Travertine floor

[woodie]

Hi all, care to give me any ideas? i've got a few of my own but we'll see. I laid a rather large travertine floor about six or seven months ago, when i laid it the floor looked and felt dry i gave it a scratch or two with my chisel and the scratches didn't darken so i primed the the floor with Bal prime APD left it all to dry out for the night and started tiling the next day, job took me about 4 1/2 days and it looked perfect after i'd finished, three or four months later i get a call saying that the tiles started popping i went and had a look and it started where the ufh was coming through from the utility room and following the course of the ufh kitchen areais fine (no ufh) as is the utility room any ideas?:incazzato:

 

[DHTiling]

What substrate is it and what adhesive did you use.?

 

[woodie]

it's that anhydrate(or whatever it's called) and i used white single part flexible:8:

 

[DHTiling]

Did you remove the laitance from the surface and how long had it been down..?

Also how did you go about priming it..?

 

[woodie]

removed the what? i primed it with a paintbrush:8:

 

[woodie]

:8:and it had been down for at least six weeks

 

[DHTiling]

Have you had no experiance with anhydrite screeds before...?.

As they are laid and dry , within a couple days or so a laitance is left on the surface and this has to be removed...mainly by sanding it off...then you use a 4:1 coat of acrylic primer and then increase the strength after each coat dries till you coat with 1:1 acrylic primer/water.

But the moisture in the screed MUST be low enough first before any prepping of the screed..0.5% if a recall correct...1mm per day for the first 50mm and 2mm per day for any thicker before you can tile direct..

Sorry to say this woodie...but you have stuffed up by the sounds of it.

 

[Ajax123]

Hi all, care to give me any ideas? i've got a few of my own but we'll see. I laid a rather large travertine floor about six or seven months ago, when i laid it the floor looked and felt dry i gave it a scratch or two with my chisel and the scratches didn't darken so i primed the the floor with Bal prime APD left it all to dry out for the night and started tiling the next day, job took me about 4 1/2 days and it looked perfect after i'd finished, three or four months later i get a call saying that the tiles started popping i went and had a look and it started where the ufh was coming through from the utility room and following the course of the ufh kitchen areais fine (no ufh) as is the utility room any ideas?:incazzato:

Oh hells bells ...Lots of stuff going on here... My first instinct is too much moisture in the screed but that may not be it so let's go from the start. Do you know who the screed supplier (e.g. Lafarge, Tarmac, Cemex etc.) and what area of the country is the job in. How deep is the screed, was the underfloor heating up and running before you tiled (assuming it is a screed embedded system.

 

[brian c]

😳😳😳

 

[woodie]

no the heating wasn't even conected up before i laid it and the builder told me 'this stuff dries very quickly and he wanted me to lay the floor two weeks after he put the stuff down, this is in south wales but i don't know the supplier. i don't know about this laitance malarky but the floor had been walked all over for at least a month before i started layin it and it had more than one coat increasing the valume of primer ,the owner had this checked out apparently and they said the floor was still damp, 'not in my opinion:8:

 

[DHTiling]

they said the floor was still damp, 'not in my opinion

Did you test it with a damp meter..?

 

[Ajax123]

Have to pop out for a while. Will deal with this when I come back.

 

[Diamond Pool Finishers]

hay ajax any chance of telling us on the forum a bit more about these screeds m8, like where there used and why ,and any stuff you know cheers fella

 

[CJ]

Laitance
A layer of weak material containing cement and fines from aggregates, brought to the top of overwet concrete, the amount of which is generally increased by overworking and over-manipulating concrete at the surface by improper finishing.



Its like tiling onto Polished Plaster Woodie.............You should have ground the top millimetre to roughen it up and loose the shine </U></I>

 

[Ajax123]

Laitance
A layer of weak material containing cement and fines from aggregates, brought to the top of overwet concrete, the amount of which is generally increased by overworking and over-manipulating concrete at the surface by improper finishing.



Its like tiling onto Polished Plaster Woodie.............You should have ground the top millimetre to roughen it up and loose the shine </U></I>

please when quoting make sure they are facts . you should NOT need to gring off the top few millimeters of any screed unless there is a problem with it. Anhyrdite and Alpha Hemihydrate screeds are no different.

 

[Big D]

I have always scuff/sand the top of anhydrite screeds so to give a key for acrlic sealer as dave said also the primer provides a barrier between addy and substrates as the products can react with each other and lead to addy failing!!

 

[CJ]

Ok.........give it a bloody good rough up then :hurray:

 

[Ajax123]

Her begineth the first lesson

Screed Design

Moisture Management is the responsibility of the main contractor and should be taken into account during the design phase of the project. I say this because I am seeing a massive upsurge in their use in the smaller end of the market where these issues are invariably ignore either through lack of knowledge and understanding or on rare occasions through mismanagement of expectations.

With any type of screed there should be a Damp Proof Membrane installed between the screed and the substrate. The cheapest and easiest way to achieve this is by placing a 1200g polythene sheet directly over the top of the major substrate....usually a concrete or block and beam floor (although other substrates clearly use screeds as well). Nature does not like inequality and tries to equilibrate them. As such moisture vapour will move from high pressure towards low pressure. A new concrete floor will take up to 12 months to dry out and a block and beam floor is unlikely ever to be fully dry. If this Damp proof membrane is missed out there is a risk that substrate moisture will therefore rise within the floor system as the vapour pressure in a room is likely to be lower than that in the floor makeup.

With heated screeds it is equally important to consider the effect of the underfloor heating. Hot things go upwards. There will always be some moisture on screeds and the effect of heating this moisture forces it upwards. so the top part of the screed will always extort the highest vapour pressure. This is important when considering flooring materials but more on this shortly.

Any insulation should then be placed in order to provide thermal resistance to the floor. The insulation should be reasonably flat and level to the substrate because if it has voids underneath and rock about this movement can cause the screed to depress and crack. Whilst not a common cause of screed cracking it does happen occasionally. The other consideration for the insulation is the level of compression it will suffer under loading. This should be taken into account again in the design phase. If the insulation is too compressible (spongy) it can compress under loading and the screed will deflect again giving the potential for cracking. For screeds which are deeper than the insulation I would suggest a minimum Compressive stress at 10% Deformation of 70 KPa and for screeds which are the same depth or thinner than the insulation of 100 KPa.

There are other design considerations regarding the insulation which are too many to go into here but include things like R Value, Environmental sustainability, fire ratings and many others - maybe another thread later

Once the insulation is selected then a second membrane which may be a separate minimum 500gauge polythene sheet or may be a prelaminated woven membrane on the insulation should be placed over the insulation to act as a slip membrane. This allows the screed to move freely and independently of the insulation so is important with heated screeds. With flowing screeds it also serves to keep the screed on top of the insulation where it belongs.

A perimeter edge strip should be placed around all edges, abutments and columns. This is normally a compressible etherfoam type material which is between 5mm and 10mm thick. This allows the screed to expand and contract in an unrestrained manner thus helping to minimise the risk of restraint cracking.

Movement joints should be placed within the screed preparation phase to take account of thermal movement with heated screeds. The maximum bay size for calcium sulphate screeds is generally 300m2 with an aspect ratio between 5 and 8 to 1 depending on the configuration of the room in question. Additional joints should be placed across all door thresholds, between independently controlled heating zones, between heated and unheated areas and abrupt excessive changes to aspect ratio e.g. an L shaped corridor. Special consideration should be given to areas where excessive thermal gain may be experienced e.g. large conservatories, glass atria and the like.

With heated screeds the polythene membrane should not be considered as a DPM because the attachment of the heating pipes using horseshoe staples make holes in it so it becomes a sieve.

Where the pipes run through the joint formers they should be sealed using a flexible sealant so as to offer a full thickness isolation joint between bays.

ALL OF THE ABOVE DESIGN CONSIDERATIONS ARE IDENTICAL FOR SAND CEMENT SCREEDS AND CALCIUM SULPHATE SCREEDS ALIKE except where I have specifically mentioned calcium sulphates. Bay sizes for heated sand cement screeds should be 15m2 according to NHBC handbook chapter 8.3.

The over riding design standards are BS8204 part 7 for flowing self smoothing screeds of which anhydrite and hemi hydrate are examples, and BS8204 part 1 for sand cement.

COMES UP FOR BREATH

 

[Ajax123]

Here beginneth the second chapter


Once all of this is designed and placed the screed can be installed with relatively little risk. I acknowledge that it sounds like hard work but then I am quoting best practice guides and corners are often cut for commercial reasons and due to a lack of knowledge (especially with joints) although anhydrite tends to be far more forgiving as it is much more flexible than cement based screed.

The design depth of anhydrite screeds should be 25mm Bonded, 30mm unbounded over a solid substrate, 35mm floating (domestic) and 40mm floating (commercial) and there should be a minimum cover to underfloor heating pipes of 25mm (most suppliers say 30mm to be safer) These depths are considerably thinner than sand cement screeds can generally be laid and the key to success in terms of drying is to keep the screed depth as thin as possible within these depth constraints.

The natural drying rate of sand cement screeds and anhydrite screeds is identical. Anhydrite systems can therefore be designed realistically to dry in a much shorter time than their traditional equivalents. However on the basis that any screed will only begin to dry once it stops getting wet the drying time is a function of the amount of extra moisture it is subjected to and also to the ambient conditions in which it is laid and "stored". The drying times are based on good drying conditions which are considered as 20oC and 60% RH with good ventilation. If the screed is placed and the room it is in not ventilated the atmospheric moisture will equilibrate with the vapour pressure within the screed and drying stops. The key here is to keep the vapour pressure in the atmosphere low and this is most easily achieved with ventilation.

Anything which affects the ambient moisture content will affect the drying of the screed both positively and negatively. One of the major advantages of anhydrite is that the atmospheric conditions can be manipulated in order to force dry the screed. This cannot be done with cement based materials because of the level of shrinkage and curl it encourages and you can end up with crazy paving.

Now to look at Calcium Sulphate Screeds specifically. There are three basic variants available in the market place all of which are available in South Wales. There is old fashioned anhydrite which forms a defined skin on its surface. This is a mixture of silt from the sand and some of the finer particulates in the anhydrite binder which rise to the surface of the screed with the bleed water as it compacts itself. This is commonly, but in my opinion, mistakenly referred to as laitance. I say this because all concretes and screeds have laitance, even sand cement screeds. In the case of these anhydrite screeds the laitance forms a thin crystalline skin which needs to be removed prior to the application of bonded floor coverings as it is not part of the screed itself. Once removed either by lightly sanding, scraping or in some cases sweeping, the surface will be open textured, sound and hard. If the surface of the screed has this skin on and it is not removed the floor coverings are likely to fail.

Secondly there are Hemihydrates and thirdly there are the new generation anhydrites which are skin free. These contain inert admixtures which stop the fines from moving to the top and thus no skin (or laitance) is formed. These screeds do not generally need to be sanded to remove skin cos it is not there. However they may need to be sanded for other reasons e.g. removal of plaster and mortar snots or other contamination likely to cause the failure of subsequent floor coverings. This is a very important distinction to make as the preparation for floor coverings i.e. the latter removal of non screed materials should rest with the flooring contractor (or suitable contract chain) whereas with the old fashioned stuff I believe it should be removed by the screed installer (again or suitable contract chain)

Laitance is much maligned and much misunderstood by the flooring industry which seems to think that to get floor coverings to stick it needs to be roughed up by Scabbling, grinding etc etc when in actual fact few flooring failures are related to the surface texture of the screed. The roughing up routine is usually an uneducated and largely vain attempt to overcome the effects of ettringite formation and moisture activity. It is in large part a wasted exercise. Lightly sanding the screed to remove surface laitance does help to speed up the drying so if it is present it is a good idea to sand it off early. Grinding and scabbling are likely to damage the screed and potential the embedded underfloor heating.

Prior to the installation of any floor coverings including primers and levellers the underfloor heating should be commissioned and run. This does a number of things, If the screed is going to move and crack this is when it will most likely happen so do you want to place 40k worth of marble on a screed then find it cracks later on splitting your tiles. It also forces a bit more moisture from the screed. Typically it will reduce a screed with 0.5% moisture down to approx 0.3% but the relative humidity at the surface will remain unchanged due to the increase in the vapour pressure within the screed at the lower content under the influence of the heating. Finally it will offer a double check as to the intactness of the underfloor heating itself.

Once commissioned the heating should be turned off prior to testing for moisture. It is impossible to determine accurately the moisture content of a screed either visually, based on the age of the screed (it may have been wetted subsequent to installation) or by scratching it with a chisel - not sure where that method came from. The correct method is to use a calibrated hair hygrometer. Other methods can be used but they present application issues. These are the Carbide Bomb Test and the oven dried sample test. I do not like either of these methods unless they are carried out correctly. You should not use electronic resistivity meters on these screeds a the presence of variable quantities of conductive salts in the surface of the screed interfere with the results and therefore yield inaccuracies. They can be used to determine the wettest areas so that the hygrometer can be placed at this point. I prefer a traditional analogue surface mounted hair hygrometer which cost around £80 ish to buy. The Final and much less scientific test is the polythene bag test which I use prior to commissioning potentially expensive accurate moisture testing. This involves a simple sheet of polythene placed on the screed for a couple of days. If when lifted the screed or the polythene has moisture on it the screed is clearly not dry.

Moisture is the single biggest issue facing the flooring industry and is the most common cause of flooring failures although is rarely directly attributable to the screed itself. It is more often introduced moisture or retained moisture present as a result of poor drying conditions.

Moisture management is the responsibility of the main contractor but the flooring contractor is wise to take some self responsibility for this because he usually has a duty of care to highlight deficiencies in the process - especially in the self build type market where the client is unlikely to have the necessary expertise to deal with these issues.

 

[Ajax123]

Chapter 3 commences.


Calcium Sulphate screeds should not be allowed to come into contact with cement based materials because of a chemical reaction which is well documented and causes the formation of ettringite salts which cause the cement to become delaminated form the substrate.

I believe that cement based adhesives should not be used on these screeds but acknowledge that they are by far the most commonly available and easiest to source. If possible the cement adhesives should be replaced by Gypsum Based materials. As no chemical reaction occurs an acrylic primer may be used. If cement based adhesives are to be used then a suitable chemical barrier should be placed between the two materials. This can be achieved in perfect circumstances with an acrylic primer but I have seen far too many failures where this type of primer is used to be entirely comfortable with it. There are good and not so good quality acrylic primers. A much more robust method of priming when using cement would be to use a water dispersible epoxy. These are more expensive and not available from all manufacturers. Any warranty for performance is clearly the domain of the material supplier but in my own experience I have seen very few failures where either gypsum based adhesives have been used or alternatively epoxy primers and cement adhesives have been used. Conversely I have seen many failures with acrylics.

Gypsum adhesives are not widely available and I am not in the business of selling someone else’s gear but if you need sourcing info PM me and I will try and help.

There are probably numerous contributory factors involved in this particular delamination event but moisture is likely to be the prime route cause.

Someone once told me screeding was simple and low tech in a building... this lot is only a bit of it.......:yikes:

The irony now of course is I just pressed the button saying post quick reply..... :lol:

Here endeth tonights sermon....:hurray:

 

[Ajax123]

hay ajax any chance of telling us on the forum a bit more about these screeds m8, like where there used and why ,and any stuff you know cheers fella

You asked for it Gooner....

 

[Colour Republic]

Where the pipes run through the joint formers they should be sealed using a flexible sealant so as to offer a full thickness isolation joint between bays.

Thanks Ajax 'Wiki' 123 you really are a wealth of info

Explain this bit to me?

EDIT oh my god there's more!

 

[Ajax123]

Thanks Ajax 'Wiki' 123 you really are a wealth of info

Explain this bit to me?

EDIT oh my god there's more!

The joint former should be cut (or preformed if available) to allow the pipes to pass through it. The "hole" through which the pipe runs should ideally be sealed with a flexible sealant (e.g. Silicon) so that the screed on one side of the hole is divorced from that on the other side thus creating a full depth full isolation joint. This is belt and braces and more often than not I don't even see these joints let alone sealing the run through.

 

[White Room]

I can get my breath back now:lol:

Some quality information Ajax

 

[Colour Republic]

After I've had a little nap, I might look up 'hair hygrometer' not got one of them, why does it need to be blond hair?


Cool I can build my own :clap:

Broken Link Removed

 

[marbleman]

let me say to Ajax123 great reply and full of information, just shows how important it is to do the job correct, I will say your advice is the best i have ever seen on this site , well done
regards
marbleman

 

[woodie]

hi Ajax123, phew that lot just blew my mind, appreciate your time and info'.
As for this laitance there was no film over the floor before i laid the tiles and it had good clean anyway to get rid of all the plaster and rubbish all over it and i am very fussy about giving the floor a good sweep before i lay tiles, as for the perimeter all that was there looked like dpm which was about a mm or two thick, the room itself was an extension to the main house off the kitchen area in which there was a blooming big patio door, the weather,well it was cold for weeks during the drying time for which i ensured there was a minimum of one days drying per mm of screed hence i could have laid it after thirty days but i left it for forty two, cheers all:8:

 

[Nardini]

Excellent advice Ajax123. Now, I only need to know what to do when ambient temperature is running at 45°C... :smilewinkgrin:

 

[Ajax123]

Excellent advice Ajax123. Now, I only need to know what to do when ambient temperature is running at 45°C... :smilewinkgrin:

Buy a fan and an ice lolly :drool5:

 

[Ajax123]

hi Ajax123, phew that lot just blew my mind, appreciate your time and info'.
As for this laitance there was no film over the floor before i laid the tiles and it had good clean anyway to get rid of all the plaster and rubbish all over it and i am very fussy about giving the floor a good sweep before i lay tiles, as for the perimeter all that was there looked like dpm which was about a mm or two thick, the room itself was an extension to the main house off the kitchen area in which there was a blooming big patio door, the weather,well it was cold for weeks during the drying time for which i ensured there was a minimum of one days drying per mm of screed hence i could have laid it after thirty days but i left it for forty two, cheers all:8:

It sounds like you were dealing with either a hemi hydrate or a skin free anhydrite. You would know pretty much immediately if it was traditional anhydrite as the skin is pretty obvious. Bear in mind in terms of the drying phase the screed will only dry when it stops getting wet. The plastering, decorating etc all put moisture into the atmosphere which will increase the smabient RH and so slow drying. The absolute key to this is ventillation as this removes the moisture from the air over the screed. It is very important to do a moistrue determination in all screeds prior to installation of coverings whether tiles or vynils, wood or carpets as the moisture movement can disrupt the adhesive properties, primer properties and the migration of disruptive salts. This latter is best minimised by the use of gypsum addy although I appreciate it is not widely available. It is however no more costly than cement. you could also use a decoupling membrane to stabilise tehe vapour pressure over the screed and divorce the tiles from it. This is a useful thing if the screed has been badly installed. I still suspect this delamination event is the result of moisture and a less robust primer adhesive combination than could have been selected.