Metals with Low Coefficient of Thermal Expansion

Hi LJ,

This seems to be a decent article on the subject:

http://www.owlnet.rice.edu/~msci301/ThermalExpansion.pdf

Mike

Mike that chart is very helpful, and in ways I had not anticipated when I posted the question.

Thanks

L.J.

John, Super Invar appears to be perfect and far better than I had any right to hope for.

I'm in touch with a vendor in Tennesee. I don't know what it costs yet but that's another issue.

Thanks for the lead.

L.J.

Talk to SCHOTT:

http://www.us.schott.com/advanced_optics/english/our_products/zerodur/zerodur.html

(but Zerodur will not be cheap)

Whoops, just noticed you said metals in the title, but in the text you said metals, perhaps - so I do not know.

The blurb says "ZERODUR^® glass ceramic can be processed using the same machines and tools as optical and technical glass (e.g. cutting, grinding and polishing)."

Somehow I don't think that accords with LJ's requirement in the OP: "... something that can be machined on a CNC Lathe ...", whether it's metal or not.

Interesting product for other apps, tho'.

In the same blurb: "Additionally, modern CNC processing equipment and a variety of grinding technologies allow the generation of complex geometries and filigree structures at the customer's request."

By the way, this is the same material that many of the biggest and best telescopes in the world are made of.

http://www.schott.com/magazine/english/archiv/download/info100/si100_07_eyes.pdf

Zerodur is not a metal, it's a member of the ceramics family.

L.J.

``need to select a substrate, metal perhaps`` is not a clear statement

Back in the 1980's we had an application requiring a thermal coefficient similar to glass for a metal component- went with a high-nickel content steel (I believe the product we used was called NiLo, but I am not absolutely certain- it was a long time ago). At any rate, the Super Invar appears to be a very similar material...

Hi, finish to optical quality is too difficult in low expansion metals.

Invar is ok but high priced (100$/Kg),

Super-Invar is better but deteriorates totally and permanent below - 20 to -40°C,

anything else is not good in TCE (above 3ppm/K).

So specify your total system: the base, the walls and where is the temperature change coming from and may be to be controlled.

Then think about glass and ceramics:

Zerodur as stated above is best (cheaper ones available too) but brittle as glass and to be machined by grinding only. (The stuff that can be machined with ordinary tools is MACOR, but this is not low TCE.)

Quartz-glass is one of the best, TCE = 0.5ppm/K and very uniform, ULE quartz-glass is still lower in TCE but non-uniform.

Cheap, sintered variety is available in many shapes above your measurements. This to be finished with a polished plate of quartz-glass to get the mirror surface, plate to be welded or glued or diffusion bonded. Maybe porosity will require filling or impregnating.

Cheapest but not so good (3 to4ppm/K) are alumina grinding wheels.

Despite high TCE - many optical objects are made from Aluminum.

RHABE

"Invar is OK but high priced (100$/Kg),"

Boy, you can that again! They quoted me a price of between 1300 and 1500 dollars for a piece 12 x 12 x 1/2".

Mitsubishi Heavy Industries has a metals division that markets a metal used in making jet turbine blades.

I'm going to poke around some more but I suspect that by the time all is said and done, I'll be using 316 Stainless or Inconel

Thanks

L.J.

If you are talking about precision optics then forget SS! There is a compilation of data about stability in the old book from

Marshall and Maringer: Dimensional Instability, Pergamon Press. and in

Batelle, Columbus (ca 1965): Stability of materials for spacecraft applications.

Heating and cooling will deteriorate geometry with cycle number, max-min-temperature and time. And TCE is near 17ppm/K - so aluminum would be much better.

If you care about gradients: fast warm up and cool down, then the situation is still worse against SS: aluminum has ten times the thermal conductivity.

Al6061 has proven its quality in many applications in optics and airbearings, get the aircraft grade. It is usually turned or milled on airbearing machines - roughness near 20nm pv, form to be discussed. Flats up to 20" no problem. Maybe some mirror blanks from military surplus (copper or copper clad for infrared reflectors) still available at lasermotion.com.

Best material with respect to relation of thermal expansion to thermal conductivity is silicon.

have success.

RHABE

You don't mention the temperature range you need to cope with but I would suggest one of the Carpenter low expansion alloys. These go down to less that 3x10-6 mm/mm/°C.

Pyrex glass has an extremely low coefficient of thermal expansion. Optical finishes can certainly be obtained since it is glass. Not sure if there is any way to machine on a CNC lathe. glass is usuall ground not machined

"Pyrex glass has an extremely low coefficient of thermal expansion."

Compared to what?

One of the biggest problems we have in optics is the amount of time it takes for the figure of a glass mirror to stablize when moved from one temperature zone to another.

Glass is horribly slow giving up its heat and it is this problem that has prompted many to make their mirrors out of borosilicate,or, if they have the money to spend, Zerodur.

The problem isn't just thermal expansion.

L.J.