Flight Question Rendezvous with ISS - Travelling in opposite direction

[jangofett287]

First pass attempt happened at 5km out. 5km is nothing when you consider the size of the XR2, the ISS, space and the speed both are going at. If we could code an autopilot to do the alignment for us, we could run orbiter at a super slow time interval, to be sure of docking.

 

[dgatsoulis]

What about Orbiter_NG?
The step command returned me these values:
dt: 0.000281 and FPS: 5441

So assuming that in Orbiter the docking port captures you at a distance of 0.1 meters, in my PC there is no way to dock with a station at a speed greater than 0.1/0.000281 = 355.872 m/s. (Using Orbiter_NG).

In order to be able to dock with a speed of 16km/s I would need a dt of 0.1/16000 = 0.00000625 seconds meaning that I'd need to run Orbiter at 1/0.00000625 = 160000 fps !!!

EDIT: :hmm: There must be something wrong with my calculation of the fps.
EDIT2: the dt and FPS change everytime I use the step command so in other words:

 

[jangofett287]

Or you could just modify the fixed time step settings under debugging parameters. Oh, and thanks for the 0.00000625 value!

---------- Post added at 08:19 PM ---------- Previous post was at 08:01 PM ----------

um, does anyone have a spare supercomputer? Running through the inline client at a fixed time step of 0.000005s per frame, with 2 vessels, I get ~80fps. This means that one second is 200000 frames long. (1/(5*10^-6) at 80fps, this works out as 2500 seconds of computer time for 1 second of simulation time. This means 1 second takes about 41 and a half minutes to compute. Even at 5500fps, we get ~36 seconds per second, but I don't think you can run LUA scripts or MFDs through the raw NG prompt, you need to add a graphics client, and there goes your framerate. I suppose I could salvage some frames by lowering graphics...:hmm:

 

[NTpspE]

Haha it seems my flight mistake has turned into a whole new experiment for orbiter

 

[garyw]

Of course, anything like this tends to stir the imagination. I was considering making it an 'around-the-moon-using-as-little-dV-as-possible' challenge

 

[insanity]

Maybe our resident challenge experts (dgat or flytandem) can think of something better, but I had a similar idea:
Take off to a retrograde LEO. Transfer around the moon and comeback to a prograde orbit for docking with the ISS with the least amount of dV.

 

[N_Molson]

Take off to a retrograde LEO. Transfer around the moon and comeback to a prograde orbit for docking with the ISS with the least amount of dV.

For DG pilots that doesn't like to do it the easy way :rofl:

Good orbital navigation exercise, though :yes:

 

[dgatsoulis]

(Shameless plug).
I have a couple of challenges like this in an addon that I'm making.

The first is named space station hopping, where you have to launch and dock with all the space stations in orbit around Earth.
In the first draft I have just the ISS and Mir and require to be completed in a small time period, (so it's just aero-surfing, not a moon transfer), but I was thinking of adding a polar OB and/or a retrograde one and leave enough time and dV for aero-surfing and a couple of moon transfers. I'll see how it works out.

The second one is flytandem's excellent Polar to GEO challenge. You don't have to actually dock with anything, just get the ship to a GEO. The script checks the orbital parameters and gives you a "Challenge completed successfully! dV used: ... m/s" score once you manage to get into GEO. Of course, there isn't enough dV to complete the challenge without using the moon.

 

[jangofett287]

Right, with some tweaking I've gotten my inline fps up to ~160. This means we now only need 1 250 real seconds (20 mins 48-ish secs) to compute the 200 000 frames required to simulate 1 second @ 5*10^-6 seconds per frame. Yes I have double checked the maths, and yes it all works out. At the above time step and traveling at 16km/s you move 0.08m per frame (16 000m/s * (5 * 10^-6)s) which should be small enough for orbiter to register the docking. Still no idea how to get the accuracy required...

 

[NTpspE]

Right, with some tweaking I've gotten my inline fps up to ~160. This means we now only need 1 250 real seconds (20 mins 48-ish secs) to compute the 200 000 frames required to simulate 1 second @ 5*10^-6 seconds per frame. Yes I have double checked the maths, and yes it all works out. At the above time step and traveling at 16km/s you move 0.08m per frame (16 000m/s * (5 * 10^-6)s) which should be small enough for orbiter to register the docking. Still no idea how to get the accuracy required...

Perhaps once the ship is close enough you could find out which docking port you are headed towards, then launch the DGIV autopilot, and override the config so that it can only use RCS thrusters? (I.e. no slowing down) not sure how you'd do this though.
Or perhaps the UAP mod from vinka could do it, again not allowing it to use anything other than RCS for alignment?

Then again, that would also depend on the accuracy of the autopilot at that speed :L

 

[jangofett287]

I thought Universal AutoPilot was made by Artlav? regardless any docking autopilots will have been written with slowing down and other 'proper rendezvous' nonsense in them. If we we're to try and write one we would need to be able to calculate how close the pass will be ahead of time, probably up to a quarter an orbit in some cases to be able to correct. we would also need to work out what our alignment would be at the meeting point relative to the docking port we are aiming for. we also need to align the station with a docking port facing prograde so we can just bang straight in. the accuracy would need to be tight. Missing even by less than 1/2 a meter would mean failure. of course in the real world, succeed or fail, coming close would mean the inevitable destruction of both vessels.

 

[icedown]

I would thing it would be a feat just to hit it( aka pass through it), let have an approach precise enough to allow docking.

---------- Post added at 23:17 ---------- Previous post was at 23:08 ----------

Perhaps once the ship is close enough you could find out which docking port you are headed towards, then launch the DGIV autopilot, and override the config so that it can only use RCS thrusters? (I.e. no slowing down) not sure how you'd do this though.

Just thought about this. All you would have to do is dump your main tanks right? Aren't rcs on a different tank? Or is this only XR?

 

[Rtyh-12]

If you disabled all perturbation sources (i.e. everything in the Launchpad as well as making Earth the only body in the Solar System) and simply reversed the orbital velocity vector, wouldn't you meet the ISS on the other side? That's my gut feeling, at least...

Perhaps better, you could give yourself a warp drive and burn retrograde, while making sure that your orbital period is a multiple of the ISS'. Because you have to pass through the point where you made your burn, assuming you do it next to the ISS (and that the burn is instantaneous) you will eventually hit the ISS. There's still the problem of attitude, both for you and for the ISS, but that's nothing in comparison...

 

[NTpspE]

(SNIP)

Yes it was made my Arltav, my mistake XD
Perhaps you could 'trick' the autopilot?

If you were able to make the ISS stop in space (without falling) and align one of the docking ports with your prograde direction, and then create a 'ghost' vessel, with the docking port infront of you and travelling in the exact same orbit as yourself, if the autopilot aligned you correctly with docking with that, once you got close to the ISS you would be at the same alignment?
(Sorry if my explanation is confusing xD)

@Icedown: Dumping the main fuel would definitely get the autopilot to only use what fuel was left (RCS) because it wouldn't be possible for it to engage main or hover engines, good idea!

 

[Rtyh-12]

I've just tried this, and I managed to get within 600 m of the station. First of all, I brought up the scenario editor and went to my ship's state vectors. Leaving the positions exactly the same, I foolishly reversed the signs of the speed, in every direction. And guess what: it somehow worked! I found myself flying away from the ISS at 14 km/s.

Then I just waited (time accelerated) until I was on the other side of the orbit. Now, even at 0.1x time acceleration, things are way too fast. So I used Cheat Engine, which, as its name suggests, is a program that helps you cheat in games. It has a speedhack function, which was probably made to speed up the game to farm money or something like that. Well, it can also slow things down. I slowed Orbiter to 1/1000x, which means that the 14 km/s I was travelling at relative to the ISS became an apparent 14 m/s, enough to leisurely watch the station flying by.

(I didn't manage to see it however. I was on the dark side of the orbit. So I'll try again, removing all celestial bodies apart from the Earth)