Really Tricky Problem?... Why are these tiles Raising? [Archive] - Ceramic Tile Advice Forums - John Bridge Ceramic Tile


View Full Version : Really Tricky Problem?... Why are these tiles Raising?

12-14-2016, 10:42 PM
Really Tricky Problem?... Why are these tiles Raising?

In a commercial kitchen, 12" square terra-cotta tiles on the floor,
and they've been there, with no problems for over five years.
The entire tiled area is about 55' X 75', including walk-in fridge and freezer.
All of a sudden, tiles in one area are raising, separating from the
concrete sub-floor, creating random apexes up to one inch high.
Barely noticeable a week ago, in the next few days it spread rather rapidly.
Some tiles are cracked, some not. We pried up / busted a large area,
about 5' X 6', trying to discover a cause underneath the problem tiles...
Found no moisture, or roots or any other evidence of a cause.
There's an average of about 3/8" of grout between tiles.
In the outlaying area, where the raising was just starting to show,
I used a masonry grinder to remove the grout, to release some lateral pressure,
and let the tiles fall back to "level", hoping to halt further raising.
Didn't solve the problem... next day, more tiles were raising at the perimeter.
Most of the Raising is happening in an area that is parallel to an exterior
(tilt-up) concrete wall that is about 12 feet from the damage,
though some of the damage is appearing perpendicular to that wall.
There are no big trees outside of, and no cracks in: that exterior wall.
Problem area is about 40 feet from the refrigerator and freezer units.
SOMETHING is causing some lateral movement... Which Is Continuing.
The guy that laid the floor has allegedly been doing this work for 25 years.
He's never seen this before,... doesn't know what's causing this problem.
Any of yous experts got some ideas? Might extra-terrestrials be involved?
Thanks very much. :talk:

Sponsored Links

Houston Remodeler
12-14-2016, 10:51 PM

Welcome to the forum.

Is it getting colder outside where you are? Can you add your geographical information to your User CP (Linked in the dark blue bar above)

Look around the perimeter. Are the tiles grouted to this outside tilt up wall ? Are the tiles grouted all around the perimeter?

WHat you are describing sounds like classic "tenting"

argile tile
12-14-2016, 11:14 PM
here's my un-professional observations:

i agree - check tile is not in contact with any walls (should be 1/8" gap all the way around), that's just code to do so, and can be fixed by doing it

but i'd say what you said sounds drastic (1" raising) and not completely clear how many instances (post photos so others can get a better idea)

my idea would be you need a building/foundation inspector to check the building is safe BEFORE worrying about the tile. secondarily, the inspector might know how to remedy the issue off-hand. it may even be a municipal issue.

your intuition may be right it may not be the tile (that a pro installed it, and he didn't think it was possible).

12-14-2016, 11:25 PM
Fast track or lack of soft joints , expansion joints , etc . or a combination of all of these . Take a look.....

Popping tiles 1 (

Popping tiles 2 (

Dave Gobis
12-15-2016, 09:02 AM
A common complaint when there is lack of movement accommodation and/or tiling over control joints. There should be a 1/4" open or soft joint at the perimeter and every 20-25' if it is interior and they don't clean it with a hose. This is not exactly covered under building code. There are countless articles and technical bulletins on this subject.

Todd Groettum
12-15-2016, 09:56 AM
Large fields we always ran soft joints every 16-20 feet depending on the tile size etc....

And concrete control joints got cut nd soft jointed that could not be figured into the layout... ( cheap insurance against this type thing)

Also with large fields we stayed out 3/16 - 1/4" from all walls...

Standard thinsets will shear much faster on polished slabs than a modified and some modifiers are better than others...Keralastic was commonly used in areas where alot of weight would be rolled over the tile...Keraset was the Minimum used...We never used straight Kerabond in Large field jobs...

Dave G would have the shear numbers ( probly knows em all by heart ) ;)

12-15-2016, 11:37 AM
Wow. Thanks for the fast replies, people.

HoustonRemodeler, we're about an hour south of Los Angeles.
And,... it never gets Really cold here, in California.
I'll confirm today, but I think that... yes, the tiles are grouted all the way to the wall.

DaveGobis, it's all interior, and it's not cleaned with a hose, but they do wet-mop it every day.

Sorry that I don't have pictures of the actual problem, but...
Yes! That's exactly what parts of this look like.
This is a screen shot from your 2nd video, Eurob...

My floor is not moving so fast that one can watch the changes happening,
and I certainly haven't heard any popping, but the end-result looks like that.

OK, to stop this further damage, we should...
1. cut the grout around the entire perimeter, where there's a wall.
2. cut the grout line every 16 feet.
3. replace that removed grout with Soft / Expansion material.
What's your suggestion for the particular Soft material to use there?

Thanks Very Much, guys.


Houston Remodeler
12-15-2016, 11:45 AM
100% silicone caulk that matches the grout from

Their RTV sealant, ColorSil, is floor rated

Dave Gobis
12-15-2016, 05:46 PM
Mark, PM me an email address and I will send you 3-4 articles on the subject. CA don't get a free pass. This photo is entitled "Death takes a Holiday". Floors or walls, I got lots more.

12-15-2016, 06:08 PM
There is tremendous stress put on the cement joints and bonding surface if things can't move as intended. FWIW, the nominal expansion rate between a slab and typical tile is about 2:1. That usually can be accommodated if you have expansion joints and a decent thinset. Keep in mind that around the freezers and underneath the stoves/cooktops, it can get really warm or cold after them running for hours on end...this can emulate being outside and having the bright sun shining on a part of the install and the next section in deep shade. Things DO expand and contract with temperature changes. Then, if you've ever seen a microphoto of thinset, you'd see that it really makes its bond by growing thin crystalline spikes that interlock into themselves and imperfections in the substrate (the slab and tile). They can only flex so much. Overextend that stretch, and they start to break off. Break off enough, and you've lost your bond. That can take some time, depending on the particular circumstances. The picture is courtesy of Mapei. When modified, it adds a coating onto the crystals, which can act sort of like a bumper, but given enough flex, and they'll still break.

12-15-2016, 11:01 PM
That picture looks like a black and white of some wild grass :D

What is presented in the picture ? A cured sample of mortar -- type of - in what state -- powder or larger piece -- ?
If modified -- polymer or latex -- , how you can distinguish in between ?

Todd Groettum
12-16-2016, 07:50 AM
Dave, i hope nobody was standing under those when they let loose OUCH!!!

Dave Gobis
12-16-2016, 08:07 AM
That is why I titled it Death takes a Holiday. Have another one someplace where a shard buries into a deck.

12-16-2016, 08:13 AM
Mark ,

This one is of the ET's involvement version of it ,

argile tile
12-16-2016, 06:13 PM
i'm sorry i mean for residential (or commercial?), because tile expands/contracts at a different rate than frame memebers, it is code and required the tile+mortar not be in contact with frame members, and be some distance from them, because it would shift/damage frame of building

12-16-2016, 06:19 PM
What is presented in the picture ? A cured sample of mortar -- type of - in what state -- powder or larger piece -- ?
If modified -- polymer or latex -- , how you can distinguish in between ?

No aggregate, no modifier, a thin layer of cement that was wetted and allowed to cure. During the curing process, cement incorporates the water into the growing crystalline structure (ever, as a kid, dumped a lump into water and watched it grow into a crystal - sold in toy stores all over? same effect when cement cures)...that's what causes the cement to be able to bond to things and hold itself together...the spikes interlock and poke into imperfections of the things your are trying to bond together when you have it thick enough, and also grow around the aggregate which is required to help keep those fragile spikes from just fracturing.

A modifier ends up wrapping the whole mess with a cushion. It can also help make the whole thing less porous. That cushion helps prevent the spikes from fracturing, but they naturally have a little bit of flex. In some tests I've read, a quality unmodified thinset ended up more flexible than a lower quality modified.

Note, what we call Portland cement is like a chocolate cake...lots of different recipes...those crystals don't all look the same. Here's another one that you might want to use in cement that had a higher compression strength.

argile tile
12-16-2016, 06:27 PM
i agree with the above, i'm no pro.

it all surprises me. spanish use mortar and terra cotta all the time - and on roofs (heavy spanish roofs are terra cotta spoons with allot of mortar and last years). many say sanded grout is the right choice.

i can only guess the movement is:

a) you have a concrete floor with sections and movement between sections, but you say no

i wonder if that floor has compression/motions in "single pour floor" (ie, when loaded with customers and at night) or ie, some areas never warm but others are heated and cooled. you'd think mortar versus concrete were not a huge difference. on the other hand: i typically see store floors tiled in sections.

b) the person above is right: that a maximum dimension between (caulk/movable) joints will hopefully solve it

in addition to that i usually see tile floor in sections in area stores...

on second look (non-professional) i see this in TCA book...

Commercial floors subject to bending the choice is to install a cleavage membrane (slip sheet) under (2") mortar with tile on top
• over structural floors subject to bending and deflection.

Movement Joint Design Essentials - expansion joint spacing

• interior—24' to 36' in each direction.
exterior—8' to 12' in each direction.
• interior tilework exposed to direct sunlight
or moisture—8’ to 12' in each direction
• where tilework abuts restraining surfaces
such as perimeter walls, dissimilar floors, curbs, columns, pipes, ceilings, and where changes occur in backing materials.

i'm certainly NOT familiar with strip mall or mall store tiling, or contractor expectations as to expansion joints

12-16-2016, 09:02 PM
The actual expansion rates of most materials is microscopic...but, as you can imagine, given a large space, it easily can become measurable. The mortar bond also has a microscopic level of flexibility, but just like many things, exceed the range, and it breaks...exceed the flexations, it breaks. That's one reason why things don't always fail immediately...they weren't flexed enough times or far enough. Get a hot day after the ground was cold, maximum differential between the slab and the tile, maximum differential variation in expansion/contraction, the bond breaks, or at least weakens. Over time, it fails.

12-16-2016, 09:20 PM
I've understood the growing crystals of cement,for years. Never seen it under microscope though lol. Seeing that pic,and understanding what materials are in Roman cement,validifies my opinion of Roman cement,and what constitututes good aggregate.

Regarding this every event I've seen/witnessed tenting of tiles,it generally occurred during and after a cold spell. Tiles tight to baseboards,grouted. I've heard the gunshot element of me on one job,and have had several customers being awaken at middle of night to what sounds like gunfire going off.

In all instances,no membrane,tile butted to baseboards,grouted,and 3 days or longer cold I.E. 32 or below. Also,grey thin set was used,and was VERY EASY to remove from slab. IMO the weak link was the thinset and the concrete,cold being the bond breaker.

12-16-2016, 10:16 PM
Rich , the tenting effect must be a more localized -- geographical zones -- where the right conditions would allow the rapid shrinking effect -- cold days -- , but I think it also have to be combined with another element , which trigger the mess .

And about Roman cement , I think they didn't get the right formula of the ancient concrete , yet . If they did and you know of a link could you post it ? Ty.

Jim , yeah , I can see the differences in the thickness of the spikes . However , I would really like to see a real microscopic view of a cured unmod and a cured premium mod , not just a pure cement cured mix. There should be more besides only pure cement forming crystalline structure in the way mortars are design .

12-16-2016, 10:30 PM
Ok Rob...lets pretend tile is thin setted to say a very very smooth surface,glass. Glass let's say is 6" this glass won't crack/break. Tiles set in a 10x10 sq ft room,butted to baseboards,grouted. When it goes through expansion and contraction cycles,what's gonna give first?

Put water in ice cube tray,and then freeze it,what does it do?

Job I'm at tomorrow,I'll get a pic of example.

12-16-2016, 10:31 PM
Jim , yeah , I can see the differences in the thickness of the spikes . However , I would really like to see a real microscopic view of a cured unmod and a cured premium mod , not just a pure cement cured mix. There should be more besides only pure cement forming crystalline structure in the way mortars are design .

A modifier just coats the cement particles. It needs to allow the spikes to form, so needs to be flexible to a point, then it can harden on its own. A difference between the four general classes of modifiers used is part of the myriad of choices the thinset manufacturer can make, along with the type of cement, the inhibitors that slow its curing rate (no inhibitors, and you get something like hydraulic cement that cures in minutes), the mechanical size and consistency of the cement particles, its formula, the size and quantity of aggregate, and the ratio of modifier to the rest of the mix. That's one reason why there are so many different thinsets and concrete mixes out there...there is a huge amount of variables you can chose when designing the mix.

My guess is that a microphoto of a modified would just look like a blob, since the modifier would just fill in between the crystalline structure and mask its structure. any good photos you could add? I stole those from your presentation. Credit where it is due.

12-16-2016, 10:34 PM
Roman concrete been around forever. Mine your lime,fire your lime and wala you have Portland. Add some aggregate from the area I.E. Pumice lava,and you have an aggregate that cement would love to intermingle/bond with.

12-16-2016, 11:20 PM
Rich , concrete is not glass . I think there is more to the tenting effect -- another catalyst element or a mix of them -- , not only from normal contraction and expansion of the substrate . In the videos linked , there is brutal tenting . Therefore , if it is just normal contraction/expansion , all of the tiles installed on concrete slabs should react in the same way . But they don't . And cold could contribute to it , but we do have cold here too and I don't see this tenting .

I was just curious what are the other elements which add to it . Being prudent in designing the assembly should prevent it from happening . But if an uncoupling is used , why the soft joints , if lateral movement is absorbed by the membrane .

12-16-2016, 11:23 PM
Jim , so what you are saying is that you can see better under the microscope when the cured sample is in a powder state an not in a blob state ?

12-17-2016, 12:11 AM
Jim , so what you are saying is that you can see better under the microscope when the cured sample is in a powder state an not in a blob state ?
You asked about viewing a picture of a modified...the modifier coats the cement particles (assuming it is mixed right), and thus, would just look like a blob and you'd not be able to see much of any detail of how the cement cures. At least that's how I understand it. The modifier helps to protect the fragile spikes in the cement matrix and in the process fills in voids that can help it become more water resistant once cured/hardened. It can aid some with the adhesion like a glue, but by varying the type and quantity of materials can have as much effect, and it is not always a clear distinction on which one ends up stronger. By varying the gauge (particle size and consistency) of both the cement and aggregates, you can affect the properties significantly. It costs more to just end up with smaller particles.

If you flex things enough, things break...the range and speed of thermal changes are probably one of the bigger triggers that help determine when a tiled floor can fail. Sun, heat from say a stove or oven, a cold slab, shutting the heat off at night, then turning it up all will thermal cycle things. Lack of expansion accommodation will eventually cause the bond to fail and the compressed material will rebound catastrophically. It acts like a compressed spring. Prior to that, the thinset bonds things together as best it can, but it can't take that amount of stress forever unless you limit it with expansion accommodation. How much you need depends on the conditions in the field.

12-17-2016, 05:47 AM
Robbie,I was using glass as an example. Smooth concrete verses rough concrete.

What answer are you really looking for? Are you waiting to hear that it's magical tile elves,causing the tenting/shearing of the bond?

12-17-2016, 09:13 AM
Very funny Rich ....... magic concrete produce tenting :yeah:

If '' brutal ' tenting is taking place on a small size room -- 10' to 16' normal room -- at a specific location , near the outside wall -- can't find the video now -- , there must be something else going on , besides regular expansion or contraction of the substrate . The stress shows opposite patterns , therefore the question(s) or the interest .

12-17-2016, 09:38 AM
Jim , my understanding was that in a powder state , a sample can not be accurately analyzed . But from the photos , it looks like you can see if ingredients , positively or not , reacted after cure in the intended way .

12-17-2016, 12:38 PM
Wow. You guys are Great. VERY knowledgeable.
Sorry to bring this thread back to the original subject, but... :twitch:
Reading thru the information that Dave G sent to me, I have a few more questions...
1. There was mention of two different types of these Soft joints...
either an "expansion" joint, or a generic "movement" joint.
The Expansion joint requires a "backer rod" before the sealant, like...
open or closed cell polyethylene foam, or closed cell butyl rubber foam.
The Movement joint requires some "bond breaker tape" before the sealant.
Question: Which of those Soft joint types should I use here?
2. The one wall that I've mentioned has large, heavy cooking stoves up against it.
Do I really need to move them out, and put a Soft joint AT the wall,
or can I just put the first Soft joint about 4 feet out from that wall?
3. The recommended sealant...
CustomĀ® Building Products Commercial 100% Silicone Caulk.
OK. No question there, I guess.

Thanks very much,people.

12-17-2016, 03:32 PM
Prior to curing, raw cement looks more like little grows the spikes as it cures. One thing that affects the overall 'feel' and consistency when mixing and spreading the stuff is how finely it is milled, and whether there are larger chunks or not. One reason the stuff becomes less 'fluid' in the bucket when it sits for awhile is that it has started to grow those spikes, making the particles larger, interlocking them, and thus harder to spread. It's also incorporating some of the water into the crystalline structure, so that also makes it thicker. When you mix it later to 'loosen' it up, you're actually breaking some of those spikes off. There's a limit on how much and how often you can do that and retain the full strength of the cured material. A broken spike just ends up as more aggregate, not necessarily helping with bonding. IOW, that portion has been 'used up' in the process.

12-17-2016, 04:19 PM
Here's some tiles that are stuck down like mad,no tenting....crack/s transferred through tile,verses bond breaking from tile or slab.

This aunt and uncles "snowbird" house in FL. House built in 05 I believe.
In 2010 appx,they called me up to fix tiles....either they were loose,cracked,or both...I don't remember. I don't even recall what was underneath,only thing I recall is having limited room/thickness to re-install room for membrane etc. (and no clue who original installers were,nor their methods)

If I recall, one entire row was removed,and reinstalled. All pieces sound solid to this day,but do see movement at outer edge of grout along the single row.

Lastly,gave up demoing tile after about ten minutes with a chisel and hammer.
Will finish it Monday with power tool :)