How does a rocket in orbit change its attitude without firing an engine?

[MetricKerbalist]

Hi KSP colleagues,

I am orbiting Kerbin with a KerbalX stock rocket.  But I have seen the phenomenon that I am about to describe with other rockets.

I am not using RCS, but I press one of the vector buttons (I think that's the correct term -- you know the buttons to make the rocket move prograde, retrograde, normal, etc.).  Also, I may press the W, A, S, or D keys, which causes the fins to change their orientation.  If you are in a vacuum, however, then the fin direction should not affect the rocket's attitude if you are not firing an engine, isn't that correct?

So I would please like to confirm that the following statement is correct:  In the real world the attitude of an oribiting rocket would not change unless you fired an engine.

Thank you.

Stanley

[DVQuill]

In the real world, rockets use Reaction Control Systems, which include both thrust engines and reaction wheels to orient the craft. While thrust engines obviously use thrust, reaction wheels use the stored momentum from the wheel, or a gyroscope system. KSP mimics these systems in that almost every command part (probe or pod) has a built-in reaction wheel system of some sort, called SAS (Stability Augmentation System, Stability Assist (System), Sickness Avoidance Scheme, or Sickness Avoidance Solution).

In KSP this system allows for control of RCS (of active), engine gimbaling, control surfaces and craft torque. You can also add parts to your craft which increase the amount of SAS control you will get. As your craft gets heavier, the standard SAS that comes with the Mk I Command Pod for example, may not be enough to reliably adjust the attitude of a ship, so adding a reaction wheel (from the Command and Control section in of parts in the VAB) may help quite a bit. 

For example, if you right click on the Mk I Command Pod, you should notice that part of the description says it has a torque of 5, whereas the Mk2 Lander Can has a torque of 15, meaning the Mk2 Lander Can will provide more SAS control overall. The reaction wheels I mentioned earlier also come with a torque value that can be found in the VAB parts list. Stock KSP currently has 3 different variants of the reaction wheels, ranging in torque value from 5 to 15 to 30. 

 

[OhioBob]

22 hours ago, MetricKerbalist said:

So I would please like to confirm that the following statement is correct:  In the real world the attitude of an oribiting rocket would not change unless you fired an engine.

(EDIT:  Note that when answering below I misread "attitude" as "altitude".  Sorry about that.  Hopefully some of it is still applicable to your question.)

To change a rocket's altitude, its velocity must change, which means there must be a force applied to the rocket to accelerate it.  This force is typically applied by firing an engine.  In KSP this is the only way to change a rocket's velocity once leaving the atmosphere.  However, in the real world the atmosphere doesn't just suddenly end like it does in KSP.  So even rockets in orbit around the earth are still subjected to a small amount of atmospheric drag.  This drag produces a force that causes orbits to decay.  So in the real world altitude can change with firing an engine.

Fins should have no affect on a rocket outside of the atmosphere because there is no air to produce aerodynamic forces.  Reaction wheels should also have no affect on a rocket's velocity.  Reaction wheels produce torque only, which causes a rocket to rotate about its center of gravity.  This changes a rocket's attitude but not its velocity as there is no lateral force applied.  RCS thrusters, however, can be used for both attitude control and translation, depending on their placement and how they are fired.  If used for translation, the rocket's velocity vector will change, and hence the size and shape of its orbit.

Edited June 28, 2021 by OhioBob

[modus]

48 minutes ago, MetricKerbalist said:

So I would please like to confirm that the following statement is correct:  In the real world the attitude of an oribiting rocket would not change unless you fired an engine.

In the world of KSP, that is correct. In the real world, there is a phenomenon called orbital decay, causing  a ship to 'loose' attitude, caused by the atmospheric drag of earth. (I am definitely not an expert on the subject)

Edit: yeah what OhioBob said

Edited June 27, 2021 by modus
NinjaBob

[OhioBob]

Just to clarify something that @DVQuill talked about...  SAS can be thought of as the computer system that controls attitude, but it does not actually produce any torque whatsoever.  To produce the torque (the turning force), reaction wheels, RCS thrusters, gimballed engines, or aerodynamic control surfaces (e.g. fins) must be provided.  In other words, to control attitude you must have both SAS and some torque producing mechanism.  SAS is a completely separate thing from that which produces the torque.

While it is true that most probe cores in KSP have both SAS and reaction wheels, don't conflate these systems.  There are some probe cores that have SAS only with no reaction wheels (Probodobodyne QBE and OCTO2 for example).  So while these probe cores have the brain to control attitude, they are unable to do so by themselves because that lack the machinery to produce torque.  They can, however, control a rocket that has, say, gimballed engines or moveable fins.

On the other hand, if you have torque producing parts but no SAS, there is no way to control those torque producing parts.  The Stayputnik probe core, for example, has no SAS.  Therefore a rocket equipped with a Stayputnik cannot be steered even if equipped with gimaballed engines or fins.  There is simply no way to command those mechanisms to move without SAS to send the commands.

One final note, a Kerbal pilot is equivalent to SAS and can perform the same functions.  SAS can be thought of as an automated or remotely control system.  If a pilot is on board and inside a command pod, he/she can perform those tasks without the need for SAS.
 

Edited June 27, 2021 by OhioBob

[18Watt]

1 hour ago, MetricKerbalist said:

So I would please like to confirm that the following statement is correct:  In the real world the attitude of an oribiting rocket would not change unless you fired an engine.

As others have mentioned, that is not necessarily true.  You need to generate a force to make an object move or rotate.   One way would be firing engines or maneuvering thrusters.

Another way is to use gyros, or reaction wheels.   These are used in the real world as well as KSP.   However, the reaction wheels in KSP are much more powerful than what you would find used in the real world.  This makes the game more playable, especially for new players.   In the real world heavy vessels would rely more on small thrusters to adjust and maintain attitude.   In KSP the reaction wheels are powerful enough that you can use them instead of thrusters.

[Rhomphaia]

18 minutes ago, OhioBob said:

While it is true that most probe cores in KSP have both SAS and reaction wheels, don't conflate these systems.  There are some probe cores that have SAS only with no reaction wheels (Probodobodyne QBE and OCTO2 for example).  So while these probe cores have the brain to control attitude, they are unable to do so by themselves because that lack the machinery to produce torque.  They can, however, control a rocket that has, say, gimballed engines or moveable fins.

On the other hand, if you have torque producing parts but no SAS, there is no way to control those torque producing parts.  The Stayputnik probe core, for example, has no SAS.  Therefore a rocket equipped with a Stayputnik cannot be steered even if equipped with gimaballed engines or fins.  There is simply no way to command those mechanisms to move without SAS to send the commands.
 

SAS is not required to control, it only gives stability assist, so if you have steerable fins, RCS, gimballed engines, or reaction wheels, then stayputnik can control those mechanisms as long as it has connection to KSC, but only manual inputs, so in space for every rotation you induce you will need an equal and opposite input to stop the rotation

Edited June 27, 2021 by Rhomphaia

[OhioBob]

22 minutes ago, Rhomphaia said:

SAS is not required to control, it only gives stability assist, so if you have steerable fins, RCS, gimballed engines, or reaction wheels, then stayputnik can control those mechanisms as long as it has connection to KSC, but only manual inputs, so in space for every rotation you induce you will need an equal and opposite input to stop the rotation

Perhaps that's true, it's been a very long time since I used a Stayputnik.  If so, I stand corrected.

Edited June 27, 2021 by OhioBob

[jimmymcgoochie]

Attitude control is done two ways- using thrusters as part of an RCS system, which uses a finite store of propellant, or using torque generated by spinning gyroscopes, which use electricity and which in KSP are referred to as reaction wheels. The reaction wheels in KSP are monstrously powerful compared to their real-world equivalents, allowing probes, rovers, landers and even space stations to rotate on all three axes without any propellant being used.

Most probe cores (except the Stayputnik, OKTO2 and RoveMate) and command pods (except the Making History KV pods and the MEM) have built-in reaction wheels for attitude control; for small probes and crewed vessels these can be enough, but you'll probably need to add some dedicated reaction wheels for your launch rockets and for larger craft unless you want to take several minutes to turn around. Reaction wheels use electricity- quite a lot of electricity- so keep an eye on your power gauge especially if you don't have a source of power generation or you could drain your batteries and end up stuck. In KSP, use reaction wheels to control your attitude (rotation in the xyz axes) and use RCS only for translation (movement in the xyz axes) when you want to make precise changes to your trajectory, such as when docking.

In real life reaction wheels are much less powerful, and are typically only used where very small/precise adjustments are needed (like for space telescopes) and/or to save propellant (e.g. long-distance interplanetary probes); there's also the International Space Station, which uses four large gyroscopic wheels to maintain its attitude or to rotate in order for a visiting spacecraft to dock. If you want to get into more detail, try these videos by Scott Manley:

https://www.youtube.com/watch?v=7Js5x4NhUxU

https://www.youtube.com/watch?v=fDp8jbP_22c

 

[MetricKerbalist]

47 minutes ago, jimmymcgoochie said:

In KSP, use reaction wheels to control your attitude (rotation in the xyz axes) and use RCS only for translation (movement in the xyz axes) when you want to make precise changes to your trajectory, such as when docking.

First, I would like to thank everyone for their answers.  I will ponder everything carefully,

Now, regarding the comment that @jimmymcgoochiemade which I am quoting.  Here I get confused between using the W, A, S, and D keys and the H, N, I, K, J, and L keys.  Are both of those sets RCS controls?

So actually, if you or someone else, could kindly expand on that remark, I would appreciate it.

Thank you.

Stanley

[jimmymcgoochie]

1 minute ago, MetricKerbalist said:

First, I would like to thank everyone for their answers.  I will ponder everything carefully,

Now, regarding the comment that @jimmymcgoochiemade which I am quoting.  Here I get confused between using the W, A, S, and D keys and the H, N, I, K, J, and L keys.  Are both of those sets RCS controls?

So actually, if you or someone else, could kindly expand on that remark, I would appreciate it.

Thank you.

Stanley

WASD(QE) are the rotation controls, they work for RCS and reaction wheels; HNIKJL are the translation controls and only work for RCS. You can control your attitude- rotate in the pitch, yaw and roll axes- without RCS, but you need RCS to translate- move forward/back, left/right and up/down laterally- as reaction wheels are only for rotating.

You can change the settings on your RCS thrusters so they only fire for specific control inputs, so if you wanted them to only work for translating and not for rotating just disable the pitch, yaw and roll toggles. You'll find that option if you right click the RCS thruster(s) and then "show actuation toggles".

Maybe I'm going into too much detail- just stick with reaction wheels unless you absolutely need to dock with something.

[fourfa]

Another option for changing spacecraft attitude without reaction engines: magnetorquers

Basically, the spacecraft can use a magnetic field generator to create a small torque relative to the planet's magnetic field.  Obviously this depends on being IN a planet's magnetic field - so won't work in deep space or around many  of the Solar System's celestial bodies that have no field to speak of.

[MetricKerbalist]

Hi KSP colleagues,

I thank all of you for your thought-provoking, informative comments.

I think that I am catching on.  I believe that the gist of what you are saying is that the reaction wheels could control the rocket's attitude even when the rocket is in orbit.

Is this a reasonable synopsis of the discussion?

Stanley

[Reactordrone]

It's the practical application of this effect,

 

[MetricKerbalist]

Thank you, @Reactordrone.  That video was highly instructive.

[MetricKerbalist]

Hi everyone,

As I say, that video was very instructive and I watched another one in the links that also taught me a lot.  For the first time I understand the meaning of the term moment of inertia -- it is the amount of mass and where the mass is spread out.

So, please, I am going to follow up with these questions:

• True, you don't have to do an engine burn to change the spacecraft's attitude while it is in orbit.  I get that now.  But you do have to change the orientation of the flywheel (if that is the correct word to use), and changing the orientation of the flywheel requires some kind of fuel, maybe using battery power.  Is that correct?
• When I see the engine throttle moving around on its gimbal, and when I see the fins changing their orientation, none of that causes the spacecraft to change its attitude in the vacuum of orbit.  Correct?  The flywheel's changing its orientation -- that is what causes the spacecraft to change its attitude.  Right?  But the flywheel is mechanically connected to the engine gimbal and to the fins.  If as an experiment you were to cut that mechanical connection, you could still change the spacecraft's attitude by changing the flywheel's orientation.  And now you wouldn't see the engine moving on its gimbal and you wouldn't see the fins changing their direction.  Is all of this correct?

Thank you for listening to my questions, and thank you for your instructive answers up to now.

Stanley

[Rhomphaia]

40 minutes ago, MetricKerbalist said:

Hi everyone,

As I say, that video was very instructive and I watched another one in the links that also taught me a lot.  For the first time I understand the meaning of the term moment of inertia -- it is the amount of mass and where the mass is spread out.

So, please, I am going to follow up with these questions:

• True, you don't have to do an engine burn to change the spacecraft's attitude while it is in orbit.  I get that now.  But you do have to change the orientation of the flywheel (if that is the correct word to use), and changing the orientation of the flywheel requires some kind of fuel, maybe using battery power.  Is that correct?
• When I see the engine throttle moving around on its gimbal, and when I see the fins changing their orientation, none of that causes the spacecraft to change its attitude in the vacuum of orbit.  Correct?  The flywheel's changing its orientation -- that is what causes the spacecraft to change its attitude.  Right?  But the flywheel is mechanically connected to the engine gimbal and to the fins.  If as an experiment you were to cut that mechanical connection, you could still change the spacecraft's attitude by changing the flywheel's orientation.  And now you wouldn't see the engine moving on its gimbal and you wouldn't see the fins changing their direction.  Is all of this correct?

Thank you for listening to my questions, and thank you for your instructive answers up to now.

Stanley

In the real world rather than changing the orientation of the reaction wheel, you use three small wheels, mounted at 90 degrees to each other, (or 4 wheels in a tetrahedral arrangement if you want a spare) by changing the speeds of these wheels you can induce rotation in the space craft in any axis you want, yes you need to power the wheels somehow. Larger spacecraft like the ISS use the system where the orientation of a larger flywheel is changed, these are called Control Moment Gyroscopes

the actuators on the fins and engines have nothing to do with the reaction wheels, they are separate systems  the only mechanical connection is that they are attached to the same spacecraft.  It is just that in KSP when you input a command then the spacecraft will use every method at its disposal to achieve that, even if those methods are ineffective such as gimballing an engine that isn't firing or angling fins in a vacuum

[MetricKerbalist]

Hi @Rhomphaia,

Thank you.  You answered my questions thoroughly and clearly.

I am so glad that I raised my original question.  I learned a tremendous amount about physics form the instructive answers that you and the others have given me.

KSP is so educational -- especially when I have knowledgeable and helpful KSP colleagues like all of you to answer my questions.

Stanley

[MetricKerbalist]

On 6/27/2021 at 1:47 PM, jimmymcgoochie said:

WASD(QE) are the rotation controls, they work for RCS and reaction wheels; HNIKJL are the translation controls and only work for RCS. You can control your attitude- rotate in the pitch, yaw and roll axes- without RCS, but you need RCS to translate- move forward/back, left/right and up/down laterally- as reaction wheels are only for rotating.

The translation issue continues to confuse me.

In the real world a reaction world cannot cause translation, can it?  In the real world you need to thrust some kind of engine to get translation, right?

In KSP, where I have not done much work with rovers, I recall that you can translate the vehicle by the HNIIKJL keys.  But isn't that completely fictional?

Please, anyone jump in.

[Rhomphaia]

16 minutes ago, MetricKerbalist said:

The translation issue continues to confuse me.

In the real world a reaction world cannot cause translation, can it?  In the real world you need to thrust some kind of engine to get translation, right?

In KSP, where I have not done much work with rovers, I recall that you can translate the vehicle by the HNIIKJL keys.  But isn't that completely fictional?

Please, anyone jump in.

Indeed reaction wheels cannot translate the vehicle, neither in the real world or KSP.

Reaction Control Systems  would normally be an umbrella term for all systems, but in KSP,  RCS specifically refers to reaction control thrusters. small thrusters pointed in various directions to control attitude or translation

The rover I remember you asking about before from the Mun Rover scenario has several of these, some powerful ones underneath to lift the rover up, and 8 sets of 4-way thrusters around the outside to allow lateral translation.

[MetricKerbalist]

But as I recall, you could get the Mun Rover to translate without using thrust.  You could just use the HNIKJL keys.  How is that possible?

[Rhomphaia]

6 minutes ago, MetricKerbalist said:

But as I recall, you could get the Mun Rover to translate without using thrust.  You could just use the HNIKJL keys.  How is that possible?

there is still thrust, you just don't need the throttle, if you have RCS thrusters then pressing for example K will cause all the RCS thrusters pointing in the appropriate direction to fire, in this example lifting the craft up

[MetricKerbalist]

Hi @Rhomphaia,

Good.  That's clear.

Stanley