Re: [Flightgear-devel] RE water crashes or landing - a change in design principle and default is suggested

[gerard r. <gh...@gm...>]

On mar 25 d=E9cembre 2007, gerard robin wrote:
> On mar 25 d=E9cembre 2007, GWMobile wrote:
> > The original post quoted below exemplifies why I beleive it is a mistake
> > to ever have crash detection for water in a flight sim however let me
> > lay it out simply.
> >
> > 1. Anyone who lands on water in a flight sim knows they are doing it. It
> > is highly likely they WANT to do it - ie have a float plane or want to
> > ditch.
> > Setting a crash default is silly. It forces people to not be able to do
> > what they want and it isn't realistic.
> >
> > 2. In reality all water is in fact landable even in
> > a non float plane. It simply acts like extremely mushy ground. It should
> > be treated like land and have a large drag component. In fact all ground
> > should have a drag componenet so pavement, grass, snow, and muddy
> > runways can be modeled - water should just have a very large drag
> > component. This would more properly simulate takeoffs and landings on
> > ground on water or snow or hard ground etc..
> >
> > Water should be treated like land - period. Any crash detection should
> > ONLY result from the speed of vertical decent during landing but frankly
> > even that should be selectable because all planes have different
> > undercarriage survivability (and again you will end up limiting
> > people.)
> >
> > We should rememeber that water crashes were an error result caused by
> > limited flight sims of the late 80's.
> > Water "Crashes" in flight sims originated
> > When BAO marketed by Microsoft added water crashes early on and it was
> > an ENTERTAINMENT feature - it caused an exciting sound and forced a
> > restart.
> > IT WAS A BAD IDEA THEN AND HAS BEEN CARRIED FORWARD BY HABIT RATHER THAN
> > REALISM ever since. It was a cheap stunt partially caused by limited
> > contact feature routines (there was only one contact routine - crash!)
> > in the EIGHTIES whether between buildings, other vehicles or water plus
> > I suspect the desire of Microsoft (or BAO Bruce Artwick) to create
> > excitement and a "feature" for amateur flyers.
> >
> > One should NEVER CRASH simply because one lands in water. One should be
> > allowed to land in water anywhere.
> > Anyone landing on water is chosing it. He either has a float plane or
> > has decided he wants to put his cessna down ignoring all reality or
> > simulating a ditching. The sim should on default allow it.
> >
> > One should ONLY crash when the rate of collision in the direction of
> > contact (in landing that is vertical speed) exceeds any reasonable
> > impact whether it be with a building, other aircraft, or in a landing.
> > That should be modeled with seperate default factors for vertical side
> > and frontal impacts - especially vertical- that an aircraft model file
> > will carry modifiers for so different aircraft structures survivability
> > can be slightly modeled without full structurally analysis.
> > This way a jungle jumper or bush plane could have say a 3 in the
> > vertical modifier key so the sim could calculate that the bush plane
> > won't "crash" unless it's vertical touchdown (rate of descent in
> > meters/sec) component is more than 3 times default.
> > If you want to get even more accurate landing without structural
> > analysis, "crashes" (unrecoverable landings) should be modeled by
> > calculating the gross weight including remaining fuel times the vertical
> > component at touchdown times the aircraft models factor modifier.
> > Anything beyond that and you need to start introducing structural
> > analysis in the sim which is a whole different ballgame.
> > George
>



Ouups some ugly mistakes (the champagne explain it :)   )

 I don't know enough with YASim FDM, however i can answer with JSBSim.

 Yes we can do what you described, since we can detect which terrain is
 under the aircraft.
 Only as a first step i have introduced to c172p an update which makes it to
 be unable to land (and to take off) on sea
In that case, because the gears are not retractable (in reality)  it crash =
the=20
NOSE "UP/DOWN" to end the course on the back ( we could tune a better=20
simulation ).

 With an aircraft which has gears  retractable , the "landing" on sea can be
 done  smoothly on the belly.
 TableData  "drag" (and "lift") can be given with the best values according
 to the water reaction.
 The values regarding landing on ground remains right.
 We have, only, to select the right TableData according to terrain type,
 which is easy to do.


 Cheers



=2D-=20
G=E9rard
http://pagesperso-orange.fr/GRTux/