Contractor Talk - Professional Construction and Remodeling Forum banner
Cancel
Create thread New Forums

proper subfloor

Jump to Latest
1 - 20 of 36 Posts
A

AustinDB

· Contractor
Joined
·
5,038 Posts
Discussion starter · #1 ·
I recently read a thread in this forum that now has me confused. I had always thought that a 5/8" + additional 1/2" installed topped w/ Ditra or 1/4" Hardi was a very good practice.

Current job has 3/4" ply on 12" joists-are there span charts to identify if the current plywood is sufficient? If the 3/4" is sufficient, is there any benefit of using 1/2" vs 1/4" Hardi? I like Ditra, but cost and ease of purchase play a factor given the height isn't an issue here.

when the TCA recommends 1 1/8", is that a minimum plywood recommendation or a 'best case' suggestion?
 
#2 · (Edited)
try this http://www.johnbridge.com/vbulletin/deflecto.pl

Basically you need a minimum L360 for porcelain tile and L720 for stone

No CBU(hardie, ditra etc..) add any structural stability to floor. You can sister joist, shorten the span with a wall, or add more plywood.

What is the span and spacing (16" apart) of said joists?

I use regular T&G real 3/4" plywood over joists with tile all the time. Most of the time when we do stone tiles we add a layer of 1/2 ply over the 3/4 then ditra.
 
#3 ·
No real benefit to 1/2" vs 1/4" cementboard... Im sure it does add some structural value, you're just not supposed to rely on that. The subfloor should be adequate by itself.

Check the span of the joists and use the Deflectolator that Kevin posted above. 3/4" ply is usually acceptable over 16"oc framing, provided you're not installing natural stone.
 
Discussion starter · #4 ·
Thanks for the link Kevin-precisely what I needed. The joist spacing is 12" and the handy dandy calculator said the system netted L/420 which is sufficient for ceramic tile.

If the spacing were 16" OC and deflection were too much given a single layer of 3/4"-how could you determine if an additional layer of 3/8" would be enough? I've always slapped down a sheet of 1/2" 'just because', but there must be a more precise way, right?
 
#6 ·
The American Wood Council has a better span calculator located here at this link.

http://www.awc.org/calculators/span/calc/timbercalcstyle.asp

You will notice that the span calculator's do not ask you for sub-floor thicknesses. Deflection ratings are for the floor joists and deflection needs to be measured in all direction and this includes in between the floor joists themselves. If you have 24" on center floor joists you can achieve a L/360 deflection rating with Spruce 2"x10" 's at 14' total span. Now if these floor joists are topped with a single layer of 5/8" plywood the chances are your plywood will deflect more in between the floor joists than they do over the span of the room. Here is where adding another layer of 1/2" plywood or 5/8" plywood helps out.

12" on center joists spans are ideal for tiling and nets a tank of a sub floor. As a tile contractor you should have a good understanding of framing requirements as outlined in this article http://www.ctioa.org/reports/fr81.html from the CERAMIC TILE INSTITUTE OF AMERICA, INC.

Here is more info on Plywood than you ever need know: http://www.fpl.fs.fed.us/documnts/fplrn/fplrn059a.pdf

If you take a common 16" on center floor joist make up you have about 14.5" in between each floor joist. If you take 14.5" and convert that to 32nds of an inch you have 464 32nds of an inch or 464/32". In a L/360 set up your plywood can not bend more than 1.3 32nds of an inch. For a L/720 rating the deflection is less than a 32nd of an inch and closer to one sixty fourth of an inch. You will not achieve this with wider joists spacing and single layers of plywood.

It's quite common online to read that cement board offers nothing to a floor assembly and this I find quite funny. A simple test at home to see for yourself the structural gain in the various subfloor options is to cut out some pieces of equal size.

Try standing on on a piece of 1/2" cement board thinset to 5/8" plywood on 16" inches.

Now try again with 1/4".

Now try again with Ditra, Spider Web

Try again with Noble Seal TS, Dal Seal TS

Once you have played around with this it's comical the posts stating that cement board adds nothing to a floor assembly. As the floor joists widen this stiffing effect of course lessens but anyone tiling a floor should be measuring this in the field. Bring a straight ruler with you to the job and pile your setting materials in one spot. You may need a simple off cut of plywood to concentrate the load in between the floor joists. Measure the sag. Remove your setting materials and recalculate.

I have done this in years past and found that unless your dealing with 3/4" plywood over 12" joists spacing the deflection is to much and a second layer of 1/2" or 5/8" plywood always is a good starting point. Then cover that with Ditra, Noble Seal TS, Wedi, or what ever you trust most.

JW
 

Attachments

#7 · (Edited by Moderator)
OK I have an Issue with all the references to span tables with out actually checking to see if the floor you are going to set on meets spec.
Span tables are for the framer/builder to choose the correct product. Our Requirements are deflection of 1/360 the length of span. VERIFY the ACTUAL floor you are about to tile meets Standard. It is simple. The deflection calculation is based on a concentrated 300# load placed center of span.

So lets say you have a 20' span, that is 240" so to calculate deflection 240/360= .66 or between 5/8 and 11/16" that is your actual "on site" pass fail standard
How to check your floor
1, Mount your horizontal laser to a wall.
2 Take a light weight tripod with an Inch scale mounted so the laser intersects it. Mark the placement of the legs so you can replace the tripod over the center of the span after you load it (see 4)
3 Move to the edge of the room and take a digital picture of the beam where it hits the rule.
4 load 300# of tile, grout,mortar in the center of the span replace tripod.
5 go back to where you took your first pic and take another digital picture of the laser line and compare deflection. If the line has moved more than 11/16 you are out of Spec Less than 5/8 and you are good. Save the pics and add the relevant data in case you need to refer to it later.
(remember the numbers i used are based on a 20' span yours will be differnet)

the same process can be to verify between joist deflection but you need to use a 4 wheeled cart the leingth of the joist span, first measurement is with wheels on joists second wheels center of span between

The book may say it is ok but the editor didint take into acount the flood the house had 5 years ago or their neighbor giving them a "deal" on cleaning up the pipes hanging too low in the basement. As professionals our job is to verify the Actual Site conditions before we lay tile.
 
#8 ·
charimon said:
...
How to check your floor
1, Mount your horizontal laser to a wall.
2 Take a light weight tripod with an Inch scale mounted so the laser intersects it. Mark the placement of the legs so you can replace the tripod over the center of the span after you load it (see 4)
3 Move to the edge of the room and take a digital picture of the beam where it hits the rule.
4 load 300# of tile, grout,mortar in the center of the span replace tripod.
5 go back to where you took your first pic and take another digital picture of the laser line and compare deflection. If the line has moved more than 11/16 you are out of Spec Less than 5/8 and you are good. Save the pics and add the relevant data in case you need to refer to it later.
(remember the numbers i used are based on a 20' span yours will be differnet)...
Click to expand...
I like it. Great execution. :clap:

Perhaps even a digital camera with a timer. You can walk out with a few bags of setting materials on your shoulder (I'm a good 230) and take a picture. Then set the camera on a timer delay and walk out.

These new home builds are so bouncy we are pricing in a good half day of blocking on each new one we tackle.

JW
 
#9 ·
Once you have played around with this it's comical the posts stating that cement board adds nothing to a floor assembly
Click to expand...

I'll assume that's directed to my post 2 above yours... Do we really need this nonsense again John? Want to poke me with a stick too because I still use Kerdi?

Of course cement board will add some structural integrity to a floor, my post said it's better to have a subfloor that is already sufficient. You're not supposed to rely on the cement board to bring it up to spec.
 
#11 · (Edited)
Splinter said:
I'll assume that's directed to my post 2 above yours... Do we really need this nonsense again John? Want to poke me with a stick too because I still use Kerdi?...
Click to expand...
You assume to much. The post is for anyone reading this thread so they understand that there is more to deflection of a subfloor than the size of the floor joist and it's span. I could care less that you use Kerdi.

Deflection is deflection. The joists span and distance between them is one part of the equation. Understanding that the plywood thickness and the deflection of that between floor joists should be checked is important.

As for your comment about using cement board to bring a floor's rating up to spec you really do need to get yourself a TCNA specification guide. Once again for those of you without a TCNA specification guide you can use;

1/4" thick cement backer board with 19/32" thick subfloor on 16" floor joists centers and get a "Residential Service Rating" TCNA Specification Detail F144-11

1/4" thick fiber cement backer board with 19/32" thick subfloor on 16" floor joists centers and get a "Light Commercial Service Rating" TCNA Specification Detail F144-11

7/16" thick cement backer board with 19/32" thick subfloor on 16" floor joists centers and get a "Light Commercial Service Rating" TCNA Specification Detail F144-11

1/4" thick cement backer board with 23/32" thick subfloor on 16" floor joists centers and get a "Light Commercial Service Rating" TCNA Specification Detail F144-11

This specification notes that if glass tile is used the service rating could be lower.

The TCNA Specification guide on page 30 specifies that subfloors both wood or concrete need to conform to the IRC for residential and IBC for commercial and that the deflection shall not exceed L/360. We can only assume that all homes are built to this minimum standard and like suggested above checking in the field is key to a successful install.

So again, why is that cement board can qualify for a tile ready surface and get a light commercial service rating if it does not add any structure to the floor? The double plywood service rating is TCNA Specification F150-11 and nets you a residential service rating with 15/32" plywood and light commercial with 19/32". This again is the second top layer of plywood.

Splinter said:
...You're not supposed to rely on the cement board to bring it up to spec.
Click to expand...
You don't. Just adding the cement board does that for you. Look it up.

JW

Source of TCNA Specifications

Image
 
#12 · (Edited)
Here we go again lol


I spoke to my cousin more about this subject a week back. Told him about my test with the ditra and wedi board. He's a structural engineer/inspector for the government in the UK. He checks highways, bridges etc etc. I Explained this whole situation to him about materials adding extra strength to subfloors and he gave me all this technical jargan but Basicly said it comes down to compression. He said the better the compression ratings for said products the stiffer the sub floor would be. He gave me an example on a concrete slab secured to plywood subfloors then another example of plywood secured to concrete floor and asked me which is stronger under center load compression. I guessed subfloor on concrete because i didnt even think about it which he said was wrong. He said because concrete can take compression extremely well when you put the concrete ontop of the subfloor it's under compression which is adding stiffness to a subfloor. He then explained that it would be very weak under the subfloor and would fail much earlier. He said this the reason that re bar is placed lower down on a structural concrete beam because the concrete is then under compression and the rebar increase tension the more the concrtete flexys. He said this is why anything with any type of compression rateing will increase the floor stiffness. I showed him the pics of the test I done and he said without even seeing the specs of said products the wedi would add vastly more stiffness to a subfloor just because of its triple layer construction. He did say that this will only apply when the materials are properly bonded though.


Hope this explains as to why sub floors feel so much stiffer with hardie, cbu and wedi compared to ditra or noble.
 
#15 ·
I guess the real question here, John, is there a "quantifiable" difference-- structurally between the 2 ratings. and if the difference between the 2 ratings is structural what is it. If the quantifiable differences are not structural than this point doesn't apply. But if they are than you have uncovered a Valid point... So what is the quantifiable differences between the 2 ratings
 
#16 ·
The real-world issue is all of these silly stats and "feelings" being attributed to what an acceptable floor is.

A light commercial rating because of CBU does NOT make a floor better.
Having a floor FEEL stiffer does not make a floor better.

An installer is doing an incredible disservice to a customer if they sell a job with the lie that a tile underlayment can make an unacceptable subfloor compliant simply by installing it.

Ratings for floor systems are for more than just deflection.

Flexural and compressive strength differ from material to material. Those are also stats that determine rating. Compressive strength has nothing to do with deflection.

If you want to drive a forklift on your floor, you probably would want 1/2" CBU instead of Ditra. If you want your kitchen tile installation to last, make sure the framed floor substructure is more than sufficient BEFORE you specify a tile underlayment.
 
#17 ·
angus242 said:
Compressive strength has nothing to do with deflection.
This is where you are wrong. Just IM'd Me cuz and he said it has everything to do with Deflection. He said the Deflection is based on a materials compressive and Tension properties. He said this is why a 20MPa concrete is stiffer than a 5MPa concrete and why some bridges have to have a much higher Compressive strength concrete to stop deflection. He said he could get into allowable flex but its not something that has anything to do with what we talking about.

He said this is something you learn in your first 6 months of uni and that my test with the ditra and wedi shows full well why a floor would be stiffer with wedi/cbu than ditra. But hey this guy designs and checks bridges for a living so i guess he know a thing or 2 about deflection.
:laughing:
Click to expand...
Click to expand...
 
#18 · (Edited)
angus242 said:
Compressive strength has nothing to do with deflection.
Click to expand...
This is where you are wrong. Just IM'd Me cuz and he said it has everything to do with Deflection. He said the Deflection is based on a materials compressive and Tension properties. He said this is why a 20MPa concrete is stiffer than a 5MPa concrete and why some bridges have to have a much higher Compressive strength concrete to stop deflection. He said he could get into allowable flex but its not something that has anything to do with what we talking about.

He said this is something you learn in your first 6 months of uni and that my test with the ditra and wedi shows full well why a floor would be stiffer with wedi/cbu than ditra. But hey this guy designs and checks bridges for a living so i guess he know a thing or 2 about deflection.
:laughing:


He just sent me a pic of why compressive strength has everything to do with deflection.

The lower cables would be what you call a "subfloor" and this would be under tension. The more layers you add the more the stiffness/deflection lowers but compresion is added more to the upper layers this would mean you would then need a material that handles compression better. Which would be your "CBU/Wedi" which in turn gives less deflection the more layers you add. If you add too many layers the layers that were under compression would then be under tension.

Image
 
1 - 20 of 36 Posts
About this Discussion
35 Replies
10 Participants
Last post:
JohnFRWhipple
Contractor Talk - Professional Construction and Remodeling Forum
posts
3.6M
members
169K
Since
2003
A forum community dedicated to professional construction and remodeling contractors. Come join the discussion about the industry, trades, safety, projects, finishing, tools, machinery, styles, scales, reviews, accessories, classifieds, and more!
Full Forum Listing
Explore Our Forums
General Discussion Tools & Equipment Business Painting & Finish Work Decks & Fencing