Heat transfer properties

Really all of the aluminum alloys are going to have about the same k.

6061 is about 20% better than 7075. Copper is about twice as good as 6061.

If you aren't heating centrally, it really doesn't matter.

6061 is much more common, the price / availability will be better.

A few observations of my own, you might know all this but still:

* To improve both strength and heat spread, use a thicker gauge plate.

* Concentrate on getting a good thermal connection between the glass plate and the aluminium too - any air in there will counteract all your hard work heating the aluminium evenly. Best option in my mind would be to have a super-flat, smooth  surface on a thick alu plate (to minimise warp), and some well-spread clamping pressure + smear of thermal grease between the glass and it. - think like you're doing a computer heatsink!

* Also, thinner glass heats up faster, and shows less temp diff between the bottom (measured) surface and top (working) surface (due to its relatively high thermal resistance and the net heat flow through the plate), but is of course more fragile. Stick to about 3mm (most plates for printers are about this size). Borosilicate is recommended as due to it's super-low thermal expansion coefficient it's far less likely to crack/distort when heated.

* Insulate the edges of the stack, I think a fair bit of heat comes out here which contributes to the cold edges.

Checked it out Ward, think I am not going as thick as he did.  I am looking at most at .25 (most likely) or .16 plate (may be easier to warp, but should heat up faster).  Using it both as a support and a heat spreader for the glass, so I don't have to worry about warping it as much.  The 18x18 inch plate is $67.00 and I can cut it to size. http://www.onlinemetals.com/merchant.cf … top_cat=60

Of course the copper would be so stylin', but at $459 per sheet, it's a wee bit expensive for my wallet.