Novice question. A few days ago I watched a TV documentary about Virgin Orbit’s failed mission earlier this year to send satellites into orbit from Spaceport Cornwall. One of the satellites shown appeared to about the size of a shoebox. It was secured inside a compartment pressing down against a spring. At the correct altitude, the door of the compartment was supposed to flip open and the spring would eject the satellite into orbit. This seems a bit of a haphazard means of placing a satellite into orbit. Would the satellite have engines of some sort allowing it to be precisely manoeuvred? Do all artificial satellites have engines?


4 Answers 4


Most satellites do have a thruster system of some sort on board. Indeed, the lifespan of a satellite is often determined by how much fuel it can carry, because once it runs dry, it'll start to slip out of its orbit due to a number of physical processes (for example, drag from the few atmospheric particles that get up that high, sunlight pressure, and the earth's slightly asymmetric gravity field). Most of the time, when the fuel is running out, we spend the last bit to kick the satellite into a "graveyard" orbit (an area set aside for nonfunctional satellites where they'll stay well away from the active orbits and are unlikely to collide with anything), or drop the satellite's lowest orbital point far enough that atmospheric drag will eventually bring it down (which may take weeks or months).

There are a lot of thruster technologies, and only some of them resemble a rocket. Hypergolic propellants have historically been popular for thrusters; in that case it's a tiny rocket where the fuel is two materials that spontaneously ignite when they're combined in the combustion chamber, which means you don't need an igniter system. The earliest satellites used cold gas thrusters (cold meaning there's no combustion — basically using an aerosol spraycan to get around) and those still get used from time to time. The new hotness is ion thrusters, which don't really look like a rocket at all. They are incredibly weak and power-hungry, but extremely fuel-efficient, so if you're willing to spend days or weeks inching your way into the proper orbit, that's an attractive option. (A notable user of ion drives is SpaceX's Starlink constellation.)

What you're looking at, though, is a cubesat, which is usually sent up as a ride-share with some larger payload that doesn't completely consume the rocket's lift capacity (for example, if you're lofting an 8000 lb satellite and the rocket can comfortably lift 8400 lbs to that orbit, you can pretty easily add in a cubesat frame and sell those slots to other people). Cubesats are typically tiny — each "unit" is a cube 10 cm on a side that would fit comfortably in your hand. Most cubesats are one unit, though some of them are two, three, or four units stuck together — but still no bigger than a shoebox. In any case, cubesats often don't have any drive system other than the spring that pops them out of the rocket, and they're intended to reach only an approximate orbit and last only a short time. "Haphazard" is a fully appropriate word. They're mostly meant for university students to get some experience with satellite development or get an instrument into orbit on the cheap, so nobody's going to be overly upset if they're off their intended orbit by a few kilometers.

Some cubesats have been pulling off some really neat tricks with orbital maneuvering, though. At that tiny size, it's possible to angle solar cells to use as a light-sail or reposition in orbit by effectively pushing off the earth's magnetic field.

  • $\begingroup$ So can these cubesats with no drive system be "aimed" in a meaningful way by adjusting the orientation of the carrying rocket? $\endgroup$
    – Peter4075
    Commented Sep 6, 2023 at 17:29
  • 15
    $\begingroup$ @Peter4075: The rocket is going 7.8 km/s. The spring mechanism looks to end up at less than walking speed, which is about 1.4 m/s. So, the velocity vector imparted by the spring is less than 0.02% of the velocity vector of the rocket. No matter where you point the satellite dispenser, it is going 7.8 km/s forwards. $\endgroup$ Commented Sep 6, 2023 at 19:14
  • 7
    $\begingroup$ @Peter4075 Not really, no, they just get approximately whatever orbit the launch vehicle was in at the time. The spring just gives it a little kick so each cubesat ends up in a slightly divergent orbit. You don't get to ask for any particular angle, it's just wherever the rocket happens to be pointed at that moment. That's the deal -- you get into space on the cheap, in exchange it's a tiny box and you don't get to specify your trajectory. $\endgroup$ Commented Sep 6, 2023 at 20:07
  • $\begingroup$ @Peter4075 they can have reaction drives for rotational aiming. And with the increasing miniaturization of ion thrusters I'd expect to see (the more expensive) cubesats to employ them as well $\endgroup$
    – Hobbamok
    Commented Sep 7, 2023 at 13:42
  • $\begingroup$ I would think a tiny gyro would be more typical for rotational control of something that small, but I suppose it's possible somebody might use cold gas for the purpose. $\endgroup$ Commented Sep 7, 2023 at 20:58

Short answer: no.

  • most small satellites (like the cubesat you saw) don't have engines. Rocket engines are complex, and therefore expensive, and these small satellites are usually used in an effort to do a low-cost mission.

  • many large satellites do use engines. These are called thrusters: small rocket engines that are used to change a satellite's attitude. Some satellites can also change their course, this requires larger engines. This is used to e.g. move the satellite into its final position after launch, and to move it to a graveyard orbit at the end of its life.

Rockets are not the only method for changing a satellite's attitude. You can also use reaction wheels, magnetorquers, or radiation pressure (vanes that stick out of the spacecraft and catch sunlight, using the tiny pressure exerted by the sunlight to rotate the satellite).

The advantage of these over thrusters is that they don't consume propellant, so they can be lighter in the long run. Space telescopes also benefit from not having exhaust gases around the satellite (Hubble, for instance: they didn't want exhaust gases to condense on the telescope optics).

  • $\begingroup$ Pressure-fed monopropellant thrusters aren't very complex at all. $\endgroup$ Commented Sep 7, 2023 at 18:00
  • 1
    $\begingroup$ Complex enough that many small, low-cost missions avoid them. $\endgroup$
    – Hobbes
    Commented Sep 7, 2023 at 20:08

A nice example of satellites that have no effectors whatsoever are the two Laser Geodynamics Satellites (LAGEOS 1 and 2). Neither has an onboard computer. Neither has any effectors. Neither has any attitude control mechanisms.

< Source: NASA, via wikipedia. This image is public domain.

The LAGEOS satellites are very simple vehicles that orbit well above the atmosphere. They are solid spheres made of a dense material but are covered with retroreflectors. The retroreflectors enable ground-based observers to ping those satellites with lasers, hence the name. The LAGEOS satellites are used to investigate the Earth's rotation and to investigate aspects of general relativity.

  • 1
    $\begingroup$ This is exactly what came to my mind when I read the question. A link to the Wikipedia page and a picture would be helpful. $\endgroup$
    – DrSheldon
    Commented Sep 7, 2023 at 20:29
  • 1
    $\begingroup$ @DrSheldon Done. $\endgroup$ Commented Sep 8, 2023 at 10:45

In space, simple rules!

This seems a bit of a haphazard means of placing a satellite into orbit.

Agreed, it does seem that way, but when done in a planned and thoughtful way it gets the job done, and in spaceflight where all kinds of things can go wrong and cause big problems.

Propulsion systems are a major headache1 - especially for cubesats and smallsats. If they are ride-sharing on other launch missions or the ISS, then they are the type that can get their job done without being in any ultraspecific orbit.

1They are substantial sources of potential mission failure, they are heavy and take up space which limits what else you can put on the spacecraft, and they are considered "stored energy devices" (reactive chemicals and/or pressurized bottles) and that really complicates the process of getting it cleared to be launched along with other cubesats when hitchhiking on other missions.

They rely on their detumbling systems (attitude measurement and control using solar-electric powered magnetotorquers) to right themselves (without using a propulsion system or expending mass) so that their cameras and other instruments can point in specified directions.

The "haphazard" releases are actually well-planned. For example when released from the ISS they are gently pushed at a controlled speed and specific direction so that they will not ever return and bump in to the ISS.

This applies to both satellites and trash destined for reentry and burnup.

If you look at a deployment from this angle and speed it up, it no longer seems so haphazard:

enter image description here

Would the satellite have engines of some sort allowing it to be precisely manoeuvred? Do all artificial satellites have engines?

Yes, many do and many don't, as pointed out in other answers. But what's really cool is that many use a third option to maneuver. It's somewhat analogous to how a glider (the kind you sit in, or say a hawk when it's not flapping its wings) maneuvers around and goes where it wants without propulsion.

The example I like the best is the constellation of eight CYGNSS satellites that were deployed together and then spread out into a great circle constellation spaced every 45 degrees. These cubesats did not have propulsion, but they did have good attitude control and big flat solar panels that could interact with the drag from the small amount of atmosphere up in LEO.

The technique is called differential drag and is discussed at length in

From Differential Drag Control Scheme for Large Constellation of Planet Satellites and On-Orbit Results and my linked answer:

Here is an example of a simulation of the use of alternating between high and low drag configurations to achieve equal spaced phasing and then to maintain it using small station-keeping drag adjustments:

Figure 8: Time-discretized high-drag commands are assigned to achieve desired slots (b) with commands

Figure 8: Time-discretized high-drag commands are assigned to achieve desired slots (b) with commands

Figure 9: Attitude modes of Dove satellite enable large drag area ratios (a) Orthographic projections with cross-sectional areas. (b) High-drag and low-drag attitudes

Figure 9: Attitude modes of Dove satellite enable large drag area ratios (a) Orthographic projections with cross-sectional areas. (b) High-drag and low-drag attitudes


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.