Fuel Usage from Datalogging - just for 'fun'
Discussion
Hi
Instead of asking questions to solve problems, I'm asking for some maths advice just for a bit of 'fun'.
I would like to be able to know the instantaneous fuel usage (in litres per minute, as we fill the car in litres and the sessions are measured in minutes), and the total amount of fuel used in a session.
I suppose that the integral of the former is the latter.
But how does one accurately work out the former? My data logger logs RPM and final injector pulsewidth, and I know the fuel pressure of 3 bar (I even log it (showing as 3.2 bar)), and the injector flow rate (0.44 litres/minute at 3 bar).
Given that my ECU is working in some form of pseudo-sequential (done without a cam sensor, and just a crank sensor, with some method of the ECU deducing which cycle the engine is one at any given moment), there are two injections per cycle. We know how long each pulse is lasting, and we know the number of pulses per second (derrived from RPM). I am also trying to correct the actual injector opening times by reducing the pulsewidth by 1 to account for the ~1ms injector opening/closing time.
But converting that into numbers is not coming to me. I'm getting confused with units and constants, and can't seem to get numbers that look about right.
The datalogger software can integrate, so I can in theory confirm the figures by comparing the total fuel used from the logger maths to what we measure after any session.
Can anyone help? Has anyone ever worked out culmulative fuel usage from RPM, injector PW, injector flow rate and fuel pressure?
Instead of asking questions to solve problems, I'm asking for some maths advice just for a bit of 'fun'.
I would like to be able to know the instantaneous fuel usage (in litres per minute, as we fill the car in litres and the sessions are measured in minutes), and the total amount of fuel used in a session.
I suppose that the integral of the former is the latter.
But how does one accurately work out the former? My data logger logs RPM and final injector pulsewidth, and I know the fuel pressure of 3 bar (I even log it (showing as 3.2 bar)), and the injector flow rate (0.44 litres/minute at 3 bar).
Given that my ECU is working in some form of pseudo-sequential (done without a cam sensor, and just a crank sensor, with some method of the ECU deducing which cycle the engine is one at any given moment), there are two injections per cycle. We know how long each pulse is lasting, and we know the number of pulses per second (derrived from RPM). I am also trying to correct the actual injector opening times by reducing the pulsewidth by 1 to account for the ~1ms injector opening/closing time.
But converting that into numbers is not coming to me. I'm getting confused with units and constants, and can't seem to get numbers that look about right.
The datalogger software can integrate, so I can in theory confirm the figures by comparing the total fuel used from the logger maths to what we measure after any session.
Can anyone help? Has anyone ever worked out culmulative fuel usage from RPM, injector PW, injector flow rate and fuel pressure?
Looks like I spoke too soon...
I have come up with this equation, and it gives instantaneous fuel flow figures in litres/minute
(('Final Injection PW'-1)/3600000)*(0.44*sqrt('Fuel Press'/3))*2*'RPM'*60
Integrating that (after dividing it by 60 to get litres/second) gives me the culmulative fuel usage in litres.
In my example session I've used, it seems to think we used 13.04 litres. In real life we measured the fuel usage as 13 litres. So I think that's fairly valid!!!
Will check with other sessions during which we measured fuel usage, and see if it agrees...
If you see an error or a mistake please let me know.
I have come up with this equation, and it gives instantaneous fuel flow figures in litres/minute
(('Final Injection PW'-1)/3600000)*(0.44*sqrt('Fuel Press'/3))*2*'RPM'*60
Integrating that (after dividing it by 60 to get litres/second) gives me the culmulative fuel usage in litres.
In my example session I've used, it seems to think we used 13.04 litres. In real life we measured the fuel usage as 13 litres. So I think that's fairly valid!!!
Will check with other sessions during which we measured fuel usage, and see if it agrees...
If you see an error or a mistake please let me know.
Indeed. I think that -1 is pretty much a fudge factor, as fuel temp, voltage compensation, the actual flow rate of the injectors (given that each one is probably slightly different) etc are all variables. And I'm sure there are many more variables than that!
But this is close enough, and seems to be about right so far. Will keep track of how it compares to real fuel usage after a few more race meetings.
But this is close enough, and seems to be about right so far. Will keep track of how it compares to real fuel usage after a few more race meetings.
Indeed I do, via GPS.
I have a channel for both instantaneous mpg (current speed over current fuelling rate), and overall mpg (cumulative distance over culminative fuel use), where the integral of speed is used for distance.
Suggests a session average of around 12mpg (including idling time in the assembly area and pitlane), and a full load mpg of around 6-8.
Always on the look out for interesting maths channels to further my understanding of what the car is doing (or not doing), what the driver is doing, and stuff to help optimise track time and performance.
I have a channel for both instantaneous mpg (current speed over current fuelling rate), and overall mpg (cumulative distance over culminative fuel use), where the integral of speed is used for distance.
Suggests a session average of around 12mpg (including idling time in the assembly area and pitlane), and a full load mpg of around 6-8.
Always on the look out for interesting maths channels to further my understanding of what the car is doing (or not doing), what the driver is doing, and stuff to help optimise track time and performance.
Gassing Station | Engines & Drivetrain | Top of Page | What's New | My Stuff