Installing Durock correctly on a wood subfloor

fossil said:

Elfin said:

...No tile pro who is willing to guarantee their work is going to install durock directly on subfloor...

Click to expand...

Perhaps we're not speaking the same language here. When I use the term "subfloor", I'm talking about tongue & groove plywood or OSB or whatever, typically 3/4" thick, specifically engineered to be applied directly to the floor joists/underlying structure. It's the continuous surface that's underneath your carpet, or your hardwood, or your vinyl, or your whatever. It's your floor, basically. If you want tile on it, then the typical way that's applied is to affix a sheet of Durock or wonderboard, or other suitable tile substrate, either 1/4" or 1/2 " thick directly to that subfloor and then set the tile. It almost sounds as if you think I'm saying it's OK to use Durock as subfloor...not! Rick

Click to expand...

Nope, Rick, that's not it at all and I know I've been quite clear. I'm saying that there needs to be *another* layer of structural support between the subfloor (the definition of which is also quite clear) and the concrete board used as substrate. Subfloor alone is simply not enough support for a tile substrate such as Durock, which adds absolutely no structural support, and will allow for too much deflection, leading to cracking tile and grout failure.

Well, Me and millions of other housholds have durock screwed to plywood with tile on top and have no failures, guess we all have a ticking time bomb! :ahhh:

Yes, I'm certain that you do. I know I do, each passing year bringing more failure to work done improperly by a previous homeowner.

Crumbling grout, hairline cracks... very common in homes. All because of improper installs. But, on the plus side, frequent tile install failures do keep reputable tilesetters in high demand, with good job security.

Elfin said:

Yes, I'm certain that you do. I know I do, each passing year bringing more failure to work done improperly by a previous homeowner.

Crumbling grout, hairline cracks... very common in homes. All because of improper installs. But, on the plus side, frequent tile install failures do keep reputable tilesetters in high demand, with good job security.

Click to expand...

Please try not to offend others with the I am always right no matter what attitude No one does that to you. Save room for people with practical suggestions. Thank you

Elfin is wrong. At least in modern construction, 3/4" T&G;ply subfloor, plus thinset, plus tile is all you will ever need.

You both can discredit me all you like; doesn't bother me a bit. All that matters is that some folks here will read what I wrote, follow the link I posted, and get solid advice on this topic before making a mistake that will haunt them for years to come.

Mikeyvon said:

Elfin is wrong. At least in modern construction, 3/4" T&G;ply subfloor, plus thinset, plus tile is all you will ever need.

Click to expand...

You give plywood and modern construction way too much credit. I pity anyone that follows this advice. I see a lot of modern construction that is done to minimum code. My former home was that way and I could feel the floor shake when the cat ran across it! That was why I chose strudifloor OSB and cross bracing when I built my current home.

OK, I went to the website for which you provided the link, and this is what I found:

"First of all, ceramic tiles (to include porcelain tiles, quarry tiles, various paver tiles and others) are hard, brittle and breakable. For them to be installed successfully on floors they must be well bonded and well supported underneath. If your floor is the least bit springy or “mushy” when you walk on it, it will be necessary to do some serious shoring up before tile setting begins. I cannot over-emphasize this. For our purposes, the tiles themselves have no structural value of their own. The use of cement backer boards or various membranes will not ameliorate an unsound floor structure.

Cement Backer Boards

Cement backer boards, such as Wonder Board http://www.custombuildingproducts.com , Durock http://www.usg.com , Hardi-backer http://www.jameshardie.com and others are used to “uncouple” a tile installation from the subfloor below. Before they are fastened, CBUs, as they are called, are bedded in thin set mortar, which is the usual adhesive used in setting floor tiles. The panels are then nailed or screwed to the subfloor following manufacturers’ specific directions. CBUs do NOT improve the stiffness or structural value of the floor."

How is that different from anything any of us have been trying to say? It's exactly what I was trying to say. Heck, if I had a house where the floor was "mushy", I'd do something about it even if tiling it was the last thing from my mind. But I'm certainly not going to argue with anyone about it...nor am I going to go back and rip out any of the myriad tile & marble installations I've done on floors, countertops, walls, showers, etc., in 5 different houses over the last four decades. Rick

Now Rick, we're talking about floors, which have nothing to do with countertops, walls or showers!! Those are an entirely different topic, are installed with very different parameters in mind, and have no bearing on what's being discussed here.

It's a small and old soundbite you have pulled from the site... partially true, needing updating. Keep looking. You'll find that uncoupling is achieved using Ditra, and that no matter how hard they try, CBUs *still* do not improve the stiffness or structual value of the floor

Elfin said:

...CBUs *still* do not improve the stiffness or structual value of the floor

Click to expand...

I agree totally, and I don't think I said anything to the contrary. CBU's are not structural materials, they add no stiffness nor rigidity, they just provide an appropriate substrate for proper adhesion of the tile. If the underlying structure has a problem, then that most certainly must be addressed before the tile project begins. But...not every light-duty residential application of tile on a floor requires some sort of extraordinary underpinning like floating a mortar bed. (And I did say "floors", too...actually some really nice ones, if I may say so myself ). Rick

Deflection doesn't mean there is a problem with the underlying structure. There is going to be deflection of some variety, no matter how sound the structure is. The issue is the rigid fragility of tile and grout, and making the substrate appropriately sound for a successful installation. And one simple sheet of plywood can make all the difference.

A side note: A lot of people are using natural stone these days, which is particularly fragile and has even higher requirements for a secure and sound substrate.

I'm talking about a latex additive made specifically for concrete. It is a bonding agent used wherever new concrete meets old or where concrete needs to bond to dissimilar material. It looks like white carpenter's glue and is often thinned down with water before applying. A quick google search for concrete latex additive will give you many examples of use.

LLigetfa said:

I'm talking about a latex additive made specifically for concrete. It is a bonding agent used wherever new concrete meets old or where concrete needs to bond to dissimilar material. It looks like white carpenter's glue and is often thinned down with water before applying. A quick google search for concrete latex additive will give you many examples of use.

Click to expand...

Just wanted to add another search term... "latex admix" is what it's commonly called as well.

If you use a "modified" thinset, this means it is already "latex-modified" and no admix is needed... however, some thinsets are more "modified" than others (usually the cheaper they are, the less admix).

We could get into the topic of whether expensive modified thinset is really necessary to use under your concrete board (it's not) but I think I've had about enough "debate" for one evening

You don't put down two different coats of anything. You buy what's called latex-modified thinset mortar, either pre-mixed ready to spread, or as a powder you mix with water strictly in accordance with the instructions on the package. Or, if you're really hard-core, you can buy the thinset and the latex modifier separate (looks like milk), and mix it yourself. Whatever way you choose to acquire that, it's what you trowel down onto the area where you're going to lay the Durock, and then you set the sheet down and use Durock screws to fasten it to the subfloor. The proprietary Durock screws will minimize damage to the surface of the Durock, and drive smoothly home, and countersink themselves. They only have to be ever-so-slightly below the surface, so don't go into "over-drive" with them. Snug 'em down pretty tight, but don't drive them through the Durock. They're really just clamps, squishing out the thinset to a uniform layer, and holding everything together until the thinset cures. Rick

Thin out the latex with water and whitewash the wood. Add a bit of latex to the mortar mix. By the time the mortar is ready, the latex application on the plywood would be set about right. Apply mortar with a 1/4" square notched trowel. Set the durock smooth side down immediately into the mortar and screw it down with bugle head screws. The screws should be flush, not really countersunk. Apply mortar to the butt joints in the durock and lay in fibreglass mesh tape. If the mesh tape is self adhesive, you could apply it before the mortar.

If the tile you are laying is not too heavily textured on the back side or irregular, the same 1/4" square notched trowel should suffice. To ensure a better bond, I like to back butter the tiles. Depending on how thick you need the mortar to be you can butter it with the smooth or the notched side. Apply mortar to the durock with the same 1/4" notched trowel. Always set the tile tight to the next one and then slide it away from the adjacent tiles so that you don't push mortar up into the grout joint. To prevent excessive drying, you can cover the tile with damp cardboard but not so wet that you dilute the mortar. Wait a day or two before applying grout and use the same damp cardboard to prevent excessive drying.

The watered down latex pretreatment of the plywood also serves to moisten it so as to not draw too much moisture out of the mortar, drying it before it has time to set. Mortar needs to set, not dry per se.

LLigetfa said:

The watered down latex pretreatment of the plywood also serves to moisten it so as to not draw too much moisture out of the mortar, drying it before it has time to set. Mortar needs to set, not dry per se.

Click to expand...

I like that idea, Is that a common practice?

webby3650 said:

Is that a common practice?

Click to expand...

I think it is but you would probably have to put it to a vote.

It's like common sense. Perhaps not so common after all.

First off, we need to be a bit nicer to each other in this thread... I am seeing complaints about personal attacks, which is not cool... I don't want to use my "Moderator Hammer", so let's be a bit nicer to each other...

Second, there is an article in the HearthWiki on building a hearth extension - I wrote a good bit of it myself as the result of research on what I needed to do to extend my own hearth...

A few factors to consider - Hearths or extensions are generally SMALL - thus deflection is much less of an issue. There is a deflection calculator on the Johnbridge tile site mentioned earlier in this thread, If I punch in my floor's specs for the entire room, there is NO WAY I could put down tiles without completely tearing the existing floor out, sistering all the joists, upgrading the subfloor etc. But if I figured the 18" wide extension I was building, the calculator wouldn't even accept a number that small, the lowest value it would take (5' IIRC) my deflection was WAY under the requirements, no problem...

I also got a great deal of information from manufacturer spec sheets, and from the JB forums - BOTH the JB forum and the Durock tech sheet on making a hearth extension said that as long as the subfloor was flat, and the deflection was low enough, there would be NO PROBLEM with putting Durock directly on the plank subfloor... (Sorry Elfin! You may be technically correct, but a plywood sheet isn't an absolute requirement) What is needed is to put the Durock down on a layer of thinset, and screw it down - as the JB forum puts it, "the screws hold it down, the thinset holds it up..." (There are specs for the thinset in the Durock application sheet, referenced in the Wiki article)

In the case of my extension, I cut out the carpet, and lifted the 1/2" particle board subfloor it was attached to, then put down a layer of thinset, a layer of roof flashing, another layer of thinset, and a layer of Durock, screwd down per reccomendations. Over that I put a layer of random size slate tiles to match the entryway that the extension merges with on one end - I needed to stay flush with the existing carpeted floor, so I didn't have the room to put down more - 2 years, no problems, looks great...

There actually is some strength to Durock BTW - if you search, you will see many instances of people building raised hearths using metal 2x4 framing covered w/ 3 or 4 layers of Durock, bonded with thinset and screwed to the framing, and tiled, w/ no other support - holds up a stove and the people installing it just fine... (Can't use plywood there due to R-value restrictions)

LLigetfa - can you provide a technical reference that describes this application for the diamond lath, and how to use it to get such a gap? Not saying it isn't true, and it does sound like it would be a useful technique, but before I'd want to reccomend it to someone, I'd want to have a link to some reference that I could hand to a code official or insurance inspector to PROVE that it is a legitimate process... (If the inspectors won't buy off on it, it doesn't work...)

Gooserider

I think people here are not fully understanding there are two types of failure here. There is the tile failure, cracks , etc. Then there is the R-value failure. Tile and thinset will not pass the exposure to heat over time. To put it simply: take what you want to make your hearth out of and see if it passes the "O/S" test. Hold it in your hand, heat the other side and see if you can figger out what "O/S" stands for............

To achive an R-value of 2.0 from common marble the hearth has to be in excess of 50 inches thick

Gooserider said:

LLigetfa - can you provide a technical reference that describes this application for the diamond lath, and how to use it to get such a gap?

Click to expand...

The hearth pad thermal isolation technique I described was designed and certified by RSF engineers. In my case I purchased the assembled product from RSF. One could design/build their own but I doubt an inspector would sign off on it without an engineer's stamp. One could however, deploy the technique to build above minimum spec for their own peace of mind.