Hacker News Reader

Comment by ortusdux | Hacker News Reader

Atmospheric drag causes a relatively speedy deorbit if not actively resisted. The starlink satellites use Hall effect thrusters to maintain their orbit. If a unit fails, it should deorbit in under 5 years. If the thrusters work, they could also force it to deorbit much faster. The risk comes from debris from a possible collision being ejected into a higher orbit.

https://en.wikipedia.org/wiki/Orbital_decay#Atmospheric_drag

https://en.wikipedia.org/wiki/Starlink#Space_debris

View on HN

> The risk comes from debris from a possible collision being ejected into a higher orbit.

Is that a thing that can happen?

My intuition is it would be possible if a rocket currently boosting to a higher orbit were to collide with something on its way there, but two objects in the same orbit colliding couldn't get enough delta-v to actually get to a higher orbit. They could maybe get "higher", but not at orbital speeds and so would rapidly decay.

I know next to nothing about orbital mechanics, so maybe someone who does can provide some better insight here.

Set a baseball on top of a basket ball and drop them both from waist high, the baseball will shoot really high into the air by stealing energy from the basketball. The same principle can occur when you start breaking pieces off of spacecraft with collisions. Also the two colliding objects might not be in "the same orbit" one could be in a polar orbit and one could be in a more equitorial orbit. One could be circular and the other highly eliptical, etc. This could boost parts of one of them into a higher orbit (though honestly it would be just as likely to slow things down).

> Ejected into a higher orbit. Is that a thing that can happen?

Apparently not as such. See sibling comment here:https://news.ycombinator.com/item?id=26809297

I am no expert either, but a paper-napkin calculation tells me yes, this can happen.

Basically, just based on impulse and energy conservation, and some assumptions on the size and number of fragments, a head-on collision of two satellites going opposite directions at the same orbit should send some pieces of debris into much higher orbit.

How likely this is to happen in reality, and thus how big a problem this could be is a subject of modeling and much more complex calculations.

It turns out that intuitive reasoning for orbital mechanics is not easy. Two satellites having a head on collision become space junk because now you have many fragments that themselves can't control their altitude or deorbit. Fragments that are ejected from the collision cannot have more than energy than they had originally, nor can they change direction very much. If pieces are ejected vertically, they are always going to have less lateral velocity than they had going into the collision, therefore will end up in an elliptic orbit with a high probability of colliding with the atmosphere at perigree.

See "can you throw a baseball from the ISS and hit the earth" (No, you can't.)

I think you're claiming that it is impossible for two colliding objects to release fragments that have more energy than either of the original objects. That's wrong.

If it helps, start by imagining two satellite-like machines colliding in interstellar space (i.e. not in orbit), where they were initially moving at 1m/s in opposite directions. Even ignoring the possibility of an explosion (unspent fuel, pressurized areas, I dunno), it's still very easy for interactions to violently fling individual parts outward at speeds higher than 1m/s.

Of course, my guess is that the possible amounts of additional delta-v are pretty low, like 100 m/s, and as such the resulting orbit would not be much higher than the original orbit. But that's just wild conjecture.

If you were to collide two LEO satellites and turn them into fragments, even if you give those fragments much higher velocity than the initial satellites, you _cannot_ launch them into a higher orbit.

Fragments that are launched down obviously hit the Earth. But an orbit is a closed ellipse, so the fragments that are launched up will _also_ hit the Earth -- they'll just go up steeply, turn around, and come down. The only fragments that won't initially hit the Earth are the ones that are ejected tangentially, parallel to the Earth's surface. Those will go into an elliptical orbit with a high apogee and perigee at the altitude of the collision. Which means they will _still _come down to that altitude and gradually lose energy.

I agree that the best case (uh, best for people who like Kessler cascades?) is that the new fragment has a perigee equal to the location of the collision, and a higher apogee. That means, yes, that it will suffer whatever degree of drag was present at the altitude of the collision.

However.

The grandparent post claimed that "Fragments that are ejected from the collision cannot have more than energy than they had originally, nor can they change direction very much", which I think is clearly what I addressed in my comment, and I stand by my correction of that, and your reply doesn't address this at all.

Also, with respect to "you _cannot_ launch them into a higher orbit", you just agreed that you can have a higher apogee than the collision. If the input satellites had perigees lower than the altitude of collision, you can also have a lower perigee. More velocity at this point on the orbit, higher apogee, higher perigee.... So how do you figure you cannot have a higher orbit?

Parts could be squeezed between high pressure zones and get shot out of the collision, but even those parts won't gain a higher orbit because the perigee will be below the original orbit.

First of all, no, the perigee need not be "below the original orbit", where did you get that? The correct thing to say is that the new orbit's perigee cannot be any lower than the location of the collision.

But more generally, how does this justify your claim that "Fragments that are ejected from the collision cannot have more than energy than they had originally"? Or were you conceding that point and making a different point?

See also my more-detailed cousin of this comment.

The key insight there doesn't have that much to do with orbital mechanics, it's more that in a collision, small fragments can come out with vastly more momentum than they went in with, which in orbits translates to a higher apogee. If you have ever bounced a small ball off a large one, you know the effect. See this video for an intuitive example https://youtu.be/2UHS883_P60

Yes, it translates into a higher apogee, but not a higher perigee. So they will still come down to the altitudes with higher drag.

Strictly speaking this is only true if you start with a perfectly circular orbit, right?

Nope, any orbit. Think of it this way: an orbiting particle (or piece of debris) with no independent control/thrust will stay along whatever orbital path it's following, round and round (modulo any further orbital decay). After a collision, the particle is on a new elliptical trajectory, but the new trajectory still passes through the location of the collision. So the lowest point of the new orbit cannot be any higher than the altitude of the collision.

Likewise, if the particle is slowed down, the highest point of the new orbit cannot be any lower than than the altitude of the collision.

An interesting feature of orbital mechanics is that a change in velocity at a given point in orbit will change the altitude of the orbit at every point except where the change took place.

The way to think about orbits is not one height but two: apoapsis and periapsis.

Debris from an explosion/collision can have an unlimited apoapsis, but the periapsis is at most the altitude of the incident.

So if it started in low orbit, it will still drop down regularly, and drag will still remove it over time.

The intuition is: all the debris from a collision now has an orbit passing through the point of collision (at that point of time). Collision at 300km means all the debris must have an orbit that passes through 300km at that moment.

The orbits will be elliptical: exactly how elliptical depends on the directions of the two objects hitting each other and the dispersion of the fragments (how much the fragments have sped up or slowed down compared with original objects, and change of direction).

The fragments the stay in orbit the longest should be those that remain closest to a circular orbit?

Reply on news.ycombinator.com