Honda EXlink Extended Expansion Linkage Engine

Honda EXlink engine intake stroke

Honda EXlink engine intake stroke

For all of you motorheads still interested in burning fuel instead of draining batteries, the Honda EXlink (Extended Expansion Linkage Engine) is worth a closer look. It's inspired by the 130 year old Atkinson cycle engine. What makes it unique? The power and exhaust strokes are 1.4 times the length of the intake and compression strokes and before you see the linkage it's difficult to figure out how that would work and unless it's in motion, you still won't get it. Another feature is that during the power stroke, the connecting rod never deviates more than 2.4 degrees from the direction of the piston compared to as much as 16 degrees in a conventional engine. The result is far less side load on the piston reducing friction to less than half of normal. Though the extra linkage adds friction, the EXlink still has the advantage of the extended power stroke.

Honda EXlink engine power stroke

Honda EXlink engine power stroke

The thermal efficiency of an engine is based on its expansion ratio. Normally, compression and expansion are identical so if you try to increase expansion you also increase compression, not always possible without inducing knock. In the EXlink, compression is 12.2:1 while expansion is 17.6:1, increasing thermal efficiency. Pretty slick.

EXlink engine components

EXlink engine components

Honda is currently using the liquid cooled, single cylinder EXlink engine in a household cogeneration unit using natural gas for fuel. The unit also recovers heat from the engine's exhaust to provide hot water. There's no mention of vehicle applications but continued development of the internal combustion engine is always good.

There are moving illustrations of the engine internals on the Honda website.

Thanks for the tip, Chris!

Link: Honda


  1. B*A*M*F says

    Were I to hazard a wild guess, the drawbacks of the engine in a vehicle would be the weight of the extra components. An even wilder guess would be that this type of engine might not change speeds very well and might have some RPM limitations.

  2. B50 Jim says

    That’s a lot of hardware thrashing around down there. But knowing Honda, they’ve probably tested it to about a zillion RPM. Still, it takes up a lot of space even though it has considerable benefits over a plain crankshaft. Might work great in a diesel where weight and space aren’t so much of a consideration. How ’bout a 2-stroke?

  3. JR says

    Scavenging a poppet valve two-stroke might be tricky.

    Atkinson cycle engines give up power density for thermal efficiency. This is a very tricky (and awesome) way to run the Atkinson cycle. Toyota and Mazda do this with a late closing intake valve on the compression stroke. In this case it seems that the trade-off is some extra rotating mass (like previously mentioned).

    Bring on the Atkinson cycle engines

  4. dan says

    That’s a novel approach to capturing the wasted motion of the crank all the while smoothing out the engine in the process!

  5. FREEMAN says

    Interesting engine. Looks big, though, on Honda’s website for a single cylinder. Anybody know of the difference in size between the two?

  6. GenWaylaid says

    I once spent some time trying to come up with an epicyclic crankshaft arrangement that would accomplish this. I couldn’t get a compact arrangement that all fit together in 3D. Honda’s rocker link seems to neatly solve that problem.

    I just have one request: Could that eccentric shaft that turns at half crankshaft rate be moved closer to the valves? Put some cams on it and build that big gear into the flywheel, and we’re getting close to the parts count of a conventional OHV engine. Just a link, a rod, and two bearings per cylinder different, so on par with the complexity of cylinder deactivation or variable valve timing.

    Of course, the simplest way to get the Atkinson cycle is still with a really long intake cam, but that does suffer some extra pumping losses. How does that compare to the extra moving parts friction here? I don’t know.

    • JSH says

      Atkinson cycle engines are used in lots of gas-electric hybrids. It is a more efficient cycle and the loss of low-end torque is compensated for by instant torque from the electric motor.

      • BigHank53 says

        There’s an Atkinson engine in every single Prius you’ve ever seen. Don’t know about the Ford and Honda hybrids.

          • Paul Crowe - "The Kneeslider" says

            I guess I wasn’t following the hybrid car technology close enough. They don’t seem to say much about the Atkinson cycle engines, at least I don’t recall seeing anything. Interesting.

    • Paulinator says

      I have old sketches of a similar layout applied to a 2-stroke. No Atkinson cycle, but longer dwell at TDC, minimal rod displacement angle, very high crank-case compression and longer power-stroke/shorter charge-transfer phase…courtesy of a jack-shaft and second connecting rod. I dropped it because of the stacked bearing tolerances, higher part-count and other self-defeating characteristics.

    • Paul Crowe - "The Kneeslider" says

      It looks like Honda made quite a big change to the original Atkinson setup with the EXlink. Look at the connecting rod swing on these animations. Honda may have been inspired by it but they changed it a lot, which makes me wonder about the internals of all of those Atkinson engines the hybrid cars are evidently using.

      • JR says

        As far as I know, pretty much standard DOHC fare. Just the intake cams have a really long duration and close “late.” Effectively shortening the compression stroke by 15-20%

  7. Bill Todd says

    [quote]I just have one request: Could that eccentric shaft that turns at half crankshaft rate be moved closer to the valves? Put some cams on it and build that big gear into the flywheel, and we’re getting close to the parts count of a conventional OHV engine. Just a link, a rod, and two bearings per cylinder different, so on par with the complexity of cylinder deactivation or variable valve timing.[/quote]

    Yes I’m sure the layout could be varied. It looks to me that, as long as the 90° phase relationship between the two cranks is maintained, the eccentric shaft could be anywhere around the crank (nor does the bell cranks need to be a right angle).


  8. Tin Man says

    Give me an old 350 Chevy, Thank You. All this over complicated, throw away High Tech may be fine for impressing people, but over time the maintence costs of all the new stuff will eat you alive. Dollar for Dollar over the life of a vehicle a simple pushrod V motor is unbeatable, either in a car or a cycle. Fire Suit On, Fire Away!!

    • JR says

      Haha… at least you’re ready for it.

      I think your desire for simplicity is not ill-placed. I agree with you on that. I do think we need to continue focusing on the new technologies for ICE’s, unless we want to risk losing them altogether.

      As far as some high tech ideas for ICE’s we’re seen lately… I think this is relatively simple and is really just more bearings and some more weight for increased thermal efficiency. This doesn’t require any more sensors or computers or high pressure injectors…. etc.

      Also, having all the regular joe’s buying little cars with efficient engines allows others of us to still run our stinky old V8’s for fun. I’ve got a Ford Fiesta with a small high compression engine… and also an 87 Foxbody Mustang that burps and farts unburned hydrocarbons likes that’s its job. Long live all of it.

      • Paul Crowe - "The Kneeslider" says

        I agree, this isn’t high tech, it’s just a different way of doing something already done 130 years ago. It’s all mechanical redesign and pretty impressive, too.

    • sluggish says

      It’s really not any more complicated than the balance shafts that have been lots and lots of engines for the last thirty years.

      That’s not a knock on pushrod V-8s. GM’s latest engine management systems are pretty astounding: who would have thought a Suburban with a slushbox could get 22mpg on the highway?

      • JR says

        Just look at how many GM LS engines go into big power engine swaps…. those things are indestructible.

  9. says

    Have a look at a simpler arrangement aimimng at the same benefits on The MECXIC engine…….
    It is arranged to give almost 100% scavenging of the combustion space, AND it utilises the extended power stroke to dump most of the exhaust gases via a 2-stroke like exhaust port too !
    Honda missed these additional features I guess …… ?

  10. says

    I wonder how effective this might be?

    I see some potential with this to a small degree, primarily at lower rpms, such as the attempts we have seen to gain more from the power stroke with a late-opening exhaust valve.
    However, the expansion is pretty much over by 90*ATDC, so it is questionable in my mind what would be recovered by a longer expansion cycle.

    Additionally, we would like to retain some decent cylinder pressure remaining for the exhaust blow-down event to reduce pumping losses as engine speeds increase.

    And, as compression ratio increases, the work performed by the expansion becomes increasingly biased toward earlier angles in the power stroke and can support even earlier exhaust valve opening angles, so I see this as more viable in low compression engines.

    It’s interesting, but I believe this is more intended for generator purposes, and not so much toward high performance vehicles.

    Perhaps I’m wrong, but that seems to be the way it is striking me at this time

    • says

      Re MECXIC ……

      Thanks for your comments Tom:
      The dummy Atkinson effect which the Toyota Prius aims at is achieved by keeping the inlet valve open after BDC – a bit of an inefficient way of reducing the inlet charge to achieve an artificially extended power stroke: big pumping losses…. sucking in a full charge then blowing out a big chunk of it !
      MECXIC dumps most of the exhaust gases via the BDC cylinder barrel ports, then purges the rest via a normal exhaust valve: smaller exhaust pumping losses, full purging as the cylinder goes right to the top (and avoids contaminating the fresh charge with the previous cycle’s combustion chamber full of exhaust gases). Reduced NOx by avoiding the re-burning these exhaust gases incidently.
      Thanks again for comments.