cracks in new travertine floor: advice needed!

[KateME13]

Hello--I am a homeowner who has recently had a new kitchen installed in a Victorian house, including a new modular travertine floor (724cm x 224cm) over a suspended floor. I have outlined the materials used and where the cracks are--if anyone has any advice, we'd be really grateful, as it has us and our builder stumped.

The materials/installation
1.Six new joists new joists are 6” x 2” C16 grade joists spanning [the width of the kitchen]
2. The old joists are 4” x 2” supported at half span.
3. There is 80 mm Kingspan insulation between the joists supported by battens on top of which are laid 15mm poly pipes for the underfloor heating. [Query: might the point at which cuts are made in the joints for the pipes to cross be weaker and allow the floor to flex?]
4. On top of the joists is laid 600 x 2400 x 22mm engineering grade tongue and groove chipboard, screwed to the joists. This is laid, beginning in the corner by the utility room door, in a brick pattern (so that the joints are multiples of 600mm from the party wall) and either 2.4 or 1.2m from the utility room wall.
5. On top of this is 6mm Tilebacker, screwed to the chipboard. This is in sheets of 900 x 1500, again beginning at the corner near the utility room door so the joints in this are multiples of 900mm from the party wall and either 1.5 or 0.75 m from the utility room wall.
6. The tiles were stuck to the tilebacker with Mapei Kerabond adhesive.
7. Liquid ‘Fire and Earth’ sandstone grout was then injected to take up any space the tiles.

The cracks
Within days of the kitchen being moved back into, a crack appeared, c.90cm from the left hand wall, which has in the ensuing 5 weeks, traveled to a length of 560cm up the floor. It is now dissected by three other cracks, one at 300cm from the top (which starts under the kitchen cabinets so not clear exactly how long it is, but travels from the bottom of the cabinet (60cm deep) to 155cm wide), one at 450cm from the top of the kitchen (which travels from under the left hand side cabinet to under the right hand side cabinet, partially through grout, 163cm from cabinet to cabinet) and one at 600cm (again, not clear where it starts under cabinet, but 135cm crack is visible). There is another lengthwise crack which travels from 615cm from the top all the way to the bottom (to 724cm), and is 185cm from the left hand wall.

Our builder has offered to replace the cracked tiles, but we would like to try and work out what has happened in order to prevent it from happening again. Grateful for any advice you might be able to give.

Thanks.

[Sir Ramic]

First its not a floating floor. Its screwed to the joists.
18mm ply would have been better and stronger than chipboard but not that much of an issue.
Were backer boards laid with gaps between for expansion ? How were they fixed ? What type were they ?
Is they any noticeable deflection in the floor when you walk on it ?
Were any joists wet when installed.
Photos can help us identify your problem if you have any.

[KateME13]

Hi--Sorry--my mistake re terminology. Will ask the builder re backer boards spacing and type. The floor over the new joists does seem a bit more 'bouncy' (and sounds quite hollow) than over the old joists. To my knowledge the joists were all dry, but will check. Will also post photos later today. Also, we live near a railway line--could the vibration from that be having an effect???

[Stewart]

as above... also, has the heated floor been turned on and if so, was it turned up gradually to avoid shocking the adhesive and timbers

[Sir Ramic]

Also I am not sure the adhesive is correct, Mapei state...
Recommendations

Use Kerabond mixed with Isolastic in the following cases:

on foamed concrete walls
on pre-cast or cast-concrete structure
over underfloor heating installations
with large-format tiles
for the installation of glass/porcelain mosaics
for the installation of stone materials as long as they are stable and moisture proof.

[Sir Ramic]

I doubt very much that the railway has anything to do with this.

[Stewart]

you say the floor sounds hollow.... did you see the fixing method of the floor tiles... did they comb the floor or use large dollops of adhesive?

[KateME13]

Hi--Will check. He did test it, but will check. The cracks have all appeared well after he tested the u.floor heating.

[KateME13]

Hi--I didn't see, no. Will ask. Lots of questions for the builder...

[Sir Ramic]

the heating should have been tested before tiling. Then it should have been turned off for 3 days prior to tiling. The floor should have about 2 weeks after tiling before the heating is turned on again and at 5 degree increments until it reaches max temp.
Failing to do these things can shock the floor and cause issues.....I am still concerned over choice of adhesive and deflection in the floor .

[whitebeam]

Was the Tile backers board fixed onto adhesive before screwing down...

[Rich]

There are so many things that could have gone wrong, I will wait to see what the builder comes back with but my feeling is that the cement boards have been screwed down without any adhesive under them?

Also is there any chance of some pics? What ever happens the tiles are going to have to be lifted, if we could get some detailed pics of the cracks, the tiles before and after they have been lifted and the floor itself, we can get a much better idea of what has happened.

[KateME13]

Hi--

Here are some photos, but they don't really do it justice! Have been in touch w builder, hopefully will have some answers tomorrow. Thanks so much for all the posts so far.P1010119.jpgP1010118.jpgP1010120.jpgP1010117.jpg

[whitebeam]

A decoupling membrane may have helped, travertine and ufh..

[Phil Hobson]

Sorry to hear of your dilemma, as above so many possible causes, pics would help. Hope you get sorted

[Rich]

A decoupling membrane may have helped, travertine and ufh..

Never a bad idea. My money on either a joint in the chipboard with a bit of bounce (not sat over a joist) or no addy under the cement boards. Will have to wait and see what Mr builder comes back with.

[bugs183]

I'm guessinfg its one of two things.
a. The floor is not secured enough to the rafters below, so is flexing.
b. and I'm with Whitebeam on this, no un/decoupling membrane has been used.For tiling over a wooden floor there is no substitute for a decoupling membrane, Trouble is not many tradesmen, tilers included have heard of these so they get sold backer boards. What you have is 'defraction within the substrate', This means whatever you are tiling to is moving, If you were to dig straight down through the tile along the crack you will find this crack will 9 times out of ten be above a joint in the backer board. Wood moves, and backer boards stress at their joints,and unless its glued with flexible adhesive and screwed at least every 100mm centres it will crack. A de or uncoupling membrane if fixed above the wood will takes the stress within itself thus relieving the stresses in the wood. It is also recommended now that any underfloor heated floor has a uncoupling membrane included, though this happens rarely due to cost.

Google Schluter Ditra Matting and you'll see info on these kinds of mats. We NEVER tile to a wooden floor without them, and we've fixed hundreds of meters with no problems.

Solution, well it may stabilise, or it may continue to crack. I'd inform the builder of your problem, then decide if you want to leave it a while for the floor to stop moving (could be damp flooring or timbers), or there really is no other way than to dig the floor up, if its flexing its needs fixing, and its too late to add a membrane because of the height difference between the old and replaced tiles.
Ooh and if you fobbed off with 'its the tiles' it's not because the cracks are in an adjacent straight line, so this proves things aren't too happy under the tiles.

[impish]

Bugs 183 - great sum up.

I've nothing to add really except the main problem (in my opinion only) is using a builder instead of a professional tile & stone fixer.

Builders build. Except mine didn't. My garage roof fell off.

[charlie1]

Always a good idea to use decoupler with such a fragile stone when tiling on wooden subfloor however, in this case I think the root problem is the flex (up&down movement) in the floor, decoupler will only address lateral expansion. I think the floor will need to come up, along with subfloor and see whets going on with these joists

[KateME13]

Hi - Thanks for all your suggestions so far -- am still trying to find out the make of tile backer and whether it was laid with gaps. (May be a dumb question but if the tiles are fixed to the backer and the backer expands, won't the tiles detach?)

Meanwhile, last night we measured out where all the cracks were in relation to the joints of the tile backer boards and in relation to the chipboard t&g boards below those. In short, the cracks (about 14 m of them in total) run almost exactly along where we think the joins of the tile backer boards must be. The only exception is where we think they cross the middle of a board to join to cracks that correspond to joints at either side (as the boards were laid in a brick pattern - hope that makes sense). Is that information helpful?

In reply to Stewart's question of yesterday -- I think they fixed the tiles with dollops of the brown adhesive and then allowed liquid grout to run under the tiles and take up any space. I'm not 100% sure about this so will need to check with the builder.

Thanks again for all your help -- I really feel as though we are getting somewhere.

[charlie1]

Hi - Thanks for all your suggestions so far -- am still trying to find out the make of tile backer and whether it was laid with gaps. (May be a dumb question but if the tiles are fixed to the backer and the backer expands, won't the tiles detach?)

Meanwhile, last night we measured out where all the cracks were in relation to the joints of the tile backer boards and in relation to the chipboard t&g boards below those. In short, the cracks (about 14 m of them in total) run almost exactly along where we think the joins of the tile backer boards must be. The only exception is where we think they cross the middle of a board to join to cracks that correspond to joints at either side (as the boards were laid in a brick pattern - hope that makes sense). Is that information helpful?

In reply to Stewart's question of yesterday -- I think they fixed the tiles with dollops of the brown adhesive and then allowed liquid grout to run under the tiles and take up any space. I'm not 100% sure about this so will need to check with the builder.

Thanks again for all your help -- I really feel as though we are getting somewhere.

Sounds like the tiles are d&d, injecting grout to fill voids is new to me! Anyone heard of this method of fixing?

[Sir Ramic]

If he has indeed Dot and Dabbed ( a process where 4 or 5 blobs of adhesive are placed on the tile or even wall or floor) then you will have voids between those dabs of adhesive, this will be a weak spot. Any movement at all will be highlighted very soon with cracking tiles.

[charlie1]

Yeah, brown adhesive too! that will bleed through the tile in time and your floor will look like its caught the measles!! ...dont sound good at all!

[Rich]

Oh dear, as has been said, if the tiles have been D&D (tiled on blobs of adhesive) then the tiling has not been done correctly and grout is not a substitute for a full contact of adhesive under the tile.

To be brutally honest it sounds like the builder has messed this up from start to finish, why was a tiler not called in to carry out the work? I have never met a builder in all my years as a tiler that has the first idea about tiling but they all say "I do a bit of tiling" or "I usually do it myself"


Nobody but an experiencedprofessionall should go anywhere near natural stone, there is so much that can go wrong.

I just hope that he is good enough to get it put right properly. Just replacing these cracked tiles will not do anything. The tiles need to be removed and the problem in the floor sorted.

Sorry to have a rant but it is veryannoyingg when we get posts like this, as Impish said, builders build things and tilers should do the tiling. I really hope it all gets sorted with as little fuss as possible.

[Phil Hobson]

Amen to that Rich

[KateME13]

Thanks so much -- I'm amazed how much help we're getting from this forum. I now have the missing information that I hope will help with the diagnosis. To recap:

* the new joists are 6” x 2” C16 grade joists spanning 274cm (the width of the kitchen) spaced at ~30cm centre to centre. (Part of the floor is supported on original joists that are 4” x 2” supported at half span.)
* There is 80 mm Kingspan insulation between the joists supported by battens on top of which are laid 15mm poly pipes for the underfloor heating. The heating was tested at full heat when the chipboard was down but before the tiles were laid. It has not been tested since.
* On top of the joists is laid 600 x 2400 x 22mm engineering grade tongue and groove chipboard, screwed to the joists. This is laid in a brick pattern.
* On top of this are 6mm sheets of 900 x 1500 Hardibacker board (I presume this is the stuff). These were laid with min 1mm gaps between the boards (up to 2-3 mm) in a brick pattern, with the gaps filled with flexible sealant. They were screwed at 200mm centres (but not glued) to the chipboard.
* The surface is modular filled and honed Travertine from Topps tiles (see link)
* This is stuck with 'dotted and dabbed' Mapei Kerabond adhesive (this is white - I did them a disservice in an earlier post by saying I thought it was brown).
* Liquid ‘Fire and Earth’ sandstone grout was then allowed to run between and under the tiles to take up any void under the tiles.
* The tiles were then sealed.

The problem: there are several cracks in the tiles, each running across several tiles and all, as far as we can tell, running along the joints between the Hardibacker boards (or linking them). There are about 14m of cracks in all and they have lengthened progressively over a period of a few weeks since the floor was laid. The kitchen units have been fitted, so taking up the whole floor would be a substantial job.

I think this is pretty much all the information that is available. Does this help to confirm or change the diagnosis? What should we do? The builder is as puzzled as we are but has been very good so far about saying that he will fix the problem but obviously we want to understand the cause so that we can ensure that when he does it won't happen again.

Thanks so much for any help you can give.

[Sir Ramic]

* this is stuck with 'dotted and dabbed' mapei kerabond adhesive (this is white - i did them a disservice in an earlier post by saying i thought it was brown).

bingo !!!!!!

[Dave]

Sorry to say this but not only a bad fixing method but that adhesive is not suitable for timber floors with or without isolastic admixture.

It is suitable for UFH on solid substrates but deffo not timber.. this is fixer installation error.

[whitebeam]

The hardie board should have been laid onto a single part flexible adhesive before they were screwed..

The dot and dab is a major issue and the flowing grout is'nt good either...you will have more cracking later because of the fixing issue..time to consult your tiler.

[KateME13]

Hi--

But can you explain why the cracks follow such long lines if the d&d'ing is to blame for the problem?

Thanks again for the supportive and constructive posts on this.

K