Eighty-five years on from the night when Dr Felix Wankel allegedly dreamed of piloting his own car powered by an engine that was half turbine and half reciprocating, the rotary engine has conspicuously failed to evolve as quickly or successfully as its piston-equipped counterparts. But the rotary engine has inherent attributes that could make it eminently suitable for running on hydrogen. Mazda is seriously looking into it. Jesse Crosse reports.

After a brief flirtation by NSU/Wankel in the 1960s with the NSU Ro80, only Mazda was prepared to battle on against the increasingly insurmountable odds of wear, high consumption and emissions that came with the territory. The last serious commercial foray with the RX7 in the 1980s was eventually defeated on the last two counts, but in 1995, Mazda resurrected its rotary concept with the Renesis engine. In 2003 the Rotary made a commercial return to niche markets in the new Mazda RX8 with emissions, wear and consumption problems apparently under control once and for all. And now, as the automotive industry moves into its third Millennium, Mazda's persistence could be about to pay off. Because by all accounts it seems the rotary engine has inherent attributes that make it eminently suitable for running on hydrogen.

Despite Ford's own activities in the field of hydrogen-propulsion, such as the alliance with DaimlerChrysler and Ballard and its own fuel cell and H2ICE programmes, Mazda has been able to pursue its own hydrogen ambition. Senior research engineer on the Hydrogen RE project, Masanori Misumi, emphasises how serious Mazda is about its alternative fuel programmes. "Since 1990", he says, "we have put more priority on hydrogen energy, conducting considerable research on hydrogen projects including hydrogen storage systems." Early attempts at developing a hydrogen rotary engine were fairly rudimentary, explains Misumi, because there were no bespoke hydrogen injectors available. "In 1991, the hydrogen was metered by a needle valve upstream from the engine, then a poppet valve in the engine itself let the hydrogen in, but fuel metering was poor."

The rapid growth of compressed natural gas (CNG) engines for commercial use during the last decade spawned the development of electronically controlled gas injectors and were exactly what the hydrogen men had been waiting for. At last, it became possible to meter the ingestion of gas into the engine accurately. In the case of the rotary engine, the conversion was relatively simple compared to a reciprocating ICE. "We put the injector in the top of the cylinder and we are proud of the fact that no other modifications were necessary - just the integration of this injector," recalls Misumi. The physical dimensions of the rotary means that unlike a reciprocating engine there's plenty of room and the RE has two hydrogen injectors delivering gaseous fuel to its combustion chamber. Using two helps to overcome the fact that hydrogen has a lower volumetric energy density than gasoline and so a much greater volume of the gas must be injected. Once injected though, there's more time for the hydrogen to be thoroughly mixed with air because the rotary engine's output shaft turns through 270 degrees in one intake cycle compared to 180 degrees for a reciprocating engine. Mixing not only occurs over a longer period, but the vigorous intake flow ensures a more uniform mixture of hydrogen and air, something that's essential for satisfactory hydrogen combustion. The Renesis is also unique in being able to run as a dual-fuel hydrogen-gasoline engine, an important feature because packaging is at a premium in the RX8 and its small hydrogen tank only gives a gaseous range of 50-60km.

But when it comes to hydrogen combustion, the rotary enjoys a much more fundamental advantage and as Misumi points out, "its intake and combustion strokes are completely separate". In each cylinder of a reciprocating engine, induction, compression, power and exhaust events all take place in the same physical space, the type of stroke that is occurring being controlled by inlet and exhaust valves, the fuel injection system and the ignition timing. When the new charge is inducted, it arrives into a hot environment and unless the mixture is set lean backfiring can occur. This contributes to the fact, says Misumi, that "output of a normally aspirated hydrogen reciprocating engine is 30 percent to 40 percent of a gasoline equivalent." Early versions of the rotary ran close to the same 'stoichiometric' chemically correct mixture setting as a conventional gasoline engine and used a conventional three-way catalytic converter. This was made possible by the cooler environment of the separate inlet chamber into which the fuel is being induced. The lower temperatures also mean injectors can be fitted with rubber seals to prevent hydrogen leakage. Consequently, its output was much higher than usual for a hydrogen ICE, the rotary reaching almost 90 percent of the power it would produce running on gasoline. However, to achieve lower NOx emissions it was necessary to settle for a halfway house. Stochiometric mixture is also known as 'Lambda 1'. The Mazda engineers found that with a leaner mixture of Lambda 1.8, NOx levels fell close to zero at 2ppm but the trade-off was a drop in power from 210PS to 110PS.

To compensate, the next generation will have both a turbocharger and hybrid drive. The turbocharger will draw on technology being developed for diesel engines, its turbine electrically-assisted at lower engine speeds (from 1,000rpm) when exhaust gasses that normally drive it are flowing more slowly. A 10kW mild hybrid drive motor in line with the engine's crankshaft will provide 10kW (13.6PS) of additional power. The target, says Misumi, is to achieve the same output as a gasoline equivalent, the cost ending up at the equivalent or slightly less than a full hybrid system. A 144-volt battery provides electrical storage and powers the stop-start system which kills the engine whenever the car comes to a halt to reduce consumption and emissions, then starts it again when the driver presses the accelerator. The electric hybrid motor draws power from the battery during low speed acceleration to boost engine power. Then, when the driver lifts the accelerator pedal, the motor becomes a generator, the force required to drive it helping to slow the car and the electricity produced being used to charge the battery in a process called 'regenerative braking.'

The use of compressed hydrogen gas is becoming a more popular alternative than cryogenic storage of liquid hydrogen because refuelling is a safer, more simple process not unlike filling a car with LPG. In contrast, the problems associated with members of the public handling liquid gas delivered at -253 degrees Centigrade are greater and may only be possible using robotic filling rigs. Liquefied gas also has the disadvantage that it evaporates or 'boils off' over time as the storage flasks inevitably warm up. The downside of compressed gas is unease over the storage pressures (despite full US and European safety certification) and the fact that at 350bar (5,000psi) the energy density by volume is lower than that of gasoline, restricting range. In common with Ford fuel cell vehicles the RX8 stores its gas in a Dynetek, alloy-lined composite tank at 350bar (5,000psi). A number of manufacturers are working on similar high pressure hydrogen storage devices drawing on experience gained in the more conventional use of CNG (compressed natural gas). 700bar tanks are under development and the energy they store at the higher pressure is claimed to be much closer to that of gasoline occupying the same volume.

Mazda is no beginner in the field of hydrogen propulsion. It developed the HR-X , a hydrogen engine study, in 1991, the HR-X2 rotary together with a hydrogen powered MX5 roadster followed in 1993 and the rotary-powered Capella Cargo in 1995. In 1997 it showed its Demio fuel cell car for the first time, with the Premacy FC-EV following in 2001 and the latest idea, the RX-8 Hydrogen RE breaking cover at the Tokyo Show last October. During the course of these projects Mazda explored various means of storing hydrogen. The Premacy FCEV carried an on-board methanol reformer and became the first of its type to be both publicly tested and government approved in Japan. In 1995, the Capella carried a metal hydride hydrogen storage device were the gas is literally absorbed into the material and then released on demand. Misumi says energy density using this approach was good and bettered that of liquid gas. But the weight penalty, some 400kg, has so far proved insurmountable. Hydrides are also slow to fill. Substantial heat is generated during the filling and release of hydrogen and they are costly.

Other companies are seriously pursuing the development of hydrogen ICE vehicles, notably Ford, which is using the global I4, 2-litre reciprocating engine as a basis in conjunction with Lysholm screw-type superchargers. BMW, having publicly eschewed an expensive fuel cell propulsion programme that in any case would not sit comfortably with a brand that occupies the prestige performance end of the market, also has an active hydrogen programme including both V12 7-Series Cars and the MINI. Overall, the outcome for hydrogen combustion engines looks promising and if they ever proliferate, will contribute to solving the biggest problem which is not the development of clean, hydrogen-consuming propulsion technology, but developing a hydrogen economy to support them.

Jesse Crosse