Space Shuttle - Development

[Thorsten]

Fairly bad idea:

* use UNIV PTG to track ISS with the -Z axis to get a star tracker fix
* end the tracking by depressing the INRTL pushbutton
* change to SPEC 34 to compute a Lambert solution
* change to OPS 202 to execute the maneuver

Since the former tracking hasn't ever been canceled, the corresponding loop is still running (we never check for that when doing OPS changes...) - which interferes with the burn targeting - which in the event gave a burn attitude totally (and I mean totally) off - which resulted in a rather spectacular cruise past ISS.

Ah well - we all learn. Took me a while to understand what happened...

[GinGin]

So I need more info on what is actually happening and what might trigger it...

Yes, I will dig the numbers next time it happens

Since the former tracking hasn't ever been canceled, the corresponding loop is still running (we never check for that when doing OPS changes...) - which interferes with the burn targeting - which in the event gave a burn attitude totally (and I mean totally) off - which resulted in a rather spectacular cruise past ISS.

Argh
It happened to me the same, not canceling the Tracking and having everything flipped apart

[Thorsten]

I've just tested a complete final sequence

* getting a star tracker fix from 35 miles out
* update state vector
* use the Lambert code to get a rendezvous solution within 35 minutes transfer time
* get RR fix during the approach
* update state vector
* defer the computation of the TIG2 parameters till ISS model is actually loaded
* then do the TIG2 burn - which then automatically takes care of the errors switching from semi-analytical to proper graviy
-> errors were O(1000 m) in position and O(0.5 m/s) in velocity, which is not as good as I'd hope, but completely viable to
* do a second CW targeting to bring us into the vbar of ISS with six minutes transfer
* errors were O(50 m) in position and O(0.3 m/s) in velocity - which is completely viable
* use COAS to get a good fix on position, use RR to get distance, null residual velocities using these tools

-> could have done proximity OPS from there, but I've already tested the fly-around and the docking.

Nice...

I wonder what targeting uncertainties the real thing had. The OMS targeting needs to be good to 10 deg only to trigger EXEC, so potentially for a 30 fps burn, that'd mean it could be as much as 5 fps laterally off. Over a transfer time of 1000 s, that'd be as much as 5000 ft miss.

Turning the Shuttle around usually gives me a 0.7-0.8 fps residual translation, so over 1000 s that'd amount to ~800 ft uncertainty.

So that might be the reason they have the manual rbar and vbar rather than solving everything by targeting software...

[GinGin]

Sweet !
I would be really interesting if next time you could take some pictures of the Rendez vous canvas plot


For Phasing, I find that I have some nice relative plot




And when Relative Height decreases, I find it a bit more hard to predict the shape of our orbit

The OMS targeting needs to be good to 10 deg only to trigger EXEC, so potentially for a 30 fps burn, that'd mean it could be as much as 5 fps laterally off. Over a transfer time of 1000 s, that'd be as much as 5000 ft miss.

As an example, I attached the burn Data and attitude ( Deadband, rate, etc) of The final Nc burn and Ti burn of STS 114





More or less 2 degrees of attitude deadband for final rendez vous burn

So that might be the reason they have the manual rbar and vbar rather than solving everything by targeting software...

There is a very interesting "article " on that,

Part 3, Chapter 26: WHY WAS SHUTTLE RENDEZVOUS AND DOCKING NOT FULLY AUTOMATED

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20110023479.pdf

[Thorsten]

And when Relative Height decreases, I find it a bit more hard to predict the shape of our orbit

ISS never has the altitude variation of a J3 orbiting object - so for small altitude separation, you get all sorts of spurious altitude changes.

Sorry about that... I'll probably work on those errors eventually, but at the moment I'm just tired of targeting problems and want to do something else for a change.

[GinGin]

ISS never has the altitude variation of a J3 orbiting object - so for small altitude separation, you get all sorts of spurious altitude changes

Ahhh, ok. I was wondering if there was a difference in gravity influence between ISS and Shuttle.
So, it gets "just" the J 2 perturbation?
If, yes, J 3 is then quite noticeable with all those small perturbations on relative plot Interesting

Sorry about that... I'll probably work on those errors eventually, but at the moment I'm just tired of targeting problems and want to do something else for a change.

Yes, no worries. I can understand that .

[Thorsten]

ISS in the far zone actually moves on a semi-analytical orbit (in altitude it's actually a perfectly circular orbit). That's a few km off J2 even - J3 you probably would not notice much.

The corrections to the analytical orbit are mainly done to include the drift of the ascending node longitude and the high-latitude deformations of the orbit such that plane corrections stay reasonable.

Since altitude corrections are far cheaper than plane corrections... it's not done.

[GinGin]

Ahh alright, I was wondering about that for a long time.

So ISS, perfect spherical gravity and Shuttle J2/3 Perturbations, hence the relative plot close to the target. That's clearer thanks.

[Thorsten]

Shuttle development is now at a stage where I consider tagging a milestone and making a release - targeting, approach and docking seems to be essentially working, I didn't see any regression for the launch, I haven't tested entry and abort yet (but will give it a try).

Any of the active pilots - please take your ride for a spin and see if stuff generally works, go through procedures, systems management, whatever takes your fancy - I won't have time to test everything either.

I've done a faster fly-around with a rotation rate of 0.4 deg/s - that already starts to get quite stressful (and propellant-consuming), but it is a fun exercise to do.

[Thorsten]

Tested 1 1/2 hours launch and insertion into a 720x620 nm orbit using the OMS kit today. PEG-4 targeting doesn't like the altitude (but then again it never has, not even the offline version of LEO targeting can handle that properly...) and I found a bug with the view shake and another one with the radiator controller power switches - both of them eliminated now.

Otherwise performance of the Shuttle was good - no system crashes, quirks or unexplained stuff. View sure is nice from such high orbits

[GinGin]

For the rendez vous, I tested in Depth and nothing major, it is working well.
I also flown several RTLS and AOA. It works like before, no breaking things here

I have just one thing which popped up in my mind, the ADI.
Would it be possible to invert the color to have it the other way around, black for the ground and grey for the sky in LVLH and Ref, like in reality ?



Not a big deal for a milestone, but we spent quite a lot of time on that instrument, and I still have to mentally focus on the correct sense and color .

[Thorsten]

Yes, I remember I promised to do that

[Thorsten]

Okay, ADI colors should be flipped now.

[Thorsten]

... and ran a quick test that PEG-4 targets for reasonably low orbital altitudes work fine.

[GinGin]

Ahhhh very nice for the ADI , thanks
Way easier for my pshychomotor skills

Anychance to have the grey part brighter ?





Alright, I tested tons of entry, contigency abort, TAL.
Really no major flaws there.
Still works nicely when within the enveloppe ( especially the ECAL and Bermuda abort, fun)

I also tested again PEG 4 with LEO for weird AOA ( Low and shallow)
That works ok ( like we said one time, atmosphere is a bit higher than expected in FG, with effects that are not calculated by leo up to 80 Nm ish )


Ah yes one thing, ADTA editing in OPS 6 ( contigency) and within MM 602
It doesn't work, need to be in the next mode MM 603 to be able to edit item 25 and 28





I tested again the LAN targeting, I would say 50 / 50 on that one ( and SSME gimbal a lot with that mode).
For me, it works much better the Launch in the right LAN window


Last thing for tonight I have in my head,
I don't find the page where you explained how to use well those lines
Could you explain me again please?

I played on trajectory loft to lower the MECO altitude ( I am trying around 350 kfeet, less loft )

Code: Select all

<srb-climbout-ang-bias-deg type="double">0.0</srb-climbout-ang-bias-deg>
  <ballistic-climb-ang-bias-deg type="double">0.0</ballistic-climb-ang-bias-deg>
  <trajectory-loft-ft type="double">0.0</trajectory-loft-ft>