Forgot your password?
typodupeerror
Science Technology

"Perfect" Mirrors Cast For LSST 114

Posted by kdawson
from the billion-pixels-of-goodness dept.
eldavojohn writes "The Large Synoptic Survey Telescope (which was partially funded by Gates & Co.) announced a world record casting for its single-piece primary and tertiary mirror blanks, cast at the University of Arizona. From the announcement: 'The Mirror Lab team opened the furnace for a close-up look at the cooled 51,900-pound mirror blank, which consists of an outer 27.5-foot diameter (8.4-meter) primary mirror and an inner 16.5-foot (5-meter) third mirror cast in one mold. It is the first time a combined primary and tertiary mirror has been produced on such a large scale.'"
This discussion has been archived. No new comments can be posted.

"Perfect" Mirrors Cast For LSST

Comments Filter:
  • by Kagura (843695) on Tuesday September 02, 2008 @07:34PM (#24852977)
    Apparently it was so awesome, they just skipped the secondary mirror and went straight to tertiary. :)
  • by Finallyjoined!!! (1158431) on Tuesday September 02, 2008 @07:35PM (#24852995)
    Let's hope the grinding is more accurate than the Hubble mirror.

    http://www.newscientist.com/article/mg12717301.000-the-testing-error-that-led-to-hubble-mirror-fiasco-.html [newscientist.com]
    • by rhyder128k (1051042) on Tuesday September 02, 2008 @08:03PM (#24853267) Homepage

      "You're *sure* about this? The calculations have to be absolutely perfect. Even the slightes..."

      "Look, take it easy. Of course we're sure. We even got hold of latest chip from Intel. Look, Pen-tee-um. It's apparently the latest thing." [fx: combined lightning/thunder clap]

    • The LSST is a ground-based telescope, so the mount can be adjusted if it's not in the right place. Besides, the engineers are, shall we say, "sensitized" to this particular error for some reason.
    • maybe they pulled a "stargate" and the company that made it was just like, "What? This is for something in space?" Btw I feel the need to mention that if you want a really, really sweet mirror of your own, crack open a buster hard drive. Those platters are smooth to what like 10nm or something? Yeah, I look totally HD in them :D
    • by budgenator (254554) on Wednesday September 03, 2008 @06:15AM (#24856885) Journal

      the Hubble mirror is extremely accurate, unfortunately the testing mechanism, the null corrector wasn't, so the Hubble mirror was ground and figured almost perfectly wrong.

    • It will be more accurate than the Hubble mirror and is tested holographically routinely to measure every detail of the surface of the mirror. When it leaves the Lab it will be one of the best surfaces in the world.
  • by QuietLagoon (813062) on Tuesday September 02, 2008 @07:47PM (#24853105)
    ... perfection is only a limitation of the measurement process used to find flaws.
    • by kmac06 (608921)
      That's not true. In this case, you can have perfect to the atomic level. And you can even measure the surface to the atomic level. Of course, this mirror is not actually perfect.
    • by hey! (33014) on Tuesday September 02, 2008 @08:04PM (#24853275) Homepage Journal

      In optics, you get to the point where further "perfection" doesn't give you any pratical benefit. That is being "diffraction limited". Diffraction limited optics are for practical purposes as "perfect" as you can get.

      For a telescope operating through the Earth's atmosphere, you run out of marginal advantage before you reach diffraction limitation. Therefore for such a system, unless special techniques such as adaptive optics are used, practical "perfection" is considerably lower.

      I don't know much about the LSST, except that it is a fast (short focal length relative to aperture) optical system. Such systems are much more difficult to get right. Long focal lengths are much more forgiving. Therefore to reach practical perfection in such an aggressive design is quite an achievement. Of course, we aren't there yet. There's three absolutely huge surfaces to grind to very price specifications. But simply casting a blank this size is a huge technical challenge. The amount of heat energy in twenty six tons of molten glass is mind-boggling. Getting it cast into a shape that can be ground and polished into an optical mirror is an engineering tour de force in itself.

      • In optics, you get to the point where further "perfection" doesn't give you any pratical benefit.
        .

        Practical benefit is not perfection, it is good enough. You are confirming my point. :)

        • Yes... (Score:2, Funny)

          by msauve (701917)
          but the summary did say that Microsoft money (via a level of indirection) was involved, so "good enough" is in its genes.
          • Re: (Score:1, Flamebait)

            by QuietLagoon (813062)
            the summary did say that Microsoft money (via a level of indirection) was involved, so "good enough" is in its genes.
            .

            If Microsoft is involved, non-responsiveness to users, bloat, and crashing are in its genes. And don't forget DRM.

            ~~~Do you have the license to look at those stars?~~~

            ~~~You may only make one copy of those star images.~~~

        • by Trogre (513942) on Tuesday September 02, 2008 @09:09PM (#24853873) Homepage

          I'm not sure "good enough" is the right term. The point is that the mirror produces results utterly indistinguishable from a mathematically perfect surface.

          Nothing to do with settling for "good enough" which usually implies a compromise has been made somewhere.

          You could say it was practically perfect in every way. I'll go stand outside now.

          • I'll go stand outside now.
            .

            No need to stand outside. You drilled right into the crux of the matter.

            Think about it. "utterly indistinguishable" What does that really mean?

            Perhaps it means that the limits of our measurement capabilities have been reached.....

            • by Muad'Dave (255648) on Wednesday September 03, 2008 @07:45AM (#24857489) Homepage
              It's not that the surface is perfect to the limit of our ability to measure, it's that the performance of the telescope _system_ is constrained by something other than the shape of the mirror (diffraction-limited). The mirror is "utterly indistinguishable" from perfect because any more perfection out of the mirror will not increase the _system's_ performance. In other words, the telescope's performance would not be enhanced at all if the mirror were replaced with a mathematically perfect one.
          • by kmac06 (608921)
            I'm sure it's not that good. Imperfections could certainly be distinguished with an atomic force microscope or electron microscope. The point is that they could not be distinguished by the intended use of this thing.
            • by cookd (72933)

              That would be the case if we were trying to focus atomic forces or electrons. Instead, since we're trying to focus certain wavelengths of light, only differences that are detectable with those wavelengths of light will imact the result.

          • by jebrew (1101907)
            That explanation is good enough for me.
  • I'm confused- I thought mirror arrays were far superior at least in part because they don't have sagging problems and can correct on the fly for atmospheric disturbances by actuating the segments of the mirror. It certainly is a hell of a lot cheaper; U Texas did it for one third the cost [utexas.edu] of this thing, and theirs is almost a meter larger in "effective" diameter.

    In fact, there are 7 or 8 telescopes larger than this [wikipedia.org], and eleven if you widen it to "larger or equal to".

    Obviously, they wouldn't have done so

    • Roger Angel likes to make mirrors this way. If they wanted to use a segmented mirror, then it wouldn't be cast in Tucson.
    • Just wait until there are four or five of these things spread across a continent and ganged together by a computer. Bigger mirrors and more mirrors both gives the advantages of both. There has to be a first one of this size, though.

    • by againjj (1132651)
      From Wikipedia [wikipedia.org]:

      Optical interferometers are mostly seen by astronomers as very specialized instruments, capable of a very limited range of observations. It is often said that an interferometer achieves the effect of a telescope the size of the distance between the apertures; this is only true in the limited sense of angular resolution. The combined effects of limited aperture area and atmospheric turbulence generally limit interferometers to observations of comparatively bright stars and active galactic nuclei. However, they have proven useful for making very high precision measurements of simple stellar parameters such as size and position (astrometry), for imaging the nearest giant stars and probing the cores of nearby active galaxies.

      • Re:why not an array? (Score:5, Informative)

        by Nyeerrmm (940927) on Tuesday September 02, 2008 @11:17PM (#24854893)

        There's a difference between a segmented or even sparse primary mirror and an interferometer.

        A telescope with a segmented mirror works almost exactly the same way as a normal telescope, except its easier to manufacture mirrors. Of course, this is with the cost of making it harder to keep aligned, and introduce unnecessary complexity for a small mirror, but as the sizes grow it becomes more and more cost effective to segment.

        A sparse mirror with a well designed layout (say a Golay array) will be very effective also as a traditional telescope. The array is designed to gather all the spatial frequencies (think of a telescope as an analog Fourier transform) with as few elements as possible. Thus, though gathering less light, it will create an image of the same resolution. Of course less light leads to lower SNRs which can be tricky and is why you don't see too many sparse systems right now.

        An interferometer, while conceptually similar to a sparse aperture system, only measures a single frequency component at a time, by taking the light from two distant telescopes and interfering them to determine the "fringes" (Young's experiment) which measure how similar the light beams are. It is thus very precise, but also very limited. Given enough time and patience you could move the relative positions of the telescopes to fill out the Fourier transform, but this is usually not very practical given that alignments need to be maintened within 10s of nanometers.

        • Do segmented mirrors also suffer some quality loss due to diffraction effects at the mirror edges? As a mathematician I expect they do, but I don't know how much of an issue it is in real observation.
          • by Nyeerrmm (940927)

            I tend to handle diffraction in terms of Fourier optics, and the effects of the aperture shapes are defined by the Modulation Transfer Function (MTF).

            This basically acts as a multiplicative mask for the Fourier transform of the image, and can be calculated as the convolution of the aperture function with itself. When you cut out small segments it has the effect of slightly reducing the MTF at whatever corresponding spatial frequency that is. The practical effects of this are minimal, with only small (~5%?

    • Re:why not an array? (Score:5, Informative)

      by edremy (36408) on Wednesday September 03, 2008 @12:45AM (#24855357) Journal
      LSST isn't interesting because of the mirror diameter, it's interesting because of its incredibly wide field of view and amazingly fast optics. This thing has a field of view of almost 10 square degrees and can image down to ~24th magnitude every 15 seconds. Nothing else built or planned even comes close. PAN-STARRS4 will be the nearest thing to LSST and it has an etendue* that's something like 1/6th of LSST, although the PAN-STARRS people like to point out it's also something like a 5th of the cost of the LSST. (* a measure of mirror diameter*field of view. Bigger is better for survey telescopes)

      The UT system isn't even the same idea- the main mirror can't even be moved in elevation and doesn't cover the entire sky- it only sees 70% of it. Hobby-Eberly is a spectroscope, designed to look at specific targets for a long time to get the spectrum of the target. LSST is a survey telescope- it's going to scan the visible sky every 3 days in multiple wavelengths, so you have to have an entirely different grade of mount, support structure and drive system. As any amateur astronomer will tell you, cheaping out on the mount will save you quite a few bucks. :^) (Although looking over the Hobby-Eberly, they did some really neat stuff with the mount to get it to track.)

      Entirely different missions, different optics, different mounts, etc etc.

    • Re: (Score:3, Informative)

      by budgenator (254554)

      this is a fast wide-field telescope, it's designed to be wide-angled and low magnification, most other scopes are narrow-field and high magnification. This one will take pictures of the whole observable sky over and over so changes over time can be easily seen, hell they could even make time-lapse movies!

  • A mirror big enough for RMS, now if only we could get him to look into it each morning...
    • by eclectro (227083)

      A mirror big enough for RMS, now if only we could get him to look into it each morning...

      The problem may not so much be getting him to look in it in the morning, the problem may be the sun rising over his shoulder...

    • by gardyloo (512791)

      Every time I see "RMS" I think "Root Mean Square -- that can't be right!" Then I remember who we're talking about, parse the words a little differently, and think, "Actually, it's entirely apropos."

  • i am sure Billy G was glad it wasn't a giant, perfect window!
  • I wonder how much that mirror will expand and contract with temperature changes. They would surely have to factor it in. Oh well, I guess they can always cut an expansion joint across the middle if it looks like cracking. :)
    • well, a 26 ton chunk of glass doesn't tend to warm up or cool down a lot, at least not over the time span of typical daily temperature fluctuations. According to the Wiki article, they poured that thing in March, and now it has finally cooled off.
  • Why is the tertiary mirror larger than the secondary? [lsst.org] That's not like any telescope that I'm familiar with.
    • by HuguesT (84078)

      The secondary is concave, it spreads the image, it does not focus it. The tertiary's job is to focus the image onto the camera.

      Actually Meade and Celestron amateur telescopes also have a spreading secondary. It allows them to have a high F/D ratio with a short body.

      Here the whole idea is to have a low F/D, I'm not sure why they need a secondary and tertiary. I'm sort of understanding that with a primary only, the resulting image would be distorted, and that the secondary/tertiary arrangement allows for a la

  • by jriskin (132491) on Tuesday September 02, 2008 @10:35PM (#24854609) Homepage

    The digital camera in this thing generates 15TB of data a day from its 3200megapixel camera. I'm assuming it has an array of sensors, but thats still a ridiculous amazing pixel count.

    • by Muad'Dave (255648)

      The digital camera in this thing generates 15TB of data a day ...

      I hope they don't have Comcast.

    • by suggsjc (726146)
      First, you were wrong [lsst.org].

      The 30 terabytes of data obtained each night

      Obviously, this is a great achievement deserving of the /. homepage...
      However, I'm more interested in hearing about how they are going to process/archive/use that much data!

      I'll be honest and say that I'd never heard (or at least remembered) anything about the LSST, so I just did a brief lookover of their site [lsst.org] and it seems like a ridiculously cool project.

      LSST will rapidly scan the sky, charting objects that change or move

      That means it will have to store multiple versions (history) to be able to do trend analysis. So at multiple TB's of data, how e

  • Mirror and Camera (Score:5, Informative)

    by stewardwildcat (1009811) on Tuesday September 02, 2008 @10:41PM (#24854659)
    I am an astronomer at the UA and the mirror is a major feat of engineering. It will be the first telescope to have the tertiary and the primary mirror on the same piece of glass. They will have to grind both parts to be perfectly aligned (point to the same place) as well as make the transition area as small as possible. The secondary mirror is a doughnut shape that will be placed above the primary and will have the $100M camera behind it. The camera itself will be the size of a small car and will be as stated before a 3.2 Gigapixel ccd. It will have 200+ 4k by 4k CCD chips that will be read out in 2 seconds. This coupled with the fact it will image the night sky in 5 colors every week will lead to petabytes of data by the programs terminus. Its basically the coolest telescope that will ever be built. ESPECIALLY since the data is set to be public (for US residents) the moment it is processed each morning.
    • Why US residents only?
      • Because it will be funded by the United States in the end. It all basically has to deal with where the money comes from. Currently its being funded through private donations until it can get the full support of the US government.
  • by dave562 (969951) on Tuesday September 02, 2008 @10:51PM (#24854701) Journal
    ...all the ants on the planet screamed out in horror at the same time, then suddenly went silent.
  • by Anonymous Coward on Tuesday September 02, 2008 @11:09PM (#24854813)

    Casting 8-meter-class blanks simply isn't that uncommon any more. The Large Binocular Telescope has a pair of 8.4-meter primaries; Subaru has an 8.3-meter; VLT has four 8.2-meter, Gemini North and South each have an 8.1-meter. Oh, and the Giant Magellan Telescope [gmto.org] is planned to have seven 8.4-meter mirrors.

    The LSST is unusual in that its light path is more "folded", hitting 3 mirror surfaces on the way to its primary camera, which means that relatively run-of-the-mill 8-meter-class blank has to be ground pretty uniquely. (And I wish them the best of luck with the process.)

    Also, its secondary mirror is absofreakinglutely huge, at 5 meters. To put this in context, just ten years ago there was only one operational telescope in the whole world with a primary mirror larger than 5 meters.

    And f/1.25 is crazy fast, yes. The newest, fastest survey scopes out there right now are VISTA at f/3.25 and Pan-STARRS PS1 at f/4. SDSS is f/5, and VLT is f/5.5.

    So there you have it - what's really cool about LSST, from a guy who drives a boring old f/10 2.2-meter. ;)

    • Re: (Score:3, Informative)

      by Shag (3737)

      So there you have it - what's really cool about LSST, from a guy who drives a boring old f/10 2.2-meter. ;)

      ...and who absent-mindedly checked the "Post Anonymously" box for no reason. Whoops.

      • Thanks for that post, still waking up so haven't RTFA yet, but your post reminded me of this [wikipedia.org]scope which although smaller is f/1.

        Andy

        • by Shag (3737)

          VATT's primary mirror is f/1. The entire optical system is, according to the same Wikipedia page you linked to, an "Aplanatic Gregorian f/9."

          If I've read the LSST web site correctly, I believe the design calls for its entire optical system to be f/1.25.

          LSST's field of view will also be much wider than VATT. As a camera user, this seems sensible to me - my short/wide lenses are "faster" than my long zooms.

    • Is your "f" notation here the same thing as for cameras? I'm used to SLRs, where "f" denotes the f-stop, the size of the lens aperture versus the focal length, with smaller numbers meaning a wider aperture, resulting in a greatly reduced depth of field (i.e., you have to be a lot more careful about focusing correctly), but also more light coming through and therefore shorter exposure times. Is this what you mean by "fast"? And why is this important? Does it allow for imaging of darker objects?

      Curious,

  • 51,900-pound mirror blank, which consists of an outer 27.5-foot diameter (8.4-meter) primary mirror and an inner 16.5-foot (5-meter) third mirror cast in one mold

    I have no idea what that means, but I want one anyway.
  • Can someone explain why a mirror this size has to weight 25 tons? Aren't there lighter materials that could be used to support the mirrors surface?
    • by bdeclerc (129522) on Wednesday September 03, 2008 @05:41AM (#24856711) Homepage

      A telescope mirror needs a number of special properties, from rigidity and weight, but also thermal stability and the ability to polish it efficiently.

      For nearly 50 years the largest mirror was the 5 meter Hale telescope, but in the late '80s materials science and casting techniques had evolved to the point where we could reliably cast larger, lighter telescope mirrors, and computing power to the point where active suspension of thinner mirrors is possible.

      However, this doesn't mean we can create weightless mirrors, and an 8.4m mirror with a short focal length and two different surfaces still requires quite a bit of internal strength. Glass still has a higher density than water.

      An 8.4m mirror has a surface area of 220 square meters, even assuming the density of water (1000 kg/m3) 25 tons corresponds to a thickness of only about 12 cm, or less than 5 inches, which is very very thin, and as I said, glass has a density higher than water, so the actual thickness would be substantially less.

      • Re: (Score:3, Insightful)

        by mapsjanhere (1130359)
        hmm, I think you did a d^2 instead of an r^2 in your area calculations. The actual area is 55 m^2. With glass density around 2.5 g/cm^3 you get about 20 cm thickness. What's an astonishing aspect ratio, about the shape of a saucer.
        • by HuguesT (84078)

          If you look at the photo on TFA, the mirror looks honeycombed and much thicker.

        • by bdeclerc (129522)

          You're right, bit of a mess-up there. Still, 20cm thick over a diameter of 8.4 meters is still pretty thin, although it's indeed honeycombed, so the actual thickness is higher, but only partially filled.

  • I scrolled through these comments without finding a single "Real Genius" reference of question about using the mirror for making popcorn. For shame, slashdotters, for shame!

"If that makes any sense to you, you have a big problem." -- C. Durance, Computer Science 234

Working...