Last week when we talked to Professor Harry Watson we concentrated on alternative fuels, which with the exception of LPG, it's fair to say are still a long way off from being widely adopted. This week we look at the development potential left in the petrol-fuelled internal combustion engine - are there big gains still possible with this technology? And what about water injection and oxygen enrichment technologies?
Firstly, Professor Watson is a fan of turbocharging.
"We had a turbocharger project with Holden in the middle Eighties for the 2-litre Camira. We took their 1.3-litre Opel engine and turbocharged it for fuel efficiency. We did a lot to manifold design, camshaft design, combustion chamber design and so on. It had identical power output and torque to the Camira engine but under steady-state driving conditions the turbo engine was 40 per cent more fuel-efficient than the larger 2-litre engine. In around-town driving conditions it was about 20 per cent better."
Professor Watson also believes that direct injection - where the fuel is injected into the combustion chamber under high pressure - has potential.
"Gasoline direct injection - or stratified charge spark ignition engines - which Mitsubishi has been producing for the Japanese and European markets now for five or six years has the potential to deliver [efficiency] improvements. There are, however, some issues.
"Much of the testing originally done for the Japanese market was over their rather lemon-type drive cycle, where you have to have an average speed of 15 km/h. You get these [efficiency] gains under those circumstances but when you have a need for higher acceleration - which we're expecting - then there may be some loss from the claimed 18 per cent gains back to about 10 per cent. But that's certainly a big opportunity."
But it's in the optimisation of existing technologies where Professor Watson sees the biggest potentials for efficiency gains.
"One of the interesting studies that we have been doing in our research work over the last three or four years is to develop [engine] optimising technologies. These are based on computer models of the engine. We've taken a range of optimising techniques, starting with neural networks - which are used quite a lot in some areas of optimisation in manufacturing processes - and now we have a method called 'swarm particle tracking'.
"For the current Ford Falcon engine we have 12 variables [in the model] which can all be simultaneously changed within limits which have been calibrated. For example, compression ratio (typically we put the limit at 16:1 on compression ratio), timing changes, the ability to model cam timing, cam phasing, cam duration, combustion chamber shape, engine speed. So there's a whole list of variables, including more obscure things like exhaust gas recirculation.
"What we're finding through this modelling work is that gains of much more than 10 per cent are possible. I think that this bodes well for significant improvements in efficiency that are going to come from optimisation of the engine. Firstly, perhaps before the engine goes into production.
"But the latest stage is real-time self-optimisation with learning strategies. Our ultimate vision for the long-term future is an engine which effectively is put together with some of the mechanical things having limitations but thereafter it will search for its own operating conditions to meet particular requirements of emissions standards and so on. That's a very long-term vision; probably a 30-year work program.
"This technology is particularly applicable when we have hydrogen-assisted ignition process where we can alter combustion duration - an extra variable, something that you can't normally control.
"What we have seen is that the numbers that come out of a simulation for the present Falcon engine show that the cam phasing is very close to what they actually use. But quite surprisingly, as you increase the power output, it seems to indicate that you go to other strategies of cam phasing which are not presently in the Falcon. So in fact instead of running very late inlet valve closing, run it very early. That way you actually trap more of the residual gas from the previous cycle and you get the same reduction in pumping work but now you get some extra stuff in there to help the emissions side of the equation.
"If we think of what's happened to the computer over the last 20 years, then it's obviously only a matter of time before such sophistication [of optimisation] can exist."
One of the other areas we asked Professor Watson about was technologies that have proved very strong in the past but are now currently out of fashion in high performance applications. What sort of technologies, then? Try water injection.
"The proof of the technology is that when performance was of paramount importance in [World War II] fighter aircraft, it was employed - and quite successfully. What it effectively does is reduce the combustion charge temperature through the evaporation of the water. We know that's one way of avoiding detonation; we know that's one way of getting more charge into the combustion chamber to get more power out of the engine. That side of it stacks up very well.
"How you evaporate the water is quite critical - getting a good break-up of the water spray in order that it vaporises very quickly. That's as much of an issue as it is for delivering petrol through a petrol injector, except - and if I can use the word - the 'evaporability' of the water is not as good as that of petrol.
"But the major issue for long-term use of water injection is that the water tends to find its way into the lubricating oil and you have diluted lubricant. In fact in my experience, engines where the lube oil suddenly starts to run out of the filler because there's so much water contamination! So unless it is well managed and there are some cycles that boil the water out of the oil, there are some problems that lead to bearing failure and corrosion. I think that is probably the reason that it is not being played with on any performance applications.
"But I put to you that with the continuing development of metallurgy and possibility of having better sensors, cycling the lube oil, controlling temperature and so on, that water injection is an area that could reasonably be revisited.
"I might just add that in one of our hydrogen cars where we needed water injection for other reasons - that being to keep the combustion chamber surfaces cool to prevent the hydrogen self-igniting - we found that a dribble feed water injection system was the best outcome, quite different from the one that we're talking about here.
"We put an exhaust T-piece [on that car] so that some of the exhaust passed through a condenser, [allowing us] to condense the water from the exhaust. There might not be any need to have a water refuelling process as part of water injection. The concept of this is that it is a built-in system that works without the driver realising that he has the benefit of such technology.
"Even ordinary [petrol engine] exhaust has got 8 per cent water!"
Another 'combustion enhancer' is the boosting of the oxygen content of the combustion air, recently covered in an AutoSpeed story ["The Latest in Intakes"]. The Caterpillar company has a patent on one version of the process, using semi-permeable membranes. Professor Watson has also worked in this area.
"We hold a patent that covers the application of semi-permeable membranes and turbochargers to any sort of engine. The benefits of using these membranes is that you can produce on board the vehicle oxygen-enriched streams. On our demonstration engine we achieved 35 per cent oxygen compared to the normal 21 per cent in air. That [configuration] used twin turbos in a V8 to provide that level of oxygen. That was a diesel engine but we have also done work with a spark-ignition engine equivalent, but not in an engine where it provided its own oxygen.
"There are some significant benefits in both engines of using this type of technology. If you are using turbocharging you have to push nitrogen and oxygen into the cylinder in the usual proportions [contained within air] in order to get the oxygen. If you just push in extra oxygen - which is possible by this means - then the compression pressures and temperatures throughout the engine cycle are reduced. This means where you have a stationary diesel engine producing maybe 16 megawatts you can actually reduce the exhaust temperature considerably for the same amount of output, or you can push up your output for the same exhaust temperature.
"There has to be a downside, and that is that the extra oxygen combines more readily with the nitrogen that remains and so you get increased oxides of nitrogen [emissions]. In order to sell this into a market that's getting ever more keen about oxides of nitrogen levels - even for stationary engines - you need to manage the control very well so that you can find a region where you're getting all of the benefits from the oxygen boosting but there's not too much oxygen available so that you get the high oxides of nitrogen. In fact water injection works extraordinarily well in that role - so combining the two technologies is a very possible solution.
"The membrane technology has come on a lot in the last ten years but it would be fair to say that the membrane that we have on our diesel engine is probably about three times the size of the normal airbox of the diesel engine. But the efficiency of these membranes - in terms of the amount of volume that you need for a given amount of oxygen to be produced - has improved quite a lot.
"It is important to recognise that in addition to the turbo and the membrane you need to have an auxiliary fan. You need to make sure that the fibres of the membrane always see fresh air, so you have to be continually drawing air across the front of the membrane. In a high velocity vehicle that might be provided naturally, but at low speed you need to have that additional item."
|