Launching Starlink with Starship

[JamesH65]

I may have missed it ...
What is the point of the VLEO birds at 340km?

Faster throughput

So lower latency.
Who needs less than 20ms?
Stock market traders?
Govt drone operators?

Gamers. Biggest market of all.

[spacenut]

We use high speed internet to keep up with our retirement accounts.  Need it really bad.  A lot of people do.  It has to be in real time. 

[TheRadicalModerate]

I may have missed it ...
What is the point of the VLEO birds at 340km?

Faster throughput

So lower latency.
Who needs less than 20ms?
Stock market traders?
Govt drone operators?

Gamers. Biggest market of all.

Throughput and latency have almost nothing to do with one another.  A 747 filled up with thumb drives, traveling from LA to Paris, has terrific throughput.  But its latency is terrible.

PS (much later): I finally couldn't resist.  A 1TB memory stick is approx. 6cmx2cmx0.5cm=.000006 m³, and weighs .02296kg.  A 747-8F has an 828m³ cargo volume and a 132.6t payload mass.  That would make the thumb drives mass-limited to 5.78E6 individual sticks, which would be 4.62E19 bits.  LA to Paris takes about 11 hours, so the throughput is 1666.7Tbps.

[su27k]

Just thought about this today: Once SpaceX starts launching Starlink using Starship, they could also recover the old Starlink satellites and bring them back to Earth instead of burning them up in the atmosphere. They can do this since the new satellites will need to be launched into the same orbit as the old satellites, so if you're replacing old satellites, the ship is in an ideal position to retrieve the old satellites, store them in the empty cargo bay vacated by the new satellites and land them on Earth.

Why would they want to do this? Two birds with one stone:

1. This way they no longer need to worry about demisablility of the satellite, thus can build them a lot bigger. If it doesn't burn up completely in the atmosphere? No problem, since vast majority will be retrieved and land back on Earth. One or two occasional failed satellites could still re-enter but that's nothing to be worried about, it would be similar to how some F9 upper stage still do uncontrolled re-entry even though it weights 5 metric tons.

2. This would completely destroy the objection that Starlink may cause changes to the atmosphere due to burning up old satellites during re-entry. The enemies of SpaceX will try to make a big deal out of this, better to start planning for it now.

[M.E.T.]

Just thought about this today: Once SpaceX starts launching Starlink using Starship, they could also recover the old Starlink satellites and bring them back to Earth instead of burning them up in the atmosphere. They can do this since the new satellites will need to be launched into the same orbit as the old satellites, so if you're replacing old satellites, the ship is in an ideal position to retrieve the old satellites, store them in the empty cargo bay vacated by the new satellites and land them on Earth.

Why would they want to do this? Two birds with one stone:

1. This way they no longer need to worry about demisablility of the satellite, thus can build them a lot bigger. If it doesn't burn up completely in the atmosphere? No problem, since vast majority will be retrieved and land back on Earth. One or two occasional failed satellites could still re-enter but that's nothing to be worried about, it would be similar to how some F9 upper stage still do uncontrolled re-entry even though it weights 5 metric tons.

2. This would completely destroy the objection that Starlink may cause changes to the atmosphere due to burning up old satellites during re-entry. The enemies of SpaceX will try to make a big deal out of this, better to start planning for it now.

Would probably be of even more benefit to the world to retrieve those much more problematic Oneweb satellites at 1000km altitude. At a nice, hefty price to Oneweb of course.

[libra]

Just thought about this today: Once SpaceX starts launching Starlink using Starship, they could also recover the old Starlink satellites and bring them back to Earth instead of burning them up in the atmosphere. They can do this since the new satellites will need to be launched into the same orbit as the old satellites, so if you're replacing old satellites, the ship is in an ideal position to retrieve the old satellites, store them in the empty cargo bay vacated by the new satellites and land them on Earth.

Why would they want to do this? Two birds with one stone:

1. This way they no longer need to worry about demisablility of the satellite, thus can build them a lot bigger. If it doesn't burn up completely in the atmosphere? No problem, since vast majority will be retrieved and land back on Earth. One or two occasional failed satellites could still re-enter but that's nothing to be worried about, it would be similar to how some F9 upper stage still do uncontrolled re-entry even though it weights 5 metric tons.

2. This would completely destroy the objection that Starlink may cause changes to the atmosphere due to burning up old satellites during re-entry. The enemies of SpaceX will try to make a big deal out of this, better to start planning for it now.

Hell of an idea. It would be like the Shuttle early business case (Hubble), except done RIGHT this time.

Talking about retrieving satellites... even before 1972 and the program starting, the Shuttle and what become Hubble were loosely tied: the idea of refurbishing and relaunch large and expensive space observatories dates from the 60's.
George Low started Hubble in 1971 and Fletcher the Shuttle the next year and thus they become even more strongly tied to each other.

Fast forward to 1990.

One of the few times the Shuttle (and Hubble)  much vaunted retrieval capability could have been really useful was when they found the mirror was flawed. What's more a perfectly fine BACKUP mirror done by Kodak was waiting on the ground.
Yet for a host of complex reasons they found instead that putting COSPAR lenses on Hubble and doing that in orbit made more sense.

To me it is one of the weirdest ironies in the Shuttle program (together with the 1971 decision of a 60 ft long payload bay heavily related to the NRO KH-9s - only for NO KH-9 to ever launch on a Shuttle because the NRO decided against it from 1975...)

[Yggdrasill]

Just thought about this today: Once SpaceX starts launching Starlink using Starship, they could also recover the old Starlink satellites and bring them back to Earth instead of burning them up in the atmosphere. They can do this since the new satellites will need to be launched into the same orbit as the old satellites, so if you're replacing old satellites, the ship is in an ideal position to retrieve the old satellites, store them in the empty cargo bay vacated by the new satellites and land them on Earth.

The only way I see this happening is if all the satellites to be returned find their way into a single orbit in immediate proximity to each other, at say 300 km altitude. But there are still massive challenges.

1. There are significant delta-V requirements in getting all the satellites to the same place. This increases the mass of the required propellant.
2. If Starship is carrying 400 satellites, and will return 400 satellites, that means you need to do 400 orbital rendezvous. That's pretty complicated. That's after dropping off the new Starlink satellites in an orbit where they will not come in conflict with the old satellites.
3. The 400 satellites to be returned will be at end-of-life, so some may be out of fuel or have defective gyros or have similar issues. If some are stuck in all sorts of orbits between 550 km and 300 km, will Starship go grab them in their respective orbits? No. Maybe you could use some sort of grabber satellite, but that has it's own problems.
4. The solar panels and visors can't be folded in, so you would need some sort of machine for doing that, or you would need to redesign the satellites.
5. You somehow need grab all the different satellites and manage to stack them neatly into the Starship. If you don't secure them properly they will move during reentry and might damage or puncture the belly and it's heat shield leading to loss of vehicle.
6. The Starlink stack only sees vertical loads during launch, while during the belly flop the stack is only subjected to horizontal load at several Gs. Engineering for these loads would complicate matters greatly.

[Redclaws]

Just thought about this today: Once SpaceX starts launching Starlink using Starship, they could also recover the old Starlink satellites and bring them back to Earth instead of burning them up in the atmosphere. They can do this since the new satellites will need to be launched into the same orbit as the old satellites, so if you're replacing old satellites, the ship is in an ideal position to retrieve the old satellites, store them in the empty cargo bay vacated by the new satellites and land them on Earth.

The only way I see this happening is if all the satellites to be returned find their way into a single orbit in immediate proximity to each other, at say 300 km altitude. But there are still massive challenges.

1. There are significant delta-V requirements in getting all the satellites to the same place. This increases the mass of the required propellant.
2. If Starship is carrying 400 satellites, and will return 400 satellites, that means you need to do 400 orbital rendezvous. That's pretty complicated. That's after dropping off the new Starlink satellites in an orbit where they will not come in conflict with the old satellites.
3. The 400 satellites to be returned will be at end-of-life, so some may be out of fuel or have defective gyros or have similar issues. If some are stuck in all sorts of orbits between 550 km and 300 km, will Starship go grab them in their respective orbits? No. Maybe you could use some sort of grabber satellite, but that has it's own problems.
4. The solar panels and visors can't be folded in, so you would need some sort of machine for doing that, or you would need to redesign the satellites.
5. You somehow need grab all the different satellites and manage to stack them neatly into the Starship. If you don't secure them properly they will move during reentry and might damage or puncture the belly and it's heat shield leading to loss of vehicle.
6. The Starlink stack only sees vertical loads during launch, while during the belly flop the stack is only subjected to horizontal load at several Gs. Engineering for these loads would complicate matters greatly.

Yeah, all of this.  And consider Starlink deployment today: a model of simplicity.  Just unlatch it and let them all float away.  Ok now consider recovery (of hundreds!), especially knowing that a significant % have gyro or thruster issues (they are EOL after all).

I think the practicality of satellite recovery - even with starship - is based around small numbers of large, expensive satellites.  Not large flocks of smaller stuff.

[ZachF]

I suspect that there will be a major redesign of Starlink sats to take advantage of Starship. I wouldn't be surprised for them to grow to 1000-2000kg and hundreds of GBps each.

SpaceX could also sell the bus and launch as a package deal too. No Geo-sat manufacturer would be able to compete with a bus + launch to GEO for a few million.

[AC in NC]

Just thought about this today: Once SpaceX starts launching Starlink using Starship, they could also recover the old Starlink satellites and bring them back to Earth instead of burning them up in the atmosphere. They can do this since the new satellites will need to be launched into the same orbit as the old satellites, so if you're replacing old satellites, the ship is in an ideal position to retrieve the old satellites, store them in the empty cargo bay vacated by the new satellites and land them on Earth.

Why would they want to do this? Two birds with one stone:

I can't understand why this would even seem desirable.

They aren't in the same orbit.  They are in the same plane.  Those a vastly different things and it makes the retrieval of even two satellites at once immensely harder.

I can't begin to imagine an automated way to properly stow a host of Starlinks.

[Eer]

Just thought about this today: ...

...

I can't begin to imagine an automated way to properly stow a host of Starlinks.

So, you'd need a run-about - enormous delta-V space-only flat bed truck with lots of bungee cords.

Question is, how do you snare them when you get "close"?  If they're not designed to be grappled, you'll likely do damage to them as you snare, rack and stack them. Think electric-induction-motor grappling hooks designed to neatly pierce and expand so you can reel them in for storage.  Or, maybe envelop them in fishnet to be reeled in.

At that point, there's very little sense in returning them to earth, though - better to eject them (or parts of them) in retro grade orbits (as propellant mass for the run-about), so they naturally deorbit more quickly.

Science Fiction, all the way. But, recovery for return and reuse? Unlikely.  Recovery for material reuse in on-orbit manufacturing? More plausible reason (the resources are already out of the gravity well, half-way to anywhere, right?)

[klod]

I suspect that there will be a major redesign of Starlink sats to take advantage of Starship. I wouldn't be surprised for them to grow to 1000-2000kg and hundreds of GBps each.

SpaceX could also sell the bus and launch as a package deal too. No Geo-sat manufacturer would be able to compete with a bus + launch to GEO for a few million.

They have an obligation before FCC to launch a certain number of satellites by a certain date.

[DreamyPickle]

Just thought about this today: Once SpaceX starts launching Starlink using Starship, they could also recover the old Starlink satellites and bring them back to Earth instead of burning them up in the atmosphere.

This would completely destroy the objection that Starlink may cause changes to the atmosphere due to burning up old satellites during re-entry. The enemies of SpaceX will try to make a big deal out of this, better to start planning for it now.

Does satellite re-entry have a real significant impact? I've spend a some minutes googling and most of what I found is relatively fuzzy statements which indicate that burning large amounts of aluminium is "not well studied" but no quantitative measures of negative impact.

A ridiculously simple solution would be to just change the structural material to steel, this would neutralize this criticism because iron is common in meteorites. If this comes with a mass penalty then SpaceX can much more easily absorb this impact than any other constellation operators.

[su27k]

Just thought about this today: Once SpaceX starts launching Starlink using Starship, they could also recover the old Starlink satellites and bring them back to Earth instead of burning them up in the atmosphere.

This would completely destroy the objection that Starlink may cause changes to the atmosphere due to burning up old satellites during re-entry. The enemies of SpaceX will try to make a big deal out of this, better to start planning for it now.

Does satellite re-entry have a real significant impact? I've spend a some minutes googling and most of what I found is relatively fuzzy statements which indicate that burning large amounts of aluminium is "not well studied" but no quantitative measures of negative impact.

A ridiculously simple solution would be to just change the structural material to steel, this would neutralize this criticism because iron is common in meteorites. If this comes with a mass penalty then SpaceX can much more easily absorb this impact than any other constellation operators.

You're right, it's not well studied and so far all the negative impacts are just hypothesis. But this is almost worse than a real, scientifically proven impact, since a hypothesis leaves infinite room for creating FUD, the impact can be exaggerated to as big as you want. And the enemies of SpaceX would just love to stop Starlink launches so that "scientists can have time to study the effect", which of course will take decades.

I thought about changing the structural material too, but steel seems too heavy, if I'm not mistaken it would increase weight by 2x for the same strength, you leave a lot of performance on the table. And it's only a temporary reprieve at best, you can bet someone will come up with a reason that burning up tons of steel in the upper atmosphere is bad too. Is it really worth it to do this just to avoid retrieval, which doesn't add a lot of overhead and may be inevitable in the long term anyway depending on how you look at it?

If you run the numbers, assuming SpaceX takes full advantage of Starship to build out Starlink, eventually the mass flow to/from LEO will rival or even surpass the total amount of meteoroids hitting Earth, it's hard to see how this can be sustained, with any structure material, even without people who intentionally want to stop Starlink: Total tonnage of meteoroids hitting Earth is about 12kt to 15kt per year. Total # of Starlink satellites proposed is 42,000, assuming future generation satellite weights 1 metric ton (same weight as the missile warning satellite they're build for SDA), then that's 42kt total. Replacement every 5 years means 8.4kt reenter every year, or 56% ~ 70% of the natural meteoroids mass influx. If you double the satellite weight to 2 metric tons (would still allow ~60 satellites to be launched in one Starship launch), then the Starlink re-entry mass flux will surpass natural meteoroids mass influx.

Remember the entire reason Starship exists is to launch crazy amount of mass into orbit, so you have to think about how to dispose of the mass launched eventually. Retrieval is a natural extension of what Starship is already designed to do anyway, seems to me this is the best long term solution.

[aero]

I believe the Starlink satellites are about 250 pounds, about 1/8 if a metric ton. Can you justify them growing by a factor of 8 to reach a full metric ton each? If anything, I would expect them to get smaller and less massive as tech evolves.

[Robotbeat]

I believe the Starlink satellites are about 250 pounds, about 1/8 if a metric ton. Can you justify them growing by a factor of 8 to reach a full metric ton each? If anything, I would expect them to get smaller and less massive as tech evolves.

They're at least twice that weight. More like 250kg.

And no, I expect them to get more massive. The better tech will enable them to be more capable.

[Yggdrasill]

They might grow a little in size, but I expect they will for the most part just add more satellites for more capacity. Having a greater number of satellites allows for better coverage, where you can see a smaller and smaller part of the sky and still have good coverage.

[M.E.T.]

They might grow a little in size, but I expect they will for the most part just add more satellites for more capacity. Having a greater number of satellites allows for better coverage, where you can see a smaller and smaller part of the sky and still have good coverage.

From an orbital congestion point of view it would surely be better to have one large 200Gb satellite than ten 20Gb sats covering a cell on the ground below.

[aero]

I believe the Starlink satellites are about 250 pounds, about 1/8 if a metric ton. Can you justify them growing by a factor of 8 to reach a full metric ton each? If anything, I would expect them to get smaller and less massive as tech evolves.

They're at least twice that weight. More like 250kg.

And no, I expect them to get more massive. The better tech will enable them to be more capable.

My bad. F9 has grown, its listed capacity to LEO is now 50,265 (SpaceX website). That allows the current load of 60 Starlinks to mass up to 837 lbm minus the dispersing system hardware mass. I don't know if the listed capacity allows for recovery or if it is for expendable mode, and neither do I know whether or not that includes fairing mass.

In any case, your number, 250 kg per satellite seems reasonable.

[DreamyPickle]

Does satellite re-entry have a real significant impact? I've spend a some minutes googling and most of what I found is relatively fuzzy statements which indicate that burning large amounts of aluminium is "not well studied" but no quantitative measures of negative impact.

You're right, it's not well studied and so far all the negative impacts are just hypothesis. But this is almost worse than a real, scientifically proven impact, since a hypothesis leaves infinite room for creating FUD, the impact can be exaggerated to as big as you want. And the enemies of SpaceX would just love to stop Starlink launches so that "scientists can have time to study the effect", which of course will take decades.

I know that lawsuits were files for this exact purpose but it seems extremely unlikely that they will win. Historically the US has not embraced the precautionary approach to regulation and seems unlikely to start now in an area with defense implications.

Europe takes a different approach and it can be seen in areas like fracking and GMO crops but this is unlikely to impact Starlink. After all, SpaceX is an US company.