Orientation of Satellites

[Morga]

How are artificitial satellites made to point to the surface of the Earth as they orbit? Is it due their spin angular momentum?

 

[rcgldr]

corrected this post:

There are small thrusters used on some sattellites to [STRIKE]orient the satellites [/STRIKE] maintain their orbits.

 

[D H]

First things first: Suppose that a specific (constant) angular velocity will keep a satellite in the desired attitude indefinitely. If that angular velocity vector is aligned with one of the principal axes of the satellite and if the satellite has that exact angular velocity, then some external torque will be needed to make the satellite veer from the desired attitude. That second if is a mighty big if; it is impossible to make a satellite's rotation rate be exactly as desired. The reason attitude control is needed is to counter internal errors in pointing and to counteract those external torques.


Most satellites do not use thrusters for attitude control, at least not as the primary means of attitude control. Many satellites do not even have attitude control thrusters. Many do not have any thrusters at all. The Hubble, for example, does not have any thrusters.

Many techniques are used to keep a satellite pointed in the right direction. Early satellites were spin stabilized. The whole satellite was set spinning either about its minor or major rotational axis. For a rigid body, rotation about either the major principal axis or minor principal axis (the principal axes with the largest and smallest moments of inertia) are stable. (This is not true for non-rigid bodies such as spacecraft with solar arrays. The only stable axis is the major axis.)

Several problems arose with these spin-stabilized satellites. Power, communications, and observation were all compromised because the entire satellite was spinning. Later satellites had two sections, one spinning rapidly for stability and the other either not spinning at all or spinning at one revolution per orbit. That was a somewhat clunky and short-lived approach. A lot of satellites still use angular momentum to provide stability, but the rotating parts are internalized inside the vehicle as momentum or reaction wheels or control momentum gyros.

Most modern spacecraft are 3-axis stabilized. Those that do have internalized spinning parts can use those devices to generate a torque on the vehicle. A forced change in either the rotation rate (momentum wheels) or in the orientation (control moment gyros) results in a torque on the vehicle. Momentum wheels or CMGs are better than thrusters because no fuel is needed. They have a lot going against them, however: They are big and bulky, and they have moving parts, drive motors, and electronics assemblies, all of which can fail. There are other approaches for attitude control, particularly if the control does not need to be all that precise.

One approach is purely passive: Put the vehicle in an attitude where the perturbing environmental torques will put the vehicle back in that attitude should the vehicle drift from the desired attitude. The desired attitude for many satellites is nadir pointing. The satellite needs to be pointing toward the Earth and it needs to rotate at one revolution per orbit to maintain that pointing. The principal perturbing torques on a satellite in low Earth orbit are gravity gradient torque and torque from atmospheric drag. One way to take advantage of these environmental torques is to make the vehicle have a longish cylindrical shape with the cylinder axis pointing radially. Aero torque will be fairly small, and gravity gradient torque will naturally keep the vehicle in this vertical orientation.

Another passive approach is to make the principal perturbing torques counteract one another. Torque from aerodynamic drag becomes a factor on larger vehicles. The International Space Station is a prime example. The favored attitude for the ISS is called Torque Equilibrium Attitude. The gravity gradient torque and aero torque on average cancel one another when the ISS is in this attitude. The little bit that is left over is easily dealt with by the control moment gyros on the ISS.

Another approach is to use the magnets. Those magnets will interact with the Earth's magnetic field and generate a torque on the vehicle. The Hubble, for example, has four magnetic torquers used for momentum management. One problem with momentum wheels is that continued use can either despin them or make them spin too fast. Using the magnetic torquers in conjunction with the momentum wheels enables keeping the momentum wheels spinning, but not spinning so fast that they tear themselves apart.

 

[rcgldr]

Info on Hubble, which uses reaction wheels for orientation when focused on objects. Since it's in low Earth orbit, it's in the outer fringes of the atmosphere, and shuttle missions are required to prevent the orbit from decaying. I'm not sure if they have decided how to end the Hubble's mission (controlled reentry and burnup).

http://en.wikipedia.org/wiki/Hubble_Space_Telescope

 

[D H]

Careful there, rcgldr. The Hubble uses gyroscopes as attitude sensors, not effectors. It uses reaction wheels (momentum wheels) as effectors.

 

[rcgldr]

Careful there, rcgldr. The Hubble uses gyroscopes as attitude sensors, not effectors. It uses reaction wheels (momentum wheels) as effectors.

Maybe there's a naming issue, based on this quote from the wiki article:

HST uses gyroscopes to stabilize itself in orbit and point accurately and steadily at astronomical targets.

However I also find this at wiki:

Momentum wheels (used in the Hubble Space Telescope) are a different type of actuator, mainly used for gyroscopic stabilization of spacecraft : momentum wheels have high rotation speeds (around 6000 rpm) and mass.

http://en.wikipedia.org/wiki/Reaction_wheel

 

[D H]

It's wikipedia. What more need be said?

Those reaction wheels on the Hubble are about as akin to gyroscopes as are the rear wheels on a car. Yes, you get a gyroscopic effect due to the rotation of the wheels. With the rotation axis fixed to the vehicle it doesn't really count as a gyro though.

 

[Integral]

Thank you DH for that execellent post.

 

[rcgldr]

It's wikipedia. What more need be said?

other articles also seem to be conflicting on the termonology.

Those reaction wheels on the Hubble ... With the rotation axis fixed to the vehicle it doesn't really count as a gyro though.

There's mention of the wheels moving at around 6,000 rpm. Is orientation changed by pivoting the axis (gyroscopic precession) of the wheels or by angular acceleration (torque) of the wheels? Since you mention fixed axis, then I assume it's angular acceleration. If it's angular acceleration, then how is the angular velocity kept within some reasonable bounds?

There's also mention of running with just 2 wheels and the possiblity of just one. Is it possible to both pivot the axis and adjust angular velocity?

 

[D H]

other articles also seem to be conflicting on the termonology.

Momentum wheels and reaction wheels are two names for the same thing.

There's mention of the wheels moving at around 6,000 rpm. Is orientation changed by pivoting the axis (gyroscopic precession) of the wheels or by angular acceleration (torque) of the wheels?

Momentum wheels are similar the rear tires on your car in the sense that the axis of rotation is fixed with respect to the vehicle. Changing the angular velocity of the wheel changes its angular momentum, and this demands an equal but opposite change in the angular momentum of the vehicle proper.

Control moment gyros use the alternate approach. They are kept rotating a constant rate. Control torque is obtained by pivoting the gyro. As some anonymous porter discovered in the Paris train station when he picked up a suitcase, a small force applied normal to a spinning gyro can result in a rather large torque. Nobel physicist Jean Perrin put a powered-up airplane gyroscope in that suitcase and left it unattended to see what fun would ensue.

Since you mention fixed axis, then I assume it's angular acceleration. If it's angular acceleration, then how is the angular velocity kept within some reasonable bounds?

Good question! The Hubble is also outfitted with magnetic torquers. Torquing the vehicle one way with the magnetic torquers and in the opposite direction with a momentum wheel that has an out-of-range angular momentum will result in zero net torque to the vehicle as a whole. There is however a non-zero torque to the momentum wheel, bringing the wheel's angular momentum back to its proper operating range. Angular momentum is of course conserved. The momentum wheel is transferring angular momentum to/from the Earth's magnetic field with the help of those magnetic torquers.

A couple of different problems arise with CMGs: Gimbal lock and saturation. The solution is to once again use some alternative mechanism for torquing the vehicle.

 

[rcgldr]

Thanks for the info. Assuming "around 6,000 rpm" is correct, that would be the mid-point of the operating range? Why not zero average rpm and allow it to rotate either way, or is there some advantage to having gyroscopic effect as well, even though the axis is fixed?

 

[D H]

Having the wheels rotating adds a lot of rotational stability to the vehicle. Multiple rotating wheels lends stability to multiple dimensions.