porcelain adhesion

[tbb]

What can go wrong with setting 2"x2" porcelain with unmodified mortar over Ditra?

The last 4 SF of tile I set in the bathroom did not adhere and I have been able to pick it off with my fingers. I found this as I was cleaning the joints in prep for grouting.

- I used 1/8" DITRA, DitraSet, DALTILE 2x2 Keystone porcelain tile.
- I washed the tile sheets for good measure and let them dry prior to setting.
- I can see the towel ridges flattened off on the top of the mortar under the tile.
- There is the slightest bit of thin set on the bottom of the tiles and washed off under water while rubbing with my thumb.
- The DitraSet dried hard and adhered to the DITRA.
- There were some tiles that were 100% embedded. No mortar ridges just flat. These were accent tiles and had been back buttered.
- Most curious is that this installation overlapped some the same tile along one edge and the DitraSet is firmly adhered to the top of those porcelain tiles. Admittedly I scored the tops of those tile.

Best guesses:
- old mortar (bag was maybe 6 months old and had been open for 3 months ... there are no hardened granules in the powder)
- thin set skinned over before pressing in tile
- thin set was too wet ... it was mixed so that it would just start to sag on a trowel held sideways
- mortar drying too fast ... this section of tile was done after a change over to heating for winter

Ditra and DitraSet can not imagine anything other than the "best guesses" above that could go wrong.

Thanks

[rmckee84]

Sounds like thinset moat likely skinned over. Did these tiles have a glued on mesh backing? If you wash the back it could have smeared glue all over the backs and then inhibited the bond.

[tbb]

Ryan,
Not a mesh as I understand it. Only 2 rubbery dabs at the sides.
I did not wash the failed tile differently than the other 15 SF that have not failed.

[rmckee84]

Got it, then definitely would lean toward either old mortar or skinned over thinset.

[Kman]

That's usually a sign that your thinset had started to dry before it was covered.

[MAPEI - Technical Service]

I don't know the root cause of your failure but it cuts to the heart of why the tile industry recommends polymer modified mortars for porcelain. The tiles you are describing typically have a flat back and close to 0% porosity, there is nothing for a mortar to mechanically 'grab' onto. When there isn't a mechanical bond with the cement, there needs to be an adhesive bond with the polymer. Since you don't have any polymer, you don't have any adhesive bond.

[tbb]

It was one of two things.
1) I back buttered the tile, which was a pain on 2x2 tiles, by pressing a thin coat of DitraSet on the back using a small putty knife.
... and
2) I covered the area with plastic, for two days, with the intent of slowing down the curing. The plastic was still holding moisture on the surface of the tile when it was removed.
---I think I read or heard that Ditra is now allowing modified thin-set (?) after years of saying it would cause a tile failure. That would be worth back checking.

I know 1 & 2 worked because one of the black accent tiles twisted sometime after being set. It not only could not be popped out I had to use a roto-hammer to break it out in pieces.

3) The two above solutions could point to not sliding/twisting the tile enough into final position so the the thin-set did not smear the back side. I would still cover the installation for a slow cure - on small tile at least.

[jadnashua]

A porcelain tile is very nearly impervious, and so is the Ditra, so covering the area won't make much of a difference in the overall bond strength in this situation.

I've found that just spreading a coat of thinset on the back of porcelain may not actually do much unless it is done just so. You need to use a fair amount of pressure to get it embedded into the pores, and, if done right, the only way to get it off is to scrub it off (assuming it's not cured!), otherwise, it sort of tends to actually ball up a little if you get what I mean, and not really stick well.

There are a couple of types of modified that will work on both Kerdi and Ditra, but, the USA ANSI spec does not differentiate that characteristic, so Schluter was unwilling to let modifieds used, since, if they contained latex, they had problems, or at least could. But, there are 3 other classes of modifiers that can be used in thinset that could work...it's just that you can't tell by the ANSI spec (the European spec does differentiate). An unmodified, installed properly over Ditra or Kerdi does work, as will certain modifieds. One reason why Schluter came out with their own is to ensure people had an easy choice of one they know will work. That makes it a bit easier when you're dealing with a tile, such as glass, where the manufacturer insists you use a modified. FWIW, they've set glass tile for centuries before modified thinsets became available...done right, it works.