Gravity turn. Do you need to do it?

[hyperformer]

when i start to do my gravity turn is usually when my first pairs of boosters, tends to be 10-13 km

[Grizzlol]

It is possible to make a launch profile straight up more efficient by having an extremely high TWR, but it would really have to be so extreme that it wouldn't be practical in a thick atmosphere like Kerbin or Earth because of the exponentially increasing drag, not to mention surviving the G-forces which would be well beyond lethal magnitude.

Edited June 2, 2013 by Grizzlol

[Liverbugg]

Scott Manley did a video comparing different altitudes of gravity turns.

[PakledHostage]

There was also the Optimal Ascent Profile for This Spacecraft challenge. The rockets we build in the game have changed a bit since then but, as far as I know, the drag model and atmospheric properties are still the same.

Zephram Kerman made a very nice plot showing the relative efficiency of different ascent profiles but the link to the plot seems to be lost. Zephram, if you read this, could you maybe re-post it?

Edit: Maybe we should host a new challenge to experiment with this subject? I don't have time to host it myself, but I'd find time to submit an entry. I saw a post by stochasty in another thread that suggested an efficiency challenge based on the FAR mod.

Edited June 2, 2013 by PakledHostage

[Mediosthenes]

Technically, the conservation of energy equation states that if you want to get to a specific altitude, it uses just as much energy if you go straight up as if you go up at an angle. The big difference is how much force it takes. Going straight up requires maximum thrust for the entire journey, while going at an angle requires less force at any given time. Because fuel used is vastly more important for rockets than total energy used, however, gravity turns are going to be more efficient, because they allow you to switch to smaller engines that burn fuel much more efficiently. Another benefit to gravity turns is that they make gravity assists practical. If you do a gravity turn, you would be able to do a fly-by of the Mun, and steal some of its momentum at no cost to you, saving even more fuel.

[Tokay Gris]

The best route is the most efficient one. That is always your task in KSP regardless of goal.

Ah, but only if the goal is difficult.

I am right now packing a interplanetary cruiser underneath my Dunalander, that packs (without payload) wooping 8500 m/s.

Thats not the wooping part. The wooping part of that cruiser is that is two thirds empty.... 7339 left out of 20160 liquid fuel.

THAT I would not call efficient. If all goes well, after I refueled it, it will return after the mission half full. At least.

And if all does not go well... Well, it will be a nice fireworks...

(A failed mission is a successful firework)

[Martijn404]

The difference is in the 150m/s orbital speed that you already have, sitting on the Launchpad, from Kerbin's axial rotation!!

By burning straight up, you are not using that speed and actually have to nullify it, hence the difference.

By doing a gravity turn, you add to this orbital speed and are ready to escape Kerbin SOI with what you've got in orbit.

As Duna rotates really slowly (thanks to the tidal lock with Ike), the difference there between the modes will be minimal (my prediction, havent tested it).

Ofcourse the timing of when to burn straight up to actually go somewhere is very difficult, so I'd always recommend an Low Kerbin Orbit first.

Edited June 2, 2013 by Martijn404

[rryy]

That's something that North Korea's rocket program would do.

What are you talking about? The Kerean Committee of Space Technology has been getting into stable orbits since, uh, last week!

[DisarmingBaton5]

I had a thought. It is not as efficient on Kerbin, but what about a REALLY low gravity world like Gilly or Bop or Pol?

[Millitron]

The most efficient is neither launching to LKO nor launching straight up. The most efficient method for escape is to launch towards LKO, but don't bother raising your periapsis. Just keep burning till you escape.

[Mercy]

Don't NEED it, but it's better.

[haltux]

http://www2.jpl.nasa.gov/basics/grav/primer.php

Check that out. As you can see, gravity can be used to increase or decrease momentum without "paying the price" as you put it.

Sorry but I am not convinced. Gravity assist by another planet is a completely different story.

As nicely explained in this web page, gravity assist is based on the fact that the planet that assists you is moving. It is just like if you hold a car in rollerblades to accelerate for free. Except that the car is a planet, and the hold is a gravity hold.

Using Kerbin gravity to help you escape from Kerbin does not work that way.

I am not too convinced either by the other posts (no offense). I'll make my own experiments, and I would also be very glad to hear a physically sound explanation.

[Darren9]

Technically, the conservation of energy equation states that if you want to get to a specific altitude, it uses just as much energy if you go straight up as if you go up at an angle. The big difference is how much force it takes. Going straight up requires maximum thrust for the entire journey, while going at an angle requires less force at any given time. Because fuel used is vastly more important for rockets than total energy used, however, gravity turns are going to be more efficient, because they allow you to switch to smaller engines that burn fuel much more efficiently. Another benefit to gravity turns is that they make gravity assists practical. If you do a gravity turn, you would be able to do a fly-by of the Mun, and steal some of its momentum at no cost to you, saving even more fuel.

That was my thought, aside from benefits of staging/engines both ways should be equal, it would be true in a vacuum but what about atmospheric effects. If you do a gravity turn in the atmosphere you cut a longer path through it and loose more to drag - how can you regain that?

[Lazro]

Remember: Escape velocity is relative to how fast you're moving in relation to the planet, so making sure you lose as little d/v as possible to drag and most importantly, gravity, is imperative, especially for interplanetary craft. A gravity turn makes sure that you lose as little fuel as possible by not having to push against gravity. The reason we don't do it immediately upon launch on Kerbin is the atmosphere. The gains from the gravity turn would be negated by drag.

[Sir Nahme]

A single full or half orbit around the planet is more efficient that powering for the entire escape trajectory. You utilize less Dv to accelerate approaching periapsis (you are already accelerating due to gravity pulling you) than it is to drive straight out (you have to provide all the thrust to escape yourself)

The loss to the extra time in high atmosphere is minimal and easily offset by the fact you are USING gravity to give you the lateral velocity

[PirateAE]

Well I've completed my testing, I'll edit my post back a bit with the rest of it but avg results, was 5700 for Eve and 5800 for Duna from orbit burn, so your saving in all cases by the looks 200-300 Dv by using gravity to assist you rather than burning straight out, as the accepted theroy is.

Edited June 2, 2013 by PirateAE
Spelling murder X.X

[haltux]

Technically, the conservation of energy equation states that if you want to get to a specific altitude, it uses just as much energy if you go straight up as if you go up at an angle.

That was my point, thanks to express it in a more formal, clear way.

[...] gravity turns are going to be more efficient, because they allow you to switch to smaller engines that burn fuel much more efficiently.

So it will make a (small) difference in KSP, confirmed by previous experiment, and no difference at all in real life, where engine weight is unsignificant.

[Fuzzy Dunlop]

Technically, the conservation of energy equation states that if you want to get to a specific altitude, it uses just as much energy if you go straight up as if you go up at an angle.

I'm not sure this is the case. I think you might have overlooked the kinetic energy of the planet. If you ascend vertically you will drag the planet after you more than if you gravity turn.

[maltesh]

Stored fuel does not directly translate to energy imparted on your spacecraft. It translates to energy imparted on your spacecraft _and_its_propellant_. If you decide to take a truly terrible path to orbit (such as gently meandering off the pad and climbing to orbit over a period of two to three hours) your remaining spacecraft will have the same balance of specific orbital energy when it gets there, but you will have to spend a hell of a lot more fuel to do so.

So it will make a (small) difference in KSP, confirmed by previous experiment, and no difference at all in real life, where engine weight is unsignificant.

Almost every single spacecraft that has ever launched from Earth to head anywhere does a gravity turn.

Edited June 2, 2013 by maltesh

[Spatzimaus]

Almost every single spacecraft that has ever launched from Earth to head anywhere does a gravity turn.

Yes, but there are practical reasons for that. Going to orbit before heading out lets you double-check systems, abort if there are problems, pick your exact moment of departure, and so on. Even if it weren't more efficient, we'd do things that way in real life on those general principles, although most of those factors don't apply in KSP. It's not only about maximizing your delta-V, it's about maximizing the overall success of the mission. Often the two are related, but not always. In a similar vein, there are occasionally times in KSP where thrusting straight up to escape is a better choice, even if it costs a little more delta-V. Debris is an obvious one; if you go straight up, any discarded stages are guaranteed to crash back down onto the planet, whereas staging after you've reached LKO leaves an empty booster in orbit. This is also relevant to the stage separation itself; if your craft is traveling in a more-or-less straight line, any radially mounted boosters will usually separate cleanly, but if you're pitching around at that moment the chances of collision go way up. (This was especially problematic for my giant U-shaped booster; it's great in that you can use struts to keep awkward payloads stable, but you can't have any serious rotation when things are separating or else it'll run into itself.)

There's also an issue of control; let's say you've got a gigantic booster lifting something large into orbit. Starting a gravity turn at low altitude is no problem, but finishing the turn once the atmosphere gets thin? That'd require RCS, and a lot of it, both to start any rotation and then to cancel it once you hit your final orientation; SAS modules don't have anywhere close to the torque needed to damp out rotations on the largest boosters. Performing the circularization burn gets really difficult in that case, as does reorienting for the escape burn (unless you IMMEDIATELY move to escape, in which case you lose the timing benefits of orbit). When I was lifting 450-ton fuel refineries to Eve, Duna, and Jool, it just wasn't worth the extra work to try to rotate around; a straight-up escape worked great, and the small amount of wasted fuel could be replenished once I arrived at the destination.

What it boils down to, IMO, is whether you have any means to refuel. In stock KSP, currently, you're pretty much limited to whatever fuel you bring along with you, so maximizing your delta-V is critical. People who use the Kethane mod often have a different philosophy; as long as I have enough fuel to not be stranded, an easy, reliable path is preferable to a complex-but-efficient one. Once the new resource system comes in, in 0.21/0.22/whenever, I think you'll see a fundamental shift in stock players. Time is money, basically; a vertical escape just takes a lot less time and attention, which gives you more time doing the parts you actually haven't done hundreds of times already. So once the major downside of that philosophy (possibly getting stranded once you run out of delta-V) goes away, things'll change.

[Entelin]

Going straight up can be more efficient than going into orbit first if your destination is another planet and you are mostly aligned to your ejection angle from the surface. There's no reason to waste dv burning in a direction other than what you need to set your prograde on your ejection angle.

[csiler2]

Great post Spatzimaus. I think you laid everything out quiet nicely.

Fuel-wise, a gravity turn is the most fuel efficient because you are not flying in opposition to gravitational acceleration. Remember, the goal of achieving orbit is to move fast enough that you miss the planet as you 'fall'. Alas, I haven't calculated anything using calc or physics equations in ten years, so I can't give you any solid math supporting or disproving the advantages of direct ascent vs. orbital ascent. Obviously the former requires more fuel, which is why NASA ditched the plan for the space race, but the later requires more time.

I think the major difficulty for direct ascent beyond fuel considerations would be timing. You'd want to makes sure the KSC launchpad is in the proper position before lift-off.

Personally, I will continue LKO rendezvous as I like the margin for error for navigation checks, etc. I'll continue to do so even after resources are implemented because I'd rather launch a smaller rocket into orbit and refuel it there rather than try and lift a lot of fuel to achieve the same task.

Does anyone know what NASA does for interplanetary probes?

[Spatzimaus]

Does anyone know what NASA does for interplanetary probes?

They usually go into a circular orbit before thrusting out-system, but they don't spend any real time there; it's all planned beforehand for timing, so there's no need to wait halfway around a nice circular orbit before reaching your optimal departure point. We don't really have that option in KSP; you launch whenever you're ready, and then fast-forward around your orbit until you hit the maneuver node.

A good diagram for the launch phase of the Voyager missions is at the end of this PDF:

http://planetary.s3.amazonaws.com/voyager_msb/VoyagerMSB02.PDF

showing the boost phase timing. Basically, the Titan rocket (a solid-fuel booster) gets you most of the way to orbit, then the Centaur rocket (liquid fuel) gets you into orbit through two separate burns (i.e., one boost one and one apoapsis circularization), with just enough elongation remaining that the Centaur will come back down after it separates. At that point, the probe slowly thrusts its way outsystem; given the dinky engines those things have, they don't really have the option of fighting against gravity the whole way out.

Notice that the document I linked is Bulletin #2, from 1977. If you want to read all 100 of them going up to 1998, they're available at http://www.planetary.org/explore/resource-library/voyager-mission-status.html

[Flixxbeatz]

Regarding about that matter, here's an old KSP video regarding about achieving escape velocity in Kerbin, courtesy of none other than Scott Manley himself.

[Temstar]

Obviously the former requires more fuel, which is why NASA ditched the plan for the space race, but the later requires more time.

No this is wrong and this misunderstanding comes from a misuse of terminology:

A Direct Ascent moon landing refers to what the spacecraft does on its way back from the moon. As opposed to Lunar Orbit Rendezvous a direct ascent mission will have the Lunar Lander take off from the surface of the moon (via gravity turn by the way, not directly up) and then either enter lunar orbit to wait for the trans-Earth Injection window, or conversely in the later First Lunar Output plan the spacecraft will go from Lunar take off to the end of trans-Earth injection in one continuous burn called "Direct Return".

A Direct Injection in context of a moon landing refers to what the spacecraft does on its way to the moon. With Apollo the spacecraft is placed into a decaying Earth parking orbit first by S-IVB, then the spacecraft is checked out and if everything looks okay the S-IVB is fired up again for trans-lunar injection. If the poop hit the fan during the launch and the spacecraft only limped into this Earth parking orbit then drag will bring the spacecraft back down to Earth well before the week worth of on-board supply runs out. As opposed to this use of a parking orbit, a Direct Injection to the moon involves a single continuous burn (aside from staging events) from the launchpad directly into lunar transfer orbit (still using gravity turn, not straight up). This kind of mission profile is a bit more delta-V efficient since no delta-V is wasted on Earth orbit circularisation. It's also more weight efficient because it means your transfer stage no longer need engine reignition and space loitering capabilities.

China's Chang'e 2 spacecraft used a Direct Injection mission profile.

If you were going to a different planet instead of the moon the idea is the same. A Direct Ejection to a planet involves boosting (via gravity turn) to LEO altitude, but instead of circularising you launch in a way so that just as you reach LEO altitude your phase angle and ejection angle lines up. So you just keep burning prograde until you reach the desired escape velocity.

Basically, weather you're going to LEO or Direct Injection / Direct Ejection somewhere, you always start with launch into a gravity turn. Then once you're at LEO altitude you either circularise to enter LEO or you keep burning for longer for Direct Injection / Direct Ejection. You never burn straight up.

Edited June 3, 2013 by Temstar