Tag Archives: 01 45 00−Quality Control

To Test or Not to Test…? A guide to field quality control

Photo © BigStockPhoto/Vichaya Kiatying-Angsulee

Photo © BigStockPhoto/Vichaya Kiatying-Angsulee

by Sean M. O’Brien, PE, LEED AP, and David Artigas, PE

When properly implemented, field and laboratory testing of buildings and their systems and components can yield a wealth of useful information about construction quality, watertightness, durability, longevity, and other critical performance criteria.

Test results can help designers better evaluate ‘as-built conditions,’ understand any problems with the installation, and develop solutions appropriate to the specific problems that prompted the testing. When improperly implemented, however, testing can yield misleading results, lead designers to incorrect conclusions, and cause unnecessary expenses related to remedial work that may not really be warranted by the in-situ conditions.

In some cases, specifying inappropriate standards or performance criteria can create confusion or incite debate between the design and construction teams, especially in the event of a perceived failure. This article reviews some of the common test methods and procedures used in contemporary construction, with a focus on how the purpose of, and results from, these tests are often misunderstood.

This site-built interior chamber creates differential pressure across a curtain wall system for a test under American Architectural Manufacturers Association (AAMA) 503, Voluntary Specification for Field Testing of Newly Installed Storefronts, Curtain Walls, and Sloped Glazing Systems.

This site-built interior chamber creates differential pressure across a curtain wall system for a test under American Architectural Manufacturers Association (AAMA) 503, Voluntary Specification for Field Testing of Newly Installed Storefronts, Curtain Walls, and Sloped Glazing Systems. Images courtesy Simpson Gumpertz & Heger

The spray rack above has been positioned on the window exterior for AAMA 502, Voluntary Specification for Field Testing of Newly Installed Fenestration Products.

The spray rack above has been positioned on the window exterior for AAMA 502, Voluntary Specification for Field Testing of Newly Installed Fenestration Products.

 

 

 

 

 

 

 

 

 

 

 

Most important (but least asked) question
With dozens of industry organizations publishing thousands of test standards for buildings and building systems, there is almost always a quick answer to the questions: “What do I test and how do I test it?” However, the more important question, and one about which designers and contractors are often less sure, is “Why do I test?”

The answer to this question will almost always dictate the best method to use, the timing of the test, the pass/fail criteria, and sometimes whether it should be performed at all. In the case of poorly specified tests, the wrong tests are often performed ‘because it was in the specifications’ or ‘because the contractor owes us testing.’ Especially with fast-track construction projects, debates over testing are often brushed aside in favor of doing whatever the specification demands, regardless of the value of that testing.

In the authors’ experience, requiring designers to explain the reasoning behind their specified test methods or procedures can be an extremely useful exercise, either during the design process or as part of pre-construction activities. In the case of a designer having correctly specified test methods, the discussion can provide valuable information to the rest of the project team, giving everyone involved a better understanding of the reasons behind the testing. For improperly specified tests, the conversation can help avoid unnecessary testing and the resulting time/expense, as well as identify the correct test methods to determine the desired information.

Flood testing of a membrane waterproofing system on a rooftop parking deck occurs the finished paving is installed.

Flood testing of a membrane waterproofing system on a rooftop parking deck occurs the finished paving is installed.

Dye is used to color water during a roof flood test; it can help link interior leaks to specific areas of the roof above.

Dye is used to color water during a roof flood test; it can help link interior leaks to specific areas of the roof above.

Windows, doors, and curtain walls
Some of the most commonly tested building components are fenestration products—windows, doors, and curtain walls. For new construction, testing is most often specified as a quality control measure to ensure the installed system(s) meet the specified performance requirements for air and water penetration resistance.

The most common requirements are for testing in accordance with various American Architectural Manufacturers Association (AAMA) standards, depending on the system being evaluated. This requires the designer or specifier to know what type of system is specified as well as the relevant performance requirements to establish the appropriate test method.

There are different standards for different components, and some contain multiple test methods or options. For example, AAMA 501.1, Standard Test Method for Water Penetration of Windows, Curtain Walls, and Doors Using Dynamic Pressure, includes a method to test curtain fenestration for water penetration under dynamic wind pressure that requires a large fan—essentially, an airplane engine/propeller—and calibrated spray racks. There is also AAMA 501.2, Quality Assurance and Diagnostic Water Leakage Field Check of Installed Storefronts, Curtain Walls, and Sloped Glazing Systems, which is much easier to perform as it uses a simple handheld nozzle to spray along gaskets and joints. In this case, specifying testing per AAMA 501 is insufficient—the specific test method from that standard needs to be called out, as the two options vary greatly in scope and complexity.

In previous versions of the AAMA 501 standard, a third option (501.3) was available to perform water leakage testing under static pressure differential (Figure 1). Despite this test being pulled from the standard and replaced by AAMA 503-02, Voluntary Specification for Field Testing of Newly Installed Storefronts, Curtain Walls, and Sloped Glazing Systems, references to the 501.3 method can still be found in specifications written today. This often results in confusion when the time comes to perform the tests.

AAMA 501.2 is intended as a field check for water leakage—a simple, economical method to verify the general quality of the curtain wall installation. For compliance with a specified level of air- and water-penetration resistance, AAMA 503 must be used. It is important to note 501.2 is solely intended for fixed glazing systems; the high nozzle pressure used would likely cause moderate to severe leakage on operable vents or window products due to the inherent limitations of seals and gaskets used in operable fenestration.

For testing window assemblies (both fixed and operable) for compliance with a specified air- and water-penetration, AAMA 502, Voluntary Specification for Field Testing of Newly Installed Fenestration Products, is typically specified. This involves building a chamber on the interior of the product to allow for negative interior air pressure and using a spray rack to wet the exterior of the window (Figure 2).

It is important to understand this test method has a specific definition of what constitutes a leak. For example, water on the sill members that does not pass the innermost projection of the window is not considered a leak, since it does not reach a point where it can damage interior finishes. This can come as a surprise to designers witnessing the test and seeing water on the sill, only to find out that by the standard’s strict definition, the window is considered to not have leaked. For this reason, some specifiers add their own language regarding the definition of leakage, but may have difficulty holding a manufacturer to this definition in the event of a dispute.

It is also important for designers to clearly specify the pass/fail criteria for windows. This information can be derived from the performance class (e.g. R, CW, or AW) and grade for the product being tested, but is often misinterpreted, leading to confusion during the test or attempts to hold installers/manufacturers to unrealistic or non-industry-standard performance criteria.

Another caveat of AAMA 502 (since the 2008 revision) as well as AAMA 503 is they only apply to newly installed fenestration products. The standards define ‘new’ as products installed before issuance of the certificate of occupancy for the building or products that have been installed for less than six months.

This recent development is often a source of debate, as this means a window installed for six months and one day, even in an unoccupied/incomplete building, is no longer subject to AAMA 502 and therefore cannot be tested for compliance with the manufacturer’s stated performance criteria. In simpler terms, the manufacturer of the window is only held to its stated performance criteria for six months. For older products, AAMA 511, Voluntary Guideline for Forensic Water Penetration Testing of Fenestration Products, contains diagnostic procedures for identifying known leaks, but is not specifically intended to evaluate in-situ performance of non-leaking windows.

When specifying testing, it is important to make the distinction between test specifications, standard test methods, and testing guides. Each of these types of documents is used for a different, but often similar, purpose. Standard test methods, such as ASTM E1105, Standard Test Method for Field Determination of Water Penetration of Installed Exterior Windows, Skylights, Doors, and Curtain Walls, by Uniform or Cyclic Static Air Pressure Difference, contains specific information on how to physically test these various components, what equipment to use, and related information.

Test specifications, such as AAMA 502, provide procedural information on the testing, the relevance/applicability of the testing, and related administrative information, and typically reference standard test methods (e.g. ASTM E1105) for the physical test procedures.

Finally, testing guides, such as ASTM E2128, Standard Guide for Evaluating Water Leakage of Building Walls, are usually more general in nature and cover a wide range of components and procedures rather than focus on one specific area of the building enclosure. Similar to test specifications, these guides include procedural/administrative requirements and reference standard test methods for the actual testing procedures. Due to their non-specific nature, including ‘compliance’ with a testing guide such as ASTM E2128 as a specification requirement is likely to result in confusion, as it can be interpreted in many ways for many different components. A testing specification and pass/fail criteria must be clearly identified in the contract documents.

This depicts laboratory testing of brick masonry for compressive strength.

This depicts laboratory testing of brick masonry for compressive strength.

Water leakage through brick veneer/cavity wall can be seen from the interior through openings in the backup wall. The veneer does not provide actual waterproofing for this assembly due to presence of drainage plane and weather-resistant barriers.

Water leakage through brick veneer/cavity wall can be seen from the interior through openings in the backup wall. The veneer does not provide actual waterproofing for this assembly due to presence of drainage plane and weather-resistant barriers.

Roofing assemblies
Leakage from roofing systems, especially in the case of low-slope assemblies, can result in significant interior damage when left unchecked. There are many different methods for testing roofs, but not all are compatible with all assembly types. Understanding which methods can be used for which systems is key to specifying the appropriate test, whether as part of a specification for quality control in new installations or as part of remedial/troubleshooting efforts.

In this article, the authors focus on large-scale testing of roof areas, as opposed to smaller-scale testing of specific detail conditions (which is most often done using spray racks/nozzles or localized flood testing). As discussed, the roof’s configuration, as well as the specific membrane type, must be considered when specifying a test method.

The most obvious method of testing a roof—flooding it with water—can be effective in some cases, but extremely damaging in others. Flood-testing is best-suited to inverted roof membrane assemblies (IRMAs) where the membrane is installed directly over the structural deck, with insulation and ballast or other overburden above. In those cases, the testing is performed once the membrane and flashings are complete but prior to the installation of any overburden (Figure 3).

For this test, which is standardized in ASTM D5957, Standard Guide for Flood Testing Horizontal Waterproofing Installations, water is ponded over the system for a period of 24 to 72 hours, during which time the interior is reviewed for leaks. The depth of water must be reviewed to ensure the structural capacity of the roof is not exceeded, as every inch of water adds approximately 0.24 kPa (5 psf) of load. This can be challenging on large or complex roofs, where the deck slope may require compartmentalizing the test into smaller areas. Any leaks resulting from this test are likely to produce only localized damage which gets repaired along with the leaking component(s).

This test procedure is not appropriate for traditional ‘membrane-over-insulation’ roof systems, since leakage through the membrane may wet (and necessitate the replacement of) large areas of roof insulation. Especially in the case of a concrete roof deck, leakage through the membrane could go unnoticed or travel far from the original location as the deck retains the water, allowing large areas of insulation to become damaged and making it difficult to determine the leak’s source. These risks can be reduced by flooding only small areas at a time (limiting the amount of water that could enter the roof), in which case the water can be dyed to provide confirmation of leak sources if multiple areas are flooded in sequence (Figure 4).

There are several test methods available for traditional insulated roofing systems that do not carry the same risk of large-scale damage. These methods typically rely on specialized equipment to detect wet insulation below the membrane. Infrared (IR) thermography uses an infrared camera that visualizes temperature differences on surfaces by measuring and processing emitted radiation.

For an insulated roof, wet insulation will tend to retain more heat and cool slower than dry insulation. Since moisture from roof leaks is often trapped in the system for an extended period, scanning of a roof with suspected or known leaks shortly after the sun has set can help identify areas of wet insulation.

The IR camera measures the surface temperature of the membrane, so this method cannot be used on ballasted roofs since the ballast (e.g. gravel) will cool off uniformly and mask any small temperature differences on the membrane below. Similarly, testing on a windy day may yield misleading results as airflow over the membrane surface may even out temperature differences or cause the wet areas to cool off to the same temperature as the surroundings before the scan is made.

IR scanning of a roof is relatively efficient since large roof areas can be surveyed relatively quickly (some companies even offer aerial surveys, which can be economical for very large, open roof areas). A second method, often referred to as electrical capacitance/impedance (EC) testing, uses handheld or rolling (push-cart) equipment that sends electrical pulses into the roof system and measures the insulation’s ability to retain electrical charge. Wet areas will tend to hold a charge for less time than dry, allowing for relative comparison between areas. Similar to infrared, this method requires an exposed roof membrane since the scanner needs to be in close proximity to the insulation to be effective. For this method to be effective, the roof membrane needs to be non-conductive, making it ineffective on most ethylene propylene diene monomer (EPDM) assemblies or on membranes with metallized reflective coatings. For both of these methods, secondary verification (i.e. roof probes) of suspected wet materials should always be specified to confirm the efficacy of the test for the specific application.

A more recently developed test method uses specialized equipment to pinpoint specific defects in the membrane. In this method, a potential difference is created between the wetted roof surface and the grounded roof deck. Any breaches in the membrane create, in effect, a short circuit in the system which can be detected using specialized equipment.

This method can be used on both traditional and IRMA systems, but—similar to EC testing—the roof membrane must be nonconductive for the method to work. For new construction, especially on traditional roof systems, a grounding screen can be added below the membrane or cover board to provide more positive leak detection and become part of a permanently installed leak detection system. This type of system can be especially beneficial for vegetated roofing assemblies where the often significant amount of overburden can make locating leakage sites extremely difficult.

This infrared image shows air leakage around a window perimeter during a whole-building test.

This infrared image shows air leakage around a window perimeter during a whole-building test.

Brick masonry and exterior walls
Brick masonry has been a common building material in the United States since the colonial period, and mass masonry walls continued to be built through the first half of the 20th century. While the basic process of brick manufacturing has not changed much, modern technology allows the creation of brick typically much stronger and has greater uniformity of properties than historic brick. Historic lime mortars typically are softer and more permeable than modern cement mortars, which allows them to absorb greater stress within the wall from expansion and contraction or enables the wall greater capacity to ‘breathe.’

Concerns with historic masonry fall under two, often related, headings: the masonry’s structural capacity and durability. While it certainly is true modern masonry manufactured and constructed to meet modern standards should result in durable construction, it is not always necessary to hold historic masonry to these same modern standards, as the historic materials often have more than the necessary capacity to provide a long service life with good performance. Also, certain properties being lesser than modern standards may prove beneficial to performance.

The International Building Code (IBC) now has requirements for masonry properties such as compressive strength and performance in shear, though that was not always the case. Current codes are written for current construction, and do not always include previsions for how historic construction ‘works’ structurally. The International Existing Building Code (IEBC) includes provisions that allow historic buildings to remain, or repairs to occur, using original or like materials, but the structural engineer and code officials must still evaluate the structure’s capacity to withstand its loads and remain safe.

Structural engineers can use both non-destructive and destructive methods to determine masonry’s strength and ability to withstand stresses (Figure 5). It is important to specify testing appropriate to both the structure being evaluated and the goal of the evaluation. While a historic mass masonry wall may not meet the letter of the current code requirements, it may have capacity that exceeds its in-service loads with an acceptable factor of safety comparable to the code. That said, one concern with mass unreinforced masonry is it typically does not perform well during seismic events. In areas of higher seismic activity, greater care must be exercised in its evaluation.

Current requirements for energy efficiency mandate the building enclosure to have a specified resistance to heat transfer. While mass masonry walls typically have a lower R-value than modern insulated wall assemblies, they have an advantage—their bulk provides thermal mass unmatched by newer assemblies comprising several thinner layers of different materials sandwiched together. This thermal mass allows the wall to absorb and dissipate heat more slowly than modern walls, slowing the interior of the building’s reaction to changes in the exterior temperature and reducing the need for supplemental heating or cooling.

Changes to the thermal properties of a mass masonry wall, such as adding insulation to the interior, or significantly increasing the interior moisture load, may affect brick performance. Uninsulated historic masonry typically allows moisture to move through the wall (i.e. ‘breathe’) while remaining above the dewpoint, since the interior heat warms the wall.

The addition of interior insulation will reduce the wall’s temperature during the colder months. Water absorbed by the brick’s exterior wythe may go through freeze-thaw cycling as a result of the wall now being colder, and interior moisture that migrates through the wall assembly may condense on the inboard side of the masonry wall, because this side of the wall now is on the ‘cold side’ of the insulation.

Historic masonry may have two advantages that will reduce the likelihood of these two events occurring.

1. Historic brick typically is more porous than modern brick. This greater porosity may allow the brick to ‘drain’ rainwater more quickly than modern brick, and the larger pores may allow the absorbed water more room to expand without causing damage.
2. Historic lime mortars are more absorptive and permeable than modern mortars, and these properties may allow the mortar to wick water rather than having it remain on the wall.

However, it must be stressed the reaction of mass masonry to the installation of interior insulation is still a topic of study among engineers and preservationists. Further, there are currently no established guidelines for insulating walls, only various opinions on the matter.

Many designers of renovation projects may equate strength with durability and specify masonry testing with this thought in mind. Great care must be exercised when considering insulating mass masonry walls, and testing of the masonry’s porosity, absorption, permeability, expansion, and relative strength (both of brick and of mortar) should be performed. Additionally, laboratory testing and evaluation to determine the relative durability of the brick, as well as its resistance to freeze-thaw damage, are a crucial part of this kind of study.

It is also critical to evaluate test results in light of numerous factors, such as the type of building/occupancy, building use, and general quality of the surrounding construction. If a sampling of brick test as SW grade (suitable for severe weathering per ASTM C216, Standard Specification for Facing Brick [Solid Masonry Units Made from Clay or Shale]) that does not necessarily mean the wall assembly in question has the level of durability required for the specific application. SW brick on a large, clear wall area will likely provide suitable performance, but the same brick installed in a shaded location (i.e. minimal drying) below a window that experiences leakage (i.e. excessive wetting) may undergo premature degradation regardless of the brick grading.

The most important question to ask when evaluating a historic masonry building is: “How has it performed thus far?” If the building shows no obvious signs of distress after several decades or even centuries of use, its testing and evaluation must begin from a position of “How does it work?” as opposed to one of “Does it meet the code requirements for modern construction?”

This understanding will include site observations and possibly onsite or laboratory testing, and research into historic construction methods and materials. Ultimately, this approach to evaluating historic masonry may lead to a more efficient and lower cost project that also can maintain the building’s character. Regardless of testing, designers who take this approach much understand when the use of the building or other characteristics of the enclosure are changed as part of renovations, the prior performance of the building may not be a suitable predictor of long-term durability.

From a water penetration standpoint, there are many different test methods available for masonry walls, but not all provide useful information. For example, ASTM C1601, Standard Test Method for Field Determination of Water Penetration of Masonry Wall Surfaces, determines water penetration at the surface of a masonry wall, but does not provide any information on how much water actually leaks to the interior (as opposed to water that is absorbed and stored by the masonry). Similarly, RILEM tubes can be used to provide relatively quick evaluations of the water absorption rate of a masonry wall.1

ASTM E514, Standard Test Method for Water Penetration and Leakage Through Masonry, provides for measurement of the actual amount of water penetrating the full thickness of the masonry, but this is a lab test not applicable to field conditions (although it is often specified—incorrectly—by designers evaluating existing masonry buildings). Field surface absorption tests may have limited use in qualitatively evaluating the change in absorption that results from installing a penetrating sealer, but are typically of little to no use in evaluating water leakage.

Neither of these tests will be of practical value for masonry cavity wall construction, where any water penetrating the exterior façade is collected in a drainage cavity and wept out of the system (Figure 6). Water leakage through a masonry cavity wall is more likely the result of a breach in the water-resistive barrier (WRB) behind the masonry, since masonry veneer systems are expected to allow water into the drainage cavity.

The authors have generally found the general guidelines from ASTM E2128, Standard Guide for Evaluating Water Leakage of Building Walls—as opposed to one specific standard test method—are helpful in establishing the right combination of testing and inspection to diagnose water leakage through masonry walls.

Air barrier systems
As far as building testing goes, the testing of air barrier systems is a relatively recent development.2 Just as with window and curtain wall testing, there are multiple test standards and guides for testing air barrier systems in both the lab and the field. One of the first points of confusion is the definition of an air barrier—a system of interconnected components including walls, windows, curtain walls, and roofs that act together to prevent uncontrolled airflow into and out of the building. While air barrier testing is often thought of as testing a wall air barrier membrane (one component of the system), it can encompass everything from single materials to the entire building enclosure.

Some of the most commonly tested building components are fenestration products like windows and curtain walls. For new construction, testing is most often specifi ed as a quality control measure to ensure the installed systems meet the specifi ed requirements for air and water penetration resistance. Photo © BigStockPhoto/Graça Victoria

Some of the most commonly tested building components are fenestration products like windows and curtain walls. For new construction, testing is most often specified as a quality control measure to ensure the installed systems meet the specified requirements for air and water penetration resistance. Photo © BigStockPhoto/Graça Victoria

Testing of actual materials, such as sheet- and fluid-applied membranes, is performed in the laboratory due to the very small quantity of air leakage being measured and the high degree of accuracy required in the measurement. Air barrier products are required by most codes to allow no more than 0.02 L/s.m2 @ 75 Pa (0.004 cfm/sf at a pressure differential of 0.3 in. of water). In reality, most sheet membranes (such as self-adhered rubberized asphalt products) exceed this criteria by an order of magnitude or greater—much too low to be reliably measured in the field.

Air barrier assemblies—essentially, air barrier materials in an as-built condition that includes laps, seams, and penetrations—can be tested in either the lab or the field. Laboratory testing per ASTM E2357, Standard Test Method for Determining Air Leakage of Air Barrier Assemblies, provides an air leakage rate for a pre-defined arrangement of air barrier products, penetrations, and a window opening (but not the window itself—an oft-overlooked element of the air barrier system).

While primarily a laboratory test used by air barrier manufacturers to demonstrate their products’ performance and code compliance, the method can also be applied to field-installed mockups. However, applying this test in the field is not as simple as installing a chamber on the interior and testing the exterior. Air leakage through the perimeter of a sample area (e.g. through a concrete block or stud wall perpendicular to the interior-exterior direction) is often impossible to isolate, and due to the relatively low leakage rates being measured, even a small amount of extraneous leakage can create a false negative test result. Using this general chamber testing approach on a qualitative basis is simpler and often more effective, since telling a contractor that the test result was 0.25 L/s.m2 (0.05 cfm/sf)—in other words, a failure—does not provide the same level of usefulness as telling them there were leaks at one membrane seam and two brick ties that need to be repaired.

Specifications for field-installed air barrier assemblies often include testing of the window as part of the assembly. While this makes sense from a practical standpoint (i.e. the connection to the actual window system is a critical transition in the air barrier), there are limitations to this test. From a practical standpoint, mockup testing of air barrier assemblies typically happens at the beginning of a project, often long before the windows are delivered to the site (or in some cases, before specific window products are even selected). Testing of the assembly with a ‘dummy window’ in place is possible, but results can be misleading since the actual connection is unlikely to be the same as what is intended for the project windows.

In cases where the dummy window is put in temporarily with sealants and sprayed-applied foam insulation, the actual leakage rate may be much lower than what will occur when the project windows are installed, giving a false positive result for the test. In the case of the project windows being available at the time of testing, the specification of pass/fail criteria for the air barrier assembly test becomes more important. The performance criteria for air barrier assemblies are based on a window perimeter being included, but not the window itself. Since most windows will experience significantly higher leakage (on an area basis) than air barrier materials and assemblies, leakage through the window may far exceed the air barrier assembly leakage criteria, even if the assembly itself, minus the window, would pass on its own. For this reason, it is important to clearly specify how and if the window is to be included in the test, and, if so, some increase in allowable air leakage needs to be included to avoid requiring a result which is not attainable in the field.

The third type of air barrier testing is whole-building testing, using blower door or similar equipment and ASTM E779, Standard Test Method for Determining Air Leakage Rate by Fan Pressurization, procedures to measure overall air leakage through the entire building—including all walls, roofs, and windows. Different codes and standards require different overall leakage rates, but 0.02 L/s.m2 @ 75 Pa (0.4 cfm/sf at 0.3 in. of water) is typical for most building and energy codes.

While this test provides a single number to describe air leakage that can be easily compared to other buildings, it has many limitations that must be considered before requiring a certain level of whole-building performance. The first major issue is at the time when the air barrier is substantially complete to the point where testing can be performed, it is also likely to be concealed by cladding and other materials that can make the detection (using IR thermography or tracer smoke) and repair difficult or impossible.

Second, the testing itself can be difficult to perform, especially on large/more complex buildings, due to the need for multiple fan systems that must all be linked together for measurement or adjustment. In tall buildings, internal fans may be needed to equalize pressure differentials over the height of the building. While most testing firms can easily come to a site and perform standard window or curtain wall tests, large-scale testing of whole buildings requires specialized equipment and a fair degree of experience and expertise to successfully execute.

Lastly, there is some debate over what is achievable in terms of air leakage through whole buildings. A designer can certainly specify the overall leakage rate needs to be 0.5 L/s.m2 (0.1 cfm/sf), but achieving that level of airtightness requires an exceptionally well-designed air barrier, as well as carefully planned execution of the construction.

This is a common mistake with all types of air barrier testing—specifying a high level of airtightness without providing the corresponding design detailing is a futile effort. As mentioned, once the building is physically ready to be tested, it is often far too late to practically implement repairs, which brings up the difficult question of “What do we do now that we failed the test?” While the industry is still working on answering that question, specifiers can help avoid problems by specifying reasonable levels of airtightness appropriate for the building design.

As with the other previously described tests, visualization techniques such as infrared thermography and tracer smoke can be used to take advantage of the pressure differential created during a whole building test and qualitatively identify air leakage sites (Figure 7).

Conclusion
While the wide variety of available testing standards means there is almost always a standard for the designer’s specific need, finding the right standard can be difficult when one does not have a firm understanding of the actual goals. Designers and specifiers should first evaluate the question of ‘why’ when it comes to testing, as the answer will typically guide them to the correct test method to follow.

Specifying both the appropriate testing method and the appropriate pass/fail criteria are necessary to provide meaningful test results and avoid the time and expense of unnecessary testing or inappropriate testing which leads to ambiguous results. A little more time spent researching test methods during the design phase and specifying appropriate methods and performance criteria can go a long way toward reducing confusion and disputes during the construction process in the field.

Notes
1 For more on RILEM tubes, see The Construction Specifier articles, “Testing the Test: Water Absorption with RILEM Tubes,” by Adrian Gerard Saldanha and Doris E. Eichburg, and “Durable Waterproofing for Concrete Masonry Walls: Redundancy Required,” by Robert M. Chamra and Beth Anne Feero in the August 2013 and July 2014 issues. (back to top)
2 For more on air barriers, see the article “Wind Load and Air Barrier Performance Levels,” by Maria Spinu, Ben Meyer, and Andrew Miles, in the July 2014 issue. (back to top)

Sean M. O’Brien, PE, LEED AP, is an associate principal at the national engineering firm Simpson Gumpertz & Heger (SGH), specializing in building science and building enclosure design and analysis. He is involved in both investigation/forensic and new design projects. O’Brien is a member of the American Society of Heating, Refrigerating, and Air-conditioning Engineers (ASHRAE), co-chair of the New York City Building Enclosure Council (BEC-NY), and a frequent speaker and author on topics ranging from building enclosure design to energy efficiency. He can be reached at smobrien@sgh.com.

David Artigas, PE, is senior staff I–building technology at SGH, specializing in building enclosure design and investigation, building science, and historic preservation. He can be reached at djartigas@sgh.com.

Specifying Movement Joints and Sealants for Tile and Stone: Reviewing current industry standards and design options

Photo courtesy Florida Tile

Photo courtesy Florida Tile

by Donato Pompo, CTC, CSI, CDT, MBA

In one way or another, all tile and stone assemblies move. Whether due to thermal or moisture movement, shrinkage, freezing, or dynamic structural movements, tile and stone installations are subjected to them all. To ensure a long-lasting installation, architects must specify the requirements for movement joint design and placement, along with the correct type of sealant for filling those joints.

A ‘movement joint’ is a general term used for all types of joints seen in construction materials that control and allow movement. Most commonly, they are known as ‘expansion’ or ‘control’ joints, but there are various categories. Generally, they contain an appropriate pliable sealant for the intended application, which is often referred to as a ‘soft’ joint.

Movement joints allow for the material in which they are placed to move without restraint; they control where the movement manifests to avoid random cracking in finish materials. An example would be the joints or separations in a concrete sidewalk. If there were no movement joints in the concrete sidewalk, then it would crack at a random point as it is subjected to shrinkage during curing, or to expansion when it is exposed to moisture (and then contraction again as it dries). Rising temperatures cause expansion, lowering temperatures cause contraction, and wet freezing conditions cause both, as the temperature drops and the moisture freezes.

There are other types of structural movement from the ground or its foundation that can cause various kinds of movement in the form of deflection. These stresses, and the resulting deformations, are compounded by adjacent materials that have a different coefficient of movement properties—the differentials can lead to serious problems, particularly over time as the respective materials go through various degrees and combinations of cycles from wet to dry or hot to cold, and so forth. Movement joints are also designed to isolate different materials from each other so they do not affect adjacent materials.

More often than not, when there is a tile (e.g. ceramic, porcelain, stone, or glass) failure a contributing factor is the lack of properly installed movement joints. In some cases the failure could have been avoided, or damage limited, if there had been properly installed movement joints. Just like concrete sidewalks, slabs, and bridges, tile and stone need to have movement joints to control the anticipated movements within a structure and the various climatic conditions it will be subjected to throughout the years.

Small horizontal movements can result in exponentially larger vertical movements. When one end of a ruler is restrained and the other end moved toward the center 3.2 mm (1/8 in.), there is a 51-mm (2-in.) rise at its apex. Photos courtesy Ceramic Tile and Stone Consultants

Small horizontal movements can result in exponentially larger vertical movements. When one end of a ruler is restrained and the other end moved toward the center 3.2 mm (1/8 in.), there is a 51-mm (2-in.) rise at its apex. Photos courtesy Ceramic Tile and Stone Consultants

Figure1b

Troubles with tile and stone
This author has seen tile floors that did not have adequate movement joints—where a portion of the floor was tented (i.e. debonded and raised) several inches off its substrate during the heat of the day, but was lying flat at night when it cooled down. For a good example of how small horizontal movements can result in exponentially larger vertical movements, one can take a 1219-mm (48-in.) metal ruler and lay it on a horizontal surface. When one end of the ruler is restrained and the other end moved toward the center 3.2 mm (1/8 in.), there is a 51-mm (2-in.) rise at its apex. In effect, this is what happens to tile floors when they tent. They are constrained at their perimeters with no movement relief, the tile is typically insufficiently bonded, and it expands for one reason or another.

Well-bonded tile floors tend to crack to relieve the stress rather than lose their hold. Properly placed movement joints allow the tile to move and control where the movement manifests (i.e. within the joint where the tile is not restrained).

Tile and stone installers may have practiced their trade and honed their skills, but they are not engineers. In other words, while installers have some responsibility in ensuring movement joints are included in the tilework, it is ultimately up to the architect to specify the appropriate design, materials, and locations.

The Tile Council of North America (TCNA) provides general movement joint guidelines for tile and stone applications in its TCNA Handbook for Ceramic, Glass, and Stone Tile Installation, listed under Detail EJ171, “Movement Joint Guidelines for Ceramic, Glass, and Stone.” TCNA states:

because of the limitless conditions and structural systems on which tile can be installed, the architect or designer shall show the specific locations and details of movement joints on project drawings.

There are industry standards that help design the appropriate movement joint layout and design for the intended application:

  • ASTM C1193, Standard Guide for the Use of Joint Sealants, which provides guidelines on how to use and install sealant joints; and
  • ASTM C1472, Standard Guide for Calculating Movement and Other Effects When Establishing Sealant Joint Width, for determining appropriate movement joint width relative to the intended application and conditions.
These two photos of the same spot show what happens when transition joints are fi lled with a hard grout rather than the soft movement joint sealant. The cementitious grout and stone crack due to expected movements within the stone and the structure. It is a good example of why we need movement joints in tilework.

These two photos of the same spot show what happens when transition joints are filled with a hard grout rather than the soft movement joint sealant. The cementitious grout and stone crack due to expected movements within the stone and the structure. It is a good example of why we need movement joints in tilework.

2CrackedGroutDueToMissingTransitionMovemtJoint

Ensuring adequate design
The appropriate design of a movement joint is based on the tile assembly’s configuration and substrate type. The substrate must be structurally sound, meet relevant code requirements, and not exceed maximum deflection limitations (ranging from L/360 to L/720, depending on the material and the application). The general rule is these movement joints should be placed at the perimeters of tile and stone installations, at all transitions of planes or different materials, and within the field of tile.

Tiles at perimeters of rooms should have movement joints. Inside and outside vertical joints on framed walls should have movement joints and not be hard-grouted (as they so commonly are, alas). Bathtub or shower receptor to wall transitions should have a movement joint. In wet areas, movement joints are important not only to control movement, but also to act as a water-stop at those transitions, providing another layer of protection.

Whether due to thermal or moisture movement, shrinkage, freezing, or dynamic structural movements, all exterior stone assemblies—like this limestone fountain surrounded with granite paving—are going to move.

Whether due to thermal or moisture movement,  shrinkage, freezing, or dynamic structural movements, all exterior stone assemblies—like this limestone fountain surrounded with granite paving—are going to move.

TCNA states movement joints for interior applications should be placed at least every 6.1 to 7.6 m (20 to 25 ft) in each direction unless the tilework is exposed to direct sunlight or moisture, which would then require the movement joints placed at least every 2.4 to 3.7 m (8 to 12 ft) in each direction. For exterior applications, movement joints should be placed at least every 2.4 to 3.7 m in each direction.

TCNA recommends the movement joint width be a minimum of 9.5 mm (3/8 in.) wide for exterior applications when the field movement joints are 2.4 m on center (oc), but recommends a minimum of 12.7 mm (1/2 in.) wide joints for exterior applications when the field movement joints are 3.7 m oc. TCNA states:

minimum widths of movement joints must be increased 1.6 mm (1/16 in.) for each 9.44 C (15 F) tile surface temperature change greater than 37.8 C (100 F) between summer high and winter low.

The aforementioned ASTM C1472 is valuable because it provides the coefficient of linear thermal movement for different materials, along with the temperature range of various geographic areas and mathematical formulas to determine the required joint widths for the respective conditions.

To ensure a longlasting limestone installation, architects must specify the requirements for movement joint design and placement, along with the correct type of sealant for fi lling those joints.

To ensure a longlasting limestone installation, architects must specify the requirements for movement joint design and placement, along with the correct type of sealant for filling those joints.

To what degree the respective substrate needs movement joints depends on the standards the substrate must meet. For instance, a plaster substrate per ASTM C1063, Standard Specification for Installation of Lathing and Furring to Receive Interior and Exterior Portland Cement-based Plaster, requires control joints installed in walls to delineate areas not more than 13.4 m2 (144 sf). The distance between control joints shall not exceed 5.5 m (18 ft) in either direction or a length-to-width ratio of 2.5 to 1.

Concrete also has standards that vary depending on its structure, thickness, and design mix.

Detail EJ171 states all underlying movement joints in the substrate need to continue through the tile assembly. Typically, this means that in addition to honoring the substrate movement joints, the tile assembly needs additional movement joints within its assembly.

If there is a mortar bed over the substrate, then the movement joint has to be continuous through it to the tile surface, which is considered an expansion joint. If the tile is being bonded to the substrate, then the movement joints not continuing up from a substrate movement joint are generic movement joints. These are often the same width as the grout joints if it was designed to work at that width. The movement joint widths within the tilework should never be narrower than the substrate joint on which it is placed.

Crack-isolation membranes
Some manufacturers of products meeting American National Standards Institute (ANSI) A118.12, Specification for Crack-isolation Membranes for Thin-set Ceramic Tile and Dimension Stone Installations, allow their membrane to cover non-structural movement joints (i.e. those that move horizontally, but not vertically) such as saw-cut or cold control joints. However, TCNA does not recommend this.

Structural expansion joints can never be covered with membranes as the vertical displacement cannot be mitigated with a crack-isolation membrane. (The membrane manufacturers require movement joints be installed within the tile assembly, and some allow those joints to not line up exactly over the substrate joints.)

The appropriate design of a movement joint is based on the tile or stone assembly’s confi guration and substrate type.

The appropriate design of a movement joint is based on the tile or stone assembly’s configuration and substrate type.

Each manufacturer of crack-isolation membranes may have different recommendations and limitations, so it is always important to follow the accompanying instructions. Some membranes are made of a bitumen material that is incompatible with certain types of sealants used to fill the movement joints.

TCNA’s Detail F125, “Partial- and Full-crack Isolation Membrane,” provide guidelines for isolating non-structural cracks with an ANSI A118.12 product. It is important to note this detail for both ceramic tile and stone applications recommends a movement joint be placed at one or both ends of the tile bridging the underlying crack, as recommended by the membrane manufacturer.

Types of movement joints
The different types of movement joints are shown in the TCNA Handbook in the EJ171 section. Expansion joints are normally considered structural joints that can possibly move vertically. They are found in concrete substrates to isolate one portion of the slab from the other, and in mortar beds as either an extension of the concrete expansion joint or just to isolate one portion of the mortar bed from the other.

A cold joint is the dividing point where two adjacent concrete pours were placed at different times. These weak points are more likely to develop cracks and have to be treated as a movement joint.

Construction or contraction joints are saw-cut concrete control joints that must also be treated as movement joints. The concrete is saw-cut at this predetermined spot, making it a weak point where the concrete will crack (rather than having it crack at a random location). There are also various types of perimeter movement joints, which are found at restraining walls or transition points from one plane to another that are all more likely to be subjected to some type of movement. Tile must be allowed to move to avoid damages.

Sealant considerations
Not only is the design of the movement joints important to a tile installation’s success, but so is the type of sealant or caulking used to fill those joints. TCNA EJ171 states a product meeting ASTM C920, Standard Specification for Elastomeric Joint Sealants, must be used to fill movement joints of all types. Such sealants include high-quality silicone, urethanes, and polysulfide materials. These types of sealants are normally rated as highly weather-resistant with high elongation properties, and high adhesion characteristics that come with 20-year commercial warranties. Too often, one finds installers using some type of acrylic, latex, or siliconized sealant, because they are easier to work with, but these sealants have low performance values and basically no warranty.

Various sealants have different physical properties and performance capabilities. TCNA and the referenced ASTM standards provide guidelines and nomenclature for designating the appropriate type, grade, class, and use for the intended application. For instance, some sealants are not suitable for foot or vehicle traffic, so one must specify “Use T” for those applications.

A traffic sealant should have a Shore A hardness of 35 or greater, which is critical because otherwise the surface could be dangerous to those who wear high heels. (Found on data sheets, Shore A is a physical property of all sealants; it indicates how hard it is in terms of resistance to penetrations or point loads.) High heels will penetrate a softer sealant and can cause a tripping hazard.

There are sealants with fire or acoustical ratings that are required for certain applications; some cannot be used in a submerged application, while others cannot be subjected to certain chemicals. Not all ASTM C920 sealants are compatible with natural stone and could cause the stone to stain. Some sealants require the surfaces to be primed after cleaning the joints and before installation. These are all important concerns to be addressed in the specification to ensure the correct material is used for the intended application.

Various sealants have different physical properties and performance capabilities. Traffi c sealants need a Shore A hardness of at least 35 to ensure those with high heels do not dig into the material, leading to trips and falls.

Various sealants have different physical properties
and performance capabilities. Traffic sealants need a Shore A hardness of at least 35 to ensure those with high heels do not dig into the material, leading
to trips and falls.

highHeelMarkOnSealant3highHeelMarksOnSealant01

It is important movement joints be properly constructed per industry standards. There are also numerous requirements sealant manufacturers specify in order for their products to perform as advertised. Sealants require only ‘two-point contact,’ meaning they are only to adhere to the two opposite sides of the movement joint for optimal performance. They are not to bond to the bottom of the joint, otherwise the sealant will not achieve the published elongation characteristics. A bond-breaking polyethylene tape or foam must be inserted into the joint prior to installing the sealant so the sealant will not bond to it.

To ensure the sealant achieves its published elongation characteristics, care must be taken to ensure it is applied neither too thin nor thick into the joint; the foam backer, installed at the prescribed depth, helps with this. The sealant must be at least 6.4 mm (1/4 in.) thick, and the width to depth ratio should be 2:1 for optimal performance. Generally, sealant companies want a minimum 6.4-mm wide joint, but 3-mm (1/8-in.) is acceptable for non-moving joints (e.g. adhered tile applications).

Additionally, movement joints must be completely filled with the appropriate backing below the sealant, so there are no voids to collect moisture. It is generally best to use closed cell foam, but some sealants require open cell to manage the sealants’ off-gassing while curing. For thin tiles—such as the 6.4-mm thick mosaics—or some of the newer large 3-mm thin porcelain tile panels, it is more problematic to try to install a bond-breaking tape in the movement joint. It is better to leave it out, since it is an adhered non-moving joint that will not require higher performance.

There are prefabricated movement joints made of metal sides and legs with plastic inserts adhered under the tiles on either side of the movement joint. There are also metal L-shapes that can be installed under the tiles on either side of the movement joint and then filled with the appropriate sealant. On one hand, these provide protection to the tile edges and the plastic inserts are conveniently replaceable; on the other hand, they restrain the tile movement since the metal angles are bonded to the substrate. This may not be a big problem if the tile is bonded well and they are installed frequent enough, but this author has seen cases where the tile was not sufficiently bonded, the movement joints were properly spaced, and the tile tented. Since tile assemblies move one way or the other, movement joints should not restrain movement.

Keystone Fashion Mall (Indianapolis, Indiana) has a striking tiled fl oor that benefi ts from having adequate consideration placed into the location and type of movement and expansion joint. Photo © Adam Novak Photography. Photo courtesy Crossville Inc.

Keystone Fashion Mall (Indianapolis, Indiana) has a  striking tiled floor that benefits from having adequate consideration placed into the location and type of movement and expansion joint. Photo © Adam Novak Photography. Photo courtesy Crossville Inc.

Circumventing aesthetic problems
Too often, movement joints are left out of installations with the common excuse being the owner did not want those ‘ugly’ joints marring their tiles. (Of course, their absence can cause even uglier failures.) When specifiers take the time to design the movement joints into the installation, they can accentuate features to make joints virtually unnoticeable.

Manufacturers of one-part silicone sealants have a broad range of colors available and on large jobs they will make custom colors to match the grout. Two-part urethane sealants can be mixed on the job by experienced sealant installers and can easily match the color of the tile grout. By placing the movement joints more frequently, they can be made narrower, matching the width of the grout.

For tile patterns with staggered joints, the designer can use the staggered grout joint (referred to as a saw-tooth joints or zipper joints) as a generic movement joint to make it less noticeable. When done well, movements are not noticeable and can enhance the installation features.

Specifying strategies
Architects should write the sealant specification for the tile and stone applications in Division 07 under “Sealants.” Still, detailed information should be provided in the Division 04 and 09 sections (for stone and tile), particularly if the tile installer is expected to install the sealant.

The following key points, as related to movement joints in tile or stone assemblies, should be included in the specification:

Part 1−General Requirements

  1. Refer to Division 07 for Movement Joint Sealants.
  2. Call out the key industry standards, which are: ANSI A108.01, Requirements for Movement Joints; TCNA Handbook for Ceramic, Glass, and Stone Tile Installation; Marble Institute of America (MIA) Dimension Stone Design Manual for Expansion Joints; ASTM C1242, Standard Guide for Selection, Design, and Installation of Dimension Stone Attachment Systems; ASTM C1193, Standard Guide for the Use of Joint Sealants; and ASTM C1472, Standard Guide for Calculating Movement and Other Effects When Establishing Sealant Joint Width.
  3. Prepare a specific quality assurance (QA) section to verify performance of the ASTM C920 sealant material and to verify it will be suitable for the intended application. Test sealant for performance per ASTM C719, Standard Test Method for Adhesion and Cohesion of Elastomeric Joint Sealants Under Cyclic Movement. Test the peel adhesion of the sealant per ASTM C794, Standard Test Method for Adhesion-in-peel of Elastomeric Joint Sealants. For stone applications, test for staining per ASTM C1248, Standard Test Method for Staining of Porous Substrate by Joint Sealants.
  4. Require a letter from sealant manufacturer stating its product is suitable for the intended use, and outlining its warranty.
  5. For larger projects, specify a sealant installation company that specializes in installing sealants on a full-time basis.
  6. Require a mockup for approval of sealant color and application.

Part 2−Products

  1. Be sure to write performance specifications. Reference specifications only call out products that meet the minimum requirements—in other words, the least-expensive products with the lowest acceptable level of performance.
  2. Call out ASTM C920 sealants. Identify the application and the respective type, grade, class, and use to the intended application. Require a primer if sealant manufacturer requires it with the sealant. Call out the appropriate polyethylene backer foam. Specify the sealant color is to be approved by the architect or owner from the mockup.

Part 3−Execution

  1. Specify the specific movement joint details, for the respective application, from Detail EJ171 in the TCNA Handbook for Ceramic, Glass and Stone Tiles.
  2. Request installers properly clean and prime movement joints as required by the sealant manufacturer. The movement joints need to be completely open and free from any obstructions.
  3. Specify movement joint layout plans, and types of movement joints and sealants, as referenced in TCNA’s Detail EJ171. Tile installers should submit Requests for Interpretation (RFIs) if they are unclear with the requirements.
  4. Specify sealant product installation as per manufacturer’s instructions and industry standards. Similarly, products must be mixed following the manufacturers’ requirements. Further, temperature limitations must never be exceeded. Shading or heat should be required, and the work protected from weather and other trades.
  5. Specify the required sealant surface profile, such as “flush,” “concave,” “recessed,” or “fillet.” Vertical surfaces can be specified to be oriented vertically, horizontally, or at any angle in between so it can control water shedding.
  6. Provide a detailed quality control (QC) plan to be implemented by a third party.
Not only is the design of the movement joints important to a tile or stone fl oor installation’s success, but so is the type of sealant or caulking used to fi ll those joints. Photo courtesy Daltile

Not only is the design of the movement joints important to a tile or stone floor installation’s success, but so is the type of sealant or caulking used to fill those joints. Photo courtesy Daltile

Conclusion
To ensure a long lasting installation, it is critical architects specify and provide the requirements for movement joint design and placement, along with the correct type of sealant or caulking for filling those joints.

In more than three decades, this author has never investigated a tile or stone failure to find all the industry standards and manufacturers’ instructions were followed. Further, the failure is never due to one deficiency, but rather many compounding ones.

The industry standards represent years of experience and scientific testing from a consensus group of industry professionals who volunteer their time and efforts to help architects, installers, and owners have successful tile and stone installations. The key to a successful tile and stone installation is to follow industry standards and to write good specifications. CSI’s MasterFormat and SectionFormat provide the structure for this. When the resulting construction documentation is used correctly and thoroughly, it limits both the designer’s and client’s risk and liability in ceramic tile, glass tile, and stone applications.

Donato Pompo, CTC, CSI, CDT, MBA, is the founder of Ceramic Tile and Stone Consultants (CTaSC), and of the University of Ceramic Tile and Stone (UofCTS). He has more than 35 years of experience in the ceramic tile and stone industry from installation to distribution to manufacturing of installation products. Pompo provides services in forensic investigations, quality control (QC) services for products and installation methods, training programs, testing, and onsite quality control inspection services. He received the 2012 Construction Specifier Magazine Article of the Year Award. Pompo can be reached at donato@ctasc.com.