[jhb] Re: Which Speed?
- From: "bones" <bones@xxxxxxx>
- To: <jhb@xxxxxxxxxxxxx>
- Date: Mon, 2 Jul 2007 17:09:32 +0100
There isn't much in the aircraft.cfg file that relates directly to the
flight model. All the hard work is in the design and the .air file - the
flight envelope, the Coefficients for lift and drag on wings, tailplanes and
fuselage, airfoil performance, control effectiveness etc etc. If you alter
any of these values when designing an aircraft then you have to test the
result through a full speed and altitude envelop to measure the resultant
changes. It's slow and boring and needs endless patience as FS has to be
rebooted each time.
It needs a fair degree of experience to observe the changes made. Ideally it
should be someone with flying experience of the actual aircraft and who
knows what it flies like. For many designers the target is to get the
aircraft to climb and cruise at reasonably accurate speeds (no mean feat)
but slow speed handling may be flawed. If fiddling with this regime results
in climb/cruise wandering off target the temptation is to leave well alone.
The problem with FS is that it is all experiment. If you are lucky enough to
have all the real world values you could possibly need and bang these into
the air file it results in an aircraft that flies nothing like the original
- and so you have to start tinkering with figures to adjust them to the
limited FS flight model. In some ways it would be better to start with a
clean sheet.
Once you have high and low speed control responses sorted out you then look
at engine performance. This can mask basic design flaws - is cruise speed
wrong because of too much/little aircraft drag or because engine power is
incorrect? Does engine application create excessive pitch/yaw moments? Next
is the much harder flight envelope testing as here you need to play with
lift and drag, damping, mass, divergent and convergent effects (dynamic
instability), trim ranges ad infinitum.. Each change needs testing and
sometimes there is a side effect that is not spotted until a lot later - and
so you go back and start again.
Some guys are really switched on and design superb performance files. Others
may just get an FS default aircraft and make minimum tweaks to get the
cruise speed range about right - or not tweak them at all. In the end it
comes down to flying each aircraft and working out your own figures.
I recall one designer who made a lot of FS98 aircraft. They were beautiful
designs and painted extremely well - but they flew dreadfully. If you kicked
in rudder the aircraft would yaw but it would continue flying in the
original direction for a good while. The aircraft mass was too great so he
had just increased rudder effectiveness to compensate. It worked in that it
made the aircraft yaw but the flight path only slowly adjusted to the new
heading. The odd thing was that people raved about the aircraft and couldn't
see anything wrong with them - but a simple ILS intercept at 8nm at 40
degrees would have revealed the flaw.
bones
-----Original Message-----
From: jhb-bounce@xxxxxxxxxxxxx [mailto:jhb-bounce@xxxxxxxxxxxxx] On Behalf
Of Gerry Winskill
Sent: 02 July 2007 16:29
To: jhb@xxxxxxxxxxxxx
Subject: [jhb] Re: Which Speed?
The Internet holds an amazing amount of unsuspected info; apart from the
Confidential A350 purchasing contracts I dropped across yesterday. Often
it's possible to unearth flight test articles, by aviation journalists.
They usually quote weights etc for the flight, then V numbers used on
that flight; specific to other conditions associated with that flight.
MTOW, range, cruise speed, ceiling etc are widely available via
manufacturers' sites.
Look into the aircraft.cfg files and the designers have obviously
visited the same sites. The discrepancy is that the aircraft don't,
often, perform to the spec set out in the file. I can understand that
with aircraft like the A340. It's more difficult to accept something
like the takeoff figures for the Default A321, where MS should have a
head start over third party designers.
In most respects what I'm doing is indeed to forget the spec of the
real aircraft and define the performance of the Fsim model; a luxury
perhaps but it keeps me off the streets. In fact that's not strictly
true, since I've just come back in from a 40 mile interlude on the
Bonnie. I'm still surprised to find that the fully loaded Default A321
takes off at 84 kias; under any conditions! I'm even more surprised when
my stall tests show that, at the same MTOW, its dirty stall speed is
110kias. So, it can take off at 25 kias below its stall speed.......odd.
Still, they are moving more towards the games market.
Gerry Winskill
bones wrote:
>Yep - you are pushing FS beyond its limitations here.
>
>Forget real aircraft values for this exercise. The issue is with the
>aircraft performance file and if the designer hasn't got this locked
>down to the performance of the real aircraft then you aren't testing an
>A350 or DC10 but a mythical design - for which a full flight test
>program needs carrying out to find it's boundaries.
>
>Any aircraft will lift off at stall speed - because this is the minimum
>speed for flying and you are accelerating past this on take off. FS
>models the low speed performance awfully so accept that designers can't
>get this right all the time and that variation exists. Another thing is
>that you say there are discrepancies between some models and the real
>aircraft but where did you get real aircraft values from? The stall
>speed varies with a huge number of factors - weight, ISA, flap setting
>are the obvious ones - but it varies between aircraft fleets too as the
>aircraft specification may change. A Boeing 737-8AS or Ryanair will
>have different figures for a Boeing 737-8K7 of Austrian Airlines. That
>is why performance manuals are written for specific aircraft - even the
>two PA31 aircraft I used to fly had different figures.
>
>The best you can do in FS is to hand fly each aircraft and determine
>your own values - forget any book figures quoted from real manuals.
>
>bones
>
>-----Original Message-----
>From: jhb-bounce@xxxxxxxxxxxxx [mailto:jhb-bounce@xxxxxxxxxxxxx] On
>Behalf Of Gerry Winskill
>Sent: 02 July 2007 11:27
>To: jhb@xxxxxxxxxxxxx
>Subject: [jhb] Re: Which Speed?
>
>
>What triggered off this Quest was the thought that I should rationalise
>the aircraft kept in my FSX Airplanes folder, so that it would hold
>only those I use, plus those required as AI. The others were moved to
>my FSX Hangar.
>
>So, I finished up with a restricted fleet of 4 GA aircraft and 8
>others. The "others" were CITATION II
>B1900D, PDMG
>DASH8 Q400, Dreamwings
>CRJ900, POSKY
>A320 (BA), IFDG
>ERJ145, Dreamwings
>A350, CamSim
>A340-313, Aerodesigns
>
>Most had a Vspeed gauge with speeds related to published weight and V
>number data but I decided to run all of them through the checks I'd
>carried out on the A350, so that Vspeeds reflect the simulated
>aircraft. I also threw into the tests the Default A321.
>
>Main criteria for selecting the fleet was what seemed to be acceptably
>realistic performance, although the Dash8 doesn't behave as well in FSX
>as it did in FS9.
>
>The results showed slight discrepancies between the takeoff speeds and
>those of the real aircraft; with two exceptions. I do the tests at MTOW
>and a minmum weight, so that the gauge has a range to use. In the case
>of two of them I didn't bother with the lower weight. At MTOW and just
>the 12 degrees of flap deployed the A340 lifted off at 108 kias! Since
>I normally fly it with elevator trim set at 10, instead of the 80 I use
>for these tests, I would rotate, at this weight, at the more realistic
>156 kias and not spot that it would have been able to lift off much
>earlier, particularly since its acceleration is what would be expected.
>I can't think of a simple way of altering the lift off speed, without
>suffering knockon effects where I don't want them. So it looks a though
>the A340 is heading for the Hangar.
>
>The Default A321, at max weight and just the first degree of flap set,
>takes to the sky at a slightly unbelievable 95 kias! It's used by AI or
>it would be dustbin bound.
>
>The expression "can of worms" starts to surface.
>
>Gerry Winskill
>
>I
>
>
>
>
>
>
Other related posts:
