> thermal conduction of air (or gas) is like a high resistance
Except when the space is large enough for convention to develop.
If the air-space is large, convection will carry a lot of cold down and around.
If the air-space is thin, no convection, but the thin space is lower resistance.
So there is some optimum space.
Vacuum of course kills convection transfer; but estimate the force of air-pressure. A window-like window will collapse. Bottles ("Thermos") do better but arn't windows. A lattice of spacers between the panes would support air-pressure, but also greatly increase direct loss through solid, and muss the view.
Good 3-pane is maybe twice as good as good 2-pane, so it is efficient of material.
I'm looking at some medium-good double-glass doors and windows here. I think loss through the wood/vinyl sash is a large part of loss through the glass, and loss through the door's non-glass parts is absurd. That's not even noticing that the door's factory weather stipping does NOT survive the average installation. Yes the sheet says install plumb and square, but obviously they don't expect that to happen because they left large gaps. The rubber may work in the factory on a table, but not in the house.
Except when the space is large enough for convention to develop.
If the air-space is large, convection will carry a lot of cold down and around.
If the air-space is thin, no convection, but the thin space is lower resistance.
So there is some optimum space.
Vacuum of course kills convection transfer; but estimate the force of air-pressure. A window-like window will collapse. Bottles ("Thermos") do better but arn't windows. A lattice of spacers between the panes would support air-pressure, but also greatly increase direct loss through solid, and muss the view.
Good 3-pane is maybe twice as good as good 2-pane, so it is efficient of material.
I'm looking at some medium-good double-glass doors and windows here. I think loss through the wood/vinyl sash is a large part of loss through the glass, and loss through the door's non-glass parts is absurd. That's not even noticing that the door's factory weather stipping does NOT survive the average installation. Yes the sheet says install plumb and square, but obviously they don't expect that to happen because they left large gaps. The rubber may work in the factory on a table, but not in the house.