On Wednesday, April 27, 2011 09:10:28 AM Ron Jensen wrote:
> On Monday 25 April 2011 12:10:26 Hal V. Engel wrote:
> > The supercharger model used by JSBSim assumes a turbo charger so the
> > power and fuel consuption curves are incorrect for an engine driven
> > supercharger where more horse power (and fuel) are used to drive the
> > supercharger.
> The supercharger in JSBSim is not very good, but it is engine driven in
> that the pressure directly varies with engine RPM unlike a turbocharger
> which varies with mass flow rate and exhaust temperature. The model just
> does not consume any power. At some point I would like to add a
> <supercharger> and <turbocharger> ability or add functionality to simulate
> an arbitrary boost device...
I knew the model was not correct for an engine driven supercharger (IE. no power consumed) so I assumed that it modeled a turbocharger. And as is often the case with assumptions I was wrong.
The key is that the model "does not consume any power". For an engine driven supercharger power consumed is substantial and goes up with altitude (assumes the same manifold pressure) and some superchargers change drive gear ratios and higher gear ratios consume more power. For example the Parckard V-1650-7 (the engine in the P-51D) uses almost 200 more HP to drive the supercharger in high gear than in low gear at full take off MP. And this makes it important to use VE and BSFC runtime functions to correctly model the HP consumed by the supercharger otherwise the power and fuel consumption curves will not be correct. The problem is particularly accute in aircraft with a very wide performance envelope like the P-51 and it may be a minor issue for a crop duster.
> > Using functions to set VE and BSFC at runtime gives you a way to get fuel
> > consumption and power curves close to correct but it does take a lot of
> > effort to get these functions tuned.
> > snip
> > this should be fairly easy to setup
> > since you would only need to do some tuning to figure out the
> > cooling-factor for the closed and open cowl flaps and then write a simple
> > function to vary the cooling-factor with the clowl flap control position.
> Learning to use JSBSim stand-alone greatly simplifies this kind of tuning.
> It allows you to make multiple, identical runs while varying only a single
> variable at a time.
I have not tried using JSBSim stand alone yet. It would definitely make testing things that require repeated testing for tuning purposes faster since it would eliminate things like start up, take off and climb out overhead which can be very time consuming when testing from with in FG. On the other hand this will not allow testing for anything that has a Nasal component. I try to limit the use of Nasal but there are some things where is is nessary.
Also in this case (IE. tuning cooling-factor) it should be possible to do this in a single FG test flight since you can vary the cooling-factor manually to test which values result in climb outs that do not over heat (too much) and cruise/decents that do not over cool the engine. For the cooling-factors settings used in the P-51D it took me perhaps an hour of testing in FG to settle on the range of values to use.