Maximum Allowable Deflection for Substrates

Problem is there is no way to field test for L/360. The number is based on engineering values. You as a setter still need to confirm that it meets the standard.

 

Scott's always trying to get in our heads to charge accordingly. Thats a good thing

 

I don't do any tile installs, but I wonder....how does a guy actually measure the deflection rate of 1/360 of a span at 300lbs? Do the math knowing the structural material dimensions, or do a deflection test?
In an older home, it becomes more difficult to know deflection....... with todays building codes, there is more 'known' info to go from.
I know the difference between the nice 'feel' of a good solid floor and one with a bit of bounce in it............... but how does one actually measure the deflection if you don't have a helper that weighs 300lbs?

 

I use span tables to figure if the floor meets L/360. Current codes require a floor to meet L/360 deflection be it dimensional joist or floor trusses on new work. On older houses using the span tables will give you a pretty good idea.

What it does not tell you is the deflection between the on-centers. I try to apply common sense to address that issue, and adding more plywood is never my first choice on a new construction job.

 

Mr. Gobis, at his school, had/has a dail guage with a plunger, he would put
some boards across some blocks, then, put it under them, put some weight on the boards, like pine, spruce, and so on,, hey Bud, how about pine plywood, good idea? sweet sweet!

 

Re: Who Is Responsible?

as scott stated, it is based on the engineering ... the enineering of the substrate/subfloor system.

In the most general terms, the MINIMUM requirements are 2"x8" joists on 16 inch centers with a maximum unsupported span of 12', with a single layer of 5/8" plywood, nailed every 6"-10" (or a minimum thickness of 1.5 inches of plywood for direct bond of tile to plywood without a membrane or suportive cbu with approved modified mortars). This gives an active deflection of L/389 which is well withing minimum approved requirement of L/360 ... but remember this is a minimum.

Of course if you increase the width of the joists, decrease spacing of the joists, decrease the unsupported span distance, and/or use laminated or more structurally sound or engineered substrate/subfloor structure, then the deflection decreases considerably.

Of course if you decrease the width of the joists, increase spacing of the joists, increase the unsupported span distance, then the deflection increases and would NOT fall within the requirements of L/360 !

the engineering of the joist width, unsupported span, and placement apart form one another all play factors in the deflection ratio.

for ceramic tile that is L/360 of the span.

In more technical terms, it should be noted that the deflection is measurable both as the length of the span, as well as between joists, in the case of a residential application. In actual practice, what this means to the mechanic is that they have to know if the sub-floor will be able to support the tile and not exceed the deflection ratio of L/360.

With joists 16" o.c. the maximum deflection that can be allowed when using L/360 is 16/360 which equals 1/25"= .04". Now how does this relate to a single layer of plywood sub-floor. When looking at the span tables published by the American Plywood Association, it will be found that a single layer of plywood 19/32" on joists spaced at 16" o.c. the maximum amount of weight that can be applied is 368 lbs before the deflection succeeds L/360 (remember 1/25").

If a concentrated load is applied between the joists using a 3" wide disk with 200 lbs applied to the same piece of plywood, the plywood can deflect, up to approximately .078" = 1/12.5". This calculation indicates that the single layer of plywood can have approximately twice the deflection acceptable for the installation of tile. In real terms, a concentrated weight of 200 lbs over a 3" wide area, is not all that difficult to attain in a situation such as a fridge or a stove being brought across a floor on a hand truck or dolly.

With all of this being said, it can now be understood where the recommendation for two layers of plywood 5/8" thick or the use of other underlayments is required. Other underlayments such as ½" cementitious backer board, or a pre-formed sheet applied membrane such as Schluter Systems, Ditra Matting can be considered.

Another consideration as well, is post-tensioned and pre-stressed concrete are often engineered with only L/240 deflection criterion. This means a slab 30 ft long, supported only at the ends, (L/ 360 = 360/360) could have deflection up to 1 ".

In contrast, the Marble Institute of America has increased the deflection ratio to L/720, which is double that of tile.

Installation techniques and systems should be evaluated and be guaranteed to respect the L/360 ratio for tile.

John bridge has an awesome tool called Deflecto ..now this isnt the final word in deflection and it isnt scientifically perfect per se, but it come VERY close and we use it alot in questionble areas and it has never let us down ... click here for it ... Deflecto

The American National Standards Institute (ANSI) in coordination with the Tile Council of North America (TCA/TCNA) sets and distributes the standards by which every job is judged and responsibility is determined in cases of inspected an/or litigated failures.

I would advise everyone doing professional tile work in the United States to purchase a copy of the The Tile Council of America's Handbook for Ceramic Tile Installation ... we use it extensively as well and it has never led us astray.

* * *​

for portions of the information in this post, the following sources are credited:
John Bridge
The Marble Institute of America
The Tile Council of America
The Terrazzo Tile and Marble Institute of Canada

 

Re: Who Is Responsible?

I re-read my post and wanted to clarify something that might be misunderstood.

the substrate/subfloor plywood layer has nothing to do with the deflection ratio overall (as far as the TCA/ANSI is concerned), but instead the rigidity of the plywood layer, which is a totally separe issue from the overall substrate/subfloor deflection, so adding additional plywood layers to a existing subfloor system which doesnt fall within the deflection ratio of L/360 IS NOT an appropriate remedy... at least i dont THINK it is.

The rigidity of the plywood layer only ensures deflection between joists, and DOES does have some secondary/tertiary effect on the overall system, but i dont think that can be used as a consideration for proper pre-installation site conditions of the substrate/subfloor in regard to L/360 deflection.

minimums for the plywood layer are 5/8" with approved membrane or cbu.

minimums for the plywood layer as 1.5" for tile over plywood installs without membrane or cbu.

i know this can be a confusing suubject for many ... and if i am wrong please let me know.

 

Re: Who Is Responsible?

Good info Floorman,. My thinking is the plywood minimum is about 1 1/8" without CBU or Ditra. Not real sure because I will not tile directly over plywood without using something between.

 

Re: Who Is Responsible?

oh absolutely a 1-1/8" thickness is MUCH better than the minimum recommended 5/8" plywood over ditra or cbu.

 

?? Adding plywood will most definitely alleviate deflection. It adds rigidity to the structure. Provided it is structrual(3/8 or better) adding plywood is the easiest way to beef up a floor.(CBU offers no rigidity benefit whatsoever)

p.s. Have any of you readers had a chance to check out the 1/4 point deflection theory set forth by the folks at Schluter? Peter Nielsen and Frank Woeste Ph. d. wrote an excellent thesis on it. The logic underlying it is "smack you in the face" type stuff. It seemed contrary to everything I thought I knew until I actually came to understand it. I could probably find it if anyone is interested.

p.p.s. I know some of you must have the article. Its in the binder they gave you at the skool.

Here is one that confuses me. Schluter says 5/8" decking is ok under Ditra on 16" centers. That seems a bit weak. There is, however, an awful lot of thinset going down when we use Ditra. Reckon that has something to do with their calculation?

Heck, the finished assembly is akin to a reinforced mudbed. I have never seen anything but 3/4 around here, so it is more a matter of trivia than anything else.

Any ideas?

I should concede that 1/2in Hardi actually does offer a nominally measureable increase in structural rigidity, but it pales in comparison to 3/8 plywood in that arena.(hell, it probably offers less than 1/4 in Baltic Birch)

This info is free for whomever desires it. All it takes is what you already have at your fingertips. Knowing beats wondering in whom to place your trust. I beats arguing also.

Knowing is a handy trump card. Seek and ye shall find. Go get it.

When it comes to matters technical, there are very few disagreements among the learned. Standards are not simply made up for the sake of creating standards. Are there logical inconsistencies within every published standard? Yes, but they are incredibly rare. Is everybody in manufacturing out there trying to ensure that failures occur so they can blame them on the installer?

NO

As an installer who never suffered repeated failures of the materials I installed for a living, I really don't get the fixation on failure that some seem to have. I focus on excellence. Failure is the last thing on my mind when I install.The bar is where you put it. Blow them away with excellence and you will never know failure.

The rules have not changed. They never will.

BTW: L/360 is a value related to the deflection of the framing members. That standard was adopted in the days of 12 and 16 centers. The advent of 19 and 24 spacing has introduced another consideration; curvature. L/360 does not even begin to address the problem of sufloor curvature between the framing members. Regardless of the overall rigidity of a designed flooring module, the problem of curvature creates instability. The L/360 deal does not make for an accurate frame of reference these days.

 

Re: Who Is Responsible?

Plywood doesn't add any strength to the defledtion in the joists system, just the deflection between the joists, if spaced 19" or above, more plywood is needed, I wouldn't tile floor even with one layer over 16" spacing, if a floor needs more strength in the flex to meet L/360 or above, then sistering, adding joists, or a cross beam under them is the only way, CBUs, membranes and thinset don't help.

 

Re: Who Is Responsible?

Get this...

I have a home of hardwood to do in the upstairs. It is a new home they are building with one of those fast paced slap them up as cheap as possible builders. I'm going in after closing, to do my work directly for the homeowner.

I have ½" plywood to put down to make the substrate meet the wood flooring requirements, because ¾OSB over trusses on 24" center, isn't going to cut it.

The sad thing is... They came in, and put down CBU and laid tile in the master bathroom and the other two baths upstairs.. I have yet to see of feel it now the tile is in, but I can just imagine. A failure waiting to happen!!

 

Re: Who Is Responsible?

Huh? How do you figure?

 

Re: Who Is Responsible?

Chuck, be patient, we'll agree on something sooner or later, until then, keep reading.

So Bud, when is this meeting taking place, I'll bring the coffee and donuts, should take awhile.

 

Re: Who Is Responsible?

How is it that plywood offers rigidity in one direction but not in the other? Is this something you have tested?