If KTM's next big thing is anything like you just said, then we know what's going on here
Of course, DTPI definitely is a useful technology. Do you know how much of your fuel savings came from better AFR control and how much was due to increased trapping efficiency? Are you injecting against the transfer flow direction, too? Seems kind of counter productive in regards of increasing trapping efficiency (but is ideal for mixing).
Generally, I think for two-strokes to have a comeback in anything else than a niche market, they need to be just as clean as four stroke engines regarding engine raw emissions.
Compared to four-strokes, two-strokes are typically:
- lower on NOx emissions (lower peak temperatures if you fire twice as often)
- much worse in HC, which is due to the scavenging losses (see below)
- worse in CO2 emissions, which is basically fuel consumption, or BSFC (grams of fuel per kW per hour)
- better or worse in particulate emissions (better with PFI, might become worse with high pressure direct injection, longer mixing time is better)
- and CO values depend on AFR setting
(Both types of engines burn oil, but it has already been shown that this can be dealt with, regarding the two-stroke, once the other problems have ben solved.)
So, scavenging losses and BSFC need to be improved. Luckily, if you reduce scavenging losses you also reduce fuel consumption.
Okay then, so the E-TEC is exactly what we are looking for? Well, not quite. The E-TEC has a DI system capable of about 35bar pressure from what I read. This is not enough to get the required mixture in the cylinder and mix it between exhaust port closure and start of ignition (~70 ish deg CA). So at higher loads, it has to inject for much longer periods of time thus generating scavenging losses just as a carbureted two-stroke. How can this be solved? Simple, more pressure. How much? Much more. How much exactly? Well, today's latest car technology uses an injection pressure of about 250 bar in their gasoline engines. These are turbocharged around roughly 20 bar (4-stroke) BMEP, which would be equivalent to 10 bar two-stroke BMEP. And they turn about 6500 rpm. That's 3250 injections per minute, let's say 1/3 of what we would need (~10k). So we're at 750 bar now, right? As we would only need half of the mixture per injection (assuming one injection per stroke), we can halve that and arrive at 375 bar and we realise that we have only about 70 deg CA to get everything done, whereas the 4stroke has more than 180°, (but uses the time available for more than one injection most of the time).
All in all, I think as soon as gasoline DI injection technology has reached a level of 500 bar in serial application (500 bar is what F1 is currently running), and CFD computing power is available abundantly, too, one could give it a try. On the other hand, adapting today's car technology one could already get a quite clean two-stroke, only that it's "clean" area of the engine map woud be somewhat revs and power limited.
Bookmarks