Anhydrite adhesive test.

[bugs183]

Hiya fellas.
I've been on a large anhydrite floor and thought i'd do a little test whilst i was there.
It is by no means scientific and was done quickly, but this does copy conditions found on site and not in a lab.

The floor has underfloor heating, has been carbide bomb tested and was dry, and has been sanded and vacuumed. I have been using Kerrkoll H40 Ideal and was told the floor doesn't need to be primed.
I had to remove a couple of stone tiles that had been laid for about 4 days, the adhesive came up clean as a whistle, but stuck well to the stone, mmm i thought, i don't like this, i need to do a quick test.

So i cut some 150x 150 porcelain floor tiles for my test pieces, and used Nicobond Gypfix, Kerakoll H40 ideal (both gypsum compatible)and Dunlop Set fast Plus adhesive, which is actually a better adhesive than it sounds!! All bedding at 10mm and back skimmed.
I chose 3 areas to tile, one unprimed, one primed with SBR and one primed with Mapei Mapeprim SP primer which is an acrylic two part. I wanted to test Epoxy but none available in time.
On the unprimed area i stuck a tile down with Dunlop, another with Kerakoll (as i've been spec to lay the rest of the floor).
The second, primed with SBR i stuck tiles down with Dunlop, Gypfix and Kerakoll. I also stuck some Ditra down with Gypfix and stuck a tile with Gypfix to the Ditra.
On the Acrylic primed area i stuck tiles down with Dunlop, Gypfix, and another Ditra sample stuck this time with Kerakoll.
I left these for 5 days.

I then hit each one on the side with a 2inch bolster.
Well the cement based samples all came up very easily in on go with no residue on the screed. Priming made NO difference.

The gypsum compatible samples, all the tiles came off the adhesive with a quick hit, the backs of the tiles were free of any adhesive but left the adhesive stuck well to the screed. To remove this from the screed really took some doing and in some places removed chunks of the screed. Priming seemed to make no difference. Adhesive seemed bone dry.

The test pieces using Ditra were interesting. The tile came up first, but stuck to the tile and showed the grid off the Ditra, but was still very dark showing the adhesive wasn't dry. Then the Ditra could be pulled off, not too easily, and again the adhesive was stuck well to the floor.

So my conclusions from this little test are:
I didn't rate the way the Cement based adhesive stuck to the floor at all, very poor, but obviously it stuck well to the porcelain.
The Gypfix and H40, both stuck really well to the screed, but not too well to the porcelain (are we back to that issue again with Porcelain), the Ditra seemed to work well stuck to the floor, but obviously if used with Gypsum compatible adhesive with porcelain above there is drying issues.
Priming seemed to make little difference.

So I won't be using cement based adhesives to these screeds, despite advise from manufacturers.
I'm very unsure about tiling direct to these screeds with Gypsum based adhesives and Pocelain tiles. There was no strength in the bond between the two at all, underfloor heating could seriously cause problems there.
I'd be happy to stick Ditra down with Gypsum based adhesives and tile on top with cement based adhesives, i see no issues with this system.

So they you go, these are from a quick test, and are my opinions ony, food for thought though!

 

[bugs183]

Incidently the customer had a cement based Lafarge poured screed called: http://www.lafarge.co.uk/AandCReadymix/Agilia Horizontal Datasheet.pdf
This dried a perfectly flat and would be a great altenative to Anhydrite, it needs to be poured at 75mm, needs to be sanded as it is waxed to aid drying,and is a little less thermal efficient, but is that so important seeing the hassle it's gypsum based brother has given us?

 

[DHTiling]

Just to say... DO NOT prime and use H40 with Ditra or similar... it is a powdered dispersion type adhesive and if you prime then moisture goes nowhere and the mat will come up... i had this happen to me..

 

[Ajax123]

That's what I've been saying all along... ;p seriously though. Not sure how to get over the porcelaine issue. I can say thaough that the pics in one of my albums which show h40ideal are used on porcs and this job has never come back to me. I gree with Dave about the priming and ditra with it though.

 

[Gazzer]

This is all assuming the correct RH values have been given of course?

 

[Andrew]

I laid a 1200x1200 porcelain, the first 30m were fixed with a gypsum based adhesive. 1 tile had to be remove to allow for incompetent plumbing. As you said, the tile came up whole(ish) but adhesive was almost part of the screed no separation at all. The remaining 120m were fixed with ditra on gypsum addy tiles set in flexible cement based.
i returned to re set (paid as it was spec'd) the first 30m over a year later, Which 80% off were loose or hollow sounding, the rest of the floor laid on ditra was as new.
Therefore that is the only method I trust!

 

[jonnyc]

bugs 183. you have obviously spent some time doing this test and i am very interested in results.
just need a bit of time to take in all points made and will come back.
we are all struggling to get our heads round this and your test will help me understand more i think.

 

[bugs183]

So i dunno what proves what really!
I'm totally aware that these compatible adhesives are disperive based, and about trapping the moisture, and also the priming considerations.
This was a test to what happened in as many scenarious as i create, albeit quickly.
But that in itself brings us right back the issue that you don't fix porcelain with dispersive adhesive.
Addressing one issue has woken another problem up!
i'm still none the wiser and certainly still have little confidence in tiling to these screeds.

 

[Ian]

It would seem that if you are fixing porcelain, the only way to go is, use gypsum based adhesive to stick a decoupler then, use cement based to fix the tiles. Now that makes the job considerably more expensive for the client.
I've only done one job on this type of screed (200m2 across 8 rooms) and the whole lot was porcelain :yikes: I did however, do a couple of test tiles and found that the adhesive had bonded very well to the substrate and the tiles, the adhesive I used was from Creative Impressions. I was at the property again last week, over a year since the tiling was done and all was well, heating was on and there were no signs of loose tiles or cracking.

 

[Ajax123]

GBTA from creative impressions is probably the rolls Royce of the gypsum adhesives. I have seen it used many times and am unaware of any failures full stop. I have even seen it used on anhydrite screed which has been at very considerably higher moisture contents than it should have been without apparent ill effects. Seen it used to stick stone, porc, ceramic and reconstituted tiles such as quartz resin.

 

[Ajax123]

So i dunno what proves what really!
I'm totally aware that these compatible adhesives are disperive based, and about trapping the moisture, and also the priming considerations.
This was a test to what happened in as many scenarious as i create, albeit quickly.
But that in itself brings us right back the issue that you don't fix porcelain with dispersive adhesive.
Addressing one issue has woken another problem up!
i'm still none the wiser and certainly still have little confidence in tiling to these screeds.

Gypsum adhesives are not based on a dispersion mechanism. They are based on hemi hydrate calcium sulphate which when added to water hydrates to form gypsum. Gypsum needs to lose moisture in order to gain strength as the moisture sits at the interfaces between the interlocking gypsum crystals and effectively lubricates them. This reduces the amount of friction between crystals and reduces the compressive and flexural strength. The reason they work so well without primers is because there is no ettringite formation a the screed adhesive interface and the unprimed screed can pull the moisture out of the adhesive helping it to gain strength quickly.

 

[bugs183]

All i know is i have a limited capacity or need to retain so much information. I just want something that sticks!
I'm a tiler that wants to go work and be happy with what i've done and sleep at night.
Its a tad annoying having to work through different adhesives, but it dies sound like Creative Impressions were the biz so i'll get onto them.

 

[The D]

So are we now saying that cement based adhesive primed or not is not suitable for fixing tiles directly to this type of sub-straight ??????

 

[Rizzle from the Portizzle]

with so many types of these screeds comming on to the market its about time the screed surpler or the adhesive surplier gave us a spec in writing each seams to be passing the buck so who should be giving us the info and standing by it .i take my hat off to buggs for doing this test for us all putting in his own time and effort and sharing with us .its some think we all need to do .its a shame we are no wiser.

 

[bugs183]

God knows!
I believe if you use an Epoxy primer with cement based adhesives it should be fine, i couldn't test that one as i had no epoxy, but i may in the future.
Yet Norcross said if i were to use Ditra they don't recomend i use their gypsum based adhesive but prime with SBR and use cement based adhesives.

Any comments on the Agilia Horizontal Ajax?? That looked like impressive stuff.