footing detail - soil? or no soil?

[Jeff Goldblatt]

Hi, 

We are trying to convince a client to install 3' of insulation along the interior perimeter of the foundation. We are trying to show the benefits with Therm. 

We have modeled a footing detail and we have a few questions. 

1) Do we need to model the soil? How would we go about this? We don't see a material option for soil in Therm. 

2) We noticed the min. auto setting for temperature is .08* degrees. Why is this? The average min. temperature for our area is more around 40* during the winter months, so we did our first three studies with 40* as our min. temperature. The fourth study is with 0* as the min. temperature. 

Any help would be great. 

Thanks! 

[Fabrizio Prato]

Hi Jeff.

Details involving soil are also dealt with in EN ISO 10211:2008. The answer to your first question is yes you should model soil cutting it at least min(20 meters ; 2,5 B'), where B' = A / 0,5P, on the outer side of the detail and downwards in the soil and at least min(4 meters ; 0,5 B'). Cutting edges must be adiabatic. Thermal conductivity of soil is generally assumed L = 2,0 W/mK.
I've seen your images and I think there is something unusual with isotherm trend... Maybe you should upload your detail.

Fabrizio

[Jeff Goldblatt]

Thanks for the reply, I have attached my therm file. I will attempt to adjust the soil to your recommendations. 

Thanks!

[Jeff Goldblatt]

Hi Fabrizio,

Attached is my therm file with the soil depth you recommended (I think, I had a little difficulty following you).

I also put the Isotherm Settings back to auto (also attached). Previously, I had changed the min to 40 and the max to 70. 

Thanks,

Jeff

On Wednesday, February 25, 2015 at 1:23:55 PM UTC-8, Fabrizio Prato wrote:

[Jeff Goldblatt]

Alright, I think I am getting closer. I was changing the display settings, not the actual temperature. I found a place to change the temperature under the boundary condition settings for Exterior Winter. I am not quiet sure what to do with the film coefficient though. 

thanks,

jeff

[Fabrizio Prato]

Hi Jeff.

Here are some images and calculations regarding your detail according ISO 10211:2008.
Therm file is duplicated, one to perform flux calculation in order to determine Psi value (with internal and external dimensions) and the other to determine fRsi and Tsi value (with Te=-15°C)
Report TB.png is a print of my personal spreadsheet using txt data exported by Therm.

Have a nice day.
Fabrizio

[Jure Lovšin]

Hello Fabrizio,

I am sorry for opening this old post, but I have some questions regarding this topic.

I do not understand how to present thermal bridges in below cases.

#1 shows TB with Psi_e = - 0,036 W/mK.

#2 shows TB with Psi_e = - 0,029 W/mK.

I am attaching my calculations and both Therm files.

Assuming case #2 is worse, but according to Building energy performance calculation, they are both still "good" thermal bridges, since Psi_e is negative.

So, the main question: What can be an argument that case #2 is worse than #1 considering TB? fRsi or just temperature distribution?

Also, what are ψe,a and ψe,t in your results and how you calculate them? Where do you use them?

BR, Jure

[Fabrizio Prato]

Hi Jure.

I looked at your therm files.

Case #1.

I obtained the same psi values for internal and external dimensions you did calculate, BUT since Tsi should be calculated using internal Rsi 0.25 (m2K)/W for all surfaces (according to EN ISO 10211, to simulate some furniture or curtains near structures) I obtained Tsi=16.5°C and fRsi=0.9, a very good result.

Case #2.

In this case I obtained a different psi value using the external dimensions because, according to EN ISO 13370, I excluded the 250mm thick concrete slab from the floor stratigraphy so the external wall dimension is 1.55 m: in this way I obtain psi(e) = -0.002 W/mK (also Uf value should be revised according to EN ISO 13370). I then calculated Tsi = 12.9°C and fRsi=0.797, a not so very good result. THIS is the fundamental fact that tells us that detail #2 is worse than #1, because fRsi=0.8 could give mold problems in case of prolonged extremely rigid outside temperatures. So the answer to yor question is: ALWAYS check fRsi value, it is the TB's ID card, regardless of the outdoor temperature.

About psi(e,a) and psi(e,t)

For me they are the psi(e,total)'s components due respectly to the heat flux to outside air (through wall) and to the ground (through floor). It could be useful for the PHPP's "ground" sheet compilation.

Two words about Therm models:
- Quad Tree Mesh Parameter: to begin I always set it to 6, some times I raise it to 7, rarely up to 8 and hardly ever to 9, for big models a high value can stop Therm (with your settings my Therm crashed the first time I run calculation)
- Maximum % Error Energy Norm: I always start simulation with the default value 10 (as Robin Mitchell has more than once written in this forum to be the right value to consider to make a correct calculation) and I sometimes try reducing it to see if total heat flow decreases.

Two words on model geometry:
- I would move more to the right the vertical cut on the internal side of the building (as you can see isotherms are not completely parallel to the floor structure faces so the cut cannot be considered adiabatic). By the way: is your building really 5m wide? If building is not a rectangle an infinitely long building have to be considered B'=A/0.5P wide and consequently the cut should be at the lower value between B'/2 and 4m.

Attached: Tsi simulations.

Ciao

Fabrizio