Hubble Space Telescope, this is your life

April 23, 2010 4:29 PM

On April 24, 1990, the Discovery shuttle launched the Hubble Space Telescope into orbit around Earth, where it's been for 20 years. This spring, NASA has been rolling out more pretty pictures, videos and even an IMAX movie in its honor. The Hubble has contributed to hundreds of studies about our universe. As we celebrate its legacy, let's reflect on a bit on its past and future.

The Hubble was initially set to launch in October 1986, but the Challenger tragedy grounded all space missions indefinitely. When it was finally launched in spring 1990, there was a flaw in the mirror, and the Hubble sent back blurry images. Though the space telescope was built with periodic upgrades and repairs, the mistakes, a result of poor management and slightly less-than-pristine construction measurements and conditions at NASA and the mirror's producer, brought new urgency to these scheduled missions. There have been five: Sts-61, Sts-82, Sts-103, Sts-109, Sts-125.

The fifth servicing mission in May 2009 was also its last. The Hubble was originally meant to last 15 years, but repairs and replacements will let it go on to collect and record data until 2014. It will re-enter Earth sometime between 2019 and 2032.

Its successor, the James Webb Space Telescope, will launch in 2014, and will continue where Hubble left off. It is part of NASA's Origins Program, which aims to discover the formation of galaxies and stars. However, because it will be orbiting 1 million miles from Earth and won't enjoy the benefit of schedule service repairs as Hubble did, NASA has only one chance to get it right (mechanically speaking).


Some trivia:
You want to know how it works?

The Hubble, of course, is named after Edwin Hubble. But his right-hand man was Milt Humason, a self-educated person who started as a mule driver during the construction of the Mt. Wilson observatory (where the two worked). Humason then was promoted to janitor, then night assistant there. Humason helped Hubble calculate the Hubble (redshift) velocity.


Previously

posted by i8ny3x (22 comments total)

Twenty years!??!?!?!!?!?! Where has my life gone!
posted by davejay at 4:37 PM on April 23, 2010

Space Day on the blue!
posted by sciurus at 4:39 PM on April 23, 2010

Amazing, zarq-style post. Thanks very much.
posted by bearwife at 4:43 PM on April 23, 2010

I used to keep a picture of this at my desk to help remind me that nothing I was doing was that important or significant, and that I should not get too caught up in it or stressed out by it.

It did not work.

But I still loved looking at it.
posted by jeoc at 4:54 PM on April 23, 2010 [6 favorites]

I had a friend at the time who was really into the space program, and he observed, rather caustically, that if you have the budget to build the telescope, you should include enough money to test it. I don't think he realized that it WAS tested, but that they ignored the results.

I dunno, I think the penalties for something like that ought to be severe... that was such a bad screwup that it very nearly wasted all the money invested in the instrument. At the very least, being required to pay for the replacement would be appropriate.

As far as I'm concerned, the Hubble paid for itself with the Deep Field image. That graphic demonstration of the sheer size of the Universe was worth the price of admission all by itself. Two decades of science? Pure gravy.
posted by Malor at 4:56 PM on April 23, 2010

Hubble is such an amazing instrument. It's given us a way to travel our beautiful, huge universe without ever leaving the ground. I'll shed a tear or two when it plunges back to earth.
posted by contessa at 5:07 PM on April 23, 2010

It's somewhat inappropriate to call the Webb a successor to HST. One of the things that everyone loves about Hubble are the visible light photographs, and the Webb is an infrared observatory, not an optical one.
posted by CheeseDigestsAll at 5:26 PM on April 23, 2010

Most of the images we're given are spruced up, color-shifted down to where we can see them. I don't think many, if any at all, of the famous images were taken with visible light. If you were actually there looking at the same spot, it would be far less impressive.
posted by Malor at 5:28 PM on April 23, 2010

Amen to the Deep Field Image.

It can't be that unrealistic to keep the HST in stable orbit and operative for another few decades. Why not? How often would it need to be serviced, anyway?

Just feels like a shame to let some wonderful technology become atmospheric debris because it's getting a little old. Also, the spectrum of visible light might be a tiny slice, but it's still combined with others for completeness, far as I remember.
posted by Jubal Kessler at 5:57 PM on April 23, 2010

Fuck you, James Webb! Hubble RULZ!!1!!!
posted by cmoj at 6:23 PM on April 23, 2010

Worth reading just to learn about Milt Humason.

(And what's up with Edwin Hubble getting sole credit for work that relied heavily on other researchers? Henrietta Leavitt I had heard of, but finding out about Milt makes it start to look like a pattern.)
posted by richyoung at 6:26 PM on April 23, 2010 [1 favorite]

Ten Things You Don’t Know About Hubble
posted by homunculus at 6:34 PM on April 23, 2010 [3 favorites]

It's somewhat inappropriate to call the Webb a successor to HST. One of the things that everyone loves about Hubble are the visible light photographs, and the Webb is an infrared observatory, not an optical one.

Yeah, I agree, visible light is cool and I'm glad they decided to save our curious, shiny, cylindrical friend. But still, it wouldn't have been the end of the world. Ground-based telescopes have gotten a lot better at seeing through the atmosphere than they were in 1990.

Most of the images we're given are spruced up, color-shifted down to where we can see them. I don't think many, if any at all, of the famous images were taken with visible light.

It's true that virtually all of the famous images are more or less false-color (that is, the red pixels represent a color that isn't quite dead-on red, and the green pixels represent a color that is greener than the red pixels but still not quite green, etc.). However, they are mostly visible light. It can see a little bit of UV and a little bit of IR, but Hubble is still mostly a visible-light telescope.
posted by Xezlec at 7:14 PM on April 23, 2010

A few obscure facts about the Hubble's flawed mirror. Kodak made a backup mirror for the Hubble. It didn't have the flaw. (Too bad they didn't drop the original by accident). They fixed the flaw by adding a correction lens on the receiving side. All new added equipment has to have this correction added.

The 'Ten Things You Don't Know' link hints that Lockheed had experience with this type of satellite even before the Hubble was launched. Gee I wonder how that could be? Lockheed offered to do a laser test of the optics before the launch, but NASA refused, saying that the test was not precise enough to catch any problems (it would have caught the flaw).

The Hubble is not designed to withstand the intensity of radiation trapped in the Van Allen belt. Fortunately it flies at an altitude below the belt. But there is a spot called the South Atlantic Anomally where the belt dips down a bit. They shut off some of the electronics when it is flying through this region.
posted by eye of newt at 8:30 PM on April 23, 2010 [2 favorites]

My mum went to school with a lady who went on to become Jeff Hoffmann's wife. Jeff Hoffmann went up on STS-61 with his space spanner to fix the blur. I've met him a few times, and in 1993 the Hoffmann family Christmas card had a picture of Dad spacewalking with Earth in the background. I stll have it somewhere.
posted by jontyjago at 8:48 PM on April 23, 2010 [4 favorites]

Most of the images we're given are spruced up, color-shifted down to where we can see them. I don't think many, if any at all, of the famous images were taken with visible light. If you were actually there looking at the same spot, it would be far less impressive.

HST's main instrument is, in fact, a visible-light collector. It is augmented by smaller instruments that see, primarily, in the UV range. Here is a good explanation about HST's abilities and how they determine the color used in the images.
posted by Thorzdad at 4:42 AM on April 24, 2010

Google's a fan of Hubble, too.
posted by mccarty.tim at 6:54 AM on April 24, 2010 [1 favorite]

Kodak made a backup mirror for the Hubble.

Kodak? I thought the original was made by Perkin-Elmer. How did Kodak get involved?
posted by kcds at 7:23 AM on April 24, 2010

Kodak? I thought the original was made by Perkin-Elmer. How did Kodak get involved?

The lead time on mirror manufacturing was so long it was thought that if anything went wrong and they had to start over, it would be a disaster for the project. So it seemed wise to have two made right from the start, by different suppliers to try to prevent the same process error from affecting both the primary and backup mirrors.

As it turned out, this was a very good decision. Kodak's backup mirror was perfect, and a process error meant the Perkin-Elmer one wasn't. The problem was that test results showing the Perkin-Elmer mirror was made incorrectly were ignored, so they used it instead of the correctly made one that was already on hand.
posted by FishBike at 7:35 AM on April 24, 2010

They fixed the flaw by adding a correction lens on the receiving side.

The big thing here was the Perkin-Elmer made a very precise wrong mirror. It was ground to the wrong shape, but it was very precisely ground to that wrong shape.

The reason for the error -- Perkin-Elmer put the null corrector they were using together wrong, moving a lens 1.3mm away from the correct spot in the optical train. Compounding this, they analyzed the mirror with two other null correctors, which, while not capable of the accuracy of the primary, both showed the spherical aberration that was being ground into the mirror. For reasons unknown, a two-to-one vote resulted in Perkin-Elmer going with the one, and the mirror was duly ground to the wrong shape.

However, it was otherwise a very well figured mirror -- typical error was on the order of 1/20th of a wavelength, and RMS error was similarly low (showing a very well ground error)

Thus: They screwed up, but they screwed up very precisely. So, a similarly precise corrective optic could restore most of the designed performance (you still have the errors and light loss of the extra optic needed to correct the primary's spherical aberration.)

So: An insanely stupid error, but a fixable stupid error. Compare this to the Soviet era BTA-6 telescope, largest in the world from 1975 to 1993 with a 6m mirror, but flaws in the mirror made the Hale 200" scope (a 5m primary mirror) a vastly more useful telescope. Other flaws also crippled the scope -- even when the primary was replaced, the scope never reached anywhere near the performance the large mirror would seem to promise, and that's why the world stuck with 5m as a maximum mirror size until Keck lept to 10m -- by using 36 smaller mirrors, rather than one large one. We later figured out some tricks that allowed thinner mirrors, and the largest mirrors, IIRC, are the two 8.4m mirrors on the Large Binocular Telescope in Arizona, and then the 8.2m mirrors on Subaru (Hawaii) and the 4 VLT telescopes (Chile.)
posted by eriko at 10:15 AM on April 24, 2010 [2 favorites]

FishBike, Eriko - fantastic info, thanks!

I once heard that the source of Perkin-Elmer's error was that they forgot to account for the fact that the telescope would be operating outside of the atmosphere. Do either of you know whether there's any truth to that?

I also heard that in fact the Hubble's mirror could have been ground with even higher precision, but the details of the necessary manufacturing processes are classified, and their use is not authorized on unclassified (civilian) projects.
posted by kcds at 6:01 AM on April 25, 2010

I once heard that the source of Perkin-Elmer's error was that they forgot to account for the fact that the telescope would be operating outside of the atmosphere. Do either of you know whether there's any truth to that?

Nope, not true.

To check the shape of the mirror, they built a device called a reflective null corrector, which was to be far more precise than previous mirror measuring instruments. While assembling this device, there's a step where they use a laser to align a metal rod correctly. Because that rod has a curved end, there's a little end cap they put on it, with a hole in the middle, to make sure that the laser only hits the very end of the rod. The cap itself is painted black to make sure the laser doesn't reflect off it instead of the rod.

Except that some of the paint had gotten scraped off this end cap, and so the laser did reflect off it instead of the rod they were trying to position. The rod ended up 1.3mm out of position, and subsequently, so did a lens that was part of the optical assembly of the reflective null corrector. The instrument was built wrong, with parts in the wrong place by 1.3mm, and so by following the results from this instrument when shaping the mirror, it came out perfectly wrong.

They actually had to make and install some spacers that were not called for in the design of the reflective null corrector in order to make this mistake. Nobody seems to have asked why this was necessary at the time. As one of the NASA researchers later pointed out, you didn't need a micrometer to spot this error, they could have found it using a plastic ruler.

So anyway, now they have this believed-to-be wonderful measuring instrument for the mirror, which is actually producing very precise but wrong results. They also have two other instruments, a refractive null corrector and an inverse null corrector, that tell them the mirror shape is wrong. These are less precise instruments, so they dismiss the results from them, because their best instrument (they think) says it's OK.

The thinking error there is that the less precise instruments should never say the mirror is wrong when the more precise one says it's OK. It's kind of like if you had to machine a part exactly one inch long, and you measure it with a micrometer, a ruler, and a tape measure. The micrometer says it's exactly 1.000 inches, but when you check with the ruler and tape measure they both read an inch and a quarter. Sure, they don't read to a thousandth of an inch, but they're saying it's way off!

That the whole optical path of the Hubble telescope performs differently in the atmosphere vs. in a vacuum is one of the things that made it hard to do an end-to-end test before it was launched. It's not the cause of the error, but did make it harder to find the error in a way that would have been sufficiently convincing.

They'd have had to arrange for a huge vacuum chamber to do the test in, and also support the mirror so that it attains the same shape that it would in zero-G. The latter was done for the mirror by itself when checking it, using a special "metrology mount" that applied carefully designed weights to the mirror to bend it into its zero-G shape. But it's very hard to do that once it's installed in the telescope.
posted by FishBike at 6:30 AM on April 25, 2010