ATM casting a 10" pyrex blank - rfc

["Andreas Derwahl" <A.Derwahl@massey.ac.nz>]

Hello everyone,

I'm a newbie to atm-ing and read this list for a couple of months until the
email load got too much to handle. Now I signed up again because I am
thinking of a way to produce a cheap blank. I have access to a lot of broken
labware glass (pyrex, duran etc. - most if not all will be borosilicate
glass) I would like to try and cast a 10" blank. I plan on building a 10" f5
dobson as a portable scope and have a very limited budget, hence the thought
of using the availiable resources (including a glassblowers oven).

I've searched the archives and found some recommendations and annealing
schemes so I came up with this plan and questions:

- mould from  1:1 plaster of paris and silica

- this I want to fill with the crushed glass. I'm not sure, though, if I
should leave the glass pieces as big as possible or crush them into smaller
pieces. I assume I will get less bubbles and a more homogenous blank if I
use larger pieces, but since the glass is mostly curved (beakers, flasks and
erlenmeyers), it might be difficult to fill the mould high enough.

- heating to 1400 to 1500 deg C and holding for 6 hours. Am I to expect
devitrification over this period?

- lowering the temperature to annealing point 565 deg C and holding for ca.
15 h (overnight). I understand the cooling rate at this stage is not crucial
(except possible devitrification)?

- lowering the temperature to 370 deg C at 0.04 deg C/min (i.e. over 3.5
days). I changed this from the recommended 0.01 deg C/min, which was said to
be a very conservative rate. Can I do this or is 0.04 deg C/min too fast (or
could I even increase the rate)?

- lowering the temperature to 320 deg C at 0.055 deg C/min (i.e. over 15 h,
overnight). Again I changed the cooling rate from 0.02 to 0.055 deg C/min.

- and finally lowering the temperature to 60 deg C at 0.14 deg C/min.

That would mean I need about a week for the whole process.

As I said, I'm a newbie and would very much like your comments on this plan.

Cheers,
Andreas Derwahl