What actually does an uncoupling membrane do?

[Eurob]

William , the uncoupling membranes can be used -- some do -- in all situations under tiles or stones installation . Usually the budget comes into play and the '' extra protections '' start to get cut from the assembly . It takes more than '' it is not necessary , trust me , I've been doing it for ...years '' to rule them out .

In fairness , not all the projects need one , but you have to evaluate them right .

 

[Jadnashua]

What's the strongest, simplest woodworking joint, where you cannot pull things apart until you destroy the material- a dovetail. That's the reason why Schluter uses that shape on their uncoupling membrane. Relying on fibers that can and are designed to tear or break loose is relying on their ultimate strength and attach the tile to the membrane with more bond strength rather than mechanical strength. How thick is the thinset layer anyways? On most uncoupling membranes, the thickness is only used as pillars to the substrate to hold the tile up, not to anchor it to the floor in a lift-off manner.

Hopefully, now that people have an idea of what an uncoupling membrane is and how it works, they can then decide whether they should be using one in their situation, and then choose which one meets their particular needs and performance requirements. There's more than one out there. They do not all perform equally. They have different features you may or may not need (such as whether they can be used also as a waterproofing layer, for example). If you can accept that porcelain and glass tile have survived for centuries before modified thinset was invented because they set those tile over an uncoupling layer, you'd understand that a modified in this situation is not required, and because the membranes tend to be waterproof (not all are), that can pose problems with modified thinsets which need to dry to attain their maximum strength and stability. You'd also know that the tile is NOT bonded to the floor via a thinset mortar bond...the whole idea of an uncoupling layer to so PREVENT that bond, so the two can move independently. The goal is twofold: lock the tile to the top of the membrane so it doesn't come off, and two, provide support to the substrate so the tile won't crack because of deflection. If your subfloor is up to specs, and you install it properly, an uncoupling membrane will provide a MUCH more reliable installation under adverse conditions than any cbu or direct bond situation where you may need a modified to complete the job. And, it is easier to cut, carry, and install; plus, it will go on quicker than any cbu installation while providing the uncoupling function. Do you need it? Only you can decide. If you're paying labor, an uncoupling membrane is likely cheaper because it goes on so much quicker. And, if you add in the screws, and tape, and maybe you have to buy dedicated cutting tools to install cbu (say you're a DIY'er - a pro would already have them), the price difference is negligible. If you want to throw in a need for waterproofing (say you want to waterproof your laundry room, or extend the waterproofing beyond your shower into the bath), you need to be careful which product you choose.

My goal was to try to educate the readers on what they are, and how they work. Now, you choose which one will work best for you.

 

[Sponsor]

 

[Eurob]

Hey John , did you follow up on the link -- Blanke -- Permat -- http://www.blankecorp.com/blanke-us...ent-systems/blankepermat/product-information/


A quote from : '' 10 Year Warranty
Importantly, Blanke•PERMAT has a 10-year warranty against damage caused by excessive vertical subfloor movement (deflection), the only such warranty in the industry. Whether installing stone on wood or cement subfloors, homeowners who include Blanke•PERMAT in their stone tile installations can rest assured their subfloors will indeed support the installation. ''


Only one layer of plywood or OSB , Permat and stone ...quite aggressive , isn't it ?


Another quote : '' Advantages

Reinforces wood substrates
Only one layer of 3/4" plywood required under most natural stone installations
Provides uncoupling / crack isolation protection on cement substrates
Only 1/8" thick - minimize transition height particularly to wood laminate floors
50 % higher compression strength and 7 x greater tensile strength than rolled uncoupling products ''


Now that you mentioned '' rebars '' ...... quite strong numbers ...... did you look at it ?

I wonder if the numbers are consistent at the transitional joints in between the mats .

 

[Vegas_sparky]

What about the wire channels with no wire? That's got to grab a hold of the thinset. No?

 

[Jadnashua]

FWIW, Schluter tested DitraXL with stone over a single layer of ply, and it got as high as a heavy rating on the Robinson floor test when TCNA conducted it depending on the type of stone. But, Schluter feels, with the variability of the natural material, it is not guaranteed to always work. Testing with say crema marfil verses a nice granite stone would give very different results. Being conservative, they require two layers to get a warranty. But, it shows what a good uncoupling membrane can do for you. Pages 28-29 discuss this and list the testing results with their respective TCNA test reports. http://www.schluter.com/media/DITRAHandbook-ENG-2013.pdf?v=201404071927 Since not only the tile, but the actual construction of the wood subflooring system is natural, put together by men (or women), there is some variability in the results. Schluter only rates their material and recommendation on being able to achieve satisfactory results 100% of the time...I've heard, some take the best result of many, and then use that, not taking into account the variability of the tile or the subflooring materials. IOW, Schluter recommends a system they know will work 100% of the time. They do not intentionally suggest a system that might work most of the time and has years of info to back that up in both NA and Europe (where it has been in use longer). When they build a test floor, they use a single sheet of ply...testing has shown that most of the time there are failures in the real world, it's where two sheets meet on the joist...none of that occurs in the Robinson Floor test. Adding a second layer properly, isolates the tile installation from those failures.

For info, the Robinson Floor test when performed with a wooden subfloor, is still sitting on a structural concrete slab. IOW, it is testing only the deflection between the joists, not the combined deflection that occurs from both the joists AND the subflooring between the joists...so, while it is a destructive test (it keeps going, adding weight and repetitions until the floor cracks), and gives a reasonable idea of the subfloor and it's consequences, it is not the whole story of whether a particular product can work in any one situation. You can build your 'floor' with any method, but it is then installed on the concrete slab as an assembly for the test. You can just see the wooden subfloor on this example.

 

[Vegas_sparky]

I'm just guessing, John. You see tile, I see wire.

 

[Eurob]

You know John , they -- Blanke -- have more than the mat .


Is this looking familiar ? -- http://www.blankecorp.com//uploads/tx_rtblanke/Installation-Instruction-Aqua-Shield-USA.pdf


http://www.blankecorp.com/blanke-usa/products/waterproofing-and-multifunctional-underlayment-systems/


No anchoring of the U channel through the waterproofing --- http://www.blankecorp.com/blanke-usa/products/edge-protectors-for-walls-and-floors/blankeaqua-glas/technical-drawing/


I have an idea of the science behind the Permat .

Funny thing is that I don't find the distributors or where to buy on their site ......I guess more searching ....

 

[Jadnashua]

I hear a buzzing about dovetail designs above.... Have you guys seen the new heating system that Jim will be pushing soon? No dovetails. No Special Reinforcing Layers. Seems to me that this new product is flawed by Jim's own definition..... And the heating system goes above the Mat. Odd really....

Have you ever TOUCHED Ditra Heat? Obviously not. That's primarily the main reason I went back to the class - to get a look at what was new and to reinforce my knowledge. You should try it sometime!

It doesn't take much of a dovetail (and there are multiple layers of them in Ditra Heat) to lock a solid material in place. As to the wire, or a hydronic pex tube, those move with the tile and thinset independently of the substrate. THey are not the primary bonding point like on the mat you discussed above. I'll rely on the thinset rather than a thin fiber, to hold the tile in place, but that's me...you can make your own observation.

If you understand the dynamics of shear stresses in a tiled assembly, you'd realize that until you exceed the flexibility of the mat to absorb the stress, the stress on the tile is NOT the same as what it takes to tear the system apart. And, you should realize that the differential movement between the tiled surface and the substrate is generally measured in thousandth's of an inch. Within that range, the flexibility of the uncoupling membrane limits the stress on the tile/thinset bond to the elasticity of the ribs on the membrane...a VERY small number in the distances indicated. A tiled floor on an uncoupling membrane is closer to a floating floor than any bonded floor. Not understanding that, means test results of 'winged' unscientific tests are nearly meaningless. The test labs have the equipment and knowledge to test these things in a meaningful manner. I seriously doubt John Whipple can unless you goes to college for years and buys many thousands of dollars of equipment. So, take any results in that vein...not worth much IMHO, and I'd bet, any of the manufacturers with material in the test as well.

 

[Jadnashua]

The steps have a slight back bevel on them...not much, but it doesn't take much.

 

[Jadnashua]

Cement based products don't compress much...it doesn't take much bevel on any surface in the mat to hold things down. And, the larger squares offer more 'air' space to allow for any expansion/contraction experienced between the tile and the subflooring. If you want to waterproof the assembly, say you want to use it in a shower, you'd need to cover it with Kerdi.

 

[Eurob]

And all these years to say the tile has to be protected with an uncoupling membrane from the temperature swings and more from the heating wires ?!

Talking about the hydronic heating.....even with the bekotec , it still requires the Ditra on top , before tile installation .

 

[Jadnashua]

The actual mass of the thinset is small, and close to the tile...the tile and thinset with the wire embedded will act together. When you have the heat underneath in the subfloor or worse, a slab, there can be a bigger differential. When you use Bekotec, you have a moderately thick mudbed screed of material, and the response and reaction times are quite different.

The reason they didn't allow the heating mats or cables on top of regular Ditra was the unknown characteristics of each different heating cable or mat, and the possible risk of there being a void, or the accuracy of the heat control. It would be very difficult for the average person to get a reliable result and in the process might risk melting the membrane or burning out a wire. Only when you can control the materials selection and installation methods can you get things to work reliably. FWIW, the highest the thermostat will allow the actual tile to get is 82-degrees F, and there's a floor sensor buried in there near a wire to manage that. It is designed as a floor warming system, NOT a room heating system, like a typical hydronic, where there would be a higher BTU use and greater potential for differential stresses.

 

[Eurob]

The reason they didn't allow the heating mats or cables on top of regular Ditra was the unknown characteristics of each different heating cable or mat, and the possible risk of there being a void, or the accuracy of the heat control. It would be very difficult for the average person to get a reliable result and in the process might risk melting the membrane or burning out a wire. Only when you can control the materials selection and installation methods can you get things to work reliably. FWIW, the highest the thermostat will allow the actual tile to get is 82-degrees F, and there's a floor sensor buried in there near a wire to manage that. It is designed as a floor warming system, NOT a room heating system, like a typical hydronic, where there would be a higher BTU use and greater potential for differential stresses.

Wow.......the millions of heating cables installed with ditra on top ........and now this ( these ) ?!

Makes me think 10 times before using them again ....and I was sure of the assembly until ...now ....damn . ....

Roberto - Have you seen the new heating mat. I did not notice if there was a slight back cut to it. Would love a third persons opinion on this.

No John , just pics . But it is more to it than the dovetails . You need a megger to test and get a proper warranty before you install the tiles over it .

 

[Eurob]

And all the other mfgs allowing the uncoupling membrane to be installed over the heating wire ?

Unless it is a melting ice heating cable ......... what are the safe temperatures of the uncoupling membranes before '' melt down " ?


Using the Flexsnap from Flextherm , as per mfg design , is the alternative .

Flexsnap -- http://www.flextherm.com/en/products/flexsnap#installation

Flextherm -- http://www.flextherm.com/en/products

 

[Vegas_sparky]

If its a high grade polyethylene is should be around 300°F. That should be well above the temperature of a heating cable encased in a cementatious product. Having the wire on top of the mat/just under the tile would produce a more even, and efficient heat transfer. Underneath the Ditra profile the voids from the lugs could add a negative insulative factor. Small yes, but not neccesarily predictable. That could affect the service life of the heating wire itself more than the perceived performance.

 

[Jadnashua]

The heating cables are conventionally either covered with SLC or thinset prior to the insulation of the uncoupling membrane...same as Ditra always has been prior to the introduction of Ditra Heat, which is slightly thicker, locks the cable into a known position, and is surrounded by thinset in any properly install. Ever read about cables or mats floating after an slc pour noticed the next day? Has it happened to you?

FWIW, if you read the instructions for most of the heat mats, they recommend you check the insulation with a megger...Schluter, being conservative, does not want even that less than 1% chance the cable is defective or damaged after install before the tile, otherwise you'd have tear the floor up to fix. It comes down to being conservative and presenting an installation that is as close to certainly working. You can nick a conventional heating system's wire easily but it won't show up as a short. But, the Schluter system's thermostat contains a GFCI, so t doesn't take much of a fault to trip it...and the only way to fix that is tear up the floor. Is that something you want to do for a client, even if it is only once? Check it first, and you're golden. Same idea with supplying a second floor sensor...they do not fail often, but if you have a spare there, it's about a 5-minute fix with no tear-out. Which would you like to do? Again, conservative.

Keep in mind that each of the 'towers' have a shaped notch to lock in the cable. Since you cannot install the cables closer than 3 towers apart, and at best you may be making a 90-degree turn around each post at the worst, you have a dovetailed notch to lock in the thinset, that notch will lock in both the cable, and when filled with thinset, lock the tile to the membrane. And, look at the size of those towers...lots of free air space for LOTS of differential movement between the tile/thinset and the substrate.

Not BS, science.

 

[Vegas_sparky]

A second thermister isn't a bad idea. I deal with them all day long. Even the best don't last indefinitely. I'm just sayin.

I would think the stat/controller can also operate off of ambient air temp? Is the logic so that a failed floor temp sensor locks out operation to protect against the +82°F limit?

 

[Eurob]

If its a high grade polyethylene is should be around 300°F. That should be well above the temperature of a heating cable encased in a cementatious product. Having the wire on top of the mat/just under the tile would produce a more even, and efficient heat transfer. Underneath the Ditra profile the voids from the lugs could add a negative insulative factor. Small yes, but not neccesarily predictable. That could affect the service life of the heating wire itself more than the perceived performance.

When you approve the uncoupling membrane on top of the wires , the word is negligible -- for more than a decade -- , not unpredictable .

And how the service life of the heating wire is affected ? It is covered by the SLU before applying the uncoupling membrane .

 

[Vegas_sparky]

Coverage of the SLU can't be controlled(floating cables). That's unpredictable. Neither can coverage of thinset if its on top, unless you burn a layer in before you set. Those materials act as heats sinks for the wire, and their characteristics are engineered into the performance. Without those materials hot spots in the cable will develop, which affect there service life.

 

[Jadnashua]

Nobody says what has been done before is wrong...it means that there is now a new way to do it that is faster, easier, and potentially more reliable. Continue to use an uncoupling membrane beneath the mat, and it will continue to work, and is totally supported by Schluter should you wish to do it that way.

The thermostat has three sensing modes, one is just ambient room air temp. But, it still limits the floor temp to 82-degrees. Again, it is not a room heating system, it is a floor warming system. In some circumstances it might be able to suffice for heating, but you'd need an independent engineer to verify your room loads, just like should be done with any heating system.