Injectors - Pulse Width to Volumetric Efficiency

[AdamQ]

I have an ECU that uses Volumetric Efficiency (VE) rather than Pulse Width (PW) to control fuelling.

I have suitable maps from other ECUs (thank you to kind people on here) that contain PW figures.

I obviously need to convert PW to VE (as a starting point at least) and that is my question - how?

The ECU manual gives (for Alpha-N fuelling):

PW = INJ_CONST * VE(tps,rpm) * AirDensity * Corrections + AccEnrich + InjOpeningTime

where INJ_CONST is "a constant for the given size of injectors, engine’s displacement, pressure 100kPa, temperatures of the intake air 21°C, VE 100%, time of injectors’ opening required to obtain the stoichiometric mixture (Lambda = 1)"

The engine is a 2.0 Duratec and the injectors are Bosch 0280 156 095 items which are, I believe, 315 cc/min (242.5 g/min).

I've Googled away and have searched the ECU forum, but have found nothing definitive (or anything that gives realistic figures - my first check gave a VE of 349% which seems somewhat unlikely).

Any assistance would be much appreciated.

Adam

[aerobod - near CYYC]

I would say that the formula for max VE = (max torque / capacity) * 1.4 is a starting point that would yield 105% for the Duratec 210bhp / 150lbft engine, good race engines will typically be approaching 130%, so as a sanity check, the R500 would be 123%.

The injector pulse width for the map I sent you is 14.76ms at the torque peak around 6300RPM at WOT. At that point lambda is around 0.85, duty cycle (for the stock Caterham R400 injectors) is 77.5%. This would lead to VE (in %) = PW (in ms) x 7.114, that can then be used as a conversion factor for the whole map. The injector opening time of 1ms or less could be applied as an offset to the PW based map first, but there is enough lambda padding at high loads that not doing this for the initial map shouldn't put the engine into a dangerous lean situation.

I would use that as a good starting point to scale the rest of the map. Then use a wideband O2 sensor to confirm readings.

[AdamQ]

Hi James,

Many thanks indeed for your comprehensive response.  That all makes a whole heap of good sense.  I've converted the values as above and all looks good.

My ECU has the facility to allow for injector opening time (as a function of battery voltage) and that's 1.13 ms at 12 V and 0.94 ms at 13 V so that would appear to be about right.

Fingers crossed ...  Thanks again,

Adam