We’ve been involved in many interesting structures, but the most challenging have often been those that seem very simple in nature. This simplicity is often deceptive; more often than not it will be the sole goal in driving the engineering challenge.
One of the interesting aspects of working with glass as the primary structure is the diversity of additional materials that are used in the typical design, from silicones and epoxy mortars, to a wide variety of interlayers, numerous coatings and methods of adding decorative effects. Throughout the evolution of the design, it is often these secondary materials, or the components fabricated from them, rather than the glass, that dictates the performance of the overall structure.
The Steve Jobs Theater (pictured above) provides one such example of how the design leads us down these investigate trails.
Without any internal or external structure, the roof had to appear to float above the landscape. Any distortions would detract from this effect. Selecting the method to curve the glass was the critical starting point; different production techniques would lead to different structural capacities, but the decision had to come from a balance with the visual aesthetic. In this case, the quality of the surface could be achieved.
The selected method was lamination bent fully toughened glass. This is a panel that is formed from layers of flat tempered glass, interleaved with a stiff interlayer which is all then forced over a form, and placed in the autoclave. At the end of the autoclave cycle when the interlayer has bonded to the adjacent glass. There is an initial geometric change when the panel is released from the form. The strain energy in the individual glass plies is redistributed across the full panel thickness. This is often referred to as spring back. The compromise of this process is that stresses are locked in the glass so there is less to use for the structural work. Given that the glass needed to support and provide lateral stiffness to a combined weight of 250 tonnes there was a lot of structural work to do in the connection to the foundation.
The preferred option was to bolt the glass but the loads required could not be achieved. The stress concentration involved with holes acts to magnify the local applied stress. The reduction in the residual structural capacity became so great that the method was discounted. Using a continuous adhesive connection across the bottom of the glass was the logical next step. The force could be distributed more evenly. Researching the various bonding technologies, we decided to proceed with a two-part structural silicone. While this is readily used for structural glazing, we understood that this would be the first building in California to use it in the load path of the structural frame. Needless to say, this type of application was not in the Local Building Code.
The initial sizing of the silicone joint used published data from Dow, and other test data we had, to satisfy stiffness and strength. In the design of the theatre though, the peak stresses came from the dynamic loads caused by seismic events, and so further testing was commissioned to understand what happened to the silicone under load cycling. The results showed that the silicone was softer than expected.
While this could be resolved in the design, we wanted to understand why. After discussing this at length with Dow’s team, this effect which varies the softness of the material as it is repeatedly stretched is known as the “Mullins” effect, and it is common to all silicones. It was explained most simply by the material scientist as the reason why stretching a party balloon several times makes it easier to blow up. Under normal glazing design conditions, such an effect on the silicone is of minimal concern and possibly negligible when compared to the safety factors. It was certainly important in our case, and the design was adjusted to accommodate these values.
We encountered a similar design evolution in the development of the K11 façade in Hong Kong. Again, the premise of the project has a very clear architectural intent, and thus an engineering problem of how to make a series of glass cylinders and the journey of starting again from choices of manufacturing.
Conceptually the facade was to be a series of cylinders, 9m high, 900mm dia., self-supporting structurally while also acting as an insulated unit, wrapping around the building on the Retail floors. We were aware of the glass tubes that Schott and other drawn glass manufacturers make, usually for laboratory use and in the chemical industries.
At the time of the project, Schott were only able to make the tubes 400mm in diameter and 3.5m long so we looked at ways they could be stacked, and overlapped with semi-circular sections. Alternatives included modifying the production facility so that the length could be increased, which was a possibility.
With that in mind, how could we ensure that the glass tube would remain safe under damage? Safety films were rejected because of the difficulty of application and, and so laminating the glass was the natural alternative. However, the laminated panels failed in the autoclave. It was considered that the variation in geometry initiated these failures.
The next preferred option was to make the cylinders from a pair of semi-circular panels. We were expecting that more consistent and accurate geometries could be achieved with this slumping process. Thermally bent glass was out of the question because the tight radius was unachievable.
Using annealed glass in a scheme that had multiple orientations of exposure required some careful assessment about the likelihood of thermal shock, as the failure stress of annealed glass is low. Dynamic solar modelling gave the worst-case differential solar exposure on each of the cylinders. These radiation values allowed the calculation of the temperature and, in turn, a computer model calculated the maximum stresses for the worst condition, which fell within acceptable limits.
Similar to the problem of isochoric pressures for IGUs produced and located at different altitudes, this system was even more susceptible to those pressures as the cavity volume was very high, and the structure very stiff. The long-term performance of the seals became a concern. We, therefore, developed a closed cavity facade system which fed dry air to the units via a ring main. This gave a consistent positive pressure to the tubes, reducing gasket performance requirements and keeping the moisture out.
These projects use glass and its production techniques to their best advantage, harmonising qualities of surface, clarity, strength to best fit the design intent.
It has been extremely exciting to be a part of achieving these new forms and structural gymnastics and the future is going to show us even more of the capabilities of glass. We are hoping the quality of glass and production will continue to improve. At our disposal, we currently have an immense range of surface finishes, coatings and laminating materials that combine for even more permutations in the final product. However, they are predominantly fixed at the design stage. The future must lie with what is loosely termed as smart glasses. One can think that the transparency, strength and size of the structural glass forms will become the carrier for some amazing technologies to provide other active or reactive qualities.
The shift to managing heat transfer and light transmission must be at the heart of this. Coatings have greatly improved and are the primary tool at the engineer’s disposal. The choice of coatings has also expanded and includes more neutral coloured coatings with less reflectivity. Improved performance has come from the specific selectivity of wavelength, but it is a still static system. A case of best fit.
Glasses with variable performance could help us achieve much more. Commentary on the varied dynamic mechanical shading options that are available are ignored as they are an additional system to the glass panel.
Electrochromic glasses such as Sage, EControl and Viewglass to name a few can switch between polar states to modify the Light transmission and G-value, with some able to achieve intermediate steps. These systems need to be actively switched from one state to another, and that transformation is relatively gradual, taking a number of minutes. The tinting of them varies but is predominantly a dark blue hue. A modified PVB material can also be used to a similar effect but in an analog way compared to Electrochromic’s digital nature. Thermochromic glass uses the suns own energy to passively change the light transmission and solar gainLiquid Crystal technology has also been developed, notably by Merck. It is an active or passive technology, controlling solar heat gain and light transmission with liquid crystal molecules changing orientation with the application of a current. The change is immediate, and the colour of the translucency can be varied by adjusting the dye; both major benefits of the technology.
In both passive and active cases, the ability of these technologies to modify translucency add the next step in advances to the glass skin. However, the translucency always has an effect on both light transmission and solar gain. Can we have a system that can control either one i.e. change the solar gain while maintaining consistent light transmission?
There are very exciting research projects that we are involved in, which aims to separate this functionality. Of that very little can be revealed, but going back to an earlier discussion of the glass acting as a carrier for technology, why not use it as a completely variable information carrier? Leaping on from the technology of phase change materials that allows computers to read and write data to optical disks, Bodle Technology in Oxford have taken this one step further and used a modulated phase change alloy in combination with various ultrathin films that change the refractive index of light to create reflected colour. Similar refraction causes rainbows and the beautiful blue Morpho Butterfly wings.
In the same way that colours can be chosen, so can particular wavelengths, adding the possibility of cutting out solar energy without effecting light transmission. This technology is currently being developed for the higher added-value market of screen technology, but it’s easy to see what this technology could do for the architectural industry, and not just in decoration or signage. The beauty of this technology is that is utilises standard coating materials and systems and so the hope is we might see this in the architectural field in the near future.
Right now, the glass industry is a very exciting place to work, to experiment and to learn. We are surrounded by fascinating new products which enhance performance and provide decoration in increasingly inventive ways. Devising new applications for these emerging technologies is only going to get more exciting.
This article first appeared in IGS Magazine’s Winter 2018 Issue – Read the full Magazine here for more thought-leadership from those spearheading the industry
Graham joined Eckersley O’Callaghan in 2004 as our first team member when the company was established. With 20 years experience as a structural and facade engineer, he has a particular passion in the development of structural glass design using sophisticated analysis and modelling tools.
Graham’s technical expertise in pioneering glass engineering has been instrumental in delivering the practice’s many challenging and award-winning glass projects, notably for Apple.
As Technical Director, Graham is responsible for the strategic operations across the company. This includes defining design processes and project delivery for project excellence. He also leads our R&D programme driven by a pursuit to explore the innovative use of material in design.