NASA didn’t design for serviceability, but rather the NRO.
There’s an old photo showing a Hubble mirror being ground by a technician in 1979. It’s a cute how-the-sausage-is-made photo, until you think about the timelines when Hubble was donated and realize it’s a photo of top secret work being done.
(1) https://en.m.wikipedia.org/wiki/File:Hubble_mirror_polishing...
You're confusing two different programmes.
Hubble was a programme dating back to the 1960s as the Large Space Telescope (LST)[0], but first given the full go ahead in 1978. It was designed and developed in the open, with the Marshall Space Flight Center as the main design location.
The donated NRO satellites were the two donated in 2012[1] and were reportedly KH-11s. Only one of those is being developed into a space telescope, the Nancy Grace Roman Space Telescope. [2]
[0] https://science.nasa.gov/mission/hubble/overview/the-history...
[1] https://en.wikipedia.org/wiki/2012_National_Reconnaissance_O...
[2] https://en.wikipedia.org/wiki/Nancy_Grace_Roman_Space_Telesc...
It's also worth adding that the Hubble mirror was made by Perkin-Elmer with a backup made by Kodak. The Kodak mirror is the one with "heritage".
Technically both Perkin-Elmer and Kodak had heritage, as Perkin-Elmer had made mirrors for the KH-9, but Kodak was making the KH-11 mirrors.
NASA initially considered a 3m main mirror for the HST, but opted for a 2.4m mirror especially to take advantage of the existing industrial base for 2.4m mirrors for the KH-11. As it was a proven industrial process that derisked the project and reduced cost compared to a 3m mirror.
Producing these large mirrors is a very specialized industry, the projects are extremely expensive, and the timelines are long. Things tend to get passed around between programs, sometimes for reasons as simple as a mirror having been completed after its original customer cancelled it. So it's hard to establish direct causality sometimes, especially since some of the "synergy" between projects was leveraged by the contractors to complete similar tasks without necessarily having direct interaction between the parent projects.
It's sort of like air surveillance radar. There's only a few companies that make them, the clients are civilian or military, but the economics mean that both the vendors and customers are incentivized to align the civilian and military systems as much as possible (including through shared-use agreements) in order to make the design and purchasing more economical. Thus the civilian designs are all modifications of the military designs and the military designs are all modifications of the civilian designs, they have essentially evolved in parallel.
I used to work on the MRO 2.4 meter telescope (at a small university observatory). Its main mirror was a Hubble design competition prototype that was transferred to the Air Force for intelligence use after the design wasn't selected. The Air Force kept it in storage for a long time and then provided it as part of their incentives to build the 2.4 meter with dual scientific and military applications. A lot of large mirrors used in astronomy and intelligence have stories like this. There's a number of 2.4 meter mirrors out there and pretty much all of them were made for either Hubble, KH-11, or as a shot at both.
Radio astronomy is no exception, there's a reason I bring up radar. Most radio astronomy observatories early on incorporated microwave electronics that were designed for weapons targeting radar---sometimes extremely directly, with military radar prototypes having been "looted for parts" by the astronomy commmunity after they were retired. Microwave electronics were very costly to design at the time and it saved everyone a lot of money to reuse everything.
That's really interesting information, thanks for sharing it.
And yes, the Hubble official history makes it clear that NASA intentionally sought to reuse industrial capacity and tooling developed for the NRO's surveillance satellite programme.
My point was only that the Hubble was not a donated KH-11, not that it didn't share the same industrial base as the NRO's satellites.
And ironically was the part that failed.
Indeed the institutional/industrial context of that photo is extremely intriguing and the stories I've heard could fill the pages of a Tom Clancy novel.
The level of secrecy at the facility was so rigorous and taken so much to heart that even years after various NRO programs were declassified (decades after program conclusion), the retired civilian personnel I've had the privilege to know were reluctant to speak openly.
I will see if I can persuade one retired engineer I know to join this discussion!
Quite, let's not bury the fact that the HST is a modified KH-11 spy satellite aimed the other way.
It is not that simple. Hubble was developed in the open by nasa civil servants who did not have access to KH-11 design. The non-optical systems are entirely different between the two platforms. Some of the contractors were shared however, and may or may not have reused IP and tooling. A lot of it is convergent evolution though. The size of the space shuttle payload bay was set by the DOD to be large enough to carry or return a KH 11 spy satellite, even though it was never used for that purpose AFAIK. The size of the Hubble was obviously the max size they could fit in the space shuttle. So they both ended up being identically sized optical telescopes, and it’s not surprising that similar design requirements end up making similar devices.
My grandfather worked on some aspect of the Hubble design (I was young; I think the timing would have put him at Lockheed M&S in Sunnyvale at that time). He never said anything about his work on it being classified. He used to say "when I retired, the stack of NDAs was this tall" and show his index finger and thumb being about 3 inches apart.
He was a very serious guy and would never talk about anything he shouldn't have. So yes, I believe the fact that he was willing to talk about Hubble meant it was done in the open. And also, he never expressed any sort of amazement that it was still up there, working. Why wouldn't it be?
Hubble was done in the open, this is an established fact. Since some of the contractors companies were the same, though usually not the same employees, it’s an open question we plebs may never know as to whether some aspects of KH-11 were declassified and shared with the workers doing Hubble.
But if we’re talking anecdotes, I worked at Lockheed as well, though not back then, and one of the stories I heard was that when the Hubble was in Sunnyvale to get tested in the vacuum chamber, one of the KH-11 workers stopped by to check it out. He spotted something that was a unique invention for the KH spy satellites and alerted the government. It took a LOT of paperwork to show that nope, it was just accidental convergent design. It was a tricky thing to sort out, but really the only way of solving that problem given the identical requirements.
> usually not the same employees, it’s an open question we plebs may never know
The impression was that my grandfather did classified work more often than not; he probably would have been one of the crossovers if it had been the case. He was exceptionally severe if any discussion came too close to whatever line existed, not that anyone else knew if or where such a line was. My mother and her siblings did not find it to be a very happy childhood, on the subject of anecdotes :). The stuff he talked about was fascinating, but you learned quickly not to ask many questions.
https://en.wikipedia.org/wiki/KH-11_KENNEN
We don't know how much of it is the same but the mirror is very similar.
> KH-11s are believed to resemble the Hubble Space Telescope in size and shape, as they were shipped in similar containers. Their length is believed to be 19.5 meters, with a diameter of up to 3 meters (120 in).[5][23] A NASA history of the Hubble,[24] in discussing the reasons for switching from a 3-meter main mirror to a 2.4-meter (94 in) design, states: "In addition, changing to a 2.4-meter mirror would lessen fabrication costs by using manufacturing technologies developed for military spy satellites".
> In January 2011, NRO donated to NASA two space Optical Telescope Assemblies with 2.4 meters (94 in) diameter primary mirrors,[53][54][55][56] similar in size to the Hubble Space Telescope, yet with steerable secondary mirrors and shorter focal length (resulting in a wider field of view). These were initially believed to be KH-11 series "extra hardware", but were later attributed to the cancelled Future Imaging Architecture program.[57] The mirrors are to be used by NASA as the primary and spare for the Roman Space Telescope.
That's not accurate. You're thinking of the Nancy Grace Roman Space Telescope [0]
[0] https://en.wikipedia.org/wiki/2012_National_Reconnaissance_O...
As someone who doesn't have a lot of knowledge here, I appreciate your corrections and sources!
And the shuttle was designed to launch and recover DoD payloads. There's a lot of "military synergy" in the early space program. No one else was interested in paying.
I have a theory that the reason they screwed up the optics initially was because they had to adapt the design from the spy satellite design.
No, the identically made mirror that kodak made in parallel was perfectly fine. the first mirror's manufacturer made a mistake
Hubble weighs 11 tons, well within the capacity to LEO of a single $70M reusable Falcon 9 launch (probably around half that cost for SpaceX).
I wonder how expensive a telescope like this would be today, both the design and actual manufacturing, and whether it would be feasible for SpaceX to shut up all the "Starlink satellites are blocking my view of the sky" complaints by launching a Hubble-equivalent space telescope (not more capable, just more modern and presumably much cheaper) and then giving out observation time on it.
Of course, it would not be easily maintainable given we no longer have the space shuttle, but if the majority of the cost was development, the manufacturing + launch costs today might make replacement cheaper than on-orbit repairs.
https://www.space.com/15892-hubble-space-telescope.html claims "Getting Hubble developed and launched cost $1.5 billion".
There's a lot of limitations with an on orbit satellite you don't run into with ground based systems. Hubble is pretty old and it's mirror pretty small at this point so there are hard limits on it's usefulness for viewing very faint objects, it's mirror simply doesn't gather as much light because it's relatively small at just under 8 feet. Another thing you can't do anywhere nearly as easily are upgrades or adding new experiments. With ground based telescopes it's relatively easy to add new experiments on the side mirror to measure new spectra or measure in a different way. That just can't be done with a space based telescope, even if the design allowed for it the costs sky rocket.
It doesn't gather as much light as ground based telescopes, but the background (for example, from airglow) is also much lower, so the SNR is pretty good.
It's not just people trying to use telescopes to do science, it's also people like me trying to do astrophotography. Time on a LEO telescope, while cool, would not help my photography.
As an astrophotographer myself, Starlink has caused me 0 issues. I'm not sure if that's what you're trying to imply. Airplanes are the biggest issue but are easily edited out.
I've written here before with some back of the envelope numbers for a hubble 2.0 + starship as the LV. It gets more interesting if you set up an assembly line for cameras in a tube like the hubble to spread the costs out, and use newer guts for the sensor payloads and on board compute. Make some assumptions about the telescopes, such as being okay to fail since we have so many up tWebb. You could even do experiments in computational photography like the Keck or the VLA to overcome the light collection limits of the mirror size. I forget the numbers I came up with but it's a lot of orbital sensors for the cost of one hubble program and like an entire space station with sensors for the James webb.
> shut up all the "Starlink satellites are blocking my view of the sky" complaints
These complaints make a lot more sense when you project forward by just a few decades. We're talking about _one_ provider already causing minor hassles. What do you think an entire commercialized segment is going to do?
Maybe we can shut up all the "LEO satellites are the best way to build a communications network" people by building out a reasonable and fair network on the surface of the earth first.
We know how much new Hubble would cost. The Roman telescope is based on donated donated spy satellite of similar size and design. It will cost estimated $3.2 billion and that doesn't include the mirror and structure.
New Hubble telescope wouldn't solve the Starlink problem. There are lots of smaller ground telescopes that have view blocked. We would need to launch a bunch of smaller, like 1m ones, to compensate. Also, there are a lot of larger telescopes that are only feasible on Earth. Good example is the 8.4m Rubin survey telescope which is only going to cost $700 million but would be lots of billions in space. That might fit in Starship but the Extremely Large Telescope won't.
Reminded me of this, not sure if anything came out of it.
US military gives NASA two better-than-Hubble telescopes (2012): https://www.theregister.com/2012/06/05/military_gives_telesc...
A democratic space telescope could be an interesting experiment. Imagine if paying $5/month got you access to all the telescope's observations, and also the ability to vote on where it points next.
Now that we don't have the Space Shuttle maintenance would be trivial. There would be no more involvement of the manned space flight division of NASA at all. The telescope designers would just need to launch something that actually worked, since there is no longer a need to justify a manned spaceflight program.
I’d rather just de-orbit every StarLink satellite, stick next gen data satellites in geostationary orbits, and outlaw more any more space trash at that scale in LEO, full-stop.
Latency's not going to stop being a problem for geostationary orbits, which is why lower altitudes get used.
> I’d rather just de-orbit every StarLink satellite, stick next gen data satellites in geostationary orbits, and outlaw more any more space trash at that scale in LEO, full-stop.
Flip side is that stuff in LEO deorbits pretty quicky, while geostationary junk doesn't deorbit in any timeframes humans care about.
It's a good thing we don't have even more ignorant tyrants in power then
Well, the smart tyrants are certainly making productive use of StarLink, to be sure.
This reminds me again of the tremendous opportunity of making a spacesuit with a five finger articulated 'hand' at the end of the sleeve. Imagine an astronaut who puts their hand into a controller at the end of the sleeve that wraps around their fingers and hand such that their hand movements are exactly replicated on the robotic 'hand' outside the end of the glove. This would revolutionize what could be done on spacewalks.
At one of the NASA tech days at NASA Ames (Moffet field), they had a space suit glove simulator where you put your hand in it and it had the flexibility that you would expect with a partial pressure spacesuit glove in vacuum. It was super hard to do anything precisely when pushing so hard just to move your fingers around. A robotic 'waldo' type device which replicated your hand movement precisely would minimize hand fatigue and allow for doing precise alignment.
Why not just use your hand through flexible gloves, like in the old days?
Nice, but maybe it would have been better to launch a rocket a day to keep the cost away.
