Shear Strain Imposed into the Structural Silicone

Here we will use a real example from the earthquake (Isidora Foster building case study presented in more detail later in the paper) and make an estimate of the shear strain that could have been imposed into the structural sealant and compare this value to the shear strain capacity that could be expected of a two – part high modulus structural sealant when subjected to strain rates higher than those used in standardized tests (Table 1). Also recognize that in four-sided SSG systems, the glass will most closely rotate about its center of mass; therefore the drift is halved as it is split between opposing corners.

Consider first the worst-case scenario of the building where:

(1) The maximum interstory drift allowed by code is reached a 33.9 mm (1.33 in.). See worst-case lateral displacement calculation on p. 14.

(2) Assume no contribution of strain lowering from the frame anchoring and/or the curtain wall frame assembly.

In this scenario, the drift is imposed into the structural sealant and the shear displacement occurring at the location of highest strain (corners) is 16.9 mm (0.66 in.). Using trigonometry, the actual elongation experienced in the 6.4 mm (0.25 in.) thick bead is 18 mm (0.71 in.), which represents a 181 % shear strain. This value exceeds that which most, if not all, two-part high modulus structural silicones currently available could be expected to accommodate at a 6.4 mm (0.25 in.) thickness and at any strain rate. This is supported by the % elongation at maximum load values reported in Table 1. A shear strain of this magnitude would have manifested itself as failure through rupture and tearing of the structural seals at corner locations with possible dislocation of glass or fallout. Because this did not occur, it is reasonable to conclude that the curtain wall and anchoring system functioned properly to accommodate displacements from the event and the structural seals experienced strains lower than those at max load (strains after maximum load but before break of the specimen are associated with tearing of the rubber).

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.

Now consider a more realistic scenario where:

(1) A 9.2 mm (0.36 in.) actual interstory drift is used (this is the drift pre­dicted to have occurred on the Isidora Foster building as estimated by fi­nite element modeling taking into consideration soil conditions, nearby measured ground accelerations, and overall building response), and

(2) assume no contribution of strain lowering from the frame anchoring and/or the curtain wall frame assembly.

In this scenario, the drift is imposed into the structural sealant and the shear displacement occurring at the location of highest strain (corners) is 4.6 mm (0.18 in.). Using trigonometry, the actual elongation experienced in the 6.4 mm (0.25 in.) thick bead is 7.9 mm (0.31 in.). This represents a 23 % shear strain, which is well below that which a two-part high modulus structural seal­ant can accommodate before damage to it occurs (reference values from Table 1 above).

Correlating this 23 % strain with a stress, we can derive from Fig. 24 (and the associated load-displacement data generated to make these curves) a stress of approximately 324-331 kPa (47-48 psi). However, Fig. 24 represents load – displacement curves generated using a very slow strain rate [the standardized strain rate of ASTM C1135; 50.8 mm/min (2 in./min)], which is not entirely rele­vant to the response of structural sealant from loading because of a seismic event. Considering the data generated at the higher tested strain rates, we can get another estimate of stress in the structural sealant. From Fig. 27, which superimposes the averages of the three tested rates in shear, it can be seen that the two faster rates closely parallel each other throughout most of the tested range. From these two faster strain rates, the associated stress at 23 % strain is in the range of 448-462 kPa (65-67 psi).

The stress estimates from the above scenarios must be tempered with the two following considerations:

(1) Both scenarios assumed that the curtain wall and anchoring system did not contribute to reduce shear displacement into the structural seals, which is unlikely.

(2) The faster strain rates used to derive the latter stress estimate are likely beyond what the structural sealant would experience in a seismic event.

Both of these considerations, taken alone or combined, would tend toward lowering the estimated stress ranges shown above. Regarding shear strain rates relative to seismic events, future work is of interest to better correlate SSG behaviour with lab and mock-up testing.