Category Archives: Spacecraft Sub-systems

Keeping the Light Attitude: Kepler’s Balancing Act

Kepler Balance

Near the middle of last year, the planet-hunting satellite, Kepler, looked like it would never Keple again.  On May 11, 2013, the second of Kepler’s four reaction wheels went out, which made precision pointing, a requirement for Kepler’s scientific mission, nearly impossible.  Some of this site’s posts have covered Kepler’s calls for ideas about how to use the gimpy satellite.  And a fantastic idea finally emerged for how to compensate for Kepler’s non-working reaction wheels.

Before we get to the solution (highlighted a ways below), there may be some of you out there wondering what exactly a reaction wheel (also known as a momentum wheel) is, and why it’s so important to satellite pointing/aiming.  This requires a lengthy, English-based, explanation, with little math (since that’s my weakness).  The main thing to understand is a reaction wheel is a part of a satellite’s navigation, guidance, and control system.  It is actually in the “control” part, since the reaction wheel is responsible for moving a satellite.

But the resulting movement is not what you’d expect.  It’s not the kind of movement that changes a satellite’s orbital path.  It doesn’t move a satellite forwards or backwards in a particular direction.  But a reaction wheel helps a satellite aim or look in a particular direction.  It helps maintain, or change, a satellite’s orientation.  In other words, a reaction wheel affects a satellite’s attitude.  Confused yet?  Maybe it’s beer time again?

Let’s use a kayak in the middle of a calm lake as an example.  It’s a cheap kayak, because to save money, you bought one without a rudder.  But it’s well balanced, and should be easy to control.  You’re sitting in the kayak, facing the front of the kayak.  You are aimed north.  Let’s also assume you are in excellent shape with great core muscles. Have you ever noticed what happens when you turn your body?  If not, then you’re about to be enlightened.

Pretend you’re holding the kayak’s paddle in both of your hands while sitting in the kayak.  Keeping the paddle parallel to the water, turn your upper torso to the right (facing east) while still sitting in the kayak.  What happens to the kayak when you do this?  It should start pointing to the left (west).  If you do the opposite and try to face west, the kayak will attempt to point east.  Congratulations, you’ve just become a reaction wheel!

The video above helps show this action/reaction in a kayak.  So, you’re moving the kayak to face different directions, but not forwards, back, or to the sides.  What you’re doing is changing the kayak’s attitude along the horizontal, or “xy” plane.  Also, you might have noticed your movements don’t have to be huge to move the kayak.  This is the essence of how a reaction wheel moves a satellite.  It also happens to be a great example of Isaac Newton’s Third Law of Motion:  “To every action, there is always opposed an equal reaction.

Reaction wheels help with pointing.  But how do they work?  If you’ve ever played with a gyroscope, or ridden a bicycle or motorcycle (particularly a horizontally opposed BMW), you’ve dealt a little bit with how reaction wheels work.  Reaction wheels are weighted wheels (amount of weight will vary), which are designed to spin, normally very fast (thousands of revolutions per minute–rpm).  When a reaction wheel spins at a particular speed (this too will vary), it resists any external force, such as solar wind, micro-gravity, etc., staying very stable.

But if you were to increase the speed of the wheel using small electric motors, the satellite, in which the reaction wheel is housed, starts spinning in the opposite direction.  If the reaction wheel’s speed is slowed, the satellite responds by changing attitude in the other direction.  With electric motors, space operators can control just how much or how little the satellite will spin, by adjusting the speed of the reaction wheel.  And satellites typically have three other reaction wheels, all mounted perpendicular to each other.  This system, when paired with a computer and other sensors, provides very refined satellite attitude movement in any direction.

Such an attitude control system is very complicated compared to small thrusters, which could also be used to change a satellite’s attitude.  But thrusters require fuel, and there are no gas stations out in space (yet!).  A satellite like Kepler would quickly run out of fuel by constantly firing thrusters to point accurately, unless the designers increased fuel capacity, thereby increasing fuel weight.  And increasing weight of a satellite on a rocket requires the rocket to also have more fuel, which means more weight.  But ultimately, it all means more money.  So reaction wheels provide a neat solution/compromise for Kepler and other satellites–until they no longer work.

the solution

Going back to Kepler then–there are only two reaction wheels left working of the original four.  This means the attitude can only be changed along the axes of those two wheels, right?  In Kepler’s case, the team came up with a great, ingenious solution, one which used something that’s already out in space to help get some control along another of Kepler’s axes–solar pressure.

Image on website. Click to embiggen.

There is apparently enough solar pressure to push against Kepler’s solar panels and keep it fairly stable when the satellite is positioned “just so.”  From what I understand, a thruster is required to fire occasionally to help keep the satellite stable, but it seems to work.  Weird, right?  But it’s apparently working well enough for NASA to propose using Kepler for a different, but equally useful, mission–K2 (read abstract here).  It will still be planet hunting, but instead of one specific region, Kepler will be looking out along the Solar System’s plane at different regions in space.

In essence, Kepler will still be planet-hunting for a while longer.  All it took was a change in attitude.




Why space matters: Imaging satellite operations, part 11—Payloads and programs—the lesson that won’t get learned.

Last lesson, you learned a little bit about satellite busses and bus-driving.  Almost all satellite busses are similar to each other.  If you looked at a RapidEye satellite bus specifications and a Digitalglobe satellite bus, you’d likely think they might be one and the same.

But the payload is different.  The satellite payload (and orbit) is very dependent on the mission the satellite is to perform.  In the case of RapidEye and Digitalglobe, the mission is taking pictures of all the goings-on all over the Earth’s surface.  But maybe the satellite belongs to NOAA, something like the Polar-orbiting Operational Satellite (POES).


POES doesn’t have just one payload, but three:  the Advanced Very High Resolution Radiometer (AVHRR), the Advanced TIROS Operational Vertical Sounder (ATOVS) suite, and the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) Microwave Humidity Sounder (MHS) instrument.  None of these payloads are close to an imagery payload, but then taking pictures is not POES’ mission.  Because POES is government-run, they have not just one, but several important sounding missions (from their website):  Data from the POES series supports a broad range of environmental monitoring applications including weather analysis and forecasting, climate research and prediction, global sea surface temperature measurements, atmospheric soundings of temperature and humidity, ocean dynamics research, volcanic eruption monitoring, forest fire detection, global vegetation analysis, search and rescue, and many other applications.

And this is typical of government satellites and satellite programs.  On almost every government satellite, there is likely to be more than one, probably more than two payloads.  If you look at the Geostationary Operational Environmental Satellite (GOES), its satellite bus hosts many different payloads:  imagery (visible and infrared), a Sounder, a magnetometer, an X-ray sensor, high energy proton and alpha detectors, and energetic particles sensor.  Some also have a solar x-ray imager (SXI), an extreme ultraviolet sensor, Emergency Position-Indicating Radio Beacon (EPIRB), and Emergency Locator Transmitter receivers.

That’s quite a few, right?  The primary reason for all of that on one single satellite bus is:  money.  The government’s programs want to advertise they are getting the most bang for the taxpayer buck, but, insidiously, this can also cause something called requirements creep.  Requirements creep (or scope creep), unfortunately, is a natural outgrowth of satellite and space acquisitions programs (actually, any program).  I won’t go much into government acquisitions programs, but will say they tend to cost a lot of taxpayer money, which is ironic, because they are always put in place with the mission of being custodians and disbursers of taxpayer money.

Reasons for requirements creep in acquisitions programs vary, but ultimately, new payloads and missions are added on, changes to the satellite bus (the design of which was finalized months or years ago) are implemented, etc.

So government satellites tend to have a lot of payloads, are very costly (acquisitions programs tend to run for years—sometimes decades), but the resulting satellites tend to be very capable (in spite of using technology that probably was designed 10 years ago—again, very long acquisitions programs).  But there’s also a lot of risk, because now a lot of money is tied up in one satellite—and what happens if that goes boom?

For more on how US taxpayer money is helping fund space systems, for NOAA, go here (it’s a start); for hosted payloads, go here; for general Department of Defense, go here and here.  Lots of issues for all of them, what the military call “a target rich environment.”

On the other hand, it appears, at least on the imagery side, that Digitalglobe and RapidEye are not only keeping things simple, with one payload per bird, but they are also making a profit.  We will start talking more about their imagery operations in the next lesson.

Why space matters: Imaging satellite operations, part 10–short bus schooling

Labor Day took its toll, but I’m back to write more (hopefully) interesting articles for you.

What a space operator should know about satellite systems and space operations fills several volumes of books, folders and checklists.  But, in essence, a space operator and space operations crew worries about three things:  the satellite ground system, the satellite bus, and the satellite payload (which typically receives a lot of a space operator’s attention).  There is additionally, the mission, but this depends very much on the payload(s).

We’ve already talked about the ground system in this lesson and this lesson.  But now we’re going to get into the satellite itself.  Starting with the satellite bus.

Satellite bus

Satellite bus

The satellite bus, depending on who you are, can be the least interesting or most interesting aspect of the satellite.  Whatever it is to you, be aware that the satellite bus is the foundation, the satellite’s infrastructure, and is what enables the satellite payload to work and talk with the space operations center in the ground system (so it’s not the picture above–although you could have a bus shaped like a bus).  The bus is an integration of a whole lot of sub-systems:  Power (solar panels, batteries), data (satellite telemetry and tracking information, antennas), thermal (radiators—active/passive/both), pointing (control moment gyroscopes, thrusters, reaction wheels), and fuel (for the thrusters).  Plus there are places on it for the payload.

And typically, if the builders and owners of these satellites are trying to buy down the risk of something going wrong, there are duplicates and backups of these sub-systems.  Just take a look at this GOES databook’s illustration on page 103.  Note the redundancies:  command receiver A/command receiver B, DSN Transmitter A/DSN Transmitter B, etc.  And that’s just one sub-system.

Remember, with some very small exceptions (Hubble), there isn’t a way to physically repair satellites once they are in orbit.   And a satellite that’s broken at the very beginning of its mission is just a very expensive and fast moving star in the sky (sometimes it’s a falling star).  In some ways, it’s quite amazing it all works, every—single—day.

What controls all of these sub-systems, is (I bet you know the answer to this one):  another sub-system.  Since we haven’t had the good fortune of raising a sentient race of space monkeys just yet, the space program has had to develop computer systems which are specifically designed to control the other sub-systems.  These computer systems are also known as control sub-systems and are the brains and the nervous system of the satellite.

There’s no mystery to these systems, as they work just like the computer you’re reading this lesson on (but they might be slower, with older hardware).  If you feel the need to learn more about computers generally, then please head over to the TWIT or Revision3 networks and get your learning on.

Since this control sub-system is the satellite’s brain and nervous system, I bet you are thinking:  “Gosh, it must be the most important sub-system on the satellite; therefore, it probably has a backup, too, maybe two.”

And you’d be right in thinking that–congratulations!  So, now you know what a basic (not short) bus is:  control, power, data, thermal, pointing, and fuel.  Maybe one day we’ll get into detail about those sub-systems, but not today—we’re trying to get you from the bus-driving space operator level, to the payload space operator level (which is more interesting to me personally, anyway).

But you do need those bus-drivers to help keep the whole thing a success…

Until the next lesson!