Rethnaraj Rambabu

UAV Siswa Challenge 2013/2014 Results Announced

noreply@blogger.com (Rethnaraj Rambabu) — Sun, 27 Apr 2014 15:00:00 +0000

The results of the UAV Siswa Challenge 2013/2014 have been announced in the DSA 2014 Exhibition which was held in PWTC, Kuala Lumpur on 16 April 2014. Our Universiti Putra Malaysia teams (Icarus and Langit Biru UPM) won Second and Third Place respectively whereas Universiti Teknologi Mara (UiTM) was announced as the Winner of the competition. Congratulations to the Namtor team from UiTM!

Even though we are not the Winner of the competition, we are glad that both our UPM teams managed to obtain top 3 ranks out of 8 teams from 5 different universities that took part. We are also happy that we managed to complete the given missions during the competition. The winner was able to perform the missions faster than us and managed to drop the sand-balloon over the required target.

Entire journey during the competition has taught us a lot new experiences in terms of team work, task distribution, problem-solving skills, decision-making, practical works, meeting people from aviation field, foreign and local government officials and etc. Besides that, we really had to learn to manage our time to make sure that our academic performance was not affected by the competition preparation. Thanks to the professors and lecturers who always adviced us and kept us motivated. Not to forget, Dassault Aviation and co-organizers, thanks for coming up with this great competition and for giving us an opportunity to explore the Unmanned Aerial Vehicle research and development. UAV Siswa Challenge certainly was an entry-point for us into the UAV field.

Not to lose the momentum, with the support from the lecturers in our Aerospace Engineering Department of Universiti Putra Malaysia, we have started more multi-rotor based research and development projects for undergraduate students. Our Aerospace Student Society (AEROS) is currently having its own special wing called Aeros UAV Research Group and consists of 60 active members of all academic years. Being the head of the group, I'm really happy with the interest shown and commitment given by fellow friends and juniors. These activities certainly would keep us all motivated and continue having passion towards Aerospace Field throughout our academic years in UPM and beyond!

Setting up "Throttle Mid Position" for Good Altitude Hold and Loiter Performance - Arducopter

noreply@blogger.com (Rethnaraj Rambabu) — Sat, 29 Mar 2014 04:33:00 +0000

Based on my experience of nearly one year involving in the development of hexacopter platform powered by Arducopter firmware, I have discovered that if the hexacopter is able to perform well in altitude hold mode followed by a good loiter performance, it wouldn't have much problems to be flown autonomously. If you have problem on that, probably one of the tricks to solve it would be try setting up the "Throttle Mid Position".

Initially, I was using six 2212-13 1000kV brushless motors with 10x4.5 props which gave me a decent thrust of 0.6kg each. During its first flight itself, the platform was able to hover nicely at 50% throttle. I only had to tune the altitude hold P value for better performance. After several crashes during following flight tests, eventually the motor shafts were severely bent.

Then, I switched to Tiger Motor MT 2216-11 900kv brushless motors. That time, my air frame weight was a bit higher (nearly 2.6kg) with additional payload attachments to the air frame such as video transmitter, payload drop mechanism, landing gear, camera and some optional components. During its first flight it hovered at 70% throttle. When I switched to altitude hold mode instantly, it would just shoot up and when I position throttle at middle, it descends.

I tried re-calibrating accelerometer, compass and radio. No improvements. Puzzled by this behaviour, I went through the Arducopter manual again to check if I missed anything. Finally I found the guide on 'Throttle Mid Position' setting and found out that hexacopter was 'underpowered'.

Basically for setting up this you have to:

1) Get your quad/hexa into a stable hover in Stabilize mode for around 30s.

2) Disconnect Li-Po and connect the APM/Crius board to the Mission Planner

3) Go to "Terminal" tab and download the latest dataflash log file. (Refer to the next image which was obtained from ardupilot.com)

4) Next, click on "Log Browse" button and load the recent log file that you've downloaded/

5) Now, search for "CTUN" messages in the data log > Search for "ThrOut" column and click "Graph This Data Left".

6) You can find out an approximation of throttle out value from the graph. In the following image which is taken from Arducopter guide, the value is roughly around 550.

7) Knowing that, now go to Config/Tuning tab > Standard Parameters > adjust the value for Throttle Mid Position value approximated above.

8) Click on "Write Parameters" button. Now, try flying your platform and check how it performs!

Personally, I adjusted the value to 700(which is actually the maximum allowed) for my hexacopter platform. After the mid throttle position has been adjusted, I was able to notice better altitude hold and loiter performance. Basically you need to take off, get into stabilize mode at your normal hover stick position (myself at 70%), switch to Altitude Hold mode and quickly position the throttle at 50% position. It will start holding its altitude nicely. Try moving around. If the copter descends while it moves, try increasing the Alt_Hold_P value till it is really able to hold at the required altitude.

Under Altitude Hold mode, if you raise or reduce the throttle stick position within the dead zone (40% ~ 60%), there wont be much effect on the hexacopter. But, if you move beyond 60%, it will raise up slowly and if moved below 40%, it'll descend slowly.

And becareful when you switch back to stabilize mode. Make sure you raise/reduce the throttle stick position to normal again accordingly. If your platform is underpowered, it'll descend at 50% stick position in Stabilize mode and if overpowered, will ascend at 50% position.

The following graph shows relative altitude variation vs time for different flight modes during the flight test.

Throttle Out value during altitude hold and loiter mode now is

UAV Siswa Challenge 2013/2014

noreply@blogger.com (Rethnaraj Rambabu) — Thu, 27 Mar 2014 16:25:00 +0000

It has been nearly one year since the UAV Siswa Challenge started. I represented Universiti Putra Malaysia under team Icarus along with three other friends of mine from Deparment of Aerospace Engineering. There were two teams (Icarus and Langit Biru UPM) selected from Universiti Putra Malaysia to be among the 10 finalists for the competition. Initially, we had to submit a proposal, followed by a closed presentation and interview session by juries from Dassault Aviation France, CTRM and MIGHT before teams of 10 finalists were chosen and awarded with complete multi-rotor development kit.

We were given nearly 8 months for building, testing and preparing hexacopter platform to conduct required missions during two rounds of final competition which was held in Universiti Putra Malaysia on 22nd and 23rd March 2014. We had learnt a lot of things throughout our journey of 8 months preparing our hexacopter platform. Beyond the technical knowledge we gained to upload the firmware, configurations to be made, performance tuning, we did also acquired a better ability of working as a team, project management, problem solving and decision making skills. Personally, myself as a person who was more comfortable working alone, came to understand the advantages of being as a team and distributing the task among the team members. Things just get easier, better and simpler solutions can be generated from a brainstorming session and collaboration between team members.

Earlier stages of the progress after receiving the kit was really slow because we had to test and understand the individual components and systems that we were provided with. Technical problems would just pop-up at anytime, where we had to crack our head thinking and searching for solutions on web. And if we don't find any solutions, our best friend, "Trial and Error" method was always there for us.

Once everything was set up, it was time to start flying. We had only one airframe with costly electronic components on-board with some spare propellers. We had zero experience in flying remote control aircrafts/ multirotors. A lift-off from the ground at that time was a great challenge for us. Afraid of losing airframe, we waited for guidance from professional flyers. Luckily there was someone who helped us with that. A local RC hobbyist who had nearly 8 years experience of flying multi-rotors lend his hand to us. After listening to some encouraging words from him, we gained confidence to take-off the hexacopter and had a decent hover flight. Well, that ended up with some bumpy landing. But, HAHA! I did land it without breaking the propellers or damaging the airframe! Continuous support from the local RC hobbyist really kept us moving when we faced big problems. Thanks to him.

Continuous flight practice built more confidence inside us. We started moving on testing different flight modes and started tuning hexacopter for good performance. Few days before technical presentation, the luck was not on our side. We had a disastrous crash. We were testing an auto mode flight. Very low experimental P value for stabilizing parameter was the reason for that. Yet, we managed to buy a new body kit and prepared the hexacopter overnight and resumed the flight test next morning. We also managed to do well during the technical presentation which was held in MIGHT office in Cyberjaya.

Our working hour shot up as the competition day was nearing. We were testing and debugging our custom made image processing algorithm and additional software functionalities to be used during the competition rounds. More and more flight test was carried out in the outdoor area. Flight performance was tuned. We tested our payload drop mechanisms.

Finally, the competition day, 22nd March 2014, arrived. We were performing well on the first day of competition. During the training session late evening on that day preparing for our challenge on the next day, we had to face another disastrous crash of our hexacopter platform when we tested our custom code to send automated commands from Ground Station via telemetry. The propellers stopped spinning on-air and platform fell like a rock from a height of 7 meters damaging our payload drop mechanism, our only landing gear and hexacopter arm. At that time, we had three problems to solve. First, we had to repair our hexacopter platform. Second, we had to recreate our payload drop mechanism. Third, we had to debug the code for sending command from Ground Stations. Fourth, we needed to test the code and make it really work or we would be disqualified from second round. Everything had to be done within an overnight.

Nearing a situation of losing hope for the second round, we tried to calm ourselves for some hours. With some moral support from our team supervisors and the local RC hobbyist, we managed to motivate ourselves. Clock started ticking. 8 more hours left. An intensive teamwork aided by quick and critical thinking of team members, we managed to revive our airframe, re-built the payload drop mechanism and debugged the code. The final flight test to verify the performance of code was done just one hour before the real competition begun. Thanking the god. It worked like charm. We managed to complete the mission for that day. Though, we had to fly a bit slower due to uncertainty in the cause of the great crash we had. Our second trial during second round was a bit shocking though, after we tuned the parameters for a faster performance, once again, the hexacopter fell from a height of 5 meters. Luckily, there was no any severe damages to the airframe.

The final results of the competition will be announced in mid-April. The entire journey and experiences gained during the commencement of UAV Siswa Challenge 2013/2014 will remain as an unforgettable moments of my undergraduate life. Here I would like to thank Dassault Aviation, Malaysian Ministry of Higher Education, CTRM, MIGHT, for coming up with an event of first of its kind in Malaysia for university students.

I would like to also thank Universiti Putra Malaysia, our team supervisor Assoc. Prof. Lt. Col. (R) Mohamed Tarmizi Ahmad, Dr. Omar Kassim Ariff, RC hobbyist Mr.Arif and my team members for their unlimited support throughout the competition.

Now, anxiously waiting for the results to be announced....

Note: More pictures are on the way!

Modified the Yak-54 Foam Model RC Plane

noreply@blogger.com (Rethnaraj Rambabu) — Sat, 01 Dec 2012 03:55:00 +0000

After the bad crash which has caused nose to break apart from the fuselage, I get to know about a few mistakes/disadvantages in the current design that I did as a beginner. Those are:- less distance between propeller tip to ground and weak motor mount to foam. As I'm still lacking of good landing technique, whenever it lands, the prop would hit the ground and results the second problem where motor mount comes out of foam.

Decreasing the prop size might not be a good solution. So, I decided to shift entire nose part up making the model to become a low wing plane and to make stronger motor mount. It's just a 'trial and error' experiment. So, I started doing a motor mount made of wood with 4 short CF tubes (which I had saved some extra pieces when I was making gliders previously).

Later, I glued it to 2 foams which has been slotted earlier for fixing the CF tubes. I glued some balsa wood at the foam sides to make a sandwich-like structure.

Finally, I glued it onto the model and gave life to a new low-wing RC model :D. I still haven't flight tested it. I'm waiting to fly it with one of my Professors who said that will bring his RC model too.

Yak-54 35" Depron Foam Model Build Log- Part 4

noreply@blogger.com (Rethnaraj Rambabu) — Thu, 29 Nov 2012 16:08:00 +0000

Continuing the build log of Yak-54 model. As per said in the post before, after a hard landing which has caused the motor mount and landing gear mount to break, I've further modified the model to have less frontal area, strengthened the landing gear attachment with foam, redid the motor mount.

Reducing the frontal area

Placing the landing gear between balsa wood and over that is ice cream pop stick on both sides.

I have tested it on the next day and it flew nearly for 45 seconds smoothly and then I lost control over the plane when it did a steep left roll followed by a crash landing which broke the fuse nose mounting :D There are two reasons for this. First, I am still a beginner rc pilot and need to master flight maneuvering skills. Second, my transmitter control sticks are a bit sensitive. Here are the picture of the model with broken nose.

I have further modified it. I've shifted the nose entirely up making the model to be a low-wing model. See it in my next post :D

Yak-54 35" Depron Foam Model Build Log- Part 3

noreply@blogger.com (Rethnaraj Rambabu) — Thu, 08 Nov 2012 17:25:00 +0000

Two days ago, I started creating a mount for motor. Glued 2 wooden sticks to wooden motor mount as shown in the next image. These sticks will be glued to the foam which is cut at half-depth. This will make motor mount structure much strong.

Next, the bottom compartment of airplane was inserted/glued with some small foam sheets to strengthen the wall structure.

Small slots were cut in the foam to install servos and control horns were added to the control surfaces such as both ailerons, rudder and elevator. These control horns and servos shall be connected using small steel rod. This will allow the servo arms to push/pull control horns attached to the surfaces.

We may use dual servos or single servo setup for ailerons. From what I read on internet, single micro-servos will have to work harder to push/pull two ailerons at same time. So, some prefer to use FLAPERON channel (CH6) for second servo and program it accordingly in transmitter so that both ailerons will move in opposite manner.

Later, I added landing gears, covered bottom surface with foam sheets and made the top cover to be open-able/ removable for the ease of battery changing. There are two kinds of landing gear types used. First is taildragger type and the second is tri-cycle alike landing gear. I have used taildragger style landing gear set, so that it will resemble the look of Yak-54 model. Besides that, there is no specific reason for me choosing it. But, many RC hobbyists recommend tri-cycle type (will write more about it next time).

And that's it. It's ready to fly.

On 7th November 2012 (Wednesday), I did a taxi test in a fairly smooth playing court and later in our university stadium running track. As a beginner in remote controlled aircraft work, I experienced some difficulties(fear) during its flight and this resulted in sudden dive followed by a crash. Luckily it was just a small take off.

After the test, I've modified it further to have better landing gear supporting structure and reduced frontal area(near motor&propeller) to increase its propulsion efficiency. And I made it to fly! Will write further about it in the next post.

Yak-54 35" Depron Foam Model Build Log- Part 2

noreply@blogger.com (Rethnaraj Rambabu) — Fri, 02 Nov 2012 13:14:00 +0000

Today, I've started working on the airframe where it is mainly built out of compressed white foam and foam with black colouring at one of its sides. We're building it in our hostel (residential college), College 10, UPM. As per stated in the first part of this build log, I've downloaded the plan/template for this model, printed it out, cut the templates(wings, stabilizers, fuselage) and pasted onto the foam. Since, I've planned to make a different fuselage structure, I did some extra work beyond those plans over that template.

Fuselage has internal width of 2 inches and gets smaller in size as it approaches the stabilizers. Thickness of the foam used is 6mm. So, here is the picture of the non-assembled electronic items, transmitter unit, and Phoenix RC Simulator which I bought from RCSmart. Captured with 2MP Nokia classic phone.

Those are the few simple tools which I used to glue, cut, stick, measure and solder. The following images show fuselage template that has been glued onto foam board and two foam parts that been cut out (left and right wall for fuselage).

Later, a foam sheet of 2" width cut out of the foam for building a compartment in between two walls of fuselage to place the electronic components nicely.

Similar step was carried out for wings, elevator and rudder.

Later, a carbon fiber tube of 4mm diameter and of 25 inches length embedded into the foam wing as shown below. It functions as wing spar to strengthen the foam wing and prevent it from fluttering much during its flight.

Then, me and one of my friend who came to my room, started gluing the structures together using CA super glue. Here its widely known as 'Gam Gajah' (or Elephant Brand Glue). We need to be careful in choosing the CA glue. From my previous experience, cheap and low quality CA glue tend to "eat" the foam structure. So, go for good medium-range-costed ones. I prefer CA glue over Epoxy glue due to the easiness of usage, quicker time of drying & bonding, don't have to mix two different tubes (epoxy&resin) together which is an extra work. But, from what I know, even though the epoxy glue tend to dry slower, it bonds very strongly to the structure. It is very much recommended to use epoxy for the wood based models.

Now, is the favorite part of mine for today. Adding hinges to the control surfaces. Three days ago, I forgot to buy nylon hinges from the RC hobby shop. As I was searching for an alternative to attach the surfaces, suddenly I thought of using plastic covers which we normally use for assignment works. I cut 6 pieces of plastic with 2" length and 1" width; folded it half; pressed on both sides until I saw white lines on them.

Later, I sliced through/in the main wing (2 places in each side) and main elevator(1 place at both sides) parts. Same done over the movable aileron and elevator parts. Then, I slotted the hinge into both surfaces and finally added CA glue at the joints of hinges. The CA glue flowed into the foam and holding both sides strongly. The surfaces are moving up and down perfectly. And the CA glue has ran out before I was able to attach rudder onto the vertical stab. Nearly, 3 small CA glue were been used.

And that is the final product of the day. Will be continuing to do this model on coming Monday. Need to get landing gear set and propeller adapter/nose cone for fixing prop onto the motor shaft.

Yak-54 35" Depron Foam Model Build Log- Part 1

noreply@blogger.com (Rethnaraj Rambabu) — Fri, 02 Nov 2012 02:01:00 +0000

The depron foam model of Yak-54 has a wingspan of 35". It is a aerobatic 3D plane and able to hover, hang vertically in flight (due to thrust produced by electric motor mounted on nose), do torque rolls and tail touches.

You can download pdf plans of this plane from link 1 or link 2. I'm going with the 35" plan in first link. We should use lighter electronic components for this kind of model which you can buy from the nearest RC hobby shops. For us in Malaysia, there are various RC hobby shops in Klang Valley but, I prefer RCSmart which is located in Amcorp Mall, Petaling Jaya because it's the nearest to my university and easiest to reach (Taman Jaya LRT station in front Amcorp Mall). We are using the following electronic components bought from RCSmart.

1) Brushless motor - Tower Pro 2408-21 black
2) Electronic Speed Control - Emax 20A ESC for brushless motors
3) 9 gram servos (x4)
4) 11.1V 3S 1300mah 25C Li-Po battery
5) Fly Sky 8-channel Receiver unit
6) FlySky TH9X Transmitter unit

The following image shows basic electronic circuit for remote controlled aircraft.

However, we're going to modify the flat single-foam fuselage shown in the plan into thicker fuselage with compartment inside it for the ease of electronic parts placement. Hoping to build something like shown in the image below.

Started building remote controlled Yak-54 Foam aircraft

noreply@blogger.com (Rethnaraj Rambabu) — Thu, 01 Nov 2012 16:58:00 +0000

Today, me and 4 friends of mine, undergraduate students of Aerospace Engineering in Universiti Putra Malaysia, have started building a small scaled Yak 54 foam model remote-controlled aircraft. A lot more friends are to join. We've started off with soldering the adapters to the cables of electronic components. I used soldering iron 6 years ago when I was studying Form 3 in SMK Telok Panglima Garang. Started off with some bad soldering, later somehow I managed to get the knowledge of soldering back. The friend of mine, Vijayan who is currently a third-year student, was better than me at soldering and helped me throughout the two hours of work today.

Making of remote controlled aircraft or known as RC aircraft/airplane, is being a hobby among people around the world. From young children to older people are involving themselves in this great hobby. A lot of videos, forums, websites and resources are on the internet to teach us how to do a RC aircraft from scratch by our own.

We're having the same passion too. We're going to do it not only as hobby to fill our free-time, but also we shall use this as a medium to understand the theories of flight that are taught to us and also those published on the books.

The main electronic components used for a basic RC aircraft are :- brushless motor, electronic speed controller (ESC), Li-Po battery, receiver module and servos. We're intending to build a classical, single motor with puller configuration, fixed wing aircraft out of compressed foam/ Depron foam.

Tomorrow, I will be posting the images and the specifications of the components and materials that we are using to build it. And I shall be sharing my experiences, project works, aeronautics knowledge and constantly update this blog with the projects' build logs and my future plans. Kindly follow up with me!