Duke Engines – 5 Cylinder 4 Stroke 3 Injector Valveless Axial Engine

Duke Engines axial valveless 5 cylinder engine

There have been a number of axial engines designed over the years with some similarities between them, but these examples from Duke Engines of New Zealand show a very impressive and well developed prototype. Arthur, one of our readers, spotted the engine on display at the recent EAA Airventure in Oshkosh, Wisconsin and thought you might be interested. The engine has 5 cylinders, 3 injectors and zero valves.

Duke Engines reciprocator and crankshaft

Duke Engines reciprocator and crankshaft

The pistons are all attached by their connecting rods to a reciprocator which rotates as a result of the movement of the pistons and is itself attached to the crankshaft which rotates in the opposite direction. As the pistons move through their strokes, they also move past the ports, injectors and spark plugs, eliminating the need for valves. It's a very ingenious arrangement.

Duke Engines internal components

Duke Engines internal components

The axial engine has a number of interesting features:

  • The counter rotating components generate very low vibration
  • Only 3 injectors and plugs for 5 cylinders plus no valve train reduces the parts count
  • It is designed to run on a wide variety of fuels
  • Lighter and more compact than similar displacement conventional engines

Watch the videos to see the internals in motion and how it works and also note the balanced coin on the running engine as it is repeatedly revved.  It's a natural for aircraft which explains the Oshkosh display, but this type of compact, low vibration engine could easily find its way into a number of applications. I like it.

Thanks for the tip, Arthur!

Link: Duke Engines

UPDATE: I added this link in a comment below but figured some of you might miss it so here it is. It's a whole series of old axial engines of various designs. Pretty neat. check 'em out

Videos below:


  1. barry says

    Fascinating, ever since the Wankel i’ve been waiting for something like this, seems to have huge possibilities in ‘planes, f’rinstance, or, at a constant speed (revs) something like that weird Jaguar with the turbo motors (x*75?) to be some sort of new hybrid. Reminds me of that rotary powered battery featured some time ago, very many thanks, Kneeslider for keeping this old head working and enjoying,

  2. Nicolas says

    Pretty cool idea, even better that it’s been built and tested.

    How do they garantee a consistent and long lasting seal between the rotating cylinders and the “head” or whatever the part containing the ports and plugs is called ?

    • scritch says

      I noticed that, too. In the past that was a big problem with Wankels, but I’ve been told since that seal technology has been developed to the point where they last a good while. I assume the same seal technology can be used on this engine.

      Based on my limited knowledge, it seems like this engine would be best used for low-torque, constant-rpm applications, as it looks like it has less torque than a standard ICE, and with no valves, it has no variable valve-timing.

      • Jason says

        I don’t understand why this would be a low torque or constant rpm engine either. The first video shows dyno output onscreen briefly as they start a run and it shows torque hit 50 at just 1500 rpm so it might not be a big displacement turbo diesel, but it makes what seems to be normal torque for just off idle. In the video they claim to already be matching reciprocating engines, so it sounds like it is making good power. And they mention being able to run it up to new, higher engine speeds (than reciprocating engines). It seems without a valvetrain and without the sudden acceleration of rod and crank at TDC and BDC, as well as control of the shape of the ramps the piston follows up and down that they could do a lot of “revs” with this engine as well…

  3. rohorn says

    There were swashplate engines (available any day now!) on display at the Oshkosh airshow when I was there…over 30 years ago. I think it was from “Dynacam” back then.

    Look up “Swashplate engines”….

    • Paul Crowe - "The Kneeslider" says

      As noted in the first sentence of my post, there have been quite a few axial engines over the years, some using swash plates, some wobble plates, they’ve been used in aircraft and even torpedoes. Whether this one is better than the rest remains to be seen but it looks like it’s pretty well designed.

      • rohorn says

        Forgive my reaction – “Swashplate (or variation of) Engine at Oshkosh” generates the same feelings as “Amazing New Engine” on the cover of Popular SciChanics. There has been at least one “Rotary V piston” engine either on display at Oshkosh and/or in Sport Aviation magazine.

        That said, it looks like a dream configuration for all sorts of vehicles. Why settle for a Square Four when you can have a Round Five? I hope it is successful.

        • Alex says

          I am sure that this engine as the engine and the Wankel, OPOC never be widespread in the production as they have many shortcomings in the design. This engine is too complicated to manufacture and he had problems in seals.

    • Justin says

      Dynacam has worked on engines like this in the past absolutely. This company was sold and sold again. A current company in the states (I forget what it’s called) owns a lot of this tech. A similar geometry was used in older torpedo designs as well. The geometry fit well inside a tube, and it didn’t matter about the seals. Only needed to run until it hit the target. Geometry like this is always found in axial pumps (but doing an opposite function of course)

      However this is different. Think of combining the fixed crank rotary engine of the classic Sopwith Camel of 1914 (I made a simple demonstration of this engine a few years ago here: http://www.youtube.com/user/GarretRavens) with the engines that Dynacam (and others) worked on. The entire block moves. Creates a whole new set of dynamic problems (think of the forces acting upon the combustible mixture, proper air delivery etc)

  4. John says

    I wonder if Mazda could run with an idea like this to reboot the the generation RX-8? … If they could machine the engine case to look like a revolver ammunition cylinder…I wonder?

  5. Jason says

    Would heat be an issue? I’m not sure how/where coolant can be circulated to the rotating inner cylinder block. Maybe air/oil cooled? Well at least heat would be more even with three hot areas (from spark plugs to exhaust ports) compared to wankel engines having one very hot side and uneven heating (causing block cracks in extreme use)

    • Medviet says

      Yeah that’s what I was thinking – ok for a plane that’s constantly moving, but how about a motorbike iddling for long periods, or moving through slow traffic? Plus, pistons move more hence genrate more friction…

  6. Mzungu says

    Not a new idea… one I first heard about with the CEM engine from Eddie Paul Industries (http://thekneeslider.com/archives/2005/05/27/cem-cylindrical-energy-module/) which simply uses rollers on a sinusoidal cam ring and puts pistons on each end of the rod, doubling the power capacity with little more space. Much like the Dyna-Cam design mentioned above, but with the pistons/cylinders rotating inside a fixed housing (where dyna-cam rotated the shaft & cam ring inside a fixed cylinder block & conventional valving on the ends.)

    Also, the CEM puts four strokes within just one revolution, where these guys appear to have 12 in one rev. With either of these designs, I would think that centrifugal force on the pistons would limit their revolution velocity… not a game ender, tho, particularly if one placed linear bearings on the piston rods to add support.

    Cool stuff with lots of potential! I wish them the best of success! (And hopefully a contract with a motorcycle manufacturer! Wouldn’t one of these be great in an R6 chassis?) :-)

  7. Alan says

    From the FAQ: Axial engines are challenging to make practicable at typical engine operating speeds.

    So it looks like there is still a way to go. Interesting, though.

  8. Bill says

    The spider has to transmit the all the crank-shaft torque to the engine casing. The natural motion of this spider is actually a figure of 8.

    The linear slipper that is shown in the animation causes a vibration (as the spider tries to push the engine casing in a figure of 8). Charles Redrup (http://www.douglas-self.com/MUSEUM/POWER/unusualICeng/axial-ICeng/axial-IC.htm#b) solved this problem on his Bristol bus engine with a crank that allowed the spider to move correctly. He rejected the alternative method of gear coupling the spider, like the Almen engines (which also allows natural figure 8 spider movement) as the teeth of these gears are subjected to considerable stress and are prone to failure.

    Bill Todd

    • DnA says

      When I watched the animation, I immediately thought the same thing: What about all the forces acting on the reciprocater? It’s more complex, but not unlike a common CV joint which of course do fail. Still, the design is brilliant and worthy of real world testing. I can see immediate uses in small power generators especially as it’s efficient with low grade fuels. I wonder if it would run on methane?

  9. Bill says

    Sorry just to clarify: The animation I’m referring to is my one on Doug Self’s site not the one of the Duke engine above. The Duke engine is using a form of toothed stabiliser which does produce the correct spider motion.


  10. OMMAG says

    The action of the pistons on the reciprocator is very similar to a hydraulic pump. I see this as a high friction and high inertia arrangement. The only advantage of this design is less moving parts…. right?

    Good for driving generators or pumps.

  11. trevor cole says

    no matter what this engine could achieve it still proves that some great designs can come from the south pacific (particularly NZ) without the influence of the great minds in the USA. remember John Britten. bloody brilliant Mate!& mostly on a shoe string budget as well,I,d like an engine to repower my 58 Peerless, it would sure make driving an old vehicle interesting eh!! great website luv it….keep it up.

  12. Leo says

    I like this ….but i hope someone can help me remember something i read here a while ago… What was the piston engine that had 2 parallel cylinders and performed the intake compression in one cylinder and the power exaust in the other… please help me .. ive been trying to remember but cant. THANKS

  13. Clive says

    Initial impression is minimal gains to existing internal combustion efficiency at the expense of poor exhaust evacuation previous to the intake stroke and high friction primary drive mechanism.

    However, in consistent rpm applications such as railroad or aircraft, the “crank shaft” could double as a rotor (with the case as a stator) to create a co-generation environment where electrical subsystems are supplanted by the engine itself, without reducing output torque. At the very least eliminating an alternator.

  14. Steve says

    I recall reading in the Phil Irving autobiography that he worked on a rotary engine similar to this in Australia in the 1920s.

  15. hoyt says

    That engine looks like my Guzzi transmission :)

    It sounds good! (a lot better than expected).

    In a motorcycle application I wonder if it would be best with a longitudinal crank orientation? (may not matter if it is compact enough…just thinking the weight of the pistons moving around the crank would make side-to-side transitions easier on a bike).

    Is it me or does this development also highlight the beauty of the boxer engine design considering its age?

  16. Paulinator says

    OK. The vids look really good at first pass…but???

    I am still mystified by the rotating cylinder, however. What kind of seal does the system have and what kind of MASSIVE FRICTIONAL LOSSES does that seal generate at rpm? The specific fuel consumption (lb/hp.hr or g/kW.hr) is a measure of efficiency that is standard in aviation. The specifications on the web site are notably qualified as projections…and the projections are no better than what current reciprocating engine technology already achieves. And then there’s a huge rotating mass that is undergoing constant and severe physical and thermal stress while maintaining tolerances for bearings and seals?

    Good luck with all that. I’ll follow the project hopefully and wish them well (I’ll also prey for the investors that are on the hook).

  17. Scotduke says

    Interesting – I remember seeing an axial type engine shown on a TV programme here in the UK in the early 80s but never heard any more about it. The cooling issues could be addressed, at least in part, by having a compressor end of some sort attached to the crankshaft. The engine could be designed so that cooling air was drawn through the block and exhausted at the rear. I’m sure with some further thinking you could figure out a way to have liquid cooling too. With air cooling though the exhaust from the compressor could be used to charge the air coming into the engine and provide supercharging, although in this instance you’d also need additional liquid cooling for the fuel system as well as for the engine in all likelihood.

  18. Artie says

    With BMW twins retaining an automotive dry clutch and separate transmission, the 180hp Duke would be a relativly easy bolt in project. Anyone interested??!

  19. alexander perez says

    in high school i had this idea to turn an ac compressor to a motor ….. im not alone

  20. Eiderskater says

    My brother shared this link..must be part of the enthusiastic inspirations we sometimes find as kids!

  21. John Miller says

    not sure if anyone else put this together but; the THRUST BEARING between the flywheel and the wobble plate where the opposition of rotation interfaces, is actually going to be subjected to twice the RPM as actual flywheel RPM, as a result of said opposition. Just an insight i had…

  22. Bob Hope says

    The last video showing the coin sitting on a revving engine, is miss leading, when the camera pans out it changes to a running revving motor without the coin on it, then it zooms back in, and magically the coin is back on it…