Board index FlightGear Development Spaceflight

Space Shuttle

Discussion about development and usage of spacecraft

Re: Space Shuttle

Postby legoboyvdlp » Fri May 18, 2018 1:45 pm

There's a reason why "NASA realized that it was far too risky to fly the Centaur on the Shuttle." :D wikipedia


Thorsten,
https://imgur.com/a/qk6xyCN
The third image definitely shows an issue. There is also a difference in the length of the particle stream as you rotate around the shuttle as seen in the first and second images.

Have you noticed the issue in the third image?

and a theoretical question,
Would this ever have a chance of reaching the ISS?
Image
User avatar
legoboyvdlp
 
Posts: 7981
Joined: Sat Jul 26, 2014 2:28 am
Location: Northern Ireland
Callsign: G-LEGO
Version: next
OS: Windows 10 HP

Re: Space Shuttle

Postby Thorsten » Fri May 18, 2018 3:57 pm

@GinGin:

Only way I found to simulate that is too reduce manually all 3 engines to 67 %, then shut down the associated APU to stuck on main valve and let an engine to 67 %, then back to Auto thrust and shut down an engine.


I've just tested the following scenario:

* nominal launch
* under autothrottle, engines get throttled to 67% in the thrust bucket half a minute into the flight
* I fail an APU
* when the engines throttle back up, the failed is stuck at 67%
* I can manually throttle the remaining two engines, but not the stuck one
* I can later fail an engine, so I have one good, one stuck at 67% and one out

My understanding is that this is the scenario which leads to the abort you describe (and you could also cause it via electric lockup). You can code that as failure mode into the mission file, so you don't even need to do anything to cause it.

Why does that not work for you?

FG allows you to throttle each engine individually, so you could throttle one engine back as you ask, and FG allows you to take engines in and out of the throttle set (see the general help for the keys), but autothrottle will be oblivious to your attempts to command a single engine and just set the value it thinks - unless the actual failure mode occurs.

@ legoboyvdlp:

The third image definitely shows an issue.


Particles don't play nice with FG's z-ordering for transparent surfaces - another manifestation of that is that they're visible through clouds. I'm a bit surprised that is news to you :mrgreen: What can I say? The OSGParticle implementation sucks, and I'd kick it out today if I had any working code for the particle system xmls we already have.

(The changing length isn't the particle system, it's the flame shader which does play nice with z-ordering - just when some weird condition is met you see the particles as well, otherwise you only see the flame shader)

Would this ever have a chance of reaching the ISS?


Sorry, no - wrong inclination and wrong longitude of the ascending node. Changing these in low orbit is insanely propellant-consuming.

Basically you need to be in the same orbital plane to better than ~0.2 degrees or you can't get there. Which is why launches have to be timed so precisely - you can only launch when the ISS orbit passes right over the launch site.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Space Shuttle

Postby legoboyvdlp » Fri May 18, 2018 4:15 pm

I launched at 51.76 which was as accurate as I could get with the dialog... even if I try 51.6 will that work, if the ISS is positioned over Kennedy?

How could I change the "wrong longitude of the ascending node" pre-launch?

Such as ->

Image

?
User avatar
legoboyvdlp
 
Posts: 7981
Joined: Sat Jul 26, 2014 2:28 am
Location: Northern Ireland
Callsign: G-LEGO
Version: next
OS: Windows 10 HP

Re: Space Shuttle

Postby GinGin » Fri May 18, 2018 5:30 pm

Yes, it was this article :)

So, I extended massively the parameters, quite impressive. Performance still good and view on Abort below 400kFeet is just immersing as hell.
However, I still get those very localise Weather Tiles around the Aircraft.

Image

You know the trick to extend them as far as my visibility, Lod and z far range ( 500 km) ?




Ok, good idea for the bucket. Yes, in the procedures, they speak about an engine lock up at 67 % which is the worst case. But it covers every lock up followed by another engine fail ( Lock up first stage, second stage, droop and TAL). It is at the bottom of contigency cue card boundaries.
I will start by trying that, lock up during the bucket leading to RTLS later on, and then another engine fail . Delicious ahah




Speaking of RTLS, I tried a couple of them.
I shut down Left engine that time, for all them, Nz hold was perfect

Some were a bit short of the runway ( after 3 mn for the abort)


Image



For those below 3 mn for RTLS activation, AP range calculation at the start of ALpha recovery was much better ( 30 % better ie. 40 Nm closer to the cap after the Nz Hold)

Image


Image



Perfect range/altitude for a standard Overhead HAC

Image


MLS work much better now, SV is almost perfect as expected once they come alive below 15 kfeet
I overran runway again, really need training


For late RTLS, I usually do the PPA manually to force AP to come back a little earlier
It fixes usually the range thing after Nz hold.

Great fun today.
GinGin
 
Posts: 1580
Joined: Wed Jul 05, 2017 11:41 am
Location: Paris
Callsign: Gingin

Re: Space Shuttle

Postby Thorsten » Fri May 18, 2018 6:46 pm

@ legoboyvdlp:

I launched at 51.76 which was as accurate as I could get with the dialog... even if I try 51.6 will that work, if the ISS is positioned over Kennedy?

How could I change the "wrong longitude of the ascending node" pre-launch?


As for the first question, you can i-load an exact number in the mission file, which is what you ought to do if you were to plan a real rendezvous mission.

As for the second, you have very limited capability to change LAS during ascent, basically you have to wait till Earth rotates the launch site into the correct inertial location and then lift off. Dependent on where ISS is at that moment, you need to do a few days of phasing (orbit at lower or higher altitude) to catch up.

Basically there's only rather limited times during which a transfer works within what the Shuttle can do - so launch windows have to be planned well in advance, and if you miss the window by a few minutes, you have to abort and try the next one. It's a bit much material to cram into a post in sufficient detail - planning a rendezvous is a handful of math. You'll get to learn it once I have added the functionality to my LEO targeting code in the tutorial. :mrgreen:

@GinGin:

You know the trick to extend them as far as my visibility, Lod and z far range ( 500 km) ?


There's a slider in the rendering options (cloud visibility range) that should give you at least 200 km. If you need more, try setting

/sim/rendering/clouds3d-vis-range

to the desired value - probably in meters (you know how to use the property browser, right?)

For late RTLS, I usually do the PPA manually to force AP to come back a little earlier


If you have even a rough table of the correction as a function of engine failure time, I can change the table that's used internally by the AP to determine PPA easily. Just 'a little' is difficult to code.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Space Shuttle

Postby wkitty42 » Fri May 18, 2018 6:59 pm

GinGin wrote in Fri May 18, 2018 5:30 pm:However, I still get those very localise Weather Tiles around the Aircraft.

[trim image]

You know the trick to extend them as far as my visibility, Lod and z far range ( 500 km) ?

i think that would be the --visibility property and/or --prop:/sim/rendering/clouds3d-vis-range

eg:
--visibility=370000
--prop:/sim/rendering/clouds3d-vis-range=370000

370000 being based on an altitude of 35000ft which gives you a horizon distance of 369km... the problem being that you can't (yet?) vary these properties based on your actual current altitude... it would be nice if they could be dynamic based on your current altitude because then memory consumption would also be dynamic to grow as you climb and shrink as you descend...



FWIW: here are my notes from my .fgfsrc file...
Code: Select all
# the following distance numbers are based on the horizon distance of 369km @ 35000ft altitude
# visibility 370000 takes fgfs up to ~9.5G of RAM usage with these current settings
# visibility 150000 takes fgfs up to ~4G of RAM usage with these current settings
--visibility=75000
#--visibility=150000
#--visibility=370000
--prop:/local-weather/config/ground-haze-factor=75
#--prop:/local-weather/config/ground-haze-factor=150
#--prop:/local-weather/config/ground-haze-factor=370
--prop:/sim/rendering/clouds3d-vis-range=75000
#--prop:/sim/rendering/clouds3d-vis-range=150000
#--prop:/sim/rendering/clouds3d-vis-range=370000


you can see the settings are for 75000km and my FG is currently using 2Gig of RAM... the lines starting with '#' are comments... i find it easier to do it this way and then move the comment leader when i want to change something ;)

so figure your horizon at your altitide and set those properties accordingly... beware of RAM usage! the more you try to see, the more RAM is needed...
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 9146
Joined: Fri Feb 20, 2015 4:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 20.04

Re: Space Shuttle

Postby Thorsten » Fri May 18, 2018 7:05 pm

370000 being based on an altitude of 35000ft which gives you a horizon distance of 369km... the problem being that you can't (yet?) vary these properties based on your actual current altitude... it would be nice if they could be dynamic based on your current altitude because then memory consumption would also be dynamic to grow as you climb and shrink as you descend...


AW sort of varies the visibility (though the model used isn't overly realistic above 100 km) - the cloud visibility range doesn't really matter once it's large, because what's drawn are impostors anyway and they hardly consume any resources - I've tested 800 km worth of clouds with the technique when I implemented it.

Extending visibility range to 300+ km is okayish if you're largely over the ocean - but with CORINE terrain in Europe, that'll kill your memory for good - so implementing a 'real' visibility model with the default terrain engine is prone to unexpected crashes and I'd never make such a model available for the casual user.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Space Shuttle

Postby GinGin » Sat May 19, 2018 5:10 am

Thanks for the indications, I will try that :)
So far, with 500 km of visibility, I am around 12 Go of Ram.

@Thorsten: indeed a little is not really quantitative :) I will do some more RTLS and write the value etc..
GinGin
 
Posts: 1580
Joined: Wed Jul 05, 2017 11:41 am
Location: Paris
Callsign: Gingin

Re: Space Shuttle

Postby Thorsten » Sat May 19, 2018 6:37 am

It occurred to me that we might equally well modify the RTLS MECO targeting code. It uses distance and site-relative speed to estimate the cutoff timing - if we know these parameters at MECO and the range difference to optimum this caused in the end, this would presumably also help for manually flown patterns - it's more general than modifying the PPA time.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Space Shuttle

Postby legoboyvdlp » Sat May 19, 2018 1:19 pm

For those of us on Windows, QTGrace is a useful plotter for LEO targeting:
Image


Also, Thorsten, about the transparency thing, the only obvious thing I noted was that you didn't seem to define a background colour, at least in the file you mentioned:

From the Canvas Snippets page:
Code: Select all
window.createCanvas().set("background", canvas.style.getColor("bg_color"));


Is that present in another nasal file?


Edit:

ah... I need more practice :D

Image

a bit hot, then a bit late to flare


slightly better:
Image

Edit 2:
Forgetting to pass a heading to --orbit :D
Image

Well... quite unusual, but we can try it!
User avatar
legoboyvdlp
 
Posts: 7981
Joined: Sat Jul 26, 2014 2:28 am
Location: Northern Ireland
Callsign: G-LEGO
Version: next
OS: Windows 10 HP

Re: Space Shuttle

Postby Thorsten » Sat May 19, 2018 3:59 pm

Also, Thorsten, about the transparency thing, the only obvious thing I noted was that you didn't seem to define a background colour, at least in the file you mentioned:


Since I don't see anything missing, it's hard to tell - does it fix the issue for you if you add the definition?

I need more practice


Yes. :mrgreen:

It's surprisingly tricky, one needs a very controlled flare and the landing airspeed is unusually fast,
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Space Shuttle

Postby GinGin » Sat May 19, 2018 4:34 pm

@Legoboyvdlp: You are discovering the joy af Approach trainer :) I am gonna to do it again also, very fun and instructive :)
We all broke a lot of wheels don't worry ahah








Visibility is better, thanks guys. However, it is funny but at high altitude, some snow appears on the land on the back corner, here at 400kfeet
When I play with the snow line, it disappears from time to time.


Image


Advanced weather is stuck at 250 km, no way to increase that in command line?


@Thorsten: I made some tests with Left or Right engine cut off, every time, I nailed the Nz Hold, no problems ( so maybe it was coming from the center one like you mentionned)
RTLS at 2:30 no problem, parameters by the book

RTLS and failure at 3:00
Vi= 6338
Outbound leg pitch at 38 ° and PPA at 52 % like in unconverged guidance table. Distance from Cap 230 Nm, Vi=7700
Inbound leg, pitch between 37 and 55 degrees ( initial unconverged pitch at 50)

Meco at 7900fps and 290 Nm from KSC

Image

NZ hold perfect (2.8 G)

Alpha transition and phugoid left me at 100 Nm and 120kfeet, perfect

What a view hehe

Image



MLS integration is far better

Image



And finally, a good landing

Image


Image



I will try now later RTLS, but guidance is really nice. Maybe I was disturbing it a bit too much with CSS.



RTLS 3:30
Vi: 7148
Outbound 33 degrees ( 34 for unconverged guidance), PPA at 53 % ( 52 %) at Vi=8100 and 196 Nm , inbound 44 ° ( 48°)

MECO at 8000 and 306 Nm

Image


Alpha Recovery at 120 nm from KSC
All good, I had to do a straight in but i worked.




Look at that painting :oops:

Image
Last edited by GinGin on Sat May 19, 2018 6:17 pm, edited 1 time in total.
GinGin
 
Posts: 1580
Joined: Wed Jul 05, 2017 11:41 am
Location: Paris
Callsign: Gingin

Re: Space Shuttle

Postby legoboyvdlp » Sat May 19, 2018 4:50 pm

Thorsten wrote in Sat May 19, 2018 3:59 pm:Since I don't see anything missing, it's hard to tell - does it fix the issue for you if you add the definition?


Not exactly too sure. I'll test it shortly and see - currently in Low Earth Orbit :mrgreen:

GinGin, could I ask how you aim your landing? Do you put the -- not sure what you call it here, but in the Airbus we call it the bird :) -- onto the circle at the end of the line extending from the runway as seen in your HUD?
User avatar
legoboyvdlp
 
Posts: 7981
Joined: Sat Jul 26, 2014 2:28 am
Location: Northern Ireland
Callsign: G-LEGO
Version: next
OS: Windows 10 HP

Re: Space Shuttle

Postby wkitty42 » Sat May 19, 2018 5:19 pm

GinGin wrote in Sat May 19, 2018 4:34 pm:Visibility is better, thanks guys. However, it is funny but at high altitude, some snow appears on the land on the back corner, here at 400kfeet
When I play with the snow line, it disappears from time to time.

that's a quirk of the weather code... you can also see it if the observer point (example tower view) is further than a certain distance from the craft... when the craft is within the distance range, the ground is normal for that area... as the craft moves further away and out of the range, the snow moves in... i've done it with the UFO over the Nevada desert in the middle of summer ;)
"You get more air close to the ground," said Angalo. "I read that in a book. You get lots of air low down, and not much when you go up."
"Why not?" said Gurder.
"Dunno. It's frightened of heights, I guess."
User avatar
wkitty42
 
Posts: 9146
Joined: Fri Feb 20, 2015 4:46 pm
Location: central NC, USA
Callsign: wk42
Version: git next
OS: Kubuntu 20.04

Re: Space Shuttle

Postby GinGin » Sat May 19, 2018 6:22 pm

@legoboy: Last phase of landing like in my screen, the circle in the HUD is the representation of the flight path ( like the bird or FPV in airbus), ie. where you go
You have to put that FPV on the virtual runway beggining ( which is around 1 Nm before the runway), 2000 feet RA you arm the gear and begin to preflare smoothly ( you can declutter the HUD to have more info )
Then 300 RA, gear down and you continue your flare to touchdown around 210 kt



http://wiki.flightgear.org/Flying_the_Shuttle_-_Final_Approach


@Wkitty: Thanks, so nothing to do for it :)
Last edited by GinGin on Sat May 19, 2018 6:23 pm, edited 1 time in total.
GinGin
 
Posts: 1580
Joined: Wed Jul 05, 2017 11:41 am
Location: Paris
Callsign: Gingin

PreviousNext

Return to Spaceflight

Who is online

Users browsing this forum: No registered users and 2 guests