How the Industry is Addressing Thermal Stress Breakage in the Spandrel Cavity

By Jordan Scott

Some say it’s an accident that must not happen. Others say it’s an issue that can be prevented during fabrication. Research conducted by several industry experts suggests that ceramic enamel weakens glass, according to Chris Barry, an industry consultant and former director of technical services for building products with Pilkington North America who has spent years researching the issue himself. However, there is disagreement about the extent of that weakening. When ceramic enamel is used in spandrel applications, which can be subjected to high levels of thermal stress, it could lead to glass breakage. Some are concerned that, in extreme cases, broken spandrel glass could fall off a building and hurt passersby below, making it imperative, from their perspective, that the issue be addressed in U.S. standards.

Thermal Stress

The growing specification of low-E coatings on spandrel glass is an effort to create harmonization with the vision area. This has led to an increase in solar gain within the spandrel cavity, says Steven Marino, manager of technical services for Vitro Architectural Glass of Pittsburgh. While the low-E coating works to keep infrared light out, it also prevents heat from dissipating. Heat loss does occur in the framing system, so the edge of the system tends to be cooler than the center of glass.

“This creates tensile stress on the perimeter of the glass. If the stress is high enough then the glass will crack,” says George Torok, principal and façade specialist for Toronto-based engineering firm Morrison Hershfield.

Many spandrel applications include a dark-colored opacifier, which absorbs more heat than light-colored opacifiers. Marino says a dark opacifier on surface four with a low-E coating on surface two can lead the inboard lite to heat up as the opacifier absorbs heat and the low-E coating reflects that heat back into the cavity.

He also says other factors can contribute to increased thermal stress such as a reflective foil backing material or a small gap between the inboard lite and the insulation or backpan.

“If the gap is not vented, then any heat that builds up in there will magnify how fast the inboard lite of the insulating glass unit (IGU) heats up,” says Marino.

He adds that, in some cases, a piece of the insulation could come loose, fall and touch the glass, creating a localized hotspot. If the spandrel cavity isn’t vented, he says this creates a high potential for heat build-up. Marino also points out that the biggest reason for glass breakage typically is associated with edge damage that can be caused at any point in the supply chain.

Glass Weakening

Barry says the weakening effect has been replicated many times and that the thicker the enamel, the larger the weakening effect. However, thinner enamel silkscreened patterns such as lines or dots don’t have as much of an impact, he says.

Barry points out that thermal stress breakage seen in enameled spandrels originates closer to the edge of the glass because that is where the greatest load combines with the weakest part of a glass lite. He emphasizes that this breakage originates not at the edge but in slightly from the edge, so it should not be confused with breakage caused by edge damage.

Marino agrees that ceramic enamel does weaken glass by some percentage, but says there’s been some debate about the degree to which glass is weakened. He explains that ceramic enamel includes glass particles and those particles can expand at a different rate than the glass substrate. However, he says it’s important for the fabricator to ensure that the opacifier and the glass are a good match. If the substrate and opacifier don’t match well, he agrees that ceramic enamel could weaken glass. Marino explains that prior to Vitro’s acquisition of PPG and when PPG was still involved in fabrication, ensuring that the opacifier and substrate matched was an important part of the quality control process.

Kate Stewart, architectural specialist with ICD High Performance Coatings of Ridgefield, Wash., explains that if the inboard lite of an IGU is broken, the spandrel system can still withstand windloads. Broken glass is more of a concern when monolithic glass is used. She says that while a broken inboard lite on an IGU may not present an immediate safety concern, the glass could go unnoticed for a while and it’s uncertain how an IGU could be impacted over time.

Tim McGee, sales manager at Glass Coatings & Concepts (GCC) of Monroe, Ohio, says the ceramic enamel issue does not come up often. He attributes the root causes of spandrel glass breakage to design and environmental factors. Examples include the absence of a gap between insulation and the inboard lite or the presence of surface flaws due to cement overspray. Both could lead to breakage.

“People have been doing testing on pristine glass but that shouldn’t be extrapolated and applied to glass in service. Glass strength degrades with time and some of those factors should be taken into consideration,” he says, adding that any concerns should be addressed during pre-construction mock-ups.

Mitigation Strategies

Several methods exist to mitigate spandrel glass breakage caused by thermal stress.

According to ball drop tests performed by ICD and verified by Morrison Hershfield, heat strengthened glass with silicone opacifiers broke at drop heights 145%-150% greater than uncoated heat strengthened glass. Whereas heat strengthened glass with ceramic enamel broke at a drop height of only 31% of the value of uncoated heat strengthened glass. Therefore, Stewart suggests architects specify a silicone-based spandrel opacifier, as tests show that while frit reduces the lateral load and impact resistance of glass, silicone
does not.

Torok says avoiding the use of ceramic enamel is too limited an approach because it’s needed for applications requiring a pattern or photorealistic image.

Kayla Natividad, architectural technical services engineer for Pilkington North America of Toledo, Ohio, wrote her dissertation on the issue of ceramic frit weakening glass, inspired by research done by Barry and Dr. Scott Norville of Texas Tech University. She suggests that, at minimum, architects specify heat strengthened glass over annealed glass, even if the opacifier doesn’t require heat strengthening.

Marino recommends addressing the issue on the design side by venting the spandrel cavity using a shadow box design and using a backpan that will prevent insulation from falling and touching the glass.

If conditions warrant it, he suggests using fully tempered glass on the inboard lite. When tempered glass breaks it typically evacuates the opening. For example, if a fully tempered inboard lite breaks it will be contained within the cavity whereas an outboard lite would fall on the street. Heat strengthened can be used for the outboard lite because it is less likely to evacuate the opening and rain onto someone below. Another risk of using fully tempered glass is the potential for spontaneous breakage, though heat soaking can mitigate this risk to some degree. Marino says that, as a general rule, Vitro doesn’t recommend fully tempered glass unless it’s being used for safety glazing purposes.

In regard to ventilation, Torok says it’s unclear if this is an effective solution. Some air is ventilated out of the glazing system’s drain openings but air also can come back in, creating a pumping effect throughout the day. He says some designers and curtainwall manufacturers put vents at the top of the spandrel to dissipate air to the outside or through the mullions out the stack joints.

Morrison Hershfield did a study about ventilation in spandrel panels a few years ago using both monolithic and IGU systems to understand the temperature conditions inside the system. They tested these units with no ventilation, with ventilation using drainage holes and with a vent hole at the top of the spandrel. Torok says that while including a vent hole at the top improved the thermal stress in the monolithic system, it did not improve
the conditions in the IGU.

Taking Action

There have been efforts by industry advocates to include the weakening of glass by ceramic enamel into ASTM E1300, Standard Practice for Determining Load Resistance of Glass in Buildings. Others, however, don’t believe it belongs in the ASTM E1300 standard. McGee says that while some say “stress is stress,” GCC believes there is a difference between stress types. Marino explains that ASTM E1300 already accounts for glass weathering.

“By saying that the glass is weakened even beyond that by ceramic frit is double dipping on weathering. In our opinion, the aspect of glass being weakened by ceramic frit is not valid as the glass already has reduced strength characteristics,” he says.

Natividad says one of the industry’s pushbacks in regard to including ceramic enamel in ASTM E1300 is that it addresses load resistance due to uniform loads, or windloads. However, she argues that stress is stress and if there’s a load resistance reduction due to stresses from ceramic enamel then that should be addressed in all standards in which spandrel design is addressed.

“You can’t say we’ll put it in a new standard because when will that come out? Part of including it in ASTM E1300, even as a footnote, is making users aware so they can use their engineering judgment,” she says. “… I think when there is a thermal stress standard it should be addressed in there but it should be mentioned in ASTM E1300 at a bare minimum. It would create more awareness for designers that are specifying glass for spandrel applications.”

Barry says Europe has addressed ceramic enamel with its EN 12150-1 standard, incorporating a weakening factor of 35-38% on the load capacity of glass depending if the glass is heat strengthened or tempered.

“It’s very disappointing to me that we can’t get over this and add it to [ASTM E1300]. I will continue to work. It’s an accident that must not happen and we shouldn’t wait for it to happen,” says Barry. “… All enameled heat-treated spandrel breaks seen to date have been in double-glazed units, with the enameled light inboard. This means the broken fragments are unlikely to immediately fall to the street as they are probably restrained by the outer
light. The important message is that heat-treated enameled glass, in service, has been shown to be significantly weaker than expected … The clear warning from the thermal stress breaks is that current U.S. enameled glass strengths for wind and other loads are too high and must be revised.”

Having the ceramic enamel issue included in a U.S. standard would mean it could be accounted for in design, says Stewart, who adds that it would allow the industry to ask questions of manufacturers and create knowledge about the degree to which the product weakens glass.

Bill Lingnell, president of Lingnell Consulting, is a member of the ASTM subtask group that has considered the inclusion of ceramic enamel weakening. He says that ASTM E1300 addresses uniform load resistance under windloads and that thermal stress won’t be included in that standard because it doesn’t fit.

“It should likely have its own standard. There’s no question in my mind that the strength of glass with ceramic enamel is weaker than glass that doesn’t have it. The thing is quantifying that part,” he says.

Lingnell explains that now that a standard addressing thermal stress on monolithic glass exists, the next step is to create one for IGUs in which the ceramic enamel issue could be addressed.

“It’s not something that’s going to happen tomorrow. I would say it’s on the wish list of things to happen,” he says.

Jordan Scott is an assistant editor for USGlass magazine. She can be reached at

To view the laid-in version of this article in our digital edition, CLICK HERE.