Below are some of the most frequently asked questions about lightweight mirrors that we have heard and answered since Dream was formed in 2003.

For FAQ's related to carbon fiber, click here.

1. Q: Aren't lightweight mirrors plagued by print through?

....A: For more than a decade before founding Dream, Shane Santi had heard the same comment, that lightweight mirrors are lighter and equalize faster but they can't be polished as smooth as solid mirrors. Dream was formed in 2003 for two main reasons. One was to optimize carbon fiber for opto-mechanical structures, since it can be designed to offer superior performance (high stiffness, low mass and extremely low CTE) compared to traditional metals. The second was to create lightweight mirrors that had just as much design and engineering effort in them as precision mirror mounts.

....Shane recognized that without changing the processing of a mirror oversimplified lightweight mirror designs had only one way to reduce print through; thicken the face. This in turn hurts the final performance of the mirror; both mechanically (self-weight deflection) and thermally (thermal time constant and figure distortion due to internal temperature gradients). A thicker face also moves the CG closer to the optical surface, which is not desirable. He recognized that actual print through, textured optical surface caused by the rib structure behind the face, is something that could be analyzed in engineering and therefore could be more highly optimized. Rib heights, widths, locations, etc., can all be optimized, through many design changes and engineering analysis iterations, and reviewed prior to a mirror being produced.

....For the past 15 years this level of modern engineering has allowed Dream to produce lightweight mirror designs that are optimized to a degree never seen before, thus providing the performance Dream knew could be achieved with intelligent design & engineering. A smooth optical surface that matched solid mirrors, without the large number of drawbacks that come from a technology that dates back to the late 1660's. Consumers need to keep in mind that any industry based on nearly 350 years of using the same technology is not going to encourage you to leave it... They are slowly shifting their solid mirrors to thinner and/or zero-expansion solid mirrors. Both trade one problem for new problems that most consumers aren't aware of yet and neither come remotely close to the performance found in Dream's state of the art zeroDELTA mirrors.

....Although Dream can design a lightweight mirror based on MRF, IBF and other fairly expensive deterministic finishing processes, most designs are created based on being able to finish Dream's zeroDELTA mirrors using conventional grinding and polishing. This article shows Dream's in-house processing, using conventional pitch-polishing and a conventional overarm machine, of a 16" f1.376 mirror that was 2.5" in edge height, 9.5 pounds and had a face that was 1/8" thick. This article shows the resolution achieved with a Dream telescope that used a similar primary mirror finished inside Dream to achieve 0.83 arc-second image resolution (raw, unprocessed single images) from the center of a town in the Northeast US. Dream's founder has been studying thermal and mechanical aspects of mirrors and opto-mechanical structures for the past 25 years.

2. Q: I thought zero-expansion solid mirrors didn't have any thermal problems, since their CTE is zero?

....A: This is a common misconception that is still being stated by some opticians. The zero CTE property of the material only "takes care" of figure distortion caused by internal temperature gradients within the bulk of the mirror itself. Zero-expansion solid mirrors, by themselves, have little to no detectable thermally-driven figure distortion as temperature changes. This type of (thermally-induced) figure distortion does occur in non-zero-expansion solid and thick-featured "lightweight" mirrors, due to the internal temperature gradients within the bulk of the mirrors. This is why it is literally impossible for non-zero-expansion solid and thick-featured "lightweight" mirrors to have perfect optical surfaces when they are not fully equalized; within +0.1°C to -0.2°C of ambient temperature. Unless of course, the figure is perfect when there is a very precise and specific internal temperature gradient, at a specific ambient temperature... But as soon as the mirror is fully equalized, the figure will change and it is no longer "perfect."

....The use of aluminum or steel mirror mounts (that do not employ flexures) with zero-expansion mirrors often causes noticable (mechanically-induced) figure distortion as temperature changes, since the mirror mount is changing at a rate hundreds to thousands of times greater than the zero-expansion mirror. This trades one problem for another. A more complex and expensive problem to deal with properly when the mirror has a zero CTE.

....The problem that a zero CTE doesn't address is thermal mass, which causes degradation at the boundary layer; the incredibly sensitive area directly above the optical surface. Using a zero-expansion material is still allowing performance loss at the boundary layer due to the thermal mass of the mirror. Misinformation cannot change this fact and degradation at the boundary layer occurs long after any visual signs of thermal issues have disappeared. Out of sight out of mind does not work in this case because the problem is still hurting performance of the mirror and system, long after any easy to notice visual cues have stopped. Numerous papers over the past 40 years have quantifed these boundary layer losses and they are substantial. This paper is an excellent example of everything just mentioned. Additional papers; 2, 3, 4, 5.

3. Q: Aren't zero-expansion materials able to achieve a super-polish but other materials can't?

....A: This question relates to the smallest-scale roughness on the optical surface, which cannot be detected with Foucault, Ronchi, Caustic, etc. It cannot be detected with full-figure interferometers either. Lyot is an inexpensive visual method that can show these smallest-scale errors but because it cannot quantify them it often leads to ambiguous results. Dedicated interferometry units that look at small areas, typically a few mm to smaller than one square mm, are ideal for quantifying these smallest-scale errors. They are more sensitive than full-figure interferometers and tend to cost between $25,000-$50,000. This page shows RMS surface roughness data from a Dream finished mirror.

....Although the term super-polish is occassionally heard and used, very few seem to know its definition. A super-polish surface averages 1Å (0.1nm) or below. It is highly recommended that when you discuss RMS surface roughness you do so in Ångstroms, not in nanometers. This helps to keep RMS figure errors, discussed in nm or waves, differentiated from RMS surface roughness. RMS figure errors are the entire surface, whereas RMS surface roughness is a very small sample (or samples from different zones of the mirror) of the smallest scale errors.

....The misconception that only a zero-expansion material has achieved a super-polish seems to originate from at least one old text on optics. A generation or more ago the finest finish on glass (non-zero-expansion) might have been 10Å but over the last 20 or more years it has been finished to a super-polish. This had nothing to do with the glass itself but with advancements of chemistry and the science behind optical finishing. Dream's in-house finishing of zeroDELTA lightweight primary mirrors averages between 6-9Å (independently tested), while secondary mirrors for Dream's Cassegrain telescopes are often finished to an average of 2Å RMS surface roughness. The industry standard for vis spectrum mirrors is 20Å. Buyers need to be vigilant with all things related to optical metrology. If an optician or company states they are at or near zero for RMS surface roughness, then you are either talking to one of a very small number of companies in the world that can literally do this (companies that are not listed anywhere) or the person you are speaking with has never had their surfaces tested. Ask to see RMS surface roughness data. If they show you an interferogram, Foucaugram, Ronchigram, etc., of the whole optical surface, that is not RMS surface roughness.

"Your company does phenomenal work. There is a lot of thought and heart that goes into your products. Dream's engineering sets their lightweight mirrors apart from competitors. Your engineering goes beyond the lightweight aspect. You focus on actual performance!"

- Ted Kamprath

40 years in professional optics, using everything from $1m & $1.5m test rooms to 144" Continuous Polishers. He's spent his career using the latest in technologies, methods, materials & science to finish precision optics.

4. Q: Some of your designs have what look like insert locations. Do your mirrors require complex mirror mounts that go up inside those locations?
....A: Over the years one or two customers have used Invar flexures up inside some of our mirrors. This eliminates the need for independent lateral support and is sometimes used due to a customer's space restrictions. But even on those mirrors that have round features like this, the mirrors have all been designed so axial support can occur at the back (plano) surface of the mirror. The majority of our mirror design library was created so they can be processed with conventional equipment, materials and techniques, while also being supported in a conventional way.

5. Q: Are you bonding or fusing to make your mirrors and what are they made from?

 ....A: Frit-bonding and fusing are lower temperature processes that can cost less but they also come with baggage that Dream was not willing to live with. Even companies with 20-30 years of experience with these technologies, let alone startups, still have bond-line issues. Buyer due-diligence can go a long way in protecting the performance of the mirror and assuring that the mirror won't change over time. Prior to developing Dream's own process this topic was researched for close to a decade. Technical papers discussing frit and fused lightweight mirrors and their bond-line issues, where far less than 100% of planned bond-lines actually bonded, made Dream desire something better. Dream's zeroDELTA lightweight mirrors are done at a higher temperature and are cast. The zeroDELTA mirrors don't have to "act" or be "virtually" anything. They are one piece that will not come apart or change figure grossly today, tomorrow or ten years from now, which is why Dream developed its own cast technology, capable of producing cutting-edge, bubble-free cast lightweight mirrors.

....Dream's zeroDELTA mirrors are made from borosilicate glass, with a final CTE of 3.16ppm/°C. This is three times lower in CTE and four times less sensitive to figure distortion than plate glass, when comparing the same thickness samples. When comparing a thick-featured "lightweight" mirror made of plate glass to Dream's zeroDELTA mirrors, the difference in figure distortion is much, much larger than just 4x.

....SiC mirrors used in space have a CTE of 2.4ppm/°C. SiC mirrors work because they are extremely stiff and they have very desireable thermal properties in; thermal conductivity and heat capacity. This allows thin-featured mirrors (due to high stiffness) to be produced that equalize almost instantly, which reduces the ability of the CTE to cause figure distortion from internal temperature gradients to or below the noise floor level. This same principle is the foundation of Dream's zeroDELTA line of lightweight mirrors. It is easier to cast mirrors with few to no sub-ribs. It is easier to cast mirrors with thick ribs. It is easier to process a mirror with a thick face. But each of these three bring lower performance to the mirror in final use and all three combine to produce a mirror with noticeably worse figure performance.

6. Q: My optical system already shows a textbook (perfect) star test. Why would I need Dream's lightweight mirror(s)?

 ....A: This paper discusses key factors that must be adhered to in order to properly conduct a qualitative star test.

....There are numerous mechanical benefits to a lighter primary mirror. This page goes over those mechanical aspects but everything starts with the primary mirror. If it is 3-5x heavier than a Dream zeroDELTA mirror, then its mirror mount has to be heavier, which forces a heavier backplate, main telescope structure, larger (telescope) mount, etc. This means the system is starting with a much heavier load and that makes maintaining an optical surface to fractions of a wavelength of light, as well as holding modern optical alignment tolerances, substantially more difficult.

....There are also numerous thermal aspects of the Dream zeroDELTA mirrors that make them perform at a higher level than is possible with both solid glass and solid glass-ceramics (zero-expansion). This paper, as well as papers by others, goes into detail about the degradation that occurs in all mirrors due to thermal issues (mirror seeing). This paper highlights all of these aspects.

"Hello Shane, I can't think of anyone who has delved as deeply into the mechanics of telescopes as you have."
Dream customer

Contact Dream to discuss your project's needs or to learn about our standard lines of telescopes.


White Papers



Buy The Best

Only Cry Once