Would molybdenum be a good mold for precise quartz or borosilicate mirrors? 2

[JRFerrell]

I need to make very large, accurate quartz/Borosilicate mirrors for telescopes. Grinding them precisely seems like it would be to much of a hassle to get right, so I would like to make a mold to pour-an-view. It's possible no matter what that I'll need to do SOME grinding, but I would like to keep any possible such work confined to extreme polishing.

So I need to have info on De-gassing of the molybdenum mold, it's co-efficient's of thermal expansion over temp ranges from 900°C to 2200°C (Annealing and melting temp ranges.), idea's on how to go about calculating an average life-span of the mold for this application, etc.

 

[Compositepro]

I don't want to discourage innovation but there is usually a reason that people who have been in a business for many decades do things a certain way, rather than some obviously easier way. In this case there are many reasons which you will discover as you investigate your idea. It will be educational.

 

[JRFerrell]

@Compositepro

I am aware that this may not be viable as a way of manufacturing these mirrors in a cost effective way, compared to grinding. I have a backup plan in place which would require 1 or two mirrors to be made through this process, after which I could deal with the mirrors the old fashioned way. I have looked into spin casting these mirrors. THAT process is not something I really want to deal with. It takes so much time that there is more of a chance to screw up. Compared to the cost of the old fashioned grinding method, it is also more expensive in terms of equipment, and man-power (techs for dealing with furnace AND motors). I did some research and found people were casting mirror, although not quite (sarcasm alert) to the finished quality. I would simply try to take it a step further. I am admittedly a noob though, so if you know of a punji pit in my path, please show me it's location. Why, if any reasons you know of, would this not work? Could I simply pour the molten quartz into the cooled down mold outside the oven, and then anneal the set in an oven at the required <1200°C?

 

[Compositepro]

Telescope mirrors require quarter wavelength accuracy. There are simply a multitude of reasons that molding can't get there. You can mold rough blanks but is that a significant cost saving over rough grinding? Pouring will not work at all. Look at pressed glass. Some of the issues are cte, thermal stresses, crystaline phase changes, mold accuracy, thermal distortion, thermal shock, cleanliness, surface finish degradation.

 

[JRFerrell]

@Compositepro

Last night, I thought about the "pouring idea" and yea, that doesn't seem like a good idea consider the extreme temp needed. I considered creating a chamber to hold the mold (to isolate it's environment), and heat and press glass into it with a robotic arm.

When grinding quartz, I'd probably end up using an appropriate (and not to mention expensive) grinding powder. And since I'd still need to do huge amounts of grinding, having started with no main profile curve already in place, that adds to the cost and time needed per piece.

I am already aware of the issues with the CTE obviously. The idea there was to find out the CTE's needed, and create a high purity mold that would reliably expand to the profile need. I am aware that is easier said than done, and possibly not feasible. Thermal stresses: Molybdenum does suffer thermal stresses, but it's creep factor is low, and it's made for taking heat. It's apparently relied on for it's durability in manufacturing sapphire, withstanding it's requirements there pretty well. Mold accuracy: Molybdenum is worked more easily than tungsten, though maybe not quite as easily as stainless steel. Also, I would be getting the mold laser polished (and possibly laser cut to create the parabolic profile after general traditional milling). Laser polishing can achieve accuracies down to .2µm on tool-grade steel. I am unsure yet what accuracies the company can get done for molybdenum, but it's should be significant. Again, I am aware the accuracy may not be spot on straight from the mold, but it would save time and money. surface deg: Yeah, I considered this, but Plansee (

http://www.plansee.com/en/index.htm)

provides molybdenum/tungsten crucibles with ultra smooth finishes to aid in release of Sapphire, as I mentioned before. Again, these are meant to have low creep despite the intense use (which shouldn't be anywhere near as intense as my intended application.). Crystalline phase change: Quartz is already used in manufacture of precision optics from molding processes. If others have already overcame the issues...well...I just need to find out how they did so...ohhhh goooooogle...find me tech papers pls. Even if it deviates quite a bit from it's original figure, it would still be a time saver.

Rayotek is already using similar process, but I am guessing they are not making telescope mirrors. Still, they definitely use molds for precision optics. They apparently do require extra working, but again, it's at a significantly lower cost of time, and money. I don't know...just trying to hash the stuff out. It would require huge amounts of money whatever the degree of finished quality the molds can provide, but it could allow me to at least churn out normally expensive, and LARGE blanks for the average consumer. If I can create a finished product though, that would be mint chocolate icing. Apparently laser polishing is still relatively new. I don't know enough about the process to understand why, but it seems like a common sense idea, yet it was only started around this past decade. Technology changes.
Common sense ideas sometimes take awhile to happen. I am just trying to research if this mold concept is possibly, like the laser-polishing, a late bloomer...

 

[EdStainless]

Mo is susceptible to sever oxidation, so a reducing or inert atmosphere would be needed.
If you are looking to mold blanks then what about BeNi?
Typically the molds would be warm (but not hot), a correct size piece of 'glass' is heated to its working temp, and then pressed into the mold. As soon as it has set it is removed and the mold is cooled for the next piece.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[IRstuff]

"Telescope mirrors require quarter wavelength accuracy."

That would actually be a relatively pedestrial mirror; a very good mirror would be better than 1/10th wave, since the wavefront errors from all the other optics aggregate into the total wavefront error, which must be less than 1/4 wave to get the desired resolution. For a really good telescope, every optical element should be on the order of 1/10th to 1/20th wave

The biggest issue with molding, which is why molded optics are only used on relatively poor optical quality devices, is the stress induced in the molding process. This stress MAY be relieved through annealing, but there's strong likelihood that the annealing will tweak the shape and require additional grinding to get into the right shape. Consider that even the metal coatings on a mirror can have enough stress to ruin the mirror's performance. We've done that, on long wavelength IR mirrors, no less, which should have been quite forgiving, but they were not.

Obviously, lots depend on what you mean by "accurate."

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[btrueblood]

Even if you mold in an inert gas environment, there will still be surface oxides on the molybdenum that will boil off at temperatures below the molding point of glass. Every time the mold is exposed to air (especially humid air), the oxides will form, and the rate of oxidation increases exponentially with temperature. Further, I'm not sure the glass (essentially oxides) would not react, when molten, with a reactive mold material like pure moly. An oxidation-resistant alloy might be a better choice.

Straight from:

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

"The mould material must have sufficient strength, hardness and accuracy at high temperature and pressure. Good oxidation resistance, low thermal expansion and high thermal conductivity are also required. The material of the mould has to be suitable to withstand the process temperatures without undergoing deforming processes. Therefore, the mould material choice depends critically on the transition temperature of the glass material. For low-Tg-glasses, steel moulds with a nickel alloy coating can be used. Since they cannot withstand the high temperatures required for regular optical glasses, heat-resistant materials such as carbide alloys have to be used instead in this case. In addition, mould materials include aluminium alloys, glasslike or vitreous carbon, silicon carbide, silicon nitride and a mixture of silicon carbide and carbon.[16]

A commonly used material in mould making is tungsten carbide. The mould inserts are produced by means of powder metallurgy, i.e. a sintering process followed by post-machining processes and sophisticated grinding operations. Most commonly a metallic binder (usually cobalt) is added in liquid phase sintering. In this process, the metallic binder improves the toughness of the mould as well as the sintering quality in the liquid phase to fully dense material.[17] Moulds made of hard materials have a typical lifetime of thousands of parts (size dependent) and are cost-effective for volumes of 200-1000+ (depending upon the size of the part)."

 

[JRFerrell]

@btrueblood Actually, Plansee once again comes to the rescue there. Those folks have thought of everything: They offer a zirconia based protective coating for their moly (and tungsten products as well I believe), which would help cut down on oxidation. Also, I am aware that I would need to use a inert atmosphere. I'll probably use Argon...if I can ever find a suitable mold material, gain the tech knowledge to pull it off, and actually receive the funding to carry it out...this would be the coolest thing ever... ::sigh::