How Fuel Injection Works

By Jeff T. aka JETHROIROC (

Reprinted by permission of

Italicized characters have been added by AFI to elaborate or further describe what the author has published.


I. Introduction:


1. Fundamental needs of a four-stroke IC engine – Fuel and air in the correct proportions accompanied by a reliable and properly timed spark.


2. Engine control A means of meeting the above needs of IC engines.

· Past Mechanical, Thermal-Fluid

· Present Electrical, Computers


3. Motivation for computer controlled engines

· Increased fuel economy

· Environmental regulation (EPA, 1970s)

· Optimum performance as restricted by environmental concerns

· The evolution of solid-state electronics

· Improved driveability/reliability

· System failure diagnosis/warnings of engine malfunction (generally via Service Engine Soon dummy light, but control problems especially easy to pinpoint by technicians equipped with diagnostic computers)


4. Engine systems that are electronically controlled/monitored

· Fuel/Air delivery system

· Ignition system

· Exhaust system



II. Greater Engine Control Subsystems:


1. Fuel/Air This system is equipped to determine intake air mass flowrate, and subsequently control fuel metering so as to ensure a stoichiometric AF mass ratio (14.7:1) in each cylinder during closed loop operation, although there is no AF ratio that minimizes all harmful combustion byproducts simultaneously (but most are optimized at stoichiometric conditions). This system seldom allows greater than a ±1.0 deviation from stoichiometric conditions instantaneously, and the average AF ratio is maintained to within ±0.05 on most modern automobiles equipped with three-way catalytic converters.


Components of modern Fuel/Air Regulation Systems:


a. Throttle Position Sensor (TPS) – Generally uses a potentiometer device to measure the instantaneous position of the throttle plate, which is mechanically linked to the accelerator pedal. This sensor is almost always positioned on the throttle body itself. Accordingly, conditions of hard acceleration and heavy engine load or deceleration will be detected by this device and the fuel/air system will increase or decrease fuel injector pulse duration accordingly. This action is “overriding” in the sense that it allows rapid maximum engine performance in the event of an evasive maneuver as dictated by the driver, at the expense of emissions and fuel economy control. Such action is permissible by the EPA, mainly for safety considerations. Note: Throttle angle may also be used in conjunction with vehicle and engine speed to initiate idle speed control operation of the fuel/air system. A bypass (IAC) valve is generally used to permit the intake of combustion air under closed throttle conditions.


b. Mass Air Flow Sensor (MAF) – Located in the air intake duct between the filter element and the throttle body, input from this sensor regulates the amount of fuel to be metered to each cylinder in an attempt to achieve the stoichiometric ratio. A derivative of the hot-wire anemometer (a heated wire cooled by the air passing over it), using a Wheatstone bridge and a variable-resistance heated element the MAF can produce a near-linear signal from which the mass flowrate of air is easily determined by the engine control module. The greater the mass flowrate of air over the sensor, the greater the voltage required to heat the wire element. The actual airflow measurement is likely the most important variable in determining the amount of fuel to be metered to the engine. Although this device is very accurate, it tends to be somewhat delicate and expensive.


c. Manifold Absolute Pressure Sensor (MAP) – Used on all “speed-density” systems (those not measuring airflow directly), this device measures the absolute pressure of air in the intake manifold using a silicon diaphragm strain sensor and the piezoresistivity phenomenon. The output of the MAP sensor is used in conjunction with intake manifold air temperature, engine displacement, RPM, exhaust gas recirculation amount and various constants to determine the amount of incoming combustion air and consequently the amount of fuel to be metered. Closed throttle plate would represent close to a vacuum in the intake manifold, whereas wide-open throttle should be near atmospheric pressure, the maximum intake manifold pressure for a normally aspirated engine.


· Speed-density method used in MAP systems is described as follows:

Recall that the mass flowrate of air is represented by the product of the air density and its volumetric flowrate. The instantaneous density is calculated by multiplying the density of air at standard conditions by the ratios of MAP and MAT data with respect to standard atmospheric conditions for temperature and pressure. The volume flow rate of air into the engine is simply the product of engine RPM and half of the displacement (it only draws one half the total engine displacement in one revolution under ideal conditions). Figure in corrections for exhaust gas recirculation and other small factors, and the mass flowrate of air into the intake manifold at any given instant is easily determined by the onboard engine computer.

· Note: Some vehicles are equipped with both MAF and MAP systems, relying on MAF data unless a malfunction is encountered, whereupon the engine control module will default to the speed-density method.

The dual systems also use MAP as an instantaneous BP sensor during the start up operation of an engine. Without the MAP sensor Barometric pressure is inferred into the ECM since there is nothing to actually measure it.

Another use of the dual system is for transient fueling operations. MAF systems without a MAP sensor have no way of knowing what the actual pressure of the manifold is. For this reason manifold pressure again has to be inferred through a series of very sophisticated software. Even with the best of software, inference of manifold pressure is not exact. A MAP sensor in the system allows for actual measurements of manifold pressure to be used along with the actual air flow measurement of the MAF sensor.


d. Manifold Absolute Temperature Sensor (MAT) – Measures the temperature of incoming combustion air to be used in those systems performing speed-density calculations to determine airflow.


e. Fuel Injectors (FI) – Solenoid actuators that deliver atomized fuel to cylinders based upon other sensor input, mainly MAF (or airflow by speed-density method) and engine crankshaft position (CPS). The volumetric flowrate allowed by fuel injector nozzles is essentially constant and determined by the pressure of the fuel system itself. Therefore, the amount of fuel actually delivered by injectors is regulated only by the duration in which they spray fuel, called the “pulse width.”


· Throttle Body Injection (TBI) This system delivers fuel in much the same way as a carburetor, generally above the throttle plate at the top of the intake manifold (usually right under the air filter cover plate in the center of the filter element). Actual fuel delivery, however, is accomplished by the engine control module and two or four fuel injectors. Therefore, this system is described as a “wet” system, as the fuel and air must travel through the intake runners together. Accordingly, this can cause some of the atomized fuel to settle out (condense), leading to somewhat ineffective and uneven charge delivery to the cylinders. The biggest advantage of this system is that the fuel is precisely metered each time, without the physical sensitivity of a carburetor.


· Multiport Injection (MPI) This system locates one or two injectors directly above each intake valve, where the fuel can be delivered very precisely. A vehicle equipped with MPI is said to have a dry system, as only air must travel through the intake runners. High fuel pressure (about 65 psi in the system and 40 psi at the injector nozzles) is also applied to sufficiently atomize the fuel discharged by injector nozzles. As one might guess, fuel condensation is eliminated by this system, resulting in more power, better throttle response and increased fuel economy. The only drawback of this system is the increased cost and complexity of a vehicle equipped with at least one injector per cylinder. Otherwise, MPI systems are superior to both TBI and carbureted systems.


· Sequential Fuel Injection vs. Batch Fire Although these are not physical injector configurations, the manner in which injector pulses are dictated is very important to engine performance and environmental variables as well. A sequential fuel injection system triggers one injector at a time following the firing sequence of the engine. Batch fire systems trigger multiple injectors simultaneously, sometimes grouping cylinders to receive fuel in “banks.” Due to the fact that batch fired injectors pulse more than once per cylinder cycle (usually twice), only half the fuel is delivered at a time. Essentially, the first pulse of fuel is fired with the intake valve closed, and then a second pulse is released just when the valve opens. SFI systems are more precise and optimize all engine performance characteristics, although such systems require more involved electronic controls.


Þ Sequential, multiport fuel injection (SMPI or SFI) is the most sophisticated means of fuel delivery as of now, and many newer vehicles are equipped with this system.


f. Ignition System (IGN) – Must provide an electric spark of the proper timing using intake manifold pressure data, engine RPM, crankshaft position and temperature measurements. This system is included here as it sometimes is not controlled by a separate module, simply because many of the important ignition timing calculation factors are stored/determined by the fuel/air system.

g. Oxygen Sensors (EGO) – An integral part of the system’s closed-loop feedback control once heated above 300oC, oxygen sensors most commonly utilize zirconium dioxide (ZrO2) for its tendency to attract oxygen ions and are generally located in more than one section of the exhaust system. To achieve the most accurate results, EGO sensors should be located at the first point where they will receive a multicylinder mix reading (usually in the tubing just beyond the exhaust manifold, before the catalytic converter), and some vehicles have more than one EGO sensor in different locations (exhaust manifolds, one past the catalytic converter). This device generates a voltage based upon the concentration of oxygen in engine exhaust and sensor temperature which is then used to indirectly relay the fuel/air system’s effectiveness in achieving the stoichiometric AF ratio, operating as a correction factor to the MAF data. It is also notable that heated EGO sensors are now used on many vehicles, allowing closed loop operation and thusly optimum system control to begin much sooner after startup.


h. Coolant Temperature Sensor (CTS) – Determines the temperature of engine coolant via direct insertion with a thermistor, usually threaded into a coolant passage in the intake manifold. This data is then used to determine the point at which the engine is warmed and a leaner mixture may be used by the fuel/air system in an open loop fashion prior to oxygen sensor warm-up. Coolant temperature is also used during engine cranking to set the starting AF ratio to a value between 2:1 and 12:1.


i. Crankshaft Position Sensor (CPS) – Recall that one complete engine cycle (four-stroke) requires a 720o rotation of the crankshaft. The crankshaft angular position can be measured as referenced to top dead center (TDC) for each cylinder, generally via magnetic or optical means. The camshaft may also be used as an indirect measurement of crankshaft position, as it rotates at ½ the crankshaft speed. Crankshaft position data is then used for ignition timing and fuel delivery timing, and may also be used to determine engine speed.



2. Ignition/Spark – Must provide a reliable and properly timed electric spark to each cylinder to ignite combustion reactants and promote proper flame propagation rather than detonation. Ignition of combustion reactants takes place before top dead center of the compression piston stroke. The ignition system operates most effectively at the maximum best torque (MBT) decided upon by engine RPM, crankshaft position, temperature and manifold absolute pressure data. The spark advance is measured in degrees before TDC, and must vary according to the type of fuel used as well as those variables previously mentioned. When the spark is advanced too far, autoignition (detonation, “knock”) of some fraction of the fuel and air mixture may occur. Recall that autoignition is generally caused by one of two things, although there are many others; fuel of an octane rating that is too low for physical engine parameters (compression ratio), or excessive spark advance. The ignition system must maximize performance under fixed AF ratio conditions as dictated by the fuel/air system. It can either function as a separate unit, or as an integrated system within the fuel/air system.


a. Crankshaft Position – Provides the direct timing signal to the ignition system and all other sensor input is essentially an elaboration on this value. Obviously, the ignition system must know the actual engine position before any spark advance can be computed!

b. Manifold Absolute Pressure – Contributes to the overall calculation of spark advance, which is generally reduced for an increase in this variable. This value is applied to a table in read-only memory (ROM) of the engine control module to determine an appropriate advance correction factor.

c. Coolant Temperature – Used with ROM tables to obtain yet another correction factor, the determination of which is beyond discussion here.

d. Engine RPM – A correction factor based mostly on engine characteristics is obtained from pre-programmed tables in accordance with engine RPM data. As a general rule, spark advance should increase with increasing engine RPM to a certain point (2500 RPM or so) and then remain close to constant in performance engines. It is a known phenomenon that flame propagation speed can increase proportionally with engine speed, but it only does so enough to avoid advance with increasing RPM in racing/high compression engines with increased turbulence in the combustion chamber (especially above 3000 RPM, where the spark may even be retarded at high engine speeds > 5000RPM). In stock cars and trucks that most of us drive (low compression, less combustion chamber turbulence), flame propagation increases much more slowly than does RPM and therefore further spark advance is essential up to around 5000RPM, by either a centrifugal and/or vacuum advance mechanism or electronic control. Although fast (about 1 millisecond), a spark still requires a finite amount of time in which to take place, and an increase in RPM shrinks this “window.” The exception to this is under idle conditions, where the spark must be advanced as well to compensate for longer combustion time under low manifold pressure conditions. At any rate, ignition science literally varies from car to car, and fuel to fuel. There is no exact method for all cars, or an exact method for any one car…only a “best timing” for a given set of conditions.


e. Knock Sensor – This device senses the presence of “knock,” or excessive cylinder pressure via magnetostriction, piezoresistivity or piezoelectric crystal accelerometers. The knock sensor is generally threaded into the cylinder block itself to sense vibration. Accordingly, the ignition is retarded when a knock is sensed and until the point at which knocking ceases. Essentially, the addition of this sensor can provide for closed-loop operation of the ignition system.

Þ You may have noticed that spark advance decreases with increasing manifold absolute pressure, but increases with increasing RPM. Ironically, manifold absolute pressure increases with increasing RPM, and this is the reason for their separate correction factors and combined use. Spark timing is still debatable and far from an exact science, and generally the spark advance at a given instant is simply a compromise between many factors that conflict one another but are summed to provide a decent result.



3. Exhaust/Exhaust Gas Recirculation – A system designed to evacuate the cylinders of spent combustion products and protect the environment from harmful byproducts, including nitric oxides (NOx), fuel remnants (HC) and carbon monoxide (CO), while redirecting a portion of exhaust gases back into cylinders for mixing with fresh environmental air and fuel. Recirculation can greatly minimize the expulsion of NOx to the environment by lowering peak combustion temperature.


a. Oxidizing Catalyst – Use permits a reduction in harmful combustion product emission via an increased reaction rate, thereby allowing better engine performance calibration under strict environmental regulations. May require the incorporation of additional environmental air to operate effectively; the effectiveness of this device is also directly related to temperature.

· Oxidizes hydrocarbons to CO2 and H2O

· Oxidizes CO to CO2

· Reduces NOx to diatomic nitrogen and oxygen


b.Three-Way Catalyst (TWC) Found in most modern systems, the TWC uses a mixture of platinum, palladium and rhodium to reduce all three major harmful emissions concurrently. The efficiency of this device is largely affected by AF, with stoichiometric conditions being the optimum working range. Although fluctuations from 14.7:1 for a finite duration are tolerable, the average AF ratio must be very near stoichiometric. This device is only effective when used in conjunction with a modern fuel/air control system Three-Way Catalyst (TWC) Found in most modern systems, the TWC uses a mixture of

Modern day catalyst systems will “light off” at approximately 500o F. This is achieved usually within the first 30 seconds or so of vehicle operation. A typical converter will run between 1000 – 1200 degrees in most normal driving modes and convert 99+ % of all the pollutants described above. Almost all of the exhaust emissions from a vehicle are generated within the first 60 seconds of operation on modern day vehicles that have been calibrated to the strict exhaust standards of today’s regulations.


c. Exhaust Gas Recirculation Valve (EGR) – Recirculates a controlled amount of exhaust gases back into the intake, lowering combustion temperature and resulting in a profound decrease in NOx even in the event that only small amounts of exhaust gas are reconsumed. Generally uses a solenoid or vacuum actuated valve that is precisely controlled by the engine computer via an exhaust and intake manifold differential pressure sensor (DPS) to provide EGR as a function of engine load. However, a decrease in performance and an increase in fuel consumption are undesirable side effects of this device and process.



III. Putting Things Together – The Modes of Operation:


1. Closed loop vs. Open Loop Control – While operating in an open loop fashion, the onboard computer functions without the input of exhaust gas oxygen sensors, and therefore will use only MAF or MAP and RPM to determine the correct amount of fuel and EGR to be metered, and the proper spark advance. When the EGO sensor warms sufficiently, closed loop control is initiated wherein a correction factor based upon EGO output is applied to the fuel injector pulse duration calculation as made in open loop operation. This is where the fine-tuning takes place.


2. Start Mode The only concern at this point is quick and reliable engine start.


· RPM is set to cranking speed

· Engine coolant is at environmental temperature

· Low AF ratio (2:1 to 12:1)

· Spark timing retarded

· No exhaust gas recirculation

· Fuel economy and emissions not under optimum control


3. Warm-up Mode The main concern at this point is a clean and fast transition from engine start to normal operating conditions.


· RPM may be adjusted by the driver almost instantly

· Engine coolant temperature rises to minimum operating value (before opening of the thermostat)

· Low AF ratio (12:1 to 14:1)

· Spark timing adjusted by ignition control system

· No exhaust gas recirculation

· Fuel economy and emissions not under optimum control


4. Open Loop Mode Fuel economy and emissions controlled and of concern, without the aid of EGO sensors.


· RPM readily adjusted by driver

· Engine coolant is warmed to operating temperature

· AF ratio roughly controlled to 14.7:1

· EGR is used

· Spark timing adjusted by ignition control system

· Fuel economy and emissions controlled without help of EGO sensors


5. Closed Loop Mode Fuel economy and emissions controlled to the closest extent possible.


· RPM controlled by driver

· Engine coolant at operating temperature

· AF ratio controlled closely at 14.7:1 ± 0.05

· EGO sensor warmed sufficiently to enter the control loop

· System resumes open loop operation if EGO fails to operate properly

· EGR system in operation

· Fuel and emissions strictly controlled


6.Hard Acceleration Mode (WOT) Maximum performance and safety of concern in this mode, with fuel economy and emissions of little consideration.

2000 and beyond emission regulations require the manufactures to test and control emissions during high load operation. There is a standard that must be met by all vehicles at this operational mode.

· Throttle plate wide open as dictated by driver

· Engine coolant temperature in normal range

· AF ratio rich (13:1)

· EGR and EGO are not used at all

· Poor fuel economy and emissions control


7. Deceleration and Idle Mode Fuel economy and emissions of primary concern, as is preventing engine stall.


· RPM dropping quickly or constant at idle speed

· Engine coolant at normal operating temperature

· AF ratio lean

· Idle mode engaged to minimize RPM fluctuations in the event that accessories are used by the driver (air conditioning, etc.)

· Emissions sometimes drastically reduced with deceleration

· EGR is not in operation

· Poor fuel economy at idle, but good fuel economy with deceleration

· Protection of the catalytic converter from over temperature conditions. A catalyst must remain below a given temperature to maintain its usefulness. During deceleration operation small misfires can take place in the combustion process allowing for some raw fuel to be exhausted and collected by the catalytic converter. This fuel is then burnt in the catalyst and can potentially over temp. the exhaust system. These misfires also can increase emissions making it necessary to “shut off” the injectProtection of the catalytic converter from ove