1.3 Rocket Ascent Profile and Gravity Turn

[A_name]

This is an in-depth tutorial, but still directed to beginner-intermediate players, on how to do a proper launch and gravity turn with the new aerodynamic model introduced as of version 1.0. This tutorial works for versions 1.0 to 1.3.

More than giving a script or set of instructions, my goal with this tutorial is for you to gain an understanding of the factors that affect your rocket's behavior during launch, so that you can apply it to any rocket you fly. For that, you'll need to go through the entire post, but I'm also including a TLDR as a "cheat sheet":

TL;DR (courtesy of @kBob)

1. Turn ON SAS and set throttle to give TWR of ~1.5.

2. Launch!

3. When your speed reaches 50 m/s, perform a pitch over maneuver (tip towards the East until pointing between 5° to 10°).

4. When SAS stabilizes (i.e. the control input arrows on the bottom left are all centered), turn it OFF. Avoid control inputs and use only throttle to control your gravity turn (throttle up to turn slower, throttle down to turn faster).

5. When your altitude reaches ~40 km, turn SAS ON. Start pitching down manually towards the horizon and adjust throttle to keep your Ap around 45 seconds in front of you.

6. When your Ap reaches the desired altitude, cut your engines, coast to Ap and circularize.

==========================================================================================

General Notes on Gravity Turn

You all probably know by this point that to get into orbit you need to go up, above the atmosphere, but you also need to go sideways (i.e. horizontally) very fast. To do this, we could launch straight up until we're out of the atmosphere, then point sideways and accelerate to orbital speed. But that would be very inefficient. We want to launch in a way that we gradually turn sideways while we ascend. This is called a gravity turn. The best way is to do a real gravity turn; that is, a turn caused by gravity and aerodynamic forces, rather than one achieved by actively turning the rocket. It is important to keep this in mind.

 

Design Items

Before even launching, you need to take these design items into consideration when building your rocket:

TWR: Your thrust-to-weight ratio (TWR) at launch should be relatively low, around 1.5. A higher TWR at the beginning of the launch makes it harder for your rocket to turn naturally, as gravity will have less influence on its trajectory, making it fly straight and screwing up your gravity turn. Keep in mind that drag losses are almost negligible in the new aero, unless your rocket is shaped like a brick or you are going extremely fast in the lower atmosphere. Thus, a slightly higher TWR of around ~2.0 in theory is more efficient, but only if the launch profile is flown correctly. The drawback is it makes your rocket less forgiving in terms of control during ascent, and it shortnes your widnow to make a pitch-over maneuver. What usually ends up happening is that you have to force the gravity turn manually, which does generate significant drag (because you expose the sides of your rocket to the airstream anytime you deviate from your prograde vector), and causes steering losses (Dv wasted on changing direction rather than gaining velocity). This reduces overall efficiency and defeats the purpose of having a higher TWR to begin with. A higher TWR also causes increased stress to the craft, inducing wobble and risking a RUD, especially when trying to maneuver. You may experience heating issues too. For these reasons, in my experience, a TWR of ~1.5 is a good sweet spot between efficiency and controllability of the rocket. Smaller and lighter rockets handle higher TWR's better than big and heavy ones and each craft will have its own sweet spot; you are encouraged to experiment. If you find your TWR at launch is too high, either use a smaller engine or just throttle down, and vice versa. As a final note, all rockets will have their TWR go up as the launch progresses due to shedding weight by burning fuel. This is normal and you should manage by reducing throttle throughout the ascent as needed (more on this below). You can check your TWR with the Kerbal Engineering Redux mod (KER) or with MechJeb, or if you're running a stock game, the G Force meter roughly doubles as a TWR meter (if the G Force meter is pointing at 1 your TWR is roughly 1, and so on).

Aerodynamic Stability: You want your rocket to be aerodynamically stable. That means that it will have a natural tendency to fly straight, instead of, say, sideways. Any object that flies through the atmosphere will naturally orient itself with its center of mass (COM) facing forwards relative to its trajectory and its center of drag (COD) facing backwards. You can see this in darts, arrows, badminton cocks, etc. Similarly, you will want to have your rocket's COM in front of your COD. To ensure this, add 3 or 4 winglets or wing surfaces with radial symmetry at the base of the rocket, and if possible cover your payload in a fairing to make it more streamlined. If your rocket insists on flipping, you need to add more/larger wings at the bottom. If that still doesn't fix it, it means your COM is shifting back too much as fuel is burnt. The heaviest part of a rocket in KSP is usually the main ascent engine(s), so the COM will tend to move back as fuel is spent. The easiest way to fix this is to add a small fuel tank at the top of the stage that's experiencing the problem and lock the tank in the VAB (right click on the tank and select the green arrows for both fuel and oxidizer). This fuel tank will act as ballast keeping your COM forward. You can unlock it manually in flight when the rest of the stage's fuel is gone so as to not waste it, and then stage as normal.

 

Ascent Profile

Once you've implemented the above design items, follow these steps for your ascent:

1. Turn on SAS and set your throttle to whatever will give you a TWR of ~1.5.

2. Launch!

2. As soon as your speed hits 50 m/s, perform a pitch-over maneuver to begin your gravity turn. To do this, tip your rocket towards the East slightly, until it is pointing between 5° to 10°. Don't start pitching over before your speed is ~50 m/s, otherwise you will likely find yourself horizontal within a few seconds, as your winglets won't be biting into the air hard enough to provide stability. The higher your thrust, the more you need to pitch over initially, because higher thrust makes the rocket want to go straight. If you're using a TWR higher than 1.5, your pitch-over should be to at least 10°.

3. As soon as your SAS stabilizes (i.e. the control input arrows on the bottom left are all centered) turn off the SAS. Watch closely for this moment, as you will have only a small window of a few seconds at most before the SAS starts trying to resist the gravity turn. Turning SAS off while it's trying to steer will cause your rocket to become unstable and lose its heading or possibly break up. You should be done with your pitch-over maneuver and have your SAS turned off by the time your velocity is around 100m/s. If you take too long and your rocket is going too fast by the time you're done, it won't want to continue turning (fast rockets like to go straight, remember?) and you'll have to force the turn manually, which is inefficient and causes stress on your craft. As mentioned above, a gravity turn should happen on its own and not as a result of control input. For particularly unwieldy rockets, you can lock SAS to prograde instead of turning it off during this phase. However, stock SAS is far from perfect and it's best to let gravity and aerodynamic forces do the steering for you. If you do use SAS, be sure to disable it before you hit 35 km to avoid you craft from jolting down suddenly when the navball automatically switches to orbit mode, which happens at around 35 km.

4. Enjoy the view while your prograde marker gradually sinks towards the horizon; your rocket will follow on its own thanks to gravity and aerodynamic forces. Try to avoid control inputs during this phase (i.e. no AWSD), just let it fly. If you need to make adjustments, use throttle. Remember, lower thrust means the rocket turns more, higher thrust makes it want to go straight. At about 10 km altitude, you should be pointing roughly to 45° and your speed should be around 500 m/s. If at 10 km altitude you're still pointing above 45°, your TWR was too high and you went too fast and/or your pitch-over maneuver was too gentle. Next time throttle down more or make a more aggressive pitch-over maneuver. On the other hand, if you're pointing below 45° at 10 km, you went too slow and/or your pitch-over maneuver was too aggressive. Next time use higher thrust or do a gentler pitch-over maneuver. If your rocket flips on its end at any point, it's not aerodynamically stable enough. See above under "Aerodynamic Stability" for possible solutions.

5. At around 40 km altitude, turn SAS back on and start steering manually; use pitch and throttle to keep your Ap around 45 seconds in front of you. Any time you're burning above the horizon, you're wasting part of your thrust to gravity instead of gaining horizontal speed; this is called gravity losses or gravity drag. In the initial stages of the launch, you can't help incurring gravity loses because you need to gain vertical speed to get out of the atmosphere. The atmosphere also means you can't steer away from the prograde vector without inducing aerodynamic drag, steering losses and/or destabilizing your rocket. However, by the time you get to ~ 40 km, you'll have enough vertical speed and the atmosphere will become negligible. Thus, at this point you want to begin gradually pitching down towards the horizon. During this phase, you will also start adjusting your time to Ap. It's most efficient to perform your orbital insertion burn right at Ap, so you want to keep it "hovering" only a few seconds in front of you. Of course, you don't want to it to get too close either, otherwise you risk passing it and falling back down into the atmosphere. A time to Ap of ~45 seconds is a good rule of thumb to balance safety and efficiency. To control your time to Ap, use pitch and throttle. If your time to Ap is more than 45 sec, throttle down a bit and point more horizontal, and vice versa. Avoid pitching below the horizon. Continue adjusting pitch and throttle until your Ap reaches your desired altitude, at which point you can cut your engines, coast to Ap and circularize. Note that as you approach orbital speed, there will be a point when your Ap will begin shooting away even if pointing straight at the horizon and no matter how much you throttle down (unless you cut the engines of course). If you reach this point, just let it go until engine cutoff; any efficiency gains from keeping your Ap near you will be negligible by then.

 

Advanced Mode

Try doing the ascent and orbital insertion in a continuous burn. This is the most efficient profile (citation needed) and it's extremely satisfying. Easier said than done, though. To pull it off, you need to allow your time to Ap to creep closer and closer during steps 4 and 5, while not allowing it to get higher than your intended orbital altitude. You do this by reducing throttle and lowering your pitch in a more aggressive manner. The closer you are to orbital velocity, the closer you can allow yourself to get to your Ap. You want to hit orbital velocity exactly at Ap. There will be much trial and error and the exact procedure will vary from rocket to rocket, but give it a try!

Edited December 10, 2017 by A_name

[A_name]

Updated for 1.0.5.

[rodentgun]

Oh goodness. This post really helped me make the best ascent I ever did. No amount of watching videos and mucking about with gravity turns worked as well as this post. It seems that letting Gravity do the work for you is the most efficient way. So many extra DeltaVs now in LKO, what to spend them all on!!! Thank you A_name, registered just to thank you for your guide.

Edited December 29, 2015 by pistolhamster
more thanks!

[jkerbalwatson]

Echoing the previous comment: this was so helpful! I can haz all the delta-v's?!

[rodentgun]

Just a follow up on my experience with this launch technique. Some of my designs dont Work at all with this. I seem to experience a lot of topsy-turvying and rocket flipping around 8 to 10 km altitude. I guess I used to SAS bruteforce me through this, but now I redesigned some of my rockets.

I went from a 109 ton rocket to propel a 13 ton lander into LKO and down to a 65 ton Vessel to launch the same lander into LKO. Done with 3200 m/s only can you believe it?

Further experience:

With my medium sized rocket of sub 100 ton the 45 seconds to apoapsis and 1.50 TWR seems too Little. My rocket appears to be aerodynamically very stable, but at this speed it drops into the horizon too quick. If I keep the time to Apoapsis at 45 I will never make it. But if I keep it at around 1 minute it seems to work better.

 

 

Edited December 30, 2015 by pistolhamster

[A_name]

Thank you all for your input!

 

6 hours ago, pistolhamster said:

Just a follow up on my experience with this launch technique. Some of my designs dont Work at all with this. I seem to experience a lot of topsy-turvying and rocket flipping around 8 to 10 km altitude. I guess I used to SAS bruteforce me through this, but now I redesigned some of my rockets.

I went from a 109 ton rocket to propel a 13 ton lander into LKO and down to a 65 ton Vessel to launch the same lander into LKO. Done with 3200 m/s only can you believe it?

 

 

For the rocket flipping, use a locked fuel tank at the top of your stage as ballast. I added some details on this above in the "Aerodynamic Stability" section under "Design Items".

The TWR of 1.5 is at launch and should be higher than that by the time your Ap starts becoming relevant. But this is a good point, I'll add to the guide.

Glad you were able to optimize your design

 

Edited December 30, 2015 by A_name

[rodentgun]

Thanks for the extra help, A_name. One thing that puzzles me is that the further I go up, the longer away my time to apoapsis seem to go. Last ascent it went to past 3 minutes as I hit my apoapsis of 82 km. I still made the ascent into LKO with around 3200 m/s, had 22 left in the main booster tank when my orbit was complete so I think my ascent was very efficient. But the numbers you cite and the ones I see don't match for some reason. Perhaps because my rockets are smaller?

 

i am using just one stage from launch to orbit for launches in Kerbin SOI so far, btw. This might also affect upper atmosphere behaviour compared to your guide.

Edited December 30, 2015 by pistolhamster

[PvtSteyr]

The 45 seconds to apoapsis is not mandatory but it is easy to remember and has the added benefit of usually getting your rocket into orbit. Your rockets, pistolhamster, have already reached the necessary velocity to reach a 82 km altitude. This doesn't necessarily mean your launch was inefficient, it just means your first stage had a higher TWR as your first stage was burning out. A constant burn to orbit requires a lot of fine tuning to achieve, it feels nice when you do it but it is a luxury when attempt to go to orbit.

Edited January 2, 2016 by PvtSteyr

[rodentgun]

Ah, thank you pvtsteyr. I have only played KSP since december so the science and fine art of ballistics is new to me. Learned what I have from this I changed the way how I launch Spaceplanes too. Instead of sharp pitching up I gently raise the nose up above Kerbin's horizon when my jet hits max thrust at 11 km. By slowly nudging the prograde vector upwards until I hit 35 degrees I have now A good 20% efficiency boost at least. More deltaVs!

 

anyways, what IS the necessary velocity to reach orbit around Kerbin? something above 2200 m/s?

[Warzouz]

Your tutorial is nice, but I wouldn't suggest to activate SAS on launchpad. On the opposite, I strongly suggest to deactivate SAS until the ship hits 30km. SAS increases wobble.

On few exceptions, I activate SAS more quickly if it's hard to keep the ship near prograde.

Further more, when the ship stops burning, switch SAS to prograde until you hit 70km.

[wumpus]

On 6/30/2015 at 1:10 PM, A_name said:

(Note 1: Some people will say 250 m/s at this phase of the flight is too slow. This is true for very light and aerodynamic rockets. With bigger crafts, drag is higher so you don't want to be going supersonic (i.e. faster than ~300 m/s) until you're higher up in the thinner air. Going too fast in the lower atmosphere will cause heating issues as of 1.0.5. Also, this approach provides for a much stabler flight in my experience, especially with bigger rockets.)

 

If this is true than the aero/physics needs another overhaul (although I suspect squad is fine with it and you should be aware of the issue and use FAR if it is too big).  While a bigger rocket might have greater drag then smaller rockets, bigger rockets (at least taller rockets) have (historically, anyway) much lower drag/weight ratios.  In other words, while there might be more overall air resistance, compared to the total weight of the thing being heaved into space it becomes less important.  This is the reason on why the Pegasus flies: as small as it is it needs to launch at 39,000ft (12,000m).  The big boys don't care and launch from sea level.

Personally, I prefer high TWR rockets (and thus higher speeds).  This has a few efficiency benefits, but don't try to do a "real" gravity turn with them: going hypersonic pretty much locks your attitude in place.  You need to angle your rocket in the right direction any way you can before getting locked in place.

[Dr.Spacebird]

I find ~30 sec. to be a pretty safe Ap margin, and that's with my (until this moment) extremely inefficient piloting.  I think it's worth pointing out that you have some control over the rate you approach Ap. with minor thrust vectoring.  Even if it's not a peak-efficiency maneuver, it beats falling back to Kerbin because you didn't leave a margin!

Thanks SO much A_name, this is the most useful efficiency advice I've found yet!

[tranenturm]

Just got to orbit with this method even after going into a spin at 40k cause I forgot the SAS. Did take me a few tries because I'm incapable of reading the nav ball and was tilting over 20 degrees instead of ten.  But the technique is awesome!  (It's my skills that suck) You just made me a better rocketeer!  Thanks. 

[lextacy]

I really love this mod. Good work. I do have a small issue with this though. Can you please include a feature so that gravity turn can respect action group launches? It take about 3 or 4 action groups to launch certain rockets.

[A_name]

On 1/2/2016 at 7:14 PM, PvtSteyr said:

The 45 seconds to apoapsis is not mandatory but it is easy to remember and has the added benefit of usually getting your rocket into orbit. Your rockets, pistolhamster, have already reached the necessary velocity to reach a 82 km altitude. This doesn't necessarily mean your launch was inefficient, it just means your first stage had a higher TWR as your first stage was burning out. A constant burn to orbit requires a lot of fine tuning to achieve, it feels nice when you do it but it is a luxury when attempt to go to orbit.

Agreed.

On 1/6/2016 at 11:08 AM, wumpus said:

 

If this is true than the aero/physics needs another overhaul (although I suspect squad is fine with it and you should be aware of the issue and use FAR if it is too big).  While a bigger rocket might have greater drag then smaller rockets, bigger rockets (at least taller rockets) have (historically, anyway) much lower drag/weight ratios.  In other words, while there might be more overall air resistance, compared to the total weight of the thing being heaved into space it becomes less important.  This is the reason on why the Pegasus flies: as small as it is it needs to launch at 39,000ft (12,000m).  The big boys don't care and launch from sea level.

Personally, I prefer high TWR rockets (and thus higher speeds).  This has a few efficiency benefits, but don't try to do a "real" gravity turn with them: going hypersonic pretty much locks your attitude in place.  You need to angle your rocket in the right direction any way you can before getting locked in place.

This is a good point. I don't know for sure that bigger rockets have higher drag. The more important factor at play though is that when flying big rockets any deviations from the prograde vector are harder to manage and are more likely to make you lose control or overheat due to exposing the side of your rocket to the air stream.

[A_name]

On 1/8/2016 at 7:00 PM, Dr.Spacebird said:

I find ~30 sec. to be a pretty safe Ap margin, and that's with my (until this moment) extremely inefficient piloting.  I think it's worth pointing out that you have some control over the rate you approach Ap. with minor thrust vectoring.  Even if it's not a peak-efficiency maneuver, it beats falling back to Kerbin because you didn't leave a margin!

Thanks SO much A_name, this is the most useful efficiency advice I've found yet!

30 seconds is doable with most rockets and even more efficient than 45 sec, but it's always easier to throttle down and coast for a while if you're over shooting than to maneuver out of a possible falling-back-down.

On 1/28/2016 at 8:12 PM, lextacy said:

I really love this mod. Good work. I do have a small issue with this though. Can you please include a feature so that gravity turn can respect action group launches? It take about 3 or 4 action groups to launch certain rockets.

Sorry man, this is not a mod, just a guide.

On 1/4/2016 at 6:22 AM, Warzouz said:

Your tutorial is nice, but I wouldn't suggest to activate SAS on launchpad. On the opposite, I strongly suggest to deactivate SAS until the ship hits 30km. SAS increases wobble.

On few exceptions, I activate SAS more quickly if it's hard to keep the ship near prograde.

Further more, when the ship stops burning, switch SAS to prograde until you hit 70km.

Yeah, SAS is questionable. You have to babysit it so it doesn't over correct during the pitch-over maneuver. I don't have the fine touch to do it manually, though. Anyway this is the method that works for me.

[RokRoland]

I just found this guide and it appears extremely helpful and might help me move away from the "enough engine and booster to lift off at the verges of overheating due to acceleration" design philosophy, which necessitates launches that point up to 20 degrees away from directly up. Specifically, the points about SAS seem to be worth serious consideration.

One thing I have been a bit baffled about is the shift of prograde when changing from surface velocity to orbital velocity. I suppose at this time the craft should be sufficiently far from atmosphere for attitude to matter much less. Then again I have a lot of experiments with spaceplanes which have good thrust in atmosphere, low thrust in vacuum and generally pretty bad drag characteristics so my thinking is probably skewed some, because I have had to consider whether to burn with the jets in vertical to get rid of drag, or horizontal to actually attain orbital speed, all the while micromanaging the shifting engine efficiency due to speed and altitude.

Those experiences aren't based on gravity turns but consider the same relationship between aero forces versus thrust amount&direction and gravity. I am hoping this guide will help me (as others have noted above) to launch them more efficiently too.

On a related note, I wish there was some kind of "Kerbal Wind Tunnel" mod, it could help you get rid of that "citation needed" part too.

 

[JacktoseIntolerant]

THANK YOU! This is exactly the explanation I've always wanted!

I've been keeping SAS on and manually forcing the turn. Now I can't wait to go fly a more efficient profile, more easily.

Question about SAS: once you have more advanced SAS available, can you have it follow prograde after the pitchover? Won't that do the same as the ideal, straight-to-orbit case, so long as you manage your throttle?

[Blaarkies]

10 hours ago, JacktoseIntolerant said:

THANK YOU! This is exactly the explanation I've always wanted!

I've been keeping SAS on and manually forcing the turn. Now I can't wait to go fly a more efficient profile, more easily.

Question about SAS: once you have more advanced SAS available, can you have it follow prograde after the pitchover? Won't that do the same as the ideal, straight-to-orbit case, so long as you manage your throttle?

Yes, that with good gimbal engines that is how you do fin-less rockets. Be careful though, your ascent profile is too steep, you will need to lower thrust...but thrust is also your control authority.

[jcddmdad]

I'm new to the game, and was wondering why my rockets either over-heated, or tumbled from the sky...I know that you're supposed to do a pitch-over maneuver, but didn't know when.  This guide is just what I needed!  Thanks!!

[KocLobster]

This is an awesome guide man, thank you. Before reading this guide, I had basically started doing what this guide suggests without fully understanding why.

[RocketSquid]

Hmm... 10 Degrees at 50 m/s? That seems, well, early. That low in the atmosphere you would experience some serious drag. I typically wait until between 20 and 30 kM to do a full gravity turn. Is there a particular size range where one is more practical, or am I mistaken?

[wumpus]

18 hours ago, RocketSquid said:

Hmm... 10 Degrees at 50 m/s? That seems, well, early. That low in the atmosphere you would experience some serious drag. I typically wait until between 20 and 30 kM to do a full gravity turn. Is there a particular size range where one is more practical, or am I mistaken?

The point of 10 degrees at 50 m/s is to get the gravity turn started.  You want to be beyond 45 degrees at 20 to 30km, so you have to start early.  The new aero model with 1.0.x isn't kind to the old "suddenly jerk your rocket over" of the earlier releases.  That said, the aero model also allows greater efficiency through reduced gravity losses by having trust-to-weight ratios higher than 1.5 (although much less than 2.0) but you will have to manually turn the rocket (and get steering losses, but they will be offset by your lower gravity losses).  As mentioned in the original post, fast rockets don't like to be steered.

The problem with going straight up (assuming a 1.5 TWR) is that you are spending three times the thurst for every bit of acceleration you get.  You want to get sideways as soon as possible so that you get as much acceleration from that thrust as you can.  Obviously, you can't just go sideways from the launch pad (unless you are on a airless moon), but 45 degrees at 15km works much better than 0 degrees at 15 km.  You could presumably try a late gravity turn with a TWR of 2.0 (or more, but I think kicker SRBs explode from overheating if you are over 2.1), but I don't know how you will find the means of turning at the speeds you will have (and you pretty much wasted all the speed you generated going up in the wrong direction).

The first thing you need to learn about orbital mechanics is that velocity is everything.  Height is only a side effect from getting your speed sufficiently high, and only needs to eventually hit 70+km (note that real rockets barely need any circularization burn.  They have achieved nearly all the needed velocity on the way up).

[fourfa]

wrong thread, no forum option to delete posts

Edited March 15, 2016 by fourfa

[max_creative]

Most of my rockets are about 3/4 as tall as the VAB. Does this still work? (I assume not)