Numerical Analysis and 1D/2D Sensitivity Study for Monolithic and Laminated Structural Glass Elements under Thermal...

Glass is largely used in architectural and engineering applications (i.e., buildings and vehicles) as a structural material, especially in the form of laminated glass (LG) sections. To achieve adequate and controlled safety levels in these applications, the well-known temperature-dependent behavior of viscoelastic interlayers for LG sections should be properly accounted for during the design process. Furthermore, the materials' thermomechanical degradation with increases of temperature could severely affect the load-bearing performance of glass assemblies. In this context, uncoupled thermomechanical finite element (FE) numerical models could represent a robust tool and support for design engineers. Key input parameters and possible limits of the FE method, however, should be properly calibrated and assessed, so as to enable reliable estimations for the real behavior of glazing systems. In this paper, FE simulations are proposed for monolithic (MG) and LG specimens under radiant heating, based on one-dimensional (1D) and two-dimensional (2D) models. A special attention is focused on thermal effects, being representative of the first step for conventional uncoupled, thermomechanical analyses. Based on experimental results available in the literature, FE parametric studies are discussed, giving evidence of limits and issues due to several modeling assumptions. In particular, careful consideration is paid for various thermal material properties (conductivity, specific heat) and thermal boundaries (conductivity, emissivity), but also for other influencing parameters like the geometrical features of samples (thickness tolerances, cross-sectional properties, etc.), the composition of LG sections (interlayer type, thickness), the loading pattern (heat transfer distribution) and the presence of additional mechanical restraints (i.e., supports of different materials). Comparative FE results are hence critically discussed, highlighting the major effects of such influencing parameters.