Frank,
The tunes have so many "two wrongs make a right" spots that I would not trust the MAF value to give you a precise idea of fuel efficiency.
Also, yes, the VVT-I changes that i have made likely affect the BSFC a bit.
Marc, I'm actually not looking at fuel economy. I'm looking at the possibility of an alternative method of estimating HP/TQ based on MAF logs, without a dyno. Other "virtual" dyno methods are inertia-based and make instantaneous force-balance calculations of rate of speed, inertial resistance, and drag, to estimate the torque. With the method that I'm exploring, the calculation is based on the energy production of the engine and its efficiency. The motivation is: "Can I make estimates of power production using the variables that I log while road-tuning (because I don't have a dyno)."
This is all preliminary and needs some refinement. In a nutshell:
1. Start with a log including at the minimum MAF, RPM, AFR. See figure 1. In this instance, I used a OBDLink MX and Torque Pro and logging nine OBD variables plus producing a fairly data-intensive display on-screen, and I'm getting a sampling rate of about 4HZ. I think I can improve this significanly, by limiting the display and the number of variables to log.
2. Chart the csv logs using MegaLogViewer (cheapware). The chart I'm interested in is MAF versus RPM. I can filter out data points that are outside of the range of interest. See Figs. 2 and 3, for raw and filtered charts, respectively. Some improvement in the data quality is possible here.
3. Use a spreadsheet to apply transformations to the filtered data points. The transformations are:
HP = MAF/(AFR x BSFC) [in consistent units]
and
TQ = HP x 5252 / RPM
The result is plotted in Figure 4, where I've also done some rough polynomial curve fitting. These HP and TQ charts seem qualitatively correct in shape and features but they need some work.
The big question is what number to use for BSFC and AFR because as we all know GIGO. I came across this thread while searching for an estimate of BSFC for the 2GR-FE.
In my charts I've used a BSFC of 0.45 lb/min.HP which is the number generally touted for naturally aspirated engines. But I see from the charts you've published here that the 2GR BSFC can vary over the RPM range. Obviously if I use the "best" BSFC from the charts that you've shared here then I get a higher estimate of HP. But what I'm really interested in is the variation of BSFC with RPM.
I believe the fueling conditions in my logs are not that far off from yours - you probably run your tunes a little bit leaner with more timing. See Figures 5 and 6. My WOT hits about 12.0 AFR/25-26 degrees and I'm about to try 12.5AFR/28 degrees if I can approach that with the tools at hand.
As an aside I've come up with a method for identifying fairly accurately the regions of closed loop control versus open loop. This is done by logging one of the STFT's for either bank. As an artifact of OBD2 logging, under open loop, the STFT is reported as zero. See Figure 7.
In the near future, I'll be looking at ways of improving the data-rate of the logs, and improving the accuracy of the power production estimates, and also, trying for some improvements in the fuel and timing maps.
:sigh:
