Manned Multicopter Project Undaunted By Crash

We have to be impressed by [amazingdiyprojects] who completely totaled their manned multi-copter build, which has been spanning over eight videos. He explained the crash in video number eight and is right back at it, learning from the recent mistakes.

When you get right down to it, this is as dangerous as this seems. However, a giant multicopter is probably the easiest flying machine for a hobbyist to build. It’s an inefficient brute-force approach, but it sure beats trying to build a helicopter from scratch. This machine is a phenomenally un-aerodynamic chair on a frame that has a lot in common with the lunar rover; with engines on it. Simple.

There are a lot of approaches to this. One of the crazier ones is this contraption with a silly amount of electric motors. [amazingdiyprojects] went with eight gasoline engines. We’re really interested in his method for controlling the rpm of each engine and dealing with the non-linearity of the response from a IC engine throttle. Then feeding that all back into what is probably the exact same electronics from a regular diy drone.

Honestly, we’re surprised it worked, and we can’t wait for him to finish it so we can see him zooming around in his danger chair. Videos after the break.

Thanks [jeepman32] for the tip!

97 thoughts on “Manned Multicopter Project Undaunted By Crash

      1. Some ideas deemed silly at the time turned out to be viable true, however most do not. The early history of aviation is littered with machines that were poor ideas that never were developed. Those that were successful were because they were better. This design is no improvement over other VTOL aircraft and introduces several more problems. It is not a step forward.

        1. No improvement over VTOL?
          Well that’s a very uneducated point of view.

          Classic powered rotary wings aircrafts of today rely on very complex mechanics that imply ridiculous man maintenance hours ratio (all the way to 4 hours of maintenance for each flight hour for army helicopters).
          The main complication being the power gear box that has to vary through each rotation the pitch of two or more (i’ll spare you the one-bladed) blades spinning at high speed (just under sonic speed at their maximum) given their very long diameter.

          Well, an electric brushless motor (100.000hours MTBF for industry-grade ones) coupled with a fixed pitch propeller sure is an interesting proposition reliability and maintenance-wise.

          Oh and when you put 8-16 or even 32 of these on an aircraft and taking in account that it can fly on only a fraction of them, it suddenly isn’t such a “stupid machine”.

          P.S : Not even talking about blown wings, boundary layer ingestion, transitioning architectures and foldable props…

          1. Please try and read for comprehetion. I don’t mind criticism but please don’t misquote what I write. I clearly stated that THIS DESIGN s no improvement over other VTOL aircraft, and it is demonstrably not. I’m all for folks pushing the envelope, but this lashup is ludicrous on so many levels that it is pathetic and shows only that the individual has ignored basic engineering and aerodynamic principles without any potential gain. Indeed large passenger capable multi-rotor aircraft are going to have an important place in aviation in the near future, however this is not going to come from nitwits randomly bolting lawnmowers to lawn chairs

          2. It’s probably not going to be the future of aviation. But dude has his own flying chair! I don’t suppose you’ve got one. I haven’t. That’s the achievement, being able to fly around the place in a home-made helicopter.

        1. Yes, if only people weren’t people. That is, if only humans weren’t social animals, and the death of a person didn’t have an impact on the survivors who knew them. Of course, if that were the case, we’d probably have been killed off as a species long before developing powered flight.

        1. There’s a difference between stupid and a bit Heath Robinson, and deadly. If one engine / rotor fails, he can come straight back down. The engines and rotors are all independent, no reason for several to fail at once.

          It’s probably a bit dangerous, but also seems awesome and amazingly fun, I think the reward is worth the risk. And that’s life!

    1. I admire your spirit. Ignore the armchair naysayers who never build anything and only have the energy to punch some keys in a keyboard to complain.

      Keep on plugging and make this work, if only to prove to yourself that you did it. That said, I would not plan on flying, rather keep it purely remote controlled and away from people, expensive stuff and animals.
      Good luck. We’ll be watching this project closely. Love your energy and enthusiasm.

  1. One has to wonder whether it wouldn’t be more efficient to run a single motor-generator and eight electric motors instead.

    The power to weight ratio of an electric motor is far greater than for a small piston engine, and they’re far less likely to fail due to mechanical failure. Keeping eight independent combustion engines in tune over any number of running hours is basically an effort in futility.

    1. I agree, it’s the same reason trains that need to move so many wheels distributed in an awkward way use diesel-electric transmissions. Carrying power from a generator through relatively thin wires with non-existent maintenance demands is much more efficient than lugging around extra steel and all the complexity that comes with making it functional. It’ll far offset the small “loss” that comes from the gas->rotation->electricity->rotation conversion

      1. The thing is, the train engine doesn’t need a power to weight ratio larger than 1 because it doesn’t need to lift all that heavy metal off the ground. Getting something heavy rolling on relatively low friction tracks needs a lot less power than lifting it up in the air.

        In order to get the amount of energy to run those motors you would need a quite hefty generator – and a correspondingly hefty and mainly heavy engine. Add in the conversion losses (the combustion engine and generator are not super efficient) and you are simply better off running 8 combustion engines instead. Also you have just added two more things that can fail to the system – and a failure of that generator means a total power failure = immediate crash. The weight and reliability issues are pretty much the biggest reasons why there are no aircraft using something similar – not even fixed wing where the power requirements are much smaller.

        The only way to make that work would be to keep that generator on the ground and power the multicopter over a thick cable. But I guess that is not that interesting option …

        The engine synchronization issue is pretty much a moot point – this sort of contraption is pretty much a Darwin award in waiting – even if he somehow managed to get the engines running in sync (it is doable, there are multi-engine helicopters and planes, Moller’s Volantor had a similar system, etc.), there is still no autorotation ability, no gliding ability, if one motor quits for whatever reason, the whole thing likely just flips and crashes because of the engine layout he had, likely maiming or killing the pilot. I am not sure what is so attractive on these multicopter designs over something like an autogyro, for example.

        1. “power to weight ratio larger than 1 ”

          That’s not how power to weight ratio works. It’s not a dimensionless unit because they are two different units.

          You’re thinking about thrust to weight ratio.

        2. “In order to get the amount of energy to run those motors you would need a quite hefty generator”

          Not true. A fast generator is a light generator for the same reason why you can make a transformer smaller by increasing the frequency. Electric generators and motors also get more efficient at high speed. We’re not talking about 60 Hz AC generators here – there’s no need for that limitation.

          The power to weight ratio of the engine also increases when 8 small engines are made into one large engine because there’s less duplication of parts. In the small engines, the cylinder displacement is relatively smaller compared to the amount of surrounding metal and casing etc.

          1. Yup, it would be much better to run just one main generator powering electric motors. Although for redundancy you might want a second one, or probably better, a battery, so you’ve enough power to come down gently.

            Using 8 engines I suppose is just simpler. If he can afford the engines, just doing the same simple thing 8 times til he has enough thrust is nice and easy. Maybe v2 can use electric transmission.

            Might be good to have 2 engine / generator sets, each powering 4 motors in an alternating pattern, so you have enough to keep the thing stable as it comes down to earth. Sure it doesn’t have to have 8 props, but it’s a nice amount for redundancy and stability. Maybe electric motors would spin faster than the engines, so he could get away with less, or, better, fly faster.

    2. I (somewhat) agree. Keeping all those engines at the perfect speed for level flight is nearly impossible. The only way I can think of would be some kind of electric clutch on each engine, but that just adds more weight.
      But, each of his eight engines is 8.6KW, and weighs 2.5Kg. And the only extra weight you need is fuel in a plastic tank. So, lets say 1Kg of fuel per engine to give us some run time.
      The energy density of the fuel is ~11X than of a lithium ion battery.
      If we try to do that with electric motors, then we need something like a 100 to 120cc equivalent motor, something like a Turnigy CA120. This motor weighs 2.7Kg. We need eight of them. Then we need eight ESC’s that can handle these monster electric motors. Something like the a Turnigy dlux 250A HV ESC will do (but it won’t give you maximum power). These ESC’s weigh 450g each. And then you need one 16s battery pack per ESC that can handle the current draw of these motors (or one gigantic battery pack). And make the battery(ies) big enough to run for the length of time you would get from 1Kg of fuel.
      No contest.
      That would be one hell of a big motor-generator setup. You would need a 68,800KW (92HP) engine, a generator, and still need speed controllers for each motor. And that is neglecting electrical losses.

      1. “each of his eight engines is 8.6KW”

        You forget that the engine is not running at full output. Piston engines are rated by peak power which is only attainable at a specific RPM, electric motors by sustained power. The engines must be oversized to the actual need, whereas electric motors can even be undersized because they can be temporarily run above spec. The only limit is overheating.

        1. The electric motors I used in my example are 8.6KW peak (depending on winding option). The motors can only produce this peak while the battery voltage is at it’s maximum.
          “Piston engines are rated by peak power” – Then he needs variable pitch props on all the motors to operate within the peak power RPM band. The peak torque of an IC engine will occur at a different RPM.
          It’s still no contest. Liquid fuel can give more power at lower overall weight for long duration flights.

          1. Lots of engine-generator sets are run at either peak power, or peak efficiency, depending on application.
            The limitation is the engine control system, ignition timing, variable valve timing, and such. These small engines dont have fancy options like this, but I have run a 100cc 4 cycle engine at peak power (4300 RPM on that particular engine) for almost 3 months continuously with proper cooling and lubrication.

          2. You can put the peak power band almost anywhere you want on 4 stroke engines. It’s purely dependent on valve timing, ignition timing, and fuel / air mixture at the desired RPM.
            If you push it too high you won’t get to it until the engine is near self-destructing, and if you push it too low it will be very difficult to start and have a really rough idle.

          3. “You can put the peak power band almost anywhere you want on 4 stroke engines.”

            But since your cylinder displacement is fixed, and cylinder pressure/compression is limited by the fuel, the lower the RPM you set it the less power you get at the peak.

            That’s why small engines like these need to run fast, and why getting lots of power out of them means that the peak power is at very high RPM, and that in turn is the reason why the engine won’t last very long if you try to run it at the maximum power band.

            Hence, the engines are oversized to the application so they don’t need to run at the peak power band, and because the guy is throttling the engines to achieve control, they must be running at somewhere around half the peak power, which means a considerably smaller and lighter electric motor would do to replace them.

        2. For example, the sustained power output of a Turnigy CA120-70 is actually 14 kW. This motor is equivalent to a 20hp+ piston engine and would likely replace two of the original engines in this setup.

      2. Not an electric clutch, just a smaller 300-500W motor only to slow down or speed up the ICE just a bit, at 300g it just add 2.4 Kg. It needs a special controller were you can also regenerate power, store or relay the energy to another motor.

    3. Agreed the best option is a mini gas turbine driving an electric generator feeding a battery pack that is buffered with super capacitors into the electric motors. If the turbine dies you still have enough power to get down safely.

      Even with the configuration he has at the moment I suspect that if he added a few control motors that were electric he would overcome most if not all of the potential variability and control lag that is inherent in internal combustion engines.

  2. Wouldnt 2 IC engines with a shaft or belt system running through the frame – with each engine driving every other prop at as constant a speed as possible – be a little easier to manage? Vary the pitch of each prop independently for control. I would think that you’d have less chance of failure in bigger engines, better efficiency and at least a slightly slower ride to the ground if one failed? Also – surely suspending the bulk of the mass under the rotors would make for a more stable pendulum problem (evolo, ehang (I think) etc). Otherwise totally rad project. Definitely keep your hands, feed and any other appendages inside the ride at all times though :p

    1. Not really, a helicopter doesn’t behave like a pendulum except because it’s controlled by software which effectively simulates a pendulum as most Flight Controllers do. This fallacy is so common that it’s named and documented though: https://en.wikipedia.org/wiki/Pendulum_rocket_fallacy.
      Another place this fallacy is seen a lot is when people talk about the use of accelerometers in the Intertial Measurement Units. Thrust is perpendicular to the prop discs and acceleration doesn’t tell you much about gravity until you reach a speed where thrusts is completely compensated by drag (for example in free fall, contrary to what most would say).

      1. It absolutely behaves like a pendulum. And accelerometers can measure the force of gravity at all times, as long as the acceleration of the craft does not exceed the range of the acelerometer.

      2. The fallacy only applies to unguided rockets, since there are imperfections in construction which lead to the trust being applied ever so slightly off (or off quite a bit) of the center of gravity.

    2. I was thinking single IC power source with a separate variomatic drive to each rotor. Electronically controlled variomatic drives operated by a flight computer.

      Electric motors won’t work in this application because batteries, duh. Pathetic energy density compared to hydrocarbons. Maybe in some far future we’ll have personal fusion reactors, or, more realistically in the short term, fuel cells, but until that day you ain’t getting airborne without emissions.

  3. I think that props should be above the chair, not at the chair level. This would aid stability and prevent limb chopping incidents. Just last year a guy lost his head when his homemade helicopter prop started suddenly. Being at the wrong height did not exactly help.

  4. He needs to redesign this so the seat is under the motors. The current design is top heavy. Anytime he has any anomaly, he’s going to end up upside down. With the weight under the motors, an anomaly will only end up with loss of power.

    Really bad design.

    1. It’d be nice if we could edit these posts.

      In reference to my comment of “. . . end up with loss of power”, he won’t end up in a serious crash. He’ll end up just descending to a rough landing (right side up).

          1. No, vectors add, so rigidly fixed multiple thrusts sum to a single vector in between alterations of a control system.

            There is a mental short cut, in that an object in free fall (accelerating, no air drag) will not experience any forces due to gravity, if you add engines then the forces on the structure of passengers depend on the thrust independently of gravity. The ‘only’ contribution gravity makes is to where it goes and how fast relative to the earth.

            To put this another way, the only reason you appear to have the same weight in a hovering helicopter as sat on earth is because there is a control system trying to maintain distance to the earth.

          2. Correct, they add. So when you have a single thrust vector (the total of all thrust vectors), with a center of gravity directly below it, then the pendelum rocket fallacy applies.

          3. The sum vector must be closely along the line through the centre of mass or instead of taking off the craft would spin. Putting the human load below the plane of the props puts the C.O.M. below the point of thrust.

            I don’t agree with your criteria, but they are satisfied.

  5. I think his ‘improvement’ of putting an RPM display onboard for all 8 motors is pretty stupid. How are you meant to make any sense of the numbers while contolling the craft? Much better to show relative RPM offsets via some bar graph or something, more immediate. In fact this info is pretty useless in flight. Much better to log the full telemetry and make sense of it after the flight, on graphs.

    Overall, I think a better and much more controllable idea would be to have one internal combustion engine, maybe 8x the power of these individual ones, power all propellers from the engine via some clever drive mechanism, maybe belts, get the RPMs up to a known high climb rate, then use collective pitch props to control the craft.

    Collective control has always allowed much quicker attitude response than RPM attitude control, it’s mechanically more ‘difficult’, but would make this a technical certainty instead of a curiosity. Because it’s not a million miles away from a helicopter, mechanically.

    1. No human can control this machine. The response of the engines is not nearly as fast as an electric motor, and there is no way a human can monitor / control eight of these at once.
      Variable pitch props and flight controller would be the only way.

  6. Do you have any sensors to detect vibration in the negative way for the motors…it seems like you may have lost a good chance to measure a motor vibration on motor loss which could be used to give the pilot/automatic controls to shut down the octocopter thingamabob

  7. Geez…”Darwin Award” indeed.

    Not to mention insanely asinine ‘design’ (if it even merits such a term – one that implies actual engineering principles and common sense went into the build). Given the body of knowledge that already exists regarding airframe design, propulsion systems, aerodynamic theory, etc. This, uh – thing, is an utter abomination.

    It’s one thing to pioneer new concepts, but this is not by any stretch groundbreaking. It’s like throwing a seat on top of a lawnmower, and revving up the engine expecting it to lift off and fly.

    It would be amusing to find out if homeowners insurance covers any mishaps resulting from blatant stupidity.

    1. I think nobody said this is groundbraking, but it’s an interesting and fun aircraft design. If it end up flying (which is kinda demonstrated already by lifting the 70kg payload with no problem) it has demonstrable advantages over most existing manned aircraft.
      What specific design flaws do you see and how could they be avoided while keeping the size, simplicity and other parameters? You mention no specifics whatsoever.

      1. The thing is it’s not “interesting and fun” it’s stupid and demonstrates ignorance of basic physics: putting the power in two long rows on the medial axis crates stability issues, the framing is not robust enough, it demonstrates a total disregard for pilot safety, and clearly there is a lack of control authority. This demonstrate any new idea, it only shows a fool with too much money out to kill himself.

        1. I disagree. All engineering is tradeoffs and finding good compromise between parameters. If your design isn’t 100% optimal in some aspect that doesn’t mean ignorance, it usually means prioritization, and in the case of motor distribution it’s totally justified. I’m surprises you’d call the project stupid because of such a stupid issue. The motor placement in a multicopter affects you more the more out of balance you are and the difference between the two rows design and all motors equidistant in a circle on an average flight is perhaps a 0.001% loss in terms efficiency and of agility (which here is probably unimportant). There are tons of multicopters out there with motors in rows, unsymmetrically, one on top of another or otherwise not ideally equidistant, and all fly perfectly. It’s like saying all cartesian-based 3D printers are terribly stupid because the forces, inertias, etc. acting on each axis are different. It’s exactly the right thing to do to put unimportant parameters farther in the queue of optimization and a sign of good engineering.
          For the record, the test flight in the video shows that the design is great and behaves just as stable as small model quads. It’s not tested for the FPV racer style maneuverability while slaloming between trees because that’s probably not a design goal. It’s also not as stable as an out-of-the-box Phantom 2, no multicopter is when you make your first flights and if you’ve never seen anything beyond an out-of-the-box DJI multicopter or have never done PID tuning on a Flight Controller, your view is skewed. The video fully demonstrates the flight capabilities are great and I can only assume that the people who still comment on the impedance of the internal combustion engines and the absolute need to use electric, have not seen the video and assume the goal is an FPV racer style agility (probably mortal for the pilot).

          1. And yet it crashed…

            Just because a design is different doesn’t necessarily mean it will be better. Given agile enough active controls and enough power, anything can be made to fly, but that does not imply that everything should. There is no demonstrable advantage in this configuration that warrants merit. This is a poorly built, poorly conceived deathtrap which is indeed as someone already said a Darwin Award waiting to happen.

          2. Flying a person on it, before full proper testing, for longer than a demonstration requires, probably is a Darwin Award waiting to happen.
            It has done remarkably well for a maiden set of flights, a usual RC build takes ~5 to 10 crashes before the thing actually looks like it can be flown and landed safely.
            I think you’re expecting the design goals require a revolutionary design. This is in not much different than the builds in the Hobbyking Beer Lift contest, although I’d say it actually beats the winners of that contest in various ways and is a major engineering feat compared. I’m sure if you’d applied you’d have won singlehandedly at the same time advancing the state of the art in efficiency and aerospace safety.

          3. I don’t expect a revolutionary design, the fact is that there are few truly revolutionary designs in any domain, but that’s beside the point. This design isn’t solving any problem, it’s not pushing into a new part of the flight envelope for this class of machine, it’s just poorly conceived, creates far more issues than it needs to, and by the looks of what we are being shown, not very well built for something that is supposed to carry a human pilot. I understand some look at this and think, gee he actually did it, many do not that want to, so kudos to him. OK but at the end of the day just taking flight isn’t enough for me to be impressed.

      1. That IS how it works with multiple thrust sources and active stability control. The “pendulum rocket fallacy” only applies to single thrust vectors with the center of gravity directly below it,.

        1. Sorry that was lazy of me. I was referring to earlier remarks I made about putting the engines in two narrow lines and the control interactions this will cause between roll and pitch.

          1. Active control is another thing. Some here claim that whether the center of gravity is above or below the center of thrust has any effect on passive stability. It doesn’t regardless of how the mass is distributed, but it’s a common mistake to make. It doesn’t matter if the seat is on top or bottom.

          2. The fallacy only applies to unguided rockets, since there are imperfections in construction which lead to the trust being applied ever so slightly off (or off quite a bit) of the center of gravity.
            This results in a slight torque about the center of gravity which cannot be balanced by the force of gravity as it acts uniformly on the rocket.
            Rotorcraft have no issue with that because they have active controls and any imperfections in alignment are trimmed by the pilot or auto-pilot.

          3. @Niko – sigh, it applies to anything with a fixed thrust vector. On both rockets and helicopters you may have a guidance system in the form of a flight computer or a human pilot, or not. That is besides the point. The point of the original comment, and similarly of the wikipedia article is that putting mass under or above the center of thrust results in difference / no difference in the torque that results from whether the two centers are aligned.

          4. The original comment was about the use of active flight control and the improvement that can be made to the stability by placing the load below props. In an orientation where the props are a sufficient distance away from the center line (and center of gravity) to allow the craft to self level for a given total vertical thrust, and allow simple correction of level by reducing power to one or more engines, instead of increasing power. This only works when all the engines are far apart. And the design in the article has them close to the center of gravity in the yaw axis, but spaced out farther in the pitch axis.
            The craft in the article could not be flown by a human. It requires variable pitch props and a flight controller.

          5. The comment says “please mount the engine frame *above* the pilot seat. The weight suspended below the drive will give you a lot more stability.”, and the second sentence is false. That’s it.
            A human might be able to control the aircraft with a basic channel mixing support similar to how so many drone/rc pilots are able to use the manual (aka “acro”) mode. But obviously a flight controller will be smoother and better at it.
            No it doesn’t require variable pitch props, how would you arrive at that? The octo in the video flies prefectly well with a payload similar to carrying a pilot and has no variable pitch, you’re saying that is impossible?

          6. It will be impossible to control for long duration. The response time of an engine is far slower than an electric motor. The human could not continuously monitor eight engines for long duration, and fly in a controlled manner. But, it’s no big deal, there are many aircraft that can’t be flown by humans without help from computers. The B-2 bomber and F-117 are examples.

  8. If he went to variable pitch props and ran the 2C engine at its peak HP RPM, then he would have instantaneous control of the power output of each motor.

    And life insurance. . . I recommend Whole Life.

  9. Rotate the engines 90 degrees and affix apparatus underneath a large egg-shaped helium balloon. You can get away with hydrogen but be advised that some German attempts proved not so successful.

    1. Neither did the american ones with helium :P
      EVERY single airship, be it a blimp or aerostat either had a serious crash in it’s lifetime or was downright destroyed by one…being lighter then air means you are going to go wherever it pushes you, weather you like it or not ;-)

  10. when life gives you lemons, make lemonade.

    when life leaves you a crate of weedeaters, then make a flying chair.

    As others have said, as long as no one else gets hurt or at risk, why TF not?

    trying to get that many gas motors to sync nicely will be like herding cats.:)

Leave a Reply to lol Cancel reply

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