Silicone sealant model

Adjacent panels are typically connected by slots and ribs, and the joints are filled by narrow silicone strips. The response of silicone sealant under imposed shear strains was studied by Dal Lago et al.

(Dal Lago, 2015; Dal Lago et al., 2017b; Negro & Lamperti Tornaghi, 2017), who performed several experiments on concrete blocks, sub-assemblies, and full-scale structures with cladding panels sealed with silicone.

The typical hysteretic response of the silicone sealant is presented in Figure 6.12 on a shear–stress versus shear–strain diagram. A relatively good agreement of the hysteretic response during the cyclic test on concrete blocks and subassembly structure was achieved. However, the experiments have shown a large scatter of the silicone properties because the material is usually not subjected to strict production control. Despite that, basic features of the silicone sealant have been identified.

According to the results of Dal Lago et al. (2017), the silicone exhibits elastic behaviour up to about 100–150% shear strain, shear strength up to 0.25 MPa, and an ultimate deformation capacity of about 200% of strain. The cyclic response of the silicone sealant is characterised by significant

stiffness degradation and progressive damage. The mean shear elastic modulus was estimated to be 0.25 MPa.

Figure 6.12: A comparison of the silicone sealant’s hysteretic response during the cyclic tests performed on concrete blocks and subassembly structure

Slika 6.12: Primerjava histereznega odziva silikona med cikličnimi preizkusi na betonskih kockah in na sestavljenem preizkušancu z dvema paneloma

Different models can be used to simulate the response of silicone sealant. The Pinching model can describe the cyclic degradation of the stiffness and strength of silicone, whereas the Elastic model assumes the completely elastic behaviour of the silicone joints.

The properties of the Pinching model were calibrated by Menichini (2019) based on experimental results from Dal Lago et al. (2017b) and are presented in Figure 6.13. The model is evaluated in Figure 6.14 by comparing the analysis with the experiment on a subassembly specimen published by Dal Lago et al. (2017b).

Dal Lago et al. (2017b) proposed that the silicone response can also be simulated with a relatively simple Elastic material model with assumed average stiffness of the silicone sealant. The stiffness of the elastic link should be evaluated by Equation 6.8, using the estimated mean initial shear modulus of silicone Gs = 0.25 MPa, width ts, depth bs and length ls of silicone strips. The equivalent stiffness of the Elastic model is plotted with the blue line in Figure 6.14 for the example of the subassembly test (Dal Lago et al., 2017b). As can be observed, the stiffness of the elastic link is somewhat larger than the initial stiffness of the silicone sealant.

𝑘_𝑒𝑞=^{𝐺𝑠∙𝑏𝑠∙𝑙𝑠}

𝑡_𝑠 (6.8)

Figure 6.13: Pinching4 model parameters (McKenna & Fenves, 2010)

Slika 6.13: Parametri materialnega modela Pinching4 (McKenna & Fenves, 2010)

Figure 6.14: A comparison of the experimental and numerical results of the silicone sealant’s hysteretic response during the subassembly test

Slika 6.14: Primerjava eksperimentalnih rezulatov in numeričnega histereznega odziva silikonskega tesnila Within the study, both models of the silicone sealant (P-Pinching and E-Elastic) were tested on the set of 15 different structures subjected to 30 accelerograms at three intensities (the selection of structures and ground motions for parametric study are presented in Section 6.1). Failure of the Pinching silicone model was defined at a shear strain of 200%. However, because the silicone

exhibits elastic behaviour up to about 100–150%, failure of the elastic link silicone model was defined at a shear strain of 150%.

Because of the higher initial stiffness and no degradation of the Elastic silicone model, the effect on the response of precast structure was larger than when using the Pinching model. The influence of the Elastic silicone model on the displacements and shear demand in the column was larger, and more failures of silicone sealant and panels were recorded. It would be more appropriate to consider the lower initial or average stiffness of the silicone sealant.

Because a relatively large scatter of silicone sealant’s mechanical properties was observed during the tests, and because the properties of silicone severely deteriorate due to climatic and ageing effects, the stiffening contribution of the silicone is not reliable and relatively limited. Because the characteristics of the silicone sealant may significantly alter due to the degradation of material (Chew, 2000), the Elastic silicone model may give too-conservative results (it overestimates the displacement and force demand at cladding connections and, consequently, also the shear demand in columns).

The silicone model P-Pinching was used in the parametric study to account for the effect of silicone sealant and to analyse the interaction of adjacent panels. The analyses considered that the silicone sealant fails during the excitation, and failure was defined at a shear strain of 200%. After the deformation capacity was exceeded, the silicone was removed from the model, and the analysis proceeded. To analyse the effect that silicone sealant has on the seismic response of the precast structure, models without the silicone sealant, N-no silicone were also included in the parametric analyses.

However, maximum displacements of structures were also relatively well estimated with the Elastic model (because of the earlier failure of silicone joints compared to the Pinching model). For those reasons and due to its simplicity, the Elastic model of silicone sealant could be suitable for use in the design.