How do aircraft fuel injection systems work?

cowman

cowman

Final Approach
Joined
Aug 12, 2012
Messages
5,280
Location
Danger Zone
Display Name
Display name:
Cowman
I was looking for a good rundown on this subject, maybe someone has a link to a site or video.

I know how it works operationally, I've flown a couple and know what the procedures are. What I don't know is the mechanics of it. Obviously unlike automotive EFI there aren't any sensors or electronic systems to determine when and how long of a fuel squirt so how is that done? I assume understanding this will help me understand exactly why I have to do all this silly fuel management stuff to get the engine started without flooding it.... and I just generally like to know how the thing keeping me in the air works.
 
ktup-flyer said:
Click to expand...
That's what I was looking for, thanks.



Also....
IT'S JUST A FANCY GAS SPRINKLER THAT JUST SPRAYS GAS WITH ABANDON
latest


my life is a lie
 
It's more of a constant dribble than a squirt. Fuel pressure at the nozzle is ~2-3 psi. Don Rivera of Airflow Performance provides a class twice a year on the fundamentals (FI-101). They are located on field in Spartanburg, SC. Great class and great food.
 
Lycoming is different than Continental. Both are mechanical. My exp system is also mechanical but lots of exp guys are using EFI. Mechanical FI is kind of Fred Flintstone compared to EFI but it works pretty well in my plane.
 
Raymo said:
It's more of a constant dribble than a squirt. Fuel pressure at the nozzle is ~2-3 psi. Don Rivera of Airflow Performance provides a class twice a year on the fundamentals (FI-101). They are located on field in Spartanburg, SC. Great class and great food.
Click to expand...

Nope. It's as much as 20 PSI. You might see 4 psi at idle. The gauge from a T210:

images


Look at the little number at the top of the fuel flow scale. 19.5 psi at redline. 2-3 psi wouldn't atomize the fuel at all and engine performance and fuel economy would be terrible.
 
Dan Thomas said:
Nope. It's as much as 20 PSI. You might see 4 psi at idle. The gauge from a T210:

images


Look at the little number at the top of the fuel flow scale. 19.5 psi at redline. 2-3 psi wouldn't atomize the fuel at all and engine performance and fuel economy would be terrible.
Click to expand...


NOPE. That is fuel flow, not pressure (not sure why it has PSI on the gauge). Where it is measured matters. That gauge is not measuring the pressure at the injector.
 
They work great. The efi systems in cars require O2 sensors to work well, those sensors are poisoned and stop working quickly when exposed to lead. I think, if a no lead aviation gasoline is ever approved we would see a rather quick development of an EFI system for airplane engines, along with increased power and reliability.
 
Hi, I know a thing or two about this.

On your Lycoming the basic description is as follows.

You have a mechanical fuel pump that will generally run around 35 psi or so to a servo (exact settings vary). This servo has a venturi (similar to a carb). The Delta-P that you get from this venturi moves a piece of rubber inside which then kicks that ~35 PSI down to a lower number. The delta P is caused by airflow going through the venturi, so this works well for measuring actual airflow.

Either before or after that metering (I forget which) is the pressure reduction caused by the mixture knob, which is essentially a variable orifice to reduce pressure.

After the servo, the fuel goes up to a fuel distributor (or spider) which then splits that fuel flow (pressure) to each of the nozzles (one per cylinder). Those nozzles have calibrated orifices to produce equal flow. This is part of why it's important to keep them clean. The nozzles have screens to allow some small amount of air in with the fuel to help atomize it. This is a continuous injection system so it's a constant dribble/spray/whatever. However the pressure at the nozzles will vary depending on the engine and the power setting. More pressure = more fuel. I forget the exact range here.

If you have a turbocharged engine the nozzles have boost pressure fed on the outside of the engine as well to prevent the boost from trying to blow the fuel out. It's important to keep these lines clear if you have them or it can cause all kinds of weird fuel flow imbalances.

Certain Lycomings have a fuel return for venting purposes to aid hot starting. I think the restart Cessnas were the main ones with that.

Continental is another ball of wax. It's a simpler concept that ends up being more complex in execution.
 
PaulS said:
They work great. The efi systems in cars require O2 sensors to work well, those sensors are poisoned and stop working quickly when exposed to lead. I think, if a no lead aviation gasoline is ever approved we would see a rather quick development of an EFI system for airplane engines, along with increased power and reliability.
Click to expand...

You will NOT see increased power from going to EFI. This is a misconception.

Reliability depends on the execution.
 
Most FI engines use higher compression so displacement for displacement FI generally is rated for higher power. There's more to it than FI alone. On the other hand, use FI in concert with electronic ignition and you can improve efficiency, but not power.
 
Ted DuPuis said:
You will NOT see increased power from going to EFI. This is a misconception.

Reliability depends on the execution.
Click to expand...

Actually I believe we would see increased power as the mixture control would be much more precise, electronic ignitions would be part of the package and margins could be reduced. Strategies could be employed to reduce cylinder head temperatures and pilots stuck in the past would bemoan the disappearance of the mixture control. Hopefully someday we will find out if I'm right or not.
 
Raymo said:
It's more of a constant dribble than a squirt
Click to expand...
upload_2019-2-14_16-51-53.png

PaulS said:
They work great
Click to expand...
At the end of the day, the fuel is not really being "wasted" since it's going into the intake manifold. Timing it and squirting it directly into the cylinder might have some advantages.. but then I would think that A.) adjustable valve timing would give a bigger benefit, and B.) having it above the intake port probably helps keep the valve decently clean
^unless I misunderstood the whole thing
 
PaulS said:
Actually I believe we would see increased power as the mixture control would be much more precise, electronic ignitions would be part of the package and margins could be reduced. Strategies could be employed to reduce cylinder head temperatures and pilots stuck in the past would bemoan the disappearance of the mixture control. Hopefully someday we will find out if I'm right or not.
Click to expand...

Well, after doing exactly that for a living for 5 years and spending around 3,000 hours on engine stands running and calibrating the things, the biggest problem we had was trying to make the public (well, marketing) realize that the rated power output of the engine did not change. Sure at cruise power you could advance the ignition timing on lower power engines, but your rated power output doesn't change.

You need to do something that changes the airflow properties or the inherent of combustion to do that (specifically a new combustion chamber), and EFI/ignition timing doesn't do that.

Tantalum said:
At the end of the day, the fuel is not really being "wasted" since it's going into the intake manifold. Timing it and squirting it directly into the cylinder might have some advantages.. but then I would think that A.) adjustable valve timing would give a bigger benefit, and B.) having it above the intake port probably helps keep the valve decently clean
^unless I misunderstood the whole thing
Click to expand...

Injection timing is an emissions issue.

Back when I did this for a living I had one setup that was so bad it was literally dropping 1/4" diameter droplets of fuel at a time into the intake runners.

Horsepower difference with that setup vs. a proper setup: 0
BSFC difference with that setup vs. a proper setup at rated power: 0

Main thing you noticed was that at idle it didn't run well.
 
Ted DuPuis said:
Well, after doing exactly that for a living for 5 years and spending around 3,000 hours on engine stands running and calibrating the things, the biggest problem we had was trying to make the public (well, marketing) realize that the rated power output of the engine did not change. Sure at cruise power you could advance the ignition timing on lower power engines, but your rated power output doesn't change.

You need to do something that changes the airflow properties or the inherent of combustion to do that (specifically a new combustion chamber), and EFI/ignition timing doesn't do that.



Injection timing is an emissions issue.

Back when I did this for a living I had one setup that was so bad it was literally dropping 1/4" diameter droplets of fuel at a time into the intake runners.

Horsepower difference with that setup vs. a proper setup: 0
BSFC difference with that setup vs. a proper setup at rated power: 0

Main thing you noticed was that at idle it didn't run well.
Click to expand...

Ted, you are focusing on one or two pieces of a big puzzle. I'm not talking about retrofits, although I think some gains could be made, I'm talking about new aviation engines, designed for EFI systems with O2 feed back loops, mass flow sensors or how ever they determine airflow these days, knock sensors, higher or lower compression, variable timing possibly. New engine in this case means something like a legacy IO-550 with new cylinders to accommodate the above, maybe new pistons and rods if needed, new intakes, new exhaust, what ever. Things that don't make sense to do when you can only read egts.
 
PaulS said:
Ted, you are focusing on one or two pieces of a big puzzle. I'm not talking about retrofits, although I think some gains could be made, I'm talking about new aviation engines, designed for EFI systems with O2 feed back loops, mass flow sensors or how ever they determine airflow these days, knock sensors, higher or lower compression, variable timing possibly. New engine in this case means something like a legacy IO-550 with new cylinders to accommodate the above, maybe new pistons and rods if needed, new intakes, new exhaust, what ever. Things that don't make sense to do when you can only read egts.
Click to expand...

EFI with variable valve timing, O2 feedback loops etc make a huge difference when your engine's power output varies widely and often. Valve timing, ignition timing, fuel flow, etc. on older car engines were a compromise between idle, acceleration and various cruise speeds. You know that V-8 lope at idle of a hot rod? That's from optimizing valve timing for higher RPM power.

Aircraft engines (as crude as they may seem) are actually pretty well optimized for 75-100% power. You'll perhaps get some fuel efficiency from better mixture control but that's about it.

Trust @Ted DuPuis on this one.
John
 
Well, as someone who did this stuff at an OEM for more than 30 years, what Ted said.

O2 feedback is important if you have catalysts and emission standards to meet. EGT is not that bad for getting where you need to go.
Getting power means running rich of peak. Even the crap mechanical systems manage to blob in the extra fuel needed with out issues. Economy comes from not running rich of peak and ****ing fuel out the tailpipe. In extreme cases, and without emission standards, you can run way lean and save a few more percent beyond "LOP" but that would require re-design of the combustion chambers along with the ignition systems. Things like variable timing really shine when you operate across a wide range of power settings - less when you spend most of the time between 65 and 76 percent.
 
PaulS said:
Ted, you are focusing on one or two pieces of a big puzzle. I'm not talking about retrofits, although I think some gains could be made, I'm talking about new aviation engines, designed for EFI systems with O2 feed back loops, mass flow sensors or how ever they determine airflow these days, knock sensors, higher or lower compression, variable timing possibly. New engine in this case means something like a legacy IO-550 with new cylinders to accommodate the above, maybe new pistons and rods if needed, new intakes, new exhaust, what ever. Things that don't make sense to do when you can only read egts.
Click to expand...

I was looking at the part of the puzzle that you mentioned, which was EFI.

Going through your list of other options, the cylinders (specifically combustion chambers), pistons, intakes, and exhausts are the items that you'd need to change. The reality is the certification barrier of getting into those items, especially the cylinders, is very high and the reason that hasn't been done is not due to a lack of EFI, it's because there's just not the money to make the investment worthwhile.

Really, all of those changes could be done using the mechanical setups we use today and it's been done. However I will grant that if you were to make a more tricked out higher compression engine, you could potentially have those tools help be a certification enabler with respect to detonation margins, etc. The engine would still run the same without them.

But you put EFI and electronic ignition on today's engines, you won't get more power, that was my point.

If you look at the Lycoming iE2, that is at its core a legacy TIO-540 with electronic fuel, ignition, spark, knock sensors, etc.
 
Raymo said:
NOPE. That is fuel flow, not pressure (not sure why it has PSI on the gauge). Where it is measured matters. That gauge is not measuring the pressure at the injector.
Click to expand...
On my Continental IO550 at 2500RPM, the metered pressure (entering the spider) is supposed to be 17.2-20psi. What would the pressure be at the injector? Divide by 6?
 
Raymo said:
NOPE. That is fuel flow, not pressure (not sure why it has PSI on the gauge). Where it is measured matters. That gauge is not measuring the pressure at the injector.
Click to expand...
Are you an aircraft mechanic with experience working on injection systems? I am, and the fuel flow gauge is a pressure gauge calibrated in gallons per hour, and sometimes pounds per hour. The gauge is connected to the flow divider where the injector lines come together. Fuel flow is a direct relationship to the pressure applied to it. And thats' why the pressure is on the gauge. It serves as a calibration point for checking accuracy of the gauge.

Go here and read all about it: http://www.flight-mechanic.com/fuel-injection-systems-part-two/

An excerpt from that:
A fuel pressure gauge, calibrated in pounds per hour fuel flow, can be used as a fuel flow meter with the Bendix RSA injection system. This gauge is connected to the flow divider and senses the pressure being applied to the discharge nozzle. This pressure is in direct proportion to the fuel flow and indicates the engine power output and fuel consumption.

That specifies the Bendix RSA system, which is used on Lycomings, but Continental does the same thing with theirs.

From https://www.accessengineeringlibrar...nts-ninth-edition/c9781259835704ch07lev1sec02

...we read:

The Continental fuel injection system is a multinozzle, continuous-flow type which controls fuel flow to match engine airflow. The fuel is discharged into the intake port of each cylinder. Any change in air throttle position or engine speed, or a combination of both, causes changes in fuel flow in the correct relation to engine airflow. A manual mixture control and a pressure gauge, indicating metered fuel pressure, are provided for precise leaning at any combination of altitude and power. Since fuel flow is directly proportional to metered fuel pressure, the settings can be predetermined and the fuel consumption accurately predicted.

Modern glass-panel airplanes use a flowmeter installed in the fuel line to the divider to measure actual fuel flow, and STC'd systems are available for steam-gauge airplanes to install the same thing. I've done a few installations of those.
 
donjohnston said:
On my Continental IO550 at 2500RPM, the metered pressure (entering the spider) is supposed to be 17.2-20psi. What would the pressure be at the injector? Divide by 6?
Click to expand...
No. It would be a hair less due to the internal drag of the injector line. Each line will get the full pressure where it connects to the divider.

You'll often see injector lines formed into coils and so on to keep the line lengths all the same to keep fuel flow identical to each cylinder.
 
Ted DuPuis said:
I was looking at the part of the puzzle that you mentioned, which was EFI.

Going through your list of other options, the cylinders (specifically combustion chambers), pistons, intakes, and exhausts are the items that you'd need to change. The reality is the certification barrier of getting into those items, especially the cylinders, is very high and the reason that hasn't been done is not due to a lack of EFI, it's because there's just not the money to make the investment worthwhile.

Really, all of those changes could be done using the mechanical setups we use today and it's been done. However I will grant that if you were to make a more tricked out higher compression engine, you could potentially have those tools help be a certification enabler with respect to detonation margins, etc. The engine would still run the same without them.

But you put EFI and electronic ignition on today's engines, you won't get more power, that was my point.

If you look at the Lycoming iE2, that is at its core a legacy TIO-540 with electronic fuel, ignition, spark, knock sensors, etc.
Click to expand...

Yup, that's what I thought you were getting at, and you are probably right about the power deal but it would be interesting to see what would happen. The possibilities are greatly increased with unleaded gas in my opinion. I would just like to see these engines get there, to electronic controls, O2 is a much better way to control the mixture than egt, there is much less hysteresis in the system with corrections close to instantaneous. Personally I would be happy not to have to f around with the mixture, more time to look outside or focus on navigating. Plus I think the engines would be much more reliable. Just a lot of plusses with very few negatives.
 
You got to remember that the job that the fuel injection system has to do in an airplane is much simpler than what it needs to do in a car. In a car, you're often constantly varying engine RPM and load while in an airplane those factors are fairly constant in any given flight regimen. So, in a car, the fuel injection must respond accurately to those changing criteria while in an airplane all it needs to do is get it close and the pilot will put it the rest of the way with the mixture control.
 
cowman said:
....
IT'S JUST A FANCY GAS SPRINKLER THAT JUST SPRAYS GAS WITH ABANDON
latest


my life is a lie
Click to expand...
Isn't this system almost identical to the Bosch K-Jetronic such as that used in the Porsche 911 up until the mid-90's?
 
asicer said:
Isn't this system almost identical to the Bosch K-Jetronic such as that used in the Porsche 911 up until the mid-90's?
Click to expand...

In a sense, I guess, though those systems used an air mass-type of flow sensor, not one based on differential pressure.
 
alfadog said:
In a sense, I guess, though those systems used an air mass-type of flow sensor, not one based on differential pressure.
Click to expand...
If you read the history and evolution of the Bosch systems from K-Jet to K-jet with lambda (oxygen sensor) through LH to motronic, all this makes sense. L Jet has a flapper air measuring system, LH has a hotwire mass airflow but still sprays all four injectors at the same time. Motronic combined pulsed sequential injection and variable timing. All the modern automotive systems are basically derived from the fundamentals of Bosch injection. I own a car shop and now I'm a pilot as well it's amazing how primitive the systems are on airplanes. Bosch had Motronic in production in 1985. KJet in the 70's. I want nothing to do with a MS carb and mags. Guess that's why I fly a rotax.....
 
asicer said:
Isn't this system almost identical to the Bosch K-Jetronic such as that used in the Porsche 911 up until the mid-90's?
Click to expand...

Probably true of any FI system that goes to the intake as oppose to direct injection into the cylinders.
I assume early Corvettes were this way as well.


Tom
 
MooneyDriver78 said:
Probably true of any FI system that goes to the intake as oppose to direct injection into the cylinders.
I assume early Corvettes were this way as well.
Click to expand...
No, most modern-ish FI systems before DI used pulse width modulation to meter the fuel, such as the Bosch L-Jetronic. Alfa Romeo's SPICA system used individual pumps per cylinder. And then there's throttle body injection...
 
alfadog said:
You got to remember that the job that the fuel injection system has to do in an airplane is much simpler than what it needs to do in a car. In a car, you're often constantly varying engine RPM and load while in an airplane those factors are fairly constant in any given flight regimen. So, in a car, the fuel injection must respond accurately to those changing criteria while in an airplane all it needs to do is get it close and the pilot will put it the rest of the way with the mixture control.
Click to expand...

The systems used in aircraft are designed for running. Ones in cars are designed for starting and transients. It's a different set of requirements. In an airplane, something that has the potential to not open and cause running issues in-flight is a much bigger concern than in a car.

asicer said:
Isn't this system almost identical to the Bosch K-Jetronic such as that used in the Porsche 911 up until the mid-90's?
Click to expand...

It may look similar but it's a lot different. The K-Jet ran at a much higher pressure (around 110 PSI, as I recall), and the injectors were actually pop-off valves, where pressure had to get above a certain pressure to then open and allow fuel in. There was some sort of injector timing but I think that more than anything had to do with being able to make starting easier. No priming with a K-jet, just turn the key and it starts. The fuel distributor was a much more complex unit.
 
How often do modern automobile fuel injection systems fail?
The EI systems use to fail regularly, nowadays they seem to be very solid. And modern injectors can clogged as easily as our old ones.


Tom
 
Again, as someone who did fuel control for an OEM for more years than I will admit to, the auto industry was dragged kicking and screaming from carburetors to feedback carburetors (talk about lipstick on a pig (yes, I was there)), to throttle body injection, to port injection, to synchronized port injection to direct injection to keep up with the demands of emission regulations. Things like the mass of fuel on the walls of the intake can absolutely clobber your HC emissions - even with port injection - during a transient cycle like the EPA test. (Clobber being a relative term - at the California PZEV levels, even a milligram of unburned HC making it through to the tailpipe is a disaster.)

Same for air metering - speed/density (engine speed, intake air pressure / temperature, assumed volumetric efficiency), vane meters, mass air. The interesting thing there is that the industry has got better at measuring and metering things like Exhaust Gas Re-circulation so many applications have gone back to the lower cost speed/density calculations.

FWIW, there is a > 10% chance I have had my finger in the fuel control of your daily driver...
 
Dan Thomas said:
Are you an aircraft mechanic with experience working on injection systems?
Click to expand...

I am and attended the Fuel Injection 101 class that Don Rivera (Airflow Performance) provides. He has been working in the business since the 60s when the Bendix system was invented.
 
Raymo said:
I am and attended the Fuel Injection 101 class that Don Rivera (Airflow Performance) provides. He has been working in the business since the 60s when the Bendix system was invented.
Click to expand...
Then something's haywire somewhere. If we check the TCDS for the IO-540, which uses the Bendix/RSA system, here: rgl.faa.gov/Regulatory_and_Guidance_Library/rgMakeModel.nsf/0/ed61cf457fee2dc686258250006ac097/$FILE/1E4_Rev_26.pdf
we see on page 7 that the maximum inlet pressures to the injectors varies from 5 psi in the B1B5 model up to to 26 to 55 psi in all the others. Similar ranges apply to the IO-360 series. Only the IO-540-B1B5 and IO-360-B1A/B1E engines use really low pressures, and they have the Simmons 530 injection system, not the Bendix/RSA. They're a different animal. http://www.vansairforce.com/community/showthread.php?t=104752
 
I'll just throw the name Hillborn into the mix, just to see what happens, lol.
 
Jim_CAK said:
I just looked at the video index on that YouTube channel. That is a lot of free content. Thanks for posting.
Click to expand...
Check out "The Finer Points", too. Really good channel that doesn't get much attention
 
You must log in or register to reply here.
Back
Top