save the Hubble

No, actually, "gyro" means "gyro" here. It's the high-precision sensors
that are giving trouble. The gyrodynes (I think) that actually produce
torque to rotate Hubble have been reliable.

I thought HST is using magnetic torque bars for manoeuvering?

Jan

In article ,
Jan C. =?iso-8859-1?Q?Vorbr=FCggen?= wrote:
...The gyrodynes (I think) that actually produce
torque to rotate Hubble have been reliable.

I thought HST is using magnetic torque bars for manoeuvering?

Magnetorquers are difficult to use all by themselves for maneuvering,
because they can generate torque around only some axes. If I've
visualized this right, the axis of the torque has to be perpendicular to
the current local direction of Earth's magnetic field -- the torquers
cannot generate rotation around the field axis.

What they *are* useful for is to solve a secondary problem. Any of the
systems using wheels basically just stores angular momentum in the wheels.
This is fine if disturbances or maneuvering motions sum to zero in the
long run. But if there is anything that constantly adds angular momentum
on some axis, and never takes it away again, you're in trouble. With
reaction wheels (the simplest case), the wheels have to spin faster and
faster to take up the added momentum; there are analogous but more subtle
bad behaviors in the other approaches. And torques are never perfectly
cyclic: there's always some imbalance somewhere that causes such momentum
buildup eventually.

So any wheel system needs, in addition to the wheels, some way of
"dumping" momentum -- a way to exchange angular momentum with the outside
world, so it doesn't build up forever. If you're in LEO, or elsewhere
with a substantial external magnetic field, magnetorquers are a very good
choice for that. They use nothing but electric power, and you can usually
get at least a small torque around the desired axis *eventually* -- the
wheels will store angular momentum temporarily, so there is no hurry -- by
just waiting for the right point in your orbit. The major alternative is
thrusters, which give results without waiting, but use up fuel and tend to
pollute the neighborhood.

MOST, the spacecraft alluded to in my signature, has reaction wheels and
magnetorquers. With minor variations on the exact wheel type, this is a
fairly standard approach for LEO spacecraft.
--
MOST launched 30 June; science observations running | Henry Spencer
since Oct; first surprises seen; papers pending. |

Steve Hix wrote:
Another point brought up during the JPL interview yesterday, the Hubble
is expected to operate at least through 2005...which means that its
planned 15-year working life will be achieved.


Odd definition of "premature death", Ace.


Well, analogous to a premature death of a 75 year-old human whose
health and general condition would lead observers to expect said human
to reach 85 with reasonable care.

/dps

In sci.space.tech Henry Spencer wrote:
In article ,
Gordon D. Pusch wrote:
The problem is that Hubble only has, what, 3 of it's
original 6 gyroscopes still working...

In this particular context, "gyro" actually means "reaction wheel."

No, actually, "gyro" means "gyro" here. It's the high-precision sensors
that are giving trouble. The gyrodynes (I think) that actually produce
torque to rotate Hubble have been reliable.

There seems to be talk on various pages about "2 gyro mode".
How degraded is this?
Were the gyros brought down, and if so, did they all fail the same way?

How long can the thing run open-loop? Are the gyrodynes "noisy", and this
has to be nulled out over the short term.

Might there be a way of just using star-tracking, although it
probably would greatly restrict the target list.

In article ,
Ian Stirling wrote:
No, actually, "gyro" means "gyro" here. It's the high-precision sensors
that are giving trouble. The gyrodynes (I think) that actually produce
torque to rotate Hubble have been reliable.

There seems to be talk on various pages about "2 gyro mode".
How degraded is this?

I don't know quite enough about Hubble's guidance to be sure, but possibly
not at all, except for perhaps more headaches for the operations crew.

IUE was successively operated on two gyros, one, and none at all, although
that was an easy case because it was in GSO and hence in continuous contact
with the ground.

Were the gyros brought down, and if so, did they all fail the same way?

They certainly brought the old ones down both times. I haven't seen a
detailed report; my understanding is that the analysis of the first batch
did conclude that they all died the same way and that the design has a
fundamental weakness.

How long can the thing run open-loop? Are the gyrodynes "noisy", and this
has to be nulled out over the short term.
Might there be a way of just using star-tracking, although it
probably would greatly restrict the target list.

I'm not sure how much the gyros are needed after locking onto a target.
The fine guidance sensors are optical in any case, and it may be just a
matter of having to get Hubble pointed in the right direction and more or
less motionless so they can lock on. *That* can probably be done with
limited assistance from the gyros, at the price of needing more help from
the ground and hence more TDRS time (to provide communications).
--
MOST launched 30 June; science observations running | Henry Spencer
since Oct; first surprises seen; papers pending. |

Henry Spencer wrote:
In article ,
Ian Stirling wrote:

No, actually, "gyro" means "gyro" here. It's the high-precision sensors
that are giving trouble. The gyrodynes (I think) that actually produce
torque to rotate Hubble have been reliable.

There seems to be talk on various pages about "2 gyro mode".
How degraded is this?


I don't know quite enough about Hubble's guidance to be sure, but possibly
not at all, except for perhaps more headaches for the operations crew.

IUE was successively operated on two gyros, one, and none at all, although
that was an easy case because it was in GSO and hence in continuous contact
with the ground.


Were the gyros brought down, and if so, did they all fail the same way?


They certainly brought the old ones down both times. I haven't seen a
detailed report; my understanding is that the analysis of the first batch
did conclude that they all died the same way and that the design has a
fundamental weakness.

If they were the ones originally built by Bendix GSD then I'm not
surprised. I knew a couple of people who worked on the ST project there
back in the late '70s that used to go on and on about the quality
control there. Hearsay, I know, but IIRC they were always talking about
the thin line between out-of-spec and "good enough". A drift test
failed? Run it again (without tweaking anything or making any changes).
It's good now? Ship it.

--
bp
Proud Member of the Human O-Ring Society Since 2003

Ian Stirling wrote in message ...
In sci.space.tech Henry Spencer wrote:
In article ,
Gordon D. Pusch wrote:
In this particular context, "gyro" actually means "reaction wheel."

No, actually, "gyro" means "gyro" here. It's the high-precision sensors
that are giving trouble. The gyrodynes (I think) that actually produce
torque to rotate Hubble have been reliable.

There seems to be talk on various pages about "2 gyro mode".
How degraded is this?
Were the gyros brought down, and if so, did they all fail the same way?

How long can the thing run open-loop? Are the gyrodynes "noisy", and this
has to be nulled out over the short term.

Might there be a way of just using star-tracking, although it
probably would greatly restrict the target list.

As Henry said, not gyrodynes but just gyros. Pointing
sensors.

Theoretically, if gyros were perfect you'd only need two
for 3-axis attitude control. If you have two orthogonal
vectors (e.g. 2 gyro axes) then you can generate a
third vector orthogonal to both using the cross product,
generating a basis set for a 3-dimmensional coordinate
system. In other words, if you know where forward is,
and where right is, then you can easily figure out where
up ought to be and thus determine your orientation.

The problem with that is that gyro's are, of course, not
perfect in the real world. With 3-gyros it's easier to
keep them all oriented pretty much correctly over
moderate lengths of time. Since, as pointed out above,
you can figure out where the 3rd gyro *should* be
pointing from the other gyro positions. But you still
need an outside source of orientation to maintain
high pointing accuracy over long time periods. With
only 2 gyros you'd need to substitute in that external
orientation source (e.g. star tracking) for the normal,
frequent gyro-only orientation checks. But Hubble has
a huge advantage here, because it looks at the sky
almost non-stop. It's Fine Guidance Sensors (FGS) do
the real pointing work for Hubble. They also provide
just the sort of highly accurate, very frequent
positioning updates that are needed to keep accurate
attitude control with only 2 gyros. So Hubble with
2-gyros has everything needed to maintain accurate
attitude control, but it'll probably take something
somewhat Rube Goldbergian to get it to work that way
properly.