4 ms·

I'm a former NASA design engineer. Reaction wheels are a real weak point in spacecraft design. They basically allow one to accurately point an observing spacecr

by lutusp 13y ago

I'm a former NASA design engineer. Reaction wheels are a real weak point in spacecraft design. They basically allow one to accurately point an observing spacecraft, but without using propellant. If you use propellant, you eventually run out and the mission is over. With electric-motor reaction wheels (a minimum of three), you can keep them going with solar energy. So in principle, they're a big improvement over the alternatives. The problem is that they're constantly in motion -- constantly. If you turn a spacecraft and bring it to a new stable position, that new position must be maintained by constant reaction wheel motion. So they're essentially running all the time, there's constant friction, and their parts and bearings eventually wear out. > Seems like either a more durable item or some sort of cartridge system where you can have like 20 spares ready to go, eject one and slot in a new one or something. You're overlooking something. To turn a spacecraft like Hubble or Kepler in a reasonable time, you must either have a massive reaction wheel, or you need to make it spin fast (it's all about momentum). Typical reaction wheels represent a compromise between mass and rotational speed. If you want to reduce frictional wear, you make them bigger, more massive, which increases the mission cost. If you need to reduce the size and mass of the spacecraft, you make the wheels spin faster, and they wear out sooner. So solving the problem by adding more reaction wheels is a non-starter -- they're too physically large and massive to just add a lot of spares.

keenerd 13y ago
> a minimum of three Correct me if I am wrong, but don't most systems use six? It is a whole lot easier to use three pairs of wheels. Pairs can be spun up from a dead stop and in the event of reduced power can be used to turn in either direction by braking action alone. Or does "a reaction wheel" imply two counter-rotating flywheels by definition? For an alternative, I've heard some good things about laser ablative thrusters. Still fairly young, but they are powered by electricity and have no moving parts. They do need reaction mass, but unlike traditional engines the mass is a solid block and can't boil off, corrode, leak, freeze or whatever problems affect liquid systems.
- lutusp 13y ago
  > Correct me if I am wrong, but don't most systems use six? Kepler was equipped with four -- three in use, one spare. The idea is that you need one reaction wheel per dimension. > It is a whole lot easier to use three pairs of wheels. All that six wheels would provide is three spare wheels, one per dimension. > Pairs can be spun up from a dead stop and in the event of reduced power can be used to turn in either direction by braking action alone. That's not how reaction wheels are normally used. Two counter-rotating wheels cancel each other out, producing zero net rotational force on the spacecraft. > Or does "a reaction wheel" imply two counter-rotating flywheels by definition? No, just one. If you have two, and if they counter-rotate, they cancel each other out. > For an alternative, I've heard some good things about laser ablative thrusters. Still fairly young, but they are powered by electricity and have no moving parts. Not nearly enough momentum to rotate a large spacecraft.
  - sigstoat 13y ago
    > Kepler was equipped with four -- three in use, one spare. The idea is that you need one reaction wheel per dimension. are you sure? i thought it was four (in a tetrahedral arrangement), but they were all constantly in use, and one failure could be tolerated.
    - lutusp 13y ago
      > are you sure? i thought it was four (in a tetrahedral arrangement), but they were all constantly in use, and one failure could be tolerated. Source: http://kepler.nasa.gov/mission/faq/ http://kepler.nasa.gov/mission/faq/ Quote: "Only 3 reaction wheels are needed to control the 3 degrees of freedom of rotation of spacecraft. But Kepler was provided with 4 reaction wheels, one extra for redundancy in case a wheel fails."
- elteto 13y ago
  A spacecraft in space has three rotational degrees of freedom (translation is a different deal), therefore you only need three wheels to be able to point the spacecraft in any direction that you want. Any more than three is just a spare.
rurounijones 13y ago
I am so far out of my depth here that it is not funny but.. Why not use wheels that use magnets instead of bearings so that there is no friction?
- lutusp 13y ago
  The same thing crossed my mind as I read the various Kepler stories in the press right now. Apparently that technology isn't ready for prime time. Remember that these components must tolerate very high vibration and temperature extremes to qualify for spacecraft use. Most permanent magnets lose their field if raised above a certain temperature (Curie temperature) and many of them aren't very tolerant of vibration. In any case, for various reasons, frictionless bearings aren't ready for spacecraft use.
moocowduckquack 13y ago
Could you do something like spin ferrofluid inside a hollow donut with some coils round it to make a reaction wheel that would take ages to wear out? No bearings and the only friction would be the fluid against the inside of the donut.
- Udo 13y ago
  That friction would be substantial, however, requiring a lot of energy just to keep going. But maybe there is potential in exploring some kind of contactless magnetic wheel suspension for space applications.
  - moocowduckquack 13y ago
    It would need more energy, but on the other hand it would be extremely simple and should last. I guess it depends on how much extra solar you would have to haul to balance the friction. Magnetic bearings are lovely, but seem fiddly.
mapt 13y ago
Laminated rolls of carbon fiber spinning at high RPM on magnetic bearings have been the state of the hype for terrestrial flywheel power applications for the last 13 years: http://www.wired.com/wired/archive/8.05/flywheel.html http://www.wired.com/wired/archive/8.05/flywheel.html Magnetic bearings have been used in fans, and I'm familiar with a bearing-analogous application in Inductrack maglev trains as well. Has this type of tech penetrated into aerospace?
- sigstoat 13y ago
  > Has this type of tech penetrated into aerospace? i worked with a spacecraft launched in 2003 that was definitely not using reactions wheels like that. (and saw a failure.)
elteto 13y ago
Why do the wheels need to be in constant motion? Is there a constant external torque acting on the spacecraft?
- lutusp 13y ago
  Assume that the spacecraft has a small rotational momentum that needs to be corrected. The reaction wheels begin to spin, and the wheels' rotational momentum corrects for the spacecraft's rotational momentum, causing a telescope to stop rotating and point accurately at a given target. Now, the only way to maintain the spacecraft's new orientation and lack of rotation is for the wheels to continue turning, forever. Such a correction might be needed simply to change the pointing of a spacecraft telescope from one target to another -- let's say, to depart one stationary angle, rotate 30 degrees, and come to a new stationary pointing angle. Such a change of pointing angle would in most cases require all three reaction wheels to change their speed of rotation, and all three wheels would have to continue spinning when the new position was reached. Remember that rotational momentum is conserved, there is essentially no friction on orbit, therefore the initial rotational momentum must be present in the spacecraft or its reaction wheels. And trace gases, micrometeorite hits, and even light pressure, over time contribute rotational momentum to the spacecraft.

by lutusp 13y ago

keenerd 13y ago

> a minimum of three Correct me if I am wrong, but don't most systems use six? It is a whole lot easier to use three pairs of wheels. Pairs can be spun up from a dead stop and in the event of reduced power can be used to turn in either direction by braking action alone. Or does "a reaction wheel" imply two counter-rotating flywheels by definition? For an alternative, I've heard some good things about laser ablative thrusters. Still fairly young, but they are powered by electricity and have no moving parts. They do need reaction mass, but unlike traditional engines the mass is a solid block and can't boil off, corrode, leak, freeze or whatever problems affect liquid systems.

lutusp 13y ago
> Correct me if I am wrong, but don't most systems use six? Kepler was equipped with four -- three in use, one spare. The idea is that you need one reaction wheel per dimension. > It is a whole lot easier to use three pairs of wheels. All that six wheels would provide is three spare wheels, one per dimension. > Pairs can be spun up from a dead stop and in the event of reduced power can be used to turn in either direction by braking action alone. That's not how reaction wheels are normally used. Two counter-rotating wheels cancel each other out, producing zero net rotational force on the spacecraft. > Or does "a reaction wheel" imply two counter-rotating flywheels by definition? No, just one. If you have two, and if they counter-rotate, they cancel each other out. > For an alternative, I've heard some good things about laser ablative thrusters. Still fairly young, but they are powered by electricity and have no moving parts. Not nearly enough momentum to rotate a large spacecraft.
- sigstoat 13y ago
  > Kepler was equipped with four -- three in use, one spare. The idea is that you need one reaction wheel per dimension. are you sure? i thought it was four (in a tetrahedral arrangement), but they were all constantly in use, and one failure could be tolerated.
  - lutusp 13y ago
    > are you sure? i thought it was four (in a tetrahedral arrangement), but they were all constantly in use, and one failure could be tolerated. Source: http://kepler.nasa.gov/mission/faq/ http://kepler.nasa.gov/mission/faq/ Quote: "Only 3 reaction wheels are needed to control the 3 degrees of freedom of rotation of spacecraft. But Kepler was provided with 4 reaction wheels, one extra for redundancy in case a wheel fails."
elteto 13y ago
A spacecraft in space has three rotational degrees of freedom (translation is a different deal), therefore you only need three wheels to be able to point the spacecraft in any direction that you want. Any more than three is just a spare.

rurounijones 13y ago

I am so far out of my depth here that it is not funny but.. Why not use wheels that use magnets instead of bearings so that there is no friction?

lutusp 13y ago
The same thing crossed my mind as I read the various Kepler stories in the press right now. Apparently that technology isn't ready for prime time. Remember that these components must tolerate very high vibration and temperature extremes to qualify for spacecraft use. Most permanent magnets lose their field if raised above a certain temperature (Curie temperature) and many of them aren't very tolerant of vibration. In any case, for various reasons, frictionless bearings aren't ready for spacecraft use.

moocowduckquack 13y ago

Could you do something like spin ferrofluid inside a hollow donut with some coils round it to make a reaction wheel that would take ages to wear out? No bearings and the only friction would be the fluid against the inside of the donut.

Udo 13y ago
That friction would be substantial, however, requiring a lot of energy just to keep going. But maybe there is potential in exploring some kind of contactless magnetic wheel suspension for space applications.
- moocowduckquack 13y ago
  It would need more energy, but on the other hand it would be extremely simple and should last. I guess it depends on how much extra solar you would have to haul to balance the friction. Magnetic bearings are lovely, but seem fiddly.

mapt 13y ago

Laminated rolls of carbon fiber spinning at high RPM on magnetic bearings have been the state of the hype for terrestrial flywheel power applications for the last 13 years: http://www.wired.com/wired/archive/8.05/flywheel.html http://www.wired.com/wired/archive/8.05/flywheel.html Magnetic bearings have been used in fans, and I'm familiar with a bearing-analogous application in Inductrack maglev trains as well. Has this type of tech penetrated into aerospace?

sigstoat 13y ago
> Has this type of tech penetrated into aerospace? i worked with a spacecraft launched in 2003 that was definitely not using reactions wheels like that. (and saw a failure.)

elteto 13y ago

Why do the wheels need to be in constant motion? Is there a constant external torque acting on the spacecraft?

lutusp 13y ago
Assume that the spacecraft has a small rotational momentum that needs to be corrected. The reaction wheels begin to spin, and the wheels' rotational momentum corrects for the spacecraft's rotational momentum, causing a telescope to stop rotating and point accurately at a given target. Now, the only way to maintain the spacecraft's new orientation and lack of rotation is for the wheels to continue turning, forever. Such a correction might be needed simply to change the pointing of a spacecraft telescope from one target to another -- let's say, to depart one stationary angle, rotate 30 degrees, and come to a new stationary pointing angle. Such a change of pointing angle would in most cases require all three reaction wheels to change their speed of rotation, and all three wheels would have to continue spinning when the new position was reached. Remember that rotational momentum is conserved, there is essentially no friction on orbit, therefore the initial rotational momentum must be present in the spacecraft or its reaction wheels. And trace gases, micrometeorite hits, and even light pressure, over time contribute rotational momentum to the spacecraft.