[Greenbuilding] windows - flush vs recessed

[John Salmen]

I am not crazy about that detail as the extension box should really be
considered as part of the window rather than as part of the structure. 

I find people get confused in working out sealing and flashing details by
thinking of the opening and the window as one detail when they are really
two details. The framed opening (or any penetration) needs to be wrapped and
sealed with both a sloped sill and an integrated head flashing as part of
the exterior wrb irrespective of the window - that way the framing elements
are completely sealed and the window floats in the assembly. The window
should not be thought of as forming part of that layer. 

If exterior insulation is used and the window needs additional support then
that should be either attached to the window and sealed to the building as
part of the window (as an extended support flange) or added to the structure
over the base flashing and sealed as an additional layer. Some window
manufacturers can supply extensions.



-----Original Message-----
From: greenbuilding-bounces at lists.bioenergylists.org
[mailto:greenbuilding-bounces at lists.bioenergylists.org] On Behalf Of John
Straube
Sent: January-16-12 9:04 AM
To: Alan Abrams
Cc: Greenbuilding; Green Building
Subject: Re: [Greenbuilding] windows - flush vs recessed

The most common detail I have seen (mostly in PH details, but  is one which
uses an undrained EIFS over the window.
The flashing you are talking about is just a rain shedding / sun shield /
esthetic finish.  Water leaks past those, and as a result leaking windows
are the number one moisture problem in buildings. The flanges and grooves
are not far enough in to deal with rain leakage in most exposed locations,
they are just the rainscreen.
The drainage plane must come under the window (aka subsill flashing - see
attached simple sketch for one mid-performance solution), regardless if it
is flanged (most common in US residential windows and a lot of Canada) or
not (most commercial, all European, many Canadian residential windows).

The low cost solution is often to make the the face of the exterior foam the
drainage plane. This works well with exterior flanged windows, but can also
work with plywod bucks.
The higher performance solution is to place the drainage plane behind the
foam (which can be rockwool sheathing in this arrangement) but outboard of
the sheathing and framing (which are the components that need protecting).

We have taken flanged windows and attached them instead with metal tabs, as
sometimes this installation method is desirable.

I dont have the NIST house drawings on my laptop, but will try to get them.
Have lots of site photos: see peel and stick membrane air and water barrier
wrapping inside the window opening.  When you add R25 on the outside of
this, you get a very low U-value enclosure.





On 12-01-16 10:26 AM, Alan Abrams wrote:
>
> One design to avoid at all costs is the one which covers the window 
> frame with exterior insulation, therbey trapping rain leaks in the 
> wall (but reducing building heat loss by as much as 1%).
>
> John
>
>
>
> what this suggests then is that the drainage plane is outside the 
> exterior insulation, right? I'm trying to recall the arrangement in 
> the NIST net zero test house, which has a conventional frame wall with 
> a membrane applied to the structural sheathing, and then two layers of 
> foil faced foam, then furring and siding. but (having loaned the 
> drawings to someone who did not return them) I can't recall where the 
> windows occurred in the section, and if the membrane was the primary 
> drainage plane.
>
> wrt to protecting inner layers of wall when windows are set in from 
> the exterior plane--note that conventional US clad windows, such as 
> Weathershield, have a groove to receive metal flashing in the 
> outermost edge of the projecting part of the extrusion.
>
> conversely, in a passive house under construction in Arlington VA, 
> which I visited yesterday, the windows had no flanges, but were 
> attached with tabs that projected toward the inside edge of the wall.
> integration with the rainscreen was not complete yet, but seemed to 
> rely heavily on butyl tape--whereas the PH detail manual--and other 
> photos I've seen--feature metal extensions at least at to cover the 
> sill. (btw, the Arlington windows were untreated vinyl by Intus, rated 
> at +/- U-0.14)
>
> -a

--
Prof. John Straube, Ph.D., P.Eng.
Dept of Civil Engineering / School of Architecture www.buildingscience.com