Fostertom was asking how I would simulate heat flow from Woo's floor into the ground.

As the house is long and narrow (about 4m wide), I'd simulate it first in 2D ie. as if infinitely long. The result would show if it might be worth doing in 3D , which with end effects is bound to be worse.
I'd use a spreadsheet - XL in my case - and create a grid of 0.1m squares on which to set out a cross section one wall up to maybe 1m high, the floor slab across to the centre of the floor and the ground below and around for perhaps 10m. (As it will be symmetrical about the centreline, the grid need only be for half the building).
The thermal properties of each square are allocated, and the start temperatures defined. Also the formulas or tables for the driving temperatures over time - ambient air, heating , ground boundaries etc. A formula is applied to each square to calculate its temperature change after a time interval, given the temperatures and properties of the surrounding squares.

Press 'Calculate' and watch the whole grid fill with error messages indicating a mistake somewhere.

Correct mistake, and repeat calculation for a series of time intervals until a week or a month or a year has accumulated. I've found a time interval in the region of 1200 sec often works. Too large and the calculation can be unstable and head off toward infinity.

Yeah I see, both the above - ST's formula being the basics of what Mike's cells contain.

Posted By: mike7on a building this narrow there is no possibility of interseasonal storage/year on year build up of temperature

In fact not necessary, for the wind/underfloor scheme, however fascinating the speculation!

Posted By: mike7Might be interesting nevertheless to see what could be achieved over much shorter time periods - I guess even in Orkney the wind drops sometimes

Just what's needed here.

Right played with LISA a bit and this is what she produced.
http://youtu.be/_Rk29Lw7sZw
The top quarter is air that varies in temperature by 5° over a time period of 12 (can be minutes, hours, weeks, months, don't matter).
The bottom 750 mm starts off at 8°.
So this would simulate a very nice spring day down here, probably a rare event in Orkney.
You will notice that there is virtually no thermal penetration in the bottom 500 mm.
The soil type is dry clay.

I like the 1st principles approach, but not wasting too much time on it as I know the answer intuitively.

I usually get stuck on something, hence the RTFM quip

Yes I agree, but having done this several times, and always got the same answers, I am quite happy to let 3rd party software do the work for me.
Now if someone wants to pay me, that is a different matter.

Posted By: Ed DaviesApproaching the same problem now I'd probably take it as an excuse/motivation to learn something about the SciPy

I have looked at that but really not confident enough with 'ordinary' Python to take it further.
There is always the Fourier approach, just depends on how much you like geometry, I hate it, but it does give some simple solutions to complex problems.

More a case that heat can be treated as a wave passing though an aperture, as it spreads out the 'front' is diminishing in magnitude or amplitude. This can be transformed/translated to normal wave mechanics.
I read a book on it when doing my Dissertation, but can't remember the title. By the second page it was into solutions to partial differential equations (I prefer partial solutions to differential equations as they are easier ) and where I understood the concepts, the mathematics was a bit beyond me. This is why calculating real shapes is difficult, as one thing changes, say the temperature, the surrounding temperatures also change, but in different proportions, even to each other.
The main thing I picked up from that book was about thermal inertia. This is a simple concept, but again, hard to apply in the real world, though it is all there (apparently).
What is wanted is a simple method that does a good approximation from known parameters i.e SHC, Conduction, x,y,z coordinates, start temperatures etc. But there does not seem to be an easy method of doing this. This could be why this topic keeps coming up as a 'solution' to lowering energy usage.

For most people, the amount of loss though a floor, and by definition, into the ground, is something that they have to live with, as not everyone can dig down and fit massive amounts of insulation. So better off adding what insulation they can on top, or after minor excavation, and say 'that's the best I can do, suck it up Princess'.

Tom
It would be like adding some extra wall area, at a different angle, I think. Realistically it is not practical though. You would end up sphere, or maybe a parabola or hyperbola.
Have you got a book on thermodynamics?