Board index FlightGear Development Spaceflight

Space Shuttle - Development

Discussion about development and usage of spacecraft

Re: Space Shuttle - Development

Postby eatdirt » Sat Apr 13, 2019 7:07 pm

Missed again, the T2 burn did not stop me :( Did compute it within the last moment to minimize errors...

My T1 nailing
Image

Beautiful braking over Europe at T2
Image


Noooooooooo!!!!!!!
Image

After the T2 burn, the HA was around 180, fine, but the HP was around 165, clearly not circular (can be seen in the first figure on the DPS). Dunno what I am doing so wrong here :(

Cheers,
Chris.
eatdirt
 
Posts: 357
Joined: Wed Aug 15, 2018 2:06 pm

Re: Space Shuttle - Development

Postby GinGin » Sat Apr 13, 2019 7:51 pm

Wow, that is a steep approach path.
From 102 to 180 Nm, stratospheric :)

I think try with a shallower angle, step by step ( longer but save/resume works well now for Iss state save )

T2 TiG solution will null the relative speed.
But it is efficient only if the two orbits are already matching very well ie. shallow approach path ( last orbit before rendez vous, differential height was around 10 kfeet in real nasa rdv profil procedures , and that works quite well in game)


In your scenario, you could have first circularise your orbit a the top of it ( 180 mn ish)
You would have been probably a bit above iss one, then slightly fall back behind her with a negative catch up,rate.
Then fire again lambert tool for pinpoint rendez vous nearby iss.
GinGin
 
Posts: 731
Joined: Wed Jul 05, 2017 10:41 am
Location: Paris
Callsign: Gingin

Re: Space Shuttle - Development

Postby eatdirt » Sat Apr 13, 2019 8:23 pm

But it is efficient only if the two orbits are already matching very well ie. shallow approach path


Thanks gingin for the explanations, much welcome. I think I did make the experiment showing that the shallow path is indeed needed, naively thought the Lambert Targetter would do the circularization for me :-/

In your scenario, you could have first circularise your orbit a the top of it ( 180 mn ish)


Yep, I naively though that the T2 burn would do it by nulling the DeltaV. My "scenario" is indeed not finished yet, I am now circularish at 185nm after a long way down and in front off ISS, got a second chance to do it with a swallow approach from above.

See? I have taken with me the payload kit to buy me a bit more time around ;)
eatdirt
 
Posts: 357
Joined: Wed Aug 15, 2018 2:06 pm

Re: Space Shuttle - Development

Postby Thorsten » Sun Apr 14, 2019 5:44 am

Image


Wow... the targeting code actually came up with THAT solution and it got you that close?

That's... pretty impressive (*pats himself on the shoulder*) because that is waaaay out of design parameters.

I'm pretty sure the main reason the T2 burn didn't do what you wanted are finite burn duration effects - if you come in with a high delta-V, you can't null that fast enough. According to some tests I've done, if you're trying to null a single axis velocity component you can compute numerically an optimal ignition time - but if it's a three axis problem with a long burn, there is no optimal ignition time that gives you the same result as an instantaneous burn in impulse approximation.

So you can always expect errors, and they'll be (a lot) worse the more velocity you have to null.
Thorsten
 
Posts: 10950
Joined: Mon Nov 02, 2009 8:33 am

Re: Space Shuttle - Development

Postby eatdirt » Sun Apr 14, 2019 9:42 am

Wow... the targeting code actually came up with THAT solution and it got you that close?


Yep :mrgreen: But I give him some nice numbers from input.

Since I was on a circular orbit at 100nm, thanks to blue box page 180, it is piece of cake to calculate the catch-up rate during phasing, and during an Hohmann transfer (2.7 degrees). With RdV nav, knowing my current "alpha", I estimated the time T1 at which alpha=-2.7 degrees. Then, I iterated 2 or 3 times with the Lamber solver to have very small DeltaZ. And in addition, calculations were fast.

This gave me overconfidence that the braking would have been perfect too ;) Finite duration burn makes sense, that's indeed hundrer ft/s to cancel.

PS: thanks for the explanation on the bugfix thread, if you point me a nas file in which the R, P, Y displays is done, I could try playing a bit with the trigo mess if you want?
eatdirt
 
Posts: 357
Joined: Wed Aug 15, 2018 2:06 pm

Re: Space Shuttle - Development

Postby GinGin » Sun Apr 14, 2019 7:46 pm

Very nice.
Some calculations and phasing before using Lambert targeting are really useful.
I have got pages of calculations for rendez vous, but it is so nice when everything come to life and you finally join iss.

To take care of each axis one by one is one of the trick there , phasing burn with pro/retro grade burn, planar change then and finally some radial shaping for fine tuning with spec 34.

Glad to see you like it :)
GinGin
 
Posts: 731
Joined: Wed Jul 05, 2017 10:41 am
Location: Paris
Callsign: Gingin

Re: Space Shuttle - Development

Postby eatdirt » Sun Apr 14, 2019 9:26 pm

It'll take a while before I really dock on ISS :) Never managed to reach a proper station keeping from my funny position and hit a bug soon after.
We should not get in and out the proximity region... I slowly drifted away, but when reaching the 5000m boundary, band, I jumped by 30km on the other X side, with tens of ft/s of speed, and velocity with other signs as well. I guess crossing the semi-analytical frame is a no-go out.

I have tried another approach, more reasonable, but I have to say that the Lambert solver never really works for me when closing in, to cancel the velocities, either in or out the proximity zone. it always return too large values of Delta V to be reasonable and wants to put me on strong elliptical orbits. How do you properly brake guys, and when? Playing with DeltaVx and gravity?

Edit: I got once more the jumping within the proximity zone from one side +x to the -x, and the velocities got reversed as well. I was well inside 2-3km max from ISS, the funny thing is that my approach was not affected, everything changed sign, that's all :)

Cheers,
chris.
eatdirt
 
Posts: 357
Joined: Wed Aug 15, 2018 2:06 pm

Re: Space Shuttle - Development

Postby Thorsten » Mon Apr 15, 2019 5:06 am

I guess crossing the semi-analytical frame is a no-go out.


Basically yes - the proximity region is made to more or less fit the semi-analytical target (as well as possible) but the reverse is not done.

You can make the proximity region 'large' though as soon as you've reached it, there's a (poorly tested) config option.

it always return too large values of Delta V to be reasonable and wants to put me on strong elliptical orbits.


For me it's usually up to 2-3 m/s when transiting from outside into the proximity region and inside the proximity region it's good up to 0.5 m/s.

The infamous jumping I've never seen - GinGin has provided good shots showing that it happens, but I don't really have an idea how to get a handle on it if I can't see it on any of my systems. :(
Thorsten
 
Posts: 10950
Joined: Mon Nov 02, 2009 8:33 am

Re: Space Shuttle - Development

Postby eatdirt » Fri Apr 19, 2019 9:25 am

I confirm that the Lambert Targetting T1 is working more than fine out of nominal, I have tried 2 direct insertions from very elliptical orbits, HA=100nm HP=180, onto DX=DY=DZ=0 and I landed at less than 1km from the station! Visual then!
The T2 burn does not work though, as expected. I did these tests to see if I could null the remaining relative velocities with the COAS after the T2 burn. No way, the residual relative speeds were about 20m/s, and in visual between ISS and the COAS this is actually quite scary. In one of the 2 tries, I passed nearby ISS at 300 meters with relative speed of 20 m/s... :shock: There is a reason for the nominal procedure....
eatdirt
 
Posts: 357
Joined: Wed Aug 15, 2018 2:06 pm

Re: Space Shuttle - Development

Postby eatdirt » Fri Apr 19, 2019 9:31 am

The infamous jumping I've never seen - GinGin has provided good shots showing that it happens, but I don't really have an idea how to get a handle on it if I can't see it on any of my systems


I've noted some thing in 2019.1.1 (that was not present in 2018.3.*), in various external views, when the Shuttle is rotating, the point of view is rotating as well but not in the continuous way as it used to be, it is made of not-so-small jumps, as if the positions were buffered and then applied. I am wondering if these two things could not be related.
eatdirt
 
Posts: 357
Joined: Wed Aug 15, 2018 2:06 pm

Re: Space Shuttle - Development

Postby Thorsten » Fri Apr 19, 2019 9:38 am

That's too high framerate - I've had that as well because the core is getting more efficient it seems. You need to throttle framerate to 60 fps in order not to confuse the view manager, it can't do 120 fps.
Thorsten
 
Posts: 10950
Joined: Mon Nov 02, 2009 8:33 am

Re: Space Shuttle - Development

Postby eatdirt » Sat Apr 20, 2019 11:39 am

That's too high framerate


Thanks, that fixed it, perfect.

The T2 burn does not work though, as expected


But, we can do more than one, and that works, amazing!!!!! It is like when the Enterprise in Star Trek leaves Warp Drive, I have just been arriving with a hell of speed in visual range, did 2 T2 burns, bang! NAILED:

Image

Sorry to pervert (again) the Lambert Targetting for of nominal situations... :mrgreen:

Cheers,
chris.
eatdirt
 
Posts: 357
Joined: Wed Aug 15, 2018 2:06 pm

Re: Space Shuttle - Development

Postby GinGin » Sat Apr 20, 2019 9:08 pm

Do you remember the paramètres before the two burn ? ( hp ha and relative speed ? )
Nice method.
In real, T2 wasn’t even use.
It was preferred to do several T 1 burn to affinate the relative position and match the target orbit by small step and forecasted relative position .
GinGin
 
Posts: 731
Joined: Wed Jul 05, 2017 10:41 am
Location: Paris
Callsign: Gingin

Re: Space Shuttle - Development

Postby eatdirt » Sun Apr 21, 2019 12:20 pm

Hi Gingin,
for that working one, it was not so crazy, I transferred from ha=175 hp=140 to ISS with a T1 burn which was quite close to a Hohmann, and then 2 T2 burns separated by less than 5 minutes, the second one being improvised.

It was preferred to do several T 1 burn to affinate the relative position and match the target orbit by small step and forecasted relative position .


This is clearly the way to go, arriving a warp speed in a random position nearby ISS is certainly not a good idea in real life :) Still have to train my prox ops, I am systematically burning all the fuel for docking!

cheers.
eatdirt
 
Posts: 357
Joined: Wed Aug 15, 2018 2:06 pm

Re: Space Shuttle - Development

Postby GinGin » Wed Apr 24, 2019 6:45 pm

Same for the prox, Here and in Orbiter, I am really not the king of Prox OPS fuel wise
Patience is key of virtue I guess :)
GinGin
 
Posts: 731
Joined: Wed Jul 05, 2017 10:41 am
Location: Paris
Callsign: Gingin

PreviousNext

Return to Spaceflight

Who is online

Users browsing this forum: No registered users and 7 guests