18. Final Project

Final Project Notes

The name of the game is Integration.

After the Final Presentation Session, I have been working to improve the functions and capabilities for my airship prototype. A few photos of the final presentation make-of can be seen below, first, aligning and balancing the airship and second, letting it float, which worked:

airship making

airship flying

After the first presentation, it was clear that some improvements needed to be made to the model in order to render it functional. These were:

Propelling: a single DC Motor and a propeller would not be enough to move the canopy, for the second version I decided to add support for a second motor and another propeller

Control: A single MCU would not be enough to deal with the DCs and the servo on its own, it needed support if it was to power both. This was addressed by making the power distribution board described in the Project Development section and connecting the motors to different MCUs, this meaning the next version would use both an arduino and a fabduino.

Mechanical Design: In order to accomodate for two DC motors and two propellers, the articulator piece made and talked about in the Mechanical Design section had to be changed, made even wider, I returned to Inventor to modify the previous design

new articulator

Since this new piece would be significantly heavier, the miniservo I used in the previous installment would not do, I chose a bigger, heavier Servo Motor, this servo had a bigger axis diameter, I measured it with a calliper and added for ther laser kerf.
new articulator fitting diameter

However, once cut, the piece fitting was too loose to work, so I took te exported design and edited it in Rhino (no need to return to Inventor to edit) just change the fitting part and cut it again.

new articulator's fittings

This time the result was as expected, and the new servo piece fits the servo perfectly, both cut pieces can be seen below, the best fit being on top:

new articulators

The next step was the assembly of the steering/propelling array, in order to do that, the DC Motors had to be bonded to the articulator piece, these was done using double-sided adhesive tape in the articulator and covering the DC Motors with insulation tape to increase their adherence, since the double-sided tape did not work on the motors alone. No more fixation elements were needed, this helping the cause of keeping things as light as possible.

bonding the articulator to the DC motors

As it can be seen in the image below, the main reason behind expanding the articulator is the length of the propellers, as a shorter distance would have caused them to interfere with each other. As I was assembling the array, I found that the shaft's diameter of one of the motors was greater than expected, so I used the electric screwdriver with a screw as a drill, after several attempts drilling in and out of the propellers attachment hole, it was able to fit in the DC motor shaft.

drilling propellers

attaching propelllers and motors to the articulator

Then, after assembling this array with the servo, it was time to test the if the steering and propelling combined worked. The motors were connected to the power distribution board mentioned in the Project Development page.

distributor board

The servo was on fed on energry from a 5V voltage regulator connected to a battery, and controlled via an Arduino board, the control signals for the DCs would come from the Fabduino board. The integrated system can be seen below.

integrated power system

In order to avoid the propellers hitting the canopy box while steering, I reduced the rotation angle from 60° to 30° by changing the Arduino IDE servo controller code, in this case, the servo would go on a three-staged arc, 74° - 89° - 104° and back.

servo control code changed

This can be seen in the video below:

After getting the System to work, I realized the weight of the electronics and motors plus the new MDF-made canopy added up to 489 grams, this was too much to be lifted, I had to find another way to make a Canopy and buy more Helium. This leads us to version 2.0.

To acheive this, I started from the foam canopy of the first version and, removed two layers of foam and added plastic tubes to attach the baloons to them. After re-wiring and pówering the MCUs, we were ready to test the motion again.

After checking the weight, it was around 283 gr, better, but still too much for the baloons I had, after this I tried several alternatives to reduce the weight and balance the frame, removing layers of foam, taking away the arduino, changing the big, heavy Servo by another one, this meaning I was able to use only one DC instead of the two planned and ditching one of the 9V Batteries and using only one, and still no success, another approach had to be taken. Which brings me to version 3:

For the frame, I cut a voronoi-inspired design I found in Ricardo Torres Academy Page. I replaced the MCUs for an Arduino Mini, a lot ligher than the Arduino or Fabduino, I kept a very small servo, a 9V battery, a voltage regulator board, and the power distributor board, though this was later removed. With this I could keep the weight around 100g and get the set to balance with the balloons. but for some reason I could not troubleshoot the Arduino Mini communicating with the servo, I had propelling power though, but adding the DC motor broke the balance and by then, I was totally out of helium.

In the end it was about two critical factors: The Helium and the Batteries, the 9V standard batteries were to heavy and we had nothing lighter in the lab to replace them, and, after working with the three versions, it seems that a full canister of helium is capable of -through balloons- of lifting around 100-150 gr, So in order to get the system to work, two whole canisters are needed, also replacing the batteries with lighter equivalents (maybe Li-po batts) would show positive effects. After learning this, I came out with a final bill of materials and costs for version n° 2 of the airship, the costs of the mainly fabbed parts are not included since they were made in the lab and the foam block was esentially waste material.

Lessons Learned

Planning, well done, can save you a lot of time.
Spiral development can really get things going by making you focus when it counts.
Persistence and well-intended help can make the difference in your project.
The electronics can certainly be improved, and in the end, a whole fabduino is too much processing power for controlling DC motors.
Balloons can be trick to handle, in order to improve the reliability of the project, a way to make a DIY Zeppelin would be more suitable.

What comes when?

I have gone that in some detail in this matter in the Applications and Implications Page

Final Words

This course have been an invaluable experience and I am deeply thankful to my instructors, the lab assistants and my fellow students, It's been a great time and I loook forward to keep improving, creating and sharing.