Within (structurally rigid) reason I don't think a laminated thermal-break section of 10mm thickness would compromise the frame at all, not least because the window will be supported across its width and depth within the reveal anyway. Any dense insulating material would suffice (?).

Back in the 80's I works managed a factory that helped develop the first mass-production of a combination upvc-aluminium window called 'Harmony', a system developed and manufactured by MDC (since defunct, I believe). The two materials were bonded and then a channel in the ali frame section filled with a foam material (can't remember it's name) that formed a rigid thermal-break. The closed side of that channel was then routed out to make the thermal break. They were very strong windows.

So the idea is sound.

If I was thinking of doing what you suggest I'd do some test-pieces first and the first thing I'd try would be contact adhesive, the three sections fastened together with the fixings illustrated in one of VH's recent pictures, showing roof insulation in place with screws passing through a plastic tube-like sleeve, the flange of which stopped it pulling through the insulation material.

(You've actually given me an idea for further development work on drawings for a 'ventilated glass' window. Excellent.)

Glad to be of help

Thanks, it's good to know it is or can be structurally sound. They will be fixed frames (non opening). The problem now is which glue and which insulation. Obviously the insulation needs to be thermally efficient as well as structurally competent.

The other thing is should I look at different glazing? Thickness will now be less of an issue. Is Ug 0.6 good enough? 28 mm units are better looking but maybe a 32mm unit would be better thermally.

Thanks again
Paddy

Tom, you can fix the corners by using mitred corners and slim slip-tenons (or even biscuits) and a belts-and-braces bracket at the corners on the wooden edges because they'll be going into the boxed reveals.

VH, or anyone - where do you get those insulation fixings from? (Hope my verbal description is good enough to identify them.)

That's certainly an option Jon. A problem is that there is moulding on the frame on the inside but I could get round that by using moulded beading. The frames would still need taken apart and it woul not improve the insulation. With 28mm units in a 47mm frame that gives 9mm for beading either side and 1mm for silicone. I am tending towards the split frames as I think that I may be able to improve my Uw. The problem is how to do it!

Well that points even more towards splitting the frame for depth if I understand you correctly. Are you saying the rebate depth should be 3G thickness plus thickness of 2 compressed pieces of glazing tape plus bead? And that it should be high enough to take the tape plus a few mm. I have to admit that I don't understand glazing in anything above SG where I use traditional linseed oil putty.

Joiner

I am looking to put triple glaze units direct into stone mullion frames is there any sealant you would recommend for this . I was going to use standard insulation foam to allow the window to expand.

Sorry John. No experience of doing that onto stone. The only (and it's a stretch) similar situation was single glazing into old oak mullions which was bedded-in on transluscent acrylic sealant and then puttied. 3g straight into stone mullions? The face onto the rebate isn't a problem with the bedding bead of sealant (and the unit sitting on packers to even it up), whichever way it's glazed (internally or externally), it's making it look "right" on what would otherwise be the glazing-bead side.

What I could have made clear in my rant to Tom (sorry mate) was that the way I do it avoids the interaction predicted between the sealant and the hot-melt more by accident than design, because that's the way I've always done it in ignorance of the science that says it's critical to avoid it. Back to the old "He didn't know it was impossible" scenario. The method I use (and which others who've now taken to using the method have adopted) had more to do with getting away with using as little as possible rather than avoiding contact with the unit's edges - although my method still advocates the use of "copious amounts", the squeezed excess guaranteeing a good waterproofing seal, which is the whole point.

Having said all that, in the early days I DID apply the sealant all round the unit and only stopped because I wondered if there was a less expensive way of achieving the same end than just throwing gallons of sealant at it, and those windows are amongst the ones still going strong.

Is there absolutely no way of getting even a minimal frame of sorts around those units, John? I'm always reluctant to trust the tape edges that units usually come with because I've seen systems fail by moisture weakening the tape edges, enough to allow water to find its way around and under the tape to the inside of the unit, but that was on a conservatory roof and it took ages to realise what was happening! Since when I always stipulate "no tape" on every unit I order.

So I'm out of ideas based on first- or even second-hand experience.

Sorry John.

So in yr further comments Joiner, am I right that you do in fact leave fresh air around the perimeter of the 2G/3G unit? If so, then I'm happy.

renewablejohn, are your stone mullions glaze-in or glaze-out? In other words, did the original putty splay face inward or outward? this is a situation I'll prob be facing myself shortly - I'm hoping they're glaze-in, but fear they're glaze-out, which is def more tricky.

One time, in 1G days, I fitted a thin 'frame' of aluminium odd-leg angle, made up welded at the corners, tight into the rebate of somewhat wavy/decayed Bath stone mullions, to provide a flat (and slightly widened - visibly, but didn't seem to matter) rebate 'floor' to direct-glaze fixed glass to, and also to tap and thread into for hinges for some timber opening lights incl draught strip against the flat Al face. This was glaze-out. Worked v well.

You'd actually form a lead U-channel (like a came) to hold the 2G/3G unit, and mastic the lead to the surround?

And this is how one of the sashes here looks (before touching up)...

[IMG]http://i41.tinypic.com/351fv3n.jpg[/IMG]

You'll notice the duplex bar! Haven't caved in, it's just that this window fronts the building and needs to match very closely with the upstairs window directly above it because it's "on show" from the lane.

Taking each of the elements of your query above...

"Do we agree on the 16mm rebate platform?" -- To suit the sightlines.

"How much clearance nec around outside of 2G/3G unit, inside the woodwork?" -- Normal (for me) 2mm glazing tolerance, with a strictly insisted upon absence of parallelism on the panes, failure to comply and units returned.

"What's minimum depth (in to out) reqd for the glazing bead?" -- 12mm for me. Bead cut to, say, 14mm (easier to handle), moulding applied, bead put through thicknesser to reduce to 12mm.

"Bead how fixed - pin or screw?" -- Neither. Cut to good snug fit and pressed into sealant which holds it in place.

"What to allow for nom thickness of the 2 mastic beds, inside and outside? In other words, how much deeper than the 2G/3G unit must the rebate be?" -- 2mm is fine. On the current windows with 16mm d/g/ gas-filled units, then 16 + 14 = 30, bead reduced to 12mm gives a slightly recessed finish which, with the sealant smoothed along it with a finger gives a good finish when overpainted.

After EWI and new 3G Russell windows:

Sightline can vary according to how "out of sight" you (though usually customer/planner/CO) want the spacer bar to be. Some people don't mind having them in sight, something which drives me nuts. A lot has to do with the colour of the spacer bar. In that sash of mine, although not obvious in the pic, the spacers are white and so can sit quite close to the sightline.

Gave the example yonks ago, about my son intending to fit d/g into his Listed house neighbouring the Nene Valley Railway yard (old railway worker's cottages), when the CO said she didn't mind double-glazing so long as she "couldn't see the foil", the 'platform' (new expression to me, but I gathered what you meant) was 18mm, whereas my standard is 15mm.

The duplex bars on my sashes are mitred into the top and bottom beads.

So to get the glass i need to measure the opening for the glass and add 15mm on each side for hight of rebate minus a couple of mm for clearance? If i want the spacer to disappear i can add more.

Even more food for thought - so what's really the minimum width for rebate platform, incl 2mm clearance beyond the glass unit edge, if you don't mind the spacer showing?

As tony says, we can start considering narrower spacer bars - but isn't there doubt as to their durability? They shd be better thermally.

Great idea with the lead flashing. My windows are only 12 inches wide and the plonker of a conservation officer insists that we replace softwood with oak frames with all bars being the width of the opening windows so 2 inches lost on either side so now only 8 inches of glass. I have one original leaded window lime mortared into a stone mullion so the precedent is set. Unfortunately I have 82 windows to replace 22 in just one room so oak frames are totally out of it. Getting back to the lead flashing would that create a heat bridge around the edge of the glazing unit or am I getting paranoid considering the amount of heat were already losing using single glaze.