How to perform the real gravity turn?

[Kitspace]

Hello!

I would like to ask a quite complex question here.

I am always using the FAR. So, it finally made me come across the idea that actually performing a real gravity assisted ascent the very same way it is done for more then fifty years in the real world would be the best way of getting my kerbals into orbit.

How to calculate what should my rocket look like, how should it be balanced, what performance should it have to take me where I need it to be after the ascent? Is it possible to get into a circular orbit just with a gravity turn without the circularization? I guess many real rockets do the job this way...

How should I fly the thing after it leaves the dense atmosphere and does not go prgrade automatically any more?

Why are my thrust to weight ratios and g forces much lower than I would expect in real world on a realistic looking launch? Is it all about the size of the planet?

Why do my rockets always tend to have their center of lift before their center of gravity even if they generally look like working rockets? To get around this without adding unrealistically looking fins that also add way too much stability and wobbling I would have to add payloads that the rockets would not lift of the ground at all.

What do they do with this in the real world? Where do real thin and tall rockets like Redstone or Titan have their center of gravity on the launch pad? I guess not very high even regarding that kerbal parts are heavier than real ones.

Can I automate the flying operations somehow in a manner similar to the real rockets?

So, out of all this post the most important general question is how to predict and calculate a naturally going gravity turn and make it end up where I need it too.

Thank you for all your help!

[Hodo]

Lots of questions.

Some of them I can answer, others I can not in good faith give you a decent answer to.

"Can I automate the flying operations somehow in a manner similar to the real rockets?"

Yes, you have two options that work quite well for this. There is Mechjeb which does have an autopilot feature which can turn it into a very automated process but it is a bit clunky at times.

Then there is KOS, which is a fair bit more complex to use but can give you a bit more accurate feel to the automation process.

"Why are my thrust to weight rations and g forces much lower than I would expect in real world on a realistic looking launch?"

Quite simply, it is a game. Kerbin is roughly 1/3rd the size of the Real Earth atmosphere, so rocket launches in KSP are closer to bottle rocket launches. The ideal thrust to weight ratio for a launch is a round 1.2-1.4:1.

As far as performing a gravity turn in FAR with stock sized Kerbin, generally I start my gravity turn at 100m/s or 1km altitude, which ever happens first. I tilt the rocket 5deg in the direction I want it to go, and gravity will gradually pull it over till you hit the higher atmosphere. You may have to adjust the crafts AoA every so often but nothing to major.

Edited June 8, 2014 by Hodo

[Kitspace]

And what about the other questions?

I am NOT talking about an actively controlled ascent path looking like a gravity turn here, which is a popular kerbal thing and are called gravity turns for some reason, I am talking about a real gravity assisted turn with no pitch control inputs like it is done in reality.

My main problem is predicting where would the gravity turn take me with my current rocket and my current pitchover speed and altitude.

Thank you

[Hodo]

And what about the other questions?

I am NOT talking about an actively controlled ascent path looking like a gravity turn here, which is a popular kerbal thing and are called gravity turns for some reason, I am talking about a real gravity assisted turn with no pitch control inputs like it is done in reality.

My main problem is predicting where would the gravity turn take me with my current rocket and my current pitchover speed and altitude.

Thank you

Ok then my last paragraph pretty much answers as best I can how to do a gravity turn, a real gravity turn in KSP. My best rocket for RO could do a gravity turn but I quite litterally copied the Areas X rocket. The trick is really in the TWR and balance of the craft. If the TWR is to high, it will not turn properly it will just spin and blow apart. To low and you have made an intercontinental missile.

[BudgetHedgehog]

Once you reach 80m/s, turn off SAS and pitch over 5 degrees in the orientation you want. Then you wait until your AP is where you want it. Well, that's the instructions, actually doing it is very difficult. Pitch control can be done with engine throttle, but unless your rocket is designed right, it'll be almost impossible. I've managed it once, but usually, I do need to make small adjustments.

[Kitspace]

That is the actual trick.

How to make your rocket right?

How to predict and calculate this gravity turn?

I want this turn not only to take me somewhere but to take me where I need to get.

I understand that this requires maths.

And I am asking where and how should I learn something about it or where can I get a tool that will do it if such thing exists.

That is the only way of launching that seems to be working good for me as most others I have tried either required multiple engine ignitions and maneuvering which is very unrealistic while getting into an simple ordinary low planetary orbit or just pushed the aerodynamic forces so high that they made my rocket fall apart.

[mhoram]

How to make your rocket right?

This writeup might answer your question.

http://www.projectrho.com/public_html/rocket/multistage.php

How to predict and calculate this gravity turn?

I want this turn not only to take me somewhere but to take me where I need to get.

I understand that this requires maths.

And I am asking where and how should I learn something about it or where can I get a tool that will do it if such thing exists.

Calculating the gravity turn could be done by simulating ascents based on real-life gravity turns.

The real-life gravity turn consists of the following stages:

1. Vertical ascent

2. Pitchover maneuver (Orientation of the ship is no longer vertical)

3. Downrange acceleration (Orientation of the ship is equal to prograde, Prograde changes because of the gravity force applied to the ship)

4. Circularization (Circularization when the orbit altitude is reached)

This gravity turn can be described (in a simplified form) by the following parameters:

- Altitude where the vertical ascent stops

- Angle of the orientation during the pitchover maneuver

- Duration of the pitchover maneuver

- Altitude of Circularization

(other parameterizations are possible as you like, for example duration of vertical ascent)

(more details could be: turnrate of the ship during pitchover, how long before reaching the orbit the circularization burn begins, ...)

With this parameterization an ascent can be simulated based on the three forces that have an impact on the rocket:

- gravity

- air resistance

- thrust

(Don't forget that planet and atmosphere rotate)

Calculating them for small timesteps the course of the ship can be approximated quite good.

And at the end of the simulation you know if the chosen parameters were good or bad.

Based on the quality of the parameters, in an iteriative procedure better parameters can be found until you get parameters that satisfy your needs.

In this post you can find a link to a program I have written that does an optimiziation of such a gravity turn with respect to fuel usage for stock KSP. I know that this is not exactly what you asked for, but it might give you some ideas.

Air resistance in FAR is quite more complex than in stock KSP, so for designing such a simulator, you will need a way to calculate it.

Edited June 8, 2014 by mhoram

[Pecan]

Sunday finished 10 minutes ago. I reluctantly drag myself away from ecstatic worship at the Church of mhoram (<- Wot? No Capital! Must be humility)

ETA: LOL - apparently I haven't given enough other people reputation to be able to give you any again yet.

[Thran]

real gravity turns are done more-or-less unguided, you pitch over and then let go of the controls (unless you're dogging into a equatorial orbit)

[Kitspace]

Thank you!

[MarvinKitFox]

Hello!

Is it possible to get into a circular orbit just with a gravity turn without the circularization? I guess many real rockets do the job this way...

Actually, none do.

All rockets heading for orbit exercise positive attitude control.

They may fly a path which seems close to a natural gravity turn, but in reality is not one at all.

[Kitspace]

Sorry, what do you mean here?

Of course rockets use positive attitude control, but while in atmosphere they just do not require any control inputs other than small corrections because gravity is doing the job. Rockets start to use their gimbal to follow the path when out of the atmosphere and their passive stability does not help any more.

At least that is what I have always thought of it.

And my original question you have quoted was about at what attitude do I have to burn once outside of the atmosphere to get into an orbit not a ballistic path where you would have to wait for apoapsis and burn again.

Most real rockets are already in a stable parking orbit, maybe not perfectly circular yet, but already safe from entering the atmosphere, while they first cut their final stage engines off. Also, most rockets do not even have capabilities of repeated ignition of their engines, their payloads have to use tug modules on hypergolic propellants to maneuver.

I really wonder how do they do it.

[Hodo]

Sorry, what do you mean here?

Of course rockets use positive attitude control, but while in atmosphere they just do not require any control inputs other than small corrections because gravity is doing the job. Rockets start to use their gimbal to follow the path when out of the atmosphere and their passive stability does not help any more.

At least that is what I have always thought of it.

And my original question you have quoted was about at what attitude do I have to burn once outside of the atmosphere to get into an orbit not a ballistic path where you would have to wait for apoapsis and burn again.

Most real rockets are already in a stable parking orbit, maybe not perfectly circular yet, but already safe from entering the atmosphere, while they first cut their final stage engines off. Also, most rockets do not even have capabilities of repeated ignition of their engines, their payloads have to use tug modules on hypergolic propellants to maneuver.

I really wonder how do they do it.

If you perform your burn correctly you actually can make a pretty decent orbit with just the gravity turn. But the trick is having the correct d/v required to do this at the desired altitude adjusting for atmosphere losses.

[Kitspace]

So, what are the right parameters, how do I do it?

How to use this tool

http://forum.kerbalspaceprogram.com/threads/58531-Finding-the-best-ascent-path-for-rockets?p=785108&viewfull=1#post785108?

I do not quite get what to do with these files.

[Hodo]

So, what are the right parameters, how do I do it?

How to use this tool

http://forum.kerbalspaceprogram.com/threads/58531-Finding-the-best-ascent-path-for-rockets?p=785108&viewfull=1#post785108?

I do not quite get what to do with these files.

Never used that plugin so don't know. But the parameters very based on the size of the rocket, the TWR of the rocket, and where the CoM is on the rocket.

If you have procedural tanks, then I suggest building an Areas I-X rocket, as it is the simplest rocket to test this stuff with.

[GoSlash27]

I think the main question is "how to create a rocket that will do a true gravity assist turn in KSP".

Basically firing up a rocket, tipping it off vertical, and (look ma no hands); a perfect orbit. Although IRL rockets still need minor corrections...

The answer is simulation, which in KSP means trial and error. Especially since #1 the aerodynamic model in KSP is hosed and #2 you can't get the precise dimensions and centers of gravity, which you need for the maths.

build a rocket that will generate the proper delta v for the payload and the mission. I can tell you how to do that if you're interested.

Then simply throttle it up and let it rip with the SAS off. Hopefully if your torque moment is long enough it'll fly straight up.

At 7Km, kick it prograde a few degrees and see what happens. If it falls over too fast, then you need to either lengthen the torque moment, increase the thrust, or adjust the altitude and/or angle of your initial pitchover. do the opposite for a rocket that doesn't tip fast enough.

Each stage must be balanced to naturally tip at just the right rate, so it's liable to take you a long time to get it just so.

AFA what a gravity turn will end up looking like, there are a few goalposts to shoot for:

-pitchover at 7Km

-22.5* off vertical at 15Km

-45*at 25 Km

maintain that rate of pitch down until you achieve the desired apoapsis. then...

- pitch= -|prograde elevation| when apoapsis = desired apoapsis

from there, the orbit should circularize itself by simply accelerating at a rate sufficient to keep the apoapsis 30 seconds ahead.

This is how I run all my launch vehicles and it's very economical for dV, but I cheat by using SAS and guiding it manually. SAS is so cheap from a mass ratio standpoint that I don't see a reason to build a rocket that doesn't use it.

Good luck!,

-Slashy

[mhoram]

How to calculate what should my rocket look like, how should it be balanced, what performance should it have to take me where I need it to be after the ascent?

I just remembered a configuration guide for rocket clusters by Temsrar for asparagus staged rockets.

He explains how to calculate the thrust needed for different stages to get into orbit.

There is also a calculator available based on this description.

So, what are the right parameters, how do I do it?

There is no single right combination of ascent parameters that fit all rockets.

Each rocket has it's own ideal ascent path.

How to use this tool

http://forum.kerbalspaceprogram.com/threads/58531-Finding-the-best-ascent-path-for-rockets?p=785108&viewfull=1#post785108?

I do not quite get what to do with these files.

These are Perl scripts. You will need a Perl interpreter to run that program. If you manage to get "liftoff.pl" to run and need an introduction guide, feel free to contact me.

I think the main question is "how to create a rocket that will do a true gravity assist turn in KSP".

Basically firing up a rocket, tipping it off vertical, and (look ma no hands); a perfect orbit. Although IRL rockets still need minor corrections...

The answer is simulation, which in KSP means trial and error.

Trial and error seems the way to go. If you find a way to automatize the building procedure of rockets to get perfect gravity turns, I would be interested.

[Kitspace]

Sorry, did not quite get what is meant under lengthening the torque moment and what is the prograde elevation?

[mhoram]

what is the prograde elevation?

Most KSP-relevant terms (including prograde) are explained on http://wiki.kerbalspaceprogram.com/wiki/Terminology.

Otherwise wikipedia or an internet-search can be used to find answers for these kind of questions.

[Kitspace]

I certainly know what is prograde. The actual question was what is the prograde elevation.

[cantab]

The one time I've done a "real" gravity turn, it was with about the simplest possible rocket, just a long SRB and the payload to circularise. I started it canted over 5 degrees on the pad and let it fly. Probe core torque, very weak on such a big thing, let me influence apoapsis but not "visibly" control the rocket.

Though it's slightly less efficient, tilting it on the pad will have the advantage of reproducability - it'll be the same angle first time, every time. That will I expect help you fine tune the other aspects of your rocket.

[T.C.]

- pitch= -|prograde elevation| when apoapsis = desired apoapsis

what is the prograde elevation?

I am also unfamiliar with "prograde elevation". From the context, I think Slashy intended the term to mean "the angle between prograde and the vertical axis"; in other words, he was saying "head toward prograde when you achieve the desired apoapsis".

With respect to Slashy, I think his post was a little misleading. In a perfect gravity turn, I believe: 1) you would (inherently) head toward prograde during the entire ascent, except during a brief early pitchover maneuver. (Unless you launch at an angle, as cantab did, in which case there is no pitchover maneuver.); and 2) you would not achieve the desired apoapsis until the very end of the gravity turn, at the same moment you arrive at apoapsis, complete your orbital insertion, and end your burn.

-TC

Edited June 11, 2014 by T.C.

[Jananton]

Since I finally decided to experience some mods and amongst others installed FAR I found that using something, that I believe could be considered as approaching a 'gravity turn', is the most secure way to get up, up and away. In a way it's much like T.C. described above. After launch and reaching 75 m/s switch off SAS and give a small nudge in the direction you want to proceed. In the mean time depending on the TWR throttle back so your speed just increases slightly from 145 m/s at approx 3000m. Now use the throttle to control the speed at which the rocket tips over, lower- if it doesn't drop enough, more throttle if the tipping over goes too fast. I still haven't managed it to perfection, but it works much better for sure then doing it the 'old' way before I had FAR installed. With FAR the 'old' way, starting the turn at 7000 to 10000m to 45 degrees, most of the time results in tipping the rocket over uncontrollably and ending the whole opration in tears. ;-)

Greets,

Jan

[GoSlash27]

Sorry, did not quite get what is meant under lengthening the torque moment and what is the prograde elevation?

For torque moment, I'm talking about the weight distribution around the center of mass. Like an ice skater with her arms spread out vs. pulled in close. tall rockets rotate slower than short rockets, so you need to balance that to make the rocket tip over at a proper rate.

For prograde elevation, I'm talking about the elevation above( or below) the horizon where "forward" is. Or in simpler terms where you're going. If your apoapsis is already correct, then you can maintain it there by pointing your nose at the opposite side of the horizon from where you're headed the same number of degrees.

This allows you to continue to accelerate to circularize your orbit without cutting/ restarting the engines.

Best,

-Slashy

[GoSlash27]

I am also unfamiliar with "prograde elevation". From the context, I think Slashy intended the term to mean "the angle between prograde and the vertical axis"; in other words, he was saying "head toward prograde when you achieve the desired apoapsis".

With respect to Slashy, I think his post was a little misleading. In a perfect gravity turn, I believe: 1) you would (inherently) head toward prograde during the entire ascent, except during a brief early pitchover maneuver. (Unless you launch at an angle, as cantab did, in which case there is no pitchover maneuver.); and 2) you would not achieve the desired apoapsis until the very end of the gravity turn, at the same moment you arrive at apoapsis, complete your orbital insertion, and end your burn.

-TC

Not so as I understand it. It would be impossible for a rocket to establish a circular orbit while only heading toward prograde. And with no air resistance, there's no reason to. As the rocket approaches the apoapsis, the nose has already fallen below the horizon. It's the increase in velocity that makes the rate of pitch *seem* to be zero; the horizon is falling away from you at the same rate that your nose is rotating.

I'm not sure I can explain this without diagrams...