News The Space Shuttle for Flightgear 3.6

[Thorsten]

For the true fans of Shuttle avionics details, here's the first appearance of the backup flight system (BFS) software, with ASCENT TRAJ 1 side by side in BFS (left) and PASS (right).

Note how the BFS displays 'fly-to' errors for comparison with the ADI, whereas PASS has the contingency abort options.

Right now BFS can't control the Shuttle but just listen to PASS (and you have to use the GPC/CRT key to request the GPC running BFS explicitly), but we'll get there eventually.

 

[Gingin]

So cool Thorsten, very good work.
You are fast to update

Will we also be able to access to SM BFS ?:idea:

It's gonna be so cool to press on BFS stick push button to avoid LOC with PASS goind mad during ascent

 

[Thorsten]

Will we also be able to access to SM BFS ?

Give it time... this is actually moderately complicated, due to the way BFS is listening to PASS commands, but PASS is not listening to input to BFS. This creates a situation in which BFS can be commanded to be in a different OPS from PASS for instance, and it all needs to be handled gracefully.

Plus, BFS actually has a couple of unique displays which need to be created.

All in all, it's *much* harder to implement plausible failures than to implement nominal procedures - I once estimated that 90% of the code I've written deals with off-nominal situations and the ability to simulate failures accurately. So there's that...

---------- Post added 07-12-17 at 02:54 PM ---------- Previous post was 07-11-17 at 05:25 PM ----------

Here's the first stab at the BFS takeover logic applied during a launch...

* IDPs 1 and 2 are automatically switched to poll GPC 5 running BFS, DPS screens change accordingly

* IDP 3 has no longer a valid polling GPC and its screens are crossed out

* MEDS screens remain unaffected

* error displays on the ASCENT TRAJ screens vanish, BFS drives ADI error needles instead

It's all fairly subtle... well, definitely not for Shuttle first-timers...

 

[Gingin]

Awesome work

I guess bfs is switched on Idp 1 and 2 because bfs Crt select switch is on 1+2 ?

Looking great !!!

 

[Thorsten]

To be honest, the switch isn't animated in the 3d mesh yet (I need to file a dozen switch requests to my 3d department soon), but its internal abstraction (the 'property node') exists and governs the selection which IDPs will be polling BFS.

Oh, and right now you can't go back to PASS. :shrug:

The problem why this is tricky is that there's just so many new possible states the avionics may be in, especially with the ability to deviate from nominal software loads and re-write the NBAT and the ability to mode the GPCs into HLT or SBY, so this has to work for exotic scenarios as well in which e.g. GPC 5 is dead, we're in OPS 3 but have one GPC run SM OPS 2 to diagnose issues with the APUs...

So there's conditionals that need to be inserted into exotic places... And the guidance algorithms need to be taught that there's a state vector from PASS and a different one from BFS, so we don't only divorce simulated truth and simulated state vector but we do this twice.

 

[Thorsten]

It's gonna be so cool to press on BFS stick push button to avoid LOC with PASS goind mad during ascent

It actually is - I just added a simulation of PASS attitude errors and made them feed the AP and instruments - it's rather instructive, you don't notice a thing at first without outside reference, because all instruments show you in perfect attitude. Then you look outside and Earth isn't where it's supposed to be, and the trajectory gradually starts deviating from nominal.

(In reality you'd get flashing BFC indicators as soon as the state vectors diverge of course - need to add that next).

And once you engage BFS, it studdenly all snaps to a different attitude and the AP tries to correct your state (or not... in my first try, I had a good 60 seconds of flight with the wrong state vector, that was beyond the ability of the AP to diagnose and fix, I had to engage CSS).

 

[Gingin]

Really nice stuff !!!
Did it work in CSS?

Quick question to stay in DPS.
I saw we can access the Store section in Spec 0

Can we do a freeze dried GPC ? Like typically with GPC 3 on orbit, going in OPS 000 with just system software and no application software ( after a Halt Sby Run or IPL) then Store in it MC (G)2 Ops without processing it, with item 47.
Then put it in sleep state via STBY and HALT to have in back up another source of orbit software?

 

[Thorsten]

Did it work in CSS?

Well, I don't have a BFS emulator, I just try to simulate its function, so since I didn't code BFS not supporting manual input yet, it actually did work.

(Shouldn't really have because BFS doesn't support CSS during ascent and, weirdly enough, no AUTO during entry).

:lol:

Can we do a freeze dried GPC ? Like typically with GPC 3 on orbit, going in OPS 000 with just system software and no application software ( after a Halt Sby Run or IPL) then Store in it MC (G)2 Ops without processing it, with item 47.

The software command exists, but the switch isn't animated yet, so you have to do it by typing commands into the console.

It's kind of pointless though since it's... not an emulator, there's no real software loaded from the mass memory storage units and so all 'software' changes are happening near instantaneously - it probably takes longer to unfreeze it than to just switch the software loadout.

 

[Gingin]

Ahah nice thanks for,the answer.

You are talking about the mode ( run stby halt) switch?

Concerning bfs and css, i was thinking that bfs doesnt support css during first stage but then it was css and no auto in 2 nd stage ( like for entry)?

Yeah, its really weird sometimes the logic, imagine a real full entry with bfs and css ahah, stressfull at least .

 

[Thorsten]

I've played a bit with the automatic NO Y JET gains and managed to fly a first entry under AP control in this mode.

This was a fairly short entry with a high target vspeed - roll reversals at a bit over 1 deg/s are not fun and screw the trajectory quite a bit. Basically a late roll reversal managed to reduce my vertical speed to zero, with a corresponding hard bank afterwards to get back onto desired drag.

I was delivered to TAEM interface around Mach 4 and immediately started S-turns, but this turned out to be manageable within normal flight rules.

Nevertheless - better keep some RCS propellant in the tanks... NO Y JET ain't feeling cozy during entry.

 

[Gingin]

Some pictures of my first really test on that amazing sim

PS Thorsten: Same Pics/Topic I posted on FG forum


I am really impressed by quality of displays and numbers of switches functionnal. APU and HYD configuration to avoid SSME lockup and launch abort:

Low pressure and full 3000 PSI Hyd circuit, let's gimbal those SSME

Nicely done for HUD Cover that we can now remove !!

And Lift OFF

Check that Flash Evap is activated, and evap out stay below 60°

I tried to play with ALS Shaders and tweaking, quite good

Very nice work on lighting, I tested many configurations

No artificials lights

The one I prefer, a mix of Back ground lights and panel lights

Meco time, next step OMS manoeuver,

Orbit is safe, perigee outside of dense atmosphere.


Propellant dump completed, a sign of it is He almost depleted

Some work waiting for OMS burn at next Apogee ( 30 mn to wait), Star trackers activation, Apu shutdown, MPS panel reconfiguration, Umbilical doors closure

OMS burn, with a failure of OMS Right ( less than 80 % combustion, no joy, I don't know why)

I tried another one, same results ( despite heaters on)

Really nice stuff with fuel injector data, showing MonoMethyl Hydrazine being heated by nozzle cooling before reaching main chamber, so great

 

[Gingin]

And final test for today with the Hydraulics recirculation Pump, really really nice.
When APU are shut down, hydraulic pressure is kept around 2600 PSI by an accumulator and a differential pressure system to allow circulation and heating of fluid during cold orbit operations.

It works really nicely in the simulation, with pump activation to maintain 2600 of pressure in every HYD circuit and recirculation.

 

[Thorsten]

@Gingin: Glad you're giving it a thorough test... There's probably several situations which I haven't run into yet...

In other matters, the first unique BFS displays are coming online - here's BFS SM THERMAL, and GNC SYS SUMM 1 (with the ullage pressure system monitors) is almost hooked up as well.

I'm unsure about a piece of logic though.

Suppose I have an MDU on IDP 3, polling GPC 3 with 1-4 running GNC OPS 1 in MM 104, 5 running BFS. I do the following:

* GPC/CRT 53 EXEC

(this makes IDP 3 poll GPC 5 running BFS, so I change to BFS MNVR)

* IDP 3 switch to SM

(this calls up BFS SM major function and will force THERMAL onto the MDU)

* GPC/CRT 33 EXEC

(this is a request to poll to GPC 3 again - which runs GNC however, whereas IDP 3 is set via switch to poll an SM GPC, of which we have none)

-> What will happen?

I first entertained the possibility that the last command typed would be discarded as invalid. However, the idea started to look less plausible since one can also use a hardware switch to make IDP 3 show BFS (or not) - and declaring a switch position invalid would be awkward from a design POV.

So my current scenario is that this produces a poll failure, the MDU gets crossed out and a MEDS error message is produced. But does anyone actually know whether this is correct or not?

 

[Gingin]

Very interesting scenario.
I might be able to answer it, just finish 200 pages of DPS workbook

To make IDP 3 poll GPS (BFS)5, it will be done by clicking on BFS display on and CRT 3+1 on pedestal.

It will display BFS on IDP 3 and if engage also on IDP 1.

You could do it eventually with the GPC/CRT function, but problem was in layman terms that it could lead in real to have two commanders GPC on one IDP ( one PASS who will not stop to command the IDP because he is not listening to the BFS and the BFS which is the new commander),due to complex behavior between PASS and BFS cohabitation. Bad things could happen then.

Here is the full technical explanation

But let's say it works well, BFS is now polling IDP 3, Major function to SM, BFS is still polling IDP 3 as all SM and GNC software are contained in BFS.

Then you want to poll again IDP 3 with GPC 3 with major mode staying in SM, right ?

What will happen then is almost like what you said, IDP 3 will look if GPC 3 is running SM software : NO. Then, IDP 3 will look in the Common Set if there is any GPC running the SM function. As you said, GPC 1 through 4 are running GNC and BFS is not in the Common Set, NO again so.

So it is an unsupported major function, current commander retains control of the IDP ( I guess GPC 3) and displays GPC MEMORY (OPS 00001) page until major function is switch back to GNC or one GPC in the Common Set is recoded via NBAT to operate on Memory Configuration 4 with SM software.

It's a bit different than a poll failure, a GPC is still commanding the IDP

To sum up , the idp will be poll by gpc 3 with only system software (ops 0) but no application software
Same effect than if you switch any of the IDP 1, 2 or 3 during ascent to SM Major Function, no GPC in the common set run SM Software and it will display OPS 0

 

[Thorsten]

That answers it rather thoroughly (and unfortunately means I have to code a few more eventualities...)

 

[Gingin]

I think I tried but I don't remember.
What will happen so far if we switch to SM during ascent ? A kind of Poll failure with cross on the concerned IDP?

BFS SM Thermal looks very nice :thumbup:

 

[Gingin]

Let's go for RTLS training, the hardest intact abort in my opinion.
It already gave me a great and though challenge with SSU, I guess it's gonna be the same here.

So many clickable switch

It looks better with the correct scenery installed, a bit of stormy weather on the last Metar, but we have the Go For Launch from MCC

View of the tower:

Breaking the Scaterred clouds layer

So nice to have OPS 6 and Spec 51 Override to play with,

OMS to RCS interconnect and OMS Propellant dump via Spec 51, check

Dump completed, Fault message for OMS low pressure and quantity, Acknowledge and fault cleared

I love those lights effect

Just passing the Vrel=0 dangerous zone, on track

Pitch Power Downed of Alpha=-2° before MECO

MECO

Nice view from the GRTLS

Unfortunately, I was a bit short by at least 50 Nm for this one, I follow the guidance, but no joy

And a beautiful Borealis Aurora to finish, great

 

[Thorsten]

Since we were discussing heating during RTLS N_Z holding entry recently, here's the answer my entry heating model spits out:

The heat shield temperature peaked at 1510 F - so that'd be quite a bit lower than during a regular entry.

(I recorded that during a test how BFS in listen mode responds to RTLS initiation - didn't really go well, but the rest of the flight did, the autopilot did its job perfectly. I messed up the landing though - need to practice my approaches again it seems...)

 

[Gingin]

Nice, which means that its normal for tiles to glow shortly during Nz hold'like in regular entry ?

 

[Thorsten]

I'd strongly suspect yes. Hitting the atmosphere with Mach 8 can't be cold, even the SR-71 gets pretty hot at Mach 3+.

(Come to think of it, there must be temperature profiles from the X-15 at the high speed flights - that'd be a similar regime - I'll try to dig up the numbers).

Edit: Okay, seems the figure of merit for the X-15 is 1200+ F, so that seems to be the same ballpark.

---------- Post added 08-02-17 at 07:33 AM ---------- Previous post was 08-01-17 at 08:55 AM ----------

Special delivery for Gingin - SPEC 21 with the star alignment being tested (IMU 2 being pulse-torqued to the reference angles).