A Review of the Behavior of Structural Silicone Glazing Systems Subjected to a Mega-Earthquake

ABSTRACT: Structural silicone glazed (SSG) curtain wall systems have offered architects and owners the ability to design a facade with unique aes­thetic features since the 1970s. While their ability to durably withstanding nat­ural weathering and even high wind locations across the globe has been well established, their performance under seismic events is less well recorded.

This paper presents a review of the performance of SSG systems following the 8.8 magnitude earthquake that shocked Chile in Feb 2010, which regis­tered within the top five recorded in past history. Field reviews of existing low-, medium-, and high-rise buildings were inspected in the aftermath of the event to evaluate the performance of the different SSG variations of this sys­tem type (two-sided, four-sided, stick, unitized, among other variations) com­bined with other facade components such as: type of glazing, glass sizes, interstory drift, width of structural grid, slab/beam rigidity, etc.

KEYWORDS: structural silicone glazing, earthquake, SSG, high strain rate behavior of structural silicone, curtain wall systems, interstory drift

Introduction

Looking at structural silicone glazed (SSG) designs from the 1970s [1], 1980s

[2] , and 1990s [3] and beyond, one will mostly find a large variation on a theme: glass bonded to metal. SSG systems can encompass one-, two-, three-, and four – side bonded designs, with glass, metal (usually aluminum), and occasionally other materials. Yet, however variable, they are all conceptually similar whereas

glass or metal plates are bonded to a metal curtain wall part utilizing a struc­tural silicone rubber adhesive as shown in Fig. 1. This silicone rubber adhesive acts as a flexible anchorage resisting wind pressures and at the same time absorbing translations and rotations due to imposed loads from: gusts, thermal expansions, and building movements (see Fig. 2). In an earthquake, in-plane rotation of inflexible glass would impart short-term shear strains of potentially large magnitude into the rubber. The most commonly used thickness of the structural adhesive in SSG systems is 1/4 in. (6.4 mm), far thicker than most “adhesives" in the traditional sense. That the adhesive in an SSG system has this thickness is what allows for the silicone rubber to stretch, rotate, and accommodate imposed strains. In a seismic event, the adhesive could experi­ence displacements that may not have already been accommodated by the cur­tain wall system. Further, the commonly used 1/4-in. thickness can be increased to improve the in-plane flexibility of the system.

FIG. 2—Di Is.

Copyright by ASTM Int’l (all rights reserved); Tue May 6 12:07:08 EDT 2014 Downloaded/printed by

With respect to the in-plane deformation that an earthquake could place into a curtain wall system, it is not hard to imagine that a flexibly bonded assemblage (i. e., SSG) could offer improved performance over non-bonded cur­tain wall systems (“dry-glazed" or gasket-type systems that rely on compression or friction) when subjected to racking displacements from an earthquake. The bonded rubber adhesive would tend to “control" movements between parts, but on the other hand, could fail through shear if overstrained. Movements into a gasketed system would be more uncontrolled, and gaskets can be dislodged introducing yet additional uncontrolled movement between moving compo­nents. The thought here is that the uncontrolled movement would lead to an earlier collision between glass and metal, and thus breakage at an earlier stage in an event.

The ASTM C1401 industry guide on SSG [4] offers some commentary on performance of these systems and suggests that there are potential intrinsic benefits to using SSG systems in seismic regions, such as:

(a) Controlling and in some cases eliminating breakage normally experi­enced during a small to moderate earthquake.

(b) Minimizing the opportunity for glass to impact the metal glazing pocket surfaces, eliminating a primary cause of breakage (in four-sided SSG systems, the lite or panel is not captured in a metal glazing pocket).

(c) Also, when a glass lite break does occur, the SSG system, because of continuous attachment of the glass edge, can retain much if not all of the broken glass, depending on glass type, and provided that the struc­tural joint retains sufficient integrity.

Some of the ASTM suggestions above are supported by Penn State Univer­sity (PSU) research studies via seismic simulations on two-side SSG systems where Memari et al. [5] conclude “that SSG systems can perform favorably in seismic events over conventional dry-glazed systems" with test results showing an approximately 65 % increase of cracking drift capacity (onset of glass cracks) when compared to a similarly sized and tested dry-glazed (non-SSG) system [5]. The research by PSU has identified that two-side SSG systems have improved serviceability and ultimate drift capacities than do dry-glazed (non-SSG) sys­tems [6].

Additional research at PSU on four-side SSG systems [7] shows yet fur­ther improvement in drift capacities than does the two-side SSG research, concluding; “Comparison of four-side SSG system test results in this study with those of dry-glazed and two-side SSG in past studies shows that four – side SSG has higher cracking drift capacity than previously tested dry-glazed and two-side SSG by 146 % and 55 %, respectively. Moreover, the four-side SSG system has higher glass fallout capacity compared to dry-glazed and two-side SSG by 54 % and 13 %, respectively." Additionally, “This study sup­ports the commonly held notion in the industry that SSG systems, in general, and especially four-side SSG configurations are less vulnerable to glass dam­age in earthquakes. This notion is predicated upon the structural sealants being applied in accordance with silicone manufacturers specifications, and that silicone material properties have not deteriorated due to in situ weather­ing effects."

Copyright by ASTM Int’l (all rights reserved); Tue May 6 12:07:08 EDT 2014

Downloaded/printed by

Rochester Institute Of Technology pursuant to License Agreement. No further reproductions authorized.