IAT related air density correction: results

[stag_matt]

I now have some practical results of experimentation to use the IAT related air density correction feature. Version used: 029l (in later versions, this feature was changed so that low temp figures cannot be input).
I have described the idea in a different thread before and others have proposed very similar solutions. The basic assumption is that the IAT sensor sees a lot of the radiated temp from surrounding hot surfaces and will not reflect true charge temp. That influence is strong at low gas speeds (low RPM) and less at higher RPM. So at low RPM the sensor will read too high (close to CLT temp) and air density correction will overcompensate (assuming low charge density = too lean).

Here is the solution that can be implemented using the still imperfect IAT related air density correction, as present in e.g. 029l and earlier versions:

-At RPM below the lower threshold reduce the density correction influence of the idle air temp sensor to zero.
-At RPM above the upper threshold apply a high percentage of air density correction from IAT
-Interpolate inbetween.


These are the values I input in the table:

TempC %
-12 94
12 98
27 101
45 105
60 108
77 112
88 115

I manually edited my airdenfactor.inc file to give a correction of only 80 percent. I.e. the density curve which follows the ideal gas law as represented in airdenfactor.inc is attenuated by 20 percent. This is just a guess. When the algorithm multiplies the above figures with those from the edited density curve, 100% air density (with an error of 1-2 percent due to bad resolution of the figures) will result at all temps. Complicated, but it works.
So, below the lower RPM threshold, my air density correction is in effect switched off. This means that at low rpm (I set the lower threshold to 800rpm) there is no longer the problem of hot start enleanment. Beyond the upper threshold (e.g. 2000rpm) 80 percent of air density correction will be applied. Interpolation inbetween. Note that the RPM thresholds in 029l are reversed.

This solution is complicated and inflexible (manually editing a .inc file is a pig) and I would love a much simpler way to get the same or a better effect.

The screen should have a table of, say, three or four pairs of "RPM" versus "air density correction applied (ADCA)". That is all you need. I use that term because "correction" has been applied to many things and is no longer clearly defined. So when ADCA is 0 percent, air density will always be 100 percent (equivalent to 20degC/70 degF), regardless of IAT. AT ADCA of 100 percent, the full IAT correction (as standard) will be applied. Between RPM values, ADCA should be interpolated.

The way to test would be to map VE at a standard temp (e.g. around 20 degC) and check at higher and lower ambient temps again. If set correctly, no re-map should be necessary.

In my installation (V8, MPI, heated manifold, NA) I have mapped at ca. 13 degC and see no variation at present temps of ca. 25 degC. This is not a vast swing but a definite improvement. Hot start enleanment after heat soak has completely disappeared. Warmup figures needed to be raised a little because the "automatic" enrichment due to lower charge temp no longer exists.

Try it and let me know your results.

Matthias