Stone Floors on Wood Framing

[aseeng]

I am looking for opinions and current practice for design of wood floor framing with stone veneer on top. I am aware of the Marble institute criteria for L/720 deflection. I understand this to be a total load requirement. Similarly I have criteria out there for tile and stone floors that talk about the 300 lb. point load deflection requirement. Further the criteria looks less stringent for concrete slabs as they behave differently than wood. Some thoughts I have;

1. It seems to me that as you lay in the stone and grout it the dead load deflection is already there prior to setting up. Wouldn't the critical deflection criteria be for live load only?
2. The different criteria for concrete slabs sounds reasonable when considering live load and how it will distribute that load. But for a given span on a one way slab L/360 results in a certain deflection and curvature that is no different than for wood framing in regards to total load. What is the difference?
3. We look at at this situation on exterior decks with high snow loading. It is one thing to size the joists for this criteria but then the beam sizing really gets out of hand when considering full snow loads. We assume the criteria should extend to the beams as well as just the joists? Correct?
4. What about steel beams? Same criteria? I know we have not strictly followed the L/720 for these members where larger live load areas are concerned.

We know there are a lot of designs out there that don't strictly meet this criteria but we are not aware of any performance issues either. Any other experiences out there?

Thanks for your replies.

 

[Archie264]

Just some thoughts in response.

1. Watch out for creep. Neglecting the dead load might work against you in that regard.

2. With concrete the live load is usually a smaller percentage of total load than with other systems. Given that the affects of live load might be better absorbed and the requirements therefore relaxed?

3. Yes, I would think it extends to the beams as well.

4. Don't know but if the purpose of L/720 is to protect the stone veneer then I would assume it would apply to a broad range of structures.

 

[dhengr]

Aseeng:
In a stone, marble, tile veneer over a conc. slab, you have a much more monolithic surface to which you are applying the tile. Even so, cracking in the conc. slab will sometimes telegraph up through the thin set and tile. Also, the general deflection (deflected shape) of the conc. slab is usually much more gradual, as this change in shape relates to the size and stiffness of the tiles. In any case, they want a very stiff substrate to protect the stiff/brittle tiles against too much concentrated flexure. In a wood floor framing system you have reverse curvature of the sub-flooring over every joist. Then you have the deflection of the joists and their supporting beams too. You might also have concentrated flexure or curvature of the joists and sub-fl. over beams, where the joists butt or lap splice. Many details show more than one layer of sub-fl. or underlayment under these large brittle tiles, primarily to minimize these localized curvatures/deflected shape changes and concentrations. The tile industry has been using these deflection stds. for a long time with reasonably good luck, and you deviate from them at your own risk. People who can afford large marble tile floors don’t like to see them turn into marble chunks or chips, and stiffening the floor system a bit isn’t a huge cost compared to the marble cost itself. Nice long beams or slabs can probably deflect quite a bit with little damage, as long as they do it gradually. The wooden sub-fl. diaphragm can’t deflect much because a tile might be cantilevering both ways over a joist, or spanning btwn. two joists and will crack. Buy the way, are your stones 6" cobbles set in a mortar bed, or something else?

 

[Archie264]

dhengr, that makes a lot of sense.

 

[aseeng]

dhengr:

We are typically looking at flagstone 2-3" thick in a mortar bed. Various plan sizes but usually 2-3 ft dimensions.

Archie264:

I agree about creep.

 

[KootK]

I suspect that the guidelines provided by the trade associations include a fair bit of CYA. For the longest time, we designed brick backup walls using h/600. Some testing was done and it turns out H/360 is just fine.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[stevenspm]

From experience I wish the people who built my house followed the l/720 deflection. When I bought my house the tile in the 2nd floor bathroom was fine but because I am somewhat larger than most people over time all the grout between tile has cracked and broken loose and a couple of tiles have cracked. This is due in part by my size but also because when they built my house they put the tile over the one layer of sub-floor and had the floor trusses designed for lesser deflection criteria.

 

[KootK]

At the risk of being a pot stirrer, I've often questioned the wisdom of using things like gypcrete and stone overlays to attempt to make wood frame construction feel like institutional (steel/concrete) space. There are limited successes but a lot of problems too. Wood systems are moisture sensitive, vibration prone, and not particularly effective at load distribution. All of these things point to problems for sensitive floor finishes.

I get it. When I get up at night to relieve myself, I also want to feel as though I'm sauntering across the lobby of the Guggenheim. It's just not always a realistic expectation in a wood frame home. And it's not that I'm down on wood. Wood frame homes are wonderfully fit for purpose. That purpose being to comfortably, modestly, and economically shelter workaday humans.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[aseeng]

And that is the trick KootK. Everyone has different expectations and perceptions of floor feel. For some only the concrete floor feel works even with carpet on it. For others the common the workaday human shelter is fine. The easy answer to my original post is to comply with the L/720 deflection total load limit. It will cost someone alot of money and may save me some money down the road. I was just trying to understand the real issues of the problem and what others are doing to address it. I suspect there is a CYA policy that the trade organizations will follow and when it comes to tile and stone installations there are many variables that will crack the final product (use of backer boards, tying mesh to the structural substrate, improper mortar mix, etc.). I just wonder how many issues are really structural deflection issues.

 

[KootK]

I'm very glad that you brought up the issue aseeng. I've already archived this thread for the next time that I have to deal with this. And I agree, there is much that goes into the install that is non-structural but will have a significant impact on the outcome. I think that all rigid flooring tiles installed over wood floor systems should be 8"x8" max and installed over a 1/2" bed of melted gummy bears.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[jayrod12]

In my house when I did the last reno I double sheathed the floor prior to installing 12x24 ceramic tiles. So far no issues, and the existing joists are 2x10 @ 16 spanning 12 feet.

My thought is the double sheathing is forcing the deflection of the joists to be more gradual across multiple joists so the deflection of a single joist may not meet requirements for the floor tile however in the larger picture over a larger area possibly the deflection is closer to allowable.

But again, that's just a thought and by no means has any sort of calculations to back it up.

 

[jayrod12]

And the amount of screws I installed the sheathing with might be able to enact some uncalculatable form of composite action between the sheathing and floor joists.

 

[XR250]

@Jayrod;

I think the dbl. sheathing is a good idea. Once worked on a job where the large tiles were cracking. Turns out one of the occupants was well over 300 lbs. I believe the localized subfloor deflection due to the point loads of his feet was the culprit.

 

[jayrod12]

I could see that. Good thing I'm not quite at 3 bills yet or maybe I can expect more problems. And my wife is like 130 lbs soaking wet so she's less of a concern.

 

[IRstuff]

Don't know if this is relevant to this subject, but roofing tiles are put in place for upwards of a month before installation, prior to final installation, to make sure that the wood below has settled into its final sag.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

Of course I can. I can do anything. I can do absolutely anything. I'm an expert!