Features Business intelligence Contracting
GLASS SCIENCE: Including inclusions

Spontaneous glass breakage is a
term used when glass breaks for no apparent cause. The most common
causes are attributed to minor damage during installation such as
chipped or nicked edges which later develop into larger breaks,
movement in the building or framing system, thermal stresses in the
glass or its inability to resist wind load.

Nickel sulphide stones exists in a number of forms such as Ni3S2, Ni7S6, Ni9S8 which are typically not a problem, but one – Ni1-xS – can be a problem because it is not stable.

All inclusion breaks will have ‘cat-eyes’, but not all breaks with ‘cat-eyes’ are caused by inclusions.

Inclusions in the float glass are relatively rare, practically invisible, and not practical to detect in a production environment.

Michael Bitterice, the senior engineer technical services for PPG Performance Glazings, presents a seminar on spontaneous glass breakage. He has announced his retirement from the glass industry this summer.

Glass breaks for a reason.

Spontaneous glass breakage is a term used when glass breaks for no apparent cause. The most common causes are attributed to minor damage during installation such as chipped or nicked edges which later develop into larger breaks, movement in the building or framing system, thermal stresses in the glass or its inability to resist wind load.

However, Michael Bitterice, the senior engineer of technical services for PPG Performance Glazings, says that inclusions in the float glass during the manufacturing process can also be the culprit. He says there are as many as 50 possible inclusion types in float glass. They include, among others: zirconia, alumina, silica and nickel sulphide. While there are a variety of inclusion types, the most widely discussed is nickel sulphide (NiS), a type of stone or crystalline blemish. Nickel sulphide stone exists in a number of forms such as Ni3S2, Ni7S6, Ni9S8 which are typically not a problem, but one – Ni1-xS – can be a problem because it is not stable.

Spontaneous breakage
“This occurs in float glass when ‘tramp’ nickel rich particles, such as stainless steel, combine with sulphur in the melting process. It’s relatively rare, practically invisible, and not practical to detect in a production environment,” Bitterice says, because they are
one-tenth of a millimetre – A 0.1mm/~0.004 inch – in diameter. The frequency is random within the float ribbon so the frequency and location of individual inclusions is practically impossible to quantify or predict in individual glass lites.

NiS stones undergo a phase change in crystalline structure that results in an increase in volume when cooled slowly from about 750 degrees F down to room temperature. In standard float glass, this expansion takes place while the glass is at annealing temperatures, so stress relaxation eliminates the tensile stress caused by the phase change.

NiS inclusions may undergo a phase transformation causing them to shrink due to the reheat/quench operations required to fabricate tempered glass. The rapid quench
traps the inclusion in its small phase, but over a period of time, these inclusions may revert back or expand to their original state. When the inclusion is located in the centre tension area of the glass, the increase in volume may cause a localized stress increase sufficient to break it several years after the glass has been fabricated, even though the glass is not under load. Such inclusions can cause spontaneous breakage in tempered glass, without any load being applied, at any time, even five or 10 years after the tempered glass has been fabricated.

“NiS undergoes a phase change, from the alpha-phase at higher temperature to the beta-phase at lower temperature, with a volume increase of approximately four percent. If the stone is in the tension zone of tempered glass and undergoes the phase change, the stress caused by the increased volume may lead to breakage,” Bitterice explains. “It doesn’t take much to get a four percent increase in the size of the growth of the stone.”

When are they a problem?
“In annealed glass, never, because any stones undergo the phase change and become stable during the slow re-annealing of the glass in the float glass manufacturing process. In heat strengthened glass it is not a problem when done properly, because the
slower cooling allows any stones to undergo the phase change and become stable. But it can occur in tempered glass, even though it’s rare, because the relatively rapid cooling may arrest the phase change of any stones, leading to possible continuation in the field much later after the glass has been installed that results in glass breakage.”

He says the problem exists in tempered glass because in the tempering process, the more rapid cool down arrests the transition from alpha (smaller) to beta (larger) which
locks the stone in. “The ‘alpha to beta’ phase change may continue with an accompanying volume growth of approximately four percent. The phase change is time/temperature dependent and this volume growth, if it occurs in the tension zone, can create a stress of 500,000psi causing breakage.

Since inspection is not practical, glass manufacturers have implemented programs of batch quality control, along with the elimination of any nickel containing materials from their raw material and glass handling systems. Automatic inspection of 100 percent of the float ribbon can also contribute to the reduction in the number of inclusions of all types present in float glass.

No guarantees
“Keep in mind that nickel sulphide inclusions in float glass are rare. No batch compositions produced in North America intentionally includes nickel and there
are procedures in place to constantly monitor raw materials. For example, nickel plates, rods and bearings have been eliminated from processing equipment, but this still
can’t guarantee 100 percent elimination of NiS inclusions,” says Bitterice.

Another method used to detect this before it leaves the factory is ‘heat soaking’, a process that can uncover some NiS inclusions present in an individual lite of glass; however, this is not 100 percent effective.

“This is an attempt to initiate or accelerate the phase change in the plant that typically exposes glass to temperatures between 280 and 300 degrees C for two to four hours. The objective is to achieve a ‘break now, not later’ result, based on the assumption that any lites with inclusions will break in the process,” he says, adding that this can be conducted on a statistical sampling basis or on entire lots.

Another solution is to specify heat strengthened, rather than
tempered, when possible. “Heat strengthened glass is adequate to meet virtually all thermal and wind/snow load requirements and it is 100 percent effective when done properly,” he says, adding that it is made on the same equipment as tempered. “But remember, it is not a safety glazing material.”

After more than 15 years, Bitterice says experts in float glass production are still debating the cause and remedy to this problem. “There is general agreement on
most of the science behind this, but not the conclusions,” he says, pointing out that advocates of heat soaking are not absolute in their support of how the process is represented. “You may see ‘avoids’ or ‘reduces’ or ‘minimizes’ breakage, but you won’t see ‘eliminates’ or ‘prevents’ or ‘guarantees’ against breakage. Heat soaking can eliminate, by destruction, some of the nickel sulphide inclusions, but it’s not a guarantee for eliminating all inclusions.”

Although it is easy to blame inclusions for spontaneous glass breakage for glass that has already been installed on site, Bitterice says the infamous ‘cat-eyes’ or ‘butterfly’ break pattern that is characteristic of these type of breaks are not always caused by them. “This is extremely rare. Out of 10,000 square feet of six millimetre glass, two pieces out of the 10,000 might be affected,” he says.

“All inclusion breaks will have ‘cat-eyes’, but not all breaks with ‘cat-eyes’ are caused by inclusions. It doesn’t prove that the break was caused by an inclusion. Glass breaks for a reason and more than likely, it broke for other reasons.”