Is thermal mass capacity impacted by the tiling process?
ChristaC
| Posted in Energy Efficiency and Durability on
I’m wondering if you can lay tlle over a concrete subfloor without compromising the thermal mass capacity of the concrete.
Would an uncoupling system, using something like Ditra, have a negative impact? Should we be looking at using sand instead?
Our building site has great orientation for passive solar, but we’d prefer tile versus a finished concrete floor.
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Search and download construction details
Replies
I would say the Ditra would have an insulating effect and would not allow as much thermal storage in the slab. There is nothing wrong with tile laid directly on concrete providing all precautions are taken regarding moisture proofing. Insulation under the slab is also needed to hold the heat. Darker colored tile will absorb more direct beam energy than lighter colors, I believe, light colors would be more reflective, heating the air instead. Be thinking about radon and methods to avoid it.
Thanks - My concern is about the tiles or grout eventually lifting/cracking without something to provide a little flexibility. It seems like most tile installation instructions bring that up.
Tile and grout work loose if the underlayment flexes, I can't think of a more durable surface than a concrete slab for holding tile, try removing it sometime. Tile, if installed properly and as long as what's underneath is solid, will last a long, long time.
According to the manufacturer's description Ditra is an uncoupling membrane to permit differential movement between a lightweight substructure (such as a wood subfloor) and a tile finish. Doug is correct that it is entirely unnecessary between tile and a heavyweight material such as concrete, you should save the money for other things.
But were you to install it the effect on the thermal capacitance of the concrete would be barely measurable. If 1/8" of Ditra were that good an insulator I'd be putting it in my walls!
I will disagree with those who say you don't need an uncoupling membrane when tiling over concrete. This is just wrong and if you plan to use passive solar (especially with sunlight on the slab) you need to use an uncoupling membrane.
Concrete often develops cracks and the uncoupling membrane reduces the likelihood of those cracks transferring to the tile. It will also reduce the stress of uneven thermal expansion on a passive solar floor. You will need to pay attention to expansion joints in the tile as well.
Ditra works well with heat transfer. I have heat mats under Ditra and there is no problem heating the tile. If you don't want to use Ditra you can use a flat membrane for uncoupling, such as from Noble.
There is good green building advice here but if you want good tile advice head over to http://www.johnbridge.com
I think the truth is somewhere in the middle here. You need to "float" a slab to allow it to shrink independently of the footings and you need to install control joints to direct the shrinkage cracks to happen at regular intervals depending on numerous debatable variables but certainly at all inside corners and every 10-15 feet in the run of large areas. When installing control joints in radiant slabs it is risky to cut them in with a concrete saw (to the tubing) so we generally use steel key-way expansion joints. Some prefer zip-strip, lots of different methods work, you just have to allow for shrinkage during curing, especially in the first 72 hours. Cutting in control joints a day or so after the slab is poured is risky in my opinion, which is why I like putting the key-way in before we pour.
This impacts tile because, once you can accurately predict where the slab is likely to crack you can place the uncoupling membrane in those locations only. (This has the side benefit of saving on the cost of the membrane) We just run a strip of Ditra, Noble or Rolled Gold or whatever membrane is your preference approximately three tiles wide in the areas over the expansion joints.
James is entirely correct. the tile will be fine so long as the concrete doesn't crack and William is correct that concrete will inevitably crack and the membrane has very little impact on heat transfer as it has negligible R-value.
To follow up on Michael's good comments you can see how Noble recommends to apply one of its membranes over expansion joints or concrete cracks.
http://www.noblecompany.com/Portals/0/PRODUCT%20INFO/Product%20Descriptions/NobleSeal%20CIS%20Product%20Decription%200409.pdf
Michael raises a very important point about leaving an expansion gap around the perimeter of the tile floor. If you are placing baseboard over the tile, leave a 1/2 inch gap around the entire perimeter. Because tilers will inevitably fill some of this gap with thinset, you can place 1/2 inch foam weatherstrip around the room or tack in a strip of 1/2 inch plywood around the room and pull it out after the tile is grouted.
Search for tile "tenting" to see what can happen without the appropriate expansion gap around a tile floor. A 1/4 inch gap may be fine unless the floor is very large.
Thank you - this is very helpful. You anticipated my next question about anticipated cracks in the cement and it's impact on the tiles. Thanks also for the recommended links.
I know a concrete contractor who guarantees his work
It is guaranteed to crack
Christa,
A poorly prepared concrete pour will have problems later on that no "membrane" is going to cover, beware of tile product salespeople. Perimerter expansion material is fine, in fact you should have a thermal break between the floor, the foundation wall and footing and have insulation under the slab. Without insulation under the slab your idea of solar thermal storage is a waste of time. Zip Strip or periodic sawcuts will help with concrete surface tension.