Comprehensive Guide to Attaining Orbit

[Excalibur]

Recently I\'ve been working on a little guide for some of the people new to KSP and the forums.

Covers the fundamentals of reaching orbit reasonably efficiently - a must for the newbie!

Check it out here on this great site...

http://xenolith.ws/?page=post&id=17

P.S. Any corrections/suggestions please don\'t hesitate to let me know!

[Thobewill10]

Very nice guide!

[Excalibur]

Thanks dude, it\'s the first in what I hope is a series of many guides for KSP. I want to contribute to the community and without any modelling/texturing/coding skills this is the best way I can think of

[0muffins0]

Now that was useful! Especially the part about atmospheric resistance below 10k.

[Excalibur]

Now that was useful! Especially the part about atmospheric resistance below 10k.

Glad to hear it

How you liking KSP? Any questions?

[0muffins0]

Glad to hear it

How you liking KSP? Any questions?

KSP = extremely fun.

What\'s the best way to physically land on the Mun? My current spaceplane design can place a satellite into Munar orbit then return just fine. However, trying to land on the Mun after placing the satellite in it\'s orbit usually ends in explosions. Wow does the Munar surface rush up to meet me! Do I do a big heavy deceleration burn then just drop down almost vertically?

[Excalibur]

KSP = extremely fun.

What\'s the best way to physically land on the Mun? My current spaceplane design can place a satellite into Munar orbit then return just fine. However, trying to land on the Mun after placing the satellite in it\'s orbit usually ends in explosions. Wow does the Munar surface rush up to meet me! Do I do a big heavy deceleration burn then just drop down almost vertically?

\'tis int it

From what I understand there\'s a couple of ways of landing on Mun. One is a direct descent; as soon as you enter Munar SOI you head for the surface. The other option is to slide into orbit after entering Munar SOI and then perform a de-orbit burn. Apparently the direct descent is the most fuel efficient (though not by much) but it\'s much harder.

I usually use MechJeb these days for my landings (getting a bit lazy I know). What I used to do though is get in a low orbit around Mun, say 15-20km. I\'d then make a nice burn de-orbit burn so that my downrange distance to landing (or impact) would be about the same as my altitude (10-20km). Kind of like a right angled triangle.

Then once I\'d descended below about 5-8km (or if my velocity exceeded 300m/s for a measure of control) I\'d make a steady burn with the intention of reducing my velocity to near zero at zero altitude (obviously it\'s not that easy manually so make adjustments as you go). Remember to keep your ship pointed towards retrograde so as to help cancel both horizontal and vertical velocity.

Main thing I\'d say is to keep your ship stable and orientated with your horizontal velocity vector so your controls actually match up with what happens on screen and don\'t let your velocity get out of control. You obviously need to be able to decelerate enough for a soft landing.

[Zephram Kerman]

Very nice guide! Everything is spelled out verbosely, yet not too complicated for the beginner. Lots of big ideas here, so the extra verbiage helps folks absorb it all.

For stage IV, you might want to mention to mouse over the PE marker. That way, you can monitor your progress and throttle down when it starts to get close. If I don\'t do that, I always overshoot.

[Excalibur]

Very nice guide! Everything is spelled out verbosely, yet not too complicated for the beginner. Lots of big ideas here, so the extra verbiage helps folks absorb it all.

Thanks, I wanted to make the it much shorter but too many big ideas to touch on as you mentioned.

For stage IV, you might want to mention to mouse over the PE marker. That way, you can monitor your progress and throttle down when it starts to get close. If I don\'t do that, I always overshoot.

Good point! I\'ve gone back and added that in, which incidentally led me to spot a spelling error so thank you.

[closette]

@Excalibur,

A good start to a plain-language guide. Wish there was something like that when I was starting out. Some of the statements about ascent in the first 10km are not quite right though. I don\'t want to toot my own horn but this post in the Goddard Challenge thread summarizes what Kosmo-not, Jebbe, Izkierka, others, and myself discovered about an efficient vertical ascent:

http://kerbalspaceprogram.com/forum/index.php?topic=7161.msg106460#msg106460

For this 'most efficient' ascent you need at least a TWR of 2, in fact the capability of 2.5-3 would be a sensible range. A TWR of 1.0-2.0 at launch makes for an excruciatingly slow and inefficient ascent (as I found out in my first attempts at rocket building).

MechJeb implements this optimal ascent scheme for the vertical part of the ascent to orbit, but the post referenced

contains an ideal speed vs. altitude table for manual flight. If you follow it closely, you\'ll end up with about 50/50 'gravity loss' vs. 'drag loss', but in practice 60/40 or 40/60 would not be too bad. (The efficiency curve has a broad maximum). If you have the choice, better to go up a little too fast than a little too slow.

And from launch, it\'s best to get up to ~120 m/s as quickly as possible - by 1000 m altitude if you can - so use full thrust (+SRBs), then throttle back if you have to.

You certainly would not want to include all this detail in the guide - perhaps mention that there is a 'compromise' speed between gravity and drag losses on ascent. But I\'ll leave the wordsmithing in your very capable hands.

[Excalibur]

Thanks closette - I\'d missed that post on my travels through the forums!

I must admit I did skim over a lot of the gory details, especially concerning the initial stage of ascent. May update the guide based on the information you\'ve shown me. What I don\'t want to do is confuse anyone or overwhelm them. Might actually add a link to that post at the end of the guide.

Many thanks for your comments though. I based much of the guide on a lot of information \'off the top of my head\' and from what I\'d learnt myself - so there were bound to be inconsistencies and gaps!

[rkman]

And from launch, it\'s best to get up to ~120 m/s as quickly as possible - by 1000 m altitude if you can - so use full thrust (+SRBs), then throttle back if you have to.

In my experience it\'s hard (and possibly inefficent) to get a large rocket (heavy payload) to do 120m/s at 1000m.

I use this simplified version of the ascent profile:

- Don\'t worry to much about the first few kilometers. Primary concern is to get off the ground.

- Aim for at least 200m/s at 10k (ideally ~250 260m/s, more than that is overkill). If less, back to the drawing board.

- Velocity should not decrease {to much} each time when the next stage kicks in. If it does, back to the drawing board.

- When exiting the atmosphere, velocity should be about 1200~1500m/s. There\'s usually enough time before reaching Ap to get near orbital velocity (~2250m/s).

About TWR:

what really matters is the average TWR of a stage over the time it\'s used. TWR of a stage improves as it burns fuel. So a 1st stage TWR of 1.5 at liftoff is likely double that or more by the time when it\'s almost out of fuel, and the average TWR for that stage is quite respectable in spite of the slow liftoff.

[closette]

Fair points - perhaps 100 m/s minimum by 1000 m is more attainable. That\'s where SRBs come in - don\'t leave home without \'em! Lots of thrust for just long enough to get you up to speed, then away they go. The initial TWR for an SRB+radial decoupler is 5.9, which can only increase the overall launch TWR of your craft if it\'s carrying any payload at all (from an ASAS module to a Mun lander).

I agree that by 10km one\'s vertical speed should be in the 200\'s (280 m/s is the optimum ascent speed at that altitude).

When staging I often see a small dip in speed. It\'s hard for this not to happen if you\'re at 100% thrust, because just before you stage compared to just after, you have more engines with nearly-empty tanks and therefore a higher instantaneous TWR. Hard to quantify but a few m/s drop is OK.

As for optimum TWR, I looked back at some calculations I did (attached, cleaned up a bit from an earlier version I posted months ago) and the efficiency vs. TWR curve in that document shows a very broad maximum around the optimum value of 2. In fact any value of TWR from 1.5-3.0 gets you at least 90% of the maximum efficiency.

I would want beginners to know that they should design for TWR of at least 1.5 at low altitude where the atmosphere is thickest. Any less, and your slow slog upwards means your 'gravity loss' will be very large and you\'ll drain your tanks without really getting anywhere or getting close to 200 m/s by 10 km.

[Excalibur]

I would want beginners to know that they should design for TWR of at least 1.5 at low altitude where the atmosphere is thickest. Any less, and your slow slog upwards means your 'gravity loss' will be very large and you\'ll drain your tanks without really getting anywhere or getting close to 200 m/s by 10 km.

Though it wasn\'t really my intention to have efficiency as the focus of the guide, might as well get the beginners off to the best possible start and get them using efficient ascent profiles from the get-go.

I\'ve added some of the info, and your table of terminal velocities as long as you don\'t mind... I did say thank you at the bottom of the guide.

As an aside, recently I\'ve been launching ridiculously large rockets, so my main focus hasn\'t been so much efficiency as just getting the thing off the bloody ground without kerploding!

[rkman]

I would want beginners to know that they should design for TWR of at least 1.5 at low altitude where the atmosphere is thickest. Any less, and your slow slog upwards means your 'gravity loss' will be very large and you\'ll drain your tanks without really getting anywhere or getting close to 200 m/s by 10 km.

Right, and that\'s where boosters might come in handy.

It\'s just that adding more thrust is not always better. Very Kerbal but not always more efficient.

I\'ve built many monstrosities just trying to get to the moon. Most of the fuel and engines on them is there because i put on to much fuel and engines to begin with. It\'s easy to fall into the trap of over-design. It ends with you either being stuck on the ground unable to make takeoff, or caught in an explosion. Both very Kerbal but doesn\'t get you to the Mun.

The TWR sweetspot and thinking about design in terms of \'fuel tanks per engine\', is not immediately obvious.

Is my gut right when it tells me gravity drag is generally more of a problem for heavy payloads that require a large heavy rocket?

I figure that a large craft with an average 1st stage TWR of ~2, making a little over 200m/s at 10k, would actually become less fuel efficient if you\'d add the fuel and engines (or boosters) needed to get it up to 100m/s at 1k. I\'m usually happy if i get to 100m/s at 5k, the next 5k is easier so i\'ll sure make 200+ at 10k.

About gravity turn:

i agree it should start at ~10k, but 'slowly turn' (east) might need some clarification.

I don\'t know if this is near optimal but i aim for 45 degrees tilt at 30k, then to zero degrees at 50k to 60k. It should have enough vertical velocity to get out of the atmosphere.

My orbits are typically at ~80k, trying to make a habit of going for ~100k because that leaves more maneuvering room for orbital intercept missions.

[Zephram Kerman]

About gravity turn:

i agree it should start at ~10k, but 'slowly turn' (east) might need some clarification.

I don\'t know if this is near optimal but i aim for 45 degrees tilt at 30k, then to zero degrees at 50k to 60k. It should have enough vertical velocity to get out of the atmosphere.

My orbits are typically at ~80k, trying to make a habit of going for ~100k because that leaves more maneuvering room for orbital intercept missions.

I\'m not sure about this, but intuition tells me the ideal gravity turn varies depending on your TWR. If your rocket is under-powered, a shallow angle should help reduce gravity drag while the airspeed tends to stay below terminal velocity anyway. An over-powered rocket should use a steep angle, to get clear of the atmosphere early, so as to put that extra thrust to work asap. This is probably not important enough for an introductory lesson, but worth discussing here.

[closette]

Unlike a purely vertical ascent (known as the 'Goddard problem'), finding the most efficient ascent-to-orbit profile is a difficult optimal control problem and unique to each design, although there is a Challenge now on the forums to get to 100km orbit with the least delta-v, and some players are doing better than MechJeb I believe (yay humans!).

Without much to back it up, I tend to start my gravity turn when the craft can no longer 'keep up' with the optimal terminal speed for vertical ascent. My ad hoc reasoning is that 'gravity drag' is then dominating over 'air drag' so it\'s time to start building up orbital speed.

By 30km altitude the terminal velocity for most craft is equal to the orbital velocity, so I try to be pitched over by at least 45 degrees by then. But not too far over, or you\'ll have to push through many km of atmosphere before getting out!

I used to be fussy about steering the craft exactly to align the thrust and velocity exactly during the gravity turn, but I learned from MechJeb that 'steering losses' are generally small compared to gravity and drag losses. It\'s even possible that such a pure gravity turn is NOT the optimal ascent profile for some rockets.

But these are details that should probably not appear in a beginners\' tutorial - better that they be discovered by practice and patience using a given launch vehicle. I\'m still learning from my mistakes!

[ferram4]

The gravity turn is only the optimal ascent profile if there is no atmosphere, though in a dense atmosphere the gravity turn makes sure there are no side forces on the rocket, since IRL rockets aren\'t built to handle shear stresses (but apparently Kerbal rockets can handle falling sideways at 150 m/s *shrug*).

I think that the optimal trajectory actually depends greatly on how much control authority the rocket has at any given stage; if it has a great deal of control authority, than going straight up until ~10k and pitching over sharply is efficient, since it trades air drag losses for smaller steering losses. If the rocket is more of a giant space station carrying behemoth, pitching over earlier (~5k) and taking a smoother trajectory is better, just because it becomes easier to control.

I tend to pitch over between 5 and 10k, depending on the rocket and I try to smoothly pitch until I\'m matching the orbital prograde marker at ~60k. I tend to try pitching over faster when it feels like the rocket is high enough that the air doesn\'t matter too much anymore and picking up speed is more important.

You should probably mention to beginners that 3 fuel tanks per large LFE results in a TWR of ~2 (not counting a command pod, decoupler or parachute ), just as a rule of thumb.

[Kosmo-not]

Did you mean to say that the gravity turn is optimal only when there is an atmosphere?

[rkman]

I think ferram means an atmosphere complicates the optimal ascent profile, without an atmosphere it would just be a pure gravity turn. What we usually call 'gravity turn' regarding launch from Kerbin is not a pure gravity turn.

I think the main difference is that on Kerbin it\'s more efficient to go straight up for the first 5 to 10km to get out of the thickest part of the atmosphere, while on the Mun it\'s more efficient to turn as soon as you\'re off the ground.

I\'ve realized that something i said before probably is not true: i said the second 5km of launch (from 5k to 10k) is easier than the first 5km. But it seems that reduced air density and improved TWR does not outweigh air drag that\'s increased due to increased speed. Doing 100m/s at 5k means you\'ll probably have a hard time getting to even 200m/s at 10k (as opposed to what i said earlier: that you\'ll easily make 200+m/s).