As I see it, one of the biggest limiting factors in engine design today is the fact that they have to drive the wheels directly through a mechanical gearbox. By being directly connected to the road wheels, potential engine design is severely restricted to the following:
1) It needs to have good torque at low engine speeds.
2) It has to run smoothly at idle.
3) It has to have a large amount of power in reserve (for accelerating, overtaking etc).
4) It has to respond quickly to changes in throttle pressure.
By doing this, engineers have argued that efficiency losses are minimised, as the energy remains mechanical straight to the wheels.
But the way I see it is that it also severely restricts engine design. So, to cut a long story short, here's the main premise of my post:
* I reckon that by removing these design limitations, we could design internal combustion engines with efficiencies close to or attaining 60% thermal efficiency
double that of a typical petrol engine).
Impossible? I'm going to argue not. Here's what I propose....
1) The engine should only be present as a range-extender for plug-in hybrids. The drivetrain should be entirely electric allowing regenerative braking etc and all motive energy comes from a large storage battery or capacitor charged up from the grid overnight.
2) Only once close to the electric range of the vehicle, the engine should start up to charge the battery NOT TO POWER THE WHEELS DIRECTLY.
3) The engine should be small, only 70bhp or so, sufficient for a top speed of 110mph.
4) Maximum power will only be limited by the battery and motor, so acceleration will NOT depend on ICE engine power, as this is transient and can be buffered by the battery. Hence small ICE need not mean slow acceleration (though full power would only be available for several minutes before being reduced back to 70hp).
Driving only a generator, the engine can therefore be designed to operate any way we like. To hit peak thermal efficiency, I would suggest building a compression-ignition (diesel) Atkinson/Miller cycle engine.
Peak thermal efficiency of a standard (10:1 expansion ratio) petrol engine is at best 25-30%
Peak thermal efficiency of a Prius (13:1 expansion ratio) petrol engine is at best 36%
Peak thermal efficiency of the best diesel engine (20:1 expansion ratio) in a road car is at best 45%
Peak thermal efficiency of big marine diesel engines can hit 50%
Therefore, why not build an engine with a hardware expansion ratio of 35:1, but a Miller cycle computer controlled inlet valve system (as used in the Prius) to lower compression ratio to say 20:1. It would therefore run by compression ignition like a diesel engine, but with nearly double the expansion ratio it could have spectacular thermal efficiency. In fact, the exhaust gases would probably be almost cool to the touch! By making the piston do more work, and take far more useful energy out of the hot gases, I don't see any reason why we couldn't hit 60% thermal efficiency.
Now of course, such an engine would not be very driveable. It would be gutless at certain rpms, and somewhat limited in terms of maximum possible rpms. But none of that matters if it's driving a generator. It can also of course run at peak efficiency all the time
, which no gearbox mounted engine can do.
Of course, the main criticism of this set up would be the inherent loss of energy going from mechanical to electrical to mechanical again. (Traditionally it's all mechanical). But, just follow these sums.
Assuming biodiesel to contain 39 kWh
per US gallon chemical energy, the engine can extract 23.4 kWh mechanical energy. With an efficient generator and motor, this could yield 22.5 kWh at the wheels
. Assuming an efficient EV drivetrain of 5 miles per kWh, that equates to 112 mpg
when driving in this mode, or approximately DOUBLE the current Prius mileage.
So what do the engine techies on here think: 60% efficiency - is it achievable?