The Alfa Romeo Experimental Wankel Engine Programme

Tales from futures past: the Alfa Romeo engine you’ve never heard about.

Display parts leftover from the Alfa Romeo Wankel programme (picture from the Author)

During the entire Sixties decade, the rotary engine as conceptualized by the German inventor Felix Wankel and developed by NSU became something of the auto industry’s darling: compact, light, powerful yet smooth, and made of few moving parts, it looked like the future.

No car company wanted to risk being left behind, so pretty much every automaker licensed the Wankel and NSU patents to start their own experiments with rotary engines. Including Alfa Romeo.

A few weeks ago, the retired engineer Giorgio Figliozzi was invited by the Alfa Romeo Museum’s curator, Lorenzo Ardizio, to shed light upon a piece of the Milanese marque’s history hardly anyone ever heard about: of course, Yours Truly wasn’t going to let this opportunity slip away…

Mr. Figliozzi patiently explaining the prototype engine’s features (picture from the Author)

The first contact between Alfa Romeo and NSU dates back to 1962, and the engineers from the two companies regularly exchanged information and experiences they made while tacking the Wankel’s main issue: the durability (or lack thereof) of the rotor’s apex seals. These crucial components of the Wankel engine design are subjected to extreme changes in temperature and pushed the material technology of the ’60s beyond its limits.

Mr. Figliozzi recalled that he and his colleagues tried a huge variety of possible materials, from graphite (which worked well but didn’t last nearly long enough) to even the good old cast iron (whose weight would cause the apex seals to actually ‘consume’ the stator due to their high centrifugal force).

By the late Sixties the Alfa Romeo experimental department had converted two cars to use as development ‘mules’: a Spider, fitted with a single-rotor 500cc engine good for 65 HP, and a 1750 saloon equipped with a bi-rotor 1000cc engine rated at 130 HP. In both cases the rotary engines were bolted to the normal production gearbox of the respective donor car: that 1750’s gearbox is still attached to the prototype bi-rotor engine today, as it can be seen in these pictures.

This slideshow requires JavaScript.

Neither car would survive the end of the Wankel experimental program in 1973: the experimental engines were taken away and stored, while the Spider and 1750 were crushed. At this point, it must be pointed out that Alfa Romeo was a small company and the budget allocated to the Wankel program was tiny: only a handful of engineers and two engine test rooms. Had there ever been a serious production intent? As Mr. Figliozzi plainly stated, no.

Albeit while the research program wasn’t fully cancelled until 1973, by the time Neil Armstrong had set foot on the Moon, the Alfa Wankel engine was doomed. At the time Alfa Romeo covered its production twin-cam engines with a 100,000 Km warranty, a distance its Wankel prototypes couldn’t even get close to.

The limited human and material resources Alfa Romeo could dedicate to its Wankel program became even more limited once the first anti-pollution laws were enacted in the USA: work on fuel injection and electronic engine management systems was then a far more pressing issue.

The oil crisis of late 1973 was then just the final nail in the programme’s coffin: it disappeared into the company’s archives, never to be seen again. Until now…

Author: Matteo Licata

I've been obsessed with cars for as long as I can remember, and I started drawing them the very moment I could hold a pencil in my hand. Being born in 1980s Turin, it wasn't long before I realized there were people actually drawing cars for a living: I remember I had a Giugiaro Design school diary, sometime in the late 80s, and every month there was a glossy picture of some of his concept cars: Aspid, Asgard, Capsula, Machimoto... From then on, my path was set: I was going to be a "car designer", come hell or high water, and have now been living the dream for about ten years. I've so far published two books about cars, with more to come. Find out more at roadster-life.com

19 thoughts on “The Alfa Romeo Experimental Wankel Engine Programme”

  1. Something which popped into my head in the wake of Matteo’s piece stemmed from his observation that for a time, the Wankel principle captured the imagination of the entire motor industry and was broadly seen as the powerplant of the future. It put me in mind of a similar example of group-thinking during the early ’90s when the Orbital Two-Stroke was all the rage and was also viewed as being ‘the next big thing’.

    Which leads me to a further rhetorical question: did any carmaker either ask their engineering leadership to look at Rotary and upon consideration, concluded it was not for them, or simply rejected it outright? There must have been at least one or two dissenters out there.

    1. What one easily forgets is how much money have been poured into the otto-cycle internal combustion engine, we’re talking hundreds and hundreds of billions if not trillions of dollars the last hundred years for them to reach the effectivity of today. Almost exactly a hundred years ago it wasn’t a given the otto-engine was the way to go. There were steam cars and electric cars, and their running costs were pretty much the same as the gas engine. It was only with the Ford Model T the gas engine really took over thanks to the oil companies and cheap gas travelled over seas in oil tankers and an enormous infrastructure of gas stations that had to be built. Research and Development on electric engines stood more or less still for a hundred years because they couldn’t beat the practicality and low cost of gas engines. It’s only in the last 20 years enough r&d have been put into electric cars for them to be a viable option and even just so. It’s not a mystery the industry looked and still looks at alternatives because only a small net gain would beat the game.

    2. “It put me in mind of a similar example of group-thinking during the early ’90s when the Orbital Two-Stroke was all the rage and was also viewed as being ‘the next big thing’.”

      It put me in mind of a similar example of group-thinking during the 2010’s and 20’s when hybrids and electrics were all the rage and was also viewed as being ‘the next big thing’.

    3. Time will tell, Angel, time will tell…

      But you are comparing very specific technologies – the Orbital or Wankel – with an entire form of propulsion here. A more apt comparison would be with, say, solid state batteries, which have been around five years away from production reality for as long as I can remember.

      Now, it could be that we never find a more efficient energy storage system, and today’s Lithium-Ion batteries are as good as it gets. In that case, I can see that EVs will reach their natural limit, because they will forever rely upon energy storage that is nasty to make and recycle, and has a poor weight:energy density ratio.

      I think that is fairly unlikely though… the model of an electric motor supplied by some kind of electrical energy is here to stay. The potential is just too compelling.

    4. Mazda invested lots of money and got nowhere regarding the Wankel’s fundamental deficiencies – terrible efficiency and dirty exhaust gases.
      There’s a fundamental difference between the Wankel and a battery powered car: the Wankel is an answer to a question nobody ever asked. Its only advantage from a customer perspective was relative smoothness and as soon as fuel prices went up and clean exhausts got priority its time was over. The BEV is a non-answer to a fundamental requirement every car user has – sufficient range and short recovery/refuelling tims, aka flexibility of use and therefore the BEV is not and will not be a replacement for the car as we know it.

  2. I’m not so sure, Dave.
    I write as I await my burgers delivered by a Renault Twizzy, a single-seater electric which needs only a small battery.
    That smoothness has not even now been matched by a diesel, no matter however much cleaned up.
    You’d expect metallurgy to have caught up by now to have found something to keep the rotor tips from crumbling. Wasn’t Mazda, with its clever MX8, on that pathway?

    1. The Twizy is not a car as we know it. It also has next to no range and takes just as long to recharge as any other BEV.

      There were several solutions to the apex seal problem.
      Mazda used sintered carbon seals and a hard chrome surface for the trochoid, NSU used nikasil for the trochoid and sintered steel for the seals. Audi’s Wieland pressed steel engine replaced the angular seals by a round design that didn’t scrape along the surface but rolled on it, fundamentally reducing surface wear.

  3. I’ve never driven a BEV, but had a Lexus NX Hybrid as a loan car when my Boxster was being serviced. The service took two days(!) because a slight weep from the sump required a new gasket to rectify, which wasn’t in stock st the dealership. This gave me ample time to gain experience of the Lexus as it was a 70 mile drive home, then back again the following day.

    I had high expectations of the NX, given Toyota’s long experience with hybrids and Lexus’s reputation for quality. What a disappointment! It was quiet and relaxing while driving gently, but as soon as you needed to accelerate, the refinement disappeared and the car felt gutless and strained. I didn’t bother to measure the fuel consumption, but the gauge fell suspiciously quickly over the 160 miles or so I drove it, and I wasn’t pushing hard. I absolutely hated it, and the thought of squandering £40k+ for a car with the mid-range performance and refinement of a bottom-rung supermini is too awful to contemplate.

    Maybe my experience was unrepresentative of other hybrids and I’d be very interested to hear from any of DTW’s readership that have had more or better experiences. BEVs may well be the future, once the charging infrastructure is adequate. They may never offer the conveniece and flexibility of IC engined vehicles, but we might just have to learn to live with their limitations.

    1. I’ve driven two hybrids, specifically two examples of the Prius, about ten years ago, and was reasonably impressed. They remain the most economical petrol cars I’ve driven – about 4.8 l/100 km trundling around rural France, about 5.5 or so on autoroutes. Very quiet and smooth in gentle driving, a bit less so if you called for more go.

  4. Good article on a little-known byway on the Wankel journey.

    However I’d disagree on how widespread interest was in the technology:

    “No car company wanted to risk being left behind, so pretty much every automaker licensed the Wankel and NSU patents to start their own experiments with rotary engines. Including Alfa Romeo.”

    There were some notable recusants – BMC / BLMC, Chrysler, and Fiat. I don’t think it’s a coincidence that all three (four if you separate BMC and Rover) were heavily involved in gas turbines some time before Felix’s engine became widely known.

    Greek Al was dismissive of the viability of the Wankel after just a look at the drawings, and advised BMC’s management against it. Much earlier, in the DDR, Motorradwerk Zschopau were given a head-start on Wankel development, possibly because of Walter Kaaden’s reputation on two-stroke racing engine development, but their rotary work didn’t progress long into the 1960s.

    The rotor tip seal matter is usually stated to be the reason for the Wankel engine’s failure to evolve into a viable alternative to the reciprocating engine, but the overriding weakness is the hopeless shape-shifting combustion chamber. High fuel consumption and heavy cooling loads should have been the clue.

    Long before the Wankel-mania of the early ’70s, LJKS weighed up the Ro80 engine’s advantages and disadvantages in CAR March 1968, and his scepticism was clear in his conclusion:

    “Wankels can be made to do lots of things, but none of the accompanying penalties would be bearable. In the end we are left with three indubitable advantages: smoothness, dimensional modesty, and cheapness of manufacture. In any car – if there be any car – in which these three considerations alone are paramount, the Wankel engine would have an assured future”.

    1. LJKS was wrong. A Wankel may be many things, but cheap to make it is not.
      There are large machined surfaces and the trochoid’s running surface in particular not only needs to be made to very tight tolerances because a Wankel’s geometry and kinematics are inextricably linked and the tools are costly because of the form of the troichoid. In a piston engine the cylinders can be drilled/milled/honed by machinery simply rotating around a fixed axis a trochoid needs tools following its form.
      With the Wieland ‘Blechwankel’ (tin Wankel) engine Audi found a solution for the cost problem by making the engine from stainless steel stampings and tubes welded together and a trochoid made from a strip of stainless steel bent into form.

    2. In addition to not being particularly cheap to make (low parts count, but the pieces are expensive) the wankel is more octane sensitive than piston engines, particularly wedge heads.

      Any economy engine has to be able to run on the cheapest grade of petrol.

  5. Could the resources used by other carmakers Wankel engine projects such as Alfa Romeo have been better spent elsewhere and potentially even saved a few from going under (e.g. NSU, Citroen, AMC)?

    Apparently a Flat-6 was studied as a possible alternative to the Wankel by some NSU technicians during the development of the NSU Ro80, only for NSU officials to put a stop to it since they thought it to be undesirable.

    Somewhat surprised Fiat and BMC / BLMC did not have their own Wankel projects, the same can be said on the latter not looking into the DAF Variomatic / CVT gearbox.

  6. Thanks, all, for a very enlightening discussion.
    I still think you could put an Ro80 on a used-car forecourt and say it had been made within the past 10 years — a brilliant design.

  7. Except for its stance. To my eyes both the Ro80 and the K70 looked like they were standing on tiptoes. I like a bit of ground clearance but those cars were ridiculously high in my opinion and it dates them.

    Just finished reading LJKS’s Some Unusual Engines, and of course the Wankel was one of them although not covered in detail. Mazda’s version was semi-reliable, but had to carry a thermal reactor to burn off the excess hydrocarbon emissions, as well as a catalytic converter here. As Robertas has pointed out and I believe I’ve mentioned before, that combustion chamber was just nowhere near ideal and excess HC due to rich mixture was the result.

    One other thing LJKS was bemoaning concerned supercharging: “If God had meant engines to be supercharged perhaps He would have shown us the way to computerise variable injection and timing”. He also covers some early attempts at variable lift. Well, now we have all those and balance shafts as well, so smoothness of the piston engine isn’t much of a factor for the customer.

    Exit the Wankel. Even Rolls Royce couldn’t tame that beast.

    1. W.O. didn’t worry about supercharging, and three Le Mans wins backed him up.
      Of course, the Blower Bentleys did need huge fuel tanks, but the faster they drank, the better the torque-to-weight ratio!

    2. Balance shafts were invented by Lanchester in c 1904.
      The best-known use was in the Lancia Delta, but they’ve now been largely eliminated.

    3. Balancer shafts aren’r really dead.
      Every VAG four cylinder has two of them, their V6 diesel engines have one.
      Further use was in BMW’s M43 four cylinder, in Benz M272 V6 and many more.
      Two litre Alfas with ‘modulare’ engines had them (on one of my 156s the belt driving them snapped, making for a very intersting driving experience).

      Audi’s A1 e-tron showcar used a Wieland type wankel for its range extender, the last time a Wankel engine was seen in a supposedly new car.
      The most successful Wankel surely must be the one produced by Fichtel&Sachs with membrane carburettor and petroil lubrication that made it able to run in whatever position it was held. It was an ideal engine for big (Stihl) chainsaws because its vibration free running put much less stress on forest workers. The F&S engine was also used for lawnmowers (anybody remember the Flymo?) and snow mobiles. It also was the base for the BSA-turned-Norton motorcycle engine.

  8. It is also worthy of consideration that the less moving parts an engine has,
    the less are its chances of a commercial survival – any decrease in total weight of steel components per car sold is apparently not good for the primary corners of the industry’s wider economic chain.

    One could argue that this worked especially against the two-stroke (the ideal engine from engineering point of view), and of course the subject
    of this article.

    Actually, the advent of multi-valve engines and DOHC in the late 80’s
    (which are, from an engineering perspective, inferior to a good 2-valve
    per cyl.engine, as the GM Small-block based LS/LT has proven so many times…) increased the number of steel components in the engines
    by a rather healthy amount – it actually doubled :

    -the nr. of valves
    -the nr. of springs
    -the nr. of retainers
    -the nr. of camshafts & cam-sprockets
    -the nr. of camshaft followers (buckets)
    -the nr. of cambelt tensioners

    Perhaps the above pattern is purely coincidental, but then again,
    we have to be aware of even a remote possibility that reducing
    the avg.weight of steel components per car sold might
    be a huge economic factor in the background.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

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