Mass Air Flow — The Science Behind It
We are all living at the bottom of an ocean; an ocean made of air instead of salt water. The earth has a vast ocean of nitrogen and oxygen encompassing it. This ocean of air is being pulled toward the earth by gravity. Gravity is the force that is created between two objects with mass that are attracted to one another. This force is proportional to the product of their masses and inversely proportional to the square of the distance between them. On earth gravity is what gives weight to physical objects. This vast ocean of air surrounding the earth has mass and therefore has weight. The weight of this ocean of air will change depending on the depth of it. Just like an ocean of water; the deeper the water the more pressure is created. This ocean of air exerts 14.7 Pounds Per Square Inch (PSI) at sea level and at 18,000 feet this ocean of air exerts 7.34 PSI, which is only half of the pressure that is created at sea level.
The internal combustion engine uses this air weight or air pressure in order to operate. When the piston moves downward, away from the head, the volume increases thus creating a low pressure area within the cylinder. This low pressure area within the cylinder sets up a pressure differential. A pressure differential is the difference in energy between a higher pressure (atmospheric air) and a lower pressure (cylinder air). High Pressure having more force always moves to a low pressure having less force, thus the surrounding air moves into the cylinder. This inrush of air into the cylinder will be used to operate the engine in several ways. The first is the volume of air (78% nitrogen, 21% oxygen, 1% other) will be compressed creating heat within the cylinder. The second is the air within the cylinder, being comprised of 21% oxygen, will provide an oxidant for the chemical reaction with the hydrocarbon fuel stock. Third the 78% nitrogen and 1% other will be heated by the burning fuel which creates the expansion of the nitrogen, thus forcing the piston downward. This, in turn, produces torque on the crankshaft.
The burning of the hydrocarbon fuel stock within the internal combustion engine is essential. This is what powers the engine so that the pumping losses of the engine and energy needed to move the vehicle can be produced. In order to properly burn the hydrocarbon fuel stock, the weight ratio of the air and fuel will be important. The proper air/fuel ratio to completely burn the fuel stock is referred to as stoichiometric. The stoichiometric ratio between the fuel and air is one where the hydrocarbons and oxygen are at a weight ratio that once they react with one another neither chemical will be present. This means that the hydrocarbons break apart becoming hydrogen and carbon. In the presence of oxygen, the hydrogen combines with the oxygen forming a new chemical; dihydrogen monoxide (H2O water). The carbon attaches to the oxygen forming a new chemical; carbon dioxide (CO2). If the hydrocarbons and oxygen are at a stoichiometric ratio and react with one another then neither of these chemicals will remain present within the combustion gases, see Figure 1. The chemical weight will be the same but the new chemicals formed during a complete reaction will be water and carbon dioxide.
In order to obtain a stoichiometric ratio between the fuel and air, the weight of the air must be known. Since the fuel stock to be combusted is known the fuel weight will also be known. However due to the load of the engine constantly changing the air weight is an unknown factor, therefore there must be a method to properly weigh the air. With the fuel injected gasoline based engine there will be three basic methods used. First is the Alpha-n method, which is the method where the Throttle Position Sensor (TPS) is the main sensor used. This is where a look up table for the throttle effected area is used to calculate the air weight entering the engine. Second is the Speed Density method, which is the method where the Manifold Absolute Pressure (MAP) sensor is the main sensor used. This is where a look up table for the absolute pressure within the intake manifold is used to calculate the air weight entering the engine. Third is the Mass Air Flow Measurement method, which is the method where the Mass Air Flow (MAF) sensor is the main sensor used. This is where a look up table for the air entering the induction system is used to calculate the air weight entering the engine.
In each of these methods an accurate air weight can be calculated. Each of these methods have advantages and disadvantages, but perhaps the MAF method has the greatest advantages. When a thermal measurement air flow device is used there is no altitude error, no significant moisture influence, no pulsation error, fast response time, and no moving parts. The fast response time from these thermal measurement devices will still have significant delay or latency. Additionally it is hard to measure air flow in unsteady conditions, such as during a transitionally event. Therefore during an acceleration the engine control program will not use the signal from these thermal measurement devices.
When using these thermal measurement device air weight is directly measured by the sensor. This is accomplished using a sensor that is located before the throttle plate, as shown in Figure 2. This sensor must be located in front of the throttle plate. If the sensor were located after the throttle plate the turbulent air flow would have a negative effect on the sensor’s output voltage, causing an erratic output voltage that cannot be used. Turbulent air is such a problem that these sensors are equipped with an inlet screen to allow the air moving through the sensor to be straightened before being measured. Some manufactures show a MAF sensor located behind the throttle plate. These sensors are called out as MAF sensors in the wiring diagrams and scan tool data streams but these are really MAP sensors because when the sensor is located behind the throttle plate the sensor is based on the speed density method.
It is important to identify which air weight method is being used. Each of these methods have different effects on the fuel injection system. For example, if you have a single exhaust with a restriction these systems would correct for this problem differently. The speed density system reads the pressure (vacuum) within the intake manifold. The exhaust restriction would cause the pressure to rise toward atmospheric pressure under load. This is interpreted by the Engine Control Module (ECM) that the air weight entering the engine is higher than what is actually entering. Thus the fuel trim would move to a negative factor (taking away fuel). Using the MAF system with this same problem would cause the air flow to drop under load which, since the MAF sensor reads the direct air flow into the engine, would read this drop in air flow correctly. Thus no fuel trim factor would be utilized. Now let’s change the problem to one of an air leak. If the speed density method were used the air actually entered the engine dropping the intake pressure. Since the air volume actually entered the engine there would be no fuel trim correction factor. If the MAF method were used and an air leak were to let air into the engine this air would not be read by the MAF sensor so the unmetered air would have to be compensated for with a positive fuel trim factor (adding fuel). As you can see, these systems operate differently from one another making it imperative that you identify the system correctly in order to accurately diagnose the fuel injection system.
The MAF sensor can be constructed in serval different ways. The first MAF sensors were vane air flow sensors where a mechanical door was pushed open by the air flow entering the engine. The Vane Air Flow sensor is based on the equation for orifice flow, where the kinetic energy of the air flow is converted to pressure energy at the vane. The vane air door is connected to a potentiometer which sends an output voltage to the ECM that is proportional to the door movement, and the door movement is proportional to the amount of air entering the engine, thus the air weight can be correctly calculated. The problem with this type sensor is that they are prone to damage from high pressure waves such as backfires.
Newer MAF sensors use an electrical circuit based on a wind speed anemometer, as seen in Figure 3. This is where a hot wire or hot film device works on the basic principle that the amount of electrical power required to maintain the heated element in a moving fluid stream at a fixed temperature above the fluid temperature is a direct function of the mass air rate past the element. This principle is based on convection. Convective heat transfer is a mechanism of heat transfer occurring due to the movement of fluids, in this case the fluid is air. The heated element is hotter than the surrounding air so when cooler air moves past the hot element the heat is transferred into the cooler air, being carried and dispersed by the air flow.
This may sound complicated but you already understand this principal in the effect of wind chill. Your body temperature is hotter than the surrounding air temperature. If the air is not moving there is not as many air molecules coming in contact with your body. Each air molecule that is colder and comes in contact with your body will take heat on. The more molecules that come into contact with your body the greater the heat transfer. If the air flow has a higher velocity then more air molecules will come into contact with your body allowing more heat energy to be removed from your body.
Now in place of your body put the electric heating element. If the heating element is hotter than the surrounding air the heat will be transferred into the air molecules. Each air molecule that comes in contact with the heated element removes heat from it. If the velocity of the air is increased then more molecules come in contact with the heated element which will remove more heat from the element. In the engine compartment the air temperature can be greater than 250°F. This means the heating element or hot wire will need to be hotter than its surroundings so the transfer of heat can be made from the hot wire to the air flow.
These heated elements will be kept at a constant temperature that, depending on the heating element type used, will be approximately 425°F to 450°F. This is accomplished with a bridge circuit from a platinum hot wire type element as shown in Figure 3. The platinum hot wire is very thin (approximately .0027 inch in diameter) so it can obtain a rapid heat transfer that will help limit the transient delay from the sensor. The bridge circuit is designed so that the heating element is kept at a constant temperature regardless of the air flow across it. When the air flow is increased, removing more heat from the heating element, the resistance of the wire drops. This causes the bridge circuit to become out of balance where the amplifier supplies more current to the circuit in order to keep the temperature consistent. Depending on the air flow, the bridge circuit supplies 500 mA to 1500 mA to the heating circuit. Since the air flow is responsible for removing the heat from the heating element the current will be proportional to the air flowing across it. This current will be read as a voltage drop across precision resistor R3.This voltage drop across R3 will be read by the ECU and converted from voltage to air weight in a transfer function table as seen in Figure 4. When this table is calibrated by the manufacturer the engine head and complete induction system, including the air filter, is used. This assembly is then put on a flow bench where air is allowed to flow through the induction system at different flow rates. The voltage output from the MAF sensor is written into a lookup table and then matched with the air flow weight that was present through the MAF sensor and induction system. When the engine is running the ECU looks up the voltage that is being produced from the MAF sensor and converts it to actual air weight. This method allows an accurate way to normalize the voltage reading to an actual air weight reading.
The density of air changes with temperature change, thus the temperature must be compensated for. This is accomplished with a temperature compensation resistor (Rk). The temperature compensation resistor has a resistance of 500 ohms so very little current will flow through this bridge leg as compared to the heating element bridge leg. The temperature compensation resistor will be made out of the same material as the hot wire; platinum. However the temperature compensation resistor will be configured out of platinum-film and will be located close to the heating element. During production there are variances within the bridge circuit that will require calibration. Resistor R1 will be used to compensate for the temperature compensation circuit and is laser trimmed to provide means to calibrate the temperature circuit. Resistor R2 will be used to compensate the bridge circuit and is laser trimmed to provide means to calibrate the bridge circuit.
Since this sensor is basically a velocity sensor the air speed is critical and is controlled by the sensor housing bore diameter. This bore diameter is engineered for the liter size of the engine that will be used. This is why a manufacture can use the same electronic sensor on a 4, 6, 8, and 10 cylinder engine. All that will be needed is to change the sensor housing bore diameter so that the air velocity is the same on each engine. For example, on a 4 cylinder engine the bore diameter will be smaller than that on an 8 cylinder engine. This method saves the manufacture from having to produce a different electronic sensor for each different engine type. Beware that someone can put the wrong sensor on the engine. For example, an 8 cylinder MAF can be installed on a 6 cylinder engine. Now the air flow velocity will be slower allowing the MAF sensor to produce less output voltage. Thus the voltage read from the MAF sensor and used to convert the air flow weight on the transfer function table will be incorrect. When this happens the fuel trim will compensate with a positive fuel trim factor that will be linear from idle, to light load, to full throttle as shown in Figure 5 (Note: the positive fuel trim factor is 21% across the engine loads). To verify this problem a Volumetric Efficiency (V.E.) test will need to be run.
A volumetric efficiency test is the best way to test the MAF sensor. The ECM takes the voltage produced from the MAF sensor and then normalizes it to an air flow weight that is used to run the engine. This MAF air weight is available on the scantool data steam. If the Engine Liter Size, Revolution Per Minute (RPM), Barometric Pressure (BARO), Air temperature (AT), and MAF air weight are known then a calculation can be made to estimate the correct air flow to the engine. For example, if a 4.4 liter engine turned two revolutions and pumped 100% of the volume it would move 4.4 liters of air. However, other factors come into play such as the BARO, AT, RPM, and the actual engine flow efficiency. These factors will reduce the V.E. from the naturally aspirated engine where (at sea-level, at 70°F AT) they are only about 89% efficient. Once these factors are known a very accurate engine model calculation can be made.
A V.E. test is then run as shown in Figure 6 on the same engine that is shown in Figure 5. This V.E. test shows the actual MAF reading in yellow and an engine model (V.E.) shown in red. This shows the actual air weight is low by 21%, and this is the same amount as the fuel trim correction factor. Remember the ECU reads the air weight first and then bases its fuel ratio calculation on this air reading. If the air weight is incorrect then the fuel delivery rate will be incorrect as well. The fuel trim factor will need to make up for the incorrect air reading that was made by the MAF sensor. This is why the V.E. test shows 21% low, and a fuel trim correction factor of 21% corrects the air weight. This is accomplished by the fuel trim correction factor multiplying the actual air weight reading, and thus corrects for the incorrect air weight.
The output voltage from the sensor can be produced in an analog format as seen in Figure 7 or that of a digital format (Note: when a digital frequency is used the transfer function table is built using frequency instead of voltage). In Figure 5 a snap throttle event from a voltage output MAF sensor has been recorded on an oscilloscope. During a snap throttle event a high air flow rate can be obtained due to the pressure differential that is created from a closed throttle plate at idle. When the throttle plate is snapped open the inrush of air into the engine can be measured as shown in Figure 7 (Note: if a Drive By Wire system is used this may not open the throttle plate to Wide Open Throttle (WOT). This inrush can be seen as point B, this is the maximum Voltage output from the MAF sensor during the snap throttle event. Use the scopes cursors and make a measurement on the voltage waveform from the idle (point A) to WOT (point B). This is the voltage difference produced from the MAF sensor. This voltage change should be greater than 2.5 volts, which is half of the 5 volt supply to the sensor.
When the engineering team produces the MAF sensor they want enough voltage change to calibrate the engine with. This voltage output from the sensor is broken down into increment that will represent the air flow from the engine. Depending on the engine type, if the engine went from idle to full load the air flow may change 250 Grams Per Second (GPS). The greater the MAF voltage change the more points can be used to calibrate this air flow. If a 5 volt power supply is used the calibration team will want at least half of the power supplied to the MAF sensor for calibration purposes. This means the minimum voltage change during this snap throttle event should be greater than 2.5 volts. In Figure 8 the MAF sensor was change on the engine from Figure 7. Now the minimum voltage change is 2.75 volts instead of the 2.33 volts shown in Figure 7. This new MAF sensors higher voltage output now matches the transfer function table in the ECM, thus fixing the fuel trim DTC’s. During your day take known good vehicles and snap test the MAF sensor on your scope. Keep the recording of this voltage change for future diagnostics. Just a little knowledge on how these MAF sensors operate, will provide you with an easy diagnoses on your next MAF sensor engine problem.
Thanks, i love reading your posts with all of the facts and information!
Thank you Bernie....as always, simply awesome.
Great info thank you, I just got the eScan elite, playing with it everday almost in every car possible. Waiting for the next class available from worldpac, the pressure is on so I can sign up
I love it! I expect you will also!
Thanks Bernie. I noticed that you mentioned that an engine is about 89% efficient at sea level. I have your eScan and I noticed that it only reads in %. Most of my techs scan tools only give them pids. I teach them to take MAF, AT, Rpm, and Fuel Trim on a full throttle pass. Then use a VE calculator. With multiple valve and VVT I'm seeing many cars go over 100% VE. Could you add maybe, an addendum about the difference between Fuel Efficiency and Volumetric Effciency?
We have scaled thousands of engines and very few engines that are naturally aspirated have greater than 100% VE @ 100% throttle opening. On average engines have about 89% VE. No doubt some of the newer engines will have a higher VE rating than 89%, but even with these engines very few exceed 100% VE. This is due to the design of these engine will be produced to be most efficient under the expected operational condition of the engine. The average passenger vehicle will run at less than 40% throttle opening for 90% of it life. This means that the engine is designed to run at less than 40% throttle opening. This is accomplished by changing the head geometry. By using smaller valves with smaller port designs the air flow velocity can be increased at lower loads. This will allow the engine to be more efficient under the normal operational conditions of the engine. However these geometry change will decrease the VE volume flow rates under full load high RPM conditions.
Volumetric efficiency is the measurement of air flow into the engine. The airflow rate therefore is the VE and is the percentage that the engine (liter size) can flow air at a given RPM. The actual air flow rate into the engine, against a percentage of expected airflow rate at a given RPM is VE. A 6 liter engine would flow 6 liters of air if the crankshaft is rotated 2 revolutions, this would be at 100% cylinder fill rate. This would be a very high airflow rate or VE percent for a normal (average) engine. A normal airflow rate would be 5.34 liters, this would be at an 89% cylinder fill rate.
Fuel efficiency is based on a given amount of a fuel stock, rated in BTU's, to produce thermal energy that can move a vehicles given weight over a given distance in time. The vehicles internal combustion engine will use the energy from the fuel to move the vehicle weight. This energy will then be rated on how efficiently the thermal energy from the fuel stock was used. Any problem will show up as a low efficiency. Low fuel efficiency ratings can be such problems as; Engine performance problems, Transmission problems, Brake dragging problems, Low tire pressures or anything that would make it hard to roll the vehicle forward thus using more of the fuel stock.
Masterfully explained, as we all have come to expect. Amazing post Bernie!