Floor part heated

[aaronhaywood]

Hi got a job to start prepping next week 60x60 porcelain going on 18ml ply with electric underfloor heating in parts!..Its multi room tiling lounge and bedroom will be heated but hall between the to will not. I need to make up the height of the subfloor for the areas not heated, was thinking 6ml insulation board for the areas heated and for the non heated areas to make up the height to top of heating wire with ply..:: latex the lot then cover all in ditra mat . Customer wants floor to be continuous so no door bars can save me, let me know wot you think

 

[Ajax123]

I think disaster waiting to happen. You cannot run an unjointed floor through door thresholds on timber part heated and part cold and not expect it to fall to bits. are there not even expansion joints between the ply sheets?? Run Forrest, run....

 

[Tile seller]

Could you perhaps have a soft joint in the thresholds?

 

[aaronhaywood]

No expansion in joints but does around outside, soft joint would be a good idea prob gunna be charcoal grout so black sili with blend in nice. My other thought is to lay insulation boards everywhere and lay heating where needed then cake it in latex a few times to level the lot

 

[Sean Kelly]

I did a quote a few years ago. The man wanted UFH from the kitchen door to the kitchen sink for his wife for when she is doing the dishes (i.e. a straight 500mm line of heated mat). I told him I would have to SLC the whole kitchen because there would be a 4mm height difference where the mat was. I did't get the job!

Talk the customer in spending a few extra quid on a mat and cover the whole floor in SLC.

 

[Dan]

Also, note that you can get 6mm, 10,20,30,40,50 and 60mm insulation boards. Cheap way to add height where needed and gives customer more insulation.

 

[Dan]

But as said you need expansion joints to go through all layers.

 

[aaronhaywood]

Thanks for the advise guys

 

[SJPurdy]

prob gunna be charcoal grout so black sili with blend in nice.

Charcoal Silicon sealant will look better with charcoal grout!

 

[Dan]

Mapei grout matches their silicones bang on doesn't it?

 

[Old Mod]

Maybe I'm reading this wrong or have missed something.
Why don't u use 6mm boards under heating then outside of heated area use 10mm back fill lower areas with SLC. Then the floor is decoupled too.

 

[aaronhaywood]

Yep that's what I'm gunna do, dident realise you can get the boards in different thickness - will order from tilefix direct

 

[SJPurdy]

decoupling works in conjunction with movement joints, it is not a substitute for them.
Consider a room that is heated so that the floor area expands slightly, the decoupling allows the floor surface (tiles and adhesive) to expand at a different rate to the substrate (still colder under the insulation) without putting a massive shear stress on the adhesive. This expansion is taken up by the compression of the perimeter movement joint (even if its only an air gap covered by skirting). It is also taken up by the compression of the movement joint strip in the doorway otherwise all the tiling in the hallway/next room would be pushed away from the heated room.

 

[Old Mod]

decoupling works in conjunction with movement joints, it is not a substitute for them.
Consider a room that is heated so that the floor area expands slightly, the decoupling allows the floor surface (tiles and adhesive) to expand at a different rate to the substrate (still colder under the insulation) without putting a massive shear stress on the adhesive. This expansion is taken up by the compression of the perimeter movement joint (even if its only an air gap covered by skirting). It is also taken up by the compression of the movement joint strip in the doorway otherwise all the tiling in the hallway/next room would be pushed away from the heated room.

That all makes perfect sense and is a great explanation!
I avoided the subject of movement joints for one very good reason.
I was always taught that if there is no expansion joint in the substrate then inserting one in the tiling is pretty much a waste of time, and that is the rule I have always followed.
Now many years on, I still struggle with the concept that if you put an expansion joint in tiling, (without there being one beneath) that it will be of any benefit.
The way I reason this is, if for example the substrate is continuous through a doorway for instance (seems to be relevant) and you have expansion from a warm room to a cooler one, the warm room cannot push toward the cooler one because there is no expansion.
So which way will it go? To my mind there is only one way it can go, up!
And no expansion joint in the tiles is going to stop the floor from effectively 'blowing'
For it to work effectively surely there must be one in the substrate too?
I'd really like to know the answer to this because it's baffled me for quite a while now.
Thanks.

 

[SJPurdy]

Yes that is something I can't understand either. If there is no movement joint in the substrate what is the point of one in the tiling unless the tile adhesive layer allows some shear movement of the tiling relative to the substrate (which most don't).
If a decoupling layer is used then this allows the shear movement. There must be a gap in the decoupling mat at the movement joint line so that each tiled area can move (expand/contract/vibrate) independently.

 

[Old Mod]

Yes that is something I can't understand either. If there is no movement joint in the substrate what is the point of one in the tiling unless the tile adhesive layer allows some shear movement of the tiling relative to the substrate (which most don't).
If a decoupling layer is used then this allows the shear movement. There must be a gap in the decoupling mat at the movement joint line so that each tiled area can move (expand/contract/vibrate) independently.

So are u saying if there is no expansion in the substrate, but one in the decoupler this will work as an expansion joint in the tiling?
Sorry if I've miss read it.

 

[SJPurdy]

Yes that is my understanding.
If you have a continuous flat substrate and you just tile it leaving a gap to be filled with backer and flexible sealant then I don't see how that can work.
However, if you have the same substrate and you fix a good decoupling membrane to it, leaving a gap where you want the movement joint; you then tile it leaving a gap above the gap in the decoupling membrane which is again filled with foam backer and flexible sealant then I can see that will work to reduce lateral stresses in the tiling.

 

[Old Mod]

Yes that is my understanding.
If you have a continuous flat substrate and you just tile it leaving a gap to be filled with backer and flexible sealant then I don't see how that can work.
However, if you have the same substrate and you fix a good decoupling membrane to it, leaving a gap where you want the movement joint; you then tile it leaving a gap above the gap in the decoupling membrane which is again filled with foam backer and flexible sealant then I can see that will work to reduce lateral stresses in the tiling.

Hmmmm ...... Don't see it I'm afraid.
The tiling isn't the problem tho is it.
It's the substrate that moves not the tiling. If the tiling was the only thing moving, then 100% no problem, I get it.
But the problem is it's the substrate that moves! It has to have some where to go!
The tiling only moves as a result of the movement in the substrate. So if the substrate has no where to go, it goes up!
Not being difficult @SJPurdy just don't see the physics in it.
And if the substrate was to shrink then the decoupler would cope anyway. (Apparently haha)
I do believe u believe what you're saying.....almost! Haha
Feels like there's a little doubt for you too! 🙂

 

[SJPurdy]

I believe what I'm saying is true for coping with movement in the tiling where the substrate is stable. for example where there is a screed greater than 8m across so that movement joint is required in the tiled layer to meet BS for a heated floor. I actually came into the post to explain that decoupling alone was not a substitute for a movement joint.
But oops I've just gone back to the original post. obviously if laying a new timber floor then there has to be gaps between the boards in places like doorways to accommodate expansion/movement in the timber floor itself and prevent bowing up of the floor.
But then if a de-coupleing membrane is fitted over the plywood then this should allow a small amount of expansion/contraction in both the tiles and substrate otherwise what is it for?

 

[Old Mod]

I don't disagree that movement joints are necessary, on the contrary, I believe that they are.
I'm just struggling in my head with the concept of there not being one in the substrate that one in the tiling alone will work.
I appreciate the effort u've gone thro to convince me too! Haha
And for that I thank u SJ
I know I'm also playing devils advocate to a certain extent 😀

I actually came into the post to explain that decoupling alone was not a substitute for a movement joint.

And I only came in to give my solution to height issues! Haha
I did twinge when u posted straight after I don't mind saying, felt like your post was a little pointed! (At me 😀)
Maybe thinking that I thought decoupling was enough.

But we now know that's not true don't we! 🙂
Good discussion tho @SJPurdy
What is ur name anyway?🙂

 

[SJPurdy]

So back to the original post:
construct rigid stable18mm ply floor primed, gapped, screw 300c/c etc.
cover with 10mm insul board (eg Wedi fixed to their spec) in tiled rooms and 12mm in hallway; idealy with a 6mm gap between them in the doorways aligned with where there will be a joint in the tiles.
Fit undertile heating wires and cover all with SLC but preventing it filling the 6mm gap.
Cover with ditra mat again leaving the 6mm gap (I think this is optional but it should give better heat distribution and more uncoupling than the insul board alone).
Tile, grout and then fill the 6mm gap with foam and matching sealant.
If it makes it easier cut out the gap in all the layers when you have tiled up to it rather than trying to guesss where it will be.
This way you have the appearance of a continuous floor with the doorway joint a bit wider (6mm rather than the normal 3mm).

 

[SJPurdy]

I did twinge when u posted straight after I don't mind saying, felt like your post was a little pointed! (At me 😀)

I just happened to be on site at that point. I aim to help or add to discussions constructively so promise it wasn't trying to be "pointed"!!
Diners ready now so have to go but I'll be back.
Steve