DFI, does anyone own it?

[theunchosen]

I've been gone for a while. I was trying to make a Stirling engine cooperate with an ICE to generate electricity to power electric motors.

I know it took a long time, and the results were mixed. Overall FE dropped(measured everything in KWHR produced per gallon fuel), but it could be done.

So I moved on for now.

The next topic that has got me fired up is Direct Fuel Injection or just DI. It occurs to me that someone here might own one, so I have questions. Please reply if you do.

Otherwise this is an open design project. If you don't know what it is Wikipedia is pretty accurate on this one. My design suspicion is that DFI can beat out hybrids in the next gen series.

My reasons are that DFI isn't necessarily Throttle dependent. In essence the gen 2 engines will be running WOT all the time and just varying how much fuel is injected. Preliminary data from research is suggesting from Ford DIE is 20% more Efficient and GM's counterpart is 30% more efficient than current ICE. 30% on a CVT civic gets you just over 50 mpg city.

I would gamble both of those numbers are liberal, considering both are suffering tanking pulic image and need someway out.

However other firms are running their own Ultra lean burns and getting close results.

My question is does Ultra lean burn occur at a lower temperature than 14.7:1 or higher.

I'm pretty confident it would not get as hot. example we have 29 grams of air in both situations 2 grams gas in stoich, and just one in lean. I'm pretty sure the lean will be a good bit cooler, but obviously above 50% of the BTUs released in the stoich chamber because we have the same amount of O2 burning.

IF, IF it runs cooler then I am pretty confident also that we can play around with the valve setup without having to worry too much about thermal warping.

I've been kicking around some ideas offline, but we weren't sure if you could expose the exhaust valves to room temperature over their lifetime without dramatically shortening their number of cycles.

Anybody?

[Christ]

The only thing I know about GDI is what I read in an article from Honda about a year and a half ago, and that was Honda engineers touting a design that could run on 60:1 AFR reliably. (Nothing further came of it, that I know of.)

I've considered attempting to make an old ICE into a GDI type engine, but I never really had the time and patience at the same time to sit and play with it. I may consider doing it on a small V-engine, like a twin cylinder lawnmower (20HP Kohler or something.)

[JoJotheTireMan]

I read an article once, don't remember the source, but it was discussing fuel vapor explosions, military/goverment document of some sort. They discussed a range of 1.3% - 6.25% fuel to air ratios are explosive. 15:1 air:fuel is about 6.25% (richest in their range) and 75:1 air:fuel is about 1.315% (leanest in their range). So, is it possible to run leaner, I would have to say yes for sure, but how much is the question. Not sure what raising the temps and pressures due to these numbers, as their discussion was on fuel vapor clouds in the air, so room temperature and atmosheric pressure.

Just some fuel for thought. Enjoy JoJo

[theunchosen]

Thanks Chris and Jojo,
Yeah so far I have not run into anyone who knows much more about it than articles or research. I think 04 was the earliest model year that saw DFI in a production car in the last two decades(for gasoline obviously diesels already do it).

I will in no way pretend to have the automotive skills to convert to DFI. I would definitely just have to find a vehicle(2009 lincoln MKs) with that engine and rip it out. Realistically I would have to wait for them to produce an L4 DFI with injection controlled throttle(probably civic will roll out with a package like this by 2010).

The reason I asked about valve temperature gradient is because I had an idea regarding increasing air flow to the cylinders(which only matters if your engine is fuel injected throttled, otherwise its like you're pushing the gas pedal down further.)

Since there is nothing special about the air before it gets into the cylinder(no fuel added till then) there is no reason you couldn't expose it to potentially very hot valve heads. Hopefully I have not overlooked something to be a serious problem, but you could use your exhaust valves as intakes and your intakes as exhausts.

No superchargers, no butterfly valves, and no complicated overpressure shemes with springs to make sure one side is open and the other is closed. Think distributor cap. Put a small cylinder above the head that has an open top and bottom. valves enter from the top and exit into the head at the bottom. Cylinder has a wide-mouthed opening larger than the air intake tubing. The cylinder would need an opening that runs a quarter or 90 degrees around. We don't want any time when the exhaust is open to the intake. . .but we want it to be a very very short transition.

Cylinder rotates at the same speed as distributor cap or the crank shaft.

> = cylinder opening
-cold air intake- > Intake stroke
___________exhaust
-cold air intake- V compression
___________exhaust
-cold air intake- < combustion/expansion
___________exhaust
-cold air intake- ^ exhaust
___________exhaust
repeat intake

Obviously the manifolds would not be 180 degrees opposed anymore. The cylinder would need to rotate quickly between exhaust and intake because it only has a few degrees of piston movement to make that happen.

Another possible solution coupled with the 90 degree bend(opposed to 180) would be to use electric motors that could very quickly cycle back and forth between near free-wheeling(compression, expansion and intake strokes) and high output(between exhaust and intake strokes).

Anyway theoretically it would give you access to all (4 valves for me) valves for intake and exhaust, thoroughly expelling all heated gases and allowing the chamber to cool more significantly, and intaking a good bit more air at less pumping losses in the intake, while also cooling any surfaces that may cause pre-ignition(or decrease the overall thermal stress and increase everything's lifetime).

Thats why I ask about DFI

[theunchosen]

oh, it would also decrease the valve movement. They all just move 4 times at the same time. So instead of two cams for each cylinder you just need one. Or if its VTEC or equivalent you just need one system for all 4 instead of two.

I haven't considered the exhaust piping, but it would go from the side of each cylinder out the back of the block. A hood scoop(small) would likely be adviseable to cool the backside of the engine(which I'm sure during the winter someone here will block off to decrease startup to maximum efficiency.) I predict engine temps will run much cooler, because the incoming air through(what grill you have left, coroplasters) will not have to heat soak across the exhaust pipes as well.

It would also radically improve the amount of power pulled out of a turbo because the pipes aren't in contact with cool air at all so no Q is lost. The turbo would not decrease FE at all because its not O2 controlled. . .its fuel direct injected, which is determined only by your foot and nothing else.

It would very likely radically improve FE(I suspect instead of being a HP mod they would become an equivalent for FE(because its adding even more air that you don't have to pay for, unlike gas)).

I suspect also that you can go much lower than 1.3% at compression. Normal air is not terrifically keen on burning. Yes it will do it but there are too many nitrogen molecules to really make it happen. Higher pressure gets the odds of having multiple O2 molecules that are almost touching up. Also during the compression the temperature alone will come reasonably close to causing the air to burn, afterall those two factors are what cause detonation or knock in your engine.