So, as further evidence that SpaceX claims of guaranteed satellite decay are bullshit, here in Jonathan Mcdowell’s Space Pages reference on failed (non-propulsive) Starlink satellites and their trajectory history: https://planet4589.org/space/stats/megacon/starbad.html
Note that the reported failed Starlink satellites at or above 500 km (Starlink 24, Starlink 48, Starlink 61, and Starlink 64 and Stalink 26 prior to what are clearly altitude-lowering maneuvers) have nearly flat orbital altitudes indicating little or no decay, and only the sats below 500 km show signficant progressive decay. Of course, we are near the solar minimum (Cycle 25 started in November 2019) and as solar activity increases the thermosphere will rise but predictions are for weak-to-moderate activity for Cycle 25. At 550 km (only approached by Starlink 24, 26, and 64) extrapolation of decay rates shows that it’ll be well beyond 25 years before the satellite orbits decay down to 500 km (in the case of Starlink 24 there is no decay evident at all over a nearly six month period).
Thanks for the good link. I’m not sure how much you can extrapolate from the current solar minimum, though. I was reading through this link:
The density difference at 550 km between solar minimum and solar max is close to two orders of magnitude, and solar max plus geomagnetic storms is over two orders of magnitude. If SpaceX’s modeling includes all this stuff, I see no reason to think that the current low rates of decay couldn’t be significantly higher as they get closer to solar max.
Starlink 24 definitely shows decay, just slow. Check the zoomed-in portion; it lost about about 2.5 km in that 6-month period. That’s small but again, there’s like two orders of magnitude difference between min and max. Even if we have a weaker solar max, say a 40x density difference, that’s still plenty to bring that 550 km down to 450 km in a couple years, after which point a fast decay is inevitable.
Starlink 24 only shows decay after the orbital lowering maneuver in late November/early December 2019. Prior to that at ~550 km it shows no decay whatsoever (actually a slight increase but I’m assuming that is just bit-toggling not real motion).
The drag caused by the thermosphere should not be confused by atmospheric drag that occurs in a continuum. In thicker atmosphere, every interaction with a particle results in that particle interacting with all of the particles around it, so there is a large amount of momentum transfer even to particles that the body does not come into direct contact with, resulting in energy-wasting form drag (low pressure region behind the body) and turbulence. ‘Drag’ at the thermosphere, however, is just direct momentum transfer from the body to a particle which has a very long mean-free length before it interacts with another particle, so drag is much less significant even in comparison to particle density. Furthermore, objects at higher orbits are actually moving slower with regard to the surrounding medium and so there is less momentum transfer per interaction. Hence, there is a dramatic difference between, say, 200 km vs 400 km, and even between 500 km vs 550 km. Nor is the expanding thermosphere due to solar activity a uniform medium; areas will expand up or not depending upon interactions with the geomagnetic field.
Looking at the small sat conference presentation you linked to, you’ll note that Slides 2, 4, 5, and 7 all show plots with altitudes less than 500 km. This is in part because the vast majority of small satellites orbit below this altitude, but also because thermospheric effects above there are not that significant. Slide 6 does show a dramatic increase in maximum density at 550 km between the minimum density at solar minimum and solar maximum, but actually only an order of magnitude between max density at solar minimum and solar maximum (unclear where their solar max data came from but we’ll take it as a good estimate for the moment). In reality, you have to account for how much time a spacecraft in a certain azimuth will spend in a high density region versus a low density one even assuming a circular orbit, or at least account for it statistically.
In any case, it is evident from the Jonathan McDowell plots that orbital decay is minimal at 550 km altitude during the solar minimum, so even an order of magnitude or two difference in mean density isn’t going to make a dramatic ramp downward during solar maximum. And having worked with spacecraft and mission requirement planning, it is generally understood that Low Earth orbits above 500 km will endure for decades or perhaps even centuries depending on the spacecraft properties and orbital azimuth, so again this fails to pass the smell test even before looking at data. I’d really like to see the details of this SpaceX analysis instead of a summary chart because I’m morally certain they made one or more assumptions that are questionable at best.
That’s fair, though the zoomed-in portion doesn’t show that area so it’s hard to tell what’s going on. Maybe I can dig up the raw data later.
Doesn’t that make things better? I’ll have to find the link again, but one paper I read said that most interactions at very low densities are specular–that is, the particle bounces off with nearly the same velocity it came in at. You end up with double the momentum transfer as compared to a normal high-density drag scenario. And they claimed a coefficient-of-drag of nearly 2.0, which is pretty bizarre but makes sense if this is what’s going on.
Sure, but I don’t see how this fails the smell test. You agree, I think, that at solar max there’s non-negligible drag. All right, say that means some satellite will take 50 years to deorbit at 550 km (or ~5 solar cycles). A satellite with 10x the ballistic coefficient should then deorbit in 5 years (or ~0.5 cycles). Though you can’t go much less than this because you might not get enough exposure to solar max.
On that we’re agreed. Really I think they have something of an obligation to present the detailed analysis. I wonder if a FOIA request would work…
Not just Starlink. There is also OneWeb, Amazon and a bunch of other projects that are going to send a bunch of satellites into orbit. There will be more and more of them. But the first satellite to remove space debris has already been launched. And similar devices are becoming very popular judging by the press releases of many space companies.
And some think that making a wooden hull for the satellite could also help solve the problem of space debris in orbit.
I assume you are referencing the demo mission by Astroscale, which uses a 175 kg satellite to magnetically link to a 17 kg “debris”. Unfortunately, their system intrinsically relies upon the target having a compatible magnetic plate which they anticipate will be installed on all satellites. This is potentially useful for retiring dead satellites but is not a general debris mitigation strategy, particularly the diverging cloud of debris resulting from a satellite collision.
I haven’t heard of using wood for satellite chassis but I have to question the utility of this. Wood has a lot of good properties but as an organic material it naturally outgasses resins (particularly adhesively bonded wood). There has been considerable work in making printed polycarbonate chassis for Cubesats and other nanosat forms but the primary rationale is being able to produce them with integrated PCBs or fluid channels to reduce weight and assembly labor. Regardless, even a piece of balsa wood moving at several thousand meters per second is going to pose a hazard to other spacecraft.
As for Starlink and similar large ‘constellations’ of smallsats and the impact upon ground-based astronomy, a recent study by the Slovak Academy of Sciences and the Consortium for Dark Sky Studies indicates that incidental backscattering of light from the proliferation of satellites in LEO could pose enough of hazard of light pollution to limit advances in ground-based telescopes even in remote, dark sky locations.
Yes, this is the mission I had in mind. If many companies have already paid attention to the problem, then there is hope that the problem will be solved. Wooden satellites were proposed by scientists from Kyoto University and the Japanese company Sumitomo Forestry.
Researchers and a private company plan to study the prospects for using wood in outer space. The first test wooden satellite can be launched into orbit as early as 2023.
The 5-foot-long composite-overwrapped pressure vessel came from the Falcon 9 rocket’s “2nd stage that did not successfully have a deorbit burn,”… The debris left a four-inch dent in the ground but no other damage has been reported
I installed my Starlink dish this weekend. Easy set up, well depending on where you have to put it. I had to drag out my climbing gear. But once installed it was pretty much plug and play.
Latency was my main issue, as I am working remotely, hooking into my work computer. It’s SO much better than my other sat internet connection which is 600-700ms. Starlink is only 30-40ms.
Heh. Yeah. But I don’t play online. Or much at all anymore. It is a relief though. I’m glad to get off the roof. It snowed a bit here this morning.
I don’t think, and haven’t really investigated how to splice that Starlink dish cable. It’s not regular coax. It has power running through it to run the motors in the dish.
I’ve been working on what to do with the 60’ of extra cable in my office. Better Homes and Gardens would not be impressed, but it’s out of my way.
Biggest difficulty was the latency for running my work (in my old office) computer. 600-700ms was a bit of a drag. Starlink is showing 30-40ms. so 20 times faster re-latency.
I really don’t stream anything (but may start). But have 100Mbs download. Which to many would be a joke. But it’s better than it was that’s for sure.
Starlink has no data caps, so that will help. I should now be able to do work Zoom meetings with video which I am not too thrilled about. But if this allows me to continue to work from home, I’m all over it.
I realise Mr Musk is evil as is anything with which he is associated, at least in the eyes of some, but SpaceX do seem to have actually given some consideration to this issue:
It’s Ethernet, and uses a proprietary power-over-Ethernet to power the dish. I don’t think it’s really intended to be spliced, but if you’re handy with a soldering iron and reasonably dexterous, you should be able to shorten it. Alternately, you could crimp on another connector where you want the end to be. IIRC, the cable has some pretty fat insulation/sheathing, but if you don’t mind it looking a little janky, you should just be able to peel it back and crimp on a connector onto the twisted pairs.
100 Mbps is no slouch. Personally, the only need I have for faster speeds is for downloading games. But that’s infrequent enough to not matter much.
How’s the reliability been? The first shell of satellites is complete in the sense that all the satellites have been launched, but I believe it’ll be a couple months before they’re all positioned.
Reliability is good. I do still get disconnected from my work computer using VMware Horizon, but I did on hughes.net too. That’s not too big of a deal, I just jump right back on. And I’ve only really had a day to test it. Neighbor has Starlink too, works remotely and loves it.