Reaction wheels and general relativity

[MSFC]

Nobody has mentioned the mystery of relativity of angular momentum yet! Although there is obviously a limit on the material strength of a rotating reaction wheel. If one simply considers the reaction wheel to be stationary and the universe to be rotating AROUND it, then the reaction wheel no longer has any internal stresses due to angular momentum (the problem is left for the rest of the universe to hold together - though fortunately the vacuum is quite rigid, they say). Of course the KSP cameras are part of the rest of the universe, so they don't pick up the rotation of the universe imposed by reactions against the stationary reaction wheel in question (kindly consider reaction wheels one-at-a-time to avoid confusion). :D

[Snark]

5 hours ago, magnemoe said:

Yes you are correct, I assumes they use the unicycle dynamic, this implies it will fail if stuck in one direction like trying to drive up an curbside. 

My understanding is that yes, it does-- a friend mentioned to me seeing someone faceplant while riding a Segway when they tried to climb a curb.

Note that a sufficiently skilled unicyclist can climb a curb, but they've got a couple of big advantages over a Segway in that regard (typically a much bigger wheel size, plus they can anticipate hitting the curb rather than merely reacting to it after the fact).

5 hours ago, magnemoe said:

Has seen some moving cubes however who use reaction wheels to generate angular momentum in an direction. Its easy to desaturate by spinning down so slow you are still standing still.

I think I've seen a toy like that, yes.

5 hours ago, magnemoe said:

Told on this forum then jumping with motorcycles or even cars they use wheel rotation as reaction wheels to keep orientation 

Now that, I believe.  Rapidly spinning wheels would have a certain amount of gyroscopic effect that I could imagine might help stabilize them on the yaw and roll axes.  And by speeding up or slowing down the wheels (e.g. by gunning the engine or applying brakes), I would imagine you could control pitch-- that would be exactly what a reaction wheel is.

The gyroscopic effect of spinning wheels helps to stabilize bicycles and motorcycles.  It's not the only thing that does so (steering plays a big role), but it certainly helps.  I recall hearing once (this is an old, dusty memory, and I heard it from someone who'd heard it, so I can't cite source and take this with as much of a grain of salt as you please) that as a demonstration, someone once built a gimmick bicycle with a flywheel mounted on it, deliberately geared to spin in the opposite direction of the wheels-- so as to cancel out (or even reverse) the gyroscopic effect of the wheels.  Result:  a bike that was really hard to ride.  Much harder to keep balance, even with steering.

As I say, the above anecdote is something I heard second- or third-hand, so I can't personally attest to its accuracy (though it sounds quite believable to me).  However, speaking as a unicyclist, I can say that gyroscopic effect and the moment of inertia of a unicycle's wheel makes a big difference to stability.  Lengthy anecdote in spoiler, of interest mainly to people who like unicycles. 

Spoiler

A typical "standard" unicycle has a wheel that's basically a somewhat heavy-duty bicycle wheel, with a diameter of usually 20" or 24", though my own is 26".  The unicyclist becomes extremely acclimated to his or her own personal ike, i.e. its dynamic handling characteristics.  Learning to ride it isn't "academic" at all-- it's all very much about muscle memory and trained reflexes.  So once you've learned to ride, you just hop on and go-- there's no real conscious thought involved, it feels as simple and natural as walking.  It's like the unicycle is part of you.

So in a situation like that, it's kinda hard to have any real awareness of "why can I ride this thing?"-- for example, does gyroscopic effect and the momentum of the wheel have a big effect on the ride?  How much of an effect?  What would it be like if it were different?  It's very hard to know, because it just "feels natural".  The unicyclist isn't aware of the moment of inertia of his wheel, any more than you're aware of the moment of inertia of your legs when you're walking.

However... if you ever have the opportunity to ride a very different unicycle from the one you're used to, then the role of the wheel's inertia can become painfully obvious. 

Back in college, I once had the opportunity to go to a unicycle convention.  It was tons of fun, many hundreds of unicyclists all getting together.  And this included a whole lot of people who had built custom rigs that were deliberately highly unusual in some fashion, and many of them were really happy to let people try them out.

So I got a chance to see the two extremes of wheel angular momentum, and it was highly educational for me.

First, the big wheel.  This is someone who built a unicycle with a giant wheel-- I think what they did was disassemble two unicycle wheels, weld the wheel rails end-to-end to make a single big wheel with double the diameter, put appropriately long spokes in place, and then (I think) take a couple of wheelchair "tires" (which are just strips of solid rubber) and attach them end-to-end around the rim.  Result:  a gigantic wheel that must have been 48" in diameter.  It was like a penny-farthing bicycle that was missing the tiny back wheel.  Anyway, I got to ride it.

Holy mackerel, that thing was fun.    For one thing, the sheer speed you can get on it (due to the bigger-diameter wheel) was quite exhilarating.  But it was also interesting to me because it was rock-solid stable, at least compared with a normal unicycle.  Once it got rolling, that giant wheel had a huge stabilizing effect, which isn't surprising.  Moment of inertia is directly proportional to mass and goes up with the square of the radius.  So if we assume the wheel had double the radius and probably around double the mass... that means its wheel would have eight times the moment of inertia of a regular unicycle's, and it really made a difference when riding.

The gyroscopic effect, of course, would only stabilize the roll and yaw axes (the latter not being that big of a deal, since unicycles aren't unstable for yaw).  However, it was really stable for pitch, too, albeit for a different reason:  the big moment of inertia meant that I could easily exert a lot of pitch-stablizing torque with my feet on the pedals, so in that aspect it was functioning as a highly effective reaction wheel.

 

Next up on the agenda was the tiny wheel.  This was a cute little gizmo.  What the guy had done was to get a teeny-tiny wheel, like only 4" in diameter or so.  Obviously you can't put pedals on that, but what he did was give it a sprocket with a bicycle chain that "geared it up" to some pedals mounted higher on the shaft.    The pedal axle was about the same height above the ground as the pedals of a 20" unicycle would be, and the gearing ratio between pedals and wheel was such that each turn of the pedals would move the unicycle about the same distance as one turn of a 20" wheel would do.

The result was a riding experience that was... odd.  In most aspects of its behavior, it mimicked a 20" unicycle perfectly.  Pedals where you'd expect them to be, and the motion / torque response of the pedals acted exactly like a 20-incher.  So it was easy to hop on and just ride.

But because the wheel was so teeny-tiny, that meant it had practically no moment of inertia at all-- meaning that the guy had, in effect, surgically removed all moment-of-inertia effects from the unicycle.  Result?  Goodness, it was twitchy.  Much less stable, meaning that I had to concentrate more and be quicker to react to small instabilities on the pitch/roll axes.  But by the same token, it was also much more maneuverable; the thing could spin on a dime and made it much easier to do quick / fancy / sharp maneuvers.

(Plus, of course, it required a really clean, flat surface to ride on.  With a tire so small, the darn thing would stub its toe if you try to ride over so much as a twig.)

 

4 hours ago, MSFC said:

Nobody has mentioned the mystery of relativity of angular momentum yet! Although there is obviously a limit on the material strength of a rotating reaction wheel. If one simply considers the reaction wheel to be stationary and the universe to be rotating AROUND it, then the reaction wheel no longer has any internal stresses due to angular momentum (the problem is left for the rest of the universe to hold together - though fortunately the vacuum is quite rigid, they say).

Mainly because it's not really a mystery.

Much of relativity is based on the concept that there's no preferred frame of reference for translational movement.  There's no such thing as "absolute velocity", because "velocity" is meaningful only in relative terms.  You can say what your velocity relative to something is, but there's no "absolute" velocity.  It's not a thing.

However, for rotation, it's a different story entirely.  Unlike translational movement, there is a preferred frame of reference for rotation.  There's such a thing as absolute zero rotation; it is possible to tell whether you're in a rotating frame of reference (and what the speed is, and around what axis) even if you're floating alone in the universe with no external bodies to compare against.

So it's not a "mystery" because it's not really a thing.

[kerbiloid]

Sorry, if they already mentioned, haven't read the whole thread.
https://www.google.com/search?tbm=isch&q=интроцикл&spell=1&sa=X&ved=0ahUKEwjVtf34hpLlAhWqpIsKHfQ1BggQBQg-KAA&biw=2144&bih=1091&dpr=0.9

Spoiler

 

 

[Tsar_bomba]

I have a question relating to reaction wheels. Using ksp Physics, why can’t you achieve forward motion in space with multiple reaction wheels?

[magnemoe]

5 minutes ago, Tsar_bomba said:

I have a question relating to reaction wheels. Using ksp Physics, why can’t you achieve forward motion in space with multiple reaction wheels?

No you can not.
You can however have an long ship, and move fuel from one end to another, rotate it and repeat, this will move the ship as its an bug regarding center of mass. 

[Tsar_bomba]

Just now, magnemoe said:

No you can not.
You can however have an long ship, and move fuel from one end to another, rotate it and repeat, this will move the ship as its an bug regarding center of mass. 

Cool

 

Um....

Spoiler

_uhow do you transfer fullfuel 

 

[Superfluous J]

3 minutes ago, magnemoe said:

No you can not.
You can however have an long ship, and move fuel from one end to another, rotate it and repeat, this will move the ship as its an bug regarding center of mass. 

This does not work. Every action you do that does not eject mass moves you around the center of mass. That includes both rotating and pumping fuel around.

[Terwin]

58 minutes ago, 5thHorseman said:

This does not work. Every action you do that does not eject mass moves you around the center of mass. That includes both rotating and pumping fuel around.

Has that bug been fixed in KSP 1.7.3?  I was under the impression that the bug was still there.

In real life it cannot work, but KSP does not model reactions from moving items inside a vessel.

(probably a good thing, as otherwise we would have instant acceleration followed by instant deceleration from moving a kerbal from one seat to another)

[Superfluous J]

3 hours ago, Terwin said:

Has that bug been fixed in KSP 1.7.3?  I was under the impression that the bug was still there.

I didn't know it existed in the first place. I've never experienced it though I've not tested for it. From my observations both transferring fuel and rotating the ship keep the center of mass as stationary as can be expected. As in, I've never observed a change in my orbit due to either.

Kerbals do not have mass inside ships. That bug still exists. But it doesn't cause any weird ship movement issues that I know of.

[Snark]

5 hours ago, 5thHorseman said:

I didn't know it existed in the first place. I've never experienced it though I've not tested for it. From my observations both transferring fuel and rotating the ship keep the center of mass as stationary as can be expected.

I believe he's right, I think this is a thing. If you have a ship in orbit that's parked right next to something (zero relative velocity) and you pump fuel from one end of your ship to the other, it doesn't move your ship relative to the other one. In the real world it would, but in KSP I think it doesn't.

I believe. It's been a while since I was in a situation that made this evident.

[Superfluous J]

20 minutes ago, Snark said:

I believe he's right, I think this is a thing. If you have a ship in orbit that's parked right next to something (zero relative velocity) and you pump fuel from one end of your ship to the other, it doesn't move your ship relative to the other one. In the real world it would, but in KSP I think it doesn't.

I believe. It's been a while since I was in a situation that made this evident.

I know the camera moves to track the COM. Never tried it with 2 ships, though, so it could be that the ship stays in the same spot while the COM moves.

Time for SCIENCE!

EDIT: SCIENCE achieved. Yes this bug is still there. The camera moves (which is why I thought it was doing it correctly) but the ship does not.

Edited October 11, 2019 by 5thHorseman

[YNM]

If someone still needs to see how reaction wheels really work like (in that you can see reaction wheels init) :

 

Edited October 13, 2019 by YNM
found the english version.

[Snark]

8 hours ago, 5thHorseman said:

EDIT: SCIENCE achieved. Yes this bug is still there. The camera moves (which is why I thought it was doing it correctly) but the ship does not.

It's worth noting that it's more of a missing feature than a "bug" per se. This isn't a question of code not performing as designed-- it's simply an aspect that they've never modeled, is all.

[k00b]

 

On 10/10/2019 at 12:17 PM, MSFC said:

Nobody has mentioned the mystery of relativity of angular momentum yet! Although there is obviously a limit on the material strength of a rotating reaction wheel. If one simply considers the reaction wheel to be stationary and the universe to be rotating AROUND it, then the reaction wheel no longer has any internal stresses due to angular momentum (the problem is left for the rest of the universe to hold together - though fortunately the vacuum is quite rigid, they say). Of course the KSP cameras are part of the rest of the universe, so they don't pick up the rotation of the universe imposed by reactions against the stationary reaction wheel in question (kindly consider reaction wheels one-at-a-time to avoid confusion).

*"rigid" albert suggests... "not a mystery" snark says... before pointing out, the curren't thing, we have, is a "concetual theory" that *"doesn't make sense" (on "mysteries")

...because physics correlate to the electromagnetism of the "spinning biome" which we live... which is spinning around the sun, which exterior is spinning around itself (- because "everything is spinning all of the time" (despite what a tardigrade may tell you)) - so you can use a singular point of reference (the centre of spin, at singular reference frame in "space time", in history for wholly pointless reason (i.e. so you know where you are if you are flying about in space, in singular system (AND ???), if you like ?.

OR you can understand about quantum entanglement verifying the "mystery" of the above... that the sun itself and everything around it is also "spinning" in space, in infinite universe, which factually does not have centre point...

"everything down to quantum particle level is spinning in precise harmony with everything else, in entire UNIVERSE"

"huh that's strange"...

https://en.wikipedia.org/wiki/Indra's_net