This is a correct methods question from someone fairly new to Therm modelling.
I am a bit unclear still on the correct methodology / process to choosing my u-factor tags and would love some help. I was looking through the THERM documentation from LBNL and its a bit confusing, but this diagram is helpful:
It seems I should have, according to this diagram:
a boundary for the frame
A boundary extending up partially from edge of glass
A boundary for center of glass
on exterior, a boundary for the frame SHGC exterior
and then where it says “none”, what does that mean? it just does not have a name? no U-Factor tag? or the tag is “none”?
Noticeably missing as far as I can see is any u-Factor tags for wall portions, Say you have a deep sill, is it advisable to have a u-factor tag line/boundary that turns the corner of the sill (blue line + orange line as one):
There is no requirement to follow these particular U-factor tag conventions but people developed them as a way to make it easier to abstract the results of their THERM models into something that can be applied broadly across a lot of different window geometries, which use effectively the same sill, jamb, or head details.
If you are not planning to use the results of your THERM model to calculate an assembly U-Factor for several different IGU + Window Frame geometries and you are not planning to use your THERM model to inform a full-building energy model, I would not worry about these conventions. It is perfectly legit to use a THERM model simply to compare the performance of different options for detailing a facade. In this case, you need only a single interior boundary condition with one U-factor tag to evaluate “how well you are doing.”
If you are trying to abstract your THERM model of a spandrel panel in a way that it can be used for several different spandrel geometries (with differing fractions of frame vs. center-of-spandrel), you can try to adopt a similar strategy that gets used for computing window assembly U-Factors (kinda like your screenshot). Generally, a frame tends to have a U-factor that is fairly uniform along it but is distinct from the center-of-glass U-Factor (or center-of-spandrel U-Factor) that you get by simply looking at the glass panes, gap layers, (and/or spandrel insulation). Because the transition between center-of-glass to the frame tends to have a U-Factor between these two, it tends to get its own U-Factor tag. So people tally up the areas of each of these regions over a particular window assembly or spandrel panel geometry, then multiply UxA and and come up with an assembly U-Factor for that particular geometry.
If you are trying to use your THERM model to inform a full-building energy simulation and all of the spandrel panels across the building are effectively the same, then the problem is simpler. You can probably just take one U-Factor tag over the opaque portions of your THERM model and just make sure that the opaque Wall Face of your energy model to which you are applying the U-Factor aligns with the geometry in the THERM model.
Or, if you are looking at some other type of thermal bridge in an opaque construction (like a parapet detail), you can look at the linear length of that condition across the model and use it to de-rate the opaque construction accordingly. Similar to what was being discussed here on the Pollination forum:
