Skip to first unread message
sun wenning
Dec 3, 2019, 9:39:19 PM12/3/19
to THERM
Hi,
I am modelling a spandrel area for part of my curtain wall system.
I found a user manual from this link below.
It changes the edge distance from 63.5mm to 152.4mm.
My question is if the total height of my spandrel (shadow box) area is less than 300mm, what edge distance should I take? Since if I take 152.4mm for my head and sill detail of the spandrel area, my U-value area weighted spreadsheet will become not valid.
I would like to know your opinion on this issue.
Many thanks!
yalin uluaydin
Dec 5, 2019, 9:26:02 AM12/5/19
to THERM
Sun,
This is actually a little bit of a shortcoming of the NFRC standard specifying the use of the area weighted method in ISO 15099. The linear transmittance method in the standard would account for the correct edge effects.
The 6” from RDH is applicable when there is a metal backpan. When a metal pan isn’t used the edge effects is limited to the 2.5” zone in my experience.
I usually compare the edge distance with the linear edge transmittance and set an edge distance. Generally ends up being +/-4”.
However strictly speaking using any other edge zone dimension other than 2.5” is not NFRC compliant.
Given the frames are close to each other you could model the whole spandrel in one simulation taking 2.5” edge zone at the frames and also put a u-factor tag at the remaining center of panel and use that as the center of panel value for that specific condition.
Yalin
This is actually a little bit of a shortcoming of the NFRC standard specifying the use of the area weighted method in ISO 15099. The linear transmittance method in the standard would account for the correct edge effects.
The 6” from RDH is applicable when there is a metal backpan. When a metal pan isn’t used the edge effects is limited to the 2.5” zone in my experience.
I usually compare the edge distance with the linear edge transmittance and set an edge distance. Generally ends up being +/-4”.
However strictly speaking using any other edge zone dimension other than 2.5” is not NFRC compliant.
Given the frames are close to each other you could model the whole spandrel in one simulation taking 2.5” edge zone at the frames and also put a u-factor tag at the remaining center of panel and use that as the center of panel value for that specific condition.
Yalin
sun wenning
Dec 8, 2019, 11:24:48 PM12/8/19
to THERM
Hi, Yalin
Thank you so much for your detailed explanation.
Can you provide me some detailed information regarding the linear transmittance method?
Is there a manual or a guide for me to read how to determine the correct edge effects using linear transmittance method?
I am running a curtain wall system for U-Value analysis where due to the design and structure there are a mixture of different types of spandrel.
Plus, can THERM handle analysis for curved glass?
Thanks!
在 2019年12月6日星期五 UTC+11上午1:26:02，yalin uluaydin写道：Reply all
Reply to author
Forward
0 new messages