Model surface temperature of uninsulated slab
145 views
Skip to first unread message
Patsy Heasly
Apr 15, 2016, 6:37:57 PM4/15/16
to THERM
Hello,
I'm trying to model the temperature at the floor surface given various floor finishes over a basement slab in Seattle that has no sub-slab insulation, (to predict condensation potential from interior moisture. I'm assuming a vapor barrier between the concrete and flooring, though it's not in the THERM model).
My exterior boundary condition is based on the response to another question in the forum about slab/wall modeling, answered by Robin Mitchell (.pdf attached), except that I used a soil temperature of 55 F.
The R-value result for just the concrete slab is higher than I would expect. Have I gone amiss somewhere in my boundary conditions or custom concrete material definition? Or is it just that the delta T between the interior and soil is not very big?
Files for the concrete and concrete with cork laminate attached (also modeling rigid insulation between flooring and concrete).
Thanks very much for your thoughts.
Fabrizio Prato
Apr 16, 2016, 1:17:57 PM4/16/16
to THERM
Hi.
(I beg your pardon but I use the international units, 'cos I'm not accustomed with US units)
To determine R-value a delta T of 1 K is enough so it is not important what external temperature or, in this case, ground temperature is used. The only thing I see in your model is that the ground BC type was set to "comprehensive" instead of "simplified" and Hc was not set to 99'999 (m2K)/W and this "improved" a little the R-value. After these little changes R-Value results 0.218 (m2K)/W which is rather probable for a 10cm thick concrete slab.
BUT to talk about the internal surface temperature I think an abstract model like yours makes little sense without the context.
As you can verify in the attached Therm model of the concrete slab, this time 5 meters wide, with modeled ground 20x20 meters L=2.0 W/(mK) and 0°C out-door air temperature (I don't know Seattle climate data, I would take the worst wintry average temperature), the ground temperature under the slab varies a lot (9°-18.9°C) and it varies consistently on the above internal surface (11°-19.1°C). So if I had to verify this concrete slab for internal surface temperature I would also model the wall and ground.
Finally, in order to prevent vapor to reach cool layers of the structure (below dew point), vapor barrier must always be installed on the warm side of insulation layer.
Fabrizio
(I beg your pardon but I use the international units, 'cos I'm not accustomed with US units)
To determine R-value a delta T of 1 K is enough so it is not important what external temperature or, in this case, ground temperature is used. The only thing I see in your model is that the ground BC type was set to "comprehensive" instead of "simplified" and Hc was not set to 99'999 (m2K)/W and this "improved" a little the R-value. After these little changes R-Value results 0.218 (m2K)/W which is rather probable for a 10cm thick concrete slab.
BUT to talk about the internal surface temperature I think an abstract model like yours makes little sense without the context.
As you can verify in the attached Therm model of the concrete slab, this time 5 meters wide, with modeled ground 20x20 meters L=2.0 W/(mK) and 0°C out-door air temperature (I don't know Seattle climate data, I would take the worst wintry average temperature), the ground temperature under the slab varies a lot (9°-18.9°C) and it varies consistently on the above internal surface (11°-19.1°C). So if I had to verify this concrete slab for internal surface temperature I would also model the wall and ground.
Finally, in order to prevent vapor to reach cool layers of the structure (below dew point), vapor barrier must always be installed on the warm side of insulation layer.
Fabrizio
Reply all
Reply to author
Forward
0 new messages