Building Homebrew VTOL Rockets

No one can deny what SpaceX and Blue Origin are doing is a feat of technological wizardry. Building a rocket that takes off vertically, goes into space, and lands back on the pad is an astonishing technical achievement that is literally rocket science. However, both SpaceX and Blue Origin have a few things going for them. They have money, first of all. They’re building big rockets, so there’s a nice mass to thrust cube law efficiency bump. They’re using liquid fueled engines that can be throttled.

[Joe Barnard] isn’t working with the same constraints SpaceX and Blue Origin have. He’s still building a rocket that can take off and land vertically, but he’s doing it the hard way. He’s building VTOL model rockets. Most of the parts are 3D printed. And he’s using solid motors you can buy at a hobby shop. This is the hard way of doing things, and [Joe] is seeing some limited success with his designs.

While the rockets coming out of Barnard Propulsion Systems look like models of SpaceX’s test vehicles, there’s a lot more here than looks. [Joe] is using a thrust vectoring system — basically mounting the Estes motor in a gimbal attached to a pair of servos. This allows the rockets to fly straight up without fins or even the launch rod used to get the rocket up to speed in the first few millseconds of flight. This is active stabilization of a model rocket, with the inevitable comments of ITAR violations following soon afterward.

Taking off vertically is one thing, but [Joe] is also trying to land his rockets vertically. Each rocket he’s built has a second Estes motor used only for landing. During descent, the onboard microcontroller calculates the speed, altitude, and determines if it’s safe to attempt a vertical landing. If the second motor has sufficient impulse to make velocity and altitude equal zero at the same time, the landing legs deploy and the rocket hopefully makes a soft touchdown in the grass.

While [Joe] hasn’t quite managed to pull off a vertical takeoff and landing with black powder motors quite yet, he’s documenting and livestreaming all of his attempts. You can check out the latest one from a week ago below.

35 thoughts on “Building Homebrew VTOL Rockets

  1. He should try hybrid engines, they have the benefit of being able to just turn them off and on without all of the complications of liquid fuel. Basically it’s a solid fuel rocket with the fuel in a cylinder shape and a separate liquid oxidizer. shut off the oxygen and the rocket extinguishes.

    1. Yeah, he’s not live-streaming. I have no idea why they’re claiming as such, given he states that he’s not livestreaming – that he’s uploading the video hours later.

      Why the hell is he pretending it’s a livestream? Is this just attention-seeking? Pretending to be like SpaceX etc?

    2. Oh my god. He’s by himself and using a radio to talk to….himself. Complete with fake squelch tones.

      And then this: “I wont’ be able to talk for very long because I got a little backed up on editing the webcast” – what? You didn’t webcast! You UPLOADED A YOUTUBE VIDEO HOURS AFTER.

    3. Oh god, the facepalming continues. Does he really not understand why his stabilizing things don’t work? They’re less effective as rocket leans. Any kid who has stuck their hand out the window of a car going down the road knows this.

    1. As far as I know (not a lawyer, and not in the US) it is not illegal to make it, you just can’t export the device or techincal details about it outside of the US without an export permit…

      I might be difficult to insure the launch. When a model rocket starts to fly sideways it won’t go very far. One that has active flight control might be able to reach quite a distance if it keeps trying to balance, and just failing.

    2. This is the US, right? Can’t comment for their laws, but in the UK, I’m pretty sure UKRA would be having kittens over this…

      I guess the *firing* of the 2nd motor would be viewed similarly to active recovery charges, or indeed, even equivalent to a 2nd stage (both ok). It’s the thrust vectoring/gimbal arrangement that is questionable.

      I’ve only ever seen one gimballed system fly before, and it was a university team with nightmare insurance. Active guidance is really frowned upon, though admittedly, I’m not sure on legality. Doesn’t sound good…

      1. UKRA would not be having kittens, because under the experimental section of the rules, gyro stabilised, gimballed rockets and fin guided rockets have been launched many times at UKRA events in the past. In fact, the Chair of the UKRA Safety and Technical Committee who preceded me, had one he used to fly regularly at events. It too, had retro rockets for landing, arranged around a centrally gimballed solid rocket motor for the ascent. Why would there be issues with legality?

        I don’t know why a University team would have had nightmare insurance, unless they did not launch as UKRA members.

        Certainly active guidance was seen as scary by some people, but the only ones who frowned upon it were those who would have been better off being traffic wardens than rocketeers. I took the attitude that if the person building it was of sound mind, and could demonstrate a high level of competence and common sense, and had a long track record of safe launches, then they could launch on one of the away pads which were located at a safe distance from the main pads.

  2. If landing is his goal, perhaps he should start with that part and just drop test vehicles from a multirotor from the appropriate height? Then he can solve the two issues of stable ascent and descent separately, without the first sabotaging the other.

  3. A good effort. I like the project. So a thumbs up there.

    But the efforts he goes to to make it seem like it’s a NASA launch is, well, weird, in a way. Talking to yourself over the radios? Activating fire suppression means blowing on the flame? I’m surprised he didn’t have a cardboard mock-up of the NASA launch control room set up out in the field.

  4. When I was a kid (1960s), I was really into model rockets. I also liked blowing them up (by stuffing the nose full of firecrackers or match heads or whatever that would be set off by the parachute ejection charge), so I mostly used cheapy throwaway rockets (cardboard fins, typing paper body, and weighted, rolled paper nosecones). And, I used the smaller, much cheaper engines.

    Well, on one of my rare launches using a real rocket (with a parachute) from a kit but with an engine that was way too small for the weight of the thing, I got an unplanned VTOL flight. The rocket rose maybe 15 feet, stalled, and dropped back to the ground–landing upright & stable on its fins! A few seconds later, the parachute popped out. And no, this is not just a fanciful memory–the friend who was with me still clearly remembers it too. We tried many times to replicate the event, but never succeeded, so I suspect that the engine used in the original may have been defective (they did occasionally explode, and I have an 8mm movie of one such incident).

    1. @Rodney McKay
      That’s one of those things that are infuriating to tell people about because either they don’t believe you or they don’t have enough experience to know how improbable it is (and thus how awesome).

      Reminds me of:
      IN A TIME BEFORE INTERNET:
      “I just thought of something obscure I’d like to know more about.”
      “That’s a damn shame.”

      1. :-) BTW, for some easily-replicated (but supremely dangerous!) awesome using one of those little engines, slit the paper shell with a razor blade and completely unwrap it, leaving just the fuel and the ceramic nozzle. Prop it at an angle against something (a curb works nicely), and ignite it (through the nozzle!!) using a fuse or a regular electrical igniter (with ???????????????? wires). You’ll get a glorious fireball shooting across the sky (so, don’t do this around dry leaves, people’s houses, etc. A big, empty playing field works nicely). I’ve used up countless engines this way, and have miraculously never done any (serious) damage to people or property (though I have bounced a few off roofs). YMMV.

  5. I’m delighted to see this kind of work being done; it’s a great experience. There are a few issues with his approach of “learn by experience,” though. A little time with Kerbal Space Program would have more quickly and cheaply taught him the lesson that off-axis thrust is fiddly at best.

    Looking through his website and debriefs of the various failures, more than once he ran into software problems caused by “poorly written libraries.” That’s my primary grief with Arduino–and sure enough, that’s what he’s using for the flight computer. If he sees this at all, my suggestion would be to ditch other people’s libraries and build your own. Don’t rely on the Arduino framework–it’s going to be too limited for anything but the most rudimentary rocket systems. Feel free to stick with AVR, if you like, though–it’s certainly up to the task! Timing-critical work requires much less overhead than what Arduino introduces.

    Out of curiosity, I wonder just how much variation there is in those hobby motors–I’d imagine it would be a devil to tune a PID if there’s enough difference between one motor and another.

    1. The real devil with the Arduino is that it has no scheduler other than some hard coded state machines or calling functions in the main loop based on timer values. I see plenty of YouTubers talking about needing to check a timer so they can perform “multitasking”. Even in the case of this guy, he mentions something about needing to change his code at apogee to prioritize the sensors. It’s really working too hard. Writing your own libraries will teach a lot more about the sensors and what they are capable of, and using an RTOS will make sure everything happens when you want it to happen. If you’re in love with the Arduino Wiring IDE, you could at least switch over to the Energia version and run the MSP432 as the CPU. It’s way faster, has an FPU, and their implementation runs below an RTOS allowing the user to just throw a pile of sketches at it and they all execute like they are the only code in the system. It’s quite clever.

  6. A 16+ minute YouTube video that in <2 minutes explains these kids have NO IDEA what they're doing. Anyway… Better luck next time guys. (Thanks HaD for the post, seeing FAIL is instructive.)

Leave a Reply to Rodney McKayCancel reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.