3.5.1 Introduction
This section of the Guide introduces the concepts of fire protection and extreme loading conditions with respect to glass structures. It only concerns the base criteria against which glass structures are designed to for extreme events. More detailed explanations are provided in Chapters 9 and 10 of this Guide.
3.5.2 Design criteria for fire protection
Many may assume that it is not possible to protect glass from the effects of fire. This is on the basis that Table 3.2 Values of edge factor ke
Glass type Edge strength factor, ke
As-cut, arrissed, or ground edgesa
Seamed edgesb
Polished edges
Float or sheet 0.8 0.9 1.0
Patterned 0.8 0.8 0.8
Polished wired 0.8 0.8 0.8
Wired patterned 0.8 0.8 0.8
Notes
a Arrissed or ground edges by machined or hand where the abrasive action is across the edge.
b Arrissed or ground edges by machine or hand where the abrasive action is along the length of the edge.
its resistance to excessively high temperatures is low and as such is commonly regarded as a sacrificial element of building structures during a fire. In addition there is the issue of thermal gain from glass as it is very well suited to transmitting heat from one space to another. This tends to result in it becoming a conduit of heat transfer between parts of the building and thus aids the spread of the fire.
As a rough guide, the temperature difference limits for glass before it suffers thermal shock are as shown in Table 3.3.
Due to the low resistance of glass to thermal shock, glass will shatter easily. During a fire the temperature will reach a transition point (around 5008C) at which the glass starts to soften and lose its stiffness.
Ultimately when the temperature becomes high enough the glass will melt.
Glass is not a combustible element and therefore does not fuel fires. Glass shatters very easily when exposed to fire, as the combination of having a low tensile capacity and a high coefficient of thermal expansion causes glass to fail when exposed to high temperatures. Glass, however, does have a lower coefficient of thermal expansion (9 106K1) when compared to adjoining materials. For example steel:
12 106K1, stainless steel: 17 106K1and aluminium: 23 106K1.
One of the key design criteria for glass when considering its fire resistance is its integrity. If the glass fails then the fire has a greater chance of spreading. It is because of this that some effort has been put into maintaining the integrity of the glass during a fire at the design and manufacture stages.
One of the means to make glass more resistant to fire is to address its chemical make-up. If 7–15% boron oxide is introduced into the glass, then its coefficient of thermal expansion drops from 9 106K1to between 3.1 and 6 106K1. This has the effect of increasing the temperature difference at which the glass will crack and also increases its softening temperature.
Another way to increase fire resistance is to use laminated glass, as it is possible to have an interlayer material that is intumescent in nature. This creates a sacrificial ply in a two-ply laminated glass pane as the ply that is exposed to the fire fails and the interlayer expands into a foam in response to the heat from the fire. This acts as an insulator from the effects of the fire for the other ply that has not failed.
Finally the introduction of wire mesh into basic annealed or heat-strengthened glass is deemed to be a valid method. While it weakens the glass due to its crack inducing properties, it does keep the glass in
place after it has cracked and therefore maintains the integrity of the pane, even after failure.
3.5.3 Design criteria for blast loading
It is unlikely that glass will remain unscathed following an explosion. However, what is controllable is whether or not it will become part of the shrapnel after a blast has occurred. In areas that could be susceptible to a blast load, glass should be designed to remain in place, even if it has cracked.
This places an emphasis on both the glass itself and the method by which it is supported. With respect to the glass, the inner pane is typically laminated basic annealed or heat-strengthened glass. When the glass cracks, the laminate material holds the glass together even if it has become permanently deformed. In order to dissipate energy the glass needs to be allowed to crack and then deform, but remain in place. When designing glass elements that must withstand blast loads, positive and negative pressures must be considered.
It is usual to provide continuous support solutions for glass that could be exposed to blast loads. This is because the force from the blast is so great that point connections become impractical due to the high local stresses that occur.
Blast design depends on both the assessed threat and on the allowable damage and injury criteria. More discussion is given in Chapter 11.
The threat can be defined in terms of charge size, usually given as TNT equivalent, and the distance away (‘stand-off’). These parameters need to be derived in conjunction with the client’s security consultant.
Damage levels to glazing can be defined with reference to ISO 169333.5, which rates performance of windows into six hazard levels A to F, depending on how far fragments fall inside the building after a blast is applied outside a standard test chamber.
These categories correspond roughly to General Services Administration (GSA) categories 1 to 5 in the USA (see Table 3.4), with these latter categories commonly used in commercially available software.
Failure strength of the glazing system will depend on a combination of:
– the strength of the glass
– the strength of the interlayer in laminated glass – any clamps or structural silicone bonding the glass
pane to the framing
– the stiffness and strength of the framing itself – the strength of the brackets and fixings.
All of these elements need to be assessed in turn.
Allowable framing deflections under blast loading are typically much higher than those for wind, and can be of the order of span/15, based on plastic hinges being formed.
Blast resistant design of glazing systems involves time-dependent loading, non-linear material properties and geometry. Appropriate specialist advice should be sought.
The subject of extreme loading conditions, which blast loads are a form of, is also discussed in Chapter 11.
Table 3.3 Thermal shock temperature limits for glass Type of glass Temperature limits (8C)
Toughened 200a
3.5.4 Design criteria for seismic loading
Glass structures must be designed to resist racking movement if they are to be subjected to seismic forces. Provided the designer allows for this movement, then any seismic events will not have a significantly adverse effect on the glass. Additional horizontal and vertical inertia loads acting on glass structures should be applied concurrently with other conventional loads. These are typically considered in a static load combination with relevant safety factors applied.
It has been found that basic annealed panes that are laminated are less likely to fall from their frames than single pane sheets. In addition, the preferred method of fixing is via silicone bonding as that is proved to be flexible enough to allow racking movement to occur without forcing the panel from its supports. This is due to the fact that silicone joints allow for the dissipation of impact
energy, which is one of the most important characteristics for dynamic loads. In curtain wall cladding systems appropriate clear distance between glass panels and frame structure is to be specified to allow for movement during seismic events.
In high-risk seismic zones extra measures are required to mitigate seismic load. An example is to support a glass structure with an isolated base support. Damping systems are also installed within the connection interface between the glass and the primary structure.
In parts of North America where significant seismic events are likely, full scale static and dynamic testing is recommended according to AAMA 501.4-09 and AAMA 501.6-093.6. This tests the performance of the curtain wall and is evaluated while being subjected to the specific static and/or racking horizontal
displacement.
Table 3.4 General Services Administration (GSA) categories for blast loads
1 Glazing does not break. No visible damage to glazing or frame
No break The glazing is observed not to fracture, and there is no visible damage to the glazing system
2 Glazing cracks but is retained by the frame. Dusting or very small fragments near sill or on floor acceptable
No hazard The glazing is observed to fracture but is fully retained in the facility test frame or glazing system frame, and the rear surface (the surface opposite the airblast loaded side of the specimen) is intact
3a Glazing cracks. Fragments enter space and land on floor no further than 3.3ft from the window
Minimal hazard
The glazing is observed to fracture, and the total length of tears in the glazing plus the total length of pullout from the edge of the frame is less than 20% of the glazing sight perimeter. Also, there are less than 3 pinhole perforations and no fragment indents anywhere in a vertical witness panel located 3m (120in) from the interior face of the specimen, and there are fragments with a sum total united dimension of 25mm (1.0in) or less on the floor between 1m (40in) and 3m (120in) from the interior face of the specimen. Glazing dust and slivers are not accounted for in the rating 3b Glazing cracks. Fragments enter space
and land on floor no further than 10ft from the window
Very low hazard
The glazing is observed to fracture, and is located within 1m (40in) of the original location. Also, there are three or less pinhole perforations and no fragment indents anywhere in a vertical witness panel located 3m (120in) from the interior face of the specimen, and there are fragments with a sum total united dimension of 25mm (1.0in) or less on the floor between 1m (40in) and 3m (120in) from the interior face of the specimen. Glazing dust and slivers are not accounted for in the rating
4 Glazing cracks. Fragments enter space and land on floor and impact a vertical witness panel at a distance of no more than 10ft from the window at a height no greater than 2ft above the floor
Low hazard The glazing is observed to fracture, but glazing fragments generally fall between 1m (40in) of the interior face of the specimen and 0.5m (20in) or less above the floor of a vertical witness panel located 3m (120in) from the interior face of the specimen. Also, there are ten or fewer perforations in the area of a vertical witness panel located 3m (120in) from the interior face of the specimen and higher than 0.5m (20in) and none of the perforations penetrate through the first layer of the witness panel
5 Glazing cracks and window system fails catastrophically. Fragments enter space impacting a vertical witness panel at a distance of no more than 10ft from the window at a height greater than 2ft above the floor
High hazard
Glazing is observed to fracture, and there are more than ten perforations in the area of a vertical witness panel located 3m (120in) from the interior face of the specimen and higher than 0.5m (20in) above the floor, or there are one or more perforations in the same witness panel area with a fragment penetration into the second layer of the witness panel GSA
condition
ASTM rating
GSA description ASTM description
3.6 Location
3.6.1 Introduction
The location of glass elements within a building affects the safety of its occupants. If glass is placed in an elevated position for example, then it has a risk of falling and injuring someone. It is therefore important that some consideration is given to this aspect of structural glass within a structure. The questions that have to be asked are: can a glass element fall, and if so, what are the consequences if it does so? If the answer to these questions is ‘yes’
and ‘harmful’ respectively, then measures must be taken to prevent it from happening.
CIRIA has produced a guide on glazing at height:
C632 Guidance on glazing at height3.7provides designers and contractors with advice on what needs to be taken into consideration when placing any glazing in a location that, should it fail, would cause harm to those beneath it.
The forms of structure that are susceptible to the risks of falling glass include:
– fac¸ades – roofs – canopies – barriers to drops – walkways/bridges – staircases.
3.6.2 Post-failure behaviour of glass at height In order to design out the risk of injury due to falling glass, an understanding of what occurs to a glass element after it has failed is required (see Figure 3.2).
Once the failure mode has been established, then it can be taken into account during the design process.
In many respects the precautions a designer should take when considering glazing at height are similar to those for blast loading, as described in Section 3.5.3.
The loads that are considered in the post-failure condition are usually self-weight and a fraction of the imposed load over a short-term period (e.g. hours to days). The glass should remain in place or fail safely, and therefore its method of support must allow for this. In the case of glass floor plates, it is also required that a failed element should be able to support the traffic of people over a certain period of time in order to allow for a safe escape from the building.
3.7 Environment
The environment in which a glass structure is to be placed has an influence on the design criteria. They can be categorised into three aspects: lighting, temperature and acoustics.
Lighting
Historically windows were the primary source of light in buildings. This forced architects to devise different ways to exploit sunlight as much as possible to ensure it flooded a building. With the invention of artificial light these design skills fell away for a while, until the late 20th century when buildings began to feature large expanses of glass. This led to a reappearance of natural light and has become a significant element of lighting design.
Consequently, the effect glass has on the passage of natural light within a building has become increasingly important. The clarity and colour of glass has a direct impact on the natural light transmittance. Clear glass has around 0.1% of iron oxide within it which absorbs the colour red from the spectrum of natural light. It is for this reason that clear glass has a green tint to it, which is more pronounced around the edges of the pane. The benefits of natural daylight over artificial lighting have been proven to be significant for the occupants’ wellbeing and comfort.
Glass can also be manufactured with a lower iron content (called low iron glass) which will reduce the green tint and allow more visible light to pass through the pane. This type of glass can be heat treated in exactly the same way as standard glass, but is significantly more expensive to produce.
The level of opacity also has an effect on the way in which natural light is dispersed within a building. The more opaque the glass is, the lower level of
luminosity is achieved within the building.
Transparency describes the amount of light that is let in through glass as well as how it is dispersed. This can be altered by applying obscuration effects, either to the glass surface or to the glass itself via patterns.
Thermal
Solar gain is a phenomenon that has become increasingly important as more expanses of glass are installed into structures. If not properly controlled it can create an uncomfortable environment to live and work in due to over-heating, or create additional requirements for cooling.
Conversely, heat loss through the glass and the structure that supports it can increase the level of heating required to maintain internal temperatures during a cold spell.
Figure 3.2 Broken glass pane within canopy
To overcome these problems a mixture of shading, protective coatings and multi-chambered glazing, along with the actual glass specifications, is employed to create the optimum environment within the building’s enclosure.
Acoustics
Sound insulation protects the building from undesirable noise pollution. Glass should attenuate any noise so that is does not annoy occupants.
BS 8233: 19993.8is the UK code of practice for sound insulation and offers some guidance on limits of sound transmission. With glass there is a relationship between sound transmission and the mass of the glass.
The choice of single and multiple glazing has an impact on the degree of sound insulation. With single glazing it is the base resonance of the glass pane that impacts its ability to transmit sound. The thinner the glass the higher its resonance and hence the likelihood of it generating sound due to it vibrating.
Additionally, the thinner the glass the lower its ability to insulate against sound from external sources. To overcome this, single panes are laminated to increase the thickness of the glass and the properties of the laminating interlayer further enhance the acoustic performance.
For the purpose of sound insulation there are
‘acoustic’ PVB interlayers, which are softer than their non-acoustic counterparts. Alternatively it is possible to use simple cast resin as an interlayer, which is known to have equivalent acoustic properties to the softer acoustic PVB material.
For multiple glazing panels there is an interaction between individual panels of glass that make up the unit. If one of the panes is thicker than the other, then the sound insulation is improved as the two panels resonate at different frequencies, which causes beneficial acoustic interference.
The choice of frame and the type of fixings to this frame will also play a part in the overall acoustic performance of the element.
3.8 Testing
Chapter 4 goes into some depth concerning the creation of models within computer simulation of glass structures. Although the level of computing technology continues to advance at a significant pace, there is still a need in many instances to carry out testing of key elements within glass structures.
Testing can be split into four categories:
developmental, strength, visual and verification.
Developmental testing is carried out during the design process. Strength testing concerns the integrity of the glass structure, while visual testing relates to the aesthetics of glass assemblies.
Verification tests are used to establish the validity of what has already been designed and, in some cases, constructed. Elements that are tested are based on computational and theoretical analysis. The results of the test are then compared against what the theoretical model predicted. If they are significantly divergent, then either the analysis is checked or the
test sample is reviewed. Figure 3.3 shows a staircase that is undergoing a load test.
EN 1990: 20023.9provides general guidelines to carrying out design that is assisted by testing.
However this code of practice has not been created with glass in mind, and the designer must be mindful of this when making reference to it.