The Engine That Didn't Change The World
by David Finlay (18 May 2005)
Under the bonnet of the Mazda RX-8 sports car lies an extraordinary piece of work. It seems at first to be solid, three-dimensional proof of a familiar saying: "Any fool can have a complicated idea - it takes a genius to have a simple one." Mazda's name for this gem, which it has spent a lot of time developing, is RENESIS, but at heart what we are looking at here is a straightforward rotary engine, the simple idea born many years ago in the mind of Felix Wankel.
The rotary is the only thing that has come close to challenging the dominance of the normal piston engine as the obvious power unit for road-going vehicles. Compared with steam engines, electric motors and gas turbines, the rotary is a big success story. Several mainstream manufacturers have used it - Mazda is the only one to have kept the faith in recent years, but both NSU (now buried deep within the Volkswagen Group) and Citroen briefly went into production with rotary engines. Even so, the piston engine remains the popular choice among manufacturers by a colossal margin.
Piston engines work in roughly the same way as a cyclist's legs do. The piston itself is a round object which slides up and down inside a cylinder. It's attached to a connecting rod which is in turn attached to a crankshaft. As the piston moves, it turns the crank. The turning of the crank is transferred through the gearbox and differential to the wheels, and that's what makes the car move.
It doesn't matter if there is one piston on its own, or if there are four or 16, or if the explosions that push the piston down the cylinder are caused by igniting petrol, diesel, methanol, nitromethane or hydrogen, or if the cylinders are in line with each other or sprout at different angles from the crank. In all these cases, piston engines operate in basically the same way.
Since so many millions of piston engines have been built, there can't be much wrong with the principle. But if you take a look at what is actually going on, you might wonder how the idea got off the ground in the first place. For a start, those pistons do an awful lot of starting and stopping. At tickover speed - say 750rpm - they go down the cylinder, come to rest, accelerate up the cylinder again, come to rest again and restart the journey downwards more than a dozen times every second. At a cruising speed of 3000rpm, it's 50 times a second. At 7000rpm (achievable by many standard road cars) it's over 116 times.
Since the effective weight of the piston momentarily increases at every change of direction, in the same way that your body seems heavier during an emergency stop, it seems inevitable that a piston engine should shake itself to bits shortly after you switch it on. Clearly, this does not happen, but you can imagine the stresses that are created and have to be dealt with by ensuring that all the components are strong enough to cope.
Another problem with the piston engine is that it's made of so many bits. As well as the pistons, rods and crankshaft, you also need valves to open and close so that fuel can enter the cylinder and exhaust gases can leave. These valves are operated by one or more camshafts which are turned by the crankshaft. The conrods have to be made in two parts so that they can be fitted to the crank, and there are bolts to join the two parts together. And so on, endlessly.
The rotary engine addresses both these problems. The rotor itself is an equilateral triangle with outwardly-curved sides. The crankshaft comes through the middle of it, so the rotor spins around the crank and inside a circular chamber.
The spark creates the explosion which pushes the rotor on its way, carrying waste gas to the exhaust port (again valveless) before collecting another fuel/air charge. As each side of the rotor is performing one task, depending on where it is in the cycle, the others are doing their own business one-third and two-thirds of the way further round.
Single-rotor engines work well, and are far smoother than single-cylinder ones. Most modern rotaries, like the RENESIS in the Mazda RX-8, nevertheless use two. And the fact that those rotors are, along with the crankshaft, the only moving parts in the engine puts the rotary way ahead of the piston engine in terms of complexity.
What's more, there is no starting and stopping. The rotors do nothing but spin - at a maxmium of well over 9000rpm in the case of the RX-8. And there's yet another benefit: the rotary is very compact and easy to package. The RX-8 unit is placed far back in the chassis, so the centre of gravity is far nearer the centre of the car than it would be if the engine had pistons. This is very good news for whoever is looking after the suspension design, and must be a major factor in the RX-8's excellent handling.
At this point the rotary seems to be standing over the piston engine with a foot on its chest and a gun pointed at its forehead. But the RX-8 is the only rotary car you can buy in the UK. Surely a brilliant engine like this should be more widely available?
Well, it has enough problems to have frightened off most manufacturers. For a start, although the rotary has hardly any components, the ones it does have are very difficult and expensive to make, and they have to be absolutely right. In particular, there must be an efficient seal between the tips of the rotors and the inside of the chamber. If the seal isn't good enough, the engine won't work, and it was a lack of reliability in this area that killed the otherwise brilliant NSU Ro80 of the early 1970s.
Even if the sealing is absolutely right, the basic principle of the rotary creates some difficulties. Whereas a single cylinder breathes in petrol and only later puffs out exhausts gases, a single rotor is constantly doing both, which means high fuel consumption and an awful lot of emissions.
A lot of noise, too, if you're not careful. The Mazda RENESIS unit is well muffled, but sporting rotaries can make quite a racket, and some of them are ear-splitting. The reason is that, as with two-stroke engines, one of the best ways of tuning a rotary is to fit a megaphone exhaust. This doesn't just fail to reduce engine noise, it actually magnifies it, with spectaular results. The beautiful little 1960s single-rotor NSU Spyder racer, which was demonstrated at Donington a couple of years ago, belied its appearance by being shatteringly loud, while the Mazda RX-7s which dominated the British Touring Car Championship in the late 1970s were even worse.
Since the rotary has a competition history (which includes Mazda winning the Le Mans 24 Hour race in the 787B sports car) it's clear that the engine can produce a lot of power. But it does this because it revs so easily. As explained in our feature, an engine's power equals its torque multiplied by its speed. Rotaries spin very quickly, but they don't produce much torque. The explosion of the fuel/air mixture skims along the top of the rotor rather than pushing heavily on to a piston. If the piston engine's operation is like the action of a cyclists' legs, the rotary does the equivalent of pushing the tyres with the flat of your hand.
Rotary engines are in this respect the exact opposite of turbo diesels. Their ability to rev isn't merely useful - it's vital if they are to produce decent amounts of power. The Mazda RX-8 has decent mid-range performance, but that's despite the basic characteristics of the rotary rather than because of them. Even then, the RX-8 really needs to be taken beyond 6000rpm before it starts to perform properly.
I can't have been more than about ten years old when I first learned about Felix Wankel and his rotary engines, and I've been fascinated by them ever since. I'm delighted that Mazda has kept the faith and brought the rotary into the 21st century with the RX-8. But even I have to admit that the rotary's place in the car industry is that of a low-volume novelty. For all its faults and crudeness, the piston engine is still the boss, and justifiably so.
