Of all moving machines, airplanes and rockets by far interest me the most. The idea of moving in 3 dimensions with such a large and heavy vehicle blows me away still. Prior to taking up this project I spent a fair bit of time learning the theory behind aeronautical design and its evolution, but I found the best way to get a deep understanding was to build one.

V1.0

To get a good understanding of professional aerospace design, I laser cut scaled blueprints of the J-3 Cub in balsa wood with parts adjusted for my own electronics mountings. Although it looked stunning and taught me a lot more about design any book could, it was a far cry from success. This first model was only able to sustain 2 flights with the main issue being strength; although the balsa was plenty light in order to stay afloat, it was incredibly brittle in compression, too brittle to land on its belly in a field of grass. With a snapped wing strut and fuselage I taped up the design for one last flight before its fatal crash.

The first issue was beyond strength was that it was actually TOO light; the plane had almost no inertia relative to the wind; this meant I had to add more weight and thus more power to maintain the desired flight characteristics. Second, the covering film I used did not yield an accurate enough airfoil, and also had a slightly higher drag coefficient yielding non-laminar flow which produces less effective lift and more drag.

So, the plan became to 3D print this airplane with a USA 35B airfoil which has much increased lift to help counter the effect of added weight at the cost of decreased aerobatic abilities

After design in Rhino3D it was time to get building again, but before I did so I also incorporated some CFD to check out how everyhting was looking and adjust acconrdingly. For the CFD I used ANSYS, although I am equally good with MATLAB and SolidWorks. This time I had also included mountings specific to my electronics so there was no movement of the internals to change the centre of gravity:

V2.0

After maidening V2 successfully, this was only the start. Now with all of the experience gained from building this I want something with more exciting aerobatic-like flying characteristics plus something a little unconventional. Although the added weight added some inertia, by making a larger scaled plane with a further back tail and longer nose I could have more inertia and more torque given the same control surface areas, and thus more maneuverability — up to an extent. V3.0 incorporates carbon fiber rods for stiffness and strength throughout, ailerons to allow banking, a more efficient internal structure utilizing the added capabilities of 3D printing, and more powerful electronics with venting to keep them cooler.

V3.0 (in progress)

Due to midterms the printing process was slightly delayed although more than 75% of the printing is finished and assembled as of now. So far the carbon fiber spars seem to add a lot of extra rigidity compared to my older models. Overall this design fits together much better thanks to more experience tolerancing, plus parts are stronger due to more experience testing infills.

I am considering including some generative design with Fusion 360 or ParaMatters (less fammiliar with the latter) to further hollow out some of the structure more effectively as opposed to the square cutouts and to iteratively modify it based on FEM results while also taking into factor the direction of the print layers as the adhesion between layers is a weak point in the tensile strength of 3D printed parts. Since I have the luxury of a 3D printer I can manufature much more complex, and thus higher performing geometries per unit mass which would otherwise be impossible with subtractive methods. I am also looking into adding a thrust vectoring compliant mechanism to the motor mount at the front in order to yield an even greater range of maneuverability. This intricate design works based on the simple rotation of two external servos, the only challenge is keeping the deforming material out of plastic deformation or from a reasonable amount of fatigue failure.

I do see a V4.0 coming within 4 months after, using the same carbon spars but potentially also a carbon body. My plan is to fully optmimize my current design with FEM but with the printing outsourced to a polycarbonate 3D printer, this way I can have thinner, lighter parts in my optimizations and I can utilize the benefits of less parts with a larger print volume. My goal is to reach a power density of 100 Watts / kg. In my design I can also remove the bulk, and thus drag, of the fuselage ‘canopy’ since it doesn’t carry any people, therefore I can make it more streamlined by even making part of the tail simply a carbon spar, removing more unneeded drag.

Further, in order to extend my coding skills and make this more than a purely mechanical design challenge, I’d really like to add a navigation system on this plane and a FPV camera. I’d like to use 3D bezier cuves (the same path defining method used in svg vector graphics) as points in real space since they are easily programmable, efficiently stored as data, and plot smooth parabolic paths the plane can easily follow along with certain maneuvers added in aswell; just like a pre-programmed airshow.

After V3 finished there was some issues with the steering abilities on the floor as the bending of the landing struts caused some twisting in the wheels and a toe-out, higher resistance and unstable stance in the front two wheels, making takeoff much harder. I am currently working with an Arduino Nano to try to program an autopilot method which uses a simple accelerometer/gyroscope sensor to cancel out any changes in pitch/roll/yaw detected while not oscillating back and forth spasmically. After that the microcontroller will be programmed to perform a figure 8 loop, for now its back to work.