While some of you may yawn when I talk optics, it is what I do for a living, and actually, it is big business here. Tucson is sometimes referred to as "Optics Valley" with all of the associated industry. One of the main reasons is the University of Arizona's College of Optical Sciences, formerly known at the Optical Sciences Center (OSC). Located on the SE corner of Cherry and University, it is shown in an IR shot here. Developed in the mid-1960s, if there is an experimental application of optics, research is likely going on here. From thin film applications, optical computing, materials research, it all goes on here. And like the Mirror Lab and the National Optical Astronomy Association (NOAO) another block to the north, it is home to an precision optical shop capable of making large optics.Of the 3 optical shops within a few blocks of each other, Steward Observatory Mirror Lab has specialized in the 6.5 and 8.4 meter diameter mirrors for large telescopes. NOAO's shop, active in the late '50s and '60s for development of telescope mirrors and instrumentation for Kitt Peak National Observatory (here near Tucson) and Cerro Tololo (down in Chile), is mostly inactive now. The OSC shop is very diverse and usually finds work that no one else will bid on - deemed too difficult or impossible to build by other optics shops. I actually got my start in optics at the OSC shop back in the '80s doing precision metrology and surface generating. The multiple shops are not in competition - in actuality they cooperate and borrow equipment and supplies back and forth quite frequently.
At the moment, the highlight of the OSC optics shop is the 4.2 meter diameter mirror for the Discovery Channel Telescope. The project is a partnership between Lowell Observatory, a private observatory located in Flagstaff, Arizona and yes, the company that brings the Discovery Channel to your cable TV system. The new Observatory will be located in Happy Jack, Arizona, an excellent site near 7,800 feet elevation 45 miles SE of Flagstaff. There are a number of interesting features of the mirror, probably the biggest challenge in fabrication is that it is only 10cm (4") thick. Because the large mirror is so thin, it is flexible, so it is extremely difficult to make into a precision surface. The solution OSC uses is 100+ passive supports to overcome the flexibility. In a recent visit, these pictures show Norm working on a zone near the center of the mirror. The thin mirror and numerous supports are seen.
The latest news I've heard is that the mirror is nearing completion in the next few weeks. The image at left shows a posted map of the surface errors as of the end of June. Opticians use interferometers to measure the surface profile of mirrors. Like a carpenter uses a ruler to measure a structure, an interferometer uses wavelengths of light as a ruler to measure mirror errors. Using these devices, errors of millionths of a millimeter can be reliably obtained. One of the neat aspects of the north end of the OSC shop is that the machine is built under a test tower, so after a polishing run, the mirror is washed and dried and can be measured without moving the mirror or support system.Ok, enough about optics and telescopes - I promise our next post will NOT be about glass, astronomy or telescopes!
seriously.... this was very interesting. had no idea what these facilities did. thank you for this introduction
ReplyDeletewhat causes the cross like appearance on the interferometer image of the mirror surface in your photo. is it some kind of support structure artifact?
ReplyDeletealso, is the edge on a large mirror less important than on a smaller one
Yes, it appears to be a diffraction effect caused by a support, though I'm not sure what it is caused by (not intimately familiar with their setup).
ReplyDeleteAnd no, the edge is important on large mirrors too, because of the huge area involved - a 1" zone at the edge of an 8.4 meter mirror is equivalent to a 20" telescope, so it needs to be good pretty much to the edge.
-Dean
I thought of another cause - supports are a little unusual in test setups, but it might be a diffraction effect if the CCD detector uses lenslets in front of the sensors to increase efficiency. I've seen similar effects in some astronomical imaging with refractor systems. Note how low level they are on the scale, really just a few nanometers, so really close to the noise level...
ReplyDelete-Dean
hmmm... interesting. my first thought was that it could be an imaging artifact like cross support of secondary on a n reflector, but could not imagine the setup. having done some seriously amateur testing with my "eye" as the camera i then thought it more likely a testing support. hadn't even considered the color scale, and thus truly minimal regardless. but yeah this is really fascinating. thanks
ReplyDeleteI'm having a bad day - I was a factor of pi off in calculating area - an inch zone at the edge of an 8.4 meter mirror is equivalent to a 36" telescope...
ReplyDeleteAs for the diffraction pattern in the data, I'm now e-mailing the metrology person to get an answer that doesn't involve "I think" in it! Check back tomorrow!
-Dean
The cross like ("New Mexico" is what we affecionately call it) is actually in the surface. - the pattern moves, so it isn't as you suspected an artifact of the detection system (which was also one of my concerns as we first began to see it!).
ReplyDeleteThe structure on the surface comes from a few things: There are 120 support points pressing up from behind the surface of a thin mirror (100 mm thick, 4300 mm diameter). Working directly above these has less give than in between them, so we had slightly more removal during initial grinding of the surface (which we are correcting now in figuring the mirror). Also, there are 12 hexagonal 1.5 meter sections that have been sawed and slumped together as a sort of puzzle assembly to create the substrate; some of the features that show up on the surface are due to slight stresses at the joints of these hexagons that are relieved as we have worked the glass.
Last, the cross like structure is a remnant of the generation process, it was present when we first received the blank. At the very beginning, the blank surface was good to about 125 microns (a diameter of a good, thick human hair), with four 60 micron bumps that followed the cross pattern you see now. I'd love to say I can leave "I think" out of my answer, but unfortunately, I am not able to do so! I think this was due to a small wobble in the bearing that was used to rotate the mirror during generation. Not bad, in my estimation (0.002" over a nearly 170" part is only about 2.5 arc seconds). -Matt N (optics shop metrology guy)
thank you for this detail. definitely interesting looking at the surface of glass in such extremely small increments to change. nice work, and yes, overall, and considering what you started with, looks really impressive.
ReplyDelete