Ditra with underfloor heating

[rj1884]

Hello all.

just a quick question. If you were laying underfloor heating would you always use Ditra Matt or similar after Self levelling or would you only use it when using particular tiles. The substrate is half original floorboards and newly laid chipboard which has already been laid prior to me visiting the job. I've also only got 25mm to play with height wise.

Thanks in advance.

 

[Dan]

I wouldn't advise tiling to chipboard. I know you can get the green stuff which is meant to not be too sensitive to water but I beleive it is still recommended that you overboard with 18mm plywood. You'll also need to check for deflection on the wood side. And use noggins under the chipboard to ensure that when you fix your plywood, it is fixed at 300mm centres. (The joists under the chipboard might not be at 300mm).

You'll also need an expansion strips where two sub-floors meet. So the heating cable can't cross that.

And perhaps then use Ditra or similar but treat the floors as separate floors, with the expansions strip between them.

Check with the tilers though as I might be missing something or a bit wrong there.

 

[A&D Tilers]

marmox concrete insulation boards or similar brand, you are better off bonding with flexible timber adhesive staggering the joints over different type or depth of floor. if you are using a self leveller ardex, aditex na, sticks to everything including bitumen. just an alternative.

 

[rj1884]

Cheers for your replies. I was going to overboard with marmox or something similar then lay the UFH then self level and tile. Would I only need to use a de-coupling Matt if the tiles are natural stone.

 

[Mr Tiler]

you wouldn't need the expansion joint if using ditra would you?

 

[rj1884]

Would I still need an expansion joint between the 2 subfloors if I'm over boarding with insulation boards.

 

[Gazzer]

Would I still need an expansion joint between the 2 subfloors if I'm over boarding with insulation boards.

I would personally.

 

[Ajax123]

It is not an expansion joint but a movement joint. With two independent substrates you can have both vertical and horizontal movement both of which can crack your tiles. Without some pretty complex calculations you can never predict the levels of movement especially in timber. Over boarding might negate this movement to a certain extent but you would be foolish not to put a joint in. Why risk a call back for the sake of a few pennies.

 

[Mr Tiler]

It is not an expansion joint but a movement joint. With two independent substrates you can have both vertical and horizontal movement both of which can crack your tiles. Without some pretty complex calculations you can never predict the levels of movement especially in timber. Over boarding might negate this movement to a certain extent but you would be foolish not to put a joint in. Why risk a call back for the sake of a few pennies.

is that the same with using ditra in this instance with 2 connecting substrates?

 

[rj1884]

Ajax I totally agree. Would you then treat the 2 areas as 2 separate floors in relation to the UFH. How would that work with the wiring of the thermostat.

 

[beanz]

Are they really 2 separate floors? Or have you just replaced some floorboards with chipboard!?! I'm not sure why there would be vertical movement, unless maybe they react that differently to the underfloor heating!?! Not someone I've sen before, or even thought about for that matter..

 

[Ajax123]

is that the same with using ditra in this instance with 2 connecting substrates?

As I understand it Ditra is not designed to negate the reflection of joints. It can help increase the bay size for the tiles but if there is a joint in the substrate it should reflect through the tiles.

 

[Ajax123]

Are they really 2 separate floors? Or have you just replaced some floorboards with chipboard!?! I'm not sure why there would be vertical movement, unless maybe they react that differently to the underfloor heating!?! Not someone I've sen before, or even thought about for that matter..

This is nowt to do with the UFH on top although this could have some effect. It is more to do with potential movement of the substrate. Vertical movement is simply explained. Timber floors move and flex when loaded. You can minimise this by the use of noggins and such likes but you cannot predict this movement very accurately especially if it's a timber suspended floor as opposed to a ground borne floor. That said I am not saying you can't bridge the joint. I'm just saying you shouldn't.

 

[beanz]

Ah, I see what you're saying.. Over boarding should negate this though, as long as you don't have join on top of join.

 

[Dan]

Whatever Alan says. Do it. He knows the score. That is all.

 

[beanz]

Whatever Alan says. Do it. He knows the score. That is all.

Im sorry, but you're not going to put a movement joint in a 3m2 floor because its half floorboard and half chipboard!?! That may or may not be what we are talking about here...

 

[Ajax123]

Im sorry, but you're not going to put a movement joint in a 3m2 floor because its half floorboard and half chipboard!?! That may or may not be what we are talking about here...

Its no one else's decision but yours what you actually do and often people do stuff and get away with it. The decision to do or not do what is technically correct will be heavily dependent on probability of subsequent issues and the commercial impact those issues might have.

If the substrate is indeed one single span floor with boards on one half and chip on the other over boarding would indeed be appropriate and no movement joint necessary. If the joint is parallel to the supporting sub structure but is transversed with some sort of reinforcement to prevent differential vertical movement then no joint would be necessary. E.g. If two joists abut each other parallel along the span where the overlayment meets but the joists are bolted or nailed together then no differential movement would occur as the joists are not independent.

If however as has been suggested ( maybe I read it wrong ) and it is two separate and independent substrates with a clearly isolated joint between them then it matters not what size the floor is there should be a movement joint in the covering. That is beyond debate because you cannot guarantee there will be no differential vertical movement.

What of course is debateable is whether or not you would actually put that joint there. With 3m2 you will probably get away with no joint because the actual level of movement in any direction is likely to be very small. And of course the cost of any potential remedial would be equally small.

Its more about how the two subsrate types are separated than the fact that there are two different materials that governs the need for a joint.

- - - Updated - - -

Whatever Alan says. Do it. He knows the score. That is all.

No pressure then...

 

[beanz]

Alan, I agree completely with your last post. Just didn't feel there was enough info from the OP to jump to any conclusions.

 

[2020Tiling]

I'm doing a similar job, 8 mtr floor half floor boards and half chipboard, porcelain tiles. My question is would I still need too use an uncoupling mat after I have over boarded with hardy and at what stage would the ufh go down with the ditra? :8:

 

[beanz]

I'm doing a similar job, 8 mtr floor half floor boards and half chipboard, porcelain tiles. My question is would I still need too use an uncoupling mat after I have over boarded with hardy and at what stage would the ufh go down with the ditra? :8:

my thoughts are, no.. You'll have flexible adhesive both sides of the hardy, which should be enough to take up the movement, IMHO.