Jim wrote
> Hi Andy,
> On the p51d fdm configuration, it looks like the substantial change was
> actually increasing the turbo multiplier from 2.0 to 5.5,  and not
> reducing the cruise speed as stated in the CVS log of March 23.  The
> cruise speed change does have an effect, but it is fairly small.
> The problem with putting the turbo multiplier up in that range is the
> manifold pressure output is directly multiplied by that number.  So full
> throttle produces an output of 164 inHG manifold pressure.  We should be
> seeing about 61 inHG at sea level for this engine.
> Setting this multiplier lower to get the correct manifold pressure with
> turbo at sea level should reduce the maximum flight level for the aircraft
> since the second stage turbo cannot currently be modeled.   On the other
> hand, using this lower value should NOT produce incorrect lower altitude
> performance since all the data I'm using is for below the 20,000 ft
> altitude where the second stage kicks in.  The drag numbers calculated by
> YASim should be more or less correct up to at least up to 20,000 ft where
> the second stage would be kicking in.
> If there is a problem that setting the multiplier to 5.5 fixes,  I suspect
> it is in the FDM design and not the P51D configuration.  Any ideas how we
> can fix or work around this?

Where do you get your numbers from for the boost? There are some
contemporary figures around for the Rolls Royce built versions of the
Mustang engine which indicate that the turbo multiplier should indeed be
around 5:


Otherwise, you won't get the correct high-altitude performance. In the
figure above extrapolate the boost curves back to sea-level, ignoring the
effect of the Boost Controller (aka wastegate), and you will see that 5 is
about right - even a bit more for the later Merlins. Don't forget that us
Brits work in psi gauge, while the ex-colonies work in psi absolute: same
thing +- 1 atmosphere. 

Although there is a bug in the current code which gives an incorrect readout
of boost pressure (I forwarded a correction to Andy some weeks ago), the
existing code gives pretty good results if you plug in the numbers right out
of the book. I have just done it for the Merlin XX and was very impressed by
the accuracy.

Of course, we still have to model the gear-driven supercharger, but again I
have forwarded some code to Andy which does this. We are still waiting to
finalise some curves to match supercharger output. I've got some good-enough
results here.



Flightgear-devel mailing list

Reply via email to