rotating spacestations
Anderson @ Mpower
to preclude theri design/use ?
I seem to remember someone saying that they were not practical, but I
don't remember what the reasons were...
--
------------------------------------------------------------------
jeffrey (j.d.) wilson
One problem is size. You couldn't launch a rotational space station
without numerous launches. If you did manage to get all of the pieces
launched, you'd also need a considerable amount of construction time.
With today's technology, this would be far too expensive and require
too much EVA time.
Another problem would be maintenance. This design would have far too
many moving parts. Moving parts break down and require EVA replacement.
This would also be too risky to attempt (see criticisms of original
Space Station Freedom design).
--
-----------------------------------------------------------------
My views, not my employers disclaimer here
) ) ) ) ) )
) ) ) ) ) )
__ ._. ----------------------------------------- ) ) ) ) ) )
_/ / \ --- ,-. '\ '\ (; ; o '\ --------- ()()()()()()
| | -- \ |_@_|/_|/ |/| / |/ ^ .-._;_ ()()()()()
_ \ / --- /| |) |) // |/ _/|_/_/_)_|_| | | _()()()()()
`-- '-` ______`'___________________________________) ) ) ) ) )
) ) ) ) )
) ) ) ) ) )
Jeff Wilson
jdwi...@bnr.ca In space, no one can hear you scream!!
ayr...@prodigy.com
------------------------------------------------------------------
William H. Mook, Jr.
lift capacity.
When NASA was spending (inflation adjusted dollars) at $30 billion a
year and was building a Saturn V rocket every few months, a big rotating
space station seemed doable. (I'll email you a separate jpeg file of
Celestial City under consideration at that time)
Of course with restricted Shuttle cargo and minimalist engineering
approaches... things look less practical.
Jim Glass ; JF ; GLASS ; x586-0375 ; (W) ; 634-000
|> Are there any problems associated with rotating stations that would seem
|> to preclude theri design/use ?
|>
|> I seem to remember someone saying that they were not practical, but I
|> don't remember what the reasons were...
|>
|> --
|> ------------------------------------------------------------------
|>
I would guess they're heavy. They have to be large to keep the rotation rate
down and the stress from rotation means they've gotta be beefy. Weight
equals expense when you have to lift everything out of the gravity well...
Jim Glass
Chris Pruett
> Are there any problems associated with rotating stations that would seem
> to preclude theri design/use ?
>
> I seem to remember someone saying that they were not practical, but I
> don't remember what the reasons were...
>
I think a big problem is that to get a decent "gravity", without making
everyone ill from the coriolis effects, one has to build a very, very
big station. Kind of expensive, I imagine.
CP
man...@bnr.ca
In article <49if6d$f...@crchh327.rich.bnr.ca>, "jeffrey (j.d.) wilson" <jdwi...@bnr.ca> writes:
> mpow...@teleview.com.sg (Anderson @ Mpower) wrote:
>>Are there any problems associated with rotating stations that would seem
> One problem is size. You couldn't launch a rotational space station
> without numerous launches. If you did manage to get all of the pieces
> launched, you'd also need a considerable amount of construction time.
> many moving parts. Moving parts break down and require EVA replacement.
It seems to me that you are making some unstated assumptions. In
particular, you are ignoring the possibility of two Stations, vaguely
Skylab-like, each launched in one piece and connected by several
tethers on-orbit.
I am not claiming there are no issues with this scenario, but it
needs to at least be considered if you want a "gravity" station.
/-------------------------------------------------------------------\
| Catherine Mancus <ca...@zorac.cary.nc.us> |
| PP-SEL, N5WVR "God is a sponge." |
\-------------------------------------------------------------------/
djenkins
>mpow...@teleview.com.sg (Anderson @ Mpower) wrote:
>>Are there any problems associated with rotating stations that would seem
>>
>>I seem to remember someone saying that they were not practical, but I
>>don't remember what the reasons were...
>>
>>------------------------------------------------------------------
>>
>
>One problem is size. You couldn't launch a rotational space station
>without numerous launches. If you did manage to get all of the pieces
>launched, you'd also need a considerable amount of construction time.
>Another problem would be maintenance. This design would have far too
>many moving parts. Moving parts break down and require EVA replacement.
>Space Station Freedom design).
Not to start a raging debate in this group, but I must point out that until we retired
the Saturn V, almost *all* space station designs were for rotating stations, backed up
by two-decades of studies that pointed out space stations should have some small amount
of gravity (and anything that needed none, i.e., crystal research, etc., should be
moved to free-flyers).
Almost all the serious designs NASA considered required exactly one launch using a
Saturn V (actually, and INT-20/21 two-stage derivative), and *minimal* EVA since they
were all 'self-deploying' into either a circular shape or a 'spoke' shape. All were
really neat designs. There was a 30-40 foot model of one of them built at Langely to
test the deployment mechanisms, and everything seemed to work fine (after some
experimentation on hinge and seal design).
Alas, no money existed to build one for real, so we ended up with Skylab, and now
International Space Station.
If I ever manage to get the time to finish it, all of this will be documented in my
Space Station book ... I have documentation on almost all the original concepts.
DJ
---------------------------------------------------------------------------
"Ignorance killed the cat, sir, curiosity was framed." | djen...@iu.net
-- Sabrina Perrault-Cadiz |
---------------------------------------------------------------------------
jeffrey (j.d.) wilson
>"jeffrey (j.d.) wilson" <jdwi...@bnr.ca> wrote:
>>mpow...@teleview.com.sg (Anderson @ Mpower) wrote:
>>>Are there any problems associated with rotating stations that would seem
>>>to preclude theri design/use ?
>>>
>>>I seem to remember someone saying that they were not practical, but I
>>>don't remember what the reasons were...
>>>
>>>--
>>>------------------------------------------------------------------
>>>
>>
>>One problem is size. You couldn't launch a rotational space station
>>without numerous launches. If you did manage to get all of the pieces
>>launched, you'd also need a considerable amount of construction time.
>>With today's technology, this would be far too expensive and require
>>too much EVA time.
>>
>>Another problem would be maintenance. This design would have far too
>>many moving parts. Moving parts break down and require EVA replacement.
>>This would also be too risky to attempt (see criticisms of original
>>Space Station Freedom design).
>
>
>Not to start a raging debate in this group, but I must point out that until we retired
>the Saturn V, almost *all* space station designs were for rotating stations,
Actually, you re-stated the point I was trying to make. If we still had
the Saturn V in service, I would not have made the post. On that note,
if we had Saturn V's still in service, we wouldn't have ISSA! My point was
simply that with today's capabilities, it would not be feasible to build a
station of the magnitude required for rotating sections.
--
//////////////////////////////////////////////////////////////////
In space, no one can hear you scream!!
Jeff Wilson
jdwi...@bnr.ca
Richardson, Tx - my opinions are...MINE.
//////////////////////////////////////////////////////////////////
Theodore W. Hall
> I think a big problem is that to get a decent "gravity", without making
> everyone ill from the coriolis effects, one has to build a very, very
> big station. Kind of expensive, I imagine.
> I had read that there was a limit to how fast a station could spin
> because the Corolis forces would make people dizzy. The limit was
> quoted ('Colonies in Space' by [some name I can't remember have to
> look it up]) as 1 RPM.
My news reader is showing at least three threads in sci.space.tech
related to rotating space stations. This thread seems to be the most
"generic" ...
There are several problems to consider in artificial gravity. Mr.
Coriolis tends to get blamed for everything, but strictly speaking
that's not correct.
Coriolis acceleration is associated with relative /linear/ motion in a
rotating reference. It's computed as a cross product of the angular
velocity of the station and the relative linear velocity of an object
within the station:
a = 2 * Omega X v
where: `v' is the linear velocity of an object relative to the station;
`Omega' is the angular velocity (radians per unit time) of the station
in inertial space; and `a' is the Coriolis acceleration of the object in
inertial space.
The Coriolis acceleration makes you heavier or lighter, depending on
whether you're walking prograde or retrograde; it pushes you
horizontally prograde or retrograde depending on whether you're
ascending or descending a ladder. (So, you want to orient your ladder
such that you're pressed into it rather than pulled away from it.)
If you're walking around the circumference of a rotating cylinder or
torus, then there's also a small relative centripetal acceleration that
adds to your weight. If the radius of the cylinder is `r', and its
rotation in inertial space is `Omega', and your rotatation within it
(prograde or retrograde) is `lambda' (or your favorite greek letter),
then your total acceleration is:
a = (Omega +- lambda)^2 * r
= (Omega^2 * r) + (lambda^2 * r) +- (2 * Omega * lambda * r)
= (Omega^2 * r) + (lambda^2 * r) +- (2 * Omega * v)
= global centripetal + relative centripetal +- Coriolis
The centripetal and Coriolis forces affect your mobility and
coordination, but they're not the main cause of dizziness or motion
sickness.
Motion sickness is caused by the relative /angular/ motion of your head.
If you rotate your head about an axis that's not aligned with the
station's rotation, you get an /angular/ acceleration in your
semicircular canals, and illusions that your visual field is rotating
around a mutually perpendicular axis. The direction and magnitude of
the effect is approximately proportional to the cross product of the
angular velocity of the station and the angular velocity of your head.
This is part of astronaut testing: sit on a spinning stool, and nod your
head up and down until you puke.
Several "comfort charts" for artificial gravity have been published
since 1962. The boundaries of the comfort zone vary according to who
you choose to believe:
Minimum acceleration (gravity)
(needed for floor traction)
0.035 g [Hill and Schnitzer, 1962]
0.3 g [Gilruth, 1969]
0.2 g [Gordon and Gervais, 1969]
0.1 g [Stone, 1973] [Cramer, 1985]
Maximum acceleration (gravity)
(for "comfort")
1.0 g [Hill and Schnitzer, 1962] [Gordon and Gervais, 1969]
[Stone, 1973] [Cramer, 1985]
0.9 g [Gilruth, 1969]
Maximum rotation rate
(limited by the onset of motion sickness caused by cross-coupling
of normal head rotations with the station's rotation)
4 rpm [Hill and Schnitzer, 1962]
6 rpm "comfort" [Gilruth, 1969]
2 rpm "optimum comfort" [Gilruth, 1969]
6 rpm [Gordon and Gervais, 1969] [Stone, 1973]
3 rpm [Cramer, 1985]
Minimum rim speed
(the tangential velocity of the station should be large relative to
your walking speed, so that your weight won't change radically
when you walk prograde versus retrograde)
20 fps 6.1 m/s [Hill and Schnitzer, 1962]
24 fps 7.3 m/s [Gordon and Gervais, 1969]
33 fps 10.2 m/s [Stone, 1973]
Maximum head-to-foot gravity gradient, or
Maximum ratio of body height to rotational radius
15% over 6 feet [Gilruth, 1969] [Gordon and Gervais, 1969]
50% over 2 meters [Stone, 1973]
6% over 6 feet [Cramer, 1985]
("0.01 g per foot", and assuming max = 1 g)
The ratio of Coriolis to centripetal acceleration is also important.
You'd like to keep it small:
2 * Omega * v 2 * v
------------- = ---------
Omega^2 * r Omega * r
`v' is a parameter of human behavior - a "given". `r' is limited by
cost. If these are held constant, then reducing `Omega' increases the
ratio of Coriolis to centripetal acceleration. So, while reducing
`Omega' ameliorates problems with dizziness and motion sickness, it
exacerbates gravitational distortion.
Sorry for the long post ... this is a special interest of mine. (This
is also my second attempt to post, since my news reader crapped out the
first time.)
- Ted
Theodore W. Hall ^
Department of Architecture '-`
Chinese University of Hong Kong '- - -`
Sha Tin, New Territories '- - - - - -`
HONG KONG \'- - - - - - - - - -`/
Theodore W. Hall
giving me grief. TWH]
is also my third attempt to post, since my news reader crapped out the
first two times.)
Frank Crary
Theodore W. Hall <twh...@cuhk.hk> wrote:
>Coriolis acceleration is associated with relative /linear/ motion in a
>rotating reference. It's computed as a cross product of the angular
>velocity of the station and the relative linear velocity of an object
>within the station:
> a = 2 * Omega X v
>whether you're walking prograde or retrograde; it pushes you
>horizontally prograde or retrograde depending on whether you're
>ascending or descending a ladder. (So, you want to orient your ladder
>such that you're pressed into it rather than pulled away from it.)
>If you're walking around the circumference of a rotating cylinder or
>torus, then there's also a small relative centripetal acceleration that
>adds to your weight. If the radius of the cylinder is `r', and its
>rotation in inertial space is `Omega', and your rotatation within it
>(prograde or retrograde) is `lambda' (or your favorite greek letter),
>then your total acceleration is:
> a = (Omega +- lambda)^2 * r
> = (Omega^2 * r) + (lambda^2 * r) +- (2 * Omega * lambda * r)
> = (Omega^2 * r) + (lambda^2 * r) +- (2 * Omega * v)
> = global centripetal + relative centripetal +- Coriolis
Not exactly.
(Omega +- lambda)^2 * r = (Omega^2 +- 2*lambda*Omega + lambda^2)*r
lambda = v/r
Where v is the prograde or retrograde velocity rather than
rotation rate. So
(Omega +- lambda)^2 * r = Omega^2 * r +- 2*v*Omega + lambda^2*r
Your second and third equations are incorrect. But if you
look your definition of the coroilis force, you will see
that your final statement is correct. However, lambda is
generally much smaller than omega, so it's easier and
more useful to simply ignore the lambda^2*r term.
(5 m/s is a reasonable, maximum speed for a person. r would be
much larger that 100 m. Omega^2*r is required to be of order
10 m/s^2. So lambda^2 = v^2/r ~ 0.25 << 10 m/s^2 ~ Omega^2*r)
Leaving out the lambda^2 term causes an error of no more
that 2.5% is calculating the effective gravity.
>Motion sickness is caused by the relative /angular/ motion of your head.
>If you rotate your head about an axis that's not aligned with the
>station's rotation, you get an /angular/ acceleration in your
>semicircular canals, and illusions that your visual field is rotating
>around a mutually perpendicular axis.
Err. I don't quite see that. Regardless of the head's orientation,
you would get a varying acceleration. The fluids in the semicircular
canals would be in hydrostatic equilibrium: Their own pressure would be
such that it would cancel out the differential rotation. This
balance occurs almost instantly, so I can't see any significant
acceleration. For the first tenth of a second after turning
one's head, maybe. But in general, no.
>...The direction and magnitude of
>the effect is approximately proportional to the cross product of the
>angular velocity of the station and the angular velocity of your head.
Now you seem to be contradicting yourself. What you are not
describing _is_ a coriolis force, except that you've left
out a factor of r. The force is
2 Omega x v = 2 Omega x (lambda x r)
Where lambda is the rotation rate of the head.
Frank Crary
CU Boulder
Theodore W. Hall
< If you're walking around the circumference of a rotating cylinder or
< torus, then there's also a small relative centripetal acceleration that
< adds to your weight. If the radius of the cylinder is `r', and its
< rotation in inertial space is `Omega', and your rotatation within it
< (prograde or retrograde) is `lambda' (or your favorite greek letter),
< then your total acceleration is:
< a = (Omega +- lambda)^2 * r
< = (Omega^2 * r) + (lambda^2 * r) +- (2 * Omega * lambda * r)
< = (Omega^2 * r) + (lambda^2 * r) +- (2 * Omega * v)
< = global centripetal + relative centripetal +- Coriolis
fcr...@rintintin.Colorado.EDU (Frank Crary) replied:
> Not exactly.
> (Omega +- lambda)^2 * r = (Omega^2 +- 2*lambda*Omega + lambda^2)*r
> lambda = v/r
> Where v is the prograde or retrograde velocity rather than
> rotation rate. So
> (Omega +- lambda)^2 * r = Omega^2 * r +- 2*v*Omega + lambda^2*r
> Your second and third equations are incorrect.
How so? Your equation is identical to my third, except for swapping the
second and third addends; and addition is commutative, even for vectors.
You also write "v*Omega" where I write "Omega*v", which doesn't matter
in a scalar equation (which this is intended to be). If it's re-written
as a vector equation, writing "Omega^2 * r" as "Omega X (Omega X r)" etc.,
then I assert that the correct order is "Omega X v".
Frank Crary continued:
> However, lambda is generally much smaller than omega, so it's easier and
> more useful to simply ignore the lambda^2*r term. (5 m/s is a reasonable,
> maximum speed for a person. r would be much larger that 100 m. Omega^2*r
> is required to be of order 10 m/s^2. So
> lambda^2 = v^2/r ~ 0.25 << 10 m/s^2 ~ Omega^2*r)
> Leaving out the lambda^2 term causes an error of no more that 2.5% is
> calculating the effective gravity.
In a large colony, yes. But there have been many proposals for rotating
space stations with r less than 100 meters and acceleration less than 1 g.
For example, von Braun suggested r ~ 38 meters (125 feet) and acceleration
= g/3, or about 3 m/s^2. So now v^2/r ~ .66 m/s^2 and Omega^2*r ~ 3 m/s^2,
and the ratio is ~ 0.22 or 22% -- I'd call that significant. My reason for
noting this component is to point out one potential problem with small
rotating stations.
I wrote:
< Motion sickness is caused by the relative /angular/ motion of your head.
< If you rotate your head about an axis that's not aligned with the
< station's rotation, you get an /angular/ acceleration in your
< semicircular canals, and illusions that your visual field is rotating
< around a mutually perpendicular axis.
Frank replied:
> Err. I don't quite see that. Regardless of the head's orientation,
> you would get a varying acceleration. The fluids in the semicircular
> canals would be in hydrostatic equilibrium: Their own pressure would be
> such that it would cancel out the differential rotation. This
> balance occurs almost instantly, so I can't see any significant
> acceleration. For the first tenth of a second after turning one's head,
> maybe. But in general, no.
Just because you "don't quite see that", doesn't necessarily mean that I'm
in err :-\ I think that you misinterpret my statement. The effect only
occurs during the rotation. The issue is angular velocity, not orientation.
When the velocity goes to zero, the effect goes away.
Take a toy gyroscope. Spin it up. Hold it in your hand. Try to rotate it
about an axis other than its spin axis. You'll feel a moment about a
mutually perpendicular axis. The same thing happens with the fluid in your
semicircular canals. The moment causes an angular acceleration of the fluid,
and a vestibular illusion of rotation about that axis. When you stop moving
your head, regardless of orientation, the effect dies down.
I wrote:
< The direction and magnitude of the effect is approximately proportional
< to the cross product of the angular velocity of the station and the
< angular velocity of your head.
Frank replied:
> Now you seem to be contradicting yourself. What you are not [sic - now?]
> describing _is_ a coriolis force, except that you've left out a factor of
> r. The force is 2 Omega x v = 2 Omega x (lambda x r) Where lambda is the
> rotation rate of the head.
No, I am not contradicting myself. What is the "r" in your formula?
there's no place for it.
First, remember that radians are a ratio of arc length to radius, and are
dimensionless: radian/s = 1/s = s^(-1)
Coriolis acceleration is a cross product of an angular velocity (1/s)
with a linear velocity (meters/s), yielding a linear acceleration
(meters/s^2).
I'm talking about a cross product of two angular velocities (1/s)
yielding an angular acceleration (1/s^2).
If you still don't believe me regarding this effect, check out these
sources:
Carl C. Clark and James D. Hardy. "Gravity Problems in Manned Space
Stations." _Proceedings of the Manned Space Stations Symposium, April
20-22, 1960_, pages 109-110. Institute of the Aeronautical Sciences,
1960.
Eugene F. Lally. "To Spin or Not To Spin." _Astronautics_, vol. 7,
no. 9, page 57, September 1962. American Rocket Society.
Ashton Graybiel. "Some Physiological Effects of Alternation Between
Zero Gravity and One Gravity." _Space Manufacturing Facilities (Space
Colonies): Proceedings of the Princeton / AIAA / NASA Conference, May 7-9,
1975_, pages 137-149. Edited by Jerry Grey. American Institute of
Aeronautics and Astronautics, 1977.
Frank Crary
>< rotation in inertial space is `Omega', and your rotatation within it
>< (prograde or retrograde) is `lambda' (or your favorite greek letter),
>< then your total acceleration is:
>< a = (Omega +- lambda)^2 * r
>< = (Omega^2 * r) + (lambda^2 * r) +- (2 * Omega * lambda * r)
>< = (Omega^2 * r) + (lambda^2 * r) +- (2 * Omega * v)
>< = global centripetal + relative centripetal +- Coriolis
>> Not exactly.
>> (Omega +- lambda)^2 * r = Omega^2 * r +- 2*v*Omega + lambda^2*r
>> Your second and third equations are incorrect.
>How so? Your equation is identical to my third, except for swapping the
>second and third addends; and addition is commutative, even for vectors.
You are quite correct. I apologize for this misstatement on
my part. I'll plead stupidity. I glanced at your equations,
and only looked at the first two terms. They weren't
the way it would be conventionally written, in order
of decreasing magnitude, so I just made a stupid mistake.
>< Motion sickness is caused by the relative /angular/ motion of your head.
>< If you rotate your head about an axis that's not aligned with the
>< station's rotation, you get an /angular/ acceleration in your
>< semicircular canals, and illusions that your visual field is rotating
>< around a mutually perpendicular axis.
>> Err. I don't quite see that...
>Just because you "don't quite see that", doesn't necessarily mean that I'm
>in err :-\
Another misunderstanding. I wrote, "Err" as it would be used
in conversation (along with "Umm...", "Huh?", etc.) meaning,
"I'm trying to see what you mean, but..." I did not intend
it to mean, "error."
>...The effect only
>occurs during the rotation. The issue is angular velocity, not orientation.
>When the velocity goes to zero, the effect goes away.
Here is where I disagree. The acceleration within the inner
ear is the same as the equation you wrote above. You agreed
that the v Omega term is a cross product, so orientation is
definitely a factor.
>Take a toy gyroscope. Spin it up. Hold it in your hand. Try to rotate it
>about an axis other than its spin axis. You'll feel a moment about a
>mutually perpendicular axis. The same thing happens with the fluid in your
>semicircular canals.
Yes. But you have only considered moving it in two directions,
both perpendicular to its axis of rotation. Try the alternative:
Lift the gyroscope. (You'll probably have to put in on a board,
book or some other, easily lifted surface.) The axis of
rotation is up/down and you will find that lifting or
lowering the gyroscope causes no torque or force. This
is analogous to moving one's head, in an O'Niel colony,
along the axis of the cylinder.
>I wrote:
>< The direction and magnitude of the effect is approximately proportional
>< to the cross product of the angular velocity of the station and the
>< angular velocity of your head.
>Frank replied:
>> Now you seem to be contradicting yourself. What you are not [sic - now?]
>> describing _is_ a coriolis force, except that you've left out a factor of
>> r. The force is 2 Omega x v = 2 Omega x (lambda x r) Where lambda is the
>> rotation rate of the head.
>No, I am not contradicting myself.
Your earlier remarks were unclear and implied (to me) that you
thought the effect was independent of velocity.
>...What is the "r" in your formula?
>there's no place for it.
It has every place. Omega and lambda are rotation rates. They
have units of 1/time. Acceleration has units of length/time^2.
So, if the acceleration depends on Omega * lambda (I believe
we are agreed on this point), then acceleration must
also depend on something else, which has units of
length. That is,
a = [length/time^2] = [1/time] * [1/time] * ?
? must be some distance, such as r. I believe that, if
you work through the algebra, and define v = lambda x r,
you will get the result I stated, above. If you
define lambda as lambda = curl v, then you might
get a different result. This may be the source of
our disagreement.
Frank Crary
CU Boulder
Theodore W. Hall
< ...The effect only occurs during the [head] rotation. The issue is angular
< velocity, not orientation. When the velocity goes to zero, the effect goes
< away.
fcr...@rintintin.Colorado.EDU (Frank Crary) wrote:
> Here is where I disagree. The acceleration within the inner ear is the
> same as the equation you wrote above. You agreed that the v Omega term is
> a cross product, so orientation is definitely a factor.
These are two different things. The Coriolis acceleration (2 Omega X v) is
related to /translational/ motion within the rotating environment. It does
depend on the orientation of translation relative to the axis of rotation -
we agree on that. There is no Coriolis acceleration associated with motion
parallel to the axis, as you noted. We agree on that also.
But with regard to motion sickness, that's not the type of motion I'm talking
about. I'm talking about, for example, sitting in a chair and turning your
head back and forth about your vertical axis while the station is rotating
about an apparent horizontal axis; turning your head about your 'z' or 'x'
axis while the station is rotating about your 'y' axis. What's important
here is the cross product of the station's angular velocity (radians per
second), and your head's angular velocity (radians per second). The issue
is your head's /angular/ velocity around your neck while you're turning it.
Not what orientation it's in when you stop turning. This is independent
of translational motion and Coriolis acceleration.
I wrote:
< Take a toy gyroscope. Spin it up. Hold it in your hand. Try to rotate it
< about an axis other than its spin axis. You'll feel a moment about a
< mutually perpendicular axis. The same thing happens with the fluid in your
< semicircular canals.
Frank wrote:
> Yes. But you have only considered moving it in two directions, both
> perpendicular to its axis of rotation.
That's right - that's exactly my point. I'm talking about rotating your
head about an axis perpendicular to the environments's axis of rotation.
That's one of the major sources of motion sickness.
Frank wrote:
> Try the alternative: Lift the gyroscope. The axis of rotation is up/down
> and you will find that lifting or lowering the gyroscope causes no torque
> or force. This is analogous to moving one's head, in an O'Niel colony,
> along the axis of the cylinder.
Agreed. Translational motion parallel to the axis results in no torque or
force. Rotational motion parallel to the axis (say, for example, facing
prograde and nodding your head up and down), also results in no force or
torque. But again, those are precisely /not/ the types of motions I'm
talking about with regard to motion sickness. Try facing prograde, and
turning your head left-right. Try facing sideways (looking out the window
of a torus, for example), and turning your head either left-right or up-down.
Now you've got a cross-product of two misaligned angular velocities (the
torus's, and your head's), accompanied by a torque around the mutually
perpendicular axis, contributing to motion sickness.
Frank wrote:
> Omega and lambda are rotation rates. They have units of 1/time.
> Acceleration has units of length/time^2.
LINEAR acceleration has units of length/time^2.
ANGULAR acceleration has units of radians/time^2. Since radians are
dimensionless, this is equivalent to 1/time^2. This is exactly the sort of
thing that results from the cross product of two angular velocities.
If you want to put an 'r' in the formula, and describe it in terms of linear
Coriolis acceleration, then the 'r' to use for this effect is NOT the radius
of the station; it's the radius of the semicicular canals.
You can think of it this way: Enter a torus that's rotating with angular
velocity Omega. Face prograde (so Omega is directed to your right). Turn
your head to the left with angular velocity lambda. The left side of your
semicicular canal has a backward relative velocity of lambda X r, while the
right side has a forward relative velocity of lambda X r, where r is the
radius of your semicircular canal. The left side gets a downward Coriolis
acceleration of Omega X (lambda X r), while the right side gets an upward
Coriolis acceleration of Omega X (lambda X r). These two forces constitute
a couple, yielding a counterclockwise moment (viewed from the original
prograde position.
You can describe it this way, but the literature on this topic doesn't. The
people doing the research describe the effect as "proportional in magnitude
and direction to the vector product of the angular velocity of the
environment and the angular velocity of the head", and don't worry about
resolving it into a couple of Coriolis forces at opposite sides of the
semicircular canal.
--
Theodore W. Hall
< The left side gets a downward Coriolis acceleration of
< Omega X (lambda X r), while the right side gets an upward Coriolis
< acceleration of Omega X (lambda X r).
I forgot a factor of 2. The Coriolis acceleration is 2 * Omega X (lambda X r)
That's what happens when it's after midnight, and I have to rush to catch the
last train home. Anyway, to reiterate the end of my previous post:
Picture yourself sitting in a rotating space station, facing prograde. The
station is rotating with angular velocity Omega (radians per second). The axis
is somewhere above your head; the vector "Omega" is directed to your right.
Now, you turn your head to your left, with angular velocity lambda (radians per
second). The vector "lambda" is directed up.
The cross product of these two vectors is an angular acceleration (radians per
second_squared); the vector is directed front-to-back; rotation about the
vector is counterclockwise (as judged from your initial prograde-facing
position).
If you don't like to think about cross-products of two angular velocities
yielding an angular acceleration, then you can, if you wish, compute linear
Coriolis accelerations on the left and right sides of your semicircular canals,
as I described in my previous post. The end result is the same: a
counterclockwise angular acceleration.
Your vestibular sense tells you that you're rotating about that axis, but your
visual sense tells you otherwise. It's this sensory mismatch that causes
motion sickness.
This is different and independent from the Coriolis accelerations associated
with linear motions such as climbing or walking. Those are not the sorts of
motions that make you dizzy.
You get dizzy when you turn your head about any axis that is not aligned with
the rotation of the station.
--
Ted Hall ^
Frank Crary
>Coriolis acceleration, then the 'r' to use for this effect is NOT the radius
>of the station; it's the radius of the semicicular canals.
>You can think of it this way: Enter a torus that's rotating with angular
>velocity Omega. Face prograde (so Omega is directed to your right). Turn
>your head to the left with angular velocity lambda. The left side of your
>semicicular canal has a backward relative velocity of lambda X r, while the
>right side has a forward relative velocity of lambda X r, where r is the
>radius of your semicircular canal. The left side gets a downward Coriolis
>acceleration of Omega X (lambda X r), while the right side gets an upward
>Coriolis acceleration of Omega X (lambda X r). These two forces constitute
>a couple, yielding a counterclockwise moment (viewed from the original
>prograde position.
>You can describe it this way, but the literature on this topic doesn't.
Ok, now I see what you meant. Sorry for taking so long to
figure it out. But is this a significant problem? The
torque is of order Omega*lambda, and the motions you
want the inner ear to be sensitive to are of order
lambda^2 or larger. So I'd only expect problems to
occur if Omega is comparable to lambda. Lambda
would be something like 20 rpm (2 sec^-1), and
the sort of rotation rates we're talking about
are around 1 rpm (0.1 sec^-1).
Frank Crary
CU Boulder
geo...@sover.net
>I wrote:
>< ...The effect only occurs during the [head] rotation. The issue is angular
>< velocity, not orientation. When the velocity goes to zero, the effect goes
>< away.
>fcr...@rintintin.Colorado.EDU (Frank Crary) wrote:
>> Here is where I disagree. The acceleration within the inner ear is the
>> same as the equation you wrote above. You agreed that the v Omega term is
>> a cross product, so orientation is definitely a factor.
You People sound very knowledgeable about spinning space stations and
Coriolis force.
Here is a new problem I haven't the knowledge to solve...
"How will clouds behave in a cylindrical spining space station. My
intuition is that water vapor will migrate to the spin axis and stay
there...
Being that it is not connected to the spinning structure...
An elevator moving from the spin axis to the hub would get a gradual
increase in centrifugal gravity correct?
So will there be problems of storms in a space habitat of this type?
Theodore W. Hall
> "How will clouds behave in a cylindrical spining space station. My
> intuition is that water vapor will migrate to the spin axis and stay
> there...
There is (or was) another thread in one of these sci.space groups
titled "O'Neill colony atmospheres". (The name "O'Neill" was
actually mis-spelled in the title, but anyway ...) The consensus
seemed to be that there probably wouldn't be much weather.
What would cause water vapor to migrate to the spin axis, other than
convection?
What would heat the vapor enough, relative to the surrounding air,
to cause it to convect?
> An elevator moving from the spin axis to the hub would get a gradual
> increase in centrifugal gravity correct?
Correct.
> So will there be problems of storms in a space habitat of this type?
If you drop a ball in a rotating space station, the ball's
trajectory will appear to deflect retrograde relative to a rotating
observer. From an inertial view, the ball falls on a straight
tangent and the observer rotates prograde. From the rotating
observer's view, the ball deflects retrograde. The higher it's
dropped from, the more it deflects and the greater its _relative_
tangential velocity (relative to the rotating ground) when it hits
the ground.
So, IFF we assume, for the sake of discussion, that clouds form near
the axis of a large rotating space colony, and that the clouds
precipitate rain, and the rain droplets have inertia and resist being
accelerated by the rotating atmosphere, then there will be a
prevailing retrograde "blow" to the rain when it hits the ground.
Whether or not it would be a gale depends on the size of the droplets
and how far they fall.
I assume that bigger droplets, with a greater ratio of mass to
surface area, would plow through the atmosphere better and hit the
ground with a greater relative tangential velocity. Lighter mist
would fall more slowly and get spun up to speed by the rotating
atmosphere by the time it reached the ground.
But, it's not evident that clouds and precipitation would occur at
all. The temperature and pressure gradients in a rotating colony
may not be conducive to earth-like weather. I defer to the
atmospheric scientists to figure that out.
--
Ted Hall ^
Theodore W. Hall
fcr...@rintintin.Colorado.EDU (Frank Crary) wrote:
> But is this a significant problem? The torque is of order
> Omega*lambda, and the motions you want the inner ear to be
> sensitive to are of order lambda^2 or larger. So I'd only
> expect problems to occur if Omega is comparable to lambda.
> Lambda would be something like 20 rpm (2 sec^-1), and the
> sort of rotation rates we're talking about are around 1 rpm
> (0.1 sec^-1).
It doesn't take much to deflect a hair cell. But, in short,
you're right: This isn't expected to be a problem at 1 rpm.
The long answer is that it depends on how fast you turn your
head and how you define "problem".
In 1960, Clark and Hardy [1] did some centrifuge studies and
observed that: "normal" head rotations may occur at up to
5 sec^-1 ; the cross-coupling threshold was 0.06 sec^-2 for
illusions and 0.6 sec^-2 for nausea. They proposed to stay
completely below the threshold of illusions, and concluded
that the station rotation should not exceed about 0.01 sec^-1
(that is, 0.06 sec^-2 threshold divided by 5 sec^-1 head
rotation), or about 0.1 RPM. At that rate, a 1-g station
would need a radius of 320,000 feet!
On the other hand, in 1970, Stone [2] assumed "normal" head
rotations of only 3 sec^-1 (rather than 5), and acceptable
cross-coupling up to 2 sec^-2 (more than 3 times the nausea
threshold predicted by Clark and Hardy), giving a maximum
station rotation of 0.67 sec^-1, or about 6 RPM.
Most of the published guidelines (that I've been able to find)
are a lot closer to Stone than to Clark and Hardy. Everyone
agrees that 1 RPM should not be a problem. In 1975, based on
experiments in a rotating room, Graybiel [3] wrote:
In brief, at 1.0 RPM even highly susceptible subjects were
symptom free, or nearly so. At 3.0 RPM subjects experienced
symptoms but were not significantly handicapped. At 5.4 RPM,
only subjects with low susceptibility performed well and by
the second day were almost free from symptoms. At 10 RPM,
however, adaptation presented a challenging but interesting
problem. Even pilots without a history of air sickness did
not fully adapt in a period of twelve days.
A "colony" sized rotating station will have a big radius and low
rotation rate anyway, so this shouldn't be a problem in anything
of that scale.
It may be a problem in smaller applications, such as manned
interplanetary vehicles. Forget about designing a habitat as
a compact cylinder with a high spin rate. Invest in tethers.
Incidentally, in small-scale applications, as an alternative to
rotating the entire habitat, another idea being explored is to
provide an on-board centrifuge or rotating bed to provide the
crew members with a daily dose of gravity when they're not doing
anything else. Rotating beds have been tested on earth with spin
rates in excess of 20 RPM [4,5]. Though people do toss and turn
somewhat while asleep, head movement is generally less than
during wakeful activities, so the bed can rotate faster. One
disadvantage of this approach in an otherwise weightless
environment is that it would require the crew to sleep "standing
up". In that respect, they would have to work harder while asleep
than while awake.
[1] Clark, Carl C.; and Hardy, James D. "Gravity Problems in
Manned Space Stations." _Proceedings of the Manned Space
Stations Symposium, April 20-22, 1960_, p. 104-113. Institute
of the Aeronautical Sciences, 1960.
[2] Stone, Ralph W. "An Overview of Artificial Gravity." _Fifth
Symposium on the Role of the Vestibular Organs in Space
Exploration_, p. 23-33. NASA Scientific and Technical
Information Division, 1973. Special Publication 115:
proceedings of a symposium held in 1970.
[3] Graybiel, Ashton. "Some Physiological Effects of Alternation
Between Zero Gravity and One Gravity." _Space Manufacturing
Facilities (Space Colonies): Proceedings of the Princeton /
AIAA / NASA Conference, May 7-9, 1975_, p. 137-149. Edited by
Jerry Grey. American Institute of Aeronautics and Astronautics,
1977.
[4] Houtchens, C. J. "Artificial Gravity." _Final Frontier_,
vol. 2, no. 3, p. 28+, June 1989. Final Frontier Publishing Co.
[5] Cardus, David; Diamandis, Peter; McTaggart, Wesley G.; and
Campbell, Scott. "Development of an Artificial Gravity Sleeper
(AGS)." _The Physiologist_, vol. 33, no. 1, supplement, p.
S112-S113, 1990. American Physiological Society.
--
Ted Hall ^
'Larry' L Gales
On 29 Dec 1995, Theodore W. Hall wrote:
> Date: 29 Dec 1995 15:00:29 GMT
> From: Theodore W. Hall <twh...@cuhk.hk>
> To: gher...@crl.com
> Newgroups: sci.space.tech
> Subject: Re: rotating spacestations
>
>
> Incidentally, in small-scale applications, as an alternative to
> rotating the entire habitat, another idea being explored is to
> provide an on-board centrifuge or rotating bed to provide the
> crew members with a daily dose of gravity when they're not doing
> anything else. Rotating beds have been tested on earth with spin
> rates in excess of 20 RPM [4,5]. Though people do toss and turn
> somewhat while asleep, head movement is generally less than
> during wakeful activities, so the bed can rotate faster. One
> disadvantage of this approach in an otherwise weightless
> environment is that it would require the crew to sleep "standing
> up". In that respect, they would have to work harder while asleep
> than while awake.
>
================================
Another interesting approach is a bicycle: in a relativly small radius
enclosure, ride a bicycle so that the centrifugal "force" (I have to put
that in quotes so the physicists won't jump on me) will glue you to
the wall at a moderate "G" force -- if you keep you head in a fixed
position, you should be able to get your exercise and your "g-ration"
for the day with, say, 30-60 minutes of cycling.
-- Larry Gales
Theodore W. Hall
>>
>> Incidentally, in small-scale applications, as an alternative to
>> rotating the entire habitat, another idea being explored is to
>> provide an on-board centrifuge or rotating bed to provide the
>> crew members with a daily dose of gravity when they're not doing
"'Larry' L Gales" <lar...@u.washington.edu> wrote:
>
> Another interesting approach is a bicycle: in a relativly small radius
> enclosure, ride a bicycle so that the centrifugal "force" (I have to put
> that in quotes so the physicists won't jump on me) will glue you to
> the wall at a moderate "G" force
Yes - that's been proposed too:
G. Antonutto, C. Capelli, and P. E. di Prampero. "Pedalling in Space
as a Countermeasure to Microgravity Deconditioning." _Microgravity
Quarterly_, vol. 1, no. 2, p. 93-101, 1991. Pergamon Press.
They proposed a system of oppositely-directed, mechanically-coupled,
counter-balanced bicycles running on parallel tracks around the
circumference of a cylinder. The mechanical coupling insures that the
two bikes are always moving at the same speed in opposite directions
relative to the cylinder, so as not to impart any reaction spin to the
cylinder itself.
(If there are only two bikes, then each is counterbalanced by a dead
weight. With four bikes, each could be counterbalanced by another bike.
Either way, half of the system rotates in one direction and half in the
opposite direction.)
Frank Crary
Theodore W. Hall <twh...@cuhk.hk> wrote:
>> "How will clouds behave in a cylindrical spining space station. My
>> intuition is that water vapor will migrate to the spin axis and stay
>> there...
>What would cause water vapor to migrate to the spin axis, other than
>convection?
Diffusion. It's a slow process, but it pretty much assures
that the mixing ratio of water vapor would, eventually, be
constant, even in the absence of convection. But that's
just water _vapor_. It doesn't say anything about the
mass of liquid water droplets. _If_ the air at the
axis was colder than the surface temperature, you'd
get water droplets condensing along the axis. That's
a sink for water, and it might cause a concentration of
water along the axis. But there is a limit to how
much water could build up in such an axial cloud.
Unless it's exactly on the axis, the droplets are going
to fall back down. So, without convection, water
vapor is diffusing up towards the axis and falling
back down as rain. Close to the axis, the centrifugal
"gravity" is very weak. (Note that the cloud is rotating
with the colony, since the drops are small and would
be dragged along with the air, which is rotating with
the station.) The weak "gravity" means that the drops
would take a long time to fall away from the axis.
So a fair amount of water could get trapped near
the axis. But how much depends on the size of the
drops (which determines their fall rate) and the
diffusion rate, which is very slow as well. For
1 mm drops, falling would take minutes. That's
easily faster than the diffusion rate. For
1 micron drops, the fall could take months.
That might be on par with the diffusion rate,
but water doesn't tend to condense into micron
sized drops. The size of the drops grows, the
longer they spend drifting down in a saturated
cloud. So I figure that the drops would
condense, grow, and fall away from the axis
fairly quickly. That would prevent any significant
concentration of water along the axis.
Of course, that all assumes that the axis was
colder than the surface. According to my
analysis, this would be very unlikely. So
the water isn't going to condense in the
first place, and there wouldn't be any
concentration of water near the axis. Condensation
may occur on a daily basis, closer to the
surface. But that's just clouds and rain.
(At night, by the way.)
But this problem of how long a rain drop would
spend near the axis has another implication.
The same drop times also apply to dust. If there
is any source of dust near the axis, the dust
tend to stay there, taking months or years to
fall down to the surface, if the dust is fine enough.
That creates a potential for a cloud of dust
along the axis. I find that there would be no
natural convection to get rid of it, so it
would just collect, to a density that depends
on the size of the dust particles and the
production rate. Since I consider this to
be undesirable, I've added it to my analysis
(in preparation...). There are ways to _force_
convection, even if it wouldn't occur naturally.
That would get rid of the dust cloud. There are easy
and hard ways to do this. Specifically, there
are natural circulation patterns. But they wouldn't
just happen on their own: Viscosity damps them
out, and no natural process would sustain them.
But in some cases, they could be forced, driven
by selective heating of the surface or some kind
of fans. You can always make air move with
fans, but if you do it in sinc with the natural
circulation patterns, you can build up a decent
amount of circulation without putting alot of
power into it. So I've worked out exactly what
you want to do with those fans, to get a desired
circulation for a minimum of effort.
>What would heat the vapor enough, relative to the surrounding air,
>to cause it to convect?
Another early result: You can't. A hot surface would cause
air to rise. But, unlike on Earth, the coriolis force
forces this motion into a fairly small, organized,
motion: Rising, moving in the direction of rotation,
falling, etc. The rising air isn't really transporting
water vapor (or anything) towards the axis, since that
air is going to move back down to the surface after
half a rotation period (30 sec. for the 1 km radius,
1 g case.) For turbulent convection, what you need
to rapidly move water vapor towards the axis, the
temperature of the air would have to decrease with
height. On Earth, that means rapidly enough for
the buoyancy of the rising air to overcome gravity:
Five or ten degrees (K) per kilometer. On an O'Neill
colony, it means rapidly enough for the buoyancy
to overcome "gravity" _and_ the stabilizing influence
of the coriolis force. That turns out to be around
one degree _per_meter_. As far as I know, no natural
force could produce that rapid a decrease in temperature,
and if one could, the people designing the colony would
work hard to prevent it.
>> So will there be problems of storms in a space habitat of this type?
>So, IFF we assume, for the sake of discussion, that clouds form near
>the axis of a large rotating space colony, and that the clouds
>precipitate rain, and the rain droplets have inertia and resist being
>accelerated by the rotating atmosphere, then there will be a
>prevailing retrograde "blow" to the rain when it hits the ground.
>Whether or not it would be a gale depends on the size of the droplets
>and how far they fall.
No. As you point out, this would only happen if the surface
area to volume ratio is small. (cross sectional area, really,
but the difference is only a factor of four, for spherical
drops.) Otherwise, the drops are picked up by the wind and
fall straight down. I find that "small enough to be picked
up by the wind" is (depending on how exact you want to be,
moving at 90% of the winds rotation rate or 99% or what)
about a 10 cm diameter. Rain drops are going to be moving
with the wind. They aren't going to affect the wind speed
by much, either.
>But, it's not evident that clouds and precipitation would occur at
>all. The temperature and pressure gradients in a rotating colony
>may not be conducive to earth-like weather. I defer to the
>atmospheric scientists to figure that out.
Well, I'm hardly an atmospheric scientist. But one of the
advantages of getting a PhD from my department is that
it's interdisciplinary: My thesis will have nothing to
do with atmospheres, but since it's a department of
"Astrophysical, Planetary and Atmospheric Sciences",
I've wound up taking a graduate-level course or
two on atmospheric sciences. That's enough for a
first-cut, rough estimate about the dynamics of
an O'Neill colony's atmosphere.
Frank Crary
CU Boulder