No S3 Kerbodyne sized reaction wheels / stabilizers?

[TrooperCooper]

Are there really no controlling enhancers for the largest hull size? I was hoping to unlock them in my career but it seems like there are none. Do I really have to control vessels consisting of the huge components by using medium sized wheels that have the radius of the Rockomax parts?

I understand that bigger wheels dont increase their controll bonus, but stacking Rocko controllers into Kerbodyne hulls causes many difficulties. It looks ugly if one squeezes them inbetween there directly. And I could imagine that the narrowing of the stack at this spot would cause stability issues. I could use the Kerbadyne=>Rockomax adapter at the top and at the bottom of the Rocko sized wheels, but it would still cause a narrow spot in the stack. Furthermore, it would stretch the size of the ship a lot, especially for larger vessels with the requirement of multiple controllers, further adding to the stability problem.

Building quite large ships based on Kerbodyne hulls, I am running into serious problems. Its perfectly reasonable to expect less maneuverability etc. with larger ships. But beeing pretty much forced to put all reaction controlers at the top end of the rockets, faaaar away from the center of mass, is kind of strange. I am new to the game and maybe I am overlooking something. I am using RCS already on my ships as well as including the vectoring help from the main engines. But the controll wheel placements / the lack of a proper large controller really baffles me. So any input would be welcome.

Is there maybe a mod for such an item?

Edited May 20, 2014 by TrooperCooper

[Commander Jebidiah]

KW rocketry has 3.75m SAS modules also your reaction wheels don't need to be at the CoM

[pincushionman]

Given the size and mass of the largest parts, and the spacecraft made out of them, it may be more advantageous to rely on RCS for attitude control rather than gigantic reaction wheels.

[Northstar1989]

Are there really no controlling enhancers for the largest hull size? I was hoping to unlock them in my career but it seems like there are none. Do I really have to control vessels consisting of the huge components by using medium sized wheels that have the radius of the Rockomax parts?

You could use winglets near the bottom of the rocket as stabilizing fins... It works great in the atmosphere, which is I assume where you're using the Kerbodyne-sized parts. You can dump them with the launch stage- you don't need nearly as much control authority once you're already in orbit...

I understand that bigger wheels dont increase their controll bonus, but stacking Rocko controllers into Kerbodyne hulls causes many difficulties. It looks ugly if one squeezes them inbetween there directly. And I could imagine that the narrowing of the stack at this spot would cause stability issues. I could use the Kerbadyne=>Rockomax adapter at the top and at the bottom of the Rocko sized wheels, but it would still cause a narrow spot in the stack. Furthermore, it would stretch the size of the ship a lot, especially for larger vessels with the requirement of multiple controllers, further adding to the stability problem.

You are correct, adding 2.5 meter parts to the middle of 3.75 meter (Kerbodyne-sized) stacks is generally a bad idea...

Building quite large ships based on Kerbodyne hulls, I am running into serious problems. Its perfectly reasonable to expect less maneuverability etc. with larger ships.

Not necessarily. It's perfectly possible to build highly maneuverable 3.75 meter stacks- although it will come at quite a mass penalty in aerodynamic control surfaces and reaction wheels... (I wouldn't suggest doing it)

But beeing pretty much forced to put all reaction controlers at the top end of the rockets, faaaar away from the center of mass, is kind of strange.

That's actually *EXACTLY* where you should be putting all your reaction wheels. The further they are from the Center of Mass, the *more* effective they are (assuming you're not having problems over-steering). Think of your rocket as a giant lever...

I am new to the game and maybe I am overlooking something. I am using RCS already on my ships as well as including the vectoring help from the main engines. But the controll wheel placements / the lack of a proper large controller really baffles me. So any input would be welcome.

Based on what you said earlier about "putting all your reaction wheels far from the Center of Mass seeming strange to you", I'm guessing you might be missing quite a few things actually. I *HIGHLY* recommend you take a look at the rocket-building tutorials on the KSP Wiki, particularly the one on "Intermediate Rocket Design" (it sounds like you're beyond "Basic Rocket Design" being useful to you, and the "Advanced Rocket Design" tutorial isn't about new concepts- only the underlying math...)

http://wiki.kerbalspaceprogram.com/wiki/Tutorial:Intermediate_Rocket_Design

Is there maybe a mod for such an item?

If you really want 3.75 meter reaction wheels, you can install "NovaPunch2" or "KW Rocketry" mod- both of which have been using 3.75 meter stack sizes LONNNNGGGGG before they were introduced to the stock game. In fact, I'd highly recommend using one of these mods (I play with NovaPunch2 myself...)

Regards,

Northstar

EDIT: Actually, I was thinking of RCS thrusters, when I said SAS works best when located far from the Center of Mass. It doesn't matter where you place SAS- though RCS still works best when placed far from the CoM. The top of your rocket is probably still the best place for your reaction wheels though- that way you keep them as you shed lower stages, and your rocket will become more maneuverable as you shed each stage...

Edited May 22, 2014 by Northstar1989

[undercoveryankee]

That's actually *EXACTLY* where you should be putting all your reaction wheels. The further they are from the Center of Mass, the *more* effective they are (assuming you're not having problems over-steering). Think of your rocket as a giant lever...

There are a lot of urban legends about reaction wheels. I put mine as near as I can to the center of the ship, and approximately symmetrical around the center if I have to spread them out, based on this comment by C7.

It's RCS where you want to think of the ship as a lever. Reaction wheel torque is more like attaching the axle of a motor to the ship at the reaction wheel: the quickest and smoothest rotation happens if you're balanced around the place where you're applying torque.

[Jimbimbibble]

Reaction wheels are equally powerful no matter where they are placed because their torque is fixed and is independent of distance to CoM, whereas RCS is more effective at the extremities.

[undercoveryankee]

Reaction wheels are equally powerful no matter where they are placed because their torque is fixed and is independent of distance to CoM, whereas RCS is more effective at the extremities.

Torque is fixed, but angular acceleration = torque / moment of inertia, and any distance between the torque source and the center of mass increases the applicable moment of inertia.

[TrooperCooper]

Thanks for all the input guys. Good food for thought. To bad I cant see a consensus on wether to move reaction wheels to the center or not. I guess I have to run some experiments for myself.

[BudgetHedgehog]

SAS wheel will always provide the same amount of torque no matter they are placed, according to the CoM. What will affect the overall torque is that SAS always wants to rotate itself. Put it at the CoM, it'll rotate the whole ship most efficiently.

Look at this:

The reaction wheels (green) are placed as far from the CoM (red) as possible. They want to spin around. On a fairly rigid ship, this won't be a problem, but on one with many docking ports and connections, the ship will bend like the picture. Of course, you'll still get some rotation because physics, but it won't be as effective and could possibly be more destructive.

[cantab]

The 2.5 m ASAS is a notorious weakpoint as it is. It can cause problems in a Rockomax stack, never mind a kerbodyne one, and I usually have to strut past it.

Instead, I suggest sticking a cubic octagonal strut on the side of your rocket and mounting a reaction wheel on that. Or you can use Editor Extensions or SurfaceNodes (shameless plug!) and ditch the cuboct strut. Use symmetry if you're concerned about balance.

[Kasuha]

Action of reaction wheels is not matter of public consensus, it is matter of physics and whether you do or don't understand it. The only thing I can say is that the KSP game physics of reaction wheels is close enough to reality (considering their torque, not matters of saturation). It does not matter where you place them or how bendy your ship is, the resulting torque on your ship is always the same. The only difference is whether your ship will rotate as a solid piece of matter or as a bungee jumping cord. Usually the former is desired and to achieve it (or get as close as possible), you need to distribute your reaction wheels evenly through the mass of your ship.

Regarding 3.75 m inline reaction wheels - if any should be introduced to the game, it should be stronger than what we already have. In real life, reaction wheels are usually WAY smaller and are mounted inside the ship along with other components, don't take up the whole diameter. KSP equivalent of that is surface mounting:

[Claw]

Torque is fixed, but angular acceleration = torque / moment of inertia, and any distance between the torque source and the center of mass increases the applicable moment of inertia.

Yes, but moment of inertia is completely separate from the distance from CoM where you are applying the torque. Applying torque at the end of a beam vs. the center of a beam has no effect on Moment of Inertia.

That's actually *EXACTLY* where you should be putting all your reaction wheels. The further they are from the Center of Mass, the *more* effective they are (assuming you're not having problems over-steering). Think of your rocket as a giant lever...

This is false. This comment comes up repeatedly and it simply isn't true from the standpoint of "effectiveness." Torque is equally as EFFECTIVE regardless of location. Placement has other side effects such as bending, but leverage does not come into play when you talk about how capable a given amount of torque is.

Reaction wheels are equally powerful no matter where they are placed because their torque is fixed and is independent of distance to CoM, whereas RCS is more effective at the extremities.

Yes. Except to say that RCS is more effective at the extremities for rotation.

The reaction wheels (green) are placed as far from the CoM (red) as possible. They want to spin around. On a fairly rigid ship, this won't be a problem, but on one with many docking ports and connections, the ship will bend like the picture. Of course, you'll still get some rotation because physics, but it won't be as effective and could possibly be more destructive.

This is the important part in terms of practicality. Torque modules can cause bending, which can be catastrophic or at a minimum, cause strange effects to your craft (depending on size and flexibility).

There are a lot of urban legends about reaction wheels. I put mine as near as I can to the center of the ship, and approximately symmetrical around the center if I have to spread them out, based on this comment by C7.

It's RCS where you want to think of the ship as a lever. Reaction wheel torque is more like attaching the axle of a motor to the ship at the reaction wheel: the quickest and smoothest rotation happens if you're balanced around the place where you're applying torque.

Yeah, spreading them out works best for most situations. But C7's comment is still misleading.

The placement does matter for reaction wheels. Generally speaking they can cause some problems if placed far from the COM. Imagine you are grabbing that point and rotating it. That is what the reaction wheels will try to do. You'll get offcenter rotation anywhere other then near the COM.

They do NOT cause offcenter rotation. There are other reasons why you want to scatter torque modules around, but the craft will always rotate around the CoM.

On a similar note, putting torque modules at the center of mass also does not guarantee that your space station won't wobble itself apart. It does tend to reduce the chances, but doesn't eliminate the possibility.

Edited May 18, 2014 by Claw

[Dispatcher]

I think that some of the differences of opinion regarding reaction wheel placement stem from the fact that most command pods have torque (reaction wheels) built in; since the pods tend to be far from the CoM during launch, its easy to get the idea that craft benefit from the placement of reaction wheels away from CoM in terms of the length of the craft. I've done this myself with acceptable results. However, my understanding is that the placement of reaction wheels closer to the CoM should be more efficient.

[cantab]

Regarding 3.75 m inline reaction wheels - if any should be introduced to the game, it should be stronger than what we already have. In real life, reaction wheels are usually WAY smaller and are mounted inside the ship along with other components, don't take up the whole diameter. KSP equivalent of that is surface mounting

Real reaction wheels are also orders of magnitude less capable than what we have in KSP. Real control moment gyros, on the other hand, are closer to what we have in KSP, and they're big and heavy enough to be stack elements for the smaller sizes. The larger radial sizes are still a bit too big, but they make for convenient construction. Not that I wouldn't like a radial CMG too.

One of the ISS's CMGs: http://en.wikipedia.org/wiki/File:ISS_gyroscope.jpg

Bit of info from Boeing: http://www.boeing.com/assets/pdf/defense-space/space/spacestation/systems/docs/ISS%20Motion%20Control%20System.pdf

[Claw]

... However, my understanding is that the placement of reaction wheels closer to the CoM should be more efficient.

Depends on what you mean by efficiency.

I couldn't find my old post on this subject, so I went ahead and recrated my old experiment (and took it 1 step further). I'll spare you my actual demo that you can do at home with a hammer, but here you are.

Javascript is disabled. View full album

So for this album, I made a completely symmetric craft fore/aft. There are two reaction wheels near the very center, two forward, and two aft. There are also three probe cores, but they don't really serve any purpose (other than to actually control).

Gravity is hacked and all reaction wheels start disabled. Probe core torque is also disabled (for all experiments). Note the position of the CoM in the SPH. Also note the starting position of the CoM in the second picture relative to the runway stripes. For all test runs, I turned only two reaction wheels at a time. Then I recorded how long it took to yaw through 360 degrees, and took note of the center of rotation (using the runway as a fixed reference point). After a test, I reverted back to launch to do it all again.

Center Two Reaction Wheels Time: 23.6 seconds

Aft Two Reaction Wheels Time: 23.8 seconds

Fore Two Reaction Wheels time: 23.4 seconds

Aft/Fore Two Reactions Wheels Time: 23.5 seconds

I only repeated each of these once (because I've done this countless times), so before you run off and conclude that putting the reaction wheels at the front buys you 0.2 seconds, bear in mind that I was manually timing inputs and 360 degrees. So for all practical purposes, a 0.2 second spread pretty much means there's no difference.

Additionally, if you look at each of the pictures you will see that the center of rotation is still around the center of mass. If we were to say "the craft wants to rotate around the reaction wheel," the pictures would be dramatically different. As it is, I challenge you to tell which one is which (without the titles). A real mind bender is to see if you can figure out if I didn't purposely mix up the pictures...

[juanml82]

Can they still cause an instable ship to wobble if placed far from the CoM, though? If they do, it would make sense to put it near the CoM. And since we're on subject, your experiment also indicate that adding more doesn't make a difference.

[Claw]

Can they still cause an instable ship to wobble if placed far from the CoM, though? If they do, it would make sense to put it near the CoM.

Yes, they can cause wobble and instability. However, they can do that anywhere since wobble is more of a function of a craft's flexibility and inertia. If you want to speak in general terms, having them way out on the end for an already bending/wobbly ship will tend to make it worse.

In my demo above, I can replace the two central fuel tanks with modular girders and make it wobble more than with the fuel tanks. If I use big fuel tanks with modular girders, I can make the whole thing come apart. And I can do that with reaction wheels in the center or at the ends, but it's definitely worse at the ends.

And since we're on subject, your experiment also indicate that adding more doesn't make a difference.

In my experiment, I only ever ran two reaction wheels at a time (that was part of the demonstration). If I were to turn on more of the reaction wheels, it would turn a lot faster. In fact (just for you ), I just ran it with all the wheels on and it took 12.2 seconds for a full 360.

[cantab]

Adding more makes a difference. Claw did all his runs using two reaction wheels each time, just a different two.

[tutrakan4e]

I have suggested this, alongside with docking ports. What are your opinions on huge docking ports?

[Kasuha]

Can they still cause an instable ship to wobble if placed far from the CoM, though? If they do, it would make sense to put it near the CoM. And since we're on subject, your experiment also indicate that adding more doesn't make a difference.

They can cause instable ship to wobble when they're placed near CoM as well. The only good approach is to distribute them evenly.

Here, have an example of wobble on a ship with reaction wheels at CoM:

[Red Iron Crown]

Kasuha, you had some screenshots of a craft with multiple reaction wheels suspended from a test rig that you showed to convince me of the reaction wheels' location thing. I can't seem to find the thread now, but they illustrated it very clearly. Do you still have them around?

Edit for post below: That's the one, shows it very clearly.

Edited May 18, 2014 by Red Iron Crown

[Kasuha]

Kasuha, you had some screenshots of a craft with multiple reaction wheels suspended from a test rig that you showed to convince me of the reaction wheels' location thing. I can't seem to find the thread now, but they illustrated it very clearly. Do you still have them around?

Found it, keyword was "convince"

Post here

Album:

Javascript is disabled. View full album

[Northstar1989]

RCS thrusters generate FORCE, not TORQUE.

The difference being, torque depends on location of the rocket.

The further from the Center of Mass RCS thrusters are located, the more torque they generate with the same force...

SAS, on the other hand, generates torque, not force, which I forgot when I made my last comment...

Therefore, RCS DOES benefit from being placed as far from the Center of Mass as possible.

SAS DOES NOT benefit from being placed away from the Center of Mass...

The following diagram, stolen from an earlier post on this subject by "mhoram", illustrates the forces generated by each type of control system.

RCS thrusters *DO* benefit from leverage, unlike SAS.

Regards,

Northstar

Edited May 22, 2014 by Northstar1989

[Dispatcher]

Depends on what you mean by efficiency.

I couldn't find my old post on this subject, so I went ahead and recrated my old experiment (and took it 1 step further). I'll spare you my actual demo that you can do at home with a hammer, but here you are. ... Additionally, if you look at each of the pictures you will see that the center of rotation is still around the center of mass. If we were to say "the craft wants to rotate around the reaction wheel," the pictures would be dramatically different. As it is, I challenge you to tell which one is which (without the titles). A real mind bender is to see if you can figure out if I didn't purposely mix up the pictures...

Nice work. For practical purposes, it seems that quantity reigns over location. I've tended to space out reaction wheels to give some kind of balance between the command pod and the starting primary engine, so sometimes thats near the CoM and at other times not.

[Northstar1989]

A little math for those of you curious:

T = r * F

Notice the term "r"? that is the DISPLACEMENT VECTOR- measuring from the Center of Mass it is essentially the distance and direction from the RCS thruster to the Center of Mass...

Regards,

Northstar