[feature request] Special EAE handling at motor start

[FranzD]

Hello,

during motor start PW is influenced significantly by EAE (see attached .msq with MS3 1.5.2 default EAE-values). At least on my start logs with varying EAE parameters (default and non def.) there is always a short enrichment followed by an usually troublesome enleanment phase because of wall film build up and down during the transition of MAP from 100 to 30kPa (what is not unexpected).
Workarounds could be to modify EAE RPM corr. to lower adhere2wall in the starting rpm range and / or increase of AfterStartEnrichment value. Both workarounds are not ideal and I really would like to have motor start untouched by EAE.

Could think of an option, to deactivate EAE for x sec after cranking phase ends. When x sec passed, initialisation of the wall film state could be done to exactely match the steady state wallf film at this point in time to start with EAE=100%: WF [ms] = PW [ms] * C_adhere / C_sucked

Thought about that possibly one could argue that EAE has to start calculation at 1st squirt due to it's design to describe wall film behaviour, which physically also happens from 1st squirt on. --> Hence EAE also has to run from the 1st squirt on to proper track wall film build up.
However, EAE does it this way, but influence on PW is obviously deactivated during cranking (good thing to not complicate start calibration even more) --> somewhat a "mixed" initialisation anyways.

Best regards, Franz


Enleanment up to 21% during motor start:

Log_Screenshot.png (40.76 KiB) Viewed 597 times

[pit_celica]

I had the same issue and my fix was to simply set at 0% the rows that are below 1000 RPM in both the "EAE SOC RPM %" and "EAE AWC RPM %" curves, effectively "disabling" the EAE correction when under 1000 RPM.

[FranzD]

Yes, I also had a similar approach. That also means that there is no acceleration correction any more at all up to 1000rpm in your case. Not really what we want, don't we?
The point is that EAE is not freely calibratable -- EDIT "any more" canceled, doesn't make sense END EDIT -- over the whole range (RPM in this case). Always being bound to not corrupt engine start too much.

Indeed, I haven't found a way to calibrate proper EAE at low rpm's in the mentioned range for cold Clt temperatures where wall film amount should be 4 times or more than at EDIT "summer Clt start temp" is wrong, I meant warm engine temp - EDIT END - numbers just guessing from logs, and are individual --> no successful start with such a calibration. Wall film amount then drops with MAP going down to 30kPa from 100 by that much that enleanment is much more severe than in the screenshot above. This COULD be ok if proper wall film built up would have happened during cranking (means that real and simulated amount of wall film fit together at crank phase ends) , but this is not the case because EAE does not influence PW at cranking. Even if it would, would it be practicable tunable?

My coolant temperature correction is very suppressed because of the mentioned issue, although this physically based EAE functionality could really shine there I think.

[Ted75zcar]

I have spent a ridiculous amount of time with EAE, and the solution I use (and recommend) is 0% added, 100% sucked at an engine speed between cranking RPM and hot idle. It takes 2 points, I use 499 and 500RPM. Obviously, 1000 defeats the purpose of EAE unless you idle over 1000.

As far as the error associated with the 0% and 100%, it is inconsequential. EAE will always act until added = sucked, so WF will catch up to where it should be, and quickly.

If you really get into the details, you can tune that initial EAE spike as part of the afterstart enrichment. It actually works pretty well.

Tip: You need to log SD as fast as you can (I use 5msec), starting the log prior to cranking to really see this stuff.

Edit: you can attempt to extrapolate CLT, but chances are you will exceed the hard-coded limits for added or sucked, so the approach I use there is similar to the above, with the transition occurring at 100F. It is such a transient condition and WUE is already open loop there, I let the time based adders handle AE for if I actually want to load the engine while under 100F, which I never do.