# How is the mass of air in the intake manifold calculated in an electronically fuel injected vehicle?

I am considering a two wheeler petrol engine. How is the mass of air entering the intake manifold calculated?

Given that I have:

• Manifold Absolute Pressure sensor
• Intake Air Tmperature sensor
• Engine Temperature sensor
• Throttle Position sensor
• Engine rpm sensor

• You're getting a fair number of downvotes for Engineering.SE. Likely you need to add more detail on exactly what it is you're having trouble understanding or calculating, or what specific problem you are looking for help solving. At first blush based on what I know about engines (which is admittedly little), the calculation of something like this is fairly complex, open-ended, and dependent on possibly dozens of factors. – wwarriner Feb 12 '16 at 4:40
• i am an electrical engineering student ,working on designing an EFI controller for a 110cc scooter engine. i have basic knowledge on how the system works as a block. it is said that the MAP and IAT sensors estimate the amount of air coming inside the manifold. what is the relation between these sensor values and amount of air? how can i make the microcontroller understand the amount of air in the manifold? – prasanna Feb 12 '16 at 4:53

An IC engine is (for these purposes) a pump. It moves a certain volume of gas through on each cycle (one revolution for a 2-stroke, two for a 4-stroke). That amount is reported as the displacement of the engine (110 cc) in your case. Then you need to determine the mass of that air in that volume.You can arrive at that using the ideal gas law and the molecular weight of the air.

$$PV=nRT$$

Values in order: Pressure($Pa$), Volume (displacement in your case$m^{3}$), number of moles, the ideal Gas constant (8.314 $J\cdot K^{-1} \cdot mol^{-1}$) and temperature (of the air $K$). Solve for $n$ and multiply by the molecular weight (~.0288 $kg \cdot mol^{-1}$ for air). That'll be the mass of air.

This is method ignores the volume taken up by the fuel itself, pressure loss between the intake manifold and cylinder etc.. It's a pretty good estimate, but won't be accurate enough to make the engine run acceptably well. You'll need to apply an empirical coefficient to tune it to the right value to control your fuel injection. That coefficient will not be constant over all RPM and other operating conditions. The good news is that the coefficient will be close to unity and won't change (significantly) over the lifetime of the engine.

That tuning process can be involved. Ideally, you'd have an air-fuel ($\lambda$ sensor on the exhaust to measure how far from stoich the mix is. Tuning with an exhaust gas temp sensor is also possible (never had to do that myself), and may be cheaper. The last option would be to tune the engine on a dynamometer (measures power), although tuning to produce maximum power will usually produce a slightly fuel rich mixture. The $\lambda$ sensors are the best way to go and can be integrated into your controller to make the system more robust, but you could borrow one temporarily to tune and run the engine without it.

Notes:

• I've specified everything in base SI units. Other units will work too, but make sure that you get the value of $R$ right and use an absolute temperature (Kelvin or Rankine, not Celsius or Fahrenheit)

• Engine temp and TP are not strictly necessary for the calculation, but TP is often used to tweak fuel delivery during changes in speed/TP.