Porcelain tiles are typically made with "porcelain" clays that have specific properties. Typically, these tiles are dense and by definitio, they have water absorption of 0.5% or less. Non-porcelain tiles have water absorption greater than 0.5%.

Because porcelain tiles have a low water absorption, they are usually frost resistant, although, not always. To know if a tile is frost resistant, you should check the manufacturer's literature.

There are also many non-porcelain tiles that can be used in freeze thaw environments and that are manufactured with properties similar to porcelain tiles.

There are both glazed and unglazed porcelain tiles. It is important to know the difference, as the glazed variety is usually a little easier to clean. Typically, glazed porcelain tiles have filled in microscopic holes that could be present in the unglazed tile. On the other hand, unglazed porcelains may have better slip resistance.

Non-porcelain tiles cover a wide range of properties. Typically they are glazed (unglazed quarry tile is the exception), and the glaze layer can be extremely durable. However, as there are differences from one glaze to another, it is important to check if the tile has been tested and to make sure the glaze hardness is suitable for your application.

In general, non-porcelain tiles are easier to bond to the floor and usually easier to cut. Porcelain tiles are harder to bond and harder to cut. While this can be relevant to the tile installer, it generally makes little difference to the end-user, so long as the installer uses the right materials.

Some people refer to unglazed porcelain tile as "through body" ( i.e., the color on the top goes all the way through). Even in extreme applications, these tiles tend not to show wear as the porcelain is quite durable (harder than granite), and the color goes all the way through.

Many glazed porcelains also have extremely good durability. Although the color in the glaze layer may be different from the body, the surface is usually sufficiently resistant to abrasion to not show wear in typical applications.

Since 1999, U.S. and European manufacturers have been using the same testing method for determining glaze wear resistance - with a value of 4 (on a scale from 0 to 5) being good for almost all applications except the most abrasive and dirty environments. However, lower ratings are also fine depending on where the tile will be used and how much traffic and outside dirt (especially sand, because it is abrasive) will be present.

A rating of 4 can be achieved if there is no visible wear (under test conditions) after 2100, 6000, or 12000 revolutions of the test equipment. A value of 3 can be achieved by passing 750 or 1500 revolutions. Usually the product specifications will indicate which value was passed when the testing was done (for example, one tile might be rated Class 3, passing 1500 revolutions, another tile could be Class 3, passing only 750 revolutions).

Whenever tile is bonded to concrete, cracks occurring in the concrete can cause cracks in the tile layer - this is often called "reflective cracking."

The TCNA Handbook for Ceramic, Glass, and Stone Tile Installation recommends that control joints in concrete carry through the tile. Clearly, this is an industry-approved, nearly foolproof, and very safe way of making sure that movement in the control joints does not cause a reflective crack.

Due to a lack of consensus in the industry regarding competing anti-fracture products and the standards of performance for such, TCNA does not, at present, recommend a method or tiling over control joints with an anti-fracture membrane. This is not to say this process will not work - it can if the right products and methods are used and the slab does not continue to curl.

If the slab does curl at the control joint (which is not uncommon), any curling that occurs after tiling may damage the tile.

Some manufacturers have proprietary products where they will guarantee a tile installation (when their products are used in accordance with their methods) over control joints, so long as vertical deflection does not occur. However, expansion joints must be used in the tile layer.

It is a frequent misconception that anti-fracture membranes allow you to eliminate movement joints - they do not. There always must be soft joints in the tilework to allow for expansion and contraction.

Typically the joints in the tile installed over an anti-fracture membrane must be placed near the joints in the concrete but not necessarily directly in line with the control joints.

Even small shrinkage cracks in concrete can be dimensionally active; continued curing of the slab can cause these cracks to expand or propagate. This type of cracking can be easily avoided in the tile layer - either by installing the tile on a mortar bed set over a cleavage membrane (method F111 for example), or by installing the tile over a crack isolation membrane using a thinset method.

The internal make-up of this membrane is such that movement in the concrete is not directly transferred to the tile. Although the membrane is bonded to the concrete and the tile to it, the membrane stretches where needed to prevent or reduce force transference. These membranes are either trowel applied or sheet applied. In many cases, multiple components or steps are part of the system. Performance varies also - it is important to check with the crack isolation membrane manufacturer regarding their installation instructions and intended use.

Some contractors have used felt paper as an inexpensive type of anti-fracture membrane. Unfortunately, this type of installation generally does not provide suitable bond strength between the tile and the floor, does not hold up to moisture, and can promote fungal growth.

Products made specifically for crack isolation are sold for everything from isolating the tile from shrinkage cracks to tiling over control joints to protecting the entire floor from potential cracking in the concrete.

When used to cover the entire floor, many manufacturers will warrant the entire floor installation, including the cost of replacing and installing new tile.

As noted previously, it is a frequent misconception that anti-fracture membranes allow you to eliminate expansion joints - they do not. There always must be soft joints in the tilework to allow for expansion and contraction. Guidelines for expansion joint placement are given in the TCNA Handbook for Ceramic, Glass, and Stone Tile Installation. The exact placement of expansion joints is a function of many items including exposure to sunlight and the range of high to low temperature, moisture, aging of the concrete (where relevant), structural movement, expected loading, and other design criteria. The manufacturer's recommendations should also be consulted and followed.

Whether or not a person should ever set thinset and tile to the glue left behind when non-asbestos vinyl tile was removed depends on the type of glue left behind.

Physically removing all coatings is the most reliable way to make sure you have an appropriate surface to which to bond tile but the old asphaltic (black) adhesives (often called "cut-back") can be very difficult to remove unless you use shot-blasting equipment. Fortunately, most grout and mortar manufacturers make a thinset which will bond to "cut back" adhesive.

Typically, the "cut back" should be scraped smooth so the tile is not sitting on ridges of adhesive - although it does not have to be absolutely flat. Note, not all thinsets will bond to cutback so it is important to make sure the thinset used is made for that purpose.

The adhesive (often called carpet mastic) used under most glue-down carpet is usually water-soluble and tends to prevent thinset from properly adhering. Should the tile get wet, the mastic may soften, negatively affecting the bond of the tile to the substrate. Thinset manufacturers recommend removing all traces of carpet adhesive if you plan to bond the tile directly to a surface that previously had glue-down carpeting.

Technically, anywhere there is a change in substrate or backing surface such as the joint between walks and floor and wall joint, caulk should be used in place of grout since these surfaces move independently of each other. However, it is important to recognize and make the end user aware of some important points.

Often, installers use grout in place of caulk for these reasons:

The caulk may not exactly match the grout color.

Even when the caulk exactly matches the grout color when installed, it may not match six months later (caulk will "age" differently from the grout).

Caulk will need to be maintained more often than grout.

Mold may grow more easily on caulk (except caulk treated with mildewcide) than on grout.

Acrylic caulks break down in horizontal wet applications. Silicone, urethane, or multi-polymer caulks are better choices but can be harder to apply.

However, when grout is used in place of caulk, the grout can cause structural and aesthetic problems.

We are often asked about epoxy grout and mortar versus conventional grout and mortar.

Epoxy grout (meeting ANSI A118.3) is quite different from cementitious grout and epoxy emulsion grout. Made from epoxy resins and a filler powder, the grout is extremely hard, durable, and nearly stain proof. Often times the bond between tiles is stronger than the tile itself.

You might wonder why this type of grout is not used all the time. First, most installers find it harder to use than cementitious grout. Also, it has a more plastic appearance which, as with all matters of aesthetics, some people like and some don't. Also, it is much more difficult to shape and slope; this can be done easily with cementitious grout and is often needed to transition from one tile to another. It may also slump in the joint hours after the floor is finished because the grout becomes less viscous initially as it heats up and cures. Lastly, it generally takes days longer to cure and must be kept rigorously clean. And it can cost three to eight times as much as cementitious grout.

There are even epoxy grouts on the market impregnated with Teflon that are both stain proof (nearly) and wipe clean incredibly easily.

It is also possible to smooth epoxy grout (with or without Teflon) before it cures in a way that leaves the surface with an extra slick plastic finish to which it is very difficult for dirt to adhere.

There are many things that can cause excessive deflection in your subfloor (and consequent cracking in the tile) or you may have a perfectly sound subfloor but not have prepared the floor properly for tile.

Here are a few of the most common questions:

Is the subfloor plywood over joists 16" on center? If not, has the installation system been designed to work with the actual type of subfloor present?

What is the span of the joists? Are they suitably sized for the span to achieve the L/360 deflection standard under the expected live and dead load? Are there any cracked, rotted, or termite damaged joists?

Was the subfloor screwed to the joists? Is there any possibility of movement between the subfloor and the joists themselves?

Does the thinset used match the conditions present (was a polymer additive used and if so was it appropriate for the subfloor?)

Was the thinset coverage satisfactory? What was the notch size of the trowel used?

Were expansion joints used in the installation to allow for normal movement?

Are any dimensionally unstable or questionable materials also in the tile/subfloor/joist sandwich? How about cushion vinyl, luaun, water-soluble patching compounds or mortar materials.

Were all layers present installed according to the applicable ANSI standards?

We are often asked what is the standard for determining the size of a grout joint - can it be bigger or smaller - does it need to be a specific size, etc.

There is a relevant standard in the ANSI A108/A118/A136.1 manual. You will see that it is not specific; however this is the only part of the standard addressing joint spacing.

ANSI A108.02, Section 4.3.8 - Nominal centerline of all joints should be straight and of even width with due allowances for hand-molded or rustic tiles.

In general, there is not a specific standard for the size of a grout joint. However, there are many relevant parameters that should be considered.

What is the amount of variation from tile to tile?
Are the edges of the tile linear or irregular (e.g. "hand-molded")?
How big is the tile?
What is the surface of the tile; can it be easily scratched?
Where is the tile being used?
Is the surface lefvel?

Ultimately, the project owner should choose the grout joint they desire, keeping in mind that a tighter grout joint will show more variation from tile to tile. Many people feel that a joint smaller than three times the average variation from tile to tile (or two times the maximum variation) tends to look irregular and poorly installed.

When evaluating grout joints, it is important to consider that the grout is used to adjust for differences in the following:

Variations in the size of the tile
Changes in the plane of the substrate
Changes in the thickness of the tile (often this applies to hand-molded tile)

Uneven grout color is not normal, but there can be a number of things that can cause blotchy grout. Also, how uneven the color appears can be somewhat subjective. Certain color grouts are not as easy to control. The easiest colors to work with are the medium grays and beiges as opposed to bright colors, whites, blacks, etc. Since most grout is cement based, it tends to turn gray with time and cleaning. Grout sealers are now recommended to reduce this effect.

Some of the things that can cause color variation include:
Too much water added in mixing the grout.
Improper cleaning of the grout and grout haze off the tile (usually by wiping too soon and with too much water).
Leaving plastic spacers in the joints.
Having the thinset adhesive fill part of the joint and showing through.
Improper mixing of the grout by using an improper amount of water.
Mixing partial bags (this is a problem if the colorant is not evenly distributed).
Using grout from a bag that was previously opened (and gained moisture during storage).
Adding water after the first mixing (rehydrating).
Failure to slake the grout.
Improper joint packing.
Inadequate cleaning of the sponges.
Using unclean water.
Defective grout itself (this happens once in a while).
Dirty tools or shoes that stain the grout.
Foot or other traffic too soon over the joints. Curing time is on the bag but can be up to three days for some grout.

Sometimes the tile itself can have a microporoisty (small pores in the tile body, especially with porcelain tile) condition that sucks the moisture out of the grout too fast or in an irregular pattern. In many cases it is hard to tell exactly what went wrong by looking at the tilework. I would suggest that you ask the tile installer what he or she thinks about the workmanship and techniques used. One can always remove the grout and replace it if that is the only acceptable solution.

Uneven grout color is not uncommon, although in this case, the darkening may be caused by water (more about that later).

This is in part because some colors are more prone to uneven color than others due to the dyes used. Also, depending on the tile, it may be impossible to allow the grout to sufficiently cure for best color uniformity and still remove it from the tile. In other words, the grout must be removed before the color has a chance to set.

However, there are many installation mistakes that can cause uneven color including improper mixing of the grout (typically by using too much water), mixing partial bags (this is a problem if the colorant was not evenly distributed), using grout from a bag that was previously opened (and gained moisture during storage), adding water after the first mixing (re-hydrating), failure to slake the grout, improper joint packing, using too much water too soon when wiping the joints, inadequate cleaning of the sponges, using unclean water, allowing traffic on the floor too soon, etc.

Where absolute color uniformity is desired, epoxy grouts are often recommended (they are much more expensive though) as they cure and clean up differently from cementitious grout.

There are no ANSI standards that address color variations in grout.

This is in part because some colors are more prone to uneven color than others due to the dyes used. Also, depending on the tile, it may be impossible to allow the grout to sufficiently cure for best color uniformity and still remove it from the tile. In other words, the grout must be removed before the color has a chance to set.

However, there are many installation mistakes that can cause uneven color including improper mixing of the grout (typically by using too much water), mixing partial bags (this is a problem if the colorant was not evenly distributed), using grout from a bag that was previously opened (and gained moisture during storage), adding water after the first mixing (rehydrating), failure to slake the grout, improper joint packing, using too much water too soon when wiping the joints, inadequate cleaning of the sponges, using unclean water, allowing traffic on the floor too soon, etc.

Where absolute color uniformity is desired, epoxy grouts are often recommended (they are much more expensive though) as they cure and clean up differently from cementitious grout.

Unfortunately, with regular cementitious grout, there is little that can be done once the grout is in place except for attempting to use a grout colorant. Grout colorants work best with grout that has not been sealed. Grout that has been sealed or has been washed with oil-based soaps (Pine Sol, Murphy's Oil, etc.) can be very difficult (or impossible) to color.

The edge of the tile also makes a difference in the success of the colorant - tiles with well-delineated edges are easier to treat than tiles with a large bevel or textured edge. When the colorant is applied, some will get on the tile. The easier it is to remove from the tile (and the better it sticks to the grout), determines in part how good the finished result appears.

Grout treated with a colorant also has a different texture than originally colored grout. On the plus side, grout colorants usually seal the grout in addition to changing its color. Typically, grout that has been treated with a colorant does not need to be sealed.

Removing grout that is adhered to a tile floor can be difficult. The type of tile greatly affects the difficulty of grout removal. Also, if the grout was polymer modified, it may be more difficult to remove.

In general, the more porous the surface, the better grout will adhere. Conversely, grout is more easily removed from dense impervious tiles (e.g. porcelain).

To remove the grout, start with an alkaline cleaner and a nylon scrub pad. Make sure to check that the scrub pad is not damaging the tile. Normal floor tile will not be affected by a using a scrub pad, but some decorative tiles do not have the same surface hardness. It is best to check your decorative in a secluded area.

If the scrub pad is not effective, there are specialty cleaners on the market that chemically attack the grout. Typically these are weak acids. As with all acids, follow the manufacturers warnings carefully and use caution. Always check the tile in an inconspicuous spot first in case the cleaner affects the tile.

Again, these specialty cleaners will not affect most floor tiles; however, it is prudent to check.

Some tile installers use stronger acids that they carefully dilute. While experienced professionals can do this, there are great risks in doing so. There is the possibility of bodily harm as well as damage to the surroundings.

Changing grout color is more commonly done but; again, the results are generally not as good as the original item. The color in grout, unlike tile, comes from liquid dispersed pigments. Obviously, these are not fired but rather become part of the cement/sand matrix. Grout is usually colored with an epoxy paint made for the purpose and sold in tile shops. When the grout is new, has not been sealed, and the edge of the grout joint is neatly defined and when the adjoining tile surface is very smooth, sometimes good results can be achieved. However, if the grout is not porous (from sealer or dirt) or the adjoining tile is rough or absorptive, it may be impossible to get a satisfactory result.

When grout has been stained to the point that it cannot be maintained or returned to its natural color, you can return the grout back to near its original color or any other color through the use of a "grout stain." Some grout manufacturers make grout colors. Others will recommend specific brands that they know work with their grout to correct color.

However, grout colorants work best with grout that has not been sealed. Grout that has been sealed or has been washed with oil-based soaps (Pine Sol, Murphy's Oil, etc.) can be very difficult to color.

Grout Stains are epoxy-based products that are specifically designed to penetrate into the grout and seal the surface with a permanent color. Once the grout has been stained, there is no need to seal it any further with a penetrating/impregnating sealer.

Prior to staining, the grout joint should be cleaned thoroughly to remove any dirt, oils, grease, or sealers with a professional strength Tile & Grout Cleaner. This can be purchased from most Home Centers or through your local Professional Floor Covering Dealer.

The edge of the tile also makes a difference in the success of the colorant. Tiles with well-delineated edges are easier to treat than tiles with a large bevel or textured edge. When the colorant is applied, some will get on the tile. The easier it is to remove from the tile (and the better it sticks to the grout), determines in part how good the finished result appears.

Also, you will want to try a test area since grout treated with a colorant does not look the same as originally colored grout. On the plus side, grout colorants also seal the grout and protect it with an "epoxy-like" finish. Typically, grout that has been treated with a colorant does not need to be sealed.

Cementitious grout, as you may have observed, is porous - it can absorb a stain. Looked at under a microscope, there is a large surface area to absorb stains. For this reason, many owners choose to seal their grout - usually the better the sealer, the more the grout joint is protected. Even better, if epoxy grout is used, it is virtually as stain proof as the tile.

Removing stains from cementitious grout is similar to removing stains from clothing. The same cleaners you might use on clothes to get out a stain should also work on grout.

Keep in mind though, that grout is based primarily of cement and sand. Sand, like glass, is unaffected chemically by most cleaners. Cement is not - rather it is alkaline based and is dissolved by acids. As baking soda and vinegar react, so do grout and vinegar.

Accordingly, it is better to clean grout with an alkaline cleaner (Spic and Span, Mr. Clean, etc.) than an acid based cleaner. There are also specialty cleaners available at most tile retailers that are designed for tile and grout. There are also cleaners with enzymes that attack stains similar to enzyme pre-soaks for laundry.

The same cleaner that works on the grout generally will work well on the tile. In fact, since the tile is usually so easy to clean, the tile can often be cleaned with water.

Just a few more important points: As the grout can absorb the soap as well as a stain, do not clean with oil or wax based cleaners (Murphy's Oil soap, Pine Sol, etc.). These products will leave a waxy or oily film in the grout. Even good alkaline cleaners, if not properly rinsed, will leave a sticky soap film. This usually attracts dirt. In fact, truly clean ceramic tile without any sticky soap film will stay very clean as tile does not tend to hold an electrostatic charge (which can attract some kinds of dirt).

The absolutely best way to clean grout is to apply the cleaner and then vacuum ("shop vac") up the dirty water. This lifts the dirt off the joint. Apply rinse water and vacuum that water up. This lifts off any remaining soap film.

Just to mention it, there are tile installers that remove very stubborn stains on grout with an acid (like straight vinegar or a stronger acid). There they have elected to dissolve the top layer of grout molecules so the stain is no longer attached to anything. While this works, it is not recommended by the grout manufacturers - needing to regrout is sometimes the result. Also, extreme care should be used when handling any acids.

Should you be unable to get your grout clean through conventional methods, you may also want to try steam. Some stains that do not respond to conventional cleaners will come clean when subjected to pressurized steam. As a last resort, some installers elect to cut out the grout and regrout. This is possible although care must be taken to not damage or loosen the tile. Generally it is not possible to grout directly over the old grout without cutting the old grout out. The same contaminants that made the old grout dirty may prevent new grout from sticking properly.

To prevent staining in the future, you should seal the grout.

Generally, sealer is a very good idea for cementitious grout (regular grout - not epoxy grout). For glazed floor tile, it is not a good idea to spray anything on the tile - the glaze of the tile will be easier to clean and longer lasting than any coating. For unglazed tile, generally sealers are recommended, although it is important to follow the recommendations of the tile manufacturer.

For cementitious grout, there are two broad classes of sealer: penetrating sealer that chemically bonds with the grout and repels water (and water based stains) and topical sealers that coat the surface of the grout and repel almost everything (until they are worn off by foot traffic).

Each type of sealer has its advantages and disadvantages. Additionally, there are hybrids on the market combining advantages.

In general, the topical sealers are less expensive but give the grout a plastic appearance. Also, they are subject to wear and tear and very sensitive to water in the grout while curing. As stated above, the plastic coating does block almost everything until it is compromised by foot traffic.

The penetrating sealers are more expensive but also more durable. There are also penetrating sealers that repel oil based stains that are even more expensive. They can be applied on the grout sooner than the topical sealers, as they are usually vapor permeable. As they do not coat the grout (but penetrate in), they do not change the large microscopic surface area. While stains don't penetrate, they can be a little harder to remove (just a little) because the sanded texture of the grout hasn't been changed.

Thinset mortar is a blend of cement, very finely graded sand, and a water retention compound that allows the cement to properly hydrate. Tile set by the thinset method is adhered to the substrate with a thin layer of "thinset" cement. The terms thinset cement, thinset mortar, dryset mortar, and drybond mortar are synonymous. This type of cement is designed to adhere well in a thin layer - typically not greater than 3/16th thick. For example, a 3/8" notch trowel will produce a 3/16th inch thick coating after the tiles are pressed in to the cement. While very minor adjustments in height can be made, this method is not appropriate for adjusting the level or flatness of a surface - rather the tile will follow the plane of the substrate.

The American National Standards Institute (ANSI) defines the properties of thinset mortar in the A118.1 specification.

Medium bed "thinset" mortars can be used to adjust for slightly larger variations in the substrate than can be accommodated with thinset mortar. They are also used with large, heavy, or thick tiles where a thicker setting bed and a coarser aggregate may be required to achieve a flat installation and to support the weight of the tile while the cement is curing.

Thick-bed installations are based on the traditional method of packing a mortar bed over a surface before installing the tile. The tile is adhered to the mortar bed either while the mortar bed is green (just beginning to dry) or after the mortar bed has cured. The mortar bed may be reinforced with wire and either set over a cleavage membrane (that allows the mortar bed to "float" free of the substrate) or bonded to the substrate; hence, the use of the terms "floating mortar bed" or "bonded mortar bed". For wall applications, metal lath is mechanically anchored to the substrate, and the mortar locks into the metal lath as it cures. The terms thick-bed installation, mortar bed installation, and thick-set installation are synonymous.

In the case of the floating mortar bed, the tile layer is unaffected by minor cracking and movement in the substrate. This can be very important in applications over concrete where cracking in the concrete could result in cracking in the tile. It is also very important in tile installations over structural slabs (not on grade) or other structures where vibration and deflection can be expected (as occurs in some exterior walls).

Mortar beds also allow for the following:
Level out unevenness in the substrate.
Create an ideal surface to which the tile can be bonded.
Incorporate slope in the tile layer if needed (e.g., slope to a drain).
Reinforcement of the substrate (usually relevant in wood framing applications).
Allow radiant hydronic tubing to be installed.
Protect metal, PVC, or CPE waterproof pans.

As bonded mortar beds do not float over the substrate, they do not offer protection from cracking or movement in the substrate. However, they do offer the same advantages listed above and in many cases can be safely installed in a thinner layer than a floating mortar bed.

Note: Anti-fracture membranes used in thinset or bonded mortar bed applications can also protect the tile layer from most kinds of cracks in the substrate.

Thinset applications are less expensive and typically faster to install than mortar bed applications. However, as the tile is bonded directly to the substrate, any variation or movement in the substrate can affect the tile. Also, there are many types of polymer-modified thinsets on the market allowing the specifier the opportunity to match the thinset properties with the project requirements. In many mortar bed installations, the mortar bed will be allowed to cure and then polymer-modified thinset will be used to bond the tile.

Thinset cement, to which polymers have been added, is commonly called latex-Portland cement mortar. In fact, this term is a bit of a misnomer. The original polymers used to modify thinset were based on latex and the term originates from their use. Today, there are over 10,000 polymers considered by cement chemists when formulating their products. Polymers such as EVA, PVA, SBR, and others are all commonly used in the industry. Many of these polymers are acrylics and not latex chemicals.

The use of these polymers allows specific properties to be imparted to the cement; commonly, freeze/thaw resistance, improved flexibility, and improved adhesion. There are also polymers used to make the cement waterproof or sufficiently elastic so that it acts like an anti-fracture membrane.

The American National Standards Institute (ANSI) set minimum performance levels for latex-Portland cement mortars in the A118.4 and the A118.11 specifications.

There has been quite a bit of debate regarding this subject in the tile industry. Originally, mortar beds were intended as a leveling and load dispersing layer over a subfloor that met the standard L/360 criteria. As such, compressive strengths could be very low (less than 1000 lbs. and high sand to cement mixes were commonplace). Also, this method allowed the floor to packed relatively quickly with short curing times.

Over the years, grout and mortar manufacturers have developed richer mixes with greater compressive strengths. Typically, these mixes also require more liquid to develop their maximum strength. In some applications, these stronger mixes are used over insulation and provide structural rigidity.

However, longer curing times and mortar bed curling (from uneven drying) have also resulted. The debate in the industry stems from differing opinions on whether the greater compressive strength is desirable over slab on grade construction or whether it provides no benefit (and could be detrimental due to longer cure times, increased shrinkage, and the possibility of curling).

There is also debate regarding the merits of greater compressive strength when using mortar beds over various suspended slab systems.

These terms are synonymous and refer to the sheet used between the mortar bed and the substrate. This sheet prevents the mortar bed from bonding to the substrate and allows it to "slip" if there should be movement in the substrate.

While this seems like a simple question, in fact many issues must be considered.

Many people suggest a minimum cure of 28 days under normal conditions, although most thinset manufacturers say you may get away with 14 days using a premium latex modified thinset. There are risks though to tiling too soon. As the slab continues to cure, it will continue to shrink as it hydrates and the excess moisture evaporates. This places the tile and thinset under compressive stress. A premium thinset can better accommodate this stress and compressive force.

Clearly, the longer you can wait before tiling, the less stress that will be applied to the tile by the concrete. Also, should the concrete crack, the crack can be treated (with an anti-fracture membrane) before it is tiled. If curling occurs (and it often does), the concrete can be ground before it is tiled. Tiling before the curling occurs can cause undesirable debonding of the tile and lippage between adjoining tiles.

If you really can't wait 14 to 28 days, you may want to consider a vapor equalization membrane. Reportedly, this membrane isolates the tile layer from the concrete layer but still allows for the vapor to leave the system. This is different from a crack isolation layer that may not allow the "green" concrete to breathe.

Recognizing that tile is a facade, movement joints are needed to eliminate stresses that can occur between the substrate and the tile due to differing amounts of expansion and contraction.

The TCNA Handbook recommends allowing for expansion and contraction in every tile installation. In small rooms, a gap at the perimeter of the room (often hidden by baseboard or shoe molding) is sufficient. For larger areas, the movement joints will be visible.

We cannot specify the exact location nor frequency of movement joints as there are many site related conditions that must be addressed; however, we do offer guidelines in Detail EJ171 in the TCNA Handbook. It is especially important to note for interior installations, movement joints are placed more frequently when moisture or direct sunlight is expected. For exterior installations, the range of temperature from summer highs to winter lows must be considered.

The intent of the guideline regarding sun exposure is to recognize that areas that get warm (or wet) may experience greater amounts of differential expansion. If the areas exposed to sunlight are warmer than surrounding areas, movement joints should be used more frequently. If the tile surface is not appreciably affected, no accommodation is needed in the joint spacing.

Only the area subjected to increased temperature needs to have movement joints more closely placed, not the entire floor if elsewhere the floor is an even temperature.

Many things can subject the tile layer to shear forces in addition to temperature and moisture. The following is a partial list:

Continued curing of the concrete substrate can put the tile in compression
Deflection and vibration of the substrate - particularly with suspended slabs

Movement joints are filled with material that allows for contraction and expansion. For floor applications, urethane, neoprene, or polysulfide are most often used in traffic areas and silicone sometimes where traffic is not a concern. Traffic areas require a sealant with a shore hardness of 35 or greater.

Tile heaved off the floor, or tented, and sometimes cracked is often a sign that movement joints were not used sufficiently. For tile over concrete, the curing of the concrete places the tile under compression.

How long it takes tile to tent is directly a function of at least three variables - the rate of concrete shrinkage, the shear strength of the thinset, and any expansive forces applied to the tile layer (for example, heat). When the tile is poorly bonded, the tile can tent very quickly. If there is a strong bond, often the grout will compress significantly before the tile will lose its bond. Of course, the type of tile is important as well. Thinset has a harder time bonding to porcelain than most other tile. At the other extreme, I have seen a saltillo installation where the tile did not tent but rather spalled as the thinset and grout were stronger than the tile.

When tile fails with a loud report, this certainly indicates that a good bond was present. Only when the shear force exceeds the strength of the bond, will the tile let go. Frequently, either the tile or the concrete will be without thinset residue - as if the thinset was not applied correctly originally. Usually, if the tile is tenting years after the installation, this was not the case. Had the thinset not been applied correctly originally, the tile would have tented long before. Rather, it is important to consider that the cleavage plane will usually occur at the thinset transition - either the bond to the concrete or the bond to the tile, depending on the relative permeability and exact composition of each.

Hence, it is common to see one surface or the other sheared clean of thinset. Even in "explosive" failures where the tile cracks and "jumps" off the floor, usually one surface is free of thinset. Clearly a good bond had been established.

With organic bonding agents and some of the polymer-modified thinsets, continued shear forces degrade the bond over time. So even when tile tents without an explosive report, the original installation may have had sufficient adhesive.

In summary, every installation should allow for movement. Properly designed installations, where expansion and contraction do not create shear forces, should have no problem for the tile to stay well adhered.

While it is impossible to speculate on the exact cause of cracks without an inspection, some reasons for cracking include but are not limited to:

In the mortar bed installation, the mortar bed is not bonded to the concrete - rather it is isolated from cracks in the concrete by the cleavage membrane. This allows the tile to "float" over the concrete.

In the thinset installation, a crack isolation/anti-fracture membrane is bonded to the concrete. Tile is bonded (with thinset) to the surface of the membrane. The internal make-up of this membrane is such that movement in the concrete is not directly transferred to the tile. The membrane compensates where needed to prevent or reduce force transference.

These membranes are either trowel applied or sheet applied. In many cases, multiple components or steps are part of the system. Performance also varies. It is important to check with the crack isolation manufacturer regarding their installation instructions and intended use.

It is a frequent misconception that anti-fracture membranes allow you to eliminate expansion joints - they do not. There always must be soft joints in the tile work to allow for expansion and contraction. The frequency of expansion joints is recommended in the TCNA Handbook.

There are many factors that can cause tile to lose its bond to the subfloor. Losing bond to the subfloor has the potential to lead to cracking in the tile layer:

Expansion and contraction, especially if movement joints were not placed sufficiently in the tile layer. Note: for outdoor, indoor but sunny, or moist installations, this is especially important.
Poor quality thinset, especially where some shear forces (from expansion and contraction or deflection) are present.
Paint or lacquer overspray on the subfloor.
Sealer applied to the subfloor.
Moisture induced deterioration of the subfloor.
Delamination of the subfloor.
Excessive deflection.
Poor thinset coverage, thinset applied in "dabs," thinset used beyond its pot life, or thinset that was disturbed as it was curing.
Moisture sensitive adhesive (affected by hydrostatic moisture or flooding).

Only an on-site inspection can truly evaluate the potential reasons for cracked tile. Please visit the TCA Team for information on these types of services.

Occasionally, a floor will sound hollow even when the tile is well bonded. This can occur when a mortar bed method is used and the mortar has delaminated from the supporting layer or when the subfloor itself is not sufficiently thick or well attached. Other systems that intentionally separate the tile layer from the substrate (such as the mortar bed with a cleavage membrane system like TCNA Handbook detail F111) should be closely examined to ascertain if hollow sounds necessarily imply that the tile is not bonded.

While a tile floor with hollow spots is not ideal, it does not necessarily mean that floor failure is imminent. On the contrary, over concrete, if there is no deflection in the floor; grout and gravity will help keep the floor in place (as long as there are sufficient movement joints in the tile and minimal shear forces). Over wood, floor failure is more likely. Movement in the subfloor could cause grout to break away from the tile, compounding the instability of the flooring.

Some contractors have tried to inject epoxy to rebond tile without reinstalling it. While this may work in a small area, it is not practical over a large area. Further, any repair that does not address the cause of the failure may not last very long.

Traditionally, the accepted minimum requirement for floor rigidity is L/360 - before the tile underlayment is installed. The L/360 standard means that the floor should not deflect more than the "span" divided by 360. If the span of the joists is 10 feet (between supports), then the deflection should not be more than 1/3" between the center and the end. Frequently, there is misunderstanding regarding deflection between joists. For example, while joist manufacturers regularly meet the standard L/360 criteria for code construction with 24" o.c. (on center) systems, these floors often have deflection between the joists exceeding L/360.

Recent research has shown tile to fail, under some conditions, when the floor is more rigid than L/360. In fact, failures at L/600 have been observed. It is for this reason that recommendations for floor rigidity are not based on deflection measurements but on empirically established methods found to work over normal code construction.

Generally there are a few possible causes for white residue on colored grout. When there is a whitish mineral residue on the grout, commonly this is caused by efflorescence. Similar to the white powder left in a drinking glass when a glass of water is left to evaporate, efflorescence is caused by minerals that are soluble in water being dissolved and transported to the surface of the grout as the water evaporates.

Typically, the minerals originate in the cement slab below the tile or in the ground below the slab.

Except in the rarest of cases, efflorescence does not occur from the small amount of minerals in water used to wash a floor. Nor when tile is installed with thinset (tile cement) are there enough soluble salts in the thinset to cause efflorescence.

Occasionally, when tile is installed over a thick mortar bed, the mortar could provide a sufficient amount of soluble salts to cause efflorescence but only if moisture is regularly passing through the mortar bed.

The next question must be where is the water coming from? Similar to the glass of water analogy, it takes a lot of water to dissolve enough minerals to be noticeable. As previously stated, typical cleaning does not provide enough water to cause efflorescence. Even saturating the grout joints with water during periodic cleaning generally does not cause efflorescence.

There are generally effective ways to minimize this problem before tiling and some less effective options after the tile is in place.

Before tiling, if regular vapor migration is detected, remedial steps should be considered before tiling. Alternatively, a vapor equalization membrane can be installed before the tile is installed. There are also companies that sell coatings claimed to reduce moisture migration; however, these may interfere with the ability of the tile cement to bond to the substrate and should only be used if both the manufacturer of the tile cement and the manufacturer of the coating will warranty the installation system.

After tiling, sealing the grout with a penetrating vapor permeable sealer may help retard the rate of evaporation. Note: some caution must be observed in selecting the sealer. Topical sealers (acrylic sealers) which coat the grout joint and are not vapor permeable may turn white from reactions between the acrylic and moisture. This is not efflorescence. Rather this is similar to the whitish haze seen with floor wax when it is applied to a damp surface.

Less common than efflorescence is the white residue that can form on polymer-modified grout if the grout is subjected to excessive moisture before the polymers coalesce. Polymer additives are often added to grout to provide superior properties, commonly improved chemical resistance, reduced porosity, improved flexibility, and freeze/thaw stability. These additives are either already in the grout as redispersible powders or are added in liquid form. In both cases, grout mixed with too much water or cleaned too soon, or cleaned with excess water can cause the polymer to migrate to the surface. In many cases (but not all), these polymers are white in color. When the excess water evaporates, the white polymer is exposed.

In areas with light-colored soils or near light-colored carpeting, occasionally light-colored residue is seen on the grout when detergents used on the tile and grout have not been fully removed. Although the detergent residue is often virtually invisible, it can combine with the soil to form a sticky film that builds up over time. Because the grout joint is commonly lower and more porous than the tile, the dirt tends to be observed in the grout joint. In severe cases, the dirt will also discolor the tile.

Very often, a tile floor can be cleaned with hot water or extremely small amounts of cleaner. Preferably, the dirty water should be vacuumed off the floor. If not possible, sufficient rinse water should be used to remove all traces of the dirt and cleaner.

Solving the problem of residue on grout depends greatly on the cause of the problem - with ongoing moisture migration through concrete being the most difficult.

Generally efflorescence is removed with an acid. However, grout manufacturers do not advocate the use of acids because they attack the cement in the grout. Used judiciously though, this can be a highly effective way of removing efflorescence. There are several products on the market for this with different acids and strengths. In all cases, extreme care should be used to protect the person applying the acid and surrounding fixtures. Generally, it is best to employ an experienced professional for this type of work. Also, acid that is too strong can strip some color out of grout (which is also undesirable) and acid generally cannot be used if acrylic sealers have already been applied to the grout.

After successfully removing the efflorescence, it can only reappear if moisture is entering the system. At this point, sealing the grout joints with a penetrating sealer may be recommended depending on the entire tile assembly. The sealer will minimize water entering from above (for example, from rain) and slow down the rate of evaporation of the water entering from below. However, they should not be used if a water sensitive material (like mastic and some self-leveling products) was used in the tile assembly and there is water migration through the substrate.

For removing a whitish residue from polymer migration, solvents or strong cleaners are needed. As there are many different polymers that could have been used in the grout, any one cleaner may have more or less efficacy. Possibly, an acid could work by removing the surface layer of grout molecules to which the polymers are attached.

For removing soapy dirt stuck to the grout, generally multiple rinsing with clear water and some brushing of the grout joints will be effective. As noted previously, vacuum extraction (for example, with a shop vac) provides the easiest and best results.

OSB is a sheathing product manufactured from wood strands glued together with waterproof, heat-cured adhesives. Sheets are assembled in cross-oriented layers, similar to plywood. Many applications where plywood was used now use OSB. APA Voluntary Standard PS 2-92 specifies performance standards for OSB and plywood.

The most current ANSI standards for tile installation (A108 - 1999) specifically exclude OSB. This is not to say that tile cannot be installed reliably over an OSB subfloor; however, there is significant debate in the tile industry regarding the conditions necessary for a long-lasting successful installation. The allowable ambient moisture level, extent of moisture related swelling before tiling, blocking and bracing and many other issues are still being debated. Each installation materials manufacturer has their own criteria and proprietary methods and requirements. Some backerboard companies will warrant installations using their products over OSB; however, their specific methods must be followed. Note: thinset manufacturers do not recommend installing ceramic tile directly to OSB. Refer to the TCNA Handbook for recommended systems that include OSB, and follow manufacturer's literature for instructions and cautions.

As with all installations, movement joints are necessary to allow for expansion and contraction in the tile layer.

Generally yes, so long as the tile is installed according to industry-accepted methods and the manufacturer rates the tile for exterior applications.

Many manufacturers apply more stringent requirements to their internal tests before they recommend a tile for exterior applications.

No, it is not the same test, although most tiles perform similarly in the EN test and the ASTM test.

We don't know of any tile that claims to be frost proof. While there may be such on the market, we don't know what criteria they are using to separate themselves from frost-resistant tile.

It is very common for frost resistant tiles (including tiles that can cycle more than 2000 times through freeze/thaw conditions) to be damaged by water and ice that sits on the tile when the water can get in the grout joints or in voids left in the setting bed.

Method F101 and F102 are both used for exterior walkways. Typically, it is important to have better than 95% mortar contact between the tile and the substrate and for the thinset or mortar to be rated for exterior applications. The grout joints should be full, compacted, and preferably sealed with a vapor permeable sealer.

Absolutely not. Many non-porcelain quarry tiles and others are rated for exterior applications. Always check with the manufacturer for "area of use" for a specific product.

This is a question that is both easy to answer and difficult. We encourage do-it-yourselfers to install ceramic tile. It is not too strenuous, unless it is a big job or is under awkward conditions. Nevertheless, there are complexities and subtleties associated with installing tile. We recommend starting by reading a good book on the subject. You may want to purchase the TCNA Handbook for Ceramic, Glass, and Stone Tile Installation (see this web site for ordering information). This Handbook contains details for industry accepted practices as well as many items of interest. It is not an installation-training manual; however. If you should decide to hire a professional to install your tile, this Handbook is a very useful tool to assure that the installer is using proper methods.

Make sure that you know your D-I-Y limitations, and do not take on a job that is too complex. Steam rooms, pools, spas, outside decks, and showers are areas among others that require proper installation or a failure can occur. You will be surprised to find detailed instructions on many products such as dryset mortar bags, backerboards, and grouts. Please read the instructions carefully, and follow them as closely as possible. Tiling a dry area is the easiest; and if the substrate is smooth and sturdy enough, it can be simple. You might start out by trying to tile such a location. Select the correct tile for the use and, of course, one you like. Layout the floor carefully to avoid small cuts along walls or other transitions. Placing the tiles on the floor before applying the adhesive (so-called dry layout) will help. You can adjust the centerlines and the grout width to avoid small cuts. These are hard to make and do not look good.

Select the correct adhesive for the tile. Don't try to save money on installation products. Use the ones recommended for the tile, type of substrate, and location. If you are using several boxes of tile, make sure they are of the same caliber (size tolerance) and shade. Usually the boxes will be marked with a code to tell you this. But it is wise to open all the boxes if there are more than one, and randomly stack the tile before installing. This is called shuffling the tiles. The space between the tiles can be close for precision tiles such as grout-edge or rectified porcelain tiles than for irregular or rustic type of tiles. Grout is joint filler. It is not designed to hold the tile in place. Watch to make sure that the tiles are level as you go. Beginners can fail to check for flatness and the job won't look good.

Some points to watch out for:

Use the correct notch-size trowel (indicated on the adhesive bag or container).
Spread enough mortar or mastic to properly bed the tile for flatness. Big tiles require more adhesive than little ones.
Always use control joints at intersecting planes and where the tile meets another materials, such as wood.
Make sure that you get all the air out from under the tiles. This is especially tricky with 10X10 inch and larger tiles. These need to be slid or rotated to get the trowel ridges flattened. If you don't do this, you will have voids under the tile and it can come loose or break.
Wait the proper amount of time before grouting (or walking on the tiles). This time is indicated on the adhesive container also.
When you grout, keep the grout on the dry side, and pack it well into the joints before striking them.
Do not use too much water to clean the grout haze off and do not clean it too soon or you can pull the grout out of the joints.
Remember that grout joints are not waterproof. If you are doing a wet-area (such as a shower), rely on the method and substrate to protect adjoining areas from water.

These are just a few hints and are not to be considered as complete instructions by any means.

Variation in the height of adjoining tiles is called lippage. This is defined in the ANSI (American National Standards Institute) standard A108.02, Section 4.3.7: "Lippage refers to differences in elevation between edges of adjacent tile modules."

The ANSI standard notes that the perception of lippage is influenced by many factors such as:

A) The allowable thickness variation of the tile modules when judged in accordance with manufacturing standards.
B) The allowable warpage of the tile modules.
C) The spacing or separation of each tile module, which would influence a gradual or abrupt change in elevation.
D) Angle of natural or manufactured light accentuating otherwise acceptable variance in modules.
E) Highly reflective surfaces of tile modules accentuating otherwise acceptable variance in modules.

Additionally, variations in the plane of the substrate will also affect lippage. In many cases, when tile is installed by the thinset method over an uneven substrate, the installed surface will not meet lippage standards.

The ANSI A137.1 standard defined allowed warpage according to the type of tile. For a paver tile (commonly used on floors) the allowable warpage is defined as follows:

5.3.1.2.6 Warpage. When measured as described in ASTM C 485, the warpage of each tile in the sample shall not exceed 1.0 percent along any edge nor 0.75 percent on either diagonal.
From this formula the allowable warpage can be determined.

No. This is a common misconception. The amount of allowable tile warpage is not used in the calculation of allowable lippage. Rather, allowable lippage is the total of the inherent (i.e. actual) tile warpage and the allowable lippage from the table. Of course, the actual warpage should not exceed the allowable warpage as calculated above.

This is really rather straightforward if you look at it as follows:

The intent of the standard is to define reasonable lippage from one tile to another. When a tile is warped but within standards, all of the corners cannot be set in the plane of the adjoining tiles. Accordingly,the tolerance for variation from tile to tile includes the actual warpage of the tile. If there is no warpage, the lippage should not exceed the value in the table.

Installation guidelines are included in the TCNA Handbook.

Once you have picked out the materials for your installation, you should consult the manufacturer for their proprietary application procedures and warranties.

While glazed tile is vapor impermeable, the grout is not. In fact, cementitious grout easily transmits water and vapor. It is highly likely in a steam room assembly that vapor can pass through the grout. It may also be necessary to insulate the wall cavity due to differential temperatures on both sides of the wall possibly causing condensation. Each manufacturer may specify a different application of the membrane. Always check with the manufacturer's instructions.

A waterproof membrane meeting the standards of ANSI A118.10 (as well as being rated and recommended by the manufacturer for the steam room application) should be used where ceramic tile is specified (bonded directly to the membrane or otherwise incorporated into the system). Always refer to the manufacturer's instructions and TCNA guidelines.

This is not at all the same as waterproof coatings used on building foundations or as might be used on wood decking. Rather, this is a specialty membrane which bonds to the substrate and to which tile can be bonded.

For thin-set applications (TCNA Handbook detail SR614), you will see that a waterproof membrane can be installed on top of, or behind, the cement board. However, some manufacturers require that the waterproof membrane be used on the front and a vapor membrane on the back. This tends to vary according to the manufacturer.

A surface application of the waterproof membrane has the advantage that the cement board fasteners do not puncture the membrane.

There are no details in the TCNA Handbook that incorporate gypsum wallboard in a steam room assembly. In general, the industry cautions against the use of gypsum wallboard in a steam room environment.

Virtually all tiles work well in steam room applications. However, as with all installations, it is worthwhile to discuss your intended use with a knowledgeable representative of the manufacturer.

B421 and B422 are shower methods in the TCNA Handbook in which a waterproof membrane is applied over an unspecified solid backing. The membrane must cover the solid backing under the entire area to be tiled. We are often asked if drywall is a suitable backing for these shower methods.

While some membrane manufacturers promote use of their membrane in such installations with drywall as the solid backing, other membrane manufacturers do not. At the same time, not all drywall manufacturers recommend their product be used in a shower, even if a membrane will be applied to the surface. Thus, the Handbook leaves the determination of a suitable membrane/backer board combination up to the membrane and backer board manufacturers to declare and up to the design professional to select/specify. Where materials manufacturers do not agree on a particular recommendation, a manufacturer is ofFering what's called a proprietary recommendation.

The International Building Code (IBC) is somewhat ambiguous on the topic. Many in the construction industry interpret it to say drywall is strictly prohibited in wet areas, regardless of whether a waterproof membrane would be applied to the surface of the drywall. Others feel the code only prohibits direct bond of tile to drywall in wet areas. Check with a local code official for your area requirements.

While this seems like a simple question, in fact many issues must be considered.

We understand from the PCA (Portland Cement Association) and the ACI (American Concrete Institute) that many variables affect the length of time a slab needs to cure and the length of time during which curling can be expected. Some of the more common variables are:

Cement - water mix ratio
Cement - sand ratio
Particle size distribution
Presence of accelerators
Curing compounds
Environmental conditions
Location of vapor membrane
Exposure to water during curing

You can see that it is not possible to definitively state how long the concrete should cure before tiling.

Many people suggest a minimum cure of 28 days under normal conditions, although most thinset manufacturers say you may get away with 14 days using a premium latex modified thinset. There are risks though to tiling too soon. As the slab continues to cure, it will continue to shrink as it hydrates and the excess moisture evaporates. This places the tile and thinset under compressive stress. A premium thinset can better accommodate this stress and compressive force.

Clearly, the longer you can wait before tiling, the less stress that will be applied to the tile by the concrete. Also, should the concrete crack, the crack can be treated (with an anti-fracture membrane) before it is tiled. If curling occurs (and it often does), the concrete can be ground before it is tiled. Tiling before the curling occurs can cause undesirable debonding of the tile and lippage between adjoining tiles.

If you really can't wait 14 to 28 days, you may want to consider a vapor equalization membrane. Reportedly, this membrane isolates the tile layer from the concrete layer but still allows for the vapor to leave the system. This is different from a crack isolation layer that may not allow the "green" concrete to breathe.

These terms are synonymous and refer to the sheet used between the mortar bed and the substrate. This sheet prevents the mortar bed from bonding to the substrate and allows it to "slip" if there should be movement in the substrate.

We are often asked about tiling exterior decks. The TCNA Handbook contains two details for decks, F103 (roof decks) and F104 (balcony decks). Additionally, many manufacturers of setting materials also make roof deck drainage systems. Often, these proprietary systems are highly modified versions of the F103 system, allowing precipitation to drain away from the tile.

There are no backerboard systems in the handbook for exterior decks; however, under some circumstances and with appropriate waterproofing, backerboard manufacturers may authorize such an installation. For details, you will need to speak with manufacturers directly as such methods tend to be highly product specific.

Many issues must be considered when tiling exterior environments:

The tiles must be rated by the manufacturer for exterior use. In addition to certain industry tests, such as freeze-thaw, crazing, and water absorption, the manufacturer also considers other properties of the clay body and how the tile was fired. Additionally, some manufacturers have proprietary testing protocols and field testing they use to further assess the suitability of tile for exteriors.

Satisfactory exterior installations must be able to drain water from the system before it gets a chance to absorb into either the tile or the layers below the tile. Since most tiles are water impermeable, the only water that can get into the system goes through the grout joints or through the substrate itself. The installation and the proper slope to drain will determine the extent to which the system saturates itself and in turn its ability to remain intact over time in freeze/thaw or high temperature environments.

The setting materials and substrate must also be appropriate for the exterior conditions present. Freeze/thaw resistant mortars and non-redispersible polymers are often necessary. Expansion joints must also be designed into the system to allow for the movement typical with large temperature variations and exposure to water.

The system must be able to accommodate wind and snow loading and other building movement. Commonly, exterior forces acting on the tile layer can be much greater than a comparable interior job. Accordingly, the installation must be suitably rigorous. Methods developed for interior installations are not necessarily appropriate for exterior installations.

The manufacturers' of tile cement (thinset) have developed a specialty thinset for setting tile that bonds well to resin-based laminates (often called plastic laminates). However, as with all tile installations, the entire substrate below the tile is important - not only the layer to which the tile is bonded.

To safely tile over laminates, the following generally should be true:

The laminate must be clean and free of wax or other bond breakers.
The laminate must be well attached.
The substrate below the laminate must meet the L/360 specification.
For counters, the substrate should be more substantial than just 3/4" plywood with a layer of laminate. Rather it must meet normal tile installation specs before you should consider applying tile directly to the laminate.
Mechanical abrasion of the laminate may be necessary to insure a good bond.

For more information regarding suitable countertop construction (for use under tile), please refer to the TCNA Handbook, details C511, C512, and C513.

Guidelines for tiling over other surfacing materials are covered in the Renovation section of the TCNA Handbook in detail TR711. When tiling over other surfaces, always make sure to check with the mortar and grout manufacturers to find the right material for bonding to the surface in question.

The manufacturer's of tile cement (thinset) have developed specialty formulations for setting tile that are designed to bond well to specific surfaces. However, as with all tile installations, the entire substrate below the tile is important - not only the layer to which the tile is bonded.

There are many kinds of terrazzo floors; some are made with cement while others use epoxy and other resins. Often the floor is coated with acrylic or wax floor finishes after the initial polishing. Many surface coatings will act as bond breakers and must be fully removed. Further, the highly polished surface of the terrazzo can present bonding problems and may need to be mechanically abraded for good adhesion (note: mechanically abrading a floor can create hazardous dusts depending on the surface being abraded - proper protective measures should be taken before undertaking such an operation).

The tile system chosen for tiling over the terrazzo must take into account the building structure and design as generally indicated in the TCNA Handbook (e.g. the methods for tiling a suspended slab are different from those for tiling concrete on grade).

The thinset chosen, if a thinset method is used, must be compatible with the type of terrazzo AND recommended by the thinset manufacturer for that purpose. Typically, epoxy thinsets are used on epoxy terrazzo and cementitious thinsets are used on cement-based terrazzo.

In general, it is best to discuss the exact application with the grout and mortar company you normally use. Inspection of the job site may be necessary as well. An anti-fracture membrane may be needed where there are cracks in the terrazzo or any potential separation between the terrazzo and the metal divider strips.

This is done regularly where there is not a floor height or wall thickness limitation and where the existing tile is well-bonded. To insure a good bond to the existing tile, certain procedures need to be followed; these are described in the TCNA Handbook in detail TR712 and TR713.

Note: Not all thin-sets (nor polymer modified thinsets) are capable of bonding directly to tile. Please consult the grout and mortar manufacturers specified for their recommended thinsets. Also, depending on the tile already installed, in some cases the tile must be mechanically abraded to insure a good bond - this is usually best determined with a "bond test."

Caution: Mechanical or chemical abrasion to tile can release fine particles which could cause harm if inhaled or ingested. Mineral analysis of the tile and glaze should be performed before performing any operation. Appropriate safety equipment should be worn at all times.

Most manufacturers’ of tile cement (thinset) have developed a specialty thinset for setting tile that bonds well to sheet vinyl. However, as with all tile installations, the entire subfloor below the tile is important – not only the layer to which the tile is bonded.

To tile over sheet vinyl, the following is generally recommended by most mortar manufacturers:

The sheet vinyl must be clean and free of wax or other bond breakers.
The sheet vinyl must be single layer only and well attached.
It should not be perimeter glued (it often is!) and it should not have a cushion or foam back.
The subfloor below the sheet vinyl must deflect less than the industry standard L/360 deflection criteria.

In all cases, we do not recommend straying from manufacturer’s recommendations. You must check with the mortar manufacturer for their specific installation and product advice.