FYI. I haven't done the parametrics to see if
this situation applies to shading of windows by shading devices, but I have
found a notable problem with direct beam reaching an interior window when the
exterior window is much larger than the interior window. While working on
a double-skin building, we found that no direct beam reached interior windows
when the outer skin window was one big surface covering 12 stories high by 8
zones/rooms of equal size wide.
The windows that did receive a direct beam
component (reported in the variable "Beam Solar Transmitted through
Interior Window" or "Zone Beam Solar from Interior Windows") were distributed in
a checkerboard approximately. To correct this, we divided the exterior
window until all interior windows reported direct beam component as
expected. The exterior window had to be no more than 4x the size of the
interior windows to have consistent results all year as expected from hand
calcs.
I don't know if this is a problem with the shading
algorithm in general or with the way it is applied to interior windows, or if
this problem was resolved between v2.2 and the latest release. Generally I
just try to keep all surfaces around the same size just to make sure. The
shading algorithm has been noted by some E+ developers as not optimal. The
Ecotect shading method is said to be the best (randomly distributed spray of
rays toward the sun to create a database of full annual conditions). I
haven't looked into the E+ algorithm enough to understand the limitations in
detail, if someone else wants to weigh in.
Daivd
|