I am indeed restarting from scratch on the plane for the fourth time, with entirely different design paradigms. This design has been so much simpler that I am ready to print parts already, after only about 6 hours of designing–rather than 60-ish from the previous design.
Type of plane: I am going to a “float-plane” design without landing gear that is intended to simply glide down to the ground (grass) as softly and gently as possible upon landing.
Simplicity: I am also aiming for the design to be as simple as I can possibly make it. The previous design became extremely unwieldy and tiresome to build and fix, with around a hundred screws. This will be (I believe) entirely screwless, instead using hot glue as an adhesive to melt the 3D-printed parts together. It has the disadvantage of disassembly not being an option after putting it together, but at this point, I just need something that can get into the air and stay there.
Control surfaces: No more hinges!! I am going full-on hot glue. The surface will be controllable entirely from bending, not from rotating. I will have a pushrod system hot glued to the servo horn.
Propeller location: Now behind and above the main fuselage, rather than in front. The blue circle indicates the area the propeller will spin through during flight.This should hopefully prevent damage to the propeller upon inevitable crashes! There are some interesting side effects of having a propeller behind, however. The main thing is that there must be a slight downward angle so that the propeller does not exert a torque forward and downward on the plane during flight. We want the propeller angle of attack to go roughly through the center of pressure of the plane, as shown by the dotted line in the image below:
Even if it’s not directly through the CP, the lever arm will be small enough to not exert much of a torque. Advantages of learning physics this fall is that I can finally start to understand some of the mechanics behind how planes are designed!
Plane body: Now a solid surface instead of a skeleton. This should improve aerodynamics. To avoid making it heavy, I am trying another novel technique: printing one layer, just like I did for the 3D-printed Christmas trees earlier this month. I was amazed by how sturdy and light they were. This means the wall thickness goes from about 3-4 millimeters with the old design down to around 0.3mm. I am hoping that this is not too flimsy, because just one layer of filament is extremely light and agile. If I can get the plane to be structurally sound, this will be a very light plane given the size of the propeller.
Tail: MUCH more simple. I’m going for a cruciform horizontal/vertical stabilizer design. This is also necessary since I have to make space for the propeller to spin, meaning the connector to the tail is very low.
It’s clear to me that currently the connector to the tail, and possibly the wings, are the two major weak points in the plane. I’m hoping they’re strong enough; if not, I may have to spend some weight penalty to reinforce them with either thicker walls or infill.
The other thing I am very proud of is the elevator/rudder design. I struggled with this idea in the previous plane, but could never get it to work. My goal is to make the rudder and elevator as independent of each other as possible, without cutting such a large notch in either of them that it affects the ability to function as an elevator/rudder.
I finally figured out how to cut notches so that I can turn the elevator/rudder pretty far over simultaneously:
Wings: ALSO an order of magnitude simpler, so I hope. I made the actual aerofoil all one piece instead of six pieces screwed together in inconvenient places. If possible, I’m also making the wingspan shorter, to accommodate only two wing segments on each side instead of 3.
I will be hot-gluing the pieces together instead of screws; hope it holds up. Since I am not Icharus and I am not flying near a heat lamp, I don’t think the hot glue will get hot enough to soften mid-air. The ailerons, as I mentioned earlier, are bendy flaps outside of the actual aerofoil, rather than being on hinges attached inside the aerofoil, as I did before.
Camera location: Right up front, right where the best pictures will be. 🙂 This also has the added benefit of being far from the electronics and battery, where EM interference would be stronger. This should allow for better FPV transmission from the plane. The air rushing past the camera will also helpfully cool it down–the WolfWhoop FPV camera I found on Amazon for $15 has very poor heat dissipation.
Electronics/Serial Wombat: Once I get the thing flying, I’m going to be using the SerialWombat 18AB, along with the SerialWombat servo breakout board. The man who designed SerialWombat, Mr. Broadwell, very graciously sent me some spare parts to try out, and I plan to report back on how effective the products are here in this blog.
Elytraflight's Epistles 