flight's concept of a 'good landing'
jim frost
around with the cessna lately and just a little while ago did a
picture-perfect landing on the pavement in front of the hangar. A
co-worker of mine did the same thing on the pavement in front of the
tower. Both of us got scores of zero even though they were beautiful
landings.
What ever happened to good landings being those you can walk away
from?
jim frost
ma...@bu-it.bu.edu
PS
haven't managed to land the f15 on the strip that runs parallel to the
runway yet, but give me some time....
G. Murdock Helms
>I take exception at the way flight rates landings. I've been playing
>around with the cessna lately and just a little while ago did a
>picture-perfect landing on the pavement in front of the hangar. A
>co-worker of mine did the same thing on the pavement in front of the
>tower. Both of us got scores of zero even though they were beautiful
>landings.
Perhaps because you're supposed to land on the *runway*.
Flight grades you on how fast you're going when you land, how much
vertical descent you've got, how far off the numbers you are, and how
much drift and heading error you've got...on the _runway_.
Note also that you can, in realtime, take off and land on grass.
Flight does not allow you to do so. Theoretically, even if your
Cessna ran out of fuel, you should be able to make a gorgeous landing
on all that grass no matter where you are, since the only obstructions
around is the runway area itself. Alas, flight doesn't work that way...
it counts all that wonderful grass as "marshes".
>What ever happened to good landings being those you can walk away
>from?
>haven't managed to land the f15 on the strip that runs parallel to the
>runway yet, but give me some time....
Try flying the 747 inverted with the cockpit under the runway sometime.
-Murdock
Rob Gabbard
Speaking of flight:
Does anyone know what the -z option is for ? Also, have you ever done
a T while in dog ? A couple of LARGE wireframe domes show up.
--
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Rob Gabbard (uunet!sdrc!crgabb) _ /|
Workstation Systems Programmer \'o.O'
Structural Dynamics Research Corporation =(___)=
U
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
jim frost
>Perhaps because you're supposed to land on the *runway*.
>Flight grades you on how fast you're going when you land, how much
>vertical descent you've got, how far off the numbers you are, and how
>much drift and heading error you've got...on the _runway_.
Yes, I know, but it upsets me a little because it won't refuel you. A
minor point, considering I probably crash a billion dollars worth of
planes a day so what's a throw-away cessna, but still...
>Try flying the 747 inverted with the cockpit under the runway sometime.
I've never tried that but my co-worker has always been amused at the
ability to do barrel-rolls and loops in a 747. I'm kind of amused at
what you can do in a cessna too. I keep hoping that some new version
will have the wings fall off as they should. I'm also waiting for the
day that I shoot someone down in dog and the plane comes apart and
falls to earth in flaming pieces. This particular effect would have
the bonus of making it easier to avoid planes you've destroyed (I hate
the shoot/dodge-explosion sequence).
Yours for a more realistic way to waste time,
jim frost
ma...@bu-it.bu.edu
D. Christopher Dunlap
> (jim frost) writes:
> >haven't managed to land the f15 on the strip that runs parallel to the
> >runway yet, but give me some time....
>
> Try flying the 747 inverted with the cockpit under the runway sometime.
>
> -Murdock
Or flying through all four buildings in one pass.
Do you folks all still use sidewinders and rockets? Try flying with
cannons only, and no "head-ons".
chris
--
-------------------
D. Christopher Dunlap email: dun...@sgi.sgi.com
Hardware Product Support
Silicon Graphics Computer Systems
jim frost
Or taking off across the runway and turning inside the tower flying an
f15 or f16. It's pretty easy with the f16 but the f15 doesn't perform
quite so well. I also like to do 8g turns at below 100ft immediately
after take-off, or to lift and immediately go into a roll. Looks
impressive in airshows!
>Do you folks all still use sidewinders and rockets? Try flying with
>cannons only, and no "head-ons".
We're not to that stage yet, although it's pretty easy to avoid
sidewinders and missiles if you know they're there (ie you're watching
your radar). One of my co-workers actually managed to shoot down an
f15 with a 747 in real combat -- quite an achievement considering the
pilot of the f15 was using missiles.
I wonder just how much CPU power goes into flight compared to other
applications for the SGI.... :-)
BTW, has anyone given though to building real controls? Even a
joystick and peddles (no throttles, etc) would be a godsend. It took
me several days to get the hang of turning with a mouse and it still
doesn't feel right.
jim frost
ma...@bu-it.bu.edu
G. Murdock Helms
>> Try flying the 747 inverted with the cockpit under the runway sometime.
>Or flying through all four buildings in one pass.
>
>Do you folks all still use sidewinders and rockets? Try flying with
>cannons only, and no "head-ons".
We usually only use sidewinders and rockets when we get too excited and
forget, or for revenge. The rest of the time, cannon only.
Steve and I are really looking forward to the big dogfight, Chris...
pick a date sometime soon. I hear Archer's just signed up
for combat also....
-Murdock (Strangman's buddy)
G. Murdock Helms
(jim frost) writes:
>Yes, I know, but it upsets me a little because it won't refuel you.
We were getting a giggle just yesterday about the way dog refuels your
jet on touchdown...even though you're still doing 300 knots! I can
just see the ground crew now, running after the jet with the hose in
one hand....
>I've never tried that but my co-worker has always been amused at the
>ability to do barrel-rolls and loops in a 747. I'm kind of amused at
>what you can do in a cessna too.
Cessnas make wonderful stealth aircraft when flown against any of the jets.
By the time the jet pilot's seen you, it's too late for him to set up on
you before he blows past.
> I keep hoping that some new version
>will have the wings fall off as they should. I'm also waiting for the
>day that I shoot someone down in dog and the plane comes apart and
>falls to earth in flaming pieces. This particular effect would have
>the bonus of making it easier to avoid planes you've destroyed (I hate
>the shoot/dodge-explosion sequence).
I've heard rumours that this may be in the future. Have you noticed on
the GTX that the rudder moves, the flames change length according to thrust,
and the landing gear fold up and down? Now if we could only have head-on
collisions.....
>Yours for a more realistic way to waste time,
Hear hear. I'm waiting for the graphic chess game that lets you choose
texture and color for the pieces and the board..including crystal...and
a changing viewpoint that you could even attach to chesspieces and get a
'pawn's eye' view of the board. Hint, hint.
-Murdock
D. Christopher Dunlap
> >Or flying through all four buildings in one pass.
>
> Or taking off across the runway and turning inside the tower flying an
> f15 or f16. It's pretty easy with the f16 but the f15 doesn't perform
> quite so well. I also like to do 8g turns at below 100ft immediately
> after take-off, or to lift and immediately go into a roll. Looks
> impressive in airshows!
Rob Mace is the king of amazing stunts. I'm not bad myself, but Rob
still managed to pull a Top-Gun on me a couple times. You'd be on his
tail about to nail him with your cannon, and he'd pull up and hit his
spoilers. You'd undershoot him and he'd nose down, reset the spoilers, hit
full throttle, and pick you off. This whole manoeuver would take about
a second and a half.
Rob was also real good at loosing people by pulling out of a steep dive
at the last moment while the pursuer buried themselves.
I used to have a good airshow where I headed down the length of the
runway, lifted off, snap-roll, touchdown, and liftoff again. Took me a
few takes though. Again, I think Rob originated this trick too.
> >Do you folks all still use sidewinders and rockets? Try flying with
> >cannons only, and no "head-ons".
>
> We're not to that stage yet, although it's pretty easy to avoid
> sidewinders and missiles if you know they're there (ie you're watching
> your radar). One of my co-workers actually managed to shoot down an
> f15 with a 747 in real combat -- quite an achievement considering the
> pilot of the f15 was using missiles.
>
> jim frost
> ma...@bu-it.bu.edu
Mark Libby used to fly a cessna and shoot down fighters of all sorts.
has anyone got into playing "arena"?
LTH network news server
>BTW, has anyone given though to building real controls? Even a
>joystick and peddles (no throttles, etc) would be a godsend. It took
>me several days to get the hang of turning with a mouse and it still
>doesn't feel right.
We have a ``Dimension-6'' -- a steering ball that senses rotation and
translation in 3 dimensions. I hooked it up to flight(1), and the
controls instantly felt much more natural. I'm sorry I can't give
you a detailed comparison between mouse and ball control.
The main problem is that the ball is polled over RS-232, which makes the
program much slower, only about 6 updates/second on an IRIS 3130.
An interesting feature is that you can turn left either by rotating the
ball CCW, or (after toggling a switch) CW. It feels differently, but
I can't say one is better than the other.
Comments please!
Dag M. Bruck
--
Department of Automatic Control Internet: d...@control.lth.se
Lund Institute of Technology
P. O. Box 118 Phone: +46 46-108779
S-221 00 Lund, SWEDEN Fax: +46 46-138118
Richard W. Webb
Speaking of some neat stunts, we have come up with a few here:
1) Ballet Pirouette (in a 747):
Take the 747 to 500 feet and fly over the runway area. While
looking at the plane from the Tower view, pull up hard. Not too
hard as this will simply cause a stall. Not too little, or you
will just climb. At a certain point, the plane is still moving
parallel to the ground, but oriented straight up and down. Sometimes
this will cause the plane to rotate VERY rapidly about its length
axis (a pirouette). I can get it about 30% of the time, and this
is with alot of practice. Usually you just flop around in the air,
and it is challengining to see if you can recover and fly off into
the sunset (if you have a 4D70GT).
2) Suicide:
We only have one machine here, so we had to come up with a Solitare
dogfight. It is REALLY hard to shoot a missile into the air and then
catch up to it and have it hit you, but it can be done. The reason
for this strange behavior is that the missile doesn't thrust for much
more than 30 seconds. After this time, it begins to slow down. If
you shoot it up at a 20 degree angle, then level off, it will be at
just about the right range when it falls back to your altitude.
An easier way is to fly over the airfield, pull up vertically,
lock onto the 90 degree vertical angle, fire a missile. It will
eventually fall down, directly in line with the point you are flying
above, so make sure you are still over the airfield when you are going
vertically. Now land back on the ground. Either taxi to the exact
point you were flying over when you launched, or use your radar to
make the "missile" pixel exactly cover your "airplane" pixel. Stop
here and wait. The missile takes between 3 and 6 minutes to fall back
down, depending on how fast you were going when you launched it.
Happy Shooting.
3) Glider:
When you start over from a crash with the "u" key, you start out
at some random point. Cut engines to 0% thrust and see if you can
land it just based on the energy of your altitude. Your altitude is
random, so sometimes it can be very challenging (especially in a 747
at 2000 ft). Make sure to go for the runway!
4) Suicide Glider:
Combine the above two. Cut engines, glide over runway, pull up
vertical, launch missile, land safely on runway, taxi to point of
impending doom (it is OK to use engines to slow down once you land
by going -100% thrust and full spoilers, then use your engines to
taxi around).
4) Core Dump (hit 1024 hundred feet):
Try to hit the ceiling of >100 Kfeet, this causes a core dump
on a 4D70GT runing 3.0 (still). You have to be going very fast at
just below 50 Kfeet then pull up to vertical without loosing much
energy. The engines quit at 50 Kfeet, but you still have enough
kinetic energy to make it (just barely!). Take the rest of the day
off if you can go from a static point on the runway (ground) to
a core dump in less than 4 minutes (it has only been done once here).
Any other good ones out there???
Be careful out there!
--
Richard W. Webb ecvax!decwrl!borealis!\
ESL Inc. MS/302 sdcsvax!seismo!- ames!esl!rww
495 Java Drive (408) 738-2888 x5729 ucbcad!ucbvax!/ /
Sunnyvale, CA 94088 SMAIL: r...@esl.ESL.COM ihnp4!lll-lcc!
Gary Tarolli
I wrote the original flight/dog programs, but haven't done much in the
last few years. Rob Mace has done the GT version and all recent mods.
Let me clear up a few things and help set the record straight on some others.
Also let me try to explain why flight doesn't always try to be real.
First - the "-z" option is probably obsolote now. Flight used to set a
governor of 20 frames per second so that very fast machines would not have
an unfair advantage. Due to a bug in swapintervals, this was removed I
think. And most recent versions now actually adjust the calculations
according to how many frames per second the pgm is actually achieving.
However, I think there's a bug in most of the released versions that make the
G-limit and Wing-stall limits come into play much too early. Future
versions should have this bug fixed. Anyway, if you're still wondering
what the "-z" option did, it used to turn off the swapinterval governor.
The "T" threat cones were written for Williams Air Force Base (in Arizona
if I remember right). The heads-up display (-h) was also written for them.
I left the "T" in there undocumented rather than take it out, just for the
hell of it. Someone added code to fire SAMs (probably just sidewinders)
when you get close. Apparently , this version has found its way to Canada.
Here's a suggestion for you solitaire players - why not record an airshow
with "dog -o" and then fly against the planes in it and see how many kills
you can record. Of course, the planes in the recorded airshow wont be
firing on you, but if you fly a tricky route in the airshow, it could make
it challenging for someone to shoot you down.
Another variation is for each person to record a 5 minute airshow and
then you can swap airshows - if you kill your opponent's plane more times
than he kills yours, you win. Sort of a a batch dogfight (yuck
Stalls: the ballet pirhouette (sp?) reported in the B-747 is just a bad stall.
Stalls are unpredictable in real-life. Since I had no good data on
how to model a real stall, I decided to take the liberty of making
it do something fun - if you stall really bad you get a random spin
based on just how bad the stall is. The worst stall is probably
when you go straight up then fall down backwards. Once I was
flying at about 100 feet and pulled a real sharp turn and somehow
stalled it real bad - I got sent spinning about 1000 RPM and needless
to say crashed. The other day I was following someone ready to
shoot him down when all of a sudden he stalled and went spinning
wildly - it was fun to watch! - but since he lost so much airspeed
i flew right by and didn't get to kill him.
Refueling: sure its tough on the ground crew to refuel you when you're going
300 knots, but would you rather have to park your plane on the runway
and really wait for the crew to service you? I think the term sitting
duck is appropriate....
Landing points: what can I say? - do you think the FAA would like it if
you landed on the taxiway at Logan or O'Hare? The points are just
some arbitrary grading system I made up to make it challenging to
land. Also, if you're dogfighting it gives you a way to recharge
if you are good enuf to land with people trying to kill you.
One more tip - if you type the '~' key - it turns off all fuel consumption
but takes away all you rockets and sidewinders. However, you can fly
forever with your canon. Since fuel accounts for a lot of a plane's weight,
its best to first burn off all but 1 pint of fuel (1 quart if you are
conservative or paranoid about running out) before using it. Of course,
you can turn fuel consumption back on, but don't expect to get your
weapons back!
I think that about covers most of the topics I've seen on the airwaves.
Remember - half the things in flight are meant to be real - the other
half are there to make dog more fun than it would be if things were too real.
Some things - like exploding planes and body parts falling to the ground
are left as homework exercises ...
Richard Bartels
>
> Stalls are unpredictable in real-life.
>
Nope. Neither are spins. The aerodynamics of both are laid out
in any good pilot's instruction book. Quite logical and quite
standard physics. As a university teacher I could never fathom why one
of my students with a pilot's license thought the flight simulator
was so uninteresting. After I got my own license, I see why.
Among other things, in real life you practise stalls and spins until you
get sick of them (sometimes literally to the point of severe nausea)
and learn through them, and sideslips, and take-offs, and landings,
and coordinated turns, and spiral dives, and etc. all of the quirks
and physical personality of a plane. The behavior of the planes
on the iris is completely phony. Part of this is evident in the
sheer caprice of their stall and spin behavior (or lack of it).
-Richard
jim frost
>on the iris is completely phony. Part of this is evident in the
>sheer caprice of their stall and spin behavior (or lack of it).
True. I first noticed this when playing with the cessna. Most
obvious was the lack of the stall that occurs immediately after
takeoff unless you do things just right. This makes takeoffs much
more interesting, let me tell you. It's virtually impossible to
*force* this under flight unless you try a low-power takeoff. Turns
are also unrealistic since rudder on the cessna has a LOT of effect,
while on flight it has little effect on the cessna and almost none on
any of the other planes. Stalls also bother me since I loose complete
coordination when it starts spinning unpredictably (something I'd
never seen before) although I have to agree with the author that this
is rather interesting. I would really like to be able to do some of
the common airshow stunts that involve stalls though.
Also, has anyone been able to put the cessna into a flat spin? I've
tried a lot and I just can't do it.
As a disclaimer, I'm not a pilot but I've spent quite some time with a
cessna flight simulator in addition to using a couple of the
microcomputer simulators (eg Microsoft). My judge of behavior is
probably somewhat flawed, especially since the lack of real movement
makes many things easier and many harder. And you just don't *try*
some things in a real plane....
jim frost
ma...@bu-it.bu.edu
Bates TAD/HRNAB ms294 x2601
I have seen "real controls" for PC's. If it is a serial device
you might be able to change dog to use it. The hardware is a complete
yoke with flap and throttle controls. It was made to be used with
MS-Flight Simulator.
--
Brent L. Bates
NASA-Langley Research Center
M.S. 294
Hampton, Virginia 23665-5225
(804) 864-2854
E-mail: blb...@aero4.larc.nasa.gov or blb...@aero2.larc.nasa.gov
Mark VandeWettering
While the aerodynamics may be laid out in a pilots book, programming it
into a realistic simulation seems to be considerably LESS trivial.
The aerodynamics are often described in a highly qualitative fashion. Getting
a flight simulator to "feel" right is probably very hard.
The point I think that Richard is missing is that flight and dog are supposed
to be GAMES. Fun. You know, blast the other guy. Figure out quirks in the
program. Is it really vital that the flight simulation be accurate when
you can have more fun zooming about?
Actually, I would like to know if people really crave realism, or they want
a video game-style simulation. I have been considering writing a flight
simulator for the sun, and right now I am leaning toward game play rather
than realism as its chief goal. Comments?
>-Richard
Mark VandeWettering
Gary Tarolli
True, most minor stalls are quite predictable and can be modelled
in simulation and practised in real life. But violent stalls, like a plane
falling backwards, or in a flat-spin, I don't believe would be predictable.
To model minor stalls you need to know the wing design - which part
of the wing will stall first and what the effect on the plane will be. Since
flight uses the same equationsmodel all its planes, this is beyond the
scope of its equations. To properly model stalls and other effects would
probably require much more than the 50 floating point operations in flight,
and would probably slow things down quite a bit if you got carried away.
The equations in flight are lousy. I'm not an aero engineer, I'm
lucky I ended up with planes that could fly period. How about one of you
aero-specialists coming up with a better flight model and posting it?
Dave Forsey
>In article <890217203...@adt.uucp>, ma...@adt.UUCP (jim frost) writes:
>> >Or flying through all four buildings in one pass.
>>
>> Or taking off across the runway and turning inside the tower flying an
>> f15 or f16. It's pretty easy with the f16 but the f15 doesn't perform
>> quite so well. I also like to do 8g turns at below 100ft immediately
>> after take-off, or to lift and immediately go into a roll. Looks
>> impressive in airshows!
>
>runway, lifted off, snap-roll, touchdown, and liftoff again. Took me a
>few takes though. Again, I think Rob originated this trick too.
A hard stunt is to start off across the runway straight towards the tower,
snap roll burying the cockpit below ground (without crashing), touching down
and lifting off *before* you reach the tower itself.
An easier but fun flight is to take off from the runway, pull immediately
into a vertical climb, cut power, continue the loop and land as close as
possible to the same end you took off from. Repeat at the other end
without stopping.
And while we're at it, why not have a version of dog that allows "manned"
anti-aicraft emplacements on the tower, buildings or ground?
The mouse controls elevation and orientation with a steady stream of 20mm
fire while a mouse button is held down. Heck, why not make them movable AA
batteries....
Dave Forsey
Mike A. Gigante
As an ex aerospace engineer this really knocked me out, as an ex-pilot, this
is what makes flight so unrealistic to fly.
I suppose that the equations were uncoupled to make it moderately fast on the
old hardware. Now that the machines are so much faster, it is probably worth
re-writing the code to use the 'correct' equations of motion.
Mike Gigante, RMIT
Gary Tarolli
I didn't leave out the equatins for coupling yaw and roll, I never could
figure them out. THe book I read didn't have them and I couldn't figure
out how to derive them myself. Also from what I could figure out, it
only is true for dihedral wings - on fighters that have nearly flat wings
(instead of a V shape) roll and yaw might not be coupled nearly as much
as on a B-747.
I would appreciate it if anyone could send me some data on this effect,
as aeronautics isn't my field.
P.S. Singer Link sells a micro-flight simulator that uses a 68000 to do
the calculations and then talks thru a serial port to an IRIS. I trust
their equations would satisfy most of you real pilots, but I couldn't get
their code and its written is assembly anyway. So real pilots - be quite
and go buy a SInger Link simulator (they take American Express...)
Joe Ilacqua
<And while we're at it, why not have a version of dog that allows "manned"
<anti-aicraft emplacements on the tower, buildings or ground?
<The mouse controls elevation and orientation with a steady stream of 20mm
<fire while a mouse button is held down. Heck, why not make them movable AA
<batteries....
<
< Dave Forsey
G.E.V.s, armed with ground to air missiles and machine guns, is
the way to go. And good 'tones' on the 4D/20 sound hardware would be
nice. But, of course, someone has to program this and I get the
impression there are too many divergent versions already.
Joe Ilacqua
Associative Design Technology