Challenge KSC to Wideawake - the "classic" speed run - 2015 iteration

[mjf]

Edit: here is the official list copied from the 2014 thread, then edited to add/change record times during 2015... :tiphat:

TOP 10 FLIGHT TIMES (as of May 18):

1) agentgonzo(playback) - 16:35
2) dgatsoulis - 17:06
3) mjf - 18:24 (set on May 16)
4) agentgonzo - 19:58
5) blixel - 20:00
6) tinfoilchef - 22:35
7) vsfx - 23:51
8) SolarLiner - 26:37


General rules:

I would like to "resurrect" the classic KSC-WIN Speed Run challenge, as it has been a minor obsession of mine for about the last month, since I restarted playing Orbiter. Beating 20 minutes is the goal for me.

To refresh:
The idea is to use a stock XR-2 Ravenstar (but can be stripped down to hold only 1 crew member and a minimal amount of fuel/oxygen to save on weight). Typical takeoff from KSC and fly approximately 8200km to Wideawake International*.

1) The old-old, ancient history thread is here:
http://orbiter-forum.com/showthread.php?t=5727

2) And a recent thread (circa 2014) is here - this is the best way to familiarize yourself with the most modern efforts to get lower times:
http://www.orbiter-forum.com/showthread.php?t=32000&highlight=classic+speed+run

You will see that a key poster in the recent KSC-WIN challenge thread is "blixel", otherwise known as David Courtney. Mr. Courtney has some excellent Orbiter Tutorials on his Youtube channel, and links are in the initial post in the 2014 thread linked to above...

===

From blixel, I learned that one can fly a "standard" run in about 20 minutes. His fastest time on this one (there are several of his, and other pilots', videos) is 20:00 even (the usual time measure is MET from 0:00 at wheels up in KSC to wheel stop on the runway in WIN). By "standard" run (more on this later) I mean a climb and acceleration to about 11-12km/sec MECO at about 70-72km altitude, flying inverted to hold within the atmosphere, then a flip with about 25% of main fuel to retro-thrust to brake to sub-orbital speed (< 7 km/sec), then flip back again to standard AOA re-entry about 40 degrees for the last 1000-1500km into WIN and (ideally) nicely lined up for a short, steep glide down to the runway.

After a bunch of practice, I have a recorded version that is just under 21:00 (note: the first time I tried it and survived, took me over 30 minutes). I will try to figure out how to post this file if people are interested (but frankly, blixel's runs are better tutorials as he has commented videos).

Diverging from the "standard" run as above, there is also what I call "crazy" runs. This involves flying even faster than 12km/sec - up to about 14km/sec. This means there is no fuel for braking, so the braking maneuver involves basically holding an attitude perpendicular to the airstream, pointing straight down with the belly of the craft facing forward, and using bursts of engine (with minimal remaining fuel) to hold one's altitude under 70km. This produces a very hot, and high-G deceleration. The best examples of the crazy runs are the original agentgonzo run, and the more recent ones by dgatsoulis.

I will try to add more posts on this in the next few days, with some of my lessons learned, checklists , etc - in case anyone wants to give it a shot. But for now, my short-term goals are:

1) get my time under 20 minutes - David Courtney is in my sights! I need to figure out how to optimize the flight to achieve this - I think it has a lot to do with how fast the deceleration/reentry can occur, since this will allow for the maximum amount of time at max cruise speed.

2) try to learn how the "crazy" reentry style works. I have tried it a few times but always get completely disoriented and lose control of the vessel - and either burn up or get flung out into space...

I will try to update this thread with any progress...

 

[Delta glider]

I just tried a run for the first time.
and it failed...
Badly!
My XR2 is now piece of the wideawake runway!:facepalm:

 

[PhantomCruiser]

You're not the first.

 

[SolarLiner]

I remembered recording myself in an attempt ... I failed hilariously!
After those I did a few other runs just under the 20 minutes mark, but it has been quite some time now and I don't think I can even remember the techniques on how to KSC-Wideawake.

Obligatory fail video:

 

[Delta glider]

:thumbup: better than me...:lol:

 

[mjf]

Update - I have added myself in with a flight I just completed this morning (April 7th), best ever time with MET about 20:30. It felt like almost an ideal run, although my landing could have come in a little faster - that probably took at least an extra 10 seconds.

I did record a playback of it, so will watch the playback to try to compare...

---------- Post added at 03:24 AM ---------- Previous post was at 03:12 AM ----------

I remembered recording myself in an attempt ... I failed hilariously!

... Obligatory fail video: ...

yes - that was a good fail video - I've had lots of those experiences...


Now to be the armchair quarterback - watching it from the start, I think your initial climb was too high too fast. For example, as you got up above 60km altitude, you didn't kill the vertical speed fast enough. Thus you blew right through 70km and had to use the rocket force to gradually get you to turn back down at around 85-87km altitude. I think by that point, you're doomed to a sub-optimal flight - and possibly a fatal one.

Then later, of course - you lose control as you go above 75km with minimal fuel to correct it. I think by the time one lets the altitude go above 72km in this challenge, one is pretty much done.

 

[mjf]

Hi everyone - as I mentioned earlier, I was influenced heavily by learning from David Courtney (blixel) on how to get the time of the speed run down to around , or ideally under, 20 minutes.

Here are his most recent videos on Youtube with the best times: run #7 and run #8:

https://www.youtube.com/watch?v=R5ppqbH02NA

https://www.youtube.com/watch?v=b1CApWuGZs0


To help myself learn, I watched these videos and manually recorded at about 30-60 second intervals the following:
- sim time / video time
- MET (mission elapsed time)
- current altitude
- current velocity
- various other things like vertical speed, pitch/AoA, distance remaining to WIN, etc.

I entered the figures into Excel and did some analysis (later, I realized I could run the flight recorder and have the data sampled and saved automatically)...

Anyway, from the analysis of blixel's flights 7 & 8, here are the graphs. The x-axis is time from wheels up, in seconds:



The various points labelled 1, 2, 3 refer to the following key aspects of the flight (see next post)...

---------- Post added at 11:08 AM ---------- Previous post was at 10:57 AM ----------

... from the previously posted image, there are 6 circled points. Recall that the time along the bottom is seconds from wheels up. Blixel flew these runs, and the videos are on Youtube (link in previous post above)...

1) at about 240 seconds (6 minutes from liftoff) we've reached our main cruising altitude of about 70 km. Orbital speeds have been reached (7000-8000 m/s);
2) we've continued to accelerate to MECO of about 12 km/sec. Altitude is held around 70-72 km - enough to keep us in the atmosphere while inverted but not low enough to burn up;
3) "the flip" - altitude is allowed to increase somewhat, the ship is flipped to face retrograde and the burn of about 25% main fuel commences. This burn will decelerate us to below 7000 m/s.
4) burn is complete - flip prograde and enter standard ~40 degree AoA reentry profile. Descent is around 200 m/s and increasing as we head down from 70 to 60 km altitude.
5) re-entry progressing - use Aerobrake to maintain course to WIN with arrival right at the island.
6) final approach and landing - wheel stop around the 12 minute mark.

I would like to make a final note on video 7. As best as I can tell, wheels up actually occurs about 10-12 seconds after the initial roll. But due to playback (I think) the MET timer starts at the roll, rather than wheels up. So I think for blixel's flight #7, the true time is 11:50, rather than 12:02. This means it would have been his best ever. I will check this and edit my post if incorrect...


Have a good weekend, everyone.

 

[dgatsoulis]

My latest run. Slight improvement on the time, but still above 17 minutes.
1026 seconds from "wheels up" to "wheels stop".

 

[Linguofreak]

Has anyone ever tried doing an around-the-world speed run?

 

[mjf]

My latest run. Slight improvement on the time, but still above 17 minutes.
1026 seconds from "wheels up" to "wheels stop".

Nicely done, dgatsoulis - another death defying run, and the time table at the top of this thread has been updated with 17:06 accordingly! Unless agentgonzo comes back to retry this, you are the man to beat on this speed run in 2015... I will study that video a few times, probably...

I notice that although you hold the 180 bank when you are "nose down" (at -85 degree pitch), as you approach the base and the hull temp starts to drop, you start to roll a bit (such that you go to about 150 degrees of bank rather than 180) - is that to give yourself a bit of sideways correction in the airstream so that you end up closer to being right over the base?

---------- Post added at 11:22 PM ---------- Previous post was at 11:17 PM ----------

Has anyone ever tried doing an around-the-world speed run?

I haven't but that is an interesting thought. Assuming a "super cruise" speed of at least 12km per second, and that we'd have to add at least 32,000 km (assuming 40,000 km around the earth minus the 8,000 to WIN) that would require about another 45 minutes of super cruise time to gain the additional 32,000 km. One would have to ensure there was enough APU fuel and O2 to handle the extended run.

I think the atmosphere would be needed to correct (using bank) for the rotation of the earth during the inverted flight, so that one could go (say) from KSC to KSC in one orbit.

Or, if done from Wideawake, we could say that such a run is a "WIN-WIN" situation! :facepalm:

 

[Dantassii]

Or, if done from Wideawake, we could say that such a run is a "WIN-WIN" situation! :facepalm:

Ugh, the puns.

Dantassii
HUMONGOUS IMS shipbuilder

 

[dgatsoulis]

I notice that although you hold the 180 bank when you are "nose down" (at -85 degree pitch), as you approach the base and the hull temp starts to drop, you start to roll a bit (such that you go to about 150 degrees of bank rather than 180) - is that to give yourself a bit of sideways correction in the airstream so that you end up closer to being right over the base?

That's actually a SurfaceMFD (and hud) artifact. As you approach a -90° pitch, the coordinates for the bank have to switch from right to left, in "reality" my wings were more or less level the whole time. You can tell by watching at 16:10 when I recenter the CoG, that I was pretty much level.

I have attached the playback of the flight shown in the video and also added a lua script with some helpful info. Unzip in your Orbiter directory and everything will go in its place.
To watch the playback, run the "1-Race to WIN_1705" scenario, located in the Playback folder.

The script checks if the goal (get to WIN and land on one of the runways) has been met and returns your time in mins:secs.
The timer starts at wheels up and is consistent whether you are flying live, or watching a playback.

You get failure messages if you touch Earth's surface anywhere outside the KSC and WIN runways, or if you break the XR2's gear during the landing.

You can use it in your runs by adding the line "Script RaceToWin" (without the quotes) in the BEGIN_ENVIRONMENT - END_ENVIRONMENT section of your scenario, right below the date.

BEGIN_ENVIRONMENT
  System Sol
  Date MJD 56521.6957058437
  Script RaceToWin
END_ENVIRONMENT

Or, if done from Wideawake, we could say that such a run is a "WIN-WIN" situation! :facepalm:

:lol:

 

[mjf]

I have attached the playback of the flight shown in the video and also added a lua script with some helpful info. Unzip in your Orbiter directory and everything will go in its place.
To watch the playback, run the "1-Race to WIN_1705" scenario, located in the Playback folder.

fantastic - downloaded and playing right now (just paused it to come back to the forums and say "THANKS!")

:thumbup:

---------- Post added at 11:46 PM ---------- Previous post was at 11:41 PM ----------

To all interested in the speed run - I had some additional private correspondence with dgatsoulis on tips and tricks for the run, and he was kind enough to give me permission to post it here for everyone's benefit...

Please see my consolidated version of his advice below.

==

The flight has 4 parts:
1.Ascent
2.Supercruise
3.Brake
4.Landing

You already know how to perform the first part. The only difference in the fast method is that you let the main engines burn longer. In my 17:15 run I let them burn 'til I had ~1900 kg left, but I had also cross-fed the RCS during the ascent, so I had ~50 kg of that.

So your goal for the ascent should be:
Alt ~71-72 km
SCRAM: All used if possible. If not, dump the remaining when the scram-engines overheat.
RCS: ~50 kg (you can go even less than that, in the vid I only needed ~20 kg)
Main: ~1900 kg. That's really pushing it, if you are just now trying this flight, go for 2000 and work your way down in subsequent flights.

The supercruise part is easy. Maintain the altitude using the trim and watching the vertical speed / vertical acceleration in surface MFD.


The brake:
At ~ 2500 km away from wideawake, pitch up, using full trim up and pushing the CoG back, to get the nose pointed towards the Earth. You'll also need to burn the main engines as soon as you do, because otherwise you'll gain altitude.
When the nose is pointed @ ~-60° hit the Kill-rot and get the CoG to about 1.6 back. The nose will keep moving to ~-85° and that's where you want it. don't pitch past -90°, 'cause you will lose control.
While you are doing all this, you also need to watch your vertical velocity and burn/stop the engines to keep the vertical vel @ about -150, -200 m/s. The temperature will tell you when to start/stop the burns.
Keep at it until AerobrakeMFD says that your flight-path ends very slightly past the target.


[NOTE: a secondary message from dgatsoulis gave some clarification/correction on the braking manoeuver - as follows...]

Full trim up and push the CoG back somewhere between -0.2900 and -0.3000.
When you reach an attitude of ~ -75° pitch press killrot and push the CoG slightly forward, somewhere between -0.1600, -0.1630. (In my previous msg I mentioned 1.6, which is incorrect). This will help you keep a steady attitude of about -85° to -87° pitch, without having to use too much RCS fuel fighting the killrot autopilot. It's a bit tricky, because clicking on the CoG button is clumsy and doesn't give you very precise control.
To have better control over the CoG use the [Alt comma], [Alt period] keyboard shortcuts. [Alt M] recenters the CoG.
You will know that you have it set right, when the killrot AP uses almost no fuel to keep the XR2's orientation steady. You may also want to make a slight adjustment on the trim.

During this whole time you also need to keep an eye on your vertical velocity (burn the main engines to keep a negative vel of ~-150 m/s) and the temperature (stop burning when the temp goes yellow/red).



The landing:
At about 1.2 km/s velocity, re-center the CoG and the trim and open the aerobrake and go for the landing.
This is where I probably messed up my flight, so you may want to be more aggressive than me if you want to get a better time. (Perhaps open the airbrake a little later and or recenter the CoG at a slightly higher velocity)

 

[Dantassii]

My latest run. Slight improvement on the time, but still above 17 minutes.
1026 seconds from "wheels up" to "wheels stop".

Orbiter 2010 - KSC to Wideawake - "The Classic" Speed Run. 17min 06 secs - YouTube

Am I seeing that right that you're pulling something like 6-7 G's during that reentry? That's a mighty long time for a human being to tolerate that sort of stress.

Dantassii
HUMONGOUS IMS shipbuilder

 

[mjf]

Am I seeing that right that you're pulling something like 6-7 G's during that reentry? That's a mighty long time for a human being to tolerate that sort of stress

I believe that a highly trained pilot (like our XR-2 crew, I'm assuming) can tolerate below 8 Gs that are positive for several minutes. Depends on how the seats can swivel. If the forces are down towards the feet this table (see bottom) would suggest 6 Gs for only about 3 minutes before blacking out; on the other hand if the force is "eyeballs in" (toward back of head) then 9 Gs for 3 minutes and 8 Gs for 10 minutes:

http://scienceblogs.com/dotphysics/2009/12/16/dangerous-jumping-calculator/


I can't imagine it would be a comfortable experience, let alone trying to fly a non-automated re-entry...

 

[Delta glider]

Fail 2! When do you start the reentry burn?:hmm:
By the way, the failure was my XR-2 disintegrating.:uhh:

 

[mjf]

Fail 2! When do you start the reentry burn?

Hi Delta glider...

I'm assuming that you are doing the "flip and burn retrograde" method as shown in blixel's videos? (For more, see links in references in the original post; this method is slightly safer (though not much) than the crazy agentgonzo/dgatsoulis "point almost straight down" method)...

If true, then for first attempts, one should transition from the inverted supercruise phase to initiating the "flip" to retrograde, at about 3300km to WIN. I have done it as late as 2700km to WIN which is what you have to do for the fastest possible runs, but starting the flip around 3300km to WIN allows for a gentler re-entry.


What happened in your fail #2 that resulted in the XR-2's destruction?

 

[Delta glider]

Thanks. Erm, well I tried to get my path centered on aerobrake MFD but as a result of that burnt up. I did the burn at 2000 km from Wideawake.:facepalm:

---------- Post added at 07:09 PM ---------- Previous post was at 07:09 PM ----------

I only burnt that far because on the first time I burnt way too far away.

 

[mjf]

just an update in case anyone is interested...

Haven't had much of a chance to play Orbiter but I have been trying a few experiments with the goal of a "sub 20 minute KSC-WIN" run...

I have quicksaved a flight where I am in supercruise mode at about 12km / sec, about 4000 km away from WIN. This allows me to practice the braking and re-entries. Using dgatsoulis' advice, I have tried the "point straight down" method of braking, rather than the braking burn, and it is working well - to the point that I can avoid dying, and if the braking is initiated about 2900km from WIN, I come about 300-400 km short. In other words, I could have initiated the braking from about 2500km away from WIN and still made it.

The key is using adequate amounts of elevator trim (the INS and DEL keys) *and* a touch of back weighted Center-of-Gravity shift, to allow for holding at an 85 degree pitch for the atmospheric braking...

Will keep updated as this quest continues...

 

[mjf]

additional update - I am extremely happy, having discovered that the XR-2 attitude hold autopilot can allow for inverted flight...

:thumbup:

to implement this, you have to be flying inverted first - say you get things roughly stable in supercruise mode. Then hit the "SYNC" button on the attitude hold autopilot. You can then adjust pitch/AoA by increments of half a degree, and you can also bank to align your target of WIN to be straight ahead.

Finally, when you get to 2500 km or less, you can straighten out (bank = 180 degrees - ie level inverted) - then you can pitch to 80-85 degrees. The autopilot will hold this attitude, making things much easier to fly as you can then concentrate on using your main engine to push you down into the atmosphere.

I strongly believe this is how I am going to get < 20 minutes in one go. Maybe by this weekend - we'll see...