Non-ideal MAT correction
 

MAT correction is driving me crazy. Setting it to 100% is DEFINITELY not working. Around 60% or so it seems to be doing OK for idle, but at higher IATs it seems to be too rich going WOT.

So here's a thought: Just like everybody else here, I'm measuring IAT at the outlet of the IC. And we assume that this is the same temperature we see in the IM. But that's not the case, if my thermodynamics professors weren't lying to me. There's a throttle plate in between. And at that plate, if it's not fully opened, we should see a cooling of the charge. And if memory serves me right, that effect changes with temperature, pressure and especially the pressure change.

Has anybody ever checked, if this cooling effect is severe enough to mess with our A/F ratios? If it's a constant factor, not depending on the temperature, it probably won't matter. For example if it's always just cooling the charge by a factor of 0.98 x T1. But if we get more or less of a cooling effect at higher temps, it will matter.

I'm probably waaay off base here, but I'm desperate....

Did your professor also teach that in reality gases do not behave "ideally"? pv=nrt is only a model.

Move your IAT sensor to avoid heatsoak from the radiator, and search for Brain's zeroed mat correction curve and/or use a Stim to zero the curve.

Air is not an ideal gas. Got it.

I've tuned last time at outside temp of 70*F. And made sure that I only allow VE to tune when IATs are below 75*F. So my VE table is correct for the 72*F that MS assumes as 100% (it was 72*F- wasn't it?)

It's not heat-soak. Even when displayed IAT is within 3 *F of outside temperature, the algorithm would pull way too much fuel at higher temperatures and at idle if the correction were 100%. We also had a few colder nights and at 100% MAT correction it would be running too rich. So for idle and low intake pressures about 60% correction seems to be working just fine. Of course the temperature range I had available for tuning so far is 60*F to 92*F or so. Not a huge range.

The observation stands that at WOT the car seems to require more correction than at small throttle openings. My best guess is that at WOT 80%-90% would be more adequate. Of course for WOT on a turbo, the reaction time of the IAT sensor is important, so there's another factor I can't easily compensate for.

Answering the next question: No. I don't have any conclusive logs. Especially none that include outside air temperature.

Why would the charge cool at the throttle plate? I doubt the air from the IC decreases in temp until it mixes with fuel inside the IM.

If anything the TB should be at a higher temp (ignoring warm up conditions) than the air moving across the plate.

Because air cools down during expansion. The pressure behind throttle plate is lower --> expansion --> lower temp.

See:
Isentropic process - Wikipedia, the free encyclopedia

Or if you don't like simplifications:
Polytropic process - Wikipedia, the free encyclopedia

Ok, fair enough. Assuming ideality again though.