looking for mechanical glass properties

[mrPelagic]

Hi
I am looking for properties of high temperature glass (borosilicate, maybe quartz).

Specifically, I am designing a pressure housing for a high temperature light. I will be using a glass hemisphere, outside pressures will reach up to about 2000 psi. The temperature differential is what concerns me. From some initial calcs, it looks like the inside surface could see up to about 300F, while the outside surface is at about 90F. at depth (2000 psi) the outside temps will be much cooler, with a corresponding reduction on the inside surface; but the differential should be similar.

I have never worked with glass before, and would appreciate any input someone may have.

 

[unclesyd]

Here is a material that might be of interest for your project.

http://www.corning.com/lightingmaterials/products/vycor.html

 

[IRstuff]

Most high temperature bulbs are quartz.

However, because of the thermal gradient, you may need something like Pyrex:

http://www.corning.com/lightingmaterials/products/pyrex.html

Their product index:

http://www.corning.com/lightingmaterials/products/index.html

TTFN

 

[NickE]

A multi-wall construction might also help you here. The inner envelope (teh evacuated one that is exposed to the high heat of the filament) could be made of Quartz, while the outside one that supports the pressure, and insulates the inner one could be made of pyrex or other material that is resistant to the stresses due to temperature and pressure.

(After I wrote this I re-read your post. It seems you already have a high temperature envelope, because the filament is much hotter than 300F)

 

[kenvlach]

It seems like there's a bit of confusion re Pyrex.

Fused silica is stronger, has greater thermal shock resistance, better optical properties, and better chemical resistance than borosilicate glass (Pyrex, Kimax).
Check the numbers:
CTE of 7.5 x 10^-7/^oC is only 1/4 that of Pyrex: 32 x 10^-7/^oC.
The only proplem with fused silica is cost and high fabrication temperature.
Pyrex as a thermally stable reflector mirror substrate became obsolete by the 1960's, when it became feasible to work large fused silica (aka fused quartz).

http://grus.berkeley.edu/~jrg/MATERIALS/node7.html

The 96% silica glasses such as Vycor are a lower cost version of pure fused silica, with nearly-as-good properties.

http://en.wikipedia.org/wiki/Fused_quartz

A very basic summary of the various glasses & their properties:

http://www.machinedesign.com/BDE/materials/bdemat7/bdemat7_3.html

Also of interest:
"Corning 7971 ultralow-expansion( ULE ) glass is fused silica doped with Ti...Results in an CTE at room temperature that is 20 times lower than that of pure fused silica." See the 8.2-meter diameter mirror substrate at

http://grus.berkeley.edu/~jrg/MATERIALS/node8.html

But, you probably don't need such an ultra-low CTE; might have problem with higher CTE mounting materials. Probably, Pyrex, Vycor, fused silica will all work if the design properly withstands the pressure, and Pyrex is easier to form. What diameter is your glass shield?


P.S In an earlier life, I quenched fused silica capsules containing samples of up to 115 atm internal pressure from 900^oC into ice-water. Capsule walls 1 mm (5 mm ID, 7 mm OD). From 800^oC, used 6x8 mm capsules. None ever broke.

 

[mrPelagic]

I don't need a low CTE. Actually, a higher one would be preferable, as the rest of my houisng is aluminum. That is not much of an issue after a bit creative mounting, however.

I think the hemisphere will be about 5 inches OD. One of the factors determining the diameter is the temperature gradient that is acceptable in borosilicate at these high temperatures. The bulb can burn at up to 1500 F, and I'm not sure how close I can get the dome to that temp.

The above numbers (300 F internal, 90F water outside) come from calcs based on a 5 inch dome. A smaller dome will be closer to the bulb, and provides less area for heat transfer, giving me a higher thermal gradient.

I am unsure of how to predict the thermal stresses due to this gradient, then incorporate them into my pressure induced stresses. Any info or ideas?

Kenvlach, your quench experience is reasuring, though!

 

[kenvlach]

A low CTE minimizes thermal stress due to temp gradient [CTE = 0, no thermal stress], although the [Δ]CTE with the aluminum housing will be huge. The inside surface of the aluminum should be mirror quality, coated with 5-8 microns of clear anodize for high IR emissivity (don't want the Al expanding too much, especially when your vessel surfaces).
Flanges, with hard rubber gaskets?
Is the housing filled with dry air or inert gas?
What wall thicknesses have you calculated for the glass?
Probably best to stick to existing designs as much as possible.

 

[mrPelagic]

good point. low cte does mean less thermal stress- my mind was on the seal interface.

the housing will be filled with dry air and dessicant. I will probobly use a high temp seal such as a fat silicone o-ring, but I have actually seen standard low temp buna seals used in a similar application.

As far as glass thickness, 1/4" wall is more than enough (something like 70,000 psi) without the thermal stress. After the thermal issue is introduced, I can only guess at this point (and reverse engineer other products To be safe, my guess is a 3/8" wall. the problem is that i'd like to engineer this, and not overbuild out of fear!