Originally Posted by serialk11r
Direct injection doesn't provide as large of an efficiency increase the way Toyota is using it with non-stratified charge. I think that article isn't exactly correct, a paper that someone linked explaining D4-S in detail explained the purpose of having both port and direct is to get a more homogeneous mix in the cylinder. D4-S does NOT use stratified charge.
With VW style stratified charge it's theoretically possible to increase part load efficiency a bit since they remove the throttle, but this creates a lot of NOx emissions and efficiency still drops at part load since at very high AFR the pressure ratio is crap.
The direct contribution of direct injection would be increased compression ratio/better spark timing, and in the case of D4-S, better combustion efficiency since the fuel is more evenly distributed. Not to say this is a small improvement, but it's not very big either, as increasing compression ratio a huge amount like from 10:1 (typical cars running on 87, with port injection) to 15:1 brings less than 10% theoretical improvement in efficiency. As far as combustion efficiency goes, I don't know how much improvement they can get over run of the mill port injected engines...
As for the rotary being like a 2 stroke, this is not a good analogy because a rotary is like 3 4 stroke engines sharing one set of ports. There is no overlap between intake and exhaust stroke, and there are 4 distinct strokes. Direct injection on a rotary wouldn't bring much improvement because the main issue with the rotary is that the combustion chamber shape is very VERY bad, and a slightly smaller but still annoying problem is that compression ratio can't be very high since the shape of the engine won't permit it. In all other ways, the rotary is actually better for efficiency. Lower friction, less losses to vibration, no losses at the valves.
I'll just remind everyone again, the KEY step to improving fuel efficiency, particularly for turbocharged motors, is variable intake duration. At idle, using late intake valve closure instead of a throttle on some motorbike engine (with the "Williams Helical Camshaft") reduced fuel consumption by 80%. Toyota, BMW, and Nissan claim over 10% overall efficiency improvement with Valvematic, Valvetronic, and VVEL. When you consider that just adding (in Toyota's case at least) an adjustable rocker to the head improves efficiency on an already pretty efficient Toyota ZR motor by more than 10%, while Subaru completely reworked their thirsty EJ motor, reducing friction by a massive 28% and reworking the combustion chamber for 10%, you see why pumping loss through the throttle is the biggest problem for efficiency (aside from gearing, which can completely kill fuel efficiency when it's retardedly short). In the case of your STi, you have very short gears, a thirsty engine in the first place, and a pretty big turbo sapping extra power at part load by creating backpressure, so it's no surprise it's not fuel efficient. Direct injection can only slightly remedy the problem.
This car can probably get into the 30s for mpgs since it's coming out in 2012, so they'll be careful not to kill the fuel economy by not having a tall cruising gear, and it's a high compression ratio N/A engine. If the Civic Si can do over 30 on the highway with its short gearing and relatively primitive engine tech, this should be able to do at least as well.
|