Broken Promises Of The Wankel Engine

Through the history of internal combustion engines, there has been plenty of evolution, but few revolutions. Talk of radically different designs always leads to a single name – Wankel. The Wankel rotary engine, most notably used in automobiles by Mazda, has been around since the late 1950’s. The Wankel rotary is an example of a design which makes sense on paper. However, practical problems cause it to underperform in the real world.

Invention and History

felixwankelFelix Wankel’s engine was conceived during a dream. In it, 17-year-old Felix was driving his car to a concert. When he arrived, he bragged to his friends that his car used a new type of engine – half turbine, half reciprocating. “It is my invention!” he told his friends. Upon waking up, Wankel became dedicated to building his engine. Though he never received a formal degree (or a driver’s license), Wankel was a gifted engineer.

Young Wankel’s checkered history includes membership in several anti-semitic groups in the 1920’s. He was also involved with the founding of the Nazi party. His conflicting views on the direction of the party lead to his arrest in 1933. Eventually released through action of Hitler himself, Wankel joined the SS in 1940. The end of the war saw Wankel spending several months in a French prison for his wartime involvement.

Work on the engine resumed in 1951 with funding from NSU Motorenwerke AG. The first working prototype was produced in 1957. Dubbed the DKM 54, this engine had a rotor and housing which rotated on separate axes. The engine was capable of great rotational speeds, up to 17,000 RPM. Maintenance was a problem though. The entire engine had to be torn down just to replace the spark plugs.


Unknown to Wankel, Hanns Dieter Paschke was called in to build a simplified version. His prototype was called the KKM 57P. This much simpler design utilized a stationary housing. It pleased everyone except Wankel who remarked, “You’ve turned my race horse into a plow mare.” The KKM design was quickly adopted and licensed. This engine is the basis of the modern “Wankel” rotary engine.

Engine operation

Piston powered engines, chiefly the Otto and Diesel cycle, are current kings of the internal combustion mountain. Piston powered engines turn reciprocating energy (the up and down motion of the pistons) into rotational energy. Wankel flies in the face of all this. A simplified Wankel engine has only two moving parts: the rotor, and the eccentric shaft.

CC-BY-SA-3.0 by Y_tambe via Wikimedia Common

The rotor is a triangle shape, but the sides bow out. Many rotors are also use cupped faces to increase combustion chamber volume. The rotor rotates within a roughly oval epitrochoid-shaped housing. The rotor doesn’t just spin, it orbits on an eccentric shaft which is analogous to the crankshaft of a piston powered engine. A fixed gear mounted to the engine case meshes with a ring gear on the rotor. The gear ensures that the rotor rotates ⅓ turn for every 1 turn of the eccentric shaft.

The points (or apexes) of the rotor create three chambers inside the housing. These chambers move with the rotation of the rotor. Fuel and air are pulled in through the intake port, compressed against the narrow side of the housing, ignited by the spark plugs. The expanding gasses push the rotor through the power stroke until the apex passes the exhaust port, which allows the spent gasses to escape.

The animation shows the process for one face. The genius of the Wankel engine is that the process is happening for all three faces in parallel. In effect, the engine has pipelined the combustion process. It would be fair to say that a single rotor Wankel engine is analogous to a three cylinder piston engine.

Commercial research and development

gmRotaryThere were numerous licensees for the Wankel engine. Just about every major manufacturer spent time researching the concept. GM created a two rotor prototype. Rolls Royce created a two stage model with low and high pressure rotors. A few companies put the Wankel into production. Curtis Wright built airplane engines, Sachs produced small air-cooled engines for everything from chain saws to snowmobiles. Norton created several motorcycles using the design. However, the only major manufacturer still working on Wankel engines for cars is Mazda. The RX series of sports cars has been synonymous with Wankel rotary engines for decades. The last model was the RX-8, discontinued in 2011. Mazda has not given up on the Wankel though, with concept cars such as the RX-Vision as proof of their continued research.

Reality sets in

So why aren’t we all driving Wankel-powered cars? The problem lies in the pitfalls of the design.

Fuel Economy: The Wankel’s combustion chamber is long, thin, and moves with the rotor. This causes a slow fuel burn. Engines try to combat this by using twin (leading and trailing) spark plugs. Even with the two plugs, combustion is often incomplete, leading to raw fuel being dumped out the exhaust port. The small 1.3 liter 232 horsepower two rotor engine in the 2011 Mazda RX-8 gets worse fuel economy (16 city / 23 highway) than the 6.2 liter 455 horsepower V8 engine used in the 2015 Corvette Stingray (17 city / 29 highway).

Emissions: The unburnt fuel, along with burned oil (described below) both result in terrible emissions from Wankel engines. The emissions problems are one of several reasons the RX-8 was pulled from production.

wankel-inside-kart-engineSealing: Rotors use seals on the faces, seals around the central port, and most importantly apex seals. The apex seal rides the wall of the housing, sealing each of the three chambers formed by the rotor. The apex seals are under extreme thermal and pressure stresses as they travel around the engine housing. Failing apex seals are the primary cause of rotary engines going down for overhaul. YouTube is littered with videos showing the rotary overhaul process.

Much like piston rings, these seals have to be lubricated. However, due to the design of the rotary engine, there is no way to keep the oil lubricating the seals out of the combustion chamber. Mazda engines include an injector pump which pushes small amounts of oil right into the engine housing, as well as into the air intake. This oil is eventually burned, causing increased carbon and emissions over the life of the engines.

Overhaul interval: Rotary engines in general don’t last as long as piston powered engines. As explained eloquently by Regular Car Reviews, the primary problem is with the seals. Browsing Mazda and rotary forums shows people rebuilding somewhere between 50,000 and 100,000 miles. However, this all must be taken with a grain of salt. The RX-7 and 8 are after all, sports cars. While some people treat them gingerly, many people drive these cars hard. Aftermarket performance parts like turbochargers will also negatively impact engine reliability.

787The Wankel rotary story isn’t completely bleak. There are some advantages to rotary engines. As mentioned above, rotary engines create more power (albeit at lower torque) than equivalent piston powered engines. They also are more reliable in the short term. With fewer moving parts, there is simply less to break. Rotary engines also tend to fail gracefully. With failing apex seals, rotary engines lose power, but will still get you home. Piston powered engines tend to fail catastrophically, blowing holes into engine blocks, spraying oil and parts all over the place. Rotary engines do well on the racetrack, that is, when they are allowed. Many racing classes (notably F1) have banned rotary engines. Of those allowed, the most notable is the Mazda 787B, which won the 1991 24 hours of Le Mans race.

What does the future hold for the Wankel rotary engine? Most likely more of the same. Mazda will continue to support the engine, and it will continue to be used in some niche fields. However, it would take a major advancement in materials and design to correct all the issues that have thus far relegated the Wankel engine to a footnote in the history of internal combustion.

168 thoughts on “Broken Promises Of The Wankel Engine

        1. Shrad probably meant some other word than “Pantone”. Pantone are a company that sell colours. Yes, very strange, but they sell consistent colours, and certify printers (commercial print shops as well as expensive desktop printers), scanners, monitors, all nice and expensive. The point being that when you choose Pantone 2728C, you know you’re getting a specific blue, the exact same blue on your monitor, printer, commercial printing, etc. They also sell colour books and swatches, with guaranteed, certified colours on them to pick from.

          It’s the only system of certifying colours that I’m aware of, and that’s most of their business.

          Pantone numbers specify a colour. Hence Annie’s joke. Then Shrad and Pez both gave what I imagine is Pantone’s office phone number, for Britain and the USA. A different kind of Pantone number! I saw what they did there.

          That’s your jokes explained for this week.

  1. There have been several rotary ICE designs that have been put forward over the years and none have made a dent in the dominance of reciprocating types with the exception of gas turbines. Possibly some of these might be superior, but much of the problem is that reciprocating geometries have had such a head start that nothing will catch them is the time combustion technologies have left for those applications.

    1. Gas turbine designs have /much/ higher material requirements, though – you’ve got parts immersed in the same temperature of gas for prolonged periods, so they’ll by default be in equilibrium with them, which means you need materials that survive that heat. They’re lighter for equivalent power, though. With reciprocating engines, the various phases that the same part participates in keeps them out of equilibrium, which basically means “cooler than the hot side of the thermodynamic cycle”, and gets you better efficiency for the same material capabilities.

      1. I was not limiting my remarks to automotive prime movers. I strongly doubt that gas turbines will ever see widespread use in cars for several reasons. Along with your quite correct observations, this type of power plant must run at high RPM and has poor throttle response and these too make it unsuitable for this service.

        1. The gas turbine is the near perfect compliment for hybridization with electric motored vehicles. Much better power to weight ration than any reciprocating engine out there right now. You can easily derate at much higher output turbine from 1megawatt down to 300KW to recharge a battery pack while still powering the car down the road especially if you put the e-motors in the wheel hubs. Eliminating transmissions, drive axles, etc. The added bonus to such a platform is that a unibody could very aerodynamic underneath(less drag) and unsprung weight could be cut by a fairly decent percentage, improving the car’s handling and stability.

          1. No actually it isn’t. The cost of this type of power plant is always going to be higher. They may find application this way in heavy vehicles but even there any advantage over Diesel is hardly worth the added expense. Again it is a case of good enough making better a hard sell, a new technology has to be a major improvement, and by orders of magnitude, to dislodge an established one – incremental improvements (as measured by cost-effectiveness) just isn’t enough.

          2. Not by orders of magnitude. 50% better would make a fine economic case.

            A microturbine is a viable alternative to a diesel engine directly on the basis of efficiency when combined with a proper recuperator, which was the problem of the original turbine car in the 70’s. They couldn’t figure out how to manufacture the heat exchanger on the cheap.

            People are also trying combined cycle designs that approach 60% efficiency, which is something no small diesel engine can reach.

          3. >”They couldn’t figure out how to manufacture the heat exchanger on the cheap”

            And still can’t which just underlines what I’m saying about poor economics. Even then 50% really isn’t good enough to move anyone to jump in simply because the risks are still very high for the potential return, at least in passenger automotive applications. Again heavy haul might be different but even then you are asking the end customer to take a leap of faith in an unproven product and thus adopters will likely be hard to find.

          4. How about CNG fuelled Wankel hybrid power trains? A hybrid vehicle is after all a mobile power plant with dual functions: when mobile the power plant is sending electricity to drive the wheels =cash outflow; when static in a garage the power plant would be exporting power at constant rpm to the grid =cash inflow. We know the +s and -s of the former, what are those of the latter? And the net of the two?
            Application more for school buses which “work” only 4 hours/day during 190 days of school year and idle the rest of the time, and not ordinary autos.
            GEEKS enlighten me!

      1. If you have an idea of how the KKM type works, think of clamping the crank of the engine stationary and then rotating the housing. The casing will turn one way and the rotors will turn the other way relative to it. Really good for balance, no counterweight needed and with modern materials science could theoretically rev to somewhere around 40k.

          1. That engine is also called a rotary engine a Gnome rotary (Gnome being a well known manufacturer of that engine) to differentiate it from the Wankel rotary.

    1. Suzuki RE5 was popular in the day. Neighbor still has one that he takes out on occasion. Nice bike.

      I raced RX2’s in the late 1970’s/1980’s. Once you understood the issues, easy to rebuild and maintain. Nothing like the sound of a rotary, nor HP. Dirt cheap to modify by cutting your own side ports. Quite a few experimental aircraft running Rotaty Engines. As the article mentions, they keep on running even if a problem.

        1. Gah, never mind, I had just missed a step during the animation. The roving ports looked like they were doing something different to me with respect to piston placement, but on second viewing I see it’s all four strokes. I’d just delete that bit above, but apparently there’s no such thing as deleting a comment here :)

      1. It seems like the way to solve that issue could be to remove the single fixed engine head, cap each cylinder with it’s own fixed head (each with a spark plug), and then put the fixed intake/exhaust port on the outside of the cylinder wall, near the top. That way, you’d only need seals around those ports.

    1. A wankelnengine is effectively a six cylinder per rotor engine compared to the traditional four stoke. Every rotation of a wankelnengine has three compbustions but only every other rotation of a four stroke combusts. This a Mazda engine is a 12 cylinder engine.

      1. Not exactly. You get 3 combustion events per rotation of each rotor, but the rotors only spin at 1/3 of crankshaft (eccentric shaft) speed. So you really only get 1 combustion event per rotor per revolution as displayed on the tachometer. This is still twice as many as a 4-stroke otto reciprocating engine, so a 1.3 liter twin rotor wankel is equivalent to a traditional 2.6 liter four cylinder.

    2. I’ve read that John DeLorean was looking into this type of engine for the DMC-12 but due to some setbacks with the company developing the engine he had to grab the PRV for production. I can’t find the thread on DMCTalk now.

      I’m glad he didn’t go with this engine. The PRV is hard enough to tinker with. I couldn’t imagine trying to find parts for a Duke.

  2. It may be fuel inefficient and it may not last as long as a piston engine, but the power you can get from it is disgusting. What other engine can whirr away happily at 10,000 rpm that is smaller than a four banger? Its like a two stroke engine that can revv even faster.

  3. “What does the future hold for the Wankel rotary engine?”

    Hopefully the same future as every other fuel burning engine plus or minus a few years: the inevitable death.

    I can’t but wonder why we don’t have “almost hybrid” cars where a smaller traditional engine serves only the purpose to generate electricity for the main electric one plus charging a reduced set of batteries.
    Yep, I know… energy conversions bring losses, and we would have two here, but… a traditional engine, when used to drive a generator, doesn’t need clutch and transmission, almost all gears can be removed (= a lot less friction) and rpm can be fixed so that the engine gets its maximum throughput with less fuel.
    I’d like to see anyone with the necessary knowledge elaborate on this to see if it would be feasible on cars. I’m sure it was done in the past and they use the same principle on some big ships, but how about cars?

        1. And given that the i3’s optional range extending engine is essentially a small motorcycle engine, I wouldn’t be surprised if a future version used an all-aluminum wankel to save weight. No need to worry about fuel economy when you’re only expecting to use it on occasion. What really matters is carrying around all of that dead weight when you’re running on batteries, and for power-to-weight nothing beats a wankel.

          1. Batteries already are a whole lot of dead weight compared to a small motorcycle engine. The battery bank required to achieve a modest range come in at around 250-500 kg.

      1. Does the Volt’s ICE engine recharge the batteries? I’ve always thought it’s either or. IOW, you either run on batteries or the ICE, but they’re separate from each other unlike hybrid cars.

        1. As I recall, the Volt is a stage 2 hybrid, meaning that the engine only generates electricity to either recharge the batteries or power the electric drive motors, or both, but there is no direct mechanical connection from the engine to the wheels. It’s essentially the same as a diesel-electric locomotive plus batteries.

          1. The drive train of the Chevy Volt can couple the ICE mechanically to the drive wheels, if required (Mode 4 of the Voltec drive system)

        2. Apparently the original design of the Volt had the engine just recharging the batteries, but they’d also built it in such a way that the generator / motor / transmission would bolt onto the engine like a conventional transaxle so it could reuse an existing FWD engine and platform. Somebody looked at the transmission arrangement they’d come up with and concluded, “If we put a a couple extra parts in here, we can let the controller couple the engine to the wheels when conditions require.” So it’s able to operate as a series or a parallel hybrid.

      2. And the problem here is the same as with all EV powertrains…the batteries. Battery tech needs a huge leap forward before EV’s can replace IC completely. The Volt is a great example of a hybrid done right.

          1. The claims are highly suspicious, considering that the car was a series hybrid with lead acid batteries. To achieve 70 MPG as claimed, the stirling engine should have been implausibly perfect.

            To achieve the figure, the car should consume no more than 560 Wh/mile, while real world examples such as the Nissan Leaf consume 340 Wh/mile, which means the stirling engine AND the generator AND the batteries would have to exceed 60% efficiency. The PbA batteries alone have a round-trip loss of 40%, which makes the claims pretty much…. bullshit.

          2. Sorry, I actually calculated it in UK gallons. The US gallons for 70 mpg would be 470 Wh/mi and the corresponding efficiency requirement from the powertrain should be 72% which no combustion engine in history has achieved.

    1. My understanding is that this is exactly how the Chevy Volt (and I’m sure other models as well) work – the drivetrain is entirely electric. The vehicle runs entirely off electric power for the first 30 miles (standard daily commute). After that a gas/diesel generator kicks in to recharge the batteries and/or directly supply electricity to the drive train.

      The inefficiencies introduced by the mechanical->electric->mechanical conversion are (at least theoretically) more than made up for by the simplified transmission & ability to use regenerative braking. I have no idea how well this works in practice.

    2. Fuel-burning engines won’t be replaced unless someone makes working and safe Ford Nucleon. When oil runs out, we will be making it from biomass, for example from algae. And those bio-fuels will be more “green” than electric cars.

      Electric cars are not as ecologic or economic as they say. Even if someone develops a battery that can store as much energy as gasoline in equivalent mass, it won’t replace internal combustion engines. Imagine charging 100kWh battery in under two minutes. And then imagine shorting such battery after car accident…

      Diesel locomotives are hybrids because their engines can operate at optimal, economic RPM. Gearbox is unnecessary, which reduces number of moving parts. Then just adjust the power going to electric motors. IIRC they used to do it by regulating power going to generator’s stator. Now they probably use inverters. I’m not an expert, I just read about this some long time ago…

      1. Re batteries: why, that short would be as bad as a tank full of hydrocarbons catching fire!

        Re diesel locos: controlled rectifier (effectively three phase PFC) on generator side, inverter on motor side. I would expect efficiencies of around 98% in both the rectifier and inverter electronics.

      2. There are safe ways to dump or contain 100kWh. There are not a lot of safe ways to dump 20 gallons of gasoline, so you’re betting your life on containment.

    3. “Hopefully the same future as every other fuel burning engine plus or minus a few years: the inevitable death.”

      Did I miss a post on HAD where someone invented a Zero Point Energy supply? If not then we’re going to be “burning” fuel of one sort or another power our transports for the foreseeable future.

      1. Aviation in particular is going to be using heat engines with liquid fuels for a long time as will heavy haul transportation, but it is very likely that passenger cars will trend away from ICE over the next few decades.

  4. The few people who run Wankel engines that I’ve spoken to have made them into what is effectively a two-stroke engine, mixing a bit of oil in with the gas to supplement the oil-injection systems. Apparently it does wonders for the seal problem, but this whole engine screams for advanced ceramic technology or some other major materials change.

    @DV82XLs comment that reciprocating engines have a huge head start is quite correct, and we should remember that it goes back to Watt’s improvements on the Newcomen walking-beam steam engine, giving the engine type a ~250 year head start. Perhaps we should check back in 2266, though I hope burning oil will be a curiosity by then.

    1. Former rotary (2nd gen RX-7 and RX-8) owner here.

      The Wankel is in some ways similar to a 2-stroke and in some ways similar to a 4-stroke but it’s really just a Wankel.

      Oil is mixed into the combustion chamber to lubricate the apex seals. The apex seals are normally carbon steel but race engines often use ceramic apex seals. If you don’t drive it like a sportscar (what a waste) your stock apex seals and engine will last 150k-300k miles.This was far more common for the lower power 1st gen RX-7s than the later turbocharged 2nd and 3rd gen RX-7s.

      The most common failure mode for a Wankel was detonation due to overboost and carbon buildup changing your compression ratio. Basically if a piston engine knocks you’ll probably get a ping and the EFI will figure it out right quick and solve the problem. The rotary Wankel has two problems here though, first, as your burn oil in your combustion chamber you build up carbon in there (which you can clean out by pouring water in your intake while your engine is running. lol. try that with a piston engine!) which increases your compression ratio, making it more prone to pre-detonation of your air/fuel mixture. Secondly, if you detonate you’ll probably blow out an apex seal which will then clatter around in your engine and, if you’re lucky, be thrown out your exhaust.

      If you have a turbocharger, however, the apex seal probably just chewed up the turbine on the exhaust side. If you used a ceramic apex seal you’ve just scratched the everliving crap out of your engine liner. If you’re a 17-year-old hooning around you’ve probably also put too big of a turbocharger on it, increasing the chance of detonation.


      Of course, with methanol injection and the right turbo you can make 500 HP at the wheels without opening up the engine on a 2nd gen RX-7. Some dude in Arizona was doing that back when I had a 7.

      1. I heard somewhere that you can burn off some of the carbon build up by redlining it once in a while. I’m not sure how accurate this is, but I sure used it as an excuse when I owned my 2nd gen RX-7.

          1. Yeah, when I had the 8 there was an on-ramp to the highway that I would gun it on in 2nd. So fun. Beeeeeeeeeee!

            (The red-line buzzer is referred to as the “shift” buzzer in RX7s and 8s because there’s no reason to shift before then in racing!)

        1. So clearly, we need to put the white supremacist organizations on the job of discovering a revolutionary new propelling technique that will solve our problems once and for all. ;)

  5. “Through the history of internal combustion engines, there has been plenty of evolution, but few revolutions.”

    False. There have been many, many revolutions. Thousands of revolutions per minute, even.

    1. “Seems to” is the operative phrase here. What many don’t understand is that there is an immense amount of engineering needed to get from a working prototype to a production unit and even then, it will be several hundreds of thousands of operating hours before the real problems emerge. Consider the first aluminum block auto engines. These did not initially do well, even though they were introduced by a mature industry that could afford to do as much predevelopment work as they did, yet once in service there were issues. And this with standard reciprocating geometry. Taking a new, unproven technology to market is just not going to happen because the risks far outweigh the potential benefits – like it or not – good enough is always the enemy of better.

      1. Aluminum was used in internal combustion engines almost from their beginning. The big problem was coming up with processes to enable aluminum to be used for the cylinder walls, without iron liners.

        1. without liners? Is there really an engine without steel cylinder lining/coating? I’m under the impression the movement is towards floating liners on aluminium or even composite casing materials.

          1. Yes, it’s been used in many engines. Most infamously in the original 4 cylinder in the Chevrolet Vega. They didn’t have the bugs worked out and had to replace the engines with one of a more conventional design. The trick turned out to be mixing silica with the aluminum then chemically etching the cylinder walls after machining so that the piston rings don;t actually slide on the aluminum. Makes boring out a worn block either impossible or a very specialized job.

        2. The point being that there were problems and it’s these sorts of issue, that don’t come up before production that make manufacturers leary of radical new designs when the existing technology is good enough to get the job done.

  6. Great article! Just wanted to point out the article reads Le Mons instead of Le Mans. Le Mons is VERY different than Le Mans and I think the legendary 787B’s team could possibly be insulted.

  7. Notable things with the 787B-first, the R26B engine employed ceramic coatings and seals and other redesigned parts, giving the car much greater reliability than the basic engine design had. It won the 24 hours of Le Mans, not of LeMons (honestly, some of the creations for LeMons qualifies as hacks, like the radial engine MR2 or the MR2/Corolla awd or the gas turbine MR2 or the two stroke miata or the harley engine prius for starters).
    A naturally aspirated R26B without the auto adjusting intakes will produce 550 at wheel Hp and 500Nm torque (or about 370 ft/lbs). It’s not amazing, but in a light sports car a four rotor or a boosted two rotor will turn tires to smoke in every gear and sound like an unhinged monster and go like a stabbed proverbial.
    Compared to many other performance engines, they really aren’t that bad. The usual thing with the engines in high performance vehicles is that they get tuned to within a few inches of their life-there is a reason why endurance racing is some of the most brutal racing on the vehicle.

    And since I’m on the topic of wankel engines, they’re also used in air compressors, range extenders for cars and the motor driven by the explosive charge in a seatbelt tensioner is a tiny wankel engine. You don’t see them, but they are in a lot more places than you expect.

    1. Yeah I was just going to comment that rotary engines are used a lot in stationary equipment. That’s actually where they work best; running at a constant RPM and relatively constant load. Varying the RPM puts more stress on rotaries so they wear faster, and they run much, much less efficient.

    2. I’ve heard of the Wankel seat belt retractors but have yet to actually see one, or even pictures of one. Has me wondering if they could be converted to air motors.

  8. I put over 120k miles on my old Mazda RX-7. The power curve was quite a bit different than the standard engine. At low revs it was a dog, so if you were racing someone (not saying I did) it was important to get a rolling start and keep the RPMs above 2000. That was where the power curve took off and I could beat bigger engines.
    Always puffed a little oil when starting. The odometer stopped working at 120k and I traded it in 2 years later. I had heard that some engines could last a very long time and somr failed at 50k.

  9. What the future? “Liquid Piston”, it’s a kind of reverse Wankel geometry (stator-rotor), advantages: better cooling of the apex seal (because they’re mounted in the case), better use of the combustion volume /cycle…
    Low weight, low number of parts, high reliability are the keys in one domain: AVIATION, where electric is nowhere to be useful, where turbines are fuel guzzlers, and piston engines are almost (injected since one decade) the same as 50 years ago.
    Car engines in aviation is an experimental thing, commercial didn’t seem to survive.

  10. One sentence in an article on Fazlur Rahman Khan mentioning that he was a muslim immigrant to the US and the comments explode about how HaD shouldn’t mention technical irrelevancies. A paragraph on how Wankel was an anti-semite, Nazi and SS member and 50 comments in everything is all about engine performance.

    Maybe all those assholes who were threatening to leave and never read HaD again made good on their threats? Dare to dream…

      1. First of all, there’s a whole range of ‘not relevant’ words that go in article titles, but only a few seem to cause outrage among the internet article title police. And people weren’t just bitching about the title either.

        Secondly, it certainly was relevant to the article. The thesis of that article was that one of America’s strengths is her openness and acceptance of people from many backgrounds. You know, the whole give us your tired etc etc schtick. That bit wasn’t about Khan specifically, but about the culture that allows people to reach greatness in what we now call STEM in part because what you can do matters more than what background you came from.

        Not sure that’s as true now as it was then, as that mess of comments demonstrates.

  11. The primary problem with the Wankel engine is the large surface to volume ratio and the shape of the combustion chamber resulting in poor burn efficiency and poor detonation characteristics. It is, though, an excellent design for a direct injection stratified charge engine where the fuel is injected as it burns. The idea is to inject just enough fuel to have a stationary flame front resulting in no “end gas” to detonate and minimal wasted fuel. As the charge is inherently swept past a fixed point in a Wankel engine, this should work well and allow high compression ratios (through turbocharging, as the combustion chamber shape makes that difficult with natural aspiration). This requires very precise timing of the fuel injection at high pressure, like a diesel but spark ignited. I have been waiting for 30 years for someone to do this. The techlology now seems to be available so does anyone know if this has been tried? With that, the Wankel would make an excellent power pack for a Volt like electric hybrid – small, powerful and (now) efficient.

    1. What if the Wankel is for a hybrid power pack being the prime mover of a genset fuelled by natural gas for primary purpose producing electricity for export when garaged (static mode) and on CNG when in secondary dynamic mode powering electric motors driving wheels?

    1. There was some work being done on the concept as late as last year in India, as I recall, at the Automotive Research Association of India, based on a press release that saw.

  12. Graphene, lasers and robot.

    Graphene for the wearing surfaces. Says here:

    “graphene can provide extreme resistance to wear (as much as four orders of magnitude reduction in wear”

    Lasers to replace spark plugs. Possible? In theory could be directed and so fast that the whole fuel/air mixture could be hit.

    Robots. Less obvious. Design an engine that could be stripped down and refurbed by a purpose built robot. Quick turnround, cheap and, if the engine then lived up to its promise might be a robot in every garage.

    However, I suspect all internal combustion will be rendered obsolete by battery breakthroughs (Lithium-air, Lithium-fluorine?) and local, cheap renewables to charge them. Low noise, very low pollution.

    1. I like the “robots” idea. By eliminating labor costs, one could use ridiculously labor-intensive techniques like replacing the rotor seals each time you change the oil. The whole thing could also be made more compact, because robots are okay with disassembling the engine to replace a spark plug.

  13. Anecdotal, but I know quite a few people who’ve had Wankel-engined Mazdas, and none of them needed engine work under 100,000 miles or even well past that. Maybe they slowly lost some power, but reciprocating engines do that as well.

    Part of the fuel economy issue must be related to the larger exposure to inside surface. A cylinder has a higher ratio of volume to surface area than a… whatever shape you call the changing combustion chamber of a Wankel. So much heat is dissipated during combustion that there’s just not as much work done.

    But a lot is due to inertia. Wankels would make better engines for light aircraft (than piston), yard machines, and drones! But very little development has occurred relative to reciprocating engines.

    1. Sorry, but “5-stroke” doesn’t make any sense at all. Since it’s a reciprocating engine, it is not possible to have a cycle with an odd number of strokes. This would more accurately be called ten-stroke, since it requires two four-stroke cylinders and one two-stroke cylinder for the scheme to work.

      1. You are correct in the conventional sense there are not 5 strokes however it is a good descriptor of how the engine works. The engine will easy fit into today’s production lines and provides better than diesel power efficiency. It is an interesting engine!

  14. I used to rebuild rx7/8 engines, while they were both labled a “13b” engine. They are very different, the rx7 (were talking fd3 here as the bulk of my experience) did indeed suffer apex seal failure as a main reason for low compression, and even a well looked after example would struggle to hit 100kmiles without a loss in compression. These engines as a rule usually only failed due to lack of compression.

    The rx8 13b is a diferent creature, mazda seemed to have the apex seal thing sorted, i have several regular customers over 100k on origional engines with perfect compression. The problem with the rx8 engines is the bearings supporting the e-shaft. These seem to wear out fast, the e-shaft starts to tilt in the housing and compression is lost due to seals running out of true. If mazda could have just kept the eshaft bearings reliable the rx8 lump could have very genuinely been a reliable engine.

    It was woefully underpowered tho. Should have been turbocharged like the fd3.

  15. A rotary is not analogous to a 3 cylinder petrol motor, for comparisons sake a Mazda 13B is closer to a 2.6lt 4 cylinder motor however for the same number of rotations as a 4 cylinder to complete the 4 stroke process only 4 faces of a rotary have been through the entire combustion process, if you take into account all 6 faces completing the combustion process it is closer to a 3.9lt engine. There’s plenty of sources available on the internet that go into great detail to explain this.

    But take it from a former rotary owner, fuel efficiency is the last thing on your mind when you have a (in my case) 290hp motor in such a compact package.

  16. Since we are listing out production Wankels, OS made a Wankel for rc airplanes. We have one new in the box that we just never got around to building the right plane for. You can hear them run on youtube though.

    1. Drooling over that (and the other OS engines) in the Tower Hobbies catalog back in the 80’s was how I first came to learn about the Wankel engine. The actual engine is discontinued, but you can still get parts for it – but the prices are insane – $239 for a rotor and housing!

    2. that engine is freakin’ great!!

      mine is/was lost somewhere in the Adelaide Hills, a few bush fires have burnt through that area so it’s probably melted. :-(

      way less vibration than any engine I’ve ever ran

  17. If the Wankel engine housing is an Oval Epitrochoid, what is the rotor? A Spherical Triangle (a.k.a., Reuleaux Triangle)? Be good, don’t be a Troll and just reply yes or no – cite proof.

      1. Perhaps, although it shouldn’t be called a “Law” when it fails under border conditions, e.g. high pressure, low volume, low temperature, or when there are significant intermolecular forces.

          1. Let me rephrase; nothing in the real world behaves exactly like an ideal gas, especially under conditions that may occur in an engine with an unusual design, so ruling out the possibility of something working based on an “ideal” gas “law” is not strictly fair. So whilst what you say may very well be true, it really requires a working model to be tested by an independent third party to be sure, which may or may not have been done by a Russian company that supposedly “stole” the idea, but unfortunately there’s nothing publicly available to know for sure either way.

          2. Regardless it does not matter if it is ‘ideal’ or not. The thermodynamic losses kill the efficiency of these compressed air systems simply because there is no practical way to deal with the heat losses both in and out. Any claims to the contrary, upon careful examination, are either based on a poor understanding of the physics or outright delusion if not fraud.

  18. Should have referred that the 787B only raced one year and Wankells are perma banned from LeMans since its single race. Such a beautiful sound :(

    Sachs also made small 250cc Wankells, used in lawn mowers and some other agricultural tools.

  19. General Motors used the Wankel to break American Motors. AMC was designing a radically new car, the Pacer. It would look unlike anything that had ever been on the road before (and other companies are still nicking design elements from it after nearly 40 years) and it would have a revolutionary engine too, the Wankel, built by General Motors under license from Mazda.

    AMC has everything ready to build the rotary engine Pacer, they’ve sunk most of their ready funds into the project when GM goes “Wankel engines? What Wankel engines? Were we going to build a Wankel engine factory and sell engines to AMC?”

    So AMC was stuck with an all new production line for a car without an engine. They had to hastily alter it to stuff their olde straight six under the hood, making the car over weight and underwhelming in performance, though for a couple of years they offered the 304 V8 as an option.

  20. Two items of trivia:
    1- As stated in the article, Mazda engineers took some pains to lubricate the apex seals with good results in that, on average, the seals showed good life. With everybody predicting apex seal failure, no attention was paid to the side seals, and, on the 12B engine (pretty confident it was the 12B and not its 12A precursor), those side seals were the primary limit on engine life.
    2- Somewhere, probably squirreled-away in a Toyo Cork Kogyo filing cabinet, are plans for a 12 series engine whose coolant was the lubricating oil rather than any, separate, water-based system. The speculation was that oil cooling might smooth thermal fluctuations and thereby assist in reducing nitrous oxides as well as make the engine assembly that much simpler. I know of no, actual prototypes being made much less tested.

  21. David Garside the developer of the Norton rotary has patented SPARCS (Self-Pressurising-Air Rotor Cooling System). This is a system that utilises self-pressurising blow-by gases as a cooling medium, absorbing higher levels of heat from the engine core and dispersing the heat by means of an external heat exchanger, providing superior heat rejection.

    Also CREEV the rotary exhaust expander unit or CREEV (Compound Rotary Engine for Electric Vehicles) for use with Wankel rotary engines. The CREEV system acts as an ‘exhaust reactor’ by consuming unburned exhaust products while expansion occurs, reducing overall emissions and improving thermal efficiency.

    There is also HCCI ign being investigated by Mazda. If they get that on a Wankel they have cracked it.

    The Wankel rotary is far from dead, things are moving on for sure. It is the perfect unit for range-extenders, which are part- time engines guaranteeing longevity. Running at a constant speed and load will clearly give superior emissions and vastly improved fuel consumption.

  22. Liquid Piston in the USA are making advances on a rotary, but not a Wankel. They turned it inside out, having the valves in the rotor. The US government has given them a grant for R&D. They have one running. It appears to have eliminated most of the problems with the Wankel rotary.

  23. Hello.

    Each combustion chamber of the Wankel Rotary engine is sealed by a “sealing grid” comprising eight different seals: two apex seals, two side seals and four button (or corner) seals.
    With eight gaps to leak from (one per button seal and side seal, and one per button seal and apex seal), the leakage of air / mixture / burned gas from a combustion chamber towards its neighbor combustion chambers is several times more than the leakage in a reciprocating piston engine.
    The working surface, whereon the seals abut and slide sealing a combustion chamber, comprises two side flat surfaces and a cylindrical surface ending on the two side flat surfaces at an angle of 90 degrees forming a corner wherein the curvature gets infinite.

    The “reverse Wankel” LiquidPiston engine has similar, if not worse, sealing issues:
    The “sealing grid” for each combustion chamber comprises two “peak seals”, two “side seals” and four “button seals” (which means eight gaps for leakage per combustion chamber).
    In comparison to Wankel’s architecture wherein all seals are mounted on the rotor, in the LiquidPiston engine some seals are “stationary” (they are on the immovable “tri-lobe” casing) while some others are “movable” (they are on the “two-lobe” rotor), and because the motion of the rotor is far from being geometrically perfect (there is, inevitably, a backlash / play in the synchronizing gearing, there is also a “play” in the bearings between the cooperating parts, there is also a deflection of the eccentric shaft due to the high pressure load and to inertia load, etc), the overall leakage cannot help being problematic, as in the Wankel rotary.

    The PatWankel rotery engine:



    uses a 3-D curved working surface to smoothly (without steep corners and infinite curvetures) bridge the two side flat surfaces of the prior-art-designs, enabling a sealing similar to that of the reciprocating piston engines.

    The gas inside each combustion chamber is sealed by one only “closed” seal, just like in the reciprocating piston engines.
    There is only one gap per combustion chamber, just like in the reciprocating piston engines.

    When you get the time take a look at


    1. It always comes down to the materials used. The “seals”. We could compare the pro’s and con’s of AK vs AR part counts. Or the virtues of JB-Weld vs future meta-materials. Truth is they didn’t have viable DLC’s (Diamond Like Coatings) that Racing Teams in the US have to send parts off to Europe to coat with a commercial, hyper-proprietary, secret sauce.

    1. JohnScnow: “Who killed the electric car?” The technology wasn’t ready. Every time somebody tried to market an electric car, the same problem came up: insufficient distance between charges. Only very recently, with the development of Lithium rechargeable batteries for laptop computers and cell phones, has electric personal transportation become practical. And the industry is thriving. There is no, and never has been a conspiracy.

      What killed the Wankel? Poor efficiency and poor reliability. The Wankel engine was introduced at a time when fuel costs were high and rising, and nobody wanted to hear about a new engine design that did nothing for this. The only advantage the Wankel gave (as well as the “Liquid Piston” you link to, probably) was high power/weight and power/volume. But the weight advantage was more than offset by low efficiency. It’s no coincidence that the only successful Wankel-engined car was a sports car.

      1. “There is no, and never has been a conspiracy.”

        Hey, I didn’t say “it was an inside job.” :D

        No, the truth was there was no incentive and no subsidy. We would have literally had to have a fleet of bicycle “Borrow, Return, Borrow another). Even to this day 1 modern diesel train is more efficient then 100 Semi trucks. Petroleum was/is cheap. War, War never changes.

        Bio-diesel works, Ethanol works (in sugarcane land/brazil, not corn/rapeseed/canola/hemp because we have subsidies not to grow to maintain commodity prices). Heck man, Plenty of good stuff out there. “Hydrogen economy”, “Liquid Nitrogen economy” even a “Ammonia economy”.

        Everything has a place when used correctly. We obsess that “that isn’t efficient”. But Efficient vs Effective are two VASTLY different things.

        Before I get too verbose or pedantic (if I haven’t already). I would conclude in a civil manner and say “Good, Fast, Cheap: You can only pick two. Good and Fast = Not Cheap, Good and Cheap = it will take time to get to, Fast and Cheap = probably not good for what one might be doing.

  24. Wankels do have a place on the roads however, its like this article forgot all the good advantages in the end… Wankels weight less for the hp as typical piston engines, they are super balanced allowing for a very low idle and a very high redline and distribute a linear power band, make relatively low amounts of torque and at a high rev range… Etc. Etc.

    Lots of us think the same: the rotary engine is perfect in racing conditions and belongs in the hall of fame for the best engines a race machine could have.

  25. Le moteur Wankel est basé sur une escroquerie aux taxes sur la cylindrée. Quoi d’étonnant de la part d’un SS Nazi ?
    Imaginez un moteur 3 cylindres 2 temps qui comporterait en bout de vilebrequin un multiplicateur par 3.
    Est ce que pour autant vous ne compteriez qu’un cylindre sur les trois et prétendriez qu’il s’agit d’un 4 temps ?
    Non, bien sûr vous vous référeriez au déplacement des surfaces de piston pour calculer la cylindrée, n’est ce pas ?
    Et vous constateriez qu’à chaque tour de vilebrequin il y a une explosion par cylindre.
    Vous avez là la réplique exacte du fonctionnement du moteur Wankel.
    3 calottes de piston = 3 faces de rotor
    3 explosions par tour de vilebrequin = 3 explosions par tour de rotor
    Un multiplicateur par 3, bien inutile, fait tourner l’arbre de sortie 3 fois plus vite = l’arbre à excentrique tourne 3 fois plus vite que le rotor.
    C’est tout à fait comparable.
    Le moteur 13b de Mazda est donc un 3.9L 2 temps, avec une efficacité très faible et peu de couple (à cause du multiplicateur par 3).
    Pas de miracle Félix !

    The Wankel engine is based on a displacement tax scam. What wonder from a Nazi SS?
    Imagine a 3-cylinder 2-stroke engine with a multiplier by 3 at the end of the crankshaft.
    However, would you only count one cylinder out of the three and claim that it is a 4 stroke?
    No, of course you would refer to the displacement of the piston surfaces to calculate the displacement, right?
    And you would find that with every revolution of the crankshaft there is one explosion per cylinder.
    Here you have the exact replica of how the Wankel engine works.
    3 piston caps = 3 rotor faces
    3 explosions per revolution of the crankshaft = 3 explosions per revolution of the rotor
    A multiplier by 3, quite unnecessary, turns the output shaft 3 times faster = the eccentric shaft turns 3 times faster than the rotor.
    It is quite comparable.
    Mazda’s 13b engine is therefore a 3.9L 2-stroke, with very low efficiency and little torque (due to the multiplier by 3).
    No miracle Felix!

Leave a 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.