Issue with Therm results

[Sadegh Hashemi]

We have been using Therm software for our thermal simulations (in the facade industry) for a couple of years, but recently we have been facing some inaccuracies and wanted to check if anyone could help us out with it.

So we were studying on a section of a curtain wall system (shown below). The frame U-value of this drawing is 6.8 W/m2K

After we added a spandrel insulation (back pan) on one side of the same section (shown below), the overall U-value reduced (as we expected) to 5.9 W/m2K. 

However when we check the U-value for the top half and bottom half of this detail the numbers seem to be unusual and possibly wrong.
The U-Value for the top half of this detail (which contains the insulation) is 1.9 W/m2K but the U-value of the bottom half has increased from 6.87 W/m2K to 9.9 W/m2K (as shown below)

So the question is why is this happening? Why is adding an insulation to the top half of this detail reducing the U-value of the bottom half? What we were expecting was that the bottom half U-value to be the same as before (or maybe slightly different) and therefore the overall U-value of the total frame to be somewhere around 4.38 W/m2K (1.9 for the top half + 6.8 for the bottom half / 2 = 4.38)

I would be thankful if anyone could help me out on this.

I am sending the Therm files in the attachments in order for the boundary conditions to be checked

Regrads,

Sadegh

[Zugur Rakhmanov]

Hi,

Check the boundary condition at the indoor surface of the glass unit. According to your model it is 'Adiabatic'.  Change that and try to check the results again.

Best,

Zugur

[Ale]

Hi,

as Zugur said, check the internal boundary conditions and also the edge tag, because in the "Section with insulation "Top half frame" you have an extension of the edge of 127mm istead of 63,5mm.

You can't obtain the overall U-value of the section with the insulation, adding half of the U-value of the frame with no insulation, they are two different things, and you can see that also from the infrared results display you showed. If you do the area weighted with the values you obtained with the two halves of the mullion (with insulation) you will have what you expected, if I have understood correctly your question.

Regards,

Ale

[Zugur Rakhmanov]

Hi,

There is one more thing.

I think it's necessary to divide the cavity inside the profile such that Nu's value will be less or equal to 1.2 (see the Therm Manual about the Nusselt number https://windows.lbl.gov/sites/default/files/Downloads/NFRCSim7-July2017.pdf).

Best regards,

Zugur

пт, 1 сент. 2023 г. в 10:00, Ale <alessia...@gmail.com>:

--
You can view the THERM forum online at:
https://groups.google.com/g/LBNL-THERM
---
You received this message because you are subscribed to the Google Groups "THERM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to LBNL-THERM+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/LBNL-THERM/f858305f-1607-4aaa-9e53-7e61bf0caa86n%40googlegroups.com.

[Zugur Rakhmanov]

During checking your models one more time I figured out several things.

First I opened two files - the Full Frame and the Bottom Half Frame. Next I checked the U-factors of both details. They was as you described in e-mail.

Next I checked the boundary conditions properties for all edges (of frame and sanpanel) in both details. And figured out several differency(see below)

Next I changed all different boundary conditions making all edges of both details the same and ran the simulation.

Results are below

Sending you both files.

Best regards,

Zugur

пт, 1 сент. 2023 г. в 12:05, Zugur Rakhmanov <zugur.r...@gmail.com>: