Crack repairs to stone floors

[markcollings]

Hi everyone. I am a Building Surveyor and have been reading your forum with interest. I am currently dealing with two properties where limestone tiling laid over heated screeds has cracked. I have uncovered the screed in both properties and found the cracking indeed extends through and seems to be as a result of shrinkage due to insufficient drying and heat cycling of the under floor heating prior to laying the stone tiling. In both instance no separating layer such as Ditra Matting was installed.

In both instances I need to specify a repair to the stone floors. In both properties this will require cutting out and replacing approximately 15 stone tiles. I am happy that the cracked tiles can be cut out and removed and new tiles bedded back in after with the floors refinished to avoid the new stones looking out of place. What I am concerned about is continued movement of the cracked screed below causing the new tiles to crack.

I have read with interest the threads relating to crack repair techniques including epoxy resin and steel needles which seems a sensible option to me. Hopefully this coupled with a flexible adhesive will help to prevent any further cracking of the replacement stone tiles. Has anyone used this technique as a repair for a damaged floor before and if so was it found to be successful? Does anyone have any trade info on products that are suitable for this type of repair.

The only other option in both instances will be to lift and relay the entire stone floor over Ditra Matting which would be incredibly costly as in both instances expensive kitchens would need to be removed.

I would appreciate any views members have on appropriate and lasting repairs.

Many thanks,

Mark

 

[jay]

hi only real way to fix is rip up and do right as cracks connot be patched with one tile span any pics:8:

 

[Ajax123]

The important thing is to determine why the screed has cracked as undoubtedly these will have reflected through the tiles. Is the screed flexing? Is it simply thermal expansion? Is it curling? there are lots of things which can cause screeds to crack and heated screeds are quite complex animals. The usual thing is that the screed has not been jointed into suitable sized bays. What type of screed is it?

If it is a cement based screed the bay sizes should not exceed 15m2 (Chapter 8 NHBC handbook) or room size whichever is smaller. If anhydrite it can be up to 300m2 but in both cases it is important that the screed is jointed at the interfaces between underfloor heating zones. This is something I coe across as a problem in many installations as installers don't seem to realise that the screed needs joints.

If the screed is repaired and it has cracked because it needs a joint then the crack will simply re-locate to the next weakest point.

I disagree with Jay in that the whole thing should be ripped up. It does depend however on the reason the cracks have appeared as to whether a resin stitch will be successful.

 

[Terry Cottar]

I have in the past retro fitted expansion joints. It involves tiles coming up but not a total refit you can usually see the weak point especially with natural stone. Limestone requires the correct filler.

tel

 

[Floor_Restorer]

Underfloor heating and cracks in Limestone go hand in hand!

Being based in The Cotswolds we see quite a lot of this. Ajax123 is right in identifying the cause and putting in measures to make sure that the repaired floor can cope with the constant expansion and contraction of underfloor heating.

With a standard screed or concrete floor that is not jointed between floor areas causes a number of instances to happen with Limestone. Often you will see cracks in doorways or narrowings, as well as in floors that do not have a floating element to them want to expand out of the weakest point. We used to install floors and any thing with heat involved had to have Ditra matting and jointing in the screed Easy to say that now..........

Also, I know this sounds stupid, but always use flexible adhesive and grout with underfloor heating. You'd be amazed at the number that haven't.....

Could you post some images?

Thanks,

Tom

 

[DHTiling]

Mark...this thread will help you understand how substrates behave...

http://www.tilersforums.com/tiling-forum/14730-understanding-tiling-shift.html

 

[markcollings]

Thanks all of you for your input. The primary reason for the cracking seems to be insufficient drying of the screed and heat cylcing prior to laying the stone tiles. The screed was laid very thick (around 125mm) and given little less than 6 weeks to dry before tiling. Also the underfloor heating was not turned on prior to laying the tiles but some 3 months later.

When the heating did finally come on it was in an uncontrolled manner with sharp rises and drops in temperature. This I suspect caused some rapid curing of the undried screed and resulted in some significant cracks appearing relatively quickly. I will post some pics next week.

In answer to AJAX123 comments about jointing, the screed is a cement screed and critically no movement joint was provided between two rooms, both of approximate size 6m x3.5m. There is a 2.5m wide opening between the 2 rooms and cracks have appeared in a Y formation extending out from either side of the opening, meeting approximately in the middle and then extending through the centre of one of the floors. I also confirm that the two rooms have independantly zoned heating.

Give the above it seems sensible to try and introduce a movement joint between the 2 sections. If anyone has any experience in retro forming movement joints in screed it would be great to here your comments.

Where cracks have formed elsewhere in the rooms (they are fairly numerous) I am suggetsing a straight epoxy and needle stitch repair.

 

[Ajax123]

Thanks all of you for your input. The primary reason for the cracking seems to be insufficient drying of the screed and heat cylcing prior to laying the stone tiles. The screed was laid very thick (around 125mm) and given little less than 6 weeks to dry before tiling. Also the underfloor heating was not turned on prior to laying the tiles but some 3 months later.

When the heating did finally come on it was in an uncontrolled manner with sharp rises and drops in temperature. This I suspect caused some rapid curing of the undried screed and resulted in some significant cracks appearing relatively quickly. I will post some pics next week.

In answer to AJAX123 comments about jointing, the screed is a cement screed and critically no movement joint was provided between two rooms, both of approximate size 6m x3.5m. There is a 2.5m wide opening between the 2 rooms and cracks have appeared in a Y formation extending out from either side of the opening, meeting approximately in the middle and then extending through the centre of one of the floors. I also confirm that the two rooms have independantly zoned heating.

Give the above it seems sensible to try and introduce a movement joint between the 2 sections. If anyone has any experience in retro forming movement joints in screed it would be great to here your comments.

Where cracks have formed elsewhere in the rooms (they are fairly numerous) I am suggetsing a straight epoxy and needle stitch repair.

All sounds sadly familiar. Where the 2 rooms meet it could be possible to retro fit a joint at this point. It would simply be a case of uplifting the tiles and saw cutting carefully - you don't want to cut into the pipes. To form a joint effectively it needs to extend approx 1/3 the depth of the screed so if they are 15mm pipes in 125mm screed you should be reasnably safe. This of course will then be reflected through the tiles. You can resin the y shaped cracks as the resulting bays will be in the right ball park and aspect ratio for this depth of screed.

Interesting point is that Y shaped cracks across dooorways often appear as a result of restraint as much as drying shrinkage suggesting the inadequacy of or the absence of the perimeter edging. Typically with underfloor heating this should be a minimum 8mm and etherfoam is best for the job.

The heating should have been commissioned prior to tiling to condition the screed. This should be not be done with sand cement for 28 days following installation. As you said, doing this in an uncontrolled manner is disastorus especialy when the screed is still significantly wet.

 

[jay]

hi if floor heating is turned on at a high setting to start with you will crack your substrate floor heating should be run very low to start as to dry out all moisture slowlly ive seen a bit of it down here with electric heating built into house slabs :8:

 

[markcollings]

Thanks Ajax,

I think cutting the movement joint in across the door opening is the best option. Can you elaborate on the perimter edging comment? I believe in this particular scenario insulation was turned up around the edge of the screed but I think this was just a thinner version of the expanded foam insulation used beneth. Can't recall the thickness.

I found some information in BS 8204-1 on screed curing. It seems that for a 125mm screed the suggested curing time is 5-6 months, therefore it seems the curing time in this instance was woefully under by about 3 months. I note that the BS recommends that heating should occur slowly once the screed is fully cured and dried i.e. early heating is not a subsitute for allowing the correct drying time hence artifical drying is not recommended.

I have also obtained the following advice about heating sequence:

"The screed should be maintained at the opertaing temperature for several days before cooling down to room temperature (but not below 15 degrees) before installing the flooring. BS EN 1264-4:2001 states initial heating up shall be carried out at least 21 days after laying the screed as follows:

- Use supply temperature of 20 degrees to 25 degrees for 3 days
-Set maximum design temperature for at least a further 4 days"

I am in the process of obtaining the BS referenced above as it seems crucial that this information is adhered to, to prevent problems from occuring. In future I will include an outline of this information in any specifications.

It seems that in order to get things right and prevent the risk of cracking the sequencing and process for forming the screed, curing and drying and heat cycling prior to laying the stone floor is essential. On busy building sites when quite often time pressures means that work is undertaken out of sequence and timings are condensed it seems that the risk of getting it wrong is quite high. I presume therefore the use of products such as Ditra Matting reduce the risk if the correct sequence and timing is not followed.

Regards,

Mark

 

[Ajax123]

BS 1264 gives some very dangerous advice if you follow it to the letter as it misses out the gradual heating up process. If you follow this advice it is likely to thermallyshock the screed which will crack it as sure as eggs is eggs.
It ought to say put on at ambient for 3 days, increase by maximum 5 degrees per day up to maximum operating temperature and the reduce by 5 degrees per day to normal operating temperature. Run heating until screed is dry. With Anhydrite the process can start after 7 days. With sand cement it should not commence for 28 days (21 days following initial curing and all cement based screeds should be cured for 7 days following installation using a suitable curing membrane - usally a sheet of polythene -although this advice is hardly ever followed).

The CFA are currently lobbying for changes to BS8203 (resillient floor coverings) to be changed at the next revision to reflect the corrct method of comissioning. As other standards come up for review I guess they will be harmonised (at least that's the hope)

As the screed dries it shrinks. It also moves about in various directions as it shrinks depending on the bay sizes. The bay size should be geared to an aspect ratio of no more than 2:1 and if heated no more than 15m2 is the NHBC recomended maximum bay size - it is here where people often say to me that the screed is reinforced with chicken wire or some such - This is completely irrelevent.

If the screed is placed around a restraint point e.g. a door liner it needs it will be shrinking in opposing directions. The effect of this is to pull at approximately 45 gegrees to the point of restraint. This causes the classic Y shaped crack you have seen here. The best way to avoid cracking in this instance is to place a joint in the screed through the door way.

If no joint is present and the heating is then turned on the effect is to create alternate expansion and contraction. The expansion strip is there to allow the screed to do this effectively unrestrainedly. If it is not there restraint occurs at the door liner also causing the screed to crack in the classic Y shape. The best material for this is expanded polyethylene (etherfoam) as it is easy to install around door liners and other re entrant conrners without breaks. This needs to be of sufficient thickness as to allow for the maximum expansion of the screed. I can't remeber off the top of my head what this is for sand cement but for anhydrite it is 0.012% - a minimum 5mm is therefore suitable for most applications increased to 8mm or 10mm for larger bays. The expanded polystyrene often used in the "cheaper" underfloor heating systems are difficult to turn round corners. They are therefore often cut into peices with a break at the corners which reduces their effectiveness considerably.

Ditra is a great product and deffinitely has a place in the market particularly for fast tracking projects or for helping to stabilise unstable surfaces. If you were uplifting the tiles in this installation and completely re laying I would suggest it in this one. However if the rules are correctly followed it would not really be necessary.

You are right that the underfloor heating being commissioned prior to the floor coverings is absolutely essential and something I actvely encourage as part of my work.

 

[Ajax123]

P.S.

People often mix up Curing and Drying - they are 2 completely different things. Cement based materials require curing. With sand cement screeds this is for a perod of 7 days. The reason is because the hydration reaction is relatively slow and the cement needs water in order to react. The curing therefore is the method of trapping the moisture in the screed to allow it to go properly hard. It is acheived by covering it with a sheet of polythene.

Drying is the exact opposite i.e. getting the moisture out of the screed. If this is done too soon i.e. the screed is not cured properly it will not have gained enough stregth internally to overcome the mechanical pressures of drying shrinkage and the screed will crack. The rate of natural drying in the UK is based on an ambient 20oC and 60% RH. It takes 1mm per day but can become eratic and extended at depths over around 75mm. It also depends on the original level of moisture and the level of compaction.

Anhydrite work completely differently and are in general much more predicatable. They are self curing and the reacton very quick being all but completed after 72 hours. They hardly shrink, don't curl and are very dimensionally stable so can be laid to much bigger bays. They are very thermally conductive compared to sand cement and so are used extensively with underfloor heating. The usual issue with these screeds is that they are laid much too thick extending drying times un-necessarly.

Here endeth todays lesson - questions on a postcard please.

Give me a call on Monday at work I should be able to sort you a copy of relevent BS - 07545 932723