STS-88 (56Kers warned)

[PeriscopeDepth]

I'd quite like to see the new landing features

Any particular points you'd like me to snap the AoA gauge for you during the approach/final? I don't know how to make videos.

Originally Posted by Lieste

I'd be more than happy to help them. Detailed bug reports, reproducible errors etc. I trained as an Aero Engineer many moons ago, and have been developing simulation flight models for a while...

But as I said they were off-hand and slightly dismissive, and they don't make it easy to get involved in the 'community'

Well I suppose your engineering background explains your precision.

Who did you send your suggestions to? And do keep in mind, talented people tend to be sensitive about their work . If you'd like, I could ask/PM a forum admin and try and get you the email addy of the proper person.

PD

[Lieste]

On finals the AOA gauge should be 'free' and mostly lie between 4 and 8 degrees, with excursions to perhaps 12 during the flare and hold-off.

In previous versions the transition from high drag (20 < AOA < 40) to low drag TAEM mode (0 < AOA < 15) was wrong, and AOA on finals would lie in the range (0 < AOA <1)

[PeriscopeDepth]

Originally Posted by Lieste

On finals the AOA gauge should be 'free' and mostly lie between 4 and 8 degrees, with excursions to perhaps 12 during the flare and hold-off.

In previous versions the transition from high drag (20 < AOA < 40) to low drag TAEM mode (0 < AOA < 15) was wrong, and AOA on finals would lie in the range (0 < AOA <1)

I'm at REI now for STS-88, so I should be flying the approach/final either today or tomorrow and will take note of the AoA for you.

PD

[Lieste]

All the above, depending on IAS, of course, and with the lowish drag this tends to be a bit high even with the airbrake fully open. (Real approaches are flown at ~65% spdbk)

[PeriscopeDepth]

Originally Posted by Lieste

On finals the AOA gauge should be 'free' and mostly lie between 4 and 8 degrees, with excursions to perhaps 12 during the flare and hold-off.

In previous versions the transition from high drag (20 < AOA < 40) to low drag TAEM mode (0 < AOA < 15) was wrong, and AOA on finals would lie in the range (0 < AOA <1)

I've been practicing my landings some tonight. AoA of 8-10 from ~15K down to the flare where I stop glancing over at it. I think you're right about the drag, speed doesn't decrease as much as it seems it should from the HUD videos on final. It's not too bad, but definitely noticeable. I fly the approach at 100% brake for the most part unless I am getting too low I'll pull it back briefly. I pull the brake back to a little under a quarter for final (seems to be how they do it on the youtube HUD videos). The mains hit at around 230 to 255 depending on how good a job I'm doing.

I do rely on the diamond too much on approach and should probably spend a little more time heads down before final.

PD

[Lieste]

Well from that it sounds like they have fixed the AOA tape not reading during the finals, but it would be nice if they mentioned it when they 'find'/fix something useful like that as well as when they add a (single use) twiddle for the cockpit lights...

I fly the TAEM and upper finals using the F3 mode, with view panned slightly right and down from the commander's seat. This way I can see the HAC, Energy, and both the Commander's MFD. I reconfigure these to be the Combined ADI/HSI and the Surfaces config (on the left one). This way I can fly the S-turns to bleed energy (HAC should only be flown subsonically IRL) transition to the HAC and fly the upper approach.
Due to the lack of collimation and lowish resolution, I can't easily read numbers on the HUD (but the diamond is clear enough, and progress is checked using the HDD, and the approach diagrams). I therefore change to the 'fixed' HUD for the touchdown phase...

The two HUDs have different fields of view, angles of projection and relative sizes of symbology (unless they've fixed it quietly... ), and the flight path vector (which should be the primary 'instrument' usable during the finals/pre-flare/short-finals/flare/hold-off) doesn't provide a consistent cue especially when the guidance mode changes near the ground. Following the velocity vector cue will result in a crash almost everytime, and it can be quite disconcerting to watch what happens to the VV when you fly the 'successful' approach that SSM-2007 prefers...

Given that the shuttle 'autoland' system uses these guidance cues it doesn't seem to be correct that following them should consistently destroy the shuttle. (On the other hand IIRC NASA flies all the shuttle approaches on manual so maybe they know something )

[PeriscopeDepth]

Payload bay doors coming shut. I've left the beginnings of the International Space Station in orbit.


Deorb burn after almost twelve days in orbit. Slows us down just enough for our orbit to decay. IIRC only a few hundred MPH.



Reentry will commence over the Pacific ocean. View of Papua New Guinea(?) from the PLT's window.


OPS30x will take us home. Kennedy Space Center is 5110 nm off the nose. We'll get there relatively fast at Mach 20+.


RCS jets pitch us up so the thermal tiles on the belly will take the brunt of the heat.


At the very edge of the atmosphere, starting to get some real decel. KSC still over 3,000 nm ahead. I think the TACAN system starts seeing it here. ~ mach 24.



Continued...

[PeriscopeDepth]

A little mood setting with a two minute youtube clip :


Orange plasma is now visible swirling around the flight deck windows.


Exterior view.


The orbiter crew passes the camcorder around.




Out of the fire, but still hot to the touch.

[PeriscopeDepth]

First roll reversal to the left off the coast off of SoCal/Baja California. 225K and just over Mach 20.


Just under 1200 nm to go to KSC, coming out of 200K and rolling right.


Sorry for the boring 2D pit pics, but it's dark and there really isn't much to see down there. 184K, Mach 13.4 . I'm in a right roll again. Endeavour is starting to issue commands to her aerodynamic surfaces for the first time in nearly 12 days. RCS jets are firing in the background.


Mach 10.4, 171K. 660 nm to go. Endeavour's computers command speed brake deployment.


Approach is next...

[PeriscopeDepth]

83K, Mach 1.45. I've just turned the autopilot off and am flying (gliding) Endeavour down to KSC manually.


Into the HAC and heading north just off the Flordia coast trying to catch the diamond.


"On at the 90." My favorite picture that I've snapped, even if it's dark and you can only see the outline of Endeavour.


I'm off center line left and low, but the diamond seems to think I'm correcting all right.


Coming out of 7K I stumble upon the glideslope.


On final and time to start my flare. I'm coming in high, but for an amateur like me this is the best landing I've done.


Chute!


Rollout.


STS-88 comes to an end. You can see the VAB and shuttle launch pads in the background.


Landing in SSM is fun as hell!

For whatever reason, SSM seems to capture the landing experience far more accurately playing the mission through rather than just skipping ahead to landing like I was doing for practice. Everything felt more "right" compared to the youtube videos. I can't explain why, but the approach and landing felt much different than just using the landing practice mode. I felt like speed bleed better than in practice mode. I notice my AoA is also substantially different on finals than it was using the practice mode (seems to be more inaccurate). The HUD VV still doesn't function properly on final though, shouldn't it uncage?.

PD

[Lieste]

The Velocity Vector should uncage once off the HAC, and it sort-of does... at least in the vertical plane.

But the guidance symbology doesn't function correctly around the VV.
IRL, the triangles pair indicates the correct glideslope and roll, in combination with the guidance diamond, and is freely displaced from the VV in the vertical plane. When correctly guiding the triangles are aligned with the VV, and the guidance diamond is gently oscillating inside the VV centre.

When off guidance, the triangles indicate desired 'ground impact point' and the diamond is swept in the direction for returning to the correct flightpath. (ie to command a roll to the right and a pull up, the diamond cycles up and to the right on the HUD. This is not a static cue, but is continually moving - possibly just jitter and noise, but also appearing like an 'attention getter'.)

During the pre-flare, the lower guidance triangles appear and sweep up towards the original guidance triangles. Once they have merged, the single guidance triangle pair remaining moves smoothly up the HUD commanding the pull-up, and guidance is followed by pitching to keep the VV between them.

After the final flare, the top arc of the VV is held on the horizon as the Shuttle settles at ~0.5 degrees descent angle. In practice, this is flown by comparison with the outer visual cues, but SSM2007 doesn't have sufficient visual information to make this very reliable, with the ground textures (especially at Edwards) being a blurry mess at times. 3-4 degree excursions of the VV without any physical cause are disconcerting at best and from my experience downright dangerous..

The closer you fly the nominal approach and guidance information, by my experience with the Demo, the more likely you are to come to grief on the final flare and touchdown, and off-nominal approaches are intrinsically harder to make consistent.... So it seems to be a choice between random (but safe) landings or consistently fatal ones Colour me confused, but shouldn't it be a choice between consistent and safe or random and risky?

[PeriscopeDepth]

Originally Posted by Lieste

The Velocity Vector should uncage once off the HAC, and it sort-of does... at least in the vertical plane.

But the guidance symbology doesn't function correctly around the VV...

Lieste, your knowledge here leads me to believe you must be a flight simulation program yourself. Truly impressive/enlightening to hear you talk about the real world shuttle specs/systems/operations and how it compares in SSM.

And yeah, I hate landing at Edwards.

The closer you fly the nominal approach and guidance information, by my experience with the Demo, the more likely you are to come to grief on the final flare and touchdown, and off-nominal approaches are intrinsically harder to make consistent.... So it seems to be a choice between random (but safe) landings or consistently fatal ones Colour me confused, but shouldn't it be a choice between consistent and safe or random and risky?

Of course. I didn't code the thing. I'm just doing my best to have fun playing it. I don't foresee NASA having a need for liberal arts educated astronauts, so it's the next best thing as of now for me.

PD

[Lieste]

Once on the ground (but while still rolling) you can see some problems with the HUD symbology and other instruments.

The F3 HUD (when zoomed out and centred) shows the datum cross just a fraction below the horizon, indicating a shuttle pitch of -1 degree or thereabouts. This is a bit low compared to the external view, which looks (non measured, but ish...) like -2 degrees, but is at least sensible...

Now if you move your head you will see that the datum cross moves relative to the runway and horizon. Ok, not a biggie, as they acknowledge that they haven't got a proper working collimated HUD, and it is passable most of the time when in the default F3 position.

However, the F6/F7 HUD shows the datum cross 3-4 degrees above the horizon, which is a serious error IMO. The Aspect ratio of the HUD is completely different and the symbology doesn't agree between them.

In the demo at least, the vertical velocity tape lurches downwards just above the ground for no obvious reason, and settles at -1200 fpm while rolling along a flat runway at considerably less than 1200 fpm forwards... On halting it sensibly returns to 0 fpm.

A feature that should be retained though is the small error in reading for the altimeter, which indicates a non zero value even once landed.

[Lieste]

Originally Posted by PeriscopeDepth

Lieste, your knowledge here leads me to believe you must be a flight simulation program yourself. Truly impressive/enlightening to hear you talk about the real world shuttle specs/systems/operations and how it compares in SSM.

I have had suggestions taken up by simulation developers over the past <eek> 25 years, including Digital Integration and Microprose when they were still going - though I promise I didn't break either of them

I've been involved with development of 3d wing code and downwash improvements for a simulation, but this project is on hold while the new engine, to accept the results of research and testing in the previous engine, is completed.

I'm currently a beta-tester for Esimgames, working on the next update for Steelbeasts ProPE.


There are heaps of technical manuals for the orbiter systems available online, and some of the cockpit videos are just possible to make out what is being displayed. Most of the systems and aerodynamic principles are deliberately similar to normal piloting techiques for the approach and landing, with the proviso that you have no power, so you have to have a 'moderately draggy' high speed nominal approach to allow the approach to be extended, and one that remains close to the desired touchdown point in case of arrival low on energy. A reserve of usable IAS and additional deployable drag, and the possibility to expand the HAC turn is required to prevent overshooting too, as you don't have the option to go around.
The whole of the shuttle range-energy-speed-drag information for various phases in the entry/approach is reasonably accurately displayed on the MPDS, and the flight instrumentation can be compared to the nominal guidance progress.