Hacker News Reader

Comment by ryandrake | Hacker News Reader

All I had to do is play a little Kerbal Space Program to understand how difficult it would be to clean up space junk in orbit. Orbit isn't like some 3D grid where you stick a roomba up there, and have it go back and forth until you get everything. Even if you have the technology that lets you grab anything you encounter, you first have to reach something. And then burn some fuel to match its orbit. And then burn some fuel to get to the next thing. And then burn some fuel to match that thing's orbit (more than before because you are now carrying more mass around).

It might be better to keep everything up there. Assuming the junk is re-usable in some way: If one day, we do manage to build some kind of industry in orbit, it will need to consume raw materials, and having all this scrap in orbit already means you saved most of the energy cost it would have taken to get all that mass up there.

View on HN

I think the much more likely method would be a laser that heats one side of an object, causing a net propulsive effect leading to eventual deorbit.

https://en.wikipedia.org/wiki/Laser_broom

Directed energy approaches are the only thing that will scale IMHO.

The heating approach is one way, but I believe something that could use the kinetic energy of ablative laser pulses would work for smaller objects. To get the correct retrograde velocity, however, I believe you would need an orbital platform and some ridiculous laser technology (which, when combined, tend to create political drag)

I also wonder about sending focused packets of ionized gas to collide with an object and reduce its kinetic energy. If you were operating in an opposite orbit you'd have a ~35km/sec closing velocity. Then spit out small targeted puffs of krypton/xenon at 50km/s, it wouldn't take a lot to knock real energy off of the target. (Stubby pencil work says 80 femtojoules per molecule or roughly 30 micrograms of gas to reduce the velocity of a 20g bolt by 1000m/s. I think.)

How would you prevent the puff of gas from diffusing into nothing? Where did you get ~35km/sec? ISS orbital speed is 7.66 km/s and it only gets slower the higher you go up.

I was thinking something similar but instead launching equivalent mass of water in the exact same orbit in the opposite direction. Right before the collision, release the water. The water would spread enough that the entire spacecraft would basically slam into a wall. The net momentum post collision should be 0. Most spacecraft debris would fall straight down to earth.

Quick caveat, I'm extremely shallow in my knowledge of any of this:

>How would you prevent the puff of gas from diffusing into nothing?

My thinking about this is that at low concentrations the gas would maintain a largely ballistic trajectory and not be subject to typical diffusion properties. Would require a purpose-built accelerator to condition for 'beam' coherency. Not sure how well that scales.

> Where did you get ~35km/sec? ISS orbital speed is 7.66 km/s and it only gets slower the higher you go up.

Major brain-o.

Ultimately I think we're on the same page of 'kinetic deorbit', just a function of what is actually feasible.

I assume you are proposing to launch these packets from some other sattelite? Then by conservation of momentum, you will be boosting your own orbit by the same amount you are diminishing the target, right? Not sure this would be a very sustainable model.

Yes it would require equal/opposite correction and lots of mass on board to send.

The gas idea is interesting, and gave me another idea. Assuming that you could track an object, I wonder how much much you could slow down/redirect debris using air resistance. The idea would be to fill a balloon with air, and then release the air at exactly the right moment just as the debris is passing. The air shoots out into the vacuum right into the path of the oncoming debris. If you got the timing and positioning just right, you could probably both slow down and redirect the debris earthward. These things could use their pressurized air for maneuvering themselves, and given that they just use air itself, you could probably devise an efficient means of refueling them via some kind of weather balloon thing where you'd refill at high atmosphere without having to go all the way back to earth.

I have no idea what I'm talking about really, but its a sort of interesting concept...

>I have no idea what I'm talking about really, but its a sort of interesting concept...

Me neither, fortunately we don't have the burden of actually having to solve this problem xD. The big bubble o' gas idea is honestly where I started and I think there's real merit to the idea...it's relatively harmless to operational satellites yet operates rather indiscriminately on debris and isn't as finicky as the directed energy options. Basically a fly trap for space.

I believe I saw a Scott Manley video recently where he mentioned that he does consulting work for people, and one of his customers apparently doesn't know about KSP because he built their simulation in something like a day and they were astonished that he was done already.

Most of the SpaceX commentators use it too.

It's always cool to me when something meant as a toy gets used for real. Like city planners practicing in SimCity or more recently CitySkylines.

For those that haven't seen the video, he consulted on the Netflix film 'Stowaway' (that comes out later this month). He whipped up a design for the proposed spacecraft in a few minutes to help the directors visualize it better.

Video: https://www.youtube.com/watch?v=WCwXJMVVdck

It should be noted that Kerbal Space Program's orbital mechanics are not all that realistic, especially when you introduce the three-body problem. The world of KSP is also scaled down to about 1/10 of real life unless you use mods. A game like Orbiter really reinforces how much harder the problems are at real-life scale.

It always struck me as weird that you have to mod KSP just to make gravity assists work accurately or Lagrange points work at all.

The default physics model in KSP was chosen for its computational simplicity, which allows for the extensive maneuver planning tools without requiring high-end hardware to continually re-run n-body simulations at several orders of magnitude faster than realtime.

You don't need high-end hardware to do an approximate simulation of a handful of gravitational sources. Also it wouldn't be terrible if maneuver nodes took a tenth of a second to update, and they weren't even in the game for a significant while.

Worth noting that the game came out in 2011, and they wanted people to be able to play it on laptops. They were aiming for really minimal computation.

I mean it's the actual craft physics that are hard to calculate and bog down computers, and you'll never be doing those and orbit fast-forwarding at the same time.

There's another game that I think really goes to show orbital mechanics in how a little for longer is better (in controlibility) than faster for shorter. It's one of those low action games, so if you're into FPS, probably not your cup of tea. There are a couple of levels specifically for the oribiting examples.

https://apps.apple.com/us/app/osmos-for-ipad/id379323382

In theory you could do it without completely matching orbits to save a lot of fuel. Just getting a close slow encounter, grabbing it and letting your orbit just be whatever the new average of your velocities is. It'd require a stronger capture system perhaps depending on the mass of your capture vehicle, but it would save a lot of fuel.

Recycling what's up there might eventually be viable but for the start we'll be shipping up raw stocks. Recycling is probably about as difficult as refining from metal rich asteroids once you consider all the coatings and paints that are added to things and having to sort all the different metals, plastics, and what have you out.

Sorry, not practical. Intercepting debris requires FAR more delta-v than you may expect. Changing the plane of your orbit is so expensive you just don't do it. Its cheaper to just launch a new satellite in a different plane. For example, to intercept hubble from the ISS (or vice-versa) you are looking at a velocity change of a bit over 3000m/sec. Compare that to delta-v of just over 9000 m/sec to reach the ISS from the surface of earth.

Lets say you are in a really, really close orbit to your target and you only need to spend delta-v of 100m/sec to intercept . If you only use 90% of the fuel you'd need to perfectly match orbits then you are still going to impact your target at 10m/sec. Thats 36 km/h, which is a very violent collision. You aren't going to just grab your target on the way past.

I know plane changes are a real problem but you can get nearly velocity intercepts without matching orbits. The timing is brief but to objects on dissimilar orbits can come to a relative standstill for brief periods when their orbits are setup properly.

Also 36 km/h isn't /that/ fast you could have a capture mechanism that absorbs that impact or where it doesn't matter like a net.

Sorry, I have no idea what you are talking about. I can't see how two objects on dissimilar orbits could possibly be at a standstill relative to each other.

I mean, if they are at a standstill relative to each other then their orbits would be identical??

Reply on news.ycombinator.com