Ceramic Tile Advice Forums - John Bridge Ceramic Tile

Ceramic Tile Advice Forums - John Bridge Ceramic Tile (https://www.johnbridge.com/vbulletin/index.php)
-   Tile Forum/Advice Board (https://www.johnbridge.com/vbulletin/forumdisplay.php?f=1)
-   -   Cracked Natural Stone Tiles II (https://www.johnbridge.com/vbulletin/showthread.php?t=16322)

t majchrowski 10-08-2004 11:30 AM

Cracked Natural Stone Tiles II
Have a couple of questions. For natural stone tile, the Marble Institute of America calls for a floor deflection of L/720 maximum. Is there a area that is okay around this maximum? Reason I ask is that we have calculated a deflection of about 1000 in my dining room/kitchen area and over 1300 in the area of a hallway. Our travertine tile has cracked at each backerboard joint in these areas. Is this deflection too tight? Could this be the cause of our cracked tile? If so, any recommendations on how to fix the problem?



Mike2 10-08-2004 11:51 AM

HI Tom.

I'll send a message for our engineer bbcamp to swing by and help you out with your deflection question.

In the meantime tell us:

1. What size of joists, how are they spaced, and what is the longest unsupported span in the tiled area?

2. Describe your sub-floor in detail.

3. Was the backer board laid over thinset? And was the backer board screwed/nailed into the floor joists?

4. Was an allowance for expansion provided around the perimeter of the room? And were expansion joints placed every 20 - 25 ft. within the field of tile itself, if in fact you have that large of an expanse.

5. What size of tile and what spacing was used?

Answers to all these questions and possibly more I haven't thought of will be most helpful in diagnosing the likely cause of your problem.

bbcamp 10-08-2004 12:10 PM

The bigger the number under the L, the better your floor is for stone.

The cracks over the backerboard joints tells the story. My money is on either no tape and mud at the backerboard joints or no thinset under the backerboard.

t majchrowski 10-08-2004 07:51 PM

Info on tile placement
For Mike,

1. The joists areTJI/Pro-150, 11/78 inch product. They are spaced 19.2 inches on center. The longest unsupported span is 14 feet.

2 & 3. The sub floor is .75 inch plywood with thinset application and .5 inch backerboard. The backerboad is stapled into the plywood. I have raised our kitchen island and looked under it. The backerboard does not appear to have mud between the seams. We just have a space between the backerboard panels.

4. Since the largest length of tile is about 20 feet, I do not see where there are any expansion joints. Do not see how any allowance for expansion exists around the perimeter of the tile either. The backerboard goes under the counters. The tile goes up to the base of the counter.

5. Have 12 inch by 12 inch travertine tile that is 1/2 inch thick with about a 1/8 inch spacing .

Thanks to Mike and bbcamp for your comments.


Jason_Butler 10-08-2004 08:33 PM

Those seams should have been taped / floated with thinset.

That's my bet...


cx 10-08-2004 09:24 PM

I don't have a table available for deflection specs higher than L/480 for the TJI Pro-150 joists, Tom, but I would have guessed you to be pretty close to the max span for L720, if not a little over. Injineer Bob maybe has the specs. Did you get your L/1000 spec from the manufacturer, or some other source?

That said, even if the joists are stiff enough, the subfloor isn't. In fact, it's not even close. The only installation method I know of that would be acceptable over that subfloor would be using Ditra and ceramic tile - and that's only because Schluter will warranty it. Not gonna make if for natural stone.

Assume that your CBU was laid such that the joints fell between the joists (correct procedure) and you could reasonably expect the cracking to appear there. Taping the joints correctly may have helped, but I don't know that it would have prevented the failure. Most manufacturers don't recommend the use of staples to install their CBU; that could also contribute to the failure if it allowed movement of the CBU.

Who made all the installation decisions on this project?

My opinion; worth price charged.

Bill Vincent 10-08-2004 09:46 PM


Originally Posted by t majchrowski
For Mike,

1. The joists areTJI/Pro-150, 11/78 inch product. They are spaced 19.2 inches on center. The longest unsupported span is 14 feet.


I have a feeling that the 19.2" with only CBU over one layer of 3/4" is the culprit. You may have astronomical numbers in the deflection department along the length of the TGI's, but between them it leaves alot to be desired. CBU adds almost NO structural strength. My own feeling is that you need atleast another 1/2" of plywood under that CBU.

Just realised Cx got here first!! Well, Cx, you know what they say about great minds!! :D

cx 10-08-2004 10:12 PM

Seems we been doin' that a lot lately, Bill. You must be gettin a lot smarter. :D

Bill Vincent 10-08-2004 10:49 PM

Stranger things have happened! :D

Mike2 10-09-2004 11:36 AM

One more thing:

What backer board has been approved for use by the TCA, or is warranted by the manufacturer, for joists spaced more that 16" OC? :uhh:

Steven Hauser 10-09-2004 11:50 AM

Well it is written as an uncoupling system. Not CBU

T_Hulse 10-09-2004 01:58 PM

Hi Tom,
I agree with everything the guys have said about possible reasons, but in my experience, it won't be quite as bad as every CBU seam cracking through unless 1 particularly bad mistake was made: no thinset under the cbu. Sounds like you've got lots of things wrong, but that would be by far the worst. Get a flashlight & look down in the spaces between the CBU panels. You should at least see the trowel ridges of the thinset used. If you only see subfloor, then this would be the worst of many mistakes here.

Without having the deflection tables either, my guess is that you would need something more like 3/4" subfloor + 1/2" plywood + Ditra to make this floor fly.

Mike2 10-09-2004 02:15 PM

I noticed that too Steven.

In the TCA handbook, F144-03 specifically uses the term "Cementitious Backer Unit" which is approved for joist spacing 16" OC. On the other hand, TCA in F148-03 uses the term "speciality uncoupling system" which is approved for use in joist systems 19.2 and greater. Are these two terms, CBU and speciality uncoupling system synonymous? I don't know for sure.

I did check the data sheets for Durock and Hardibacker, both say for joist spacing not to exceed 16" OC.

T_Hulse 10-09-2004 02:19 PM

Speciality uncoupling system means Ditra or equal. Backer boards are definately excluded from that category. :)

cx 10-09-2004 02:43 PM

Yeah, Mike, that's one of them clever little wording games they use to get a procedure published for a specific manufacturer's specific product without appearing to have published a procedure for a specific manufacturer's specific product. In that particular case, you can certainly use any product on the market, so long as it's exactly like Ditra. Which, of course, includes ........... Ditra. :rolleyes:

That's my understanding of it, anyway. If you look closely at F147 and F148, don't that look a lot like a cross-section of Ditra?

Actually, I wonder if all the new Ditra knock-offs are also approved under those methods now that there are two or three other waffle-type membranes out there.

How does that work? :confused:

t majchrowski 10-09-2004 09:53 PM

Thanks to all

Really appreciate all the input.

The general contractor made all the decisions on the subfloor. He said he checked with his architect and everything was fine. He also claims to have done this type of installation many times without a problem. Maybe our eyesight is better than others.

I don't understand all the deflection issues, but do realize that the 19.2 inch spacing between joists and 3/4 inch plywood doesn't meet specs.

Now to try to get the problem fixed. Wonder if we should go with a good ceramic tile instead of all the work to upgrade the subfloor for stone tile. Oh, well.



Bill Vincent 10-09-2004 10:14 PM

Tom, your contractor might want to read the following article:


Quite frankly, I don't think ceramic would make it on what you have now, either. Not at 19.2" spacing. If he's done this before with no problems, he's been lucky.

cx 10-09-2004 10:47 PM

Well, I'm sure don't nobody wanna hear my architect opinions again, but...... :rolleyes:

One or both of them shoulda known better than to do that installation, Tom. If a competant architect was involved with the designing, he would have known the use of CBU over 3/4 inch plywood over 19.2" joist spacing was not an industry approved procedure even for ceramic tile, let alone natural stone. Couple that with an incorrect CBU installation and I agree with Bill, your chances of even ceramic tile surviving are poor.

If you want your stone (as I presume you were promised by the builder in your contract), you'll need to take up the existing floor at least down to the plywood subfloor, add another layer of plywood, correctly installed, then a substrate of one kind or another, correctly installed, then tile. If that's gonna cause finished floor height problems, we can discuss that, too.

The article Bill posted, by Dave Gobis, is a good one for you to present to the builder and architect. It's simple, clear, and athoritative. Dave has tested a lot of floors in his time and knows whereof he speaks.

My opinion; worth price charged.

t majchrowski 10-10-2004 06:10 PM

Thanks for all the info. I have the article by Dave Gobis. I will pass it on to the contractor and tile installer. See what they have to say. I will hold back on all your advice until I get a response. Wish me well!!


t majchrowski 10-25-2004 10:21 PM

Natural Stone Cracked Tile III
Continuation of a long story. First, thanks for all the fine inputs to my first two posts. I have given info to the general contractor. I'm the guy with cracked tile running linear accross the width of two rooms and a hallway, with the joists 19.2 inches apart, 3/4 inch plywood, thinset with backerboard. Deflection is calculated using standard techniques as over 1000. Your past comments say to look at the deflection in the plywood between the joists and to make sure the backerboard is screwed, not staples down. Also should look into supports under the stone between the joists and adding another 1/2 inch of plywood. The contractor says he has found people who fix the problem by just taking down to the plywood and adding two layers of backerboard perpendicular to each other. Does this sound right? In our earlier discussions, there were concerns that the subfloor was not strong enough for the stone. Wouldn't adding another layer of non-supporting backerboard over the existing subfloor just add more weight and maybe make the problem worse? Or, is running the backerboard perpendicular in two layers okay despite the added weight.

Hope I am getting to the end of this adventure.


cx 10-25-2004 11:31 PM


I merged your new thread with the previous one. Please keep all your questions for this project on this thread so anyone who wants to comment will be able to follow the history. Your thread will pop up to the top of the queue every time you make a post to it, no matter how old it gets. :)


Would you please invite your GC to come by here and talk with us?

Tearing out everything down to the subfloor is a good start. Adding two layers of CBU makes no sense at all. He is obviously unaware even that CBUs have no grain and it makes no difference which direction they are laid - they still don't count appreciably toward the strength of the subfloor, although common sense recognizes you gain something. Why would he go to all the trouble to correctly install two layers of CBU instead of adding a sufficient second layer of plywood and then an approved substrate? That's if, in fact, any CBU manufacturer will specify a correct method for that second layer - but, hell, first you gotta get one to approve using their CBU over that joist spacing at all.

Try to 'splain to him that he needs a subfloor that meets the 1/720 deflection standards between the joists before the installation of a tiling substrate, and the substrate hasta be approved for the application. If you have joists with a three inch wide top chord, that's a deflection of no more than .022 inches with a really, really fat guy standing on one foot right inna middle. That's a very stiff floor, and he ain't gona get it with only one layer of 3/4 plywood.

And I think I still had a question about how y'all determined the L/1000+ rating on that joist structure.

I think you're still being led down the primrose path, Tom. Don't go.

My opinion; worth price charged.

Steven Hauser 10-26-2004 07:34 AM


Hi there,

First and foremost, it doesn't matter what the deflection calculation is along the truss because as it has been clearly delineated the primary issue is the fact that backerboard manufacturers don't advise using their products on truss systems. nor floor joists spaced wider than 16" O.C.

That aside, I think it is time to print out the responses you have and present them to the contractors.

My .02

t majchrowski 10-26-2004 12:35 PM

Deflection Calculations and Subfloor Info
CX and Folks,

I had a fellow from Trus Joist run the deflection calculations on his own program. I believe this is the industry program. Also had an independent engineer run the calculations on his own program. Both came up with about the same results. I saw the independent engineer enter the info for his program. The inputs were type of trus, length of trus, separation of trus, width of plywood, number of supports under the trus, including the end walls, and maybe something else. I believe these calculations just gave the deflection of the trus. Anything else, way beyond my comprehension!!

Hope this is what you asked for.

Also, just found some info indicating my subfloor is not 3/4 inch plywood, but either 23/32 inch OSB or 1 1/4 inch 1.3E Timberstrand LSL. These are listed as accessories on my floor plan. Don't know which goes where. Does this make any difference?

I will give this entire thread to the general contractor and the tile contractor. Hope they see fit to ask questions on this forum.



cx 10-26-2004 07:34 PM

Well, we'll feel better about the joist deflection now. :)


Also, just found some info indicating my subfloor is not 3/4 inch plywood, but either 23/32 inch OSB or 1 1/4 inch 1.3E Timberstrand LSL. These are listed as accessories on my floor plan. Don't know which goes where. Does this make any difference?
Yep, that could make a difference, but only if we know qual es qual. The OSB option is still just a 3/4 inch subfloor. The LSL would be substantially better. You'd still hafta contact the manufacturer to determine the deflection, though. But first you need to find out what's actually on the floor, eh?

And then, even if the deflection is better than 1/720, there's gonna be at least one guy jump in here and say it don't matter on acoounta you still gotta have double-layer subfloor for that stone. But that will be a fun argument. :D

My opinion; worth price charged.

T_Hulse 10-26-2004 07:54 PM

You still gotta have a double layer subfloor for that stone. :)

Bill Vincent 10-27-2004 06:15 AM

Absolutely. AS well as having L/720 along the framing, you also need it between. Double layering is a must. Even if using a membrane, all membranes, including Ditra, require that the floor meet the deflection standards PRIOR to installation of the membrane.

cx 10-27-2004 07:15 AM

Aw, c'mon, Vincent, I promised the guy a fun argument. :rolleyes:

If he has that injineered flooring, he'd have 1 1/4 inches of subfloor between joists, likely meeting 1/720, but it will be a single layer.

Now, would you kindly enter into the argument as presented? :shades:

Bill Vincent 10-27-2004 07:49 AM

Well, as "presented", the floor should do fine, then, for Ditra or 1/2" cementboard, however, I'd still prefer the protection that double layering the plywood gives.

T_Hulse 10-27-2004 10:48 AM

All right CX you finally baited me into it. :stick:
There is a 2 layer subfloor requirement for all stone floors. When you add your final underlayment layer (CBU, membrane, or mud) that makes 3 layers total. The reasons you need two layers of subfloor are that stone is super sensitive to movement, & because deflection is not our only movement concern. There is also expansion/contraction. The overlapping 2 subfloor layers limit the effects of this by tying the whole floor together structurally. Imagine you found a subfloor material as thick & stiff as concrete. Let's just say you could put concrete up there. You still have a cold joint every 4x8 feet, and none of the underlayments let you just plow through & ignore cold joints. So a super stiff 1 1/4" panel is great for deflection, but it does nothing to tie the floor together structurally.
Tom please make sure your builder understands that cement board is not a structural panel. They will sag, crumble, & pull right through screws when subjected to structural stress because that's not what they were designed for. They're just an underlayment made to help isolate the tile from floor stress (not reduce the stress) and to provide a good glueing surface.

You need 2 layers of subfloor, 1 layer of underlayment for stone. :)

All times are GMT -6. The time now is 09:05 AM.

Powered by vBulletin® Version 3.7.4
Copyright ©2000 - 2022, Jelsoft Enterprises Ltd.
Copyright 2018 John Bridge & Associates, LLC