I know its not the most greenest of insulation materials, but in my particular situation it really is the only alternative if I want to see my green oak rafters from below. I see that Actis still do a version here in France which uses sheep wool and they claim a u value of 6.1. What do you think?

Actis? Incorporating sheep wool in a multifoil? Do tell us more

http://www.uvalue.ie/html/pdfs/actis/Trisolaine_data_sheet.pdf

You could probably fold a piece into a nifty hat that would keep your head warm and keep you safe from harmful radiation.

Have we considered boundary effects -- strange things happen at boundaries an there are lots of them in multilayer insulation.

Biff,

Would pseudoscience include the under-cafetiere-test-method..?

Seriously though, I'm sure I came across some research once which looked at typical installations of rigid foam (gaps, loose boards, broken boards, missing boards etc all included) and produced some U-values... But can't for the life of me remember where. Has anybody else come across these tests would be great to get my hands on the results again, its bugging me that I cant find it...

J

Oiseaux, if you do not have any air-gaps either side of this insulation, it will achieve a U-value of at best 1.25W/m^2K (i.e. the same as a still air layer 20mm thick). As with all the thin foil insulations, most of the resistance to conduction is in the big air-gaps either side of the blanket, rather than the blanket itself. Note that Actis' data are from an unspecified in situ test that must have had large air-gaps.

Biff, the internal layers will have some effect (see my post http://www.greenbuildingforum.co.uk/newforum/comments.php?DiscussionID=125&page=5#Item_15 ). The effect of the internal layers is largest for rapid temperature fluctuations ("high frequencies"). But at those higher frequencies, the transmittance is much lower anyway than it is in the steady-state, so the additional benefit of the inner layers is partly academic. One really has to do the calculations for specific examples to be sure about what is important and what is not - there are no "black or white" answers, unfortunately.