[watsisname]

?  Not sure what you mean.

Excellent post btw, Herra.

[Aardwolf]

Oh boy, another article with an extremely dumbed-down title!

[General Battuta]

Oh boy, another article with an extremely dumbed-down title!

It is a mark of your peculiar brand of intellect that you require the meaningful information in the thread to be presented to you in the title.

[Bob-san]

http://news.stanford.edu/news/2010/august/sun-082310.html

Interesting.

[Aardwolf]

Oh boy, another article with an extremely dumbed-down title!

It is a mark of your peculiar brand of intellect that you require the meaningful information in the thread to be presented to you in the title.

Hey, at least I'm not making false assumptions about what the content of the article says based on it this time! Instead, I'm just not going to bother reading it, and I'm going to be cynical about it all the while

[NGTM-1R]

other possibility, either Keck or VLT is, or was during the times of the observations, slightly out of tune.

I applaud the thought of questioning/checking the accuracy of the instruments, but first let's note the following things

- *both* telescopes gave a nonzero result.
- the observations were discontinuous and made over the course of several months.
- A detailed analysis of errors is included in the publication, as is customary.

If this finding was due to an inaccuracy of the scopes, I'd imagine we'd have noticed said inaccuracy by now from other research done with them.

It's actually pretty likely, because we HAVE noted inaccuracy in both of them. Keck's and the VLT's adaptive optics setup involves deforming the mirror to eliminate distortion and has had a number of unintentional consequences that have reduced its accuracy, focus, and overall usefulness for very deep sky objects. It's a mechanical problem and would require stripping down and rebuilding the scopes. The 36-meter scope will probably fix the problem, but until then...

[Mika]

It's actually pretty likely, because we HAVE noted inaccuracy in both of them. Keck's and the VLT's adaptive optics setup involves deforming the mirror to eliminate distortion and has had a number of unintentional consequences that have reduced its accuracy, focus, and overall usefulness for very deep sky objects. It's a mechanical problem and would require stripping down and rebuilding the scopes. The 36-meter scope will probably fix the problem, but until then...

Me being a researcher in related field in Optics, I have always been a little bit curious about seeing cancellation with adaptive optics. I know I'm nitpicking about the terms, but it doesn't remove distortion as it is pretty much constant and doesn't depend on temperature that much. It removes atmospheric seeing effect (statistically at least). But, due to mirror being supported from a couple of hundred points, it makes me wonder about the cancellation/introduction of astigmatism and perhaps coma - and the exact location of the image. Anyways if you have stuff about finding the error, I'd be delighted to read it.

Makes me wonder if the theoretical people are going a bit too fast if they published stuff of changing alpha instead of checking the telescope first.

[Mika]

And about the observed sun and radioactive decay effect: I say more people to test this stuff. If confirmed, I don't know what would be implications of that, but it would be interesting.

Actually, more people to test that changing alpha stuff also. Especially with different telescopes.

[watsisname]

Oh boy, another article with an extremely dumbed-down title!

It is a mark of your peculiar brand of intellect that you require the meaningful information in the thread to be presented to you in the title.

Hey, at least I'm not making false assumptions about what the content of the article says based on it this time! Instead, I'm just not going to bother reading it, and I'm going to be cynical about it all the while

If you can't read past the title of an article, then don't post about it.  We're having an interesting discussion here.

[achtung]

Oh boy, another article with an extremely dumbed-down title!

It is a mark of your peculiar brand of intellect that you require the meaningful information in the thread to be presented to you in the title.

Hey, at least I'm not making false assumptions about what the content of the article says based on it this time! Instead, I'm just not going to bother reading it, and I'm going to be cynical about it all the while

If you can't read past the title of an article, then don't post about it.  We're having an interesting discussion here.

He probably spends time on Slashdot.

[NGTM-1R]

Anyways if you have stuff about finding the error, I'd be delighted to read it.

I'll ask around a bit, but it's mainly what I've heard with my semi-docency at Palomar Observatory. As I gather it has to do with the fact that Keck and the VLT, which both use individually actuated and deformed 1-meter mirror sections rather than a single solid mirror, have difficulty pointing the mirror properly at the instruments with the adaptive optics engaged. Apparently there have been nights where Keck simply can't get the mirror to point at all while using the adaptive optics.

Palomar, with a single solid mirror, has it easier in getting the scope mirrors to properly point at the instruments. Why lessons from one are not applicable to the other is something you would know better than I.

[Mika]

I'll ask around a bit, but it's mainly what I've heard with my semi-docency at Palomar Observatory. As I gather it has to do with the fact that Keck and the VLT, which both use individually actuated and deformed 1-meter mirror sections rather than a single solid mirror, have difficulty pointing the mirror properly at the instruments with the adaptive optics engaged. Apparently there have been nights where Keck simply can't get the mirror to point at all while using the adaptive optics.

Palomar, with a single solid mirror, has it easier in getting the scope mirrors to properly point at the instruments. Why lessons from one are not applicable to the other is something you would know better than I.

If you have something more, I'd like to read it. I'm starting to have hutch about why does behave as it does. The telescope consists of segmented mirror parts that are planar, but small enough that the larger curved surface can be approximated with them accurately enough, isn't it? The problem with segmentation is that each mirror is free to move within the limits imposed by the support structure, and small differences in the surface normal direction might cause large deviations in the spot location after a dozen meters. With a single mirror that is deforming, the supporting structure cannot cause as large deviations as the mirror surface itself is still connected to the neighboring support points.

[Herra Tohtori]

If you have something more, I'd like to read it. I'm starting to have hutch about why does behave as it does. The telescope consists of segmented mirror parts that are planar, but small enough that the larger curved surface can be approximated with them accurately enough, isn't it?

No, the parts of the mirrors are parabolic and they are guided to direct light into a single focus, and they are individually actuated to counter small enough atmospheric refraction changes in the image during long exposure, as far as I understand the point of adaptive optics.

But I suspect the term "planar" was from a bad choice of words rather than being uninformed about the geometry of the mirror surface itself, yes?

The problem with segmentation is that each mirror is free to move within the limits imposed by the support structure, and small differences in the surface normal direction might cause large deviations in the spot location after a dozen meters. With a single mirror that is deforming, the supporting structure cannot cause as large deviations as the mirror surface itself is still connected to the neighboring support points.

Deformation, and difficulty in synching the mirrors to focus within required precision. It's a challenge from the perspective of the machinery as well, not just the deforming of the mirrors while they are being actuated continuously.

[Mika]

No, the parts of the mirrors are parabolic and they are guided to direct light into a single focus, and they are individually actuated to counter small enough atmospheric refraction changes in the image during long exposure, as far as I understand the point of adaptive optics.

But I suspect the term "planar" was from a bad choice of words rather than being uninformed about the geometry of the mirror surface itself, yes?

Actually no. I said planar, which is actually due to being uninformed about the telescope. And I have no problems about confessing that. Though I had my reasons, the first being the thought never occurred that somebody would try to do adjustments for a piece that has not only curvature, but changing curvature. Why is this a nasty thing? Principal ray path is not well defined in that case (yes you can put it in easily to a computer but to do it in real life is completely different). So it sounds like asking for trouble to me at least because then all the adjustments need to be even more accurate - and along all five or possibly even six axes! If anyone of you guys have ever tried to align a system by giving it movement along all six degrees of freedom, you know what I'm talking about. After that I made a mental note of designing optics with as little amount of adjustments as possible.

Second reason is that I recall reading about plans to make a telescope with individually adjustable planar segments and assumed this to be it, but I guess I mixed the telescopes. I guess the other one worked at different wavelength for starters.

EDIT^2: After correcting the typos, I still find typos. Bad day for writing English, or that the brain is secretly pondering something else.

[Herra Tohtori]

Due to the low ratio between their thickness and their diameter, the VLT primary mirrors will be rather flexible and sensitive to various disturbances, requiring permanent control of their optical shape.

Active optics consists in applying controlled forces to the primary mirror and in moving the secondary mirror in order to cancel out the errors. The scheme was developped by ESO for the 3.5-m New Technology Telescope (NTT) and is now applied to the VLT. The system must essentially compensate for static or slowly varying deformations such as manufacturing errors, thermal effects, low frequency components of wind buffeting, telescope inclination, ... It is also used when changing between Cassegrain and Nasmyth foci.

The mirror blanks are produced by spin-casting. The process (figure 3) starts with the casting of approximately 45 tons of glassy Zerodur into a concave mold. Thereafter the mold is transported onto a rotating platform where it is spun until solidification. When the temperature has decreased to about 800 ºC and the viscosity is such that the blank will retain its meniscus shape, it is brought into an annealing furnace where it is cooled down to room temperature in about 3 months.

So, basically they are cast into paraboloid shape meniscus to be specific) and then polished, but require support to stay in optically correct shape, which also makes the active optics possible.

So yeah there are a lot of variables, not just the active optics commands but general stability upkeep.

[Mika]

I just realized an additional thing: it is a paraboloid mirror. Had it been spherical, that would give some leeway as spherical surface can be considered degenerate for decenter and tilt, but with a paraboloid one cannot do that. The other thing is that off-axis paraboloids tend to have a rapidly decreasing imaging performance with increased field angles. While spherical surfaces don't give as good theoretical imaging performance, they tend to do a lot better tolerance wise.

I'd hate to be the guy who designs the adjustment routines for that thing.

[Herra Tohtori]

I just realized an additional thing: it is a paraboloid mirror. Had it been spherical, that would give some leeway as spherical surface can be considered degenerate for decenter and tilt, but with a paraboloid one cannot do that. The other thing is that off-axis paraboloids tend to have a rapidly decreasing imaging performance with increased field angles. While spherical surfaces don't give as good theoretical imaging performance, they tend to do a lot better tolerance wise.

I'd hate to be the guy who designs the adjustment routines for that thing.

The main mirror in VLT (well, all the four VLT units) is one piece. Keck telescopes have segmented primary mirrors. Sadly, I couldn't find as accurate descriptions about the Keck's structure and optics as I could for the VLT. Here's some stuff; http://keckobservatory.org/about/mirror/

[Mika]

The mirror blanks are produced by spin-casting. The process (figure 3) starts with the casting of approximately 45 tons of glassy Zerodur into a concave mold. Thereafter the mold is transported onto a rotating platform where it is spun until solidification. When the temperature has decreased to about 800 ºC and the viscosity is such that the blank will retain its meniscus shape, it is brought into an annealing furnace where it is cooled down to room temperature in about 3 months.

Meniscus in this context doesn't mean what you think it means (or at least I think so). Take a look at here at the picture where they show the the effect of the bending factor for a single lens, titled "types of single lens". Meniscus lens is usually considered to be a lens which has the same sign for the radius of curvature (and curvature cannot be INF) on both surfaces. I don't think they have made the bottom of the molding cup flat, but I could be wrong.

But man, talk about colossal lens. Usual diameter limit is around 2 meters, after which glass starts to lose its shape due to gravity. Astronomical projects always tend to challenge current technology, and this one is at its best. No wonder Schott is doing that!

[Mika]

The main mirror in VLT (well, all the four VLT units) is one piece. Keck telescopes have segmented primary mirrors. Sadly, I couldn't find as accurate descriptions about the Keck's structure and optics as I could for the VLT. Here's some stuff; http://keckobservatory.org/about/mirror/

   

It is surprisingly hard to find accurate descriptions of those things. VLT seems to have a rather good documentation available for everybody. So hats of to Schott. Though I would still be interested in hearing the "that didn't work" parts, which tend to teach you best. In both manufacturing and design. I wonder how they did the tolerance analysis of the design in that VLT case.

Bah, I missed a lot of Painkiller playing due to this thread. Now off to that. Can't think optics all the time, after all.