UKenGB

Thanks for the replies. I'm going with edge foam. Better be safe than sorry. 😀

Because I was hoping for confirmation of my analysis, or a reason why I'd got it wrong. I can't help but doubt that all new workshops have edge foam installed when flow screed is laid to provide the smooth and level floor they want. Also, most suppliers of this sort of thing only quote its necessity for heated floors, without mention of the unheated alternative. My feeling is that without the foam will be stronger, but maybe I'm missing something vital and hoped for confirmation either way.

What expansion? The blocks walls sit on the concrete base. The screed will be bonded direct to the concrete so all will expand together. I just don't see any expansion causing any movement between the 3 parts of this equation, concrete base, screed and walls. An 'unbonded' or 'floating' screed on top of insulation and heating pipes etc, yes I get it. Must allow room for different expansions. But this is just concrete blocks on a concrete base with a concrete screed and no differential heating involved. They'll simply expand all together. Hence why I mostly read that for unheated floors, expansion is not the issue. Edge foam in that case is purely for thermal insulation, which in my case is also irrelevant. So I'm questioning the need for edge foam as seems to me the structure will be stronger if all elements are bonded together. If the screed is moving independently of the concrete base, somethings gone wrong with the bond between the 2.

This is for a new workshop floor. To make it smooth and level, I'm using a flow screed onto the concrete slab (with adhesion promoter/seal on first). However, do I need perimeter edge foam? The floor is entirely unheated so why would the concrete slab and bonded screed expand at different rates? I've also seen stated that with unheated floors, edge insulation is all about preventing heat transfer between floor and walls, but in this case they'll basically be at the same temperature so that would not be an issue. Apart from the (not huge) cost of 100m of perimeter edge foam, seems to me that allowing the screed to bond fully to not only the concrete base, but also into the roughness of the concrete block wall has to be a good thing. Preventing such bonding surely results in a weaker structure - as long as thermal expansion is not going to create a problem. I will need to seal up any gaps to prevent the screed from flowing away where it should not, but is perimeter edge foam required, or would it be better to let the screed fully bond not only to the concrete base but also the block walls?

Planning on using something like Kingspan panels for a workshop building, each panel about 5m long. How bendable are they? I realise they are pretty rigid, but how much can they actually bend over that length? The problem is that the eaves 'plate' is possibly 10mm higher than the plane of the purlins which are at about 1.2m spacing. I doubt the panel will flex 10mm in 1.2m, so I'd have to pack up the nearest purlin but what about the next one, or the one beyond that? I might not need 10mm, it may end up only 8 or even 6mm. Just trying to get some idea of what might be possible. Anyone any experience of these panels and can give me some idea of how much they can flex?