Development of the modern racing two-stroke engine is in a pitiful state at the moment; the hamster is dead, but the wheel is still going round. The two-stroke is the finest piece of kinetic art, combining the highest principles of mechanical, thermodynamic, and acoustic engineering, and it’s being done to death in a political putsch which has nothing to do with sporting competition and everything to do with money-grubbing expediency. As the knife is plunged into the type of engine which has made GP bike racing the most thrilling sport in the world for a generation, we should remember that though it has been around for 120 years or so, it's being cut down in the prime of life, not advanced old age, as the lap times from the IRTA tests are showing us. The two-stroke's peculiar good and bad points - light weight, small size, high power output, mechanical simplicity, low cost, high fuel consumption, low internal friction, poor emissions, lack of engine braking - make it just perfect for a racing bike (and not much else) so its development has been driven solely by the needs of bike racers for a century.
The leading pioneer was Alfred Scott, 1874-1923. In four successive TTs before the First World War, Scott motorcycles managed four lap records and two successive wins. They would probably have managed a hat-trick had not a magneto failed in the final stages of the last race.
Mr Scott didn't think of his bikes as racers, but they were certainly not cheap, aimed at being the ultimate Gentleman's Carriage – a Super Blackbird of the day, and he was well aware of the good publicity which TT success would bring. The Scott had a well-made chassis noted for its good handling, and the engine bristled with ideas that were well ahead of their time – it was a water-cooled parallel twin with an oscillating inlet valve doing the same job as a disc valve on a modern Aprilia, which made it much more tractable than the piston-ported screamers that would follow sixty years later. If you put something like a Scott Flying Squirrel next to a Yamaha RD350LC, there will be more similarities than differences. A rosy future was predicted for the bikes, which were made in Shipley, Yorkshire, ee bah gum tha knows, far away from the established centre of light engineering in the West Midlands.
Unfortunately, the TT successes made little impact on the motorcycling public, which soon had other things on its mind - like four years of the most hellish conflict in the history of mankind. Scotts concentrated on making three-wheeler machine-gun carriages through the war, and the solo bike business never quite picked up again afterwards. Though post-war Scotts had many advanced features, they were let down by some basic simple faults (like a poor crank design which made the bottom end weak) which the paying punters would never forgive them for. Hence, the TT success was never followed by massive sales.
Although two-strokes continued to be manufactured in large numbers, they were not considered as high-performance machines in this country. The world leader in racing two-strokes was shrouded in mystery until the 1960s, because it was behind the Iron Curtain in East Germany. The DKW company was an offshoot of the mighty Auto-Union conglomerate, a powerful association of German automotive industries which was one of two excellent ideas to come from the most influential person of the twentieth century, Adolf Hitler (the only other good thing he ever did was to shoot himself). The DKW factory churned out very well put-together two-stroke bikes - so good that, after the war, the Brits nicked one of their designs and turned it into the BSA Bantam. Just pause for a moment and imagine an utterly broken and defeated Germany, where anything could just be taken away for free - the Americans took a team of space-rocket scientists, the Russians took half the country, we took drawings of a 125cc moped.
As well as having the bikes and even two-stroke cars in full production, DKW also ran their own Grand Prix bike team. The team manager behind the racing efforts was Walter Kaaden, one of those natural Top Bananas whose tireless work and enthusiasm motivate the whole team around them. Deeks ploughed a lonely furrow as the only two-stroke GP team worth mentioning, which always makes development difficult because you don't have anyone to learn from, copy off, or compete directly against, but they nevertheless produced some startlingly fast (and noisy) bikes.
The Germans really had all the good ideas at this stage of the game; the next revolution came from a Dr Schneurle, who first took out a patent pertaining to the scavenging of two-stroke engines by what he called the "loop" process in 1925. All two-stroke racing engines of the modern era have relied on this system. Essentially, it means that instead of blowing fresh charge through ports in one side of the cylinder and getting the exhaust gases to waft out of the other side (like the Scott) or feeding fresh charge into the bottom of the cylinder and letting the exhaust out of valves in the head (common on big industrial diesel engines), the Schneurle system fires mixture out of transfer ports to one part of the cylinder wall opposite the exhaust, where it goes up towards the head (aided by the rising piston) and then comes down the other side to complete the loop before jumping down the exhaust port as the falling piston uncovers it. This loop-scavenge system was the last word in piston-ported engine design, because no significant changes have been made to it for decades.
However, the fundamental problem with any piston-ported engine is the symmetrical timing. Any port controlled by the piston is going to close the same number of degrees before TDC as it opens after TDC. You can make the timing longer or shorter by moving the port up and down the barrel, but the timing will still be symmetrical about TDC. Many ideas were tried to get around this problem (like the early DKW one of having two pistons sharing the same combustion chamber but slightly out of phase) but none was entirely satisfactory until the tireless Walter Kaaden perfected three excellent ideas – two for the inlet side, and one for the exhaust.
Asymmetrical exhaust port timing has been a racing two-stroke Holy Grail for as long as there have been piston-ported two-strokes. Using poppet valves for the exhaust of a racing engine was no good, because they couldn't be made to move quickly enough (twice as quickly as on a four-stroke, of course) and the flow characteristics were nowhere near as good as with a piston port. Any sort of rotary or reed valve was out of the question because it would just get fried by the hot exhaust gas. A racing engine needs a piston port, but it also wants asymmetrical timing. How could they both be achieved?
The biggest flash of genius in the history of two-stroke development was making the exhaust port timing asymmetrical not by using a mechanical valve, but by using the wave action of the exhaust gas itself, to create a plugging effect. The credit for this idea – on which the whole success of two-stroke racing engines is founded - goes to a Russian engineer, Michel Kadenacy, who pioneered the use of exhaust pipe geometry to help with engine scavenging processes on both two and four stroke engines. Any exhaust ramming or boosting is now known as the "Kadenacy Effect" in his honour. Walter Kaaden relied heavily on Kadenacy's help in developing the DKWs, and all the work at Queen's University, Belfast has been related to exploiting the effect pioneered by this farsighted Russian, whose genius should be remembered every time we see the classic two-stroke expansion chamber layout – header, divergent cone, central part, convergent cone, and tailpipe.
After Walter Kaaden and Michel Kadenacy had shown the way, many other engineers and tuners put huge efforts into developing exhaust pipe effects, which probably remain the single most important aspect of two-stroke race tuning. Great advances were made by Dr (now Professor Emeritus) Gordon P Blair of Queen's University, Belfast. Dr Blair was the first to exploit the fact that the tedious and repetitious calculations of wave actions required to analyse exhaust systems, called "nets", were ideally suited to the crude digital computers then just becoming available. He went on to become the world authority on the design of performance two-strokes by using computational techniques rather than riffler files.
I have met Prof Blair briefly, and am pleased to report that he is every person's impression of their favourite uncle and a mad professor all in one, wearing a tweed jacket with leather patches on the elbows, a pipe in one pocket and a tobacco pouch in the other, scattering Swan Vestas around the place and always talking with infectious enthusiasm about racing bikes. What a great bloke; and it's good to know that his work is now being continued by another great bloke and bike fan, Professor Robert Fleck, who has done so much for QUB-backed bike racing teams over many years.
Once engineers felt that they had the exhaust port behaviour under proper control, attention turned to inlets. It's strange that simple piston-ported inlets with their symmetrical timing lasted so long, reaching their apogee in 1978 when Mick Grant was riding a Kawasaki KR750 around the Isle of Man. I say that it's strange because Deeks were running inlet systems with both reed valves and disc valves with some considerable success before the war. Thus, all the ingredients of the latest engines had been built and tested before 1940; they just needed matching up together correctly in a single package, and refining. As the power outputs rose, so the cooling problems increased, until water-cooling became universal in the 1970s.
As far as I can tell, the first package with all the bits in place ever seen west of the commie bloc (a British water-cooled disc-valve reverse-cylinder two-stroke) was designed and built by my dad in his back shed, in 1957. Housed in a Greeves Silverstone chassis, it won too many clubby races to remember, and finished second in a national race – something which the Old Man still kicks himself about, because the only way the dozy pillock avoided winning was by falling asleep whilst in an easy lead and then allowing himself to be pipped at the post. The chassis of this bike was sold years ago, but he still has the engine.
The way that the world lead in two-stroke racing engines passed from the East Germans to the Japanese is the most amazing story, real Michael Caine stuff with men in long coats whispering "Red cows over Moscow" in remote forests infested with KGB thugs. The central character was Ernst Degner, a works DKW grand prix racer who was also the star development engineer. It would have been relatively easy for him to defect to the West at a GP meeting, but since he was married that option wasn't open to him because the Stasi always made sure they kept tabs on his wife whenever he was abroad. His only chance was to be smuggled over the border in the boot of a car, choosing a remote area where there was a good chance that the guards would be a bit dozy and not too diligent in their searches - nor too trigger happy. Once into the free world Ernst and his skills were snapped up by the works Suzuki GP team. The combination of Ernst's knowledge and Suzuki's hard work and financial backing was a winner.
Mr Degner's defection and Prof Blair's work fortunately coincided with a great revival of interest in two-strokes as high-performance racers rather than utility putt-putts. The resurgence was basically caused by Yamaha and Suzuki trying to out-flank the mighty Honda. In the 1960s, new factory teams had no chance of entering four-stroke GP racing with any hope of success, because they'd be up against years of experience built up by two massive industrial empires – Honda, and Agusta. To move in with a different type of engine was a chance to manage a sneaky head-start. Mr Honda had seen the writing on the wall and bottled out in plenty of time instead of staying around to be beaten by stink-wheels. It was a pragmatic decision, which Mr Honda justified quite well by pointing out that his company saw no reason for making production two-stroke engines, so what was the point in racing them?
This set the scene in development terms pretty much to the present day. Nothing has been seen on racing two-strokes recently that wasn't tried by Deeks all those years ago, with the two exceptions of power valves and electronic fuel injection.
The Yamaha power valve, a sort of rotating barrel with a cut-out which effectively altered the height of the exhaust port according to engine revs, was a huge leap forward in development terms, and all other two-stroke producers were more or less forced to copy it – but they had to do so very carefully to avoid infringing Yamaha's patent rights. Suzuki had their adjustable ramp (twin stage on the RGV), Honda had the ATAC system, Rotax and Aprilia used a simple sort of sluice gate, Kawasaki had their KIPS system, and so on. These systems were really only a refinement of the simple piston port, because although they could vary the timing, they couldn't make it asymmetrical. Therefore, although peak power wasn't enhanced much, the engines became much less peaky and the bikes therefore more easy to ride.
Peakiness was the problem with the electronically fuel injected engine, according to Mick Doohan who called the shots with HRC at the time. The EFI two-stroke was probably axed a bit too quickly as a result of Mr Doohan's opposition; a more far-sighted development rider would have persisted with it much longer, but Honda was content to abandon it for political reasons since other teams were moaning that EFI would be too complex and expensive (but prototypical four-strokes are all right, though).
Electronically injected two-stroke engines are fiendishly difficult to sort properly (enough to bankrupt Bimota) and therefore have been raced only twice – by Honda Racing Corporation, and me. I was first. My converted RS125 was debuted a few weeks before Shinichi Ito was first seen in public on the injected NSR500. I finished about 10th in a clubby meeting at Brands Hatch, and abandoned the project soon afterwards in a Doohan-esque fit of impatience, though with hindsight I can see that I should have continued. I wasn't impressed with it at the time, but the poor result at Brands might have been because it was my first visit there (complete with orange novice bib) and also because I had a bad cold, rather than lack of power from the engine. The way my injection worked wasn't very smooth, but it was about a thousand percent better than the Bimota V-Due which was released several years later, though of course I wasn't to know that at the time. What really finished me off was Dr Robert Fleck of QUB very kindly giving me a copy of Honda's technical paper on their system, which was so far ahead of anything I could do in my back shed that I went back to fiddling with my 1929 KTT Velocette, in the sure knowledge that no Honda was going to beat me in the Historic Racing Championship for pre-1934 bikes.
The Velocette company, Birmingham’s finest, was no slouch when it came to two-strokes, with the 250cc GTP of the 1930s being a fine performer. Marketed by Velo as a no-frills commuter, owners found that the light, nimble machine would keep up with the four-stroke superbikes of the day if thrashed mercilessly, and early models were bedevilled with seizing problems as a result. Velo eventually cured this with a landmark development – the idea of a special lathe to turn the pistons slightly oval as well as tapered, so that they'd end up perfectly true and round when distorted by the heat. This feature has been incorporated on pretty much every piston since.
Velocette also contributed to two-stroke development many years later, by reviving the reed-valve on their Viceroy scooter in about 1965, which had a very advanced (for the day) 250cc two-stroke flat-twin engine. The scoot was supposed to be an up-market model (a Silver Wing or Yamaha T-Max of the day) but it was far ahead of its time, and the buying public couldn't be persuaded to buy a scooter for the price of a bike. The Viceroy died a death, but not before the Japanese had managed a good look at the way the reed valves discharged directly into the crankcases.