Hi Farbrizio and other Therm-users,
thank you for your answer!
I have to calculate with DIN EN ISO 10077-1, -2 and DIN EN 12 631 to get an NFRC100 certified u-value.
So i have to calculate my glass, opaque and edge-u-values.
I just added my therm file "Corner Detail Down" to show in which way i calculated it now:
I get an u-value of 0,8677 W/m²K at a length of 128,841mm. I added at the adiabatic boundary condition type the "u-factor-surface" to "frame" to get this result.
Do you think this is correct now?
Furthermore i added an other therm-file "Corner Detail Top" with an similar issue at an Corner detail.
When i splitted the Detail, because there is a vertical and horizontal section, i mentioned that the edge values differ from 5,2599 W/m²K to 2,4396 W/m²K
I think the reason is, because the projected direction of the heat-flux was first the double size because of the edge of the vertical and horizontal section.
So i calculated the u-factor surface just to the middle of the Aluminium Profile ( see in "Corner Detail Up section up 20° tilt")
Should i also use the adiabatic boundary conditions in this case as frame-"u-factor-surface" to get the right amount of the heat-flux?
Do you think this is the right strategy?
Christian