D r e a m

Thin Solid Mirrors

Using a solid mirror that is thinner than the traditional 6:1 aspect ratio will make the solid mirror equalize faster, compared to a 6:1 solid mirror of the same material. Unfortunately it comes at the expense of stiffness and the "gain" in thermal time constant is similar to comparing the difference in top speed of two turtles, relative to a Formula 1 car. A thinner solid mirror is a Band-Aid "solution" on a much, much larger problem.
* What happens to the stiffness of the solid mirror when you make it half as thick? The stiffness of the mirror becomes 4x lower than it was before; 12" diameter by 2" thick (6:1) is 4x stiffer than a 12" diameter by 1" thick (12:1).
* What happens to the stiffness of the solid mirror when you make it 1/4 as thick? The stiffness of the mirror becomes 16x lower than it was before; 12" diameter by 2" thick (6:1) is 16x stiffer than a 12" diameter by 0.5" thick (24:1).
* How do the 12" examples above compare to a 6" diameter solid mirror that is 1" thick? A 12" diameter by 2" thick (6:1) is 4x lower in stiffness than the 6" diameter by 1" thick (6:1) mirror. Both are 6:1 aspect ratio but the 12" is 4x lower in stiffness. When we compare the 12" by 0.5" thick (24:1) to the 6" diameter by 1" thick mirror we find the 12" is 64x lower in stiffness. To see charts and additional information, click here for a short paper on aspect ratio & stiffness.

Glass is like any other material, when the height is decreased, the stiffness will decrease. Whether it is a steel "I" beam, wooden floor joist or glass mirror, everything will have lower stiffness when the height is reduced.
If bending of the optical surface can't be detected in situations where the thin, solid mirror has substantially lower stiffness, is far easier to bend and engineering analysis shows it is bending in large amounts, concluding that it is not bending is wishful thinking. Common sense says a higher resolution (better) test is needed.
The "precision" of easy and cheap tests can be substantially worse than believed. Even data from a $450,000 interferometer still has to abide by the laws of mechanics. The interferometer doesn't know how to properly test the large optic but the operator needs to. An ignorant (lack of expertise) operator may subtract out more than just noise, thus leading to data with higher uncertainty.
One thing to keep in mind regarding optical testing. It is often done with the optical axis horizontal. This places the face of the mirror in shear. For a thin solid mirror it has high stiffness in shear, relative to axial stiffness. Axial being the thin solid mirror zenith-pointing (optical axis vertical). So even in situations where the optical metrology equipment used in horizontal testing is high enough in resolution, the data will not show how weak the mirror is.
At fractions of a wavelength of light, all mirrors bend. Everything from freeware FEA to full-blown custom modeling and FEM/FEA to actual telescope tests using a camera, all easily show that overly thin solid mirrors are bending by large amounts and they are even more sensitive to over-constrained conditions. The lower the stiffness of the mirror, the easier it is to over-constrain and warp the optical figure due to the mirror mount. For all optics larger than 1-2" in diameter, the mirror mount is an integral part of the optical surface, not separate from it. With proper optical metrology and a desire to learn, this is easily discovered.

Dream has tried 16.5:1 and 13.2:1 aspect ratio thin solid glass mirrors in the 16" diameter range. Even the 13.2:1 was abandoned because it was so readily distorted by the mirror mount. This was using Dream's in-house fabricated carbon fiber mirror mount, which was essentially athermal to the mirror. Using an inadequately designed (little to no real mechanical engineering and no flexures) mount made from aluminum, which is eight to hundreds of times higher in CTE than the mirror material chosen, is even worse. Dream's results are based on both testing of the full telescopes and engineering analysis. Bending was readily evident in both the physical and the virtual worlds. Believing in, something for nothing, is wishful thinking. Performance lives in the details, not in baseless and easy to discount claims.

The vast majority of optical test reports are not accounting for real-world mirror mounts, or fina-use mirror angles. In order to achieve the same performance as the 6:1 mirror requires a mirror mount of far greater complexity, especially as the mirror diameter gets larger and larger. Even when a more complex mount is created, physically mounting the low-stiffness mirror properly is an additional challenge, because the mirror's optical surface is so readily distorted.
Thin solid mirrors are much more likely to have astigmatism ground and polished into them. The largest red flag statement a buyer can ever hear is that a mirror has no errors. Without exception this cannot live in the real world and statements like this persist because buyers have not pushed for proper testing or done their own research to more deeply understand optical metrology. This leaves the buyer living on faith instead of facts and creates a market where claims abound.

The argument that these fundamental, detectable and easy to understand realities of a physical object are not happening is akin to ignoring proper optical alignment, structure stiffness and critical focus. If a difference can't be seen, the test undoubtedly does not have the accuracy that was assumed. Simultaneously the instrument is often swimming in thermal and mechanical errors. If we aren't exposed to better performance, the status quo seems fine because it hasn't been compared to something better. Intelligent buyers know there will always be a better mousetrap.

It doesn't take a world-class site to detect these types of problems, and consequently anyone can see improvements when these issues are properly addressed. This tends to turn site seeing on its head because often it was the instrument itself with the largest thermal and/or mechanical errors. Traditional opticians are married to solid glass mirrors. For them it is another sale of 167 year-old technology. Going thinner is simply a shiny bow on an old technology, which slightly improves the thermal aspect(s) while providing you with an unwanted side effect; loss of stiffness.

For Dream it has always been about improved performance, not traditional performance, because we have always made the full, high-performance instruments, not just a mirror blank or a portion of the structure. Dream has never let others stand in the way of performance gains, logic and sound engineering, which is why it has become such a vertically integrated company over the past 15 years.

Move on to the Mechanical benefits page

Q u a l i t y

S t a r t s

A t

T h e

S u b s t r a t e


White Papers



Copyright © 2003 - 2018 Dream, All Rights Reserved.