@cstross In terms of long term infrastructure, that gives a use case for "space manufacture" in the form of a permanent structure for receiving, recycling, & processing satellites. Pull old satellites out of orbit, open them up for a refurbish with updated tech & refuel, pop them back into orbit; no need to go in & out of the gravity well per satellite, simply send & recieve bulk materials to the station.
Finally a plausible excuse for an L5 O'Neill cylinder!
@HighlandLawyer That might work eventually, but remember, empty Starships have to go down anyway, and it'll be a long time (if ever) before they're manufacturing the important bits for Starlink sats in space (the electronics).
I keep calling them cubesats, but the Starlink satellites are approx 250kg and cost around $250,000 to build, so, assuming they can be returned and upgraded, cost-wise, this might work as vehicles such as Starship will often be full going up, and almost empty returning.
@TonyJWells No, Starlink 1 is about 250kg—Starlink 2 is 2000kg! It has to support laser coms for sat-to-sat networking and a big enough antenna to talk to cellphones at ground level. I suspect Starlink 3 will be even bigger.
@TonyJWells Bear in mind SpaceX have also pre-announced a "stretched" Superheavy/Starship stack about 25 metres taller(!), and a likely 3rd generation version using Raptor 4 (Raptor 3 is just about flying) that's even bigger. Starship v.1 has 100 tonnes to LEO payload; if Starship v.3 can deliver 250 tonnes that'll make for far fewer refuelling flights per moon/Mars mission.
Didn't they just (in the last month or so) announce that that has been downgraded to about 40 to 50 tonnes, hence the need to get on with the stretched versions? [¹]
This worries me. A drop to, say, 80 tonnes would be quite understandable but more than halving the expected payload after three flights with performance measurements and on the cusp of the first full flight (assuming the next one meets the objectives, as seems quite possible) makes it seem like a panicky response to things which haven't been properly communicated internally.
@edavies@TonyJWells I missed that: it might make sense if they realized the heart shield was under specified and they needed a much heavier one? Or other affordances to qualify it for human spaceflight.
If Elon wanted to be a trillionaire, they'd expand StarLink to include our Moon. Probably just an extra 20 some satellites at first.
Then they would own an information toll road that every Moon-ambitious party on Earth would have to weigh against the cost of building their own from scratch.
@Phosphenes Too late: NASA's already contracting with Nokia to supply LTE phone service on the moon! First proof-of-principle base station is going up there in the next few months.
What you're suggesting pretty much can't be done, in terms of orbital mechanics. In order to achieve global coverage from LEO satellites, they need to be in a bunch of different orbital planes.
(And also, in LEO it takes a lot of fuel to do orbital plane changes.)
So that means it's impossible for a Starship upper stage to pick up more than a few of them for return.
It makes more fiscal sense to just let them burn up in the atmosphere and let others pay the externalities.
@isaackuo Bear in mind Starlink sats have ion rockets. And yes, they orbit in constellations in different orbits and inclinations—but the point is, there's a whole bunch in each constellation.
(I'm also betting on those "externalities"—atmospheric heavy metal pollution—being illegal in the not too distant future.)
The ion thrusters on Starlink satellites definitely have better performance than typical satellite thrusters, but even so ... it takes a lot of delta-v to do plane changes in LEO.
It could take much less delta-v to move up to a graveyard orbit (where atmospheric drag is negligible).
@isaackuo I'm not expecting them to make plane changes—but to park up in a graveyard orbit until there are enough to justify sending a Starship with extra fuel to fetch them back down again. (Bring them back wholesale, in other words, not one at a time. With 45,000 on orbit and a ten year design life they'd need to replace 12.5/day, or an entire constellation every 5 days. With a 5 year life? Near-daily launches just for maintaining numbers.)
@iinavpov@isaackuo That's what they said five years ago, about the current 4-5000 Starlink sats now in orbit overhead. (How much does the entire world’s cellphone infrastructure cost?)
Your ethernet cables don't go obsolete in 5 years, though.
But maybe I'm misunderstanding. I thought the idea was that the Starlink satellites would need to be replaced after 10 years (or 5 years maybe) due to obsolescence. In other words, it's not about keeping them in use, just deciding what to do with them at EOL.
@isaackuo I'm foreseeing the technology improvement curve. Starlink is improving rapidly right now, but there's going to be an inflection point in 5-15 years when the pace of improvement slows right down. These things are basically cell towers in space, and apart from fuel for station-keeping the tech will eventually stabilize except for the payload. I figure most of the bus will be reusable, if refueled. And the bits that need replacing are the least likely to be manufactured in space.
There has been, for some time, attempts at coming up with schemes to refuel satellites. So far, this hasn't led to anything successful; the orbital mechanics of the situation is pretty daunting.
But for a Starlink specific solution, it could make sense to design supply modules that a Starlink satellite rendezvous with to get more propellant and/or a new payload comms package.
No need to return stuff to Earth, just receiving supplies/upgrades.
@isaackuo I'm betting that the bandwidth demands will spiral far more rapidly than the satellites will support without actual electronics upgrades. At which point you've got serviceable satellites but obsolescent payloads.
Add comment