Building (And Crashing) my Homemade Surveillance Drone

Hatchet is my 3D printed RC Plane with an autonomous thermal camera turret. This is about its design and first flight attempts. Hatchet is EDF Powered, has deployable wings, and uses X-form control surfaces and thrust vectoring to steer. It is going to be used to film rocket launches from the air!

I hope you enjoy this overview of its design. Please let me know what changes you think I should make in order to make this aircraft launch and fly successfully!

0:00 Intro
1:06 Design Overview
7:25 Try Brilliant!
8:36 Materials and Testing
9:33 Foreign Object Damage
10:31 Repairs!
11:26 Launch Attempt 1
12:02 Launch Attempt 2
13:07 Launch Attempt 3
14:09 Improvements
16:21 Outro

I do not provide code, CAD or PCB files for my projects.

[TJMartinek]

I just imagine some guy in the CIA building taking out a cigarette every time one of these videos is posted.

[dark_matter]

As someone who also jumped into the RC hobby with 0 experience and decided to build an extremely complex model, this video is amazing. You just dialed everything to 11.

[jonasprusek]

Ah yes, my favourite non-military engineer, building non-autonomous, IR guided, non-loitering munitions / non-cruise missile / non-air-intercept drone / non-autonomous recon platform.
Jesus at this point you are just building next gen autonomous drone.
Also i belive that you have exact ideas what can be put into that free space.

[Chretze]

13:20 "I don't think I have yaw stability enough" my brother in Christ, I don't think you have enough thrust

[BPSspace]

9:51 "in terms of fan blades... we have no fan blades" 😭😭😭

[_hanz]

Watching this made me realize that I have to learn to get good at CAD, holy smokes the visuals and design part is so cool!!

[kylebedrich]

I’ve been designing and building competitive RC planes for a couple years in SAE aero and here is my brain dump:
1. Thrust needs to be increased, EDFs are not known for their static/low speed thrust, especially those small ones. (EDIT: Now thinking about it, adding a folding propeller in the front would help increase thrust and reduce wiring weight since your batteries are in the front also) Make sure you have an adequate thrust to weight ratio. Decreasing takeoff weight is crucial since your wing area is a design constraint. Increase powertrain battery voltage if you can, you’ll have to sacrifice capacity to save weight though.
2. Use an analysis tool like XFLR5 instead of ecalc, it uses actual airfoil data and VLM to solve for lift, drag, and moment along CG. See how much lift you’re making at takeoff speeds near stall angle. Check CG also, it might be unstable (especially on launch 3)
3. Increase aft control surface size, they look pretty small and won’t do much at low speeds. The EDF should help, but looking at the video it’s not enough. The planes longitudinal inertia seems really high especially with the heavy batteries up front and large fuselage.
4. I believe you are close to/a little unstable in yaw because you don’t recover very fast at 13:11. You have almost equal stabilizing and destabilizing side area on the fuselage and the control surfaces don’t seem to help with that since they’re moving to control pitch and roll. Increase your aft side area, this will help you recover from a large sideslip which seems to be the biggest issue on takeoff at lower speeds. At a 90 degree sideslip on the 3rd attempt the plane seemed like it had no intention of recovering from it. This could be done by adding fixed fins that are not your control surfaces. You will also dutch roll during these thrown takeoffs if you dont have yaw damping.
5. Add ailerons because you have little roll authority at lower speeds. Roll authority increases if your control surfaces are further outward- those in the back aren’t gonna do much if you have a large lateral inertia from the wings, but this all depends on how heavy the wings are. Your plane rolled a lot during these takeoffs and didn’t seem to want to recover. Dihedral won’t solve this problem. Include a flap too for higher lift during takeoff.
If you have any questions I’d be happy to answer in the reply comments or if you have a discord.

[zizouxyz_playz]

- Rocket boosters?

- Magnetic rail launcher?

- Improve thickness for wing (areas near/touching the fuselage)?

- Improve nose cone material?

Thats all i have for ideas

[ahmedkamel]

As someone who builds a lot of RC planes, I’d recommend adding ailerons, and possibly flaps, to improve lift and control at lower speeds during takeoffs and landings. As you mentioned, the plane seems too heavy relative to the available thrust—using lightweight filaments would help a lot. Also, it’s unlikely you’ll be able to manually stabilize the plane quickly enough during takeoffs and landings, so I highly recommend adding a flight controller for automatic stabilization. Finally, it looks like the elevator authority might not be sufficient with just the vectored thrust setup.

All that said, this plane looks absolutely stunning, and I can’t wait to see what you do with v2!

[realitynotfictionii]

People like this absolutely amaze me. There are so many talented humans out there.

[dallenlofgreen]

Correct me if I'm wrong, but Hatchet is meant to be initially launched via rocket with the EDF just keeping it in the air. As such, redesigning it to be hand-launched seems like a step in the wrong direction. I think a catapult is a better idea, as it can provide the higher launch speeds that are needed without having to deal with the complexities of rocket launches.

[nickname]

This is so sick! One of the coolest DIY projects I've seen

[gamingwithflight]

As someone who has been in the rc plane hobby for a while, I highly recommend you make the control fins larger, at any low speed those fins are basically a suggestion.

[anactualeggplant]

Oh good! I was worried we would never get to loitering munitions but here we are

[tenlittleindians]

Cool project. I started modeling airplanes in the 60's.

It's built way to heavy to fly as a hand launch machine given the thrust to weight ratio.

The lower fins are a terrible idea for a reusable vehicle that belly lands.

A takeoff and landing skid similar to model sailplanes would provide a place to mount a tow hook for rubber slingshot style launches.

Another idea would be to bend up a wheeled dolly out of coat hangers that would cradle the airplane during rolling ground launches and fall away after take off.

To get an idea what these cradles look like; look up control line speed racer videos.

I think the speed of these things will amaze you.

[Twangaming]

13:15 the way your friend is throwing it here is forcing the plane to pitch up. If he can hold it at COM and guide it into the air under the planes own power, you’ll be able to keep control. Alternatively maybe a rail like the V1 can help you eliminate variables here. Looking forward to seeing more!

[dark_matter]

Loved the video! I'm an Aerospace Engineer and I recently jumped head-first into a similar-ish project without any sort of RC experience. My system is purely meant for relaxing FPV flights, but I think I have a lot of useful insights as I encountered pretty much everything you did in this video. First of all, for RC planes, getting off the ground is the single biggest hurdle. Your mass right now is excessively high. Generating lift is proportional to velocity squared. So at low to no velocity, you reducing weight isn't just beneficial, it's a requirement. Then, I believe the choice of an EDF is a major mismatch with what you're trying to accomplish. I know, it looks really cool (trust me, I also started with an EDF), but their thrust at low speeds is really poor and in general, thrust is limited by their small diameter. As such, I would instead opt to swap the engine out for a propeller in a pusher configuration. This will bring you much more thrust due to the larger diameter prop you can use, and it will also easily double or tripple your flight time since propellers are far more efficient than EDFs. Rocket launches take some time to set up, so if you want to monitor the entire process, having some margin on your flight time will be much appreciated. Unfortunately, the addition of a propeller will make your vision of a rocket-launched plane very difficult. It's a conflict of requirements, and you might have to make concessions. It might be worth employing a proper System Engineering approach and noting down your principle requirements and order them in terms of their importantce. Noting key and killer requirements will make the system a lot more functional in the end. IMPORTANT NOTE: There is a high likelihood that a rocket-launched plane like this is very much illegal to build and operate. (Hobby) Rockets are legal because they generally only go up. As soon as you aim to transfer to horizontal flight and actively steer them, you are building a cruise missile! Finally, for your control surfaces, I would generally ditch the thrust-vaning control. Its addition to controllability on top of the main fins is likely negligible.

[d_savage]

Definitely reduce takeoff weight at least for early test flights. A rail launcher would be a game changer for getting you off the ground consistently with enough velocity to get enough lift out of the small wings. In addition, a rail launch would let you spin up your edf right before you launch. That way you are taking full advantage of your vectored thrust right as you leave the rail, when your control surfaces are going to really struggle giving you enough control authority.

[johnnylego]

I love these kinds of videos, Thankyou for the detailed explanation!

[Charlie-gfmv]

1. Take off weight is too large, well, at least for the wing area you have chosen which is pretty much fixed if they are going to fold out of the fuselage eventually.
2. CG and wing placement seem to be mismatched, dangle it on a string from the mean chord, quarter chord point as a sanity check.
3. Control surfaces are too small for realistically any control during takeoff, and the fact you do not have ailerons would mean that they would need increasing in size compared to a normal set. For roll control matching ref area isn't sufficient as the moments are what counts, that close to the fuselage the moment will be tiny.
4. I'm unsure about the thrust vectoring, there is so much vorticity coming from that ducted fan that redirecting that flow is probably doing nothing apart from obstructing it, which leads me to the final point..
5. I don't think there is any thrust being made tbh. Compare just holding the plane at full throttle to what the fan is expected to produce.

Really cool project, I hope you can make it fly!