After my article a few days ago ‘The Sportbike Engine of the Future: Camshaft Design‘, a few readers emailed me asking if I had forgotten about a few other interesting valvetrain-related systems currently being used or developed. Of course I hadn’t forgotten about anything, I simply felt that these other systems were complicated enough to warrant their own article.
Let’s start with pneumatic valve springs. Many people, having heard that Formula 1 cars use “pneumatic valve actuation” (not an accurate description, by the way), may think that the valves on F1 cars are both opened and closed by air pressure (pneumatically). In fact, these engines still use traditional camshafts, the only difference being that the valves are closed pneumatically rather than by spring pressure.
In any four-stroke engine, the valves must be pushed open for the required length of time, and then pushed shut before they get themselves smashed by the rising piston. The opening is usually accomplished by pressure on the top of the valve from the camshaft lobe (or a rocker arm actuated by the camshaft lobe), while the closing has traditionally been accomplished by a spring (or sometimes a pair of springs, a small one inside a larger one), pushing upwards on a ‘retainer’ which is clipped to the top of the valve.
The traditional spring-closing system has a few problems, all of which can be eliminated by pneumatic closing systems. First, the higher the engine is revved, the harder and faster the cam lobe ‘smacks’ the valve down – meaning that higher RPM engines require stiffer springs to ensure that the valve is forced back to the closed position equally quickly. Stiffer springs create more drag on the engine, as the spring pressure resists against the cam lobe’s attempt to force the valve open. At REALLY high RPM, even a very stiff spring can be unable to close the valve quickly enough, leading to valve to piston contact, which is reliably reported to be extremely unhealthy for any engine.
Another problem is that springs wear out. Over time, a spring being compressed and released thousands of times per minute will gradually lose its tension and thus its ability to close the valve rapidly. Again, the result will be valve-piston contact.
Pneumatic valve closing systems simply replace the spring with a very small cylinder containing compressed gas (typically nitrogen, as it is less temperature-sensitive than oxygen), which is kept at a pressure of around 100psi. This cylinder pushes the valve back up in the same means that a conventional spring does, but without some of the drawbacks of traditional valve springs. The system is not particularly high-tech, and has been in use in F1 since the mid-1980s.
How useful would pneumatic valves be on a motorcycle? Not very. First of all, even the highly developed systems used in F1 have some amount of gas leakage from each cylinder, and thus each car must carry a reservoir of pressurized nitrogen which constantly refills the cylinders. If one should run out of nitrogen, the cylinder stops closing the valve – you know what happens then. The requirement to carry an external reservoir, and the consequences should it go empty, make this type of system more useful on race vehicles than street machines – and anyway it would be difficult to find space to stash a large reservoir on today’s tiny sportbikes (it would also add unneeded weight).
Pneumatic valve closing wouldn’t do much for bikes, anyway. The reason F1 cars use it is to allow higher RPM, which in a race motor usually translates to more peak power. However, as I discussed in the ‘Camshaft Design’ article, producing more power at higher RPM requires a camshaft optimized to perform at those RPM, which means trade-offs in low- and mid-range power and torque. Thus sportbikes intended for street use would achieve little by switching to pneumatic valve closing, even if engineers did find an efficient way to apply it to motorcycles.
A more interesting technology is that of hydraulic valve actuation. Several companies have developed true ‘camless’ systems, in which each valve is opened and closed by a hydraulic actuator entirely controlled by the ECU. These systems have been used at revs as high as 15,000 RPM, and continuing advances mean that 20,000 RPM could be possible soon.
Entirely computer-controlled valve actuation would bring a host of benefits to any engine, as it would mean that a virtually infinite variety of ‘profiles’ could be used, each optimized for the specific conditions (RPM, load, etc.). Valve lift, duration, and opening/closing speed could be varied to a virtually unlimited extent.
Why hasn’t this been applied to motorcycles (and cars) already? The number one reason is cost. Being a relatively new technology, the components are of course still fairly expensive. In addition, programming the ECU to take full advantage of the available options (as far as varying lift, duration, etc.) would require an incredible amount of time and money to be spent in research and development – basically, the manufacturer would have to test thousands, or even millions, of ‘virtual’ cam profiles, under thousands of varied combinations of conditions (RPM and load). Controlling this system would also require an ECU with computing power far beyond that of anything currently found in a production motorcycle.
Another problem arrises when one considers the power source for this system – a hydraulic pump. The pump would have to be driven by the crankshaft (sapping power), and if it failed, the consequences could be disastrous. It would again add weight and take up space, both valuable commodities on a modern sportbike.
Despite all these drawbacks, it is likely we will eventually see hydraulic valve actuation on a production motorcycle – just not for 5 or 10 years. Once some of the potential pitfalls are engineered out, it is likely that the manufacturers will test these systems both in the lab, and on the track (probably in MotoGP), before bringing them to production.
