Dynamic Airflow changes WOT causing lean issues...what is going here? E67 CTSV2

[adam112]

So Im having an issue with WOT fueling changes...even from one gear to the next....sometimes fuel is within a percent or less, then other wot blasts and the dyn. airflow value changes by a couple percent, and wala, Im lean...(or rich here and there)

I know this has to be a parameter influenced by MAT value, but logs show it only changed a couple deg C from gear to gear in the attached and yet for same maf hz (11600 for ex.) its got different dynamic air values and is pulling lean.

Ive also had very rich hits intermixed, with no direct correlation to ambient temps...though the air is getting dryer...so Im just confused.

Any input is appreciated. My MAF curve needs some smoothing perhaps, but Ive got logs with same airflow model that are tight to commanded, and then swings of a couple percent each way, again with no maj diff in IVT, ITT, or MAT jumping out.

Running an intake with 7:1 honeycomb 3 in in front of MAf with a hard 90 from there to behind fog light...

[sgod1100]

First off....why are you revving your engine to 6700 rpms?? Is the motor stock besides the intake? That seems extremely high for the stock valve train

[adam112]

First off....why are you revving your engine to 6700 rpms?? Is the motor stock besides the intake? That seems extremely high for the stock valve train

Full port
Flex
Forged rotating assembly (HED built...see Jeff Hagaman V...)
2.7/9.55 making 13ish lbs
Johnson 2110s
GP 2.5 cam
Manton custom 11/32 pushrods
Manton Trunion Upgrade
Shifted at 6200 ish I guess lazily?
Accelerator Position at 6500 is 0%...it over revs when shifted cause manual.

Any help on my actual question??

[sgod1100]

Try zeroing out your prediction coeffiecients. I've read that when modified they can do goofy shit to the a/f ratio

[RobZL1]

+1 to that. Zeroing ALL of the prediction coefficients fixed a few odd issues I was having. Only other thing I can see in your logs is the injector tip temp jumps to the next cell in the chart between pics 1 and 2, and pic 3. In pics 1 and 2, the ITT is right at the 122*F cell, in pic 3 the ITT is right at the 131*F cell.

In the 122*F logs (pics 1 and 2), you are pretty close to commanded lambda, but in the 131*F log (pic 3) you are a few percent richer than commanded. Perhaps you need to remove a little from the 131*F cell in the ITT table?

I wouldn't worry about the error in pics 1 and 2 personally- looks pretty close to commanded to me!