It would be interesting to know the mass of those satellites. Could SpaceX make a stretch fairing?
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
It would be interesting to know the mass of those satellites. Could SpaceX make a stretch fairing?
dhanson865
Well-Known Member
They only have to deploy 1600 for the full network of 550km satellites that cover to 53 degrees north and south (with ground usability to about 55 degrees)
at 60 per launch that milestone is 1 launch every 2 weeks for a year (or once a week for 6 months)
The 2nd and higher layers are for covering further north and south past 55 degrees and for less optimal north south routes like sending data from New York to Capetown South Africa that wouldn't go nearly as fast on the 550km layer since it's optimized for east/west travel.
Keep in mind Elon says this will be usable when they hit about 400 satellites in the 550km layer and could get paying customers when they get to 800 satellites in that layer. The actual quote was
combined with another quote
by SpaceX vice president of satellite government affairs Patricia Cooper.
That puts the milestones even earlier. We don't need 12,000 satellites to use this network. We need hundreds of satellites and it starts the ball rolling.
Last edited:
dhanson865
Well-Known Member
More thoughts about the quote from Elon.
Minimum ground station beam angle is 40 degrees by regulation if that helps the math (keeps FCC in the US and other organizations in other countries happy).
I'm assuming that's why Elon said it'd take about 800 or so satellites before it'd be reasonably usable.
I guess we'd have to discuss what he means by minor and moderate.
60 per orbital plan means one every 6 degrees in the sky in one direction.
7 orbital planes (assuming even spacing) gets you about 51 degrees apart so loss of signal several times a day for brief periods?
13 orbital planes (assuming even spacing) gets you about 28 degrees apart so it should be continuous coverage by then. But it's 780 satellites, slightly shy of the 800.
14 orbital planes (assuming even spacing) gets you about 26 degrees apart. Gets you above 800 even if you had to ditch a few.
Still if you can put another 60 up in a weeks time no customer is going to complain about the service knowing it'll literally improve every week or two.
So what is the end goal compared to the above?
26 orbital planes gets you to about 14 degrees apart.
27 orbital planes gets you to about 13 degrees apart.
One of those two is the final goal. I'm not sure which.
Either way the roughly 800 satellites is half coverage in bandwidth but likely 100% coverage in ground area serviced.
Minimum ground station beam angle is 40 degrees by regulation if that helps the math (keeps FCC in the US and other organizations in other countries happy).
I'm assuming that's why Elon said it'd take about 800 or so satellites before it'd be reasonably usable.
I guess we'd have to discuss what he means by minor and moderate.
60 per orbital plan means one every 6 degrees in the sky in one direction.
7 orbital planes (assuming even spacing) gets you about 51 degrees apart so loss of signal several times a day for brief periods?
13 orbital planes (assuming even spacing) gets you about 28 degrees apart so it should be continuous coverage by then. But it's 780 satellites, slightly shy of the 800.
14 orbital planes (assuming even spacing) gets you about 26 degrees apart. Gets you above 800 even if you had to ditch a few.
Still if you can put another 60 up in a weeks time no customer is going to complain about the service knowing it'll literally improve every week or two.
So what is the end goal compared to the above?
26 orbital planes gets you to about 14 degrees apart.
27 orbital planes gets you to about 13 degrees apart.
One of those two is the final goal. I'm not sure which.
Either way the roughly 800 satellites is half coverage in bandwidth but likely 100% coverage in ground area serviced.
Last edited:
Interesting photo.
One thing I noticed, is the new PAF (payload adapter fitting). I believe the standard one has a maximum of 10,886 kg, and I suppose this wider lower one would allow SpaceX to increase that substantially. (Maybe also usable for Falcon Heavy?) Being lower, I guess it also helps with fitting as many satellites as possible into the usable space of the fairing.
I'm not sure exactly what the maximum payload mass is on a Falcon 9 to LEO with drone ship landing, but this must really be pushing the limits. Assuming a 34% performance hit vs expendable (same as 5.5 tons to GTO reusable vs 8.3 tons to GTO expendable), that means a maximum of around 15 tons. So each satellite should have a maximum mass of around 250 kg. That compares to around 400 kg for the development prototypes, so they've really managed to shrink them down.
It will be interesting to see whether the landing burn will use three Merlins. That would be an indicator they are really pushing up against the maximum payload mass.
Last edited:
These sats are only going to 340 miles vs the original 715 altitude versus 1,200 miles for LEO. Earth's radius is 3,960, so 4,300 miles vs 5,160, requiring 9.5% additional orbital velocity (20% more energy). Potential energy required is only 25% or so due to lower altitude. So total energy is only 30% of an LEO launch for the same mass.
Not sure on the split between first and second stage though. Other factor is circularization requirements on the second stage.
They may also trade off efficency for capacity by reducing acceleration to reduce the load on the adapter. May even need to do a longer single engine burn to shed fuel.
I'm not sure exactly what altitude the 22.8 tons to LEO are specified for, but I think it's something like a 185 km altitude at 28.5 degrees inclination. Going up to 550 km should then be a ~9% performance hit from my calculated 15 tons capability.
But it's entirely possible the satellites are released at a fairly low elliptical(?) orbit, and then have to find their way into the correct orbit using the ion thruster. And then if the ion thruster fails, they deorbit within a few days.
185 km is the baseline perigee for eccentric GTO orbits. LEO is 2,000 km (1,200 mi) and is done either as a two step or direct injection. ISS (Dragon) is 409km (254 mi) at 51.64 degrees.
https://www.spacex.com/sites/spacex/files/falcon_users_guide_0419.pdf
Look at the table on page 5/6 here: https://sma.nasa.gov/LaunchVehicle/assets/spacex-falcon-9-v1.2-data-sheet.pdf
The 22.8 tons is specified at 185 km and 28.5 degrees inclination.
Gotcha, that is more informative than their user's guide.
From this calculator, the total energy difference between the two (185km and 570km) does not seem much, Space Calc (Calculators) - Ian Mallett - 2019 .However, it only deals in net energy, chemical is much different.
I really think they want to do these on reusable F9's. Especially once (if?) they figure out how to recover the second stage and fairings, recurring launch cost is going to be pretty insignificant, and especially if the starlink flights use older used stages that have already statistically nibbled away at their useful life. While certainly surmountable by brute force, the big issue with launching more satellites at once is that you need to do more orbital plane insertions.
It will be interesting to see what the propulsion solution is, if any. Small delta-V electric systems are still going to be a recurring cost upper, mass upper, and a major and high risk element in the supply chain; a completely crazy notion would be to [again] completely eliminate propulsion and use the F9 second stage to phase each and every satellite in each and every orbit. On the flip side, a small delta-V EP solution is kind of elegant, because the satellites are going to have relatively large solar arrays anyway for when the the comms payload is turned to 11. From a power budget perspective, during initial orbit insertion operations (raising/phasing) your comms amplifiers are off, so there's plenty of available juice for an electric propulsion system.
I would gamble that its not worth the additional expense in tooling, as I'd also gamble that they're pushing the load limits of the stack already. A larger fairing will be heavier (and if its wider, more aerodynamic losses too), so its one of those deals where the bigger it gets, the worse it gets. Sorta.
For per-unit mass, traditionally when you approach a constellation like this you start with a traffic model, which backs you in to the number of orbital planes and number of satellites you need. You can then use that number to back into the number of satellites per launch, and then that becomes your per unit mass target. While it wouldn't surprise me to find that exercise was de-prioritized for Starlink in favor of a more ground up approach, I have to think that they at least referred to some mass target that is backed out of some launch vehicle performance standard. A ground up approach can totally miss the mark, because designing for mass is expensive.
The flip side of all of that is, near as I can tell, the stack of 60 is volume limited. Since it is extremely unlikely (and unlike SpaceX) to exactly hit the sweet spot of volume and mass on the first launch, it is extremely likely that the mass budget is positive.
I'm really really interested to see what the actual control solution is for these satellites. Gravity gradient would be my guess (as noted earlier), but I would also bet that other non traditional solutions are used. We talked about torque rods upthread; some smaller spacecraft actually use clever routing of internal circuitry/harnessing (I suspect Starlink has little to no round-wire harness...) as a free, if not inefficienct use of earth's magnetic field in lieu of actual torque rods. Those concepts include clever layout of the strings on the solar arrays, where presumably Starlink's [presumably] large solar arrays could provide a useful amount of magnetic torque. SpaceX could also leverage solar and/or RF pressure (probably not the latter), novel conops modes that span more than one orbit, and then also a high fidelity orbital decay model. With such a crazy amount of assets on orbit, a lot of the traditional stuff can be thrown out the window as long as you can accurately determine/predict location and attitude. You don't need to rely on those couple of satellites flying overhead pointed precisely if you have 50 satellites where maybe 1/2 or 3/4 are pointed in the right direction, so to speak.
Per previous tweets, 2nd stage will remain expendable and moist recovery of fairing is sufficient.
Elon Musk on Twitter
Elon Musk on Twitter
What's the widest PAF SpaceX has used to date? What do they use to bolt dragon onto F9? Certainly lower and fatter helps with both loads and volume...and while it does look like there are tie-downs to the top circumference of this PAF, there's always going to be a degree of sub-optimization from the square-peg-in-a-round-hole game.
Do we know the total planned capacity of the system? Is it still 5G? Do we have any insight into the traffic model and customer base, above and beyond the nebulous 'rural/underserved' and 'low-latency' markets?
dhanson865
Well-Known Member
the FAQ on reddit says
**What is the Bandwidth of the entire system?**Total available bandwidth after 12,000 satellites are in operation would be 12k*20 = 240,000 Gbps
So at 800 satellites you have 16,000 Gbps, and at 1600 satellites you have 32,000 Gbps.
if by 5G you meant LTE for phones, no this system does not communicate with cell phones. Ground stations are pizza box sized, not something you can carry around and there isn't 5G signal in the ambient just waiting to be soaked up by cell phones.
As to the market I'd say it's every single business and home user on the planet. ALL OF THEM. The goal is to have internet better than current ground services.
I thought I read somewhere that the user experience was--at least in layman's terms--equivalent to 5G.
Disclaimer: While I know a few things about satellites, I am definitely a layman when it comes to the dee-bees.
Gotcha, it's more like the currently satellite internet providers than a cell phone, but with a few differences:
Current providers are geosynchronous, high latency, with a parabolic antenna and typically limited uplink speed.
Starlink is a flat phased array antenna, low orbit, low latency, with high uplink speed. Moving sats mean more opportunities for a decent signal versus requiring a clear southern(ish) view.
In a dense region like a city, the beam covers too much area and the end terminal count would be too high for good bandwidth as a general consumer product (like cell phones at a sporting event). In the rural areas, it will rock.
Certainly that's a technological impossibility, as terrestrial solutions will always be cheaper, faster, and quicker to implement new technology.
The question is really about what traditionally terrestrial markets will be better served by a satellite network, and especially how Starlink's market will differ from the traditional satellite solutions.
At least for [so called] first world regions with strong terrestrial solutions in the population centers, most studies point toward what is basically the 'burbs as the major revenue generating market. Sure everyone in the boons wants connectivity, but the density just isn't there to make a strong business case. And typically in the dense population centers there's plenty of terrestrial solutions that blow satellites out of the water. So the studies show that the big market is people that work in the population centers but live far enough away that they're beyond the reach of the cutting edge technology.
For developing regions, cell towers are faster/ cheaper infrastructure than land lines, and Starlink (once established) will be faster/ cheaper than building cell towers.
dhanson865
Well-Known Member
No, not really.
Check out
There is no amount of money or new tech that can send data faster underground or underwater or through the air. What starlink will do is send data through space. The speed of light through vacuum is faster than the speed of light through gasses, liquids, and solids. Physics makes this unassailable by ground tech.
so in something like 20 out of 24 cases it'll be clearly better, 21 and 22 might be slightly better, 23 being about the same and 24 being worse. All in all it shrinks the world.
If you've ever played a first person shooter that has an east server and a west server in the US to keep the game playable. Imagine ping is so reasonable that you could play that same game on the US server, UK server, or Asia server and not care as much about the distance.
Now take that to streaming services, and data centers, and AWS, and google, and any global or international business you can possibly think of. If it's superior in 85-90% of the use cases every one will want to use it.
Kids, gamers, impatient people, business people, travelers, you name it. Hell, I live in a large enough city to have multiple internet options and I'm ready to switch sight unseen.
From 55 degrees north (southern Canada, central Europe) down to 55 degrees south (everything but Antarctica) it'll be competitive with any fiber based internet and will likely force fiber based internet to drop prices.
Anything worse that a fiber connection would be relegated to an even lower tier losing value / pricing power in the process. Cable modems, DSL, old school satellite internet. All that stuff gets devalued day 1 that a customer has access to Fiber and gets devalued again when they get access to Starlink.
Think about it, India, China, Australia, Philippines, South America, Central America, Africa, Middle East. It covers pretty much any place not called Scandinavia or Antarctica.
That's a huge market. Pretty much any bandwidth they put up there will sell out, guaranteed. If you don't live in Scandinavia or Antarctica you'll be seriously looking at this for your future internet.
It doesn't matter if you are in the largest city in your country or out in a field hundreds of kilometers from the next human.
It might matter if you can afford it, we don't now how pricing will play out for 3rd world countries. But for anyone that currently has anything like stable internet, this will compete from a technical standpoint if it's even half as good as I expect.
Last edited:
