In piston engines at fuel/air mixtures above 0.095, the expected relationship
between fuel/air mixture and fuel flow appears to break down. Here is a plot
of ENG_FUEL_FLOW_GPH vs. RECIP_MIXTURE_RATIO (data collected by simconnect)
for the Flysimware Chancellor:
As fuel/air ratio rises above 0.095, EGT continues to fall as expected, but
the fuel flow abruptly stops rising and begins a gradual decrease. The pattern
becomes more exaggerated at excessively rich mixtures. At around 0.110
fuel/air ratio, the fuel flow decreases more rapidly, and above 0.120, the
decrease is even faster. Here is a plot from the stock G36 Bonanza:
It appears that fuel flow is
serving as some kind of proxy for engine power or perhaps fuel burn rate at
very rich mixtures. Why does this happen? Is there a way to tune fuel flow for
more realistic values at high mixture ratios?
I have two suggestions related to the fuel mixture behavior I noted in my
original post: 1. Could the full rich fuel-air ratio be made adjustable?
Currently, it appears to be hard-coded to something around 0.090 or 0.095. For
one engine I was trying to model, the real-world full rich fuel/air ratio was
near 0.105 (my estimate based on operating data). If full rich is limited to
0.090, it is not possible to tune the engines.cfg to produce correct fuel flow
for this engine under both full rich and lean operation. 2. Could the fuel
flow logic be improved so that very rich mixtures at high altitudes do not
result in a loss of fuel flow? Even if combustion stops at excessively rich
mixture, unburned fuel will still flow through the engine and be carried out
the exhaust. The current logic results in two problems:
- Too little fuel is lost from the tanks at high altitudes and rich mixtures compared to reality.
- Leaning the mixture at altitude causes an increase in fuel flow in MSFS until fuel flow reaches a peak and decreases again.
And, good-old Leaning Bug was inherited too… It’s more intense the higher
you fly - as you start leaning the Fuel Flow initially starts rising, instead
of promptly decaying… :-/
