the subject of Navigation
algorythims in the Tactic Missile Systems-cruise missiles department. During the summer
practice i got a chance to work on Matlab-Simulink, in particular various Simulink navigation
algorythims. Also in the end of my internship I build an another algorythm for restorance of
missing navigation data in sequance. In addition to these as ROKETSAN interns we
participated number of lectures and presentations about various topics which I will be
explaining them in detail in this report. In my opinion this summer practice was very
educational in terms of engineering approach I experienced in the company.

2
COMPANY
Description:
The name and location of the company I conducted my summer practice is ROKETSAN
located at (Lalahan, Ankara Türkiye). The following link is the website adress of the
company.

https://www.roketsan.com.tr/tr

Brief History and main areas of business:

Roketsan was founded in 1988 by the Turkish government with the purpose of improving and
designing rockets and missiles in Turkey. The initial founding reason of the company is to
contribute Stinger European Common Production Program. The company successfully
finished the given task of producing a composite solid propellant engine. In 1991, the
company got its first building in Elmadağ Ankara, which is still in use, and many other
facilities, manufacturing, and test zones. In the last 34 years company has produced many
missiles and missile systems such as Cirit, MAM-l, TEBER, which are short in range missiles
or drop missiles. The company also makes long-range missiles with turbo engines such as
SOM and ATMACA. There are also many other projects ongoing, So with all these, it is easy
to say that ROKETSAN is by far the most experienced rocket and missile company in
Turkey.

Organization Structure:
Organization Chart of the company is showed with Figure-1 in Appendix part.

Facilities:
The 2 main facilities of company is

• ROKETSAN Elmadağ Ankara

Also there are other small facilities like

• ROKETSAN Bilkent CYBERPARK Ankara

3
Employee profile:
The compony mostly consist of engineers in the areas of Mechanical Engineering, Electic
Electonic engineering, Aerospace engineering, Computer Engineering which most of them
having Bachleor’s Degrees and Masters in their areas. In addition to these, there are
employees with Physics and Chemistry professions.

4
Summer Practice Activities

First Week
On the first day of the internship, we take an introductory seminar for learning the
ROKETSAN’s history rules and working zones. After the seminar is finished, we are
distributed to our departments. As I mentioned in the Introduction, my department was
decided by Human resources as Tactic Missile Systems-cruise missiles. Also, I met my
provisor Nazmi KANKAYA, a senior engineer in cruise missiles. He is working on the
subject of Navigation algorithms of the cruise missiles that the company manufactures.

In the last hours of the first day, my provisor introduced the Simulink program to me. Also, I
have taken information about the navigation algorithms, how they work, and their use in
Roketsan.

In the following days, I kept using Simulink, and I realized that it has lots of useful features,
especially like this type of algorithmic scheme.

On the third day, we again attended a seminar in which the company introduced in a more
detailed way. In this seminar, we learned about the job opportunities, the process of
recruitment, the employee profile, etc.

The fourth and fifth days of the first week is reserved for Occupational health and safety
education prepared by ROKETSAN Academy. The lectures were online, and they were quite
detailed in every way.

During the first week, I also tried to finish my first homework given by my provisor. I tried to
build a basic model of a rocket using Simulink. The program calculates the trajectory of the
rocket using the contribution of basic forces such as engine thrust, gravity, drag, etc. The
program also considers the mass change due to propellent spent during flight. However, it is
not realistic to use such a program for detailed analysis since the angle of attack changes
“orientation” are not simulated well, and for a moving fin rocket, it is even more useless.
However, I learned many basics of the Simulink program in the very first week by these.

5
Second Week
In the second week, Before going into the detailed navigation algorithms, my provisor asked
me to model a 3-body problem using Simulink. The three-body problem considers three
celestial bodies that each one has the gravitational force applied by the other two bodies, and
it plots the trajectories of all three using initial positions and initial velocities of the bodies.
The program I successfully built models the Sun Moon Earth in the XY plane, and the output
of the program can be used to show that one year consists of 12 months by observing the
moon's trajectory.

After finishing relatively small and easy tasks, I finally started to learn general navigation
algorithms used for modeling and observing the trajectories of moving objects. These
algorithms use IMU data taken from the flight computer during flight or after the flight to
calculate navigational parameters such as position (location on earth "coordinates"), velocity,
acceleration, angular orientation, and angular accelerations at each time separately. These
pieces of information can be used by flight computers to correct other internal measurement
units' errors or tune them during flight. Also, these data can be used for test purposes to
observe how well-performing the navigation algorithm is used and how well the other flight
computer components perform. (like GPS ).

The book I used in the second week and following weeks is the Noureldin, A., 2013.
Fundamentals of inertial navigation, satellite-based positioning, and their integration.
Heidelberg: Springer. My provisor recommends this book there are variety of Navigation
algorithms actually and the ones I learned in the second week are Internal Navigation System
in ECEF, internal Navigation System in inertial frame, and in local-level frame.
In general, the days of the second week of my internship is mostly spent with understanding
the Navigation algorithm's idea and its general use in aircraft.
On the last day of the second week, we took one more seminar on Skill management,
Roketsan's applications for a master's degree, EQ and IQ, and future in business life, etc.

6
Third Week
The third week of my internship is mainly spent building navigational algorithms in Simulink.
I also get the chance to use real Missile test data; even though it includes 30 seconds of IMU
data, it is excellent to see my Algorithm works well with the given IMU data. After working
with easier navigational algorithms, the one We The details of my navigational Algorithm
can be explained as follows.

It takes the instant IMU data from Gyroscopes and accelerometers. For velocity calculations,
the Coriolis acceleration gravitational acceleration and frame transformed IMU acceleration
values are integrated over the timespan. As given in Figure-2

Using these total felt acceleration values, we can calculate transport rate, velocity, and
position using the blocks given in Figure-3. Transport rate can be explained as the rotational
rate of the frame over the Earth's surface. And their calculation formulas are given in Figure3.

Next, we will calculate the Earth rate, which is the Earth's rotation rate times the
instantaneous rotation matrix. Both Earth rate and Transport rate values will be contributing
to linear accelerations and angular accelerations.

For calculations of the attitudes, angular orientations, and angular velocities, we have lots of
options. To be honest, this part was the most challenging section of my internship due to lots
of building mistakes and algorithmic mistakes. I kept getting the wrong results for days.
However, since it worked in the end, I can confidently explain the idea of attitude kinematics.

In my Algorithm, there are three ways mentioned in the book to implement attitude
calculations. These are Euler's method, the Direct Cosine Matrix method, and the Quaternion
method. I used all three of them to calculate the attitude of the Navigation system.

You can see all of the methods' pictures from Figure-4 to 6, and you can see the attitude
blocks in Simulink in Figure-7. All of these blocks are used for calculating the attitude of the
system and instantaneous Rotation matrix since it is calculated by the current orientation.

In the end, what we will get is the Navigation algorithm given in the mentioned book as in
Figure-8.

Other than navigation algorithms in the third week, we attended a field trip on Roketsan's
Elmadağ and Lalahan facilities. On these trips, we get the chance to see real rockets and their
overall production processes. We learned very detailed applications of manufacturing,
assembling, working principles guidance systems, tuning of flight computers to flight
environment, etc.

7
Fourth Week
In the final week of my internship, I finished with the navigation algorithm and talked more
detailed about these types of algorithms with my provisor. Building these systems correctly
for a cruise missile is very important for a flight computer to reach correct navigational data
during flight and achieve its primary objective of hitting the target. In these systems, errors
tend to snowball at each step, so even one minor mistake could lead to enormously big errors
with time.

After this talk, I received my final task of building a Simulink program for restoring the
missing navigational telemetry data. The idea of restoring comes from the fact that when the
telemetry data we are receiving is missed by the receiver for one step only, there is an
analytical solution for that one missing line look at Figure-9. The building idea of this
program is to ask what position and orientation will I be in the missing step using the previous
acceleration gyro and orientation data. After that, I will ask what linear and angular
accelerations do I need to change my orientation and position so that I reach the next non-
missing data.

I created an algorithm that works both forward and backward as the Navigational algorithm
with this idea. In this case, I could only implement Euler's method for attitude calculations;
even though I tried DCM and quaternion methods too, they didn't work quite well, probably
due to my inadequacy. You can see some detailed explanations In Appendix Figure-10-12.
After looking at these figures, you will see that the required angular and linear accelerations
values are calculated exactly backward.

After finishing with this, I intentionally distorted the given data so that there won't be any two
sequent distortion. The distortions are all randomly selected to model real-life situations.

The tests of the finalized program were very successful. When I compared the actual data and
calculation data, there is maximum difference of 10−6 even though it is an analytical solution
when I asked about this to my provisor he said that it could happen.

In the end I built succesfully a program for restoring the data inside missing telemetry
package and my provisor was saying that it will be really used in Roketsan (“probably he
mentioned himself”). With these I was very happy and satisfied to contribute even a little to
Turkish defence Industry. I will remember this program as my first contribution to the Turkish
Defance Industry.

In the final day there was an event of farewell where we recieved gifts and certificates from
Roketsan Academy. Then we give our enterence cards back and my internship ended at 6 of
september

8
CONCULUSION
To conclude, in this summer practice, I specifically worked on Navigation algorithms and
Matlab Simulink. I first did some introductory projects. Then we continued with the
navigation algorithm with three different methods of attitude calculations. Then I continued
with the program for Restoring missing telemetry data, which will be used in Roketsan as my
piece of work. Also, by attending various seminars, I learned lots of information on the
recruitment process and how things work in Roketsan. In addition to these, I got a chance to
see missile manufacturing processes and learned lots of detailed information. To sum up, this
was a good experience for me, and I am delighted with this internship as a last-year student.

9
APPENDİX:

10
Figure 1 Organization Char of Roketsa
 Figure 2 blocks for acceleration and velocit calculations

Figure 3 Algorythm for caluclating velocity position and Transport Rate

12
Figure 5 DCM method for attitude calcualtions

13
e

Figure 6 quaternion method for attitude calcualtions

Figure 7 three different attitude calcualtion blocks that I built

14
Figure 8 The one missing column of telemetry data between two non missed parts

Figure 9The one missing column of telemetry data between two non missed parts

15
 Figure 10 Figure showing the sequance for calculating missing IMU acceleration value

Figure 11 Figure showing the sequance for calculating missing IMU gyro value with reverse euler’s method

16
Figure 12 Figure showing the general structure of backwards navigational algorythm

17