From http://www.greenbuildingforum.co.uk/newforum/comments.php?DiscussionID=8333&page=1

So as not to hyjack that discussion, thought I'd copy the relevant bit here. It would be good to see what others have concluded

Working link here http://www.buildingforafuture.co.uk/spring06/25-50.pdf

Unfortunately not any help to me. My thermal modelling days are over for the time being

Is always useful to collect data from hard to get at places, put several in.

I have several thermocouple probes in the ground under and near my basement. One was in the centre of the basement a meter below the slab.

I'm using Therm from LBNL. The UI is idiosyncratic (and the modelling is steady-state only, so no inter-seasonal stuff - but that's OK for me as I'm not using ground coupling etc.), but once you get used to it, it's quite a useful tool.

2D only tho', which can be a pain with some details (e.g. sloping roof to wall junction). If you're going to attempt to use it, then google for the tutorials - there are a couple of handy screencasts...

I've done part of a Fakro rooflight cross-section (looking at different flashing options), and have some wall and floor details in-progress.

I'll try and post some results here in the next couple of days...

Tim.

Posted By: JSHarrisOK, I'll try and stick as many sensors in as I reasonably can. The data is bound to be useful to someone and it may well help me to determine how accurate the fabric heat loss predictions I have are against reality. The cost of the sensors is low, the main challenge will be getting the wiring tough enough to stand up to conditions on site during the build, I suspect.

I'm intending to do the same so any feedback on this will be useful.

No one else done any computer modelling of floor slabs then?

Posted By: SteamyTeaNot of a floor slab, but could be something to do when I have a spare 5 minutes" alt="" src="http:///forum114/extensions/Vanillacons/smilies/standard/wink.gif" >

or 5 years

Posted By: Mike GeorgeNo one else done any computer modelling of floor slabs then?

I had a go using Excel a while back, spurred on by Tony's efforts with heat storage. I could only simulate 2D or 3D solids of revolution - ideal for lighthouse builders - but it worked well enough for my purposes and gave me a better idea of how heat flows throught the ground. It was cumbersome to use and easy to make errors, plus I lacked the computer knowhow and motivation to take it any further.

A slightly different tack might be to employ one of the open-source finite-element models, see: http://www.cfd-online.com/Wiki/Codes - ELMER and OpenFOAM look like good contenders for modelling heat transfer (conductive, radiative, convective) through solids, as well as fluids.

Admittedly, ST's 5 minutes is more likely to be 5 weeks or months - but would furnish skills that could transferred to many of the problems debated within this forum. I think I need to 'retire' early...

(thanks djh - I'm not here often enough to remember how to HTML/text links consistently...)

We had a discussion about modeling heat flows into the ground before. Here's a post I made a while ago:

See page 193 onwards of http://www.nrel.gov/docs/fy08osti/43388.pdf for a description of how BASECAL and BASEIMP (the models used in esp-R and hot2000) work and what the modeling assumptions are in terms of water table, slab thickness etc. These numerical models have been validate against real buildings over multi-year periods.

Natural Resources Canada has made the BASECALC program available for download on its own: http://canmetenergy-canmetenergie.nrcan-rncan.gc.ca/eng/software_tools/basecalc.html


See also: http://www.esru.strath.ac.uk/Programs/ESP-r_capabilities/3Dcond.html

Paul in Montreal.

Posted By: Mike GeorgeI realise we discussed this before but the main discussion was [i believe] on the old forum which is now long gone.

The post I re-posted was from 2010 on this forum.

Posted By: Mike GeorgeThe importance of such fudges in terms of their affect on results is a moot point. A further limitation was [is?] the inabilty of the software to model the affects of ground temperature over several rather than a single year.

If you search for esp-r and ground heat flow modeling, there are quite a few papers that come up. I have a hunch esp-r (or maybe it was RETSCREEN) was used to model the Drakes Landing Solar Community's ground store too.

Paul in Montreal.

Thanks again Paul, but I think you misunderstand what I'm looking for. The original discussion I was referring to was relating specifically to whether or not adding insulation to solid floors was cost effective. At the time, the discussion revolved around two publications in 'Building for a Future'. The majority view was that the work I had done was flawed (ha ha), and Exeter Univeristy attempted to replicate the work. One of their conclusions was that the software I used was not appropriate for the task at hand, as it was incapable of accurately modelling the complex heatflows, for reasons I give above, amongst others.

This is also a moot point, as it comes down to how important the limitations are, and whether or not making allowances for them [by fudging] is acceptable.

As I am no longer doing thermal modelling, my interest is more curiosity rather than anything else and is limited to what I can glean from other peoples conclusions, rather than the heavy reading you posted.

Would the power losses for a volume of material of known R-value and known [x,y,z] size look something like this?
Imagine it is a cube with x,y,z dimensions being equal.
Most of the losses are at the heat input end, gradually diminishing as the cube gets bigger (volume to surface area getting greater. The Delta T is currently set at 1K. R-Values are between 0.05 W.m^-1.K^-1 and 2 W.m^-1.K^-1, or somewhere between Aerogel and Concrete. If it looks right, and if anyone can verify it even better, I shall try and work out a way to allow irregular shapes to be calculated.