Back in the carburetor days, it was common knowledge that every time you pressed on the gas pedal, the accelerator pump would squirt raw gasoline into the venture to compensate for RPM lag when your foot put out the call for power, which contributed to poor MPG rating.
However with the advent of direct fuel injection, your foot (via the throttle body) only controls the amount of air coming into the cylinders, with the computer regulating the amount of gas being injected based on a programmed air/gas ratio.
My belief is that other than to overcome the inertia load, which should theoretically be equal between a carbureted or FI engine under identical circumstances, pumping the accelerator on a FI engine while underway should have minimal, if any, effect on MGP since you are not pumping raw gas into the engine. Anyone with thoughts on this?
Steady state operation is the most fuel efficient. The engine sensors tell the engine management computer that the throttle position has been changed long before the varying air flow has an effect on cylinder filling. www.wikipedia.com has some good articles on fuel injection, sensors, and engine controls.
On a FI engine pushing the throttle down tells the computer that you want to accelerate and it will increase the fuel to the injectors. It won’t “over-fuel” like a carb engine, but it will call for more fuel. The most efficient way to drive is to keep the throttle steady.
Older fuel injection systems boost the fuel rail pressure as the vacuum drops. Even newer returnless fuel systems have a small spring loaded piston that operates on vacuum to boost fuel pressure. The TPS also tells the computer you are trying to accelerate and extends the injector on length. So, the PUMP still exists!
it was common knowledge that every time you pressed on the gas pedal, the accelerator pump would squirt raw gasoline into the venture
Common knowledge? That’s just plain wrong unless you didn’t properly adjust the linkage for the pump. Slow tip in should result in NO accelerator pump output. It’s only needed when the tip in rate exceeds the mass flow change rate and results in a temporary lean condition. If you accelerate slowly, the AP shouldn’t even be operating. And the amount of output is adjusted by a cam so that the further you press the gas pedal the more fuel is injected. It’s not an all or nothing proposition. A properly adjusted carb works very well and is not the gas hog you imply. They suffer from lack of dynamic adjustability and that is where FI really shines.
pumping the accelerator on a FI engine while underway should have minimal, if any, effect on MGP since you are not pumping raw gas into the engine
If you think that problem is totally eliminated by virtue of FI alone, you’re mistaken.
Both carburetion and fuel injection enrich the fuel/air ratio when the engine is under load. The carburetor does it directly via the accelerator pump, while the FI-engine does it electronically as a result of various sensor feedback to the computer. The effect on mpg ought to be very similar for both cases.
And yes, in each case you are squirting “raw gas” into the air stream. How else do you get it vaporized?
Opera House’s vacuum controlled fuel injection system brough back memories, but I’ll assume you’re talking post-1970. (just ribbing you OH)
I’ll also assume that by “direct fuel injection” you’re actually referring to multiport injection, as direct injection systems are just coming on the market. (diesel guys, I know already, I know already, but we’re talking gas engines here). Direct injection sprays gas directly into the cylinders rather than at the intake ports.
With fuel injected engines, the pressure of the fuel system is kept high and the amount of fuel delivered to the cylinders is controlled by adjusting the “pulse width” of the injectors. Injectors are more like on/off switches than rheostats. They’re either fully opened or fully closed. How long they’re kept open for each pulse determines how much fuel is sprayed. The pulse width is controlled by the computer, which uses inputs from the “engine demand” sensors including the Throttle Position Sensor. The whole process happens in milliseconds, almost instantaneously. And remember that the injectors are right behind the intake valves, so there’s no time lag from the spray having to travel through a manifold system. And the “feedback” from the upstream oxygen sensor and the computer’s response are also almost instantaneous. so there’s no gas wasted.
In short, yup, the inertia problem (the energy needed for the change in velocity from speed X to speed Y) is the same, but the fuel is more accurately controlled with fuel injection. None is wasted and blown out as unburned hydrocarbons as it is in a carburated engine.
The real world difference? Who really knows. My buddies 1968 327 Camaro got 25 mpg…
TT, thanks for your comments. You are correct in that the amount of raw gas entering the venturi area is proportional to correcting the temporary lean condition at acceleration until normal venturi action can pick up the load. And yes, during normal (easy foot) driving, the amount of raw gas being introduced by the accelerator pump is minimal if any. My point was that every time you pressed the accelerator to get more power, and depending on how far and fast you depressed the throttle, a proportional amount of raw gas was pumped directly into the air stream. This condition does not occur with FI engines, which in-of-itself makes for better fuel management resulting in better MPG.
Under normal driving the gas is brought into the venturi by lowered atmospheric pressure, while the accelerator pump actually forces the gas in by visible squirts. You can easily see this action by peering into the carb and actuating the throttle linkage. I don?t believe this has anything to do with any adjustment in the accelerator pump mechanism.
The point I was exploring is whether FI engines suffer from the ill affects of ?pumping? the accelerator pedal in the same way as carbureted engines. I believe not.
My 1967 Cougar with a HP-289 got between 14 - 18 mpg. It could pass everyting but a gas station.
My point was that every time you pressed the accelerator to get more power, and depending on how far and fast you depressed the throttle, a proportional amount of raw gas was pumped directly into the air stream
while the accelerator pump actually forces the gas in by visible squirts. You can easily see this action by peering into the carb and actuating the throttle linkage. I don?t believe this has anything to do with any adjustment in the accelerator pump mechanism.
These seem to be a contradiction. The point is, the amount of fuel squirted into the venturi is controlled by the cam position/adjustment. If properly adjusted, it will not squirt in any more fuel than is required for smooth acceleration. Many people were ignorant of this adjustment or just didn’t care. Gas was cheap.
FI must do the same thing but it can be more efficient under varying conditions. The main difference is it is under computer programming control. They can add hysteresis to the control loop. A loose carb based analogy would be mechanical versus vacuum secondaries. The FI programming can be more tolerant of “requests” for power and apply it smoothly regardless of the ramp speed of the input. It can also decide to go WOT if the conditions dictate. It still has to accomodate the lean tip in condition. You’d like to think that intelligence in control leads to improvements…
Some of that extra fuel a carb accel pump squirts in is needed because when the throttle tips open and manifold vacuum drops some fuel collects/condenses in the intake manifold and can even form a small puddle if the manifold isn’t warm enough. It’s a transient condition that last for just a fraction of a second. Emission control carbs usually have a dashpot on the throttle to slow down its closing to reduce a rich condition when the manifold vacuum increases and the fuel collected in the manifold suddenly vaporizes. One of athe advantages of multipoint FI over single point is that it eliminates fuel collecting in the intake manifold, simplifying fuel management under changing conditions.
What you are referring to is acceleration enrichment. EFI also incorporates acceleration enrichment by several different schemes:
- Load rate of change; the faster the engine load changes the more fuel is added
- Throttle rate of change; the faster you push the throttle peddle, the more fuel is added
- Throttle position
The fuel droplets in carburated engines also coalesce (SP?) into larger droplets as they travel the manifold on the way to the cylinder. And, carburators operating on far lower pressure differentials between the fuel and the spray sides of the orafice, they produce bigger droplets (like slowly squeezing the Windex spray as opposed to squeezing it hard and fast). And, of course, a fine mist burns better than a larger-droplet mist…you know, the surface area in contact with oxygen things.
The other nice thing about multipoint FI is the lack of lag. As soon as the throttle plate is opened the piston pulling in air feels an immediate drop in resistance and is immediately pulling in more air. In a carburated system, the added fuel is still way back at the carburator. In multiport FI, the added fuel is right behind the valve! Granted the timeframes are small, but just the same…
Watch that foot! Those injectors can really push the fuel. It’s why there is so much power from so small an engine. You are asking for more fuel when you push the pedal and it will be delivered in a hurry.