EVA on canadarm and other issues

[Chris Van den Bossche <chris.vdb@xxxxxxxxx>]

Hi, orbiterfans.

First of all thanks a lot to Martin and all the other contributors for this 
fine version crammed with all sorts of new toys to play with...

I played with the new Atlantis based scenarios and got these experiences:

-It is apparently impossible to grapple the carina satellite with the 
canadarm, although the config file says " SAT_GRAPPLED 1" the satellite 
stays in free floating mode. It is also not possible to stow the sattelite 
back in its cradle for return to Earth. (After jettison, I got it back in 
the cradle using only the Atlantis RCS system but the sat did not "lock 
in") Would be nice to be able to rendez-vous with Carina using Atlantis, 
grapple the sat with Canadarm, stow it in the cradle and bring it back home.

-While trying various key combinations with the satellite repair scenario, 
suddenly an astronaut appeared atop the canadarm. (actually i noticed it 
when the cargo doors were closed for re-entry, sitting on the side of the 
shuttle)
Nice, but this is apparently an undocumented function, and I don't know 
which key i used to get it, so does anyone know the key to get the 
astronaut on the Canadarm and back?

-The original ISS resupply mission scenario crashes on startup with a 
null-pointer error. (i use windows XP pro)
I modified the scenario to start with focus on the other ship, landed on 
earth, then i changed manually to the STS-110. It worked then. Grappling 
works here, but apparently the MPLM module is always locked with Atlantis, 
it can even stay firmly afloat above the Shuttle when re-entering...

I also noted other things (and wishful thinkings) while using Orbiter:

-I tried weird things like landing with two docked vessels (the GL-01 and 
the PB-01) and after landing i suddenly found myself in a fast outbound 
trajectory. I let it run and oribiter finally crashed when i was almost 
half a lightyear away in deep space...

-I accidentally used x10000 while docked to ISS and suddenly me and the 
whole station were also suddenly on a fast outbound trajectory. Less than a 
minute later at x10000 the ISS passed by Neptune's orbit....

Okay, we can live with that since these things should be avoided in a 
"serious" simulation....

A really annoying issue:
-The x10000 timewarp, which is really needed for interplanetary flight, 
destroys all low earth orbits very rapidly. So it is impossible to arrive 
from Mars and dock with ISS since by the time you arrive at earth, the ISS 
orbit has completely gone mad, even crashing the program when you try 
selecting it as a target on arrival.
I know this is due to the inaccuracies introduced in the Newtonian model 
when using the x10000 at a high gravity gradient (like for all objects in 
LEO), so it is difficult to fix directly.
I have a proposal for a workaround: If it is possible to get an estimate of 
the gravity gradient on a given object, why not lock all tight orbits to 
the last computed good orbital elements and keep these elements constant as 
long as the x10000 is in effect? (i mean to calculate postition and 
velocity vector from fixed orbital elements instead of calculating orbital 
elements from newtonian dynamics)

This would introduce an error because station orbits must be able to change 
when you treat it a s a vessel, but at least it keep the stations in the 
correct orbit even when using x10000 for a long time. It wouls also fix the 
ejection problem on accidental x10000's when in LEO
I remember the really old-style space station based on fixed orbital 
elements. Crude, but its orbit never went wrong, not even when returning 
from a moon trip.
The gravity gradient on a vessel in interplanetary transit is very low so 
here the normal newtonian model can stay in effect.

Another thing that should be easy to implement:

Why is the minimal FOV still limited to 20°? Direct3D certainly allows for 
accurate viewing with much smaller FOV angles.
I suppose any astronaut would at least pack a small telescope in his 
baggage (if the ship isn't already equipped with a navigation scope) A 
small amateur telescope can easily get a x200 magnification, (about 0.1° 
FOV) which would mean you can easily see Saturn's rings and moons. (would 
be a great idea for a Hubble Space Telescope module, to be able to really 
look at the solar system with it)

Related to that, it would be nice of all planets had a star-like appearance 
in the sky whenever they are too small to be rendered in 3D.  (ever seen 
Venus in the evening sky?)
Brightness can easily be estimated using distance, albedo and phase. 
Additionally, it would be nice to have labels on the planets in 
"planetarium" mode so one can find em easily.

Apart from looking at the planets from earth, it would enable the 
possibility for "fast trajectories", to have a spacecraft travel between 
planets in a matter of a few days of real-time (instead of months and 
years) using continuous forward thrust the first half and reverse thrust 
the second half. But one has to be able to see the target planet to 
accurately point the ship, sicne the normal orbital instruments cannot be 
used for that.
I once did such a fast trip home using the docking HUD for a trip from 
Neptune to earth with an unlimited-fuel deltaglider.The old-style docking 
HUD always showed the position and distance of ISS (and thus Earth), even 
from Neptune! With new style docking HUD this no longer works. :-(

And finally, just a cosmetic issue: The sun glare should in fact disappear 
when the sun itself is eclipsed, either by a planet or a vessel. Is it 
possible to detect this easily? Would be also nice if a corona appeared 
instead of the glare when the sun's disk is covered fully, so one could see 
(natural or artificial) solar eclipses like they should be seen...


Chris.
______________________________________

Thank God our brain doesn't run on Windows.


--
Orbiter discussion list:
- mail orbiter-request@firedrake.org with subject "unsubscribe" to leave
  or ask the listmaster (@firedrake.org)
- archives at http://lists.firedrake.org/orbiter/