Maths question: Calculating Fuel flow from engine data

Frozennova · 28 Nov 2016 at 21:33

I'm currently making a small Pi based Python script to pull engine data from an aftermarket ECU in car and cannot for the life of me anywhere find how I would calculate fuel flow

I have all of the sensor and hardware data (Engine speed, Injector pulse width in milliseconds, injector flow rate, injector dead time) I think I need but im having trouble working out how to get the end result I'm looking for

Diddums · 28 Nov 2016 at 21:36

Careful, you risk danger to the manifold.

Chuk_Chuk · 28 Nov 2016 at 21:45

Fuel flow between what two points?

Frozennova · 28 Nov 2016 at 22:00

Chuk_Chuk said:
Fuel flow between what two points?

Maybe I should have worded it a little differently.

I want to calculate the volume of fuel injected

Danm54 · 28 Nov 2016 at 22:41

Your injectors will be rated at cc's/min at a set fuel pressure, you may be able to find data on flow rates at different pressures too but you need to know what pressure you're running at.

With the cc per minute and pulse width in milliseconds you can work out how much fuel is being injected per pulse if you do some wizardry.

Then see how many pulses you are running per revolution, times the amount of injectors gives you cc per revolution.

Times by rpm and you have fuel volume at whatever rpm.

Maybe....

ianh · 28 Nov 2016 at 22:43

Danm54 - Beat me to it!

Clarkey · 28 Nov 2016 at 22:49

Not really that simple because rail pressure is directly linked to manifold pressure which is going to make your flow rate all over the place in practice. You would need to factor that in.

itchy · 28 Nov 2016 at 22:57

Also temperature, as heat decreases viscosity.

Danm54 · 28 Nov 2016 at 23:02

Heat changes the volume of fuel but not sure about viscosity or flow rate to any noticeable difference.

Not sure how accurate you could get it.

itchy · 28 Nov 2016 at 23:13

Danm54 said:
Heat changes the volume of fuel but not sure about viscosity or flow rate to any noticeable difference.

Not sure how accurate you could get it.

As the heat expands the volume of the fuel and viscosity will decrease, as its on the verge of turning into a gas depending on the pressure.

From what I know fuel flow is more efficient in a warm engine than a cold one.

So what OP is asking has many facets that are way beyond me and I know a lot about nothing.

Frozennova · 29 Nov 2016 at 08:41

Think I have it, or at least something that spits out some sane values anyway

Code:

lph = (RPM * ((injectorFlow / 60000 ) * (PW - DeadTime)) * 4) * 60 / 1000

Plugging some values taken from a data log at part throttle and 31mpg

Code:

lph = (1845 * ((650 / 60000 ) * (2.2 - 1)) * 4) * 60 / 1000 = 5.75

Clarkey said:
Not really that simple because rail pressure is directly linked to manifold pressure which is going to make your flow rate all over the place in practice. You would need to factor that in.

If it works out to be a problem I have the MAP sensor value available to offset the injector flow rate

Edit: forgot to account for the fact that all injectors are fired per engine cycle which is 2 revolutions

So the correct formula should be

Code:

lph = (RPM/2 * ((injectorFlow / 60000 ) * (PW - DeadTime)) * 4) * 60 / 1000