Posted By: Mike GeorgeDon't think so Tom, not multi-directional (360 degrees) through multiple different materials at the same time - which is what happens in reality.

Except, in 3D it is 129,600 degrees, as each degree on the 2D plane has 360 degrees of movement.
Then you have to calculate every property that many times. So if you do temperature gradient, moisture, thermal conductivity and specific heat capacity, you need to do 518,400 calculations for each 'slice' of each material.
This is why finite element analysis is slow.
Much easier to go down the monitor and estimate route, which is what I suspect most models are based on.

Maybe the 5c quoted by Viking is a bit of an exaggeration but I still cannot see the issue of insulating the void?

Posted By: SteamyTeaExcept, in 3D it is 129,600 degrees, as each degree on the 2D plane has 360 degrees of movement.

Yes, I meant in a single plane..... but the correction gives more weight to my point :)

Posted By: Mike GeorgeNope, simulation software's cannot model that. It can only pick a slice or where one known object intersects with a homogeneous object in a sphere surrounding it - can we agree on that?

I suggest reading up on finite element analysis. It can be used for heat transfer analysis as well as stress analysis, or indeed both together, or pretty much any other property. It's just that it's computationally intensive.

Posted By: Mike George So you got he 5 deg figure in THIS case how? Presumably by calculation.... Can you share your calculations please so we can all understand the justification for your claims. While you are at it can you also provide authoritative justification for your calculation method please? How about ONE single reference to an authoritative study of beads in a floor void ANYWHERE in the world?

I drilled thermocouples at 45 Degrees into the joists and measured the temperature and compared it to the temperature beneath an uninsulated floor. The house was also dry-lined so the heat-loss point at the wall/floor junction was through the joists. The heat-loss could also be picked up externally showing hot spots in the wall which were 3 Degrees hotter than the rest of the wall.
There's no reference to pumping the void beneath a wooden floor with EPS beads anywhere else in the world because I'm the first to do it!

Posted By: Mike GeorgeMy experience doesn't matter here -anyone who has done a simple temperature gradient calculation can see a hole the size of a bus in your argument. . I'm not the one making wild claims and giving advice contrary to Building regulations and Best Practice - presumably the writers of such documents have sufficient experience to know what they are talking about....

A temperature gradient in every wall will show a dew point in the wall but what's the point if there's no moisture present to condense? If you stick a moisture probe 10mm into external insulation on a wall there's no humidity present but that's exactly where the Condensation Risk Analysis says it will occur, why is that? If there was moisture present at the joist ends then the wicking power of the wood and masonry would wick it away before it does any damage. How did best practice become best practice? its when people like me push out the boundaries! I can show you 70+ projects where pumping the subfloor with EPS beads works and the joists haven't rotted. Can you show me one project where a timber floor with pumped EPS bead in the void rotted?

Posted By: Mike GeorgeWith regards to THERM it cannot predict the unique and multi directional heat and vapour movement through a joist end any more than WUFI or TAS can.....You need real life observation and study for that(and I don't mean 'I've done 70 houses so it must be okay for everyone to do the same') - even then the limited conclusions you would get from a PROPER SCIENTIFIC study would only apply to a single scenario.

I know all about Therm and its limitations, I was the first to mention it on this forum, it disregards water vapour diffusion from the inside, assumes an airtightness of 3ACH and if you plaster or put a water repellent on the outside of the wall every wall build up passes.

Posted By: Mike GeorgeIt's just missing the point. It makes no difference. Neither does it really matter how warm the joist is relative to the rest of the *floor*where it meets the masonry. It is about the Dew Point and where it occurs. Where the joist becomes embedded in masonry there is only insulation adjacent to it - not surrounding it - so the temperature gradient will decrease sharply from the joist into the masonry. At this point the external temperature becomes more important. If it is extreme for the UK - say - 5 deg, then the temperature difference is high and the likelihood of the Dew point occurring within the masonry will be very high. Hence condensation and risk of localised timber degradation.

If in a worse case scenario moisture was present and a few drops condensed on the joist ends, the temperature would have to be sub zero for this to occur which is too cold for fungus/mould to grow, by the time it heats up to the 12 degree temperature where fungus/mould can grow the moisture would have evaporated or wicked away. But because the vents are closed up its very airtight beneath the floor so there's no moist air present to condense. Anyway the heat-loss from the house keeps the joist ends warmer and drier than they were before so this never happens!

Posted By: Mike GeorgeI have posted pictures of such degradation on other threads which have been dismissed as not being relevant as the floor in the examples is not insulated - this is not the case as the insulation makes NO DIFFERENCE to the likelihood of the dew point occurring within the wall make up adjacent to a joist. And if the joist is already partially degraded then the risk of failure is increased. You cannot know this without destructive uncovering and visual examination of each and every joist end.

We drilled into the joist at an angle to put in the thermocouple and measured the humidity of the material that was removed, its moisture level was 12%, better than before!

Posted By: Mike GeorgeThe risk to joist ends and other intersections between sub floor timbers and masonry (stub walls etc) is primarily controlled by historical knowledge and laterly regulations about water and vapour movement - timber degradation due to dry rot (for example)CANNOT occur where the timber moisture content is less than say around 18 %. This risk of this being designed out by copious sub floor ventilation.

Where did you cut/paste this from? Better way to keep the joists healthy is to pump the void.

Posted By: Mike GeorgeSo the idea of surrounding wooden joists with insulation while leaving the joist ends in the masonry is completely at odds with both historical design and current building regulations - and any professional recommending/advocating it is wide open to litigation should there be ANY damage to the building which results from their advice. It's cowboy practice.

It's exactly what happens when many flat roofs are built and the internal walls are drylined, the insulation is put between the rafters and the rafter end sits on the cold outer wall. I don't see many of them rotting! I suppose that's also a cowboy practice, Eh Tonto!

Posted By: djh

Posted By: Mike GeorgeNope, simulation software's cannot model that. It can only pick a slice or where one known object intersects with a homogeneous object in a sphere surrounding it - can we agree on that?

I suggest reading up on finite element analysis. It can be used for heat transfer analysis as well as stress analysis, or indeed both together, or pretty much any other property. It's just that it's computationally intensive.

Hmm, I'm not sure your comment relates in any way at all to what I said.......... And with regards to reading text books I think using and teaching a Simulation software package (TAS) for more than ten years qualifies me to give an opinion on this.............

Working link to my reference. Basic methods but reliable and compliant by definition with Building Regulations http://regulations.completepicture.co.uk/pdf/Building%20Regulations/Construction%20Details/Limiting%20thermal%20bridging%20and%20air%20leakage-%20Robust%20construction%20details%20for%20dwell.pdf

Not really Saint, otherwise radiators wouldn't work?

Posted By: VictorianecoNot really Saint, otherwise radiators wouldn't work?

They don't if you cover them in insulation

Posted By: Mike George

Posted By: djh

Posted By: Mike GeorgeNope, simulation software's cannot model that. It can only pick a slice or where one known object intersects with a homogeneous object in a sphere surrounding it - can we agree on that?

I suggest reading up on finite element analysis. It can be used for heat transfer analysis as well as stress analysis, or indeed both together, or pretty much any other property. It's just that it's computationally intensive.

Hmm, I'm not sure your comment relates in any way at all to what I said.......... And with regards to reading text books I think using and teaching a Simulation software package (TAS) for more than ten years qualifies me to give an opinion on this............." alt="" src="https:///newforum/extensions/Vanillacons/smilies/standard/cool.gif" >

FEA simulation software uses a 3D mesh that the user gets to make as detailed as necessary for the model. So you certainly don't have to pick a slice or use homogeneous spheres. So no I don't agree with what you wrote about simulation software.

I have no idea of the limitations of TAS as a specific case, but if it has those restrictions then it clearly isn't a suitable tool for the job under discussion.

Posted By: Victorianecoit would just take a longer time?

That's what insulation is.

Some do and it still works

Not when the heat source is in the small space,

Posted By: Viking Housedid you spot the details where the joist and rafter was sitting into a cold wall, recipe for rot in your opinion.

Can you give me the full reference (page numbers) to that please so I can comment specifically on what you are asking?

Posted By: Viking HouseI quietened you with 5 out of my 7 replies that must be a record,

Indeed. A new record for straw man arguments in a single post! - Like I said I no longer comment on those. Strangely though this is another one of yours and I find myself replying...... I guess it needed to be pointed out to readers who may not know your form....

Posted By: djhFEA simulation software uses a 3D mesh that the user gets to make as detailed as necessary for the model. So you certainly don't have to pick a slice or use homogeneous spheres. So no I don't agree with what you wrote about simulation software.

The way I see it is that the manuals for these softwares tell you all kinds of things... as can the people that use them. What they don't tell you is how you have to fudge the input to *make* the software do what you want it to do- which waters down the output; sometimes to a level where it is totally unreliable.

I used the Tas example as I am very familiar with it. The manual says you can model heat loss dynamically through ground constructions...well yes you can.... But..... and without getting myself in to a libelous situation I can tell you for a fact that it all depends upon the level of detail you want to model. When you start thinking of (for example) what a random stone wall adjacent to a floor (which you need to know) actually contains in reality - various types of stone; mortar made from who knows what; air voids; intentionally placed timber (such as joist ends) unintentionally placed timbers etc....1. you have no way of knowing what to input and 2. even if you did you cannot model it.

And this example applies to all software. You may indeed be able to model a 3d mesh in Therm or whatever, but how big can you go with this?

How many joist ends in what floors? orientated how? ventilated how?
What surrounds EACH AND EVERY one of them? Air? DPC? stone? Brick? Mortar?
What is the water table like?
What are the external ground levels like?
Are there any cracked drains adjacent to the wall which the joist are embedded in?
Are there any broken gutters/rainwater pipes deluging the walls in isolated areas?
Are there any other leaking pipes? Under floors or in voids or wherever?
Is there any existing wet rot/dry rot in any of these elements of construction?
Are there any other unknown interventions which have taken place in the buildings lifetime?


And then of course there's the small matter of how you know the answers to the above questions in the first place?

Have you carefully demolished the building piece by piece to find out? because that is what you would need to do to get the software input in the first place.

In order to justify extrapolating single applications into general guidance applicable to all housing you would need to be able to MANY MANY different modesl(if relying on models alone) to this level to gain ANY kind of Certification. Clearly it is not possible (in fact it's ridiculous) otherwise we could rip up the building regs and all do whatever our lovely software suggests

Give me answers to my points above and I will agree with you and bow down to the great gods Viking and Therm