Posted By: CWattersSo whats the best insulation to use on the inside of a hot tin roof? One that blocks radiant heat or both radiant and conducted?

The point is I don't see how the dominant transport mechanisim in the existing uninsulated structure effects the choice of insulation type being added.

Agree. How the heat was transmitted through the thickness of the tin, or any other material, doesn't determine what mix of conduction/convection/radiation is emitted from its under-surface.

It could have been a sheet of Onduline (sawdust bound with bitumen, with lots of micro-pores within the wood fibre and maybe the bitumen too) or slate (also full of micro-pores, I'd guess) whose under-surface gets almost equally hot. The tin would have purely conducted, but I'd say that radiation would dominate within the interstices of the Onduline.

What mix of conduction/convection/radiation is emitted from the under-surface of either of them, is determined by what happens below it - e.g. is the sheet 'looking at' an airspace, of what thickness, or is it in contact with a substrate. Irrespective of whether it's tin or anything else above.

Posted By: CWattersthat appears to be the implication of the "60% is by radiation" argument. (eg a slate roof transports 100% of the heat through the slate itself by conduction so a radiant barrier is usless?).

I'm not saying that at all. I'm talking about the mechanisms of heat transport within the thickness of an insulant (or most other building materials), which make its behaviour different, in real life, from what's conventionally assumed/understood/got from artificially steady-state testing.

Posted By: SteamyTeaPhysicists approximate from their data quite a bit, much of it because of the accuracy of the equipment used

That's fine as long as the results so obtained are useful and correspond with reality. In the slide-rule era, concepts and theories had to be pragmatically simplified, eliminating much complexity, in order to be practically calculable. In the number-crunching computer era, that's no longer so necessary - tho there's always further areas of complexity that have to ignored, even by super-computers.

The problem comes when those approximations are revealed to be inadequate, when new combinations of materials or new uses place novel emphases on the assumed/glossed-over parts of the theory. As in radiation-blocking insulants, or the opposite, carbon-blacked/radiation-absorbing insulants (aerogel, 'Platinum' EPS). Does the scientific community then promptly review its assumptions and readily correct course? No way - the old assumptions are gospel and challengers are heretics. The scorn and fury of the defence is a sure sign of emotional crutches of routine and security being kicked away, jobs and reputations threatened by new thinking.

It's not true that experimenters are careful not to influence results to fit their expectations. For example the outrageously biassed investigation by the hack 'scientists' of NPL into multifoils. By using the very testing methods that CMM (Confed of Multifoil Manufs) was showing as inadequate and misleading, they 'proved' that multifoils don't work, exactly as they expected. As a sop to CMM, they allowed hotbox temps to vary, and still got neg results. But, in studied ignorance of the physical principles and mechanisms that CMM were suggesting, the rate-of-change of said 'variation' was so glacial, that it amounted to steady-state. They didn't check with CMM as to what 'dynamically varying' (i.e. just like real life) should mean, to be relevant. In general, the hotbox is a way of carefully selecting/eliminating parameters so as to produce results that still correspond with the 60yr old slide-rule-era assumptions and approximations of building science, upon which BRE's reputation was built (and then squandered by Mrs T).

Most recently, those troublesome experimental anomalies and 'inaccuracies' that had to be approximated or statistic'd out, have been discovered to be due to the fundamental non-deterministic chaotic behaviour of systems, in all areas (except, apparently, building physics) - and behind that the unpredictabilities of quantum behaviour. Everything around us is in perpetual chaotic ferment, even with no apparent perturbative input to set it going. Accepting this, new ways have been developed to understand the semi-predictable nett 'patterns' that emerge from that chaos. That offers exciting new potential to control, manipulate or take advantage of sophisticated kinds of physical behaviour newly revealed, and the development of new physical theories couched on the 'pattern' level (rather that classic Newtonian 'billiard-ball' level), which incorporate and reproduce the results of traditional/approximated theories (like building science's versions of heat-transmission theory) but are equipped to manipulate strange but useful behaviours that have been traditionally eliminated as anomalies.

Posted By: fostertomresults so obtained are useful and correspond with reality

Case of whose reality really.

Posted By: fostertompromptly review its assumptions and readily correct course?

Yes, generally they do, called the peer review process.

Posted By: fostertomoutrageously biassed investigation by the hack 'scientists' of NPL into multifoils

I don't know these report but would be interested in reading them.

Posted By: fostertomapproximated or statistic'd out

Very powerful tools as long as everyone is clear exactly what was done and the reasons why. An example being a typical repeatable train journey, the statistics over many months will say what the mean speed was, it can also tell you what the excepts where.

Posted By: fostertompotential to control

This is the area that has really changed, we know have much better control of over things since moving away from the analogue to the digital (price had plummeted as well). A bi-metallic thermostat is a good example, a cheap one probably had a 5 deg window, an expensive one possibly 0.5, now a cheap microprocessor can measure and control hundreds of temperatures easily to a much finer resolution. I still wonder why they are not used more.


Philosophy of science is an interesting area, not one I really wish to enter, but the basic theories are valid and applied.


Well I have put some tests out and I suspect that the order will be Tile the hottest, followed by just tinfoil, insulation(if it is not swamped, this may need investigation) then tin foil with insulation wedged in between.
That is the hypothesis, the data will show if it is correct.

Posted By: SteamyTea A tile with tinfoil on the inside, a tile with just thin insulation and one with the same insulation but with that insulation foiled both sides should do it. What do we think?

Woud be interesting to compare three tiles, one with a layer of foil underneath and the other a layer of newspaper and a control. Repeat with tiles swapped around unless the tiles are similar colour.

There's been a bunch of speculation and ill-informed rumour-mongering in this thread about how science lags behind reality and ignores new facts that arise. I was finally stung into doing a little digging to show the facts. So I was investigating the history of foam insulation - and I'll come back to that in another message, because it's a longer story - and I came across a paper about multifoils that I couldn't resist.

A while ago I dug around and discovered that multifoils are excellent when used for insulation IN VACUUM. They're used on spacecraft and they're used in high-vacuum cryogenic equipment. But their performance drops off as pressure is increased and there's a trade-off so foam insulation is better at atmospheric pressure. Google "multilayer insulation" and perhaps "cryogenic" to find lots of information. But this time I found a paper that makes it clear that DYNAMIC considerations are also fully considered ...

http://accelconf.web.cern.ch/AccelConf/e00/PAPERS/TUP5A13.pdf

... because the entire measurement technique is based on the dynamic response!

I'll paraphrase the method in the hope of making it easier to understand but please read the paper before declaring it's rubbish.

The basic idea is that you take a piece of insulation and keep it at a fixed temperature (20 C in this case). Then you shine a light at a small part of the 'front' surface (one watt in this case) and vary the intensity of the light sinusoidally. Then you measure the slight sinusoidal variation in temperature that appears on the 'back' surface of the insulation. By comparing the magnitude and timing of the variation on the back and comparing it with that of the variable light on the front, you can work out the physical properties of the insulation.

It sounds like magic; the temperature variations being measured are around 10 microkelvins and they're extracted from the normal random fluctuations using a box of electronics called a 'lock-in' amplifier. But the technique is real and the amplifiers can achieve performance like this. I've used them myself to measure other small things.

So far from science ignoring dynamic radiative effects; they're a fundamental basis of the techniques used to measure the static properties!

Apparently the insulation used on existing space suits will have to be redesigned for a mars trip because they rely the vacuum of space removing gas from between the layers and reducing the conductance. Article here..

http://www.bme.utexas.edu/docs/kd/220.pdf

says the "..heat lost trough the current insulation is expected to be intolerable at Mars pressures.The presence of interstitial gases compromises the insulating properties of multilayer materials that were designed for use in a radiation dominated environment."

"The target thermal conductivity to provide adequate insulative properties for the suit is 0.005W/mK."

Given that insulation on current suits is only about 0.1" thick and has to be flexible that sounds challenging.

Indeed - aerogel, 'platinum' EPS and multifoils are all playing similar games. Both aerogel and multifoil give 'impossible' results in the hotbox - in opposite directions.

Biff

Fits in with the cooling law nicely by the looks of it, just as it should

Well' I've tried v hard but I don't understand what Chapters 3, 4 and 5 mean, so am not throwing in the towel yet!Is anyone up for a bit of correspondence, to help me understand the terminology of these chapters?

Hi could you re save that as a Zip file or something else.
Cant download somehting to open with as a) have to pay, b) work computor
thanks
Cheers
Mike up North

Nope nothing
error - cannot find source directory??

Keep trying

Yes it downloads. using rar extension you get the usual option to find the prog to open it with and changing ext to zip it is recoognised as a zip, but on opening gives the error message. This is winzip v10.

I will try it on another machine

Cheers

Paul

I re saved it as an 2003 but it went from about 700k to over 2.1mb, know idea why that happened, sometimes technology is a mystery to me.

DJH

Was a simple polystyrene box with a roof tile on top. I then added different insulation under the tile and measured the temperatures as the solar input varied. Disregard Solar Shiny and Solar Matt as they are left over from a previous test.

As for oocalc not reading it have you tried Find and Replace, or a copy of Excel, there are enough versions floating around.

Derrick

All the test were done side by side. The temp drops are from maximums, mean temperatures are not so impressing with means of:

Standard Tile=21.8
Foiled Tile=20.2
Insulated Tile=20.6
Foiled Insulated=19.2