How to do a good gravity turn

[Domax]

So, when I look at the Delta-V chart, going into LKO must "use" 3400 m/s of Delta-V. But me when I get to orbit I "use" approximately  I use at least 4200 m/s of Delta-V. I know it's quite impossible to do as the Delta-V chart say but I find that I use too much Delta-V for LKO and I think that the problem is from my gravity turn. So do you have some tips to do an efficient gravity turn ?

 

Thank you

 

Fly safe

[Human Person]

I usually aim to hit 45 degrees at 10km altitude. This applies to a medium TWR rocket. When your first stage has more thrust, you can be a little more aggressive, if it has less thrust you have to fly a higher arch. 

I had a small rocket with 3 thud engines on the first stage and a terrier 2nd stage that could get to LKO with just about 3000m/s. It could get away with hitting 45 degrees at 5 km altitude.

also, in order to reduce drag you should fly your turn as smooth as possible. Ideally you give your rocket a little initial sideways  nudge and then set SAS to prograde,

Edited February 23, 2018 by Human Person

[Rocket In My Pocket]

Also make sure you are heading out over the ocean.

You will benefit from the rotation of the planet this way and net a bunch of free Dv.

I typically rotate my rocket to face the correct way in the VAB, so I don't have to rotate before performing my turn every time I launch.

Edited February 23, 2018 by Rocket In My Pocket

[Streetwind]

Welcome to the forums!

Perhaps you can describe us your current technique? Remember, the best way to get an answer on the internet is to say something wrong!  

But yeah, in general it's about two things, just like that other guy who is definitely not a dinosaur in disguise said: starting early but very gently, and hitting certain "milestones" of altitude and tilt angle. The term "gravity turn" implies that gravity turns you, not your steering input; the skill in flying a good gravity turn, then, is to set up the initial conditions (the "pitchover maneuver") so that your rocket will fly itself without the need for further input. And yes, this is pretty hard to do with most rockets, and you are just human, not a computer. You'll probably be adjusting manually throughout the ascent. I know I often do. It's the end result that counts, which is reaching orbit with an adequate expenditure of dV.

If you want to practice, I always recommend the following training rocket: mk1 pod, mk16 parachute, 1.25m heatshield, decoupler, two FL-T800, swivel engine. It is intentionally just a little difficult to handle for a newcomer, but perfecty capable in a veteran's hands. Fly this rocket into an 80x80 km orbit, then return the Kerbal safely to the ground. If you can manage that, you can fly well enough to get the numbers on the dV maps.  

Edited February 23, 2018 by Streetwind

[Zhetaan]

1 hour ago, Domax said:

I use at least 4200 m/s of Delta-V

Well, you're not wrong about it being inefficient.  However, that does not automatically mean that the turn is suspect.  Rocket design and aerodynamic efficiency play a role, as well.  This can range from the obvious issues such as pancake-shaped and heavily-greebled versus smooth and needle-shaped rockets ploughing through the atmosphere to less obvious issues such as a low thrust-to-weight ratio in your first stage.

I'll assume that you have a good shape with no extra parts studded to the surface and a 'good' first-stage TWR of at least 1.2-1.3 on the pad.  Try to keep your TWR below 1.6 until you have these turns figured out; too-high thrust at the start requires a more aggressive turn and it's better to understand why that is abnormal than to begin with it and assume that it is normal.

Usually, I launch straight up until I get to either 100 m/s or 1000 m altitude, which generally occurs at about the same time; this is to establish some momentum and also to give the fins at the base of my rocket some wind to bite, which generally stabilises the rocket's flight.  Turning earlier than that is valid, but it's more difficult to control because it's easier to overshoot the proper angle.

At the end of the straight-up portion, I pitch the rocket about three to five degrees east and hold it there until the prograde marker pitches that far east as well.  Some people advise setting SAS to prograde; I will tell you that a properly-designed rocket doesn't need it.  You can shut SAS off entirely and it will fly to space on its own.  Setting SAS to prograde is generally needed for either feel-good value for you or to help an otherwise-good but slightly-unstable rocket make it to orbit.

@Human Person is right about aiming for about 45 degrees at about 10 km; use this as a rough guide, but don't revert if you hit 45 degrees at 11 km or something like that.

For the rest of the rocket, I try to design stages to burn for about two minutes each with the possible exception of the third stage, since that's usually my orbital manoeuvring stage.  Your second stage TWR doesn't need to be more than about .8 and it shouldn't need fins if you have a gimballing engine.  Your third stage doesn't need to be more than .5 if you're using it to circularise; it can be a goodly amount less if you circularise on your second stage.

Have a look at @Norcalplanner's Cheap and Cheerful Rocket Rules of Thumb for some ideas; the thread is old and has some formatting problems, but the ideas are solid.  This link gives a better step-by-step guide on gravity turns, but it advises a pad TWR of 1.5 and an initial pitch of 5 - 10 degrees at 50 m/s.  As you can see, there is some room for variation of style in gravity turns; the best thing for you to do is to go into Sandbox mode and try it a few times.

Also, welcome to the happy collection of (only moderately) mad rocket scientists!

Edited February 23, 2018 by Zhetaan

[Flavio hc16]

I can tell you that when you become very good you can reliably reach orbit with 3200m/s of delta v, for some lucky rockets and perfect gravity turn, as @Human Person said, you can even do with 3k, 3.1k of delta v.

That said,  tips to get your rocket to be more efficient:

1) Make them aerodynamic!! IF your rocket are as aerodynamic as a brick, you will have to fight hard against drag for the biggest part of your ascent ( up to 25-30k of altitude), so a very elongated and long shape is very useful. And don't get too carried away with control surfaces (flap, wings etc...), if your rocket flips or is unstable, 95% of the time you are designing it wrong

2)Don't exaggerate with your TWR at launch, and that's a mistake that a lot of people ( me too) do/did in their early days, if real rocket have a twr at the launchpad between 1.2 and 1.6, there is a reason, and you should do the same ( IMHO 1.2 is on the low side, i always aim for 1.35, and 1.6 is my upper limit), and that reason is drag: you will start accelerating a lot where the atmosphere is thick ( under 20kms) and your ship has to waste a lot of energy on just getting through the atmosphere .

3) that is a consequence of the former rules, one of the most efficient ways to get in orbit is to doing a somewhat slow burn that never stop firing up until you have reached your AP, so you should try to aim for the coasting period to be as short as possible ( the period where your engine is not firing during ascent) or,  putting it in another way: your rocket should make a  3.4k delta v maneuver in 2.30-3 minutes time.

4) point at your east ( 90° on the navball) so you use the kick from the rotation of the planet, and that are another 175 m/s that you save from circularization.

 

That's how I do for my  average rocket ( something that start with 1.35 twr and has the burnout at 3.5-4 ) 

lift off, going straight up, at 120m/s and 1200ms of altitude i point East with an inclination of 80°, and from there on i stick my rocket to the prograde marker using sas,

at 6500 ms of altitude, i reach Mach 1 and my rocket is pointing at 65° of inclination (always pointing east prograde)

at at 13.000meters i breach mach 2 and are pointing at 50°

at 22k meters i'm at 1 km/s and pointing 40°

at 45 kms i'm at 1.5 km/s pointing 20°

i stop my burn at 2 km/s , with 10° of inclination  and 60 kms of heigh, then do the 40 sec coasting to 75 kms and spend the remaining 250 m/s of delta v to circularize

[Domax]

Ok,

I read all of your comment and I think I found the issue. I always have a TWR of 2 or higher so the prograde marker stay higher than the nose of my rocket. 

When I launch a rocket I wait until 100 m/s and I start to turn to about 10 degrees. When I hit 12000 m I'm about 70 degrees.

It's funny because I bought the game in April 2013 (yes I'm going to have the Making History Expansion for free !!) and I'm still not capable of making an efficient LKO. My rockets have always way more Delta-V than they should because none of my manoeuver are efficient. I can go every where but I know that my efficiency is not efficient  I don't have much knowledge on how to do certain type of manoeuver (Gravity Turn, Sucide Burn etc...)

[Laie]

1 hour ago, Flavio hc16 said:

2)Don't exaggerate with your TWR at launch,

Question is, when are you exaggerating? Consider this example of making orbit with much less than 3km/s. It's a expensive booster with a dummy payload, but I think one can learn of it nonetheless:

 

Gravity turn for dummies: tip your rocket over, a little, right after launch, then just follow prograde until you reach apoapsis.

Slightly advanced: most everyone has his own recipe for how to initiate the turn. The key element is that you stop interfering pretty early, and either let SAS take over in prograde mode, or (if you rocket can fly straight) just let go of all controls.

The higher your TWR, the more you can tip over at first.

Typically, the shallowest ascent you can get away with will consume the least delta-V. It's not that aerodynamic drag is wholly unimportant, but more often than not the dV savings from an aggressive turn will outweigh the drag losses.

The ultimate limit to how shallow you can go is that if you overdo it, your rocket will burn up during launch. The example above is seriously pushing that limit.

 

[Kryxal]

Also, higher TWR means lower delta-v costs to orbit ... but remember that fuel is cheap, engines are expensive.  It can cost less to use a lower TWR, and maybe use boosters for that extra "kick" off the launchpad.

[OhioBob]

I second what @Kryxal just wrote.

My advice is to try to minimize the delta-v that a particular rocket requires to get to orbit, but don't design your rocket solely for the purpose of lowering the delta-v it requires to get to orbit.  The first part of that means you're flying your particular rocket as efficiently as possible, which is good.  The second part means you could be making some ill-advised design choices.  Delta-v is not necessarily the best metric by which to measure how well your launch vehicle performs, so don't obsess over trying to reduce it.

As Kryxal said, trying to get to orbit for really low delta-v values generally means that you require a high TWR, which means big powerful engines.  It is usually best to trade big expensive engines for cheap fuel.  This means you'll have lower TWR and greater gravity losses, but it will cost less.  A rocket that takes 3400 m/s to get to orbit could very well be a better launch vehicle from an economics standpoint than one that takes 3200 m/s to get to orbit.
 

[Flavio hc16]

2 hours ago, Laie said:

Question is, when are you exaggerating? Consider this example of making orbit with much less than 3km/s. It's a expensive booster with a dummy payload, but I think one can learn of it nonetheless:

 

Gravity turn for dummies: tip your rocket over, a little, right after launch, then just follow prograde until you reach apoapsis.

Slightly advanced: most everyone has his own recipe for how to initiate the turn. The key element is that you stop interfering pretty early, and either let SAS take over in prograde mode, or (if you rocket can fly straight) just let go of all controls.

The higher your TWR, the more you can tip over at first.

Typically, the shallowest ascent you can get away with will consume the least delta-V. It's not that aerodynamic drag is wholly unimportant, but more often than not the dV savings from an aggressive turn will outweigh the drag losses.

The ultimate limit to how shallow you can go is that if you overdo it, your rocket will burn up during launch. The example above is seriously pushing that limit.

 

3

there is a problem with what you are saying: dimensions.

Now i explain  myself:0

1) if you start to try your approach  on bigger rocket that can carry more useful stuff, lets say, for long duration mission or to carry a big rover somewhere or an entire base ( especially if you have life support and planetary base, for example) you will see that make a rocket that slim and that tall will start to be uncontrollable, it will usually try to flip or  it will start to bend/broke ( your rocket is 1.25 meter in diameter and 30+ meters high, try to do the same with a 3.75 diameter rocket and i'll love to see how that fly), even with autostruts. And the higher twr on the launchpad will make matters even worst.

2) power scale: you could do that thanks to the fact that the vector is a super power-dense engine, is a 1.25 meters engine with 1000kn of thrust, but if we are not using some absurd engines from mods, there is no other engine that can come close to that thrust/meter^2 of surface: the twin boar has half that power density, the mammoth even a little less than that ( it has 4 engines, but doing an octaweb spacex' style you will fit 9 engines on the bottom of a  3.75 meters part, so the power is 4000kn/9000kn), and then there is the problem that meanwhile your power is following a quadratic scale, your weight is following a cubic scale, so unless you start doing clipping and stuff like that, you can't make a rocket that aerodinamic or that it has that high of twr at the launch AND be controllable.

[CoreI]

8 hours ago, Domax said:

I always have a TWR of 2

One thing that you might be having an issue with is the atmosphere. That's a really high TWR if you're staying at full throttle the whole time. What'll happen is you will be going at 500 m/s at 10 km altitude. At this point the atmospheric drag starts trying to murder the rocket, and you lose a LOT of Delta-V from fighting the atmosphere. What I've found is the best thing to do is throttling during the early parts of the flight so that when you hit 10km your moving at ~350 m/s. This conserves fuel for where the atmosphere is thinner. Once you reach 10km throttle as high as you want.

As for the gravity turn, don't ask me. I've been playing for a few years and still never get it perfect.

[Spricigo]

Orbiting is just falling and misisng the ground. We accomplish it by going sideway so fast that our fall don't keep up with the curvature of Earth Kerbin. We just don't do it at the surface level to avoid drag (collision with the ground is just a severe kind of drag)

That define the first requirement for a good launch to orbit: using most of the energy to build up horizontal momentum (mass*velocity).

The second requirement is to minimise what can reduce our momentum. IOW reduce gravity, drag and cosine loses.

[Laie]

20 hours ago, Flavio hc16 said:

1) if you start to try your approach  on bigger rocket that can carry more useful stuff,

2) power scale: you could do that thanks to the fact that the vector is a super power-dense engine,

You didn't even try, did you?

That the vessel is shaped like a knitting needle is only the icing on the frosting. The key feature is going sideways early and with gusto -- as a rule of thumb, the shallowest ascent you can survive tends to be the most energy-efficient one. That general principle still holds true on more ordinary rockets. Like, ones looking like real-life launchers with fairings being a good deal wider than the booster and so on.

However, as has been pointed out several times already, energy-efficient does not equal cost-efficient; going to such extremes as in the video is something one only does for a challenge. Then again, under the KSP price model, with fuel being rather expensive relative to engines, it often is worthwhile to take off at TWRs of 1.5 or more.

[davidpsummers]

On 2/23/2018 at 3:53 AM, Domax said:

So, when I look at the Delta-V chart, going into LKO must "use" 3400 m/s of Delta-V. But me when I get to orbit I "use" approximately  I use at least 4200 m/s of Delta-V. I know it's quite impossible to do as the Delta-V chart say but I find that I use too much Delta-V for LKO and I think that the problem is from my gravity turn. So do you have some tips to do an efficient gravity turn ?

 

Thank you

 

Fly safe

Note: The number calculated in the VAB is not adjusted for changes in your engines Isp at different pressures.  The default is "vacuum" and your ship almost certainly has less delta-V than listed.  I find that the "vacuum delta-V" is more like 3800 m/s.

If you are using Kerbal Engineer, you can hit "atmosphere".  But even that will give you the sea-level atmospheric pressure, which is the worst case (you rocket will spend most of its time at a lower pressure.  

You can try and play games with look at the delta-V of the different stages at different attitudes and estimate how much time it spends at those altitudes and get a better correction.  That is, IMO, more annoying than fun.