h a l f b a k e r y
Resident parking only.
add, search, annotate, link, view, overview, recent, by name, random
news, help, about, links, report a problem
or get an account
Double glazing could be made a better insulator if the
convection of the gas within the double glazing could be
The obvious way to suppress convection is to break the
into lots of smaller cells using baffles or walls. The problem
is, sticking anything inside the double
glazing cavity is going
detract from the uniform transparency of the window.
My solution is to use an array of fibres (whiskers) extending
between the panes of glass.
So I hear you immediately ask two questions: 1. why are
whiskers going to suppress convection?, 2. won't the
increase the opacity of the window?
The answer to these questions is that the whiskers are very
thin and fairly densely arrayed. I hypothesise that if the
diameter of the whiskers is less than the wavelength of light
(e.g. 200nm), the fibres will be essentially invisible in an
analogous way that 'moth-eye' anti-reflective coating is
invisible. If the whiskers are packed densely (e.g. 1µm) the
convection will be almost completely suppressed.
The whiskers could be formed by, for example, starting with
an array of liquid polymer droplets between adjacent panes
glass and drawing the panes apart.
Similar idea. However aerogel tends to be translucent rather than transparent. [xaviergisz, Jan 03 2013]
Moth-eye anti-reflective coating
[xaviergisz, Jan 03 2013]
Advanced glazing and transparent insulation
See in particular page 18 onwards. [xaviergisz, Jan 03 2013]
Development of windows based on highly insulating aerogel glazings
[xaviergisz, Jan 04 2013]
similar, but with nanofibers parallel with the window panes (my idea has the fibers extending orthogonally to the window panes). [xaviergisz, Jan 08 2013]
||Won't the whiskers conduct heat from one pane to the other?
||//I hypothesise that if the diameter of the
whiskers is less than the wavelength of light (e.g.
200nm), the fibres will be essentially invisible//
||And I hypothesize that every single one of these
fibers would shatter the moment the window was
moved, to say nothing of the effects of thermal
||Second, if the panes are properly spaced,
convection is a minimal issue, since there isn't
sufficient air space for the movement to occur.
||If the whiskers were hexagonal in cross-section and
perfectly stacked, I would't worry about breakage, but
wouldn't it be so much (relatively) simpler to just seal a
vacuum between the panes? That way you'd have zero
thermal transmission (except through the material of the
frame) and zero opacity.
||Vacuum "filled" double pane windows do exist, but
require regular pumping out (in-gassing) and can
lose the seal. Given that it is never a perfect
vacuum, you do have some conduction losses, but
they are minimized.
|| In fact window design is usually a trade off
between increasing the spacing to minimize
conduction and limiting it to get rid of
convection. This idea would allow larger spacing
without convection, but since it would produce
thermal bridges between the panes, the
advantage would be lost.
||Trust me, however, when I say that 200 nm
threads of glass, 3-6 mm long, will break under
their own weight in normal handling, so you would
end up with a typical double pane window, with a
large amount of dangerously fine glass dust in the
bottom if it ever shattered.
||Also, it would be far cheaper to simply upgrade to
triple glazed windows (which do exist) than to use
||14.7 PSI just isn't that high, and glass is fairly tough.
They aren't perfectly evacuated, but they do exist.
The real problem is that with to great a thermal
gradient the seal cracks.
||Since triple-pane windows were mentioned, suspended film windows should also be mentioned. Those often have two layers of film between the outer glass panes, breaking the cavity into 3 separate chambers to reduce convection.
||Yes, but as I said, glass is fairly tough. It's stronger
than human skin, and that can survive true vacuum,
and has a lot larger surface area.
||//human skin, and that can survive true vacuum//
||That's somewhat misleading. Humans in a vacuum do
not experience a pressure *differential* across their
||Any attempt to create such a differential (for
example, by holding one's breath before being put
under vacuum) is rather balloonishly unsuccessful.
||//Second, if the panes are properly spaced, convection is a
minimal issue, since there isn't sufficient air space for the
movement to occur.//
||I disagree. Convection appears to be a significant factor
||//Trust me, however, when I say that 200 nm threads of
glass, 3-6 mm long, will break under their own weight in
||OK, are polymer nanofibers any better?
||That article doesn't suggest that convection is a
major player, although it does suggest it is a
factor, something I never denied. You'll note that
it indicates that for a properly spaced window,
the frame material is far more of a factor than
convection. This factor would be greatly
increased by the sort of bridging fibers you
||You'll also note it indicates an optimum distance
where heat transfer is minimized which is the
point where you get the minimum conduction
without inducing significant convection (fig. 9).
At distances narrower than that point, still air
conduction dominates, which your threads would
not help. Only at distances larger than that point
does convection become a factor.
||Thus I don't deny that being able to increase the
spacing further without inducing convection
would be desirable, but given that the optimum
point is already pushing the limit of thickness for
most residential windows, there is limited benefit
to this. Even where there is a benefit it is better
served by the addition of an additional planar layer
(either triple pane, or [scad]'s film which I will
admit I had not encountered before). These
retain optical clarity and sub-convective air layers
without the cost or complexity of nanoscale
materials, and without adding thermal bridges.
||Page 18 and onward are primarily (with the
exception of the monolithic aerogel section) are
primarily focused on translucent rather than
transparent applications, and in doing so they are
looking mostly at radiative rather than convective
or conductive heat transfer.
||Well leave it o MechE to ruin it for the rest of us.
Man that is one smart dude.
If the goal is to see outside, with out letting heat
in/out then a workaround might be a 50" LED screen
powered by solar panels outside, fed images from a
camera. Screen only turns on when you are in the
room and the Blinds are open. Plus you could get Facebook updates while spying on your neighbors.
||//You'll also note it indicates an optimum distance where heat transfer is minimized which is the point where you get the minimum conduction without inducing significant convection (fig. 9). At distances narrower than that point, still air conduction dominates, which your threads would not help. Only at distances larger than that point does convection become a factor.//
||Even taking into account the thermal conduction of the whiskers themselves, I still think they are going to provide an insulative advantage at any spacing of double glazing.
||Figure 9 shows a minimum heat transfer coefficient of the gas within the double glazing to be 2 W/m²K for air or 1.2 W/m²K for krypton. Aerogel (which, because of it's air-trapping qualities, is an analogous composition to the whiskers) has a heat transfer coefficient of less than 0.7 W/m²K (see most recent link).
||Once again, that's based on sufficiently thick
window spacing that natural convection takes
over as the dominant energy transmission. Again,
I don't argue that reducing convection helps in
those ranges. Nor do I argue that if you can
reduce convection, increasing the over-all spacing
is beneficial. I do argue that the increase in
thermal transmission from glass bridging that gap is
more severe than the benefits from increased
spacing provide. I argue this especially since
adding an additional pane of material (film or glass)
will achieve the same affect without any bridging,
and with a much lower cost than a nano-
structured material. You'll note in section
18.104.22.168.3 of your earlier link, it mentions a U value
around 0.5 W/m²K for a film spaced window, an
improvement over even
||And the material structure of Aerogel is not
functionally identical to straight whiskers, because
Aerogel has a very convoluted material path,
reducing, but not eliminating, the bridging effect
as shown by the lower U value of the film
||if evacuated double glazing exists, does each glass
pane not deform tremendously? Wouldn't the window
end up being a weird biconcave lens?
||//Once again, that's based on sufficiently thick window
spacing that natural convection takes over as the dominant
||No, I was basing that on the optimal spacing such that
convection is minimised (the minimum point on the curves
in Figure 9).
||//I argue this especially since adding an additional pane of
material (film or glass) will achieve the same affect
without any bridging, and with a much lower cost than a
nano- structured material. You'll note in section 22.214.171.124.3
of your earlier link, it mentions a U value around 0.5
W/m²K for a film spaced window, an improvement over
||But that is for a double glazing that is more than twice as
thick as the aerogel comparison. It is also using layers of a
'heat mirror' (thus I'm not sure of the transparency) and
uses Xenon gas filling (which is expensive and requires
good gas sealing).
||//And the material structure of Aerogel is not functionally
identical to straight whiskers, because Aerogel has a very
convoluted material path, reducing, but not eliminating,
the bridging effect as shown by the lower U value of the
||I'm guessing that it is the nanoscale as much as the
convoluted path that reduces the conductivity of the
aerogel (since I intuit a 'convoluted material path' will not
reduce the thermal conductivity of a material at macro
||One advantage of my idea is that it doesn't require a gas
seal to trap the gas within the double glazing. This may
increase the longevity of the window, since it is my
understanding that double glazing can lose its seal integrity
after about 15 years.
||// No, I was basing that on the optimal spacing
such that convection is minimised (the minimum
point on the curves in Figure 9).//
||Sorry, the "that" in my statement was the aerogel
paper. It discusses the use of aerogel in slabs
thicker than optimum, which of course increases
insulation properties. To increase insulation in
any air filled space, you have to increase the
thickness without inducing convection. Your
original idea only provides benefits if the pane to
pane distance is increased. Aerogel allows thicker
panels without convection, and without getting
excessively heavy, hence it's advantages.
||And yes, a convoluted path will reduce thermal
conductivity in the macro scale. The critical
factor is that the thickness of a material affects its
thermal conductivity. It's logical enough that a one
meter bar of aluminum will conduct less heat from
one end to the other than a 1 cm bar. What's less
obvious is that the same
holds true (ignoring transfer through air and a few
other factors) if that 1
meter bar is bent to fit within that 1 cm space.
||And the section I referenced does not refer to
krypton fill or anything similar. It is simply a
quadruple glazed window, except the center two
panes are thin plastic film rather than glass. The
film is optically transparent, so there are no issues
there. Once again, this is simply to allow a thicker
air space without convection and without
excessive weight gain, exactly the same as your
idea. And it isn't sealed, at all. Seals are not
required for double pane windows at all, although
they are often used because they limit moisture
(they are required for gas filled, but again this
isn't). What is required is a sufficiently minimal air
exchange to avoid convection into the room, but
this won't be affected by a cracked seal.
||OK, point conceded on the conductivity of convoluted path
||//And the section I referenced does not refer to krypton fill
or anything similar.//
||I thought you were referring to the bottom entry on the
table in chapter 126.96.36.199.4 on page 16. The 0.5 W/m²K
example on page 15 is presumably thicker than 20mm so
it's not an apples to apples comparison.
||My point on that is that there's no advantage to
reducing convection unless the airspace is thicker
than the optimum point, because it's essentially a
||//[the aerogel paper] ... discusses the use of aerogel in
slabs thicker than optimum, which of course increases
||I still don't get what we're arguing about. If I wanted a
reasonably thin (15-20mm) window with the *best*
insulation properties (and perhaps with the added benefit
of good longevity), a monolithic aerogel filling would be
the best choice. If I also wanted excellent transparency,
then a variation on aeorogel (like this idea) would be an
option worth investigating.
||I'm not saying this is definitely going to work (e.g. I'm not
sure it would be more transparent than aerogel), but I don't
think its a completely absurd starting point of
||// *best* insulation properties//
||It would be vacuum, that's my point, although that
does require maintenance. Aerogel in that thickness
is not significantly better than air, and may be
||But (assuming my numbers quoted are correct) aerogel is
three times better than air (2 W/m2K vs 0.7 W/m2K).
||Okay, maybe a little, but only because convection is
starting to take over at that point. The same
thickness with a film middle pane should be more
efficient than a simple double pane. What I think
I've been neglecting is that aerogel also sharply
reduces brownian mixing, which, again, this idea
||Regardless, it's critical that your central material not
produce a bridging effect, which this idea will.
||I've come up with a small variation on the idea to overcome
the thermal bridging problem.
||Instead of the whiskers extending from one pane of glass to
the other, they could instead be extend only from one pane
only and not quite touch the other.
||The whiskers would need something to prevent them from
drooping. If the whiskers were given a negative charge the
electric fields would repel thus keeping them erect. The
opposite pane of glass could be given a positive charge to
also help pull the whiskers straight and taut. The charge
could be provided by a battery or the whiskers could be
made of an electret material.