The glass ‘butterfly effect’; The risk of glass failure due to nickel sulphide inclusions
During the life cycle of glass panels, breakage can occur due to a number of factors and many of those factors can be difficult to identify with any degree of certainty once the glass is broken. However, some types of breakage leave telltale clues, which can help an appropriately qualified and knowledgeable person determine the cause of failure. One such case is that of nickel sulphide (NiS) inclusions.
It is often desirable to modify float glass panes from their original annealed state, in order to enhance certain characteristics, such as structural strength and resistance to thermal stress. The most common way to achieve this is to thermally toughen the glass, which involves a precisely controlled heating and cooling process. This process increases the glass stress capacity and changes its fracture mechanism from large sharp shards to small blunt dice, which are less likely to cause injury. Hence, thermally toughened glass is considered to be a safety glass.
Unfortunately, this treatment of the glass can potentially introduce a new failure mechanism, inclusion failure, which is often attributed to the presence of Nickel Sulphide crystals within the material. Nickel sulphide crystals are very small impurities (often circa 100-300μm) that can be present in all glass types, entering the batch of raw materials during manufacture.
Nickel sulphide crystals are non-passive, meaning they can change state and consequently volume. At everyday operating temperatures for glass, the crystals are stable in their beta state. However, the toughening process raises the temperature of the glass (and any inclusions) above 600˚C and during this regime NiS crystals are transformed to a temporary alpha state. The crystal in its alpha state has a lesser volume than that of the beta state and as the crystal slowly returns to the beta state (often over a number of years) the increasing volume of the crystal can trigger a sudden and catastrophic failure of the glass without warning. Generally, the larger the nickel sulphide inclusion, the greater the change in volume and therefore the greater the risk of glass breakage.
Failures due to nickel sulphide inclusions are notable for the distinctive ‘butterfly’ fracture pattern they produce, as well as the spontaneous and seemingly random nature of their occurrence. However, whilst failure due to nickel sulphide inclusions tends to produce the recognisable butterfly pattern, other failure mechanisms can also result in a similar effect and therefore the presence of a butterfly pattern does not automatically mean a nickel sulphide inclusion is the culprit.
If a toughened glass failure has occurred and the presence of a nickel sulphide inclusion is suspected, a sample of the broken glass containing the fracture epicenter (butterfly) can be sent to a laboratory for identification using an electron microscope (SEM) with Energy Dispersive X-ray (EDX) analysis capability.
Failure of glass due to the presence of nickel sulphide inclusions is an issue gaining notoriety with the construction industry, affecting developers, contractors and insurers alike. Due to its failure mechanism and subsequent breakage pattern, toughened glass suffers a substantial loss of structural integrity post fracture. Where the glass is used to provide guarding, such as in a balustrade application or full height glazing, this introduces a significant risk of falling from height for building occupants. Whilst toughened glass breaks into safe dice, these small fragments tend to form clumps which can fall as much larger pieces from the building.
In addition to the potential health and safety risks, the economic risk should also be considered when evaluating the use of toughened glass. In many instances, the warranty provided by glass processor specifically omits failure due to the presence of nickel sulphide inclusions from its cover. Factor in the possible tenant dissatisfaction, long lead times and the potential effect on the public perception of the development, and the combined risks associated with the use of toughened glass may prove the be unpalatable.
Whilst failure due to nickel sulphide inclusions has been acknowledged for over 50 years, a method of completely preventing NiS crystals entering the mixture of raw materials has yet to be devised. There is a general perception within the façade industry that four tons of toughened glass is likely to include one NiS inclusion of sufficient size to result in spontaneous glass breakage, however Wintech has yet to see any evidence that substantiates this popular belief. Furthermore, it can be potentially misleading to apply this type of statistic when assessing the risk of failure as, in Wintech’s experience, certain batches of glass can suffer much higher rates of failure than others.
Reducing the risk
Where the risk of spontaneous breakage and subsequent potential for falling glass is unacceptable, Wintech would recommend an alternative to toughened glass is sought. This may lead to annealed, heat strengthened, chemically toughened or laminated glasses being utilised instead. In all instances, a fully developed and detailed glass risk assessment should be undertaken by a competent person to ascertain and consider the residual risks associated with the selected glass type.
Where toughened glass must be used, the risk of spontaneous breakage due to the presence of nickel sulphide inclusions may be reduced by implementing a process called heat soaking. Heat soak testing is a destructive test regime which aims to trigger the phase change of NiS inclusions, breaking the glass which contains inclusions of critical size, prior to the glass leaving the processors factory. The process is not 100% effective and glass containing NiS inclusions of a critical size may still remain after the heat soaking is complete. The European Standard to which heat soaking is carried out, BS EN 14179, states that glass which has successfully passed through the heat soak process should have less than one critical NiS failure per 400 tons, on average.
As such, it is generally Wintech’s recommendation that where the use of toughened glass is unavoidable, it should be subjected to a heat soaking regime and that evidence of the correct implementation of this test is demonstrated by the glass processor. Furthermore, full traceability of the glass, from initial batch production through to installed location, should be provided.
Wintech have over 35 years of unmatched expertise in all facets of façade design and the practical application of bespoke and standardised façade solutions. We can provide expert advice on all aspects of the selection, use and residual risks associated with all glass types.
Article courtesy of Dr. Robert Edmondson and James Smith of Wintech.