How does petrol injection work?

[Jan Olieslagers]

I have a fairly good understanding of how engines work, both carbureted or diesel. But I cannot see how, in an injected petrol engine, there is still mention of a mixture setting. In a diesel, one sends an ample quantity of fuel to the injector, and a return line takes care of the excess, simple and efficient. Must I understand that, in an injected petrol engine, the quantity of fuel injected is metered in function of both the throttle setting, and the mixture setting compensated for air pressure?

Apologies if the question is dumb. Pointers to a good explanatory web site are just as welcome, it's not that I don't want to do any thinking on my own.

[Capetonian]

HowStuffWorks "How Fuel Injection Systems Work" (http://auto.howstuffworks.com/fuel-injection.htm)

[mm_flynn]

Jan,

My understanding of a road diesel is that they always run LOP (except when their airfilter is clogged and they are belching out black smoke) and the power is set by varying the amount of fuel introduced into the cylinder using the fuel servo (i.e. mixture control) and this is controlled by the power peddle (frequently and incorrectly called the throttle as it does not throttle the intake, but may throttle the exhaust to simulate the same peddle dynamic of a petrol road vehicle).

For an injected road vehicle, (electronic fuel injection) typically fuel is injected into a point in the intake manifold (or sometimes into the cylinders like aero engines). The amount of fuel injected is set based on the oxygen level in the exhaust to maintain a target mixture. As far as I can tell, this target is a rich of peak target in auto's. The power is regulated by throttling the intake airflow.

In aero engines we don't have computer mixture management systems. We also have two very different operating modes. A - Maximum sustainable power without regard to efficiency (i.e. climb power), B- Cruise power where we want maximum economy. A is set by default (all levers forward), B- Is set by hand. A further difference between auto and aero engines is that the air volume in an aero engine changes substantially without any change in engine settings as we climb and descend.

[Jan Olieslagers]

OK, think I got it:

-) there is still a (butterfly?) valve regulating the ingress of air into the manifold, under control of the "power" lever
-) in the manifold, fuel is injected, mostly close to each individual inlet valve
-) fuel injection is metered to produce a given air/petrol mixture, as determined by manifold pressure, temperature, and probably the setting of a "mixture" lever.

Modern engines have electronic ECU's to control all this, though I understand early injection engines (RR Merlin?) did this mechanically.

Thanks for the replies!

[FlyingStone]

For an injected road vehicle, (electronic fuel injection) typically fuel is injected into a point in the intake manifold (or sometimes into the cylinders like aero engines). The amount of fuel injected is set based on the oxygen level in the exhaust to maintain a target mixture. As far as I can tell, this target is a rich of peak target in auto's. The power is regulated by throttling the intake airflow.

Most aircraft piston engines, except new turbodiesel engines use indirect fuel injections into the intake port by means of a continuous fuel flow through the injector. If it would be injected directly into the cylinder, it would require precise timing and amount of the fuel injected, which would almost in all cases require electronically controlled fuel injection - not something you'de see in a 1960 aircraft piston engine design :)

Rich of peak target in automotive engines is changing nowadays, so in latest VW FSI engines, which use stratified fuel injection, target is quite well lean of peak EGT - comparable with those of diesel engines. But the piston design, valve design and timing (variable) are much more advanced than your usual Lycontinenal.

My understanding of a road diesel is that they always run LOP (except when their airfilter is clogged and they are belching out black smoke)

Modern turbo diesel engines run quite deep LOP at low to medium loads (e.g. while cruising), but if you put the "pedal to the metal", ECU recognises throttle position via the potenciometer as 100% demanded load and adjust mixture to correspond to rich of stoichiometric to achieve maximum available power. You can see that the mixture is indeed rich in the color of exhaust gases, which is almost pure black (unless the car is equipped with diesel particulate filter - I think it's required from Euro IV on) - indication of partially burn fuel in the exhaust - something that wouldn't happen if running LOP.

The simplest explanation for fuel injection in aircraft engine works goes like this. You have an engine without carburettor, but you retain the throttle valve, which allows you to control the mass air intake flow to the engine (of course, you only see the "pressure", not the actual mass flow). The fuel goes from the fuel tanks via the fuel selector to the fuel pump (usually backed up with an electric fuel pump, which can be low-speed or combined low and high-speed). From fuel pump, fuel goes to the fuel manifold (usually on the top of the engine), which then splits the fuel line into smaller lines, one of which goes to each cylinder (4-cylinder engine has 4 outlet ports in the fuel manifold + 1 for controlling fuel pressure). The lines that go from the fuel manifold end up in a fuel injector, which is inserted into the intake port of each cylinder and provides continuous fuel flow when the engine is running. Each injector is supposed to be (at least TCM and Lycoming say so) tuned so the fuel flow from the injector matches the air intake flow through each intake port, but in reality they are usually quite far off (unless you have GAMI injectors). The fuel flow through the fuel injector depends on the throttle and mixture position if you have TCM engine, which basically means: full throttle and full rich mixture (takeoff power) = maximum fuel flow. If you use Lycoming engine, the thing gets a bit more complicated, since the injectors try to adjust for increasing altitude and thus reduced air pressure and density by "comparing" air and fuel pressure, but it's far from what one would consider optimum - if you climb with all three levers forward, the Lycoming engine would still run richer and richer and the TCM engine would be just a bit worse.

[ShyTorque]

Electronic fuel injection for spark ignited/petrol road vehicle engines uses a feedback loop, ie. a voltage from an oxygen sensor in the exhaust system is fed back to the electronic controller.

The more oxygen in the exhaust (= a lean mixture, not enough fuel put in) the higher the voltage produced by the sensor and the controller holds open the electronic injector(s) for longer to bring the mixture back to where it's required.

Unfortunately the oxygen sensor cannot tolerate leaded fuel (lead salts from combustion contaminate and "kill" it). So this type of system can't be used on an AVGAS burning engine and the pilot needs to get more involved in setting the mixture for himself.

AVGAS still contains a lot of tetraethyl lead, despite the "LL" designation. It's only low lead compared to what it used to contain. To put this in perspective, it contains about five times the lead content of "old fashioned 4 Star" petrol. For racing purposes only, of course, one gallon of AVGAS in a five gallon tank of cheap pump unleaded gives you a similar octane rating to that of 4 Star. AVGAS must not be put in a modern catalytic convertor equipped engine because it will also ruin that, as well as the O2 sensor.

[ShyTorque]

Silvaire, yes, correct but I disagree about the effects of disconnecting (or a failure of) the O2 sensor.

I once went for a test drive in a BMW, with a view to buying it from the garage. It was immediately noticeable that there was something not right about the performance of the car, despite there being no OBD fault code showing. The effect was a "surging" sensation, then a hesitation under acceleration. I agreed to buy the car but insisted that the garage would be responsible for fixing the problem. They agreed, but all they kept doing was checking the fault codes and they were stumped. I suggested they should change the O2 sensor but they couldn't see that this would make any difference. To cut a long story short I eventually bought a new sensor myself, took it to them and asked them to change it, or the car would go back to them under the Sale of Goods act. I also said that if it cured the fault I would expect them to reimburse me for the new sensor. They agreed but were very sceptical. They fitted the new sensor, the car's performance was immediately transformed and they reimbursed me. The old sensor was probably still putting out a very weak voltage so the OBD didn't pick it up as a fault.

However, we were originally discussing "mechanical" aircraft fuel systems designed and built almost forty years ago. :)

[OpenCirrus619]

Take a look at: Continuous Flow Fuel Injection (http://www.csobeech.com/files/ContinuousFuelFlow.pdf)

It explains the approach Continental took.

OC619

[rasti121]

One thing I don't understand is why Rotax engines don't have mixture "knob" yet still can climb to 12000 feet while delivering reasonable power (could climb more) - I have been demonstrated this (to my amazement) in an ultralight (Eurofox). Obviously there is some kind of mixture control, else the amount of petrol would stop the combustion at FL100 air pressure (about 75% of sea level), but I'm not aware of any FADEC or similar to do the job.

[IO540]

In the Bing implementation the slide is sealed by a very thin rubber diaphram and if it cracks the cylinder(s) being fed by that carb make(s) no power.

That is reportedly a not unknown failure on the Conti altitude compensated fuel servos.

The engine stops.

You need to be quick on the mixture lever and pull it right back.

And it still doesn't work properly, because while it is doing altitude compensation, one actually needs to be measuring mass flow of the air, in order to maintain the correct stochiometric mixture.

Metering mass flow on the fuel is trivial because it is liquid, and any constant volume pumping device (e.g. a gear pump) will deliver a constant mass flow. But measuring mass flow on the air is a lot more complicated if you want to do it to the accuracy required to maintain stochiometric combustion from SFC to say FL120 or more.

One can use a simple altitude compensated carb to make the engine work ok to FL120 without any pilot adjustment, but it won't be running optimally.

[ShyTorque]

UK Triumphs used Stromberg CV carbs, not Zeniths. But yes, a split diaphragm means idle rpm only because there is no direct connection between the throttle butterfly and the actual airvalve. I preferred the SU due to the lack of the diaphragm and still have a car using one of those carbs and three spares in the garage, for other projects.

The SU will be replaced by fuel injection system very soon, I'm working on that project right now.

I've still got a spare Stromberg diaphragm in my cupboard from a Hillman Hunter that I sold in early 1982! Probably a bit crispy by now :O

One of my favourite words, diaphragm. :ok:

[ShyTorque]

In UK, where the Triumphs were actually made, Strombergs were known as Strombergs. Zeniths were Zeniths, and were non-CV, if you get my drift. That Wiki article does actually show a photo of a set of Strombergs, and names them as such.

[miroc]

I have a fairly good understanding of how engines work, both carbureted or diesel. But I cannot see how, in an injected petrol engine, there is still mention of a mixture setting. In a diesel, one sends an ample quantity of fuel to the injector, and a return line takes care of the excess, simple and efficient.Diesel engines use QUALITATIVE regulation, petrol engines use QUANTITATIVE regulation.

Qualitative regulation in diesels regulates the power output by amount of fuel injected. The amount of air is always the same. There is nothing like incombustible mixture like with the petrol engines. Mixture knob is not needed at all. The return line is for a different reason here compared to return line in petrol engines.

Quantitative regulation in petrol engines regulates the power by reducing the amount of induced air. The amount of gas has to be reduced too to keep the mixture combustible. With reduction of air mass the amount of gas is reduced automatically in carburetors by their construction (venturi - less air, lower speed, less suction, less fuel). The imperfection of this automatism and the wish of pilot to run richer/leaner can/shall be set by mixture knob.

The injected engine replaced the one "injector" in carburetor by one injector per cylinder just before the induction valve. The compensation for air mass is made differently. The mixture knob is still here for the pilot to decide how rich the mixture should be.

Must I understand that, in an injected petrol engine, the quantity of fuel injected is metered in function of both the throttle setting, and the mixture setting compensated for air pressure?Throttle setting gives the amount of air allowed to enter the cylinders. The technical realisation is by changing the pressure (rather suction) of induced air. To this mass of air (number of oxygen molecules) the amount of gas is adjusted by some automatic mechanism. Otherwise a simple increase of power would be a very complex task for the pilot to keep the mixture in combustible region.

The mixture knob compensates the imperfection of this mechanism and allows us to adjust the mixture more precisely.

Miroc

[Jan Olieslagers]

Diesel engines use QUALITATIVE regulation, petrol engines use QUANTITATIVE regulation. Hehe, must be an engineer to apply that kind of vocabulary. (meaning nothing but good!) I think the answers were already quite sufficient, but it is nice to see the same facts confirmed in a quite different wording - or so I read you. Thanks indeed! But none less to the other responders, of course.