Ceramic Tile Advice Forums - John Bridge Ceramic Tile

Welcome to John Bridge / Tile Your World, the friendliest DIY Forum on the Internet


Advertiser Directory
JohnBridge.com Home
Buy John Bridge's Books

Go Back   Ceramic Tile Advice Forums - John Bridge Ceramic Tile > Tile & Stone Forums > Professionals' Hangout

Sponsors


Reply
Thread Tools Search this Thread
Unread 07-14-2004, 11:20 PM   #1
cx
Moderator emeritus
 
cx's Avatar
 
Join Date: Jul 2001
Location: Boerne, Texas
Posts: 91,922
Herr Gobis: Long Joist Spans

Dave:

Nearly two years ago you posted the following in a discussion of engineered floor joist deflections and some ongoing testing:
Quote:
There are three research projects in progress that I know of, two by Universities, relative to this issue. We have trusses manufactured at L/600 in a 40' span in which tile failures occur. Is it that the tile has exceeded the maximum curvature? Or is it that the vibration ( racking ) of the floor system is too great? Or a combination of both? I will let you know when I know. These are very expensive studies involving a tremendous amount of time and effort. But, if I am permitted, you will see the results here.
Have you results of that testing that you are permitted to share with the masses?
__________________
CX

Y'ALL NEW VISITORS READ THIS HERE!
cx is offline   Reply With Quote
Sponsored Links
Unread 07-15-2004, 10:00 AM   #2
ninemile
Tile Saw Goddess
 
ninemile's Avatar
 
Join Date: Sep 2002
Location: St. Paul, MN & Savannah, GA
Posts: 508
Thanks CX!

Dave, inquiring minds want to know all about span lengths and engineered trusses. Give us some guidelines please!
__________________
Colleen Staton
Geometric Tile &
Nine Mile Properties, LLC
ninemile is offline   Reply With Quote
Unread 07-16-2004, 04:02 AM   #3
Dave Gobis
da Man!
 
Dave Gobis's Avatar
 
Join Date: May 2001
Location: Racine, WI.
Posts: 5,581
I have some information I can post when I get back in the office maybe later today or this weekend. Little lengthy to type out.
Dave Gobis is offline   Reply With Quote
Unread 07-16-2004, 07:43 AM   #4
Dave Gobis
da Man!
 
Dave Gobis's Avatar
 
Join Date: May 2001
Location: Racine, WI.
Posts: 5,581
TILE FLOORS

Introduction

Though the history of tile dates back many centuries, there are still many unknowns concerning the proper design of wood floor systems supporting ceramic or marble tile installations. Most of the tile industry's recommendations result from in-service floor performance and minimal testing has been completed on full scale tile-floor systems. Therefore, this discussion summarizes some of the current recommendations of the tile industry and reviews some of the unknowns.

Please note that Trus Joist does not allege that the recommendations contained in this document will preclude problems with installed tile floor systems. In addition, there are many conditions not covered in this document that may cause potential performance issues or tile floor failures (cracking of the grout, tile, etc.).


Design of Tile Floors

Tile floor assemblies can be broken up into 4 basic categories:
· Structural Elements (i.e., floor joists, plywood subfloor, concrete subfloor, etc)
· Substrate (i.e., plywood underlayment, cementitious backer unit (CBU), mortar bed, etc)
· Bonding materials and grout (thinset)
· Tile

Tile material is generally hard and brittle and is designed to provide a long lasting floor surface. Therefore, tile must have a support structure made up of the substrate and structural elements to carry the applied load. Bonding materials are used to adhere the tile to the substrate. Loads placed upon the tile floor generate shear stresses in the bonding material and compression and tension stresses in the tile and grout. Stresses are also developed in the tile and at the tile/substrate interface due to dimensional expansion and contraction caused by changes in moisture content and temperature.

Bonding materials can be enhanced with polymer additives to improve the mechanical bond and lower the water absorption of the mortar and grout. Other systems allow the tile to move independently of the substrate. These crack isolation systems are intended to work by reducing the transfer of stresses in the substrate to the tile, bonding materials, and grout.



Tile Floor Failure Modes

Tile floor failures can occur in the substrate, tile, bonding materials or grout and may include one or more of the following (as defined by ASTM C627):

· Popped-up grout joint: Situation where the grout joint remains solid, but has broken loose from the installation.
· Cracked grout joint: A crack in the grout that occurs parallel or perpendicular to the grout line.
· Powdered grout joint: Loss of grout material.
· Loose tile: Loss of tile bond at the substrate/tile interface. This situation creates a hollow sound when the tile is tapped with a metal object.
· Chipped tile: Tile chips or cracks within one tile depth from the tile edges.
· Broken tile: A crack in the tile that cannot be categorized as a "chipped tile."

Design References

Many sources of standards and specifications exist for the ceramic tile and stone industry, including the Tile Council of America (TCA), the Marble Institute of America (MIA), the Ceramic Tile Institute, ASTM and ANSI standards, specifications from the Construction Specifications Institute, and model building codes.

The recommendations provided by the Tile Council of America (www.tileusa.com) and the Marble Institute of America (www.marble-institute.com) are of particular interest to designers. Every year, the Tile Council of America publishes a new edition of the Handbook for Ceramic Tile Installation containing tile specifications for tile installations over wood I-joists. Every two years, new details are considered by the handbook committee for inclusion into the manual. Details are approved by the consensus of the committee. Systems considered for inclusion into the handbook must be well-supported by field experience, laboratory testing and manufacturers' warranties. The Marble Institute of America provides detailed deflection, span and spacing specifications for floor systems with a marble tile covering.


Robinson Test

In order to simulate the in-situ performance of a tile floor system, the tile industry uses the Robinson-type floor tester (ASTM C-627). This test characterizes the performance of a 4' square tile floor assembly by applying an increasing amount of load to three caster wheels equally spaced in a circle with a radius of 15". A drive mechanism turns the wheels over the floor at a constant angular velocity. The test induces vibration in the floor as the wheels roll over the grout joints and investigates grout/tile durability as well as subfloor deflection between joists. While the ASTM C-627 test does provide some useful information about system performance and tile durability, it does not duplicate the in-situ performance of a tile floor system. In the test configuration, joists are fully supported along their length, restricted from rotation along the bottom flange and the entire system is blocked and supported in a way that stiffens the underlayment.
Table 1 describes the loading schedule for the Robinson test. Deflection is measured mid-span between the joists under the wheel path and the performance of the floor is measured by the number of cycles reached before a failure in the grout, tile or mortar occurs. Residential floor systems must pass cycles 1-3. Light commercial floor systems must pass cycles 1-6 and moderate commercial floor systems must pass cycles 1-12. As the wheel hardness increases, the pressure applied to the tile and the likelihood of grout powdering and tile chipping increase.



Table 1. Robinson Test Loading Schedule
Cycle Type of Wheels Load per wheel, lb Duration of test, h Total Revolutions
1 Soft Rubber 100 1 900
2 Soft Rubber 200 1 900
3 Soft Rubber 300 1 900
4 Soft Rubber 300 1 900
5 Hard Rubber 100 1 900
6 Hard Rubber 200 1 900
7 Hard Rubber 300 1 900
8 Hard Rubber 300 1 900
9 Steel 50 ½ 450
10 Steel 100 ½ 450
11 Steel 150 ½ 450
12 Steel 200 ½ 450
13 Steel 250 ½ 450
14 Steel 300 ½ 450

Deflection

Deflection in a floor system occurs along the joists, in the sheathing between joists and in the movement of one joist relative to another. The induced strain in the substrate creates shear stresses in the bonding material and compressive/tensile forces in the grout and tile.

The Tile Council of America's (TCA) deflection criteria recommendations are stated in the 2002 Handbook for Ceramic Tile Installation as:

"Design floor areas over which tile is to be applied to have a deflection not greater than 1/360 of the span when measured under 300 lb. concentrated load."

The L/360 deflection criterion evolved from historical performance of tile installations over sawn lumber joists in which the live load deflection is limited to L/360. However, many sawn lumber floor joist spans are limited by strength rather than deflection. Therefore, many sawn lumber floor joists deflect less than what is allowable, providing false assurance that L/360 is acceptable (not to mention that most floors rarely “see” the full live loads that they were designed to withstand). As mortar beds became less common, it was apparent that subfloor deflection between joists was critical. The tile industry recognized this and promoted the L/360 deflection criteria for both the joist and subfloor design.

Even though the deflection criteria specified in the TCA Handbook refers only to a concentrated load, the intent is to maintain the L/360 deflection criterion for the joist and subfloor under the uniform total design loads of the application and for a 300 lb. point load applied to the subfloor.

It is the opinion of Trus Joist that limiting the amount of curvature in the floor system will reduce tile and grout cracks not related to installation errors. It should also be noted that deflection is not the same as curvature. Table 2 shows an example of this concept for the L/360 deflection criterion. As the span is reduced, the radius of curvature becomes smaller when using the same deflection criterion. A smaller radius of curvature will increase the induced bending stress in the tile.

Table 2: Radius of curvature as a function of span length for a deflection criteria of L/360
Span (ft) L/360 Allowable Deflection (in) Associated Radius of Curvature (in)
20 0.667 10,795
16 0.533 8,646
12 0.400 6,480
2 0.067 1,075
1 0.033 545
*Table assumes simple span conditions

Table 2 indicates that the transverse deflection of the subfloor is more critical due to the smaller radius of curvature, which induces a higher bending stress in the tile. This stress is further compounded by the multiple span condition of the subfloor, which induces tensile stresses in the tile surface. Full scale tests of properly installed tile floor systems support this rationale.

Differential Deflection

Differential deflection is the deflection of one joist relative to another. Situations that cause differential deflections are varied, but need to be considered when designing either a marble or ceramic tile floor system. The article, "Field Investigations of Engineered Lumber Products," written by Glyn R. Boone, P.E. and published in the spring 2002 edition of Wood Design Focus reviews some actual floor plans that resulted in excessive differential deflection. Larger on-center joist spacings may allow differential deflection issues to develop depending on the type and location of loads that are applied to the floor. Situations that cause excessive differential deflection need to be avoided or accommodated for in the floor design.

Moisture
Trus Joist products are intended for dry-use conditions only. Moisture contents outside of this range will negatively affect both solid-sawn and engineered wood I-joists. Sustained moisture conditions in excess of dry-use levels will produce a higher initial deflection and greater long-term deflection due to creep. Once the member returns to dry service conditions, the deflection induced is essentially permanent and the member will perform normally under additional load. Moisture in the floor system will also cause stresses to develop because of the differing expansion rates of the tile, grout and substrate.

Few tile installations are actually waterproof, so unless a waterproof membrane is installed, moisture from repeated heavy wetting of the floor may become trapped in the floor system and increase the potential for problems.

Radiant Heating Systems

Radiant heating systems are often installed as part of a tile installation. They can be in the form of heating pipes, low voltage electric mats or electric cable systems. While excessive heating and cooling cycles can create problems from differential expansion and contraction, radiant heating systems do not approach temperatures high enough to pose problems with tile floor installations. Anti-fracture membranes are often used in these applications since the heating pipes create weak points in the concrete. The anti-fracture membrane may help to isolate the tile system from concrete stresses caused by lateral movement.

Sound Rated Assemblies
Building codes require floor assemblies that separate dwelling units or guest rooms from each other and from public spaces to provide insulation from airborne and impact sound. Appendix Chapter 12, Division II of the 1997 Uniform Building Code requires residential floors to meet an airborne sound transmission class (STC) of 50 and an impact transmission class (IIC) of 50. Various companies produce sound control products for use with tile that have been verified through testing to meet or exceed the building code requirements. Refer to the manufacturer's literature for a list of approved sound rated assemblies.

Marble Tile
Marble floor coverings create additional concerns since marble contains natural veins which results in poor bending strength in comparison with ceramic tile. In addition, marble surfaces contain dust and must be installed properly or a poor bond will result. Because of this dust, the stone must be wiped clean or the mortar can be applied to both the surface to be covered and back-buttered to the back of each stone.


Design Recommendations

Deflection Criteria:

Ceramic Tile: L/360 total load deflection limit

Stone: L/720 total load deflection limit, 7/32" max. (13'-1" max. span when using L/720 criteria)

O.C. joist spacing:

Ceramic Tile: 16" max. for 5/8" plywood subfloor and ½" underlayment, 24" max. for ¾" plywood subfloor and ¾" underlayment. Trus Joist recommends 16” maximum joist spacing. Mortar beds and cementitious backer units (CBU's) are acceptable as wood underlayment substitutes for joist spacings up to 16”.

Stone: 16" max. for 5/8" plywood subfloor or combination of ¾" plywood subfloor and cementitious backer unit underlayment

Creep: When the dead load exceeds 1/3 of the live load, use a 1.5 deflection factor with the long term component of the design load

Composite Action: When calculating the allowable floor span, do not include a composite action factor. The increased span can cause measurable and perceptible vibration.



Common material weights (psf):
Gypsum board (per 1/8" thickness) 0.55
5/8" subfloor 2.0
¾" subfloor 2.5
¾" ceramic or quarry tile 10.0
1" mortar bed 12.0
¼" CBU underlayment 2.0
½" CBU underlayment 3.0



Tile Floor Installation Guidelines

Subfloor/Underlayment:
· Allow a gap of 1/8" between sheets for subfloor and underlayment.
· Install face grain of subfloor perpendicular to joists.
· End joints must occur over framing members. Subfloor edges must be tongue-and-groove.
· Glue subfloor to joists and underlayment to subfloor using a construction adhesive.
· Currently, ANSI does not allow the use of engineered panels (OSB) for the subfloor or underlayment.
· Use a subfloor with the appropriate span rating.
· Use an underlayment type that promotes a proper bond with the thin-set mortar. Protect the underlayment from damage or contamination.
· Avoid overloading the floor beyond the design dead and live loads.
· Avoid situations in which differential deflection exceeds the deflection criteria. Using a closer joist spacing or doubling up joists in the transition area may be used to reduce differential deflection.

Mortar bed/Tile system:
· Follow manufacturer's instructions when mixing mortars. Mortars mixed with improper amounts of water or additives may not bond properly.
· Do not allow the mortars to skin-over prior to placement of the tile.
· Beat tile into mortar bed for proper bonding.
· Use proper trowel notch size for the application.
· Use sufficient setting material under the tile (especially under large unit tiles).
· Back-butter large format tiles to insure a minimum of 85% continuous coverage in dry areas and 95% continuous coverage in areas exposed to water such as bathrooms.
· Allow adequate time for tile to set before grouting.
· Allow sufficient time for curing before construction traffic is allowed on the floor (refer to Manufacturer’s recommendations).

Movement Joints:
· Install movement joints where tilework abuts restraining surfaces such as columns, pipes or perimeter walls and at cold joints (These joints can be located by inserting a pen into the joint. If the joint is soft, it is a movement joint.).
· Follow the TCA recommendations in EJ171 for the proper design, installation and placement of expansion joints in the tilework.

Expansion/Contraction
· Avoid situations that result in excessive shrinkage or expansion (such as repeated drying/wetting or heating/cooling cycles) of the underlayment causing tiles to lose bond.



References:

American National Standards Institute. American National Standard Specifications for the Installation of Ceramic Tile. 1999.

American Society for Testing and Materials. ASTM C 627-93, Evaluating Ceramic Floor Tile Installation Systems Using the Robinson-Type Floor Tester. 1999.

APA - The Engineered Wood Association. Plywood Design Specification. 1998.

Ceramic Tile Institute/Ceramic Tile Institute of America. Ceramic Tile Institute's Tile Manual. 1991.

Hua G, Taylor SB. Dynamic performance of wood framed floor systems with poured toppings.

Proceedings of the World Conference on Timber Engineering; 2000. Whistler, British Columbia, Canada.

International Conference of Building Officials. 1997 Uniform Building Code. 1997.

Marble Institute of America. Installation of Modular Stone Floor Tile: Thin Set Method. 2002.

Tile Council of America. 2002 Handbook for Ceramic Tile Installation, 2002.
Dave Gobis is offline   Reply With Quote
Unread 07-16-2004, 09:16 AM   #5
bbcamp
Registered User
 
Join Date: Jul 2002
Posts: 30,274
Quote:
Even though the deflection criteria specified in the TCA Handbook refers only to a concentrated load, the intent is to maintain the L/360 deflection criterion for the joist and subfloor under the uniform total design loads of the application and for a 300 lb. point load applied to the subfloor.
AH-HA! Just as I thought!




Quote:
Composite Action: When calculating the allowable floor span, do not include a composite action factor. The increased span can cause measurable and perceptible vibration.
This is why you should use L/480 when evaluating I-joists and trusses. The manufacturer's span tables usually assume composite action of the subfloor glued and nailed to the joists (Those sneaky devils!).




Nothing different about long spans (say, over 15 feet)?

Nothing different about large format tiles (bigger than 16 inches)?
bbcamp is offline   Reply With Quote
Unread 07-16-2004, 09:46 AM   #6
ninemile
Tile Saw Goddess
 
ninemile's Avatar
 
Join Date: Sep 2002
Location: St. Paul, MN & Savannah, GA
Posts: 508
Ok my eyes are crossed tryin to read all of the above can someone please tell me if I can do a tile floor with the following conditions.

19.2" O.C., 2"x4" trusses, 18" high. Every 10ft is a 2x10 running through them to tie them together. 12"x12" porcelain with some smaller cut pieces for a border. 3/4" subfloor (ply ext grade) and I'll add 1/2" more ply under the Ditra matte.

Span is 25ft which is what concerns me, building is 2 years old. Location of tile is in the corner I'dd get more info off the trusses but I'm out of town now.
__________________
Colleen Staton
Geometric Tile &
Nine Mile Properties, LLC
ninemile is offline   Reply With Quote
Unread 07-16-2004, 11:47 AM   #7
John Bridge
Mudmeister
 
Join Date: Mar 2001
Location: Rosanky, Texas
Posts: 68,140
Send a message via AIM to John Bridge
Colleen,

The article doesn't address overall joist span, but I'll tell ya I have installed a bunch of floors with the specs you mention. I've also installed floors over two foot centers where the subfloor is a single layer of 1- 1/8 in. Sturd-I-Floor (tongue and groove flooring plywood). With your two layers and Ditra, I don't see how you can go wrong.

I've taken to installing the second layer of plywood in the manner explained in Woeste's article. The butt joints do not occur over the joists, and the screws do not go into the joists, only into the subfloor.

Bob, Dave,

I still think the biggest problem we have is that no one seems willing to come out and put a label on a piece of plywood when it comes to the amount of deflection or curvature that can be expected. You can see what an article like this does for someone in the field like Colleen. It doesn't even come close to telling her whether her floor is going to work. (It'll work, Colleen.)
John Bridge is offline   Reply With Quote
Unread 07-16-2004, 11:55 AM   #8
John Bridge
Mudmeister
 
Join Date: Mar 2001
Location: Rosanky, Texas
Posts: 68,140
Send a message via AIM to John Bridge
Also, Colleen,

I checked with Woeste as to whether gluing the underlayment to the subfloor might somehow short-circuit the idea of allowing the underlayment to "float." Woeste says it doesn't, so in view of the astonishing increase in stiffness that can be attained by gluing layers of plywood together, I would go for the gusto. Glue and screw your 3/8 to the 3/4. I will do it on my next job.
John Bridge is offline   Reply With Quote
Unread 07-16-2004, 12:10 PM   #9
bbcamp
Registered User
 
Join Date: Jul 2002
Posts: 30,274
Colleen, you may have overlooked my reply to your other thread. Your floor is good, so have at it!
bbcamp is offline   Reply With Quote
Unread 07-16-2004, 12:33 PM   #10
ninemile
Tile Saw Goddess
 
ninemile's Avatar
 
Join Date: Sep 2002
Location: St. Paul, MN & Savannah, GA
Posts: 508
John, thank you for answering my question! I will glue and screw the heck out of that 1/2" underlayment.

While I enjoy the engineering listed above; I was at the point where all I wanted is a Yes or No so I could get the flooring job to go with the Kerdi $6k fancy shower job!

It's too bad the Deflecto meter can't be revised to include truss info so we don't have to ask the questions or make the wrong proposal to a customer. I did try a Google search but got lost in all the data. I want to make sure that every project I do will last another 100 years, especially since most of my work is in historical buildings and I ain't cheap.
__________________
Colleen Staton
Geometric Tile &
Nine Mile Properties, LLC
ninemile is offline   Reply With Quote
Unread 07-16-2004, 01:34 PM   #11
Dave Gobis
da Man!
 
Dave Gobis's Avatar
 
Join Date: May 2001
Location: Racine, WI.
Posts: 5,581
Just from experience, I know L/900 has been recommended in long spans with 24" and two layers of plywood.

John B, you are going to get your wish on panel ratings soon from the lumber side. Till the wood boys are done, the tile boys can't do much. There are some ASTM's coming. I would suspect it will be several years from an optimist view, when you may see L/360 disappear. As Dr Woeste points out, and his peers agree, the issue is curvature. Currently, ASTM C-627 ( the Robinson Floor Tester ) is the only device used on a 300# load. The wood industry uses primarily uniform load, not point load. It is all coming together, just takes time.

Last edited by Dave Gobis; 07-16-2004 at 01:43 PM.
Dave Gobis is offline   Reply With Quote
Unread 07-16-2004, 01:45 PM   #12
John Bridge
Mudmeister
 
Join Date: Mar 2001
Location: Rosanky, Texas
Posts: 68,140
Send a message via AIM to John Bridge
You have to remember the Deflecto-lator was designed and built by engineers. Dave Misevich built it with Bob Campbell's help. Do you think for a minute those guys are going to design themselves out of a job?

L/900. Here again, Dave, it doesn't do us a lot of good when we have no way of knowing how much subfloor, etc, L/900 represents. I'm glad to hear the plywood folks are finally coming around.
John Bridge is offline   Reply With Quote
Unread 07-16-2004, 02:02 PM   #13
John Bridge
Mudmeister
 
Join Date: Mar 2001
Location: Rosanky, Texas
Posts: 68,140
Send a message via AIM to John Bridge
and speaking about subfloors, I've got to show you this one. Not only did the painters use this room as their shop, but check out the butt joint. When we pulled up the rug, we found there were no nails in one sheet of plywood (OSB, actually) and very few in the one that abuts it. Carpet layers didn't seem to mind.
Attached Images
 
John Bridge is offline   Reply With Quote
Unread 07-16-2004, 05:20 PM   #14
Steven Hauser
STT Owner
 
Steven Hauser's Avatar
 
Join Date: Mar 2003
Location: Greenville, South Carolina
Posts: 5,156
I don't know if I should ressurect some of my posts about l/900 and failures due to curvature and failures pointing at the joints between the plywood.

Yep, the plywood guys can start but the tile guys will have to finish the study.
__________________
Clean
Sealed
Safe
Products and Practices Perfected™
STT Sealers with Attitude!®
Steven Hauser is offline   Reply With Quote
Unread 07-17-2004, 01:36 PM   #15
Bill Vincent
Tile Contractor
 
Bill Vincent's Avatar
 
Join Date: Jan 2003
Location: Bridgton, Maine
Posts: 8,631
Quote:
Originally posted by John Bridge
Also, Colleen,

I checked with Woeste as to whether gluing the underlayment to the subfloor might somehow short-circuit the idea of allowing the underlayment to "float." Woeste says it doesn't, so in view of the astonishing increase in stiffness that can be attained by gluing layers of plywood together, I would go for the gusto. Glue and screw your 3/8 to the 3/4. I will do it on my next job.
Wouldn't that be ONLY if you're going to use CBU as a bonding surface? I would think that if the tile were to be laid over the underlayment that this would defeat the purpose of the double layer subfloor.
__________________
Bill

"Whether you think you can, or you think you can't, you're probably right."

http://www.creativeceramicandmarble.com
Bill Vincent is offline   Reply With Quote
Reply

Stonetooling.com   Tile-Assn.com   National Gypsum Permabase


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -6. The time now is 06:42 AM.


Sponsors

Powered by vBulletin® Version 3.7.4
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.
Copyright 2018 John Bridge & Associates, LLC