Tutorial STS-Discovery Reentry and Landing

ADSWNJ

Scientist
Addon Developer
Joined
Aug 5, 2011
Messages
1,667
Reaction score
3
Points
38
OK - first time trying to put a tutorial together, so go easy on me!

This is a tutorial I was looking for when first trying out the Shuttle. After flying the XR-2 and XR-5, I found the Shuttle really hard to handle, and the lack of thrust to help out at the end made it doubly hard. So I've tried many times to land this beautiful ship, and having now landed it 5 or 6 times without problem, I figured this may be of use to some of you.

If you want to follow along, here's the starting scenario:

Code:
BEGIN_DESC
STS-Discovery Shuttle Reetry Tutorial from ADSWNJ
END_DESC

BEGIN_ENVIRONMENT
  System OFSS-Sol
  Date MJD 55918.1999599720
END_ENVIRONMENT

BEGIN_FOCUS
  Ship STS-Discovery
END_FOCUS

BEGIN_CAMERA
  TARGET STS-Discovery
  MODE Cockpit
  FOV 90.00
END_CAMERA

BEGIN_HUD
  TYPE Orbit
  REF AUTO
END_HUD

BEGIN_MFD Left
  TYPE User
  MODE BaseSync
  Target KSC_-_Kennedy_Space_Center
  Reference none
  Data -1.40804 0.497782 0 1 0 8
END_MFD

BEGIN_MFD Right
  TYPE Map
  REF Earth
  POS 0.00 0.00
END_MFD

BEGIN_SHIPS
STS-Discovery:Shuttle
  STATUS Orbiting Earth
  RPOS -3861768.41 -5072031.30 -2564139.27
  RVEL 2053.264 2010.375 -7051.046
  AROT -15.70 15.50 -54.28
  VROT 0.00 0.06 0.00
  RCSMODE 2
  AFCMODE 7
  PRPLEVEL 0:0.307293
  NAVFREQ 0 0
  UMMUCREW Capt-Steve_Lindsay-50-70-70
  UMMUCREW Pilot-Eric_Boe-47-70-70
  UMMUCREW MS-Alvin_Drew-49-70-70
  UMMUCREW MS-Steve_Bowen-47-70-70
  UMMUCREW MS-Michael_Barratt-52-70-70
  UMMUCREW MS-Nicole_Stott-48-70-70
  CONFIGURATION 3
  OV- 103
  RENDER 
  RMSARM 
  ODS 
  CARGODOOR 1 1.0000
  KUBAND 1 1.0000
  POSTLAND 0 0.0000
  SSMES 1 1.0000
  ADTA 0 0.0000
  PETD 0 0.0000
  SETD 0 0.0000
  ARMTILT 0 0.0000 0 0.0000
  PRADIATOR 1 1.0000
  SRADIATOR 0 0.0000
  GEAR 0 0.0000
  PAYLOAD_MASS1 3.0000
  PAYLOAD_MASS2 0.0000
  PAYLOAD_MASS3 0.0000
  ARM_STATUS 0.5000 0.0151 0.0162 0.5000 0.5000 0.5000
  ARM_SEQ1 0.5468 0.3240 0.3153 0.1258 0.4342 0.5000
  ARM_SEQ2 0.7315 0.5539 0.7881 0.2691 0.4943 0.5887
  ARM1SET 
  ARM2SET 
  PL1_OFS 0.000000 0.000000 0.000000
  PL1_DIR 0.000000 1.000000 0.000000
  PL1_ROT 0.000000 0.000000 1.000000
  PL2_OFS 2.550000 2.150000 3.850000
  PL2_DIR 0.000000 1.000000 0.000000
  PL2_ROT 1.000000 0.000000 0.000000
  PL3_OFS -1.850000 1.800000 11.750000
  PL3_DIR 0.000000 1.000000 0.000000
  PL3_ROT 0.000000 0.000000 1.000000
  PL4_OFS 0.000000 1.000000 -5.500000
  PL4_DIR 0.000000 -1.000000 0.000000
  PL4_ROT 0.000000 0.000000 1.000000
  PL5_OFS 2.200000 1.800000 -6.250000
  PL5_DIR 0.000000 1.000000 0.000000
  PL5_ROT 0.000000 0.000000 1.000000
  PL6_OFS 1.850000 1.800000 11.750000
  PL6_DIR 0.000000 1.000000 0.000000
  PL6_ROT 0.000000 0.000000 1.000000
  KEEL_CAM 0.145000 -0.500000 -1.356000
  OMS 
  CALLOUT 
END
BayLights:PBLights
  STATUS Orbiting Earth
  RPOS -3861768.41 -5072031.30 -2564139.27
  RVEL 2053.264 2010.375 -7051.046
  AROT -116.58 34.24 -108.86
  VROT 0.00 0.06 -0.00
  ATTACHED 0:15,STS-Discovery
  AFCMODE 7
  NAVFREQ 0 0
  LIGHTS 0.00
END
RMSLight:RMSLight
  STATUS Orbiting Earth
  RPOS -3861769.37 -5072030.39 -2564146.29
  RVEL 2053.264 2010.375 -7051.046
  AROT -116.58 34.24 160.73
  VROT 0.06 -0.00 -0.00
  ATTACHED 0:16,STS-Discovery
  AFCMODE 7
  NAVFREQ 0 0
  RMSL 0.00
END
END_SHIPS

BEGIN_ExtMFD
END

BEGIN_uap
END

I was coming back from the OFSS, so hence the rather high 500km orbit for the de-orbit burn.

The tools we are going to use are BaseSync (for the deorbit), Glideslope (for most of the situation awareness), Surface MFD (for the vertical speed and acceleration, and other miscellaneous things like static and dynamic pressure), and GPC MFD for Pitch, Roll and Yaw data, and for another view of the vertical profile.

First up, let's set up for the deorbit burn, and then execute. As you can see, we are on the final orbit, about 2/3 of an orbit from Kennedy. The cross-range is 692km, but as we will see, the shuttle is well able to fly that course correction on the way down the reentry.

On BaseSync, go to DEO to set up the deorbit parameters. I selected ANG 1.0 degrees, ANT 55 degrees, ALT 95 km. Depending on various things (e.g. your altitude), you may wish to vary these, but this set worked well for this scenario. So - enter these parameters, and you will see a white burn-line appear on the display, around about the 7 o'clock point, with the yellow Kennedy up at 12 o'clock.

Zoom time with T and R to get close to the burn time (i.e. our green line comes round to the white line). We are close to retrograde anyway, but engage the retrograde AP to align us precisely. Let the time tick down until the green line disappears completely under the white burn line. At this point, burn the OMS engines (NUMPAD + and then CTRL key to lock them 100%) until the BT (burn time) gets to zero. If you are used to more powerful engines (i.e. less realistic?), then these OMS engines feel like they take for ever to slow you down! As you get close to zero, throttle back on the engines (CTRL NUMPAD -) and cut-off (NUMPAD *) at BT = 0.0. Don't worry about any more accuracy then +-1 sec either way.

(As an aside - I don't know if the burn should be equalized around the white line - e.g. burning 3-4 degrees early to have equal burn time before and after the white line. If it should, then probably we need less anticipation, and you get the same effect anyway.)

Bring up Orbit on the left MFD, hit DST (to switch to Altitude), and check the new Periapsis. If it's between 55-65km, then the burn was good and we are on the way home.


Next job - secure the ship for reentry and get the attitude right. Flip the ship prograde with the prograde AP (autopilot). Ctrl U to stow the antenna, Ctrl L to stow the radiators, and once all stopped moving, then K to close the bay doors. Have a look outside using F1 to see everything move into place.

So we are now starting to drop in altitude, through 300km, on our way to an interface at around 80-100km. Time now to set up our displays, HUD and then get the initial entry attitude right.

Put Glideslope up on both the left and right MFDs.

(Tip: you may get a crash to desktop as Glideslope enumerates the base station data. See here for the discussion and answer: http://orbiter-forum.com/showthread.php?t=21187)


On the left Glideslope MFD, hit PB or NB to cycle the base-selector to Canaveral RWY 15. Then hit MOD until it shows the vertical profile entry corridor display (i.e. a yellow altitude trace and a green velocity trace running from right to left across the MFD). On the right, hit MOD until it shows the tape display, with the 5 blue columns of numbers. Set units to your choice of Metric or Imperial with UNT. Personally, although it's the shuttle, I prefer metric, as that just feels right for all the other Orbiter data.

If you have screen real-estate (e.g. a second screen, or a big enough main screen, then hit Ctrl-F4 and bring up 2 Ext MFD's.

(Tip: If you don't see ExtMFD in your Ctrl-F4 dialog, then save current state with Ctrl-S, quit Orbiter, restart, and make sure ExtMFD is ticked on the Modules tab, then re-start from your saved state).

On one Ext MFD, bring up Surface. This is not strictly necessary, but I found that the VACC and VS on Surface are much easier to read than the ones on the Glideslope tape. (It's the same data though.) On the other, bring up GPC MFD, hit OPS, mouse over to the main screen and select 3 to bring up the
ENTRY OPS 301 screen. This is also not strictly necessary, given we are flying a Glideslope entry, but I really like the numeric Pitch, Roll and Yaw data on this display (P, R and Y fields in the bottom right corner).

On the HUD, bring up Surface display. Use Z and X as you wish to get a good field of view. (I personally like 90 degrees until I have the runway in sight, so I get the most situational awareness.)

Click off your prograde AP to conserve fuel, and zoom time to get down to around 110km.

At this point, I want to give you a bit of theory for what we are about to do. You are in the space shuttle, which has very tight angle of attack (also called "alpha") for the reentry. The range is around 37-43 degrees until we get quite low and slow (say 50km / 2000 m/s and below). Too steep or too shallow an alpha and you put the plasma heat onto parts of the shuttle that cannot handle it (e.g. the windshield or the tail). Breaking the windshield or burning your tail off would really mess up your whole day, so we don't want to do that. OK - but here's the other issue... if you try to come in wings level, then you will develop around 3-4 m/s2 of lift on the shuttle. I.e. you will not descend anywhere near fast enough. The answer is to roll really aggressively (e.g. to 90 degrees or even more), so that the lift vector from the wings acts as a forwards brake rather than as a vertical lift. But as you do this, you will gradually cruise away from your target direction. So the reentry profile is a series of S-turns, somewhat like a skier coming down a double black diamond slope in a series of jump-turns to keep the downhill speed under control. Throughout these S-turns, our goal is to maintain 40 degrees angle of attack, so the airflow over the shuttle is as if we were wings-level, and we don't get cooked. Fortunately, we have a pretty awesome autopilot to assist with keeping the AoA at 40 degrees and the Yaw angle at zero.


OK- theory over. Ready for the descent? Good. Time-warp until your altitude is around 110km. Hit prograde AP to bring us back tight on track and null out any yaw. Then click off the prograde. Now, we need to think which way we need to roll. Look at the tape display (right MFD) delta azimuth (DelAz) line along the bottom. This is the angle from your velocity vector to Kennedy. If it's on the left, then we want to roll left. If it's on the right, roll right. The goal throughout the reentry is to keep an eye on this and each time it gets to around 15 degrees (or 5 degrees later on as we get closer), reverse your roll to come back the other way. So - now you have figured out which way to roll, if the prograde is good for the first roll, then great, else roll over 180 degrees to point the other way. Then pull up to 40 degrees pitch, keeping the yaw to zero (i.e. keeping the velocity vector in the middle of the pitch ladder). The P, R and Y data from GPC MFD is handy here.

By now, you should be around 100km, and you should be starting to see a flicker of dynamic pressure on the Surface MFD, as the first stray molecules of atmosphere zip past us at mach 20+. On Glidslope, hit STR. This engages the Glideslope autopilot, that keeps Yaw = 0 and tries its hardest to keep AoA = 40. (You will need to help it out mid-descent, as it struggles to do this by itself).

As we descend towards the atmosphere entry interface, you will notice the left vertical Glideslope display yellow line (alt) coming down to join the desired profile, and the green line (speed) going flat across. Until we hit the atmosphere, we cannot influence these lines, so just watch them and see where they end up.

On the tape display, apart from the DelAz that I mentioned above, you have an AoA display on the left (blue is the desired target, white is the actual). The STR autopilot will try to keep it at 40, but you will need to help it out with some NUMPAD-2 particularly in the middle of the descent when the air pressure is trying it's hardest to push the nose down to kill us! On the 5 blue tapes, the first (vrel) is your relative velocity (i.e. linked to the green line on your left display), then vrelacc, which is the acceleration of your relative velocity, then your altitude, then vertical speed and finally vertical acceleration. Your goal is to try to keep the actual values and the target values all across the middle of this display.

OK - let's try to make that easier to understand. Your actual goal is to fly the vertical profile on the left MFD, so the yellow and green lines run roughly along the target lines, WHILST watching the DelAz and rolling to the opposite side every minute or two to keep the DelAz to +- 15 degrees. If you do this, the blue tapes will trend into the middle of their targets. SO focus on the vertical ascent trend-lines, and see the effect on the tapes, rather than vice-versa. To control the yellow alt line, you roll the shuttle between 60 and 100 degrees on each side. The more you roll it, the more you accelerate the downward velocity. E.g. at around 90 degrees roll, you will get around 2 m/s2 of vertical acceleration, and if you are not careful, your vertical velocity will pick up very quickly. So you need to roll to get the yellow line onto the target, then unwind the roll by 10-15 degrees until the slope of the yellow line matches the target (e.g. -20 to -100 m/s), and then you roll a bit more until the vacc is zero (so the v-speed remains steady). You can see the v-speed and v-acc current values on the blue tapes, but you may find that the Surface MFD has a nicer display to read. So that's the yellow line. On the green line (speed) - you have less control of this, It's driven by drag though, so if you are above target speed, then you can pull up the nose a bit (say to AoA of 43), and below speed, drop a bit (say to AoA of 37) to vary the drag a little. Don't worry too much if you are higher speed, as you can bleed it out in the final "HAC turns", and if you are a bit low speed, then ride a bit above the yellow altitude line for a while to balance it out.

Below around 50km, or when your RCS thrust runs out, switch off the RCS. This activates your flight control surfaces. By this stage, you have more than enough airflow to control the shuttle without thruster jets. (Besides - the thruster propellant is nasty stuff, so running the tanks dry earns you a beer of two from the ground crew!).

So that's about it from 100km to 30km. Ride the yellow and green curves. Watch your AoA and help the STR autopilot to keep the nose up. And watch your DelAz and reverse the roll each time the error gets above say 15 degrees. (Don't worry if it goes higher - you can recover even up to 30 degrees offset given enough height).

As you get a few hundred km from Kennedy, (range on the blue tape display), switch the tape MFD over to the Glideslope horizontal display. You should see a track line to a circle, then to a faint green line. The circle is the "Heading Alignment Cone", and the faint green line is the runway. On the right of the horizontal Glideslope display is a blue desired glideslope angle (20 degrees), and your current angle (less than 20 degrees). You are now in the "TAEM" phase (Terminal Area Energy Management), where you will bleed off any excess speed, align onto the runway heading and land. The idea here is to fly shallow S-turns down the yellow track (say +- 5 degrees) on the horizontal display, up to the circle, then depending on your energy state (i.e. if your glideslope is steep enough), run a circle or two round the HAC, then land. If you have managed your energy nicely, you should have just enough for one circle. If you are high energy (green line is high on the vertical display), then you will need to do a slightly larger circle than the track HAC circle. Be CAREFUL not to pull back too hard in the HAC turn to try to keep to the exact track, as you run the risk of bleeding off too much speed and then crashing short of the runway.

OK - so as you get to the circle, have a quick look around with the mouse to locate the runway and get your bearings in your head. If you are used to flight sim landings, remember that this looks very weird because your glideslope is so steep. Compared to a flight-sim Cessna or a Jumbo your shuttle has very different aerodynamics!

As you roll out of the HAC turn onto final approach, hit CTRL-G to arm the gear, then G to deploy the gear. (If you are low energy, keep the wheels up until the last few seconds, but if on energy or greater, then put the wheels out earlier to act partially like airbrakes.) Tidy up your final bearing, and then put the velocity vector right onto the start of the runway. You should get the "you are cleared to land" then the altitude call outs (assuming Orbiter Sound installed and working). Get to say 100 meters height, then bring your nose up smoothly, keeping the velocity vector just below the horizon (say 2-3 degrees), and your nose above. Touchdown, then use , and . keys to control the brakes. Apply one or the other to brake differentially to steer down the centerline.

Wheel stop. Congratulations Orbitnaut, you have just landed the Shuttle.

That's how I did it. Let me know if this works out for you too. If anything is unclear, let me know and I'll edit the tutorial to explain more.
 
Last edited:

Hurricane

Grinfeld Aerospace guy
Joined
Sep 16, 2011
Messages
211
Reaction score
0
Points
0
Wow. Impressive, thorough, detailed. A few pictures would surely help! :thumbup:
But a few words on the last part:
Actually the space shuttle's approach and landing isn't quite the way you said, it is a very straight-forward procedure that is done almost the same every time. I recommend checking out F-Sim Shuttle to learn it, it is very helpful and also extremely fun! [And frustrating!] :tiphat:
 

ADSWNJ

Scientist
Addon Developer
Joined
Aug 5, 2011
Messages
1,667
Reaction score
3
Points
38
Thanks Hurricane. That F-Sim looks great - I'll download it later for a play. What struck me from the F-Sim trailer video is how short of the runway the pilot lines up (on the steep part), and then pull up into a shallow descent for the final half-mile or so. I'll have to try that. In both the shuttle and the XR-5, I have always tried to put the velocity vector on that start of the runway and then balance the rotation and descent when over the runway. Doing it short like this looks smarter ... i.e. get the pre-flare done first, then glide it in.
 

Hurricane

Grinfeld Aerospace guy
Joined
Sep 16, 2011
Messages
211
Reaction score
0
Points
0
I'm glad you found it interesting! :)
Well, the pilot lines up short of the runway to bleed off that extra energy-it is worth noting that you come in for the preflare at around 307 keas usually, and touchdown at an airspeed of no 195-205 keas. That's a lot of speed to burn off!
 

ADSWNJ

Scientist
Addon Developer
Joined
Aug 5, 2011
Messages
1,667
Reaction score
3
Points
38
That F-Sim is very nice. Definitely useful to get a feel of the real HAC turn and landing approach. (I would *love* to have those alignment corridor rectangles on the Orbiter HUD.) Can't get the final flare right though! Keep dropping it down 1100 feet from the touchdown, 214 kts (my only "yellow")!

Lessons for Orbiter Shuttle from F-Sim ... from about 30,000 ft (~9100 m), you are at around 275 kts (~140 m/s), on a 15 degree glideslope, with HAC roll angles of around 45 degrees. As the glideslope gets to 20 degrees, you roll onto final and pick up speed to around 309 kts (~160 m/s). The aim-point is quite a way short of the runway (eg. a half-runway length), to a pre-flare where you bleed the speed off for final touchdown.
 

Hurricane

Grinfeld Aerospace guy
Joined
Sep 16, 2011
Messages
211
Reaction score
0
Points
0
Well with a bit of coding the shuttle HUD is possible to have in Orbiter (;
The aimpoint for the OGS is the first set of PAPI indicators, the aimpoint for the preflare is the second set of PAPI indicators and the aimpoint for the IGS is just "above" the touchdown zone (marked by the 2 darker rectangles). Make sure that your nose is always aligned with the little triangles on the sides. Those triangles mark what the flight director would do to make a safe landing. Also make sure to make the final flare as smooth as possible so that you'd come to a safe mainwheel touchdown at ideally exactly 200 kts and a v/s of less than 2 m/s. Of course that less is better!
Another thing, I'd recommend you to listen closely to the chatter in F-Sim as it helps you understand what is done at every single step of the landing.
If you know it by heart, than you're qualified for landing the shuttle (in Orbiter!)
~Oz.
 

kwan3217

Addon Developer
Addon Developer
Donator
Joined
Apr 29, 2008
Messages
115
Reaction score
0
Points
16
Location
Geosynchronous Orbit
Wow, I'm pleased that people are using my addon, and ashamed that so many people are having trouble with it, especially without me finding out about it.

I still need to write that deorbit burn calculator, and read glideslopes from a file as has been requested. Sounds like I need to throw this base bug fix onto the stack as well.

Here at St Kwan's Home for the Terminally ADD, we are proud of our customer service, and will jump right on these problems, some time between now and... Oooh, shiny! It's off to another project for me.
 
Last edited:

ADSWNJ

Scientist
Addon Developer
Joined
Aug 5, 2011
Messages
1,667
Reaction score
3
Points
38
Kwan - my huge respect to you, sir! I love Glideslope, and there's nothing to be ashamed of re having trouble. It would be a real honor if you could take me through the source code some day (never coded against the Orbiter API, but coded most types of systems in my 20 years of system programming).

I would love to have a numeric screen version of the tapes, with digital outputs of the actual, desired and deltas on each input, the P R Y numbers, and some quantification of on-energy versus off-energy. Same for AoA and Azimuth offsets. Make the text change color depending on degree of error (magenta/cyan going red/amber, etc). Help me get into this and I'll help do some of this work with you?
 
Top