Calibration of hyperelastic material models for structural silicone and hybrid polymer adhesives for the application of bonded glass. (10th September 2020)

• Record Type: Journal Article
• Authors: Van Lancker, Bert; De Corte, Wouter; Belis, Jan
• Is Part Of: Construction & building materials. Volume 254(2020)
• Publication Date: 2020-09-10
• Subjects: Bonded glass -- Two-component structural silicone -- Two-component hybrid polymer adhesive -- Hyperelastic material models -- Mullins effect; Building materials -- Periodicals; 624.18
• DOI: 10.1016/j.conbuildmat.2020.119204
• Languages: English
• ISSNs: 0950-0618

Abstract

Highlights: The material behaviour of three silicones and a hybrid polymer is investigated. Multiple deformation states, and both monotonic and cyclic loading are studied. Constitutive models are calibrated for individual and combined deformation states. Material models for stress-softening are calibrated for several deformation states. Calibrated material parameters are available for implementation in FE-software. Abstract: To enable the introduction of new developments and concepts of bonded glass in façades, finite element modelling is a frequently used method. The implementation of the complex material behaviour of suitable adhesives into such software, however, can be labour-intensive. Usually, hyperelastic material models need to be calibrated, i.e. theoretical mathematical expressions are fitted to experimentally obtained data. In this research, three two-component structural silicones and one two-component hybrid polymer adhesive were selected. Experimental tests on H-specimens in correspondence with ETAG 002 were performed for tension, compression and simple shear at different haul-off speeds of the testing machine, more specifically 1 mm/min, 5 mm/min, 50 mm/min and 500 mm/min. The stress–strain response of the adhesives altered with different test speeds. The material response under cyclic loading regime was investigated as well. These tests confirmed the presence of strain softening, also referred to as the Mullins effect, in all selected adhesives. Next, twenty isochoric hyperelastic material models were calibrated with the experimental data of the monotonic load tests at a speed of 5 mm/min for single, dual and triple deformation states. A theoretical estimator of the relative quality of the material models, more specifically the Akaike Information Criterion, allowed for the selection of the models that fit the experimental data the best. For all investigated adhesives, the polynomial based hyperelastic material models provided the best fit. This paper presents the results of the experimental programme that enable researchers to compare numerical and analytical analyses with these results. Designers and engineers can compare the structural performance of the adhesives in correspondence with ETAG 002 and assess the importance of cyclic loading regimes in the design of bonded glass. The calibrated hyperelastic material models allow engineers to numerically implement hyperelastic material behaviour of the studied adhesives in correspondence with ETAG 002. The best-performing material model for the considered application needs to be selected and the numerical model needs to be validated using the results of additional experiments on the considered bonded glass application. Especially the calibrated material models for multiple deformation states are valuable to model complex stress distributions in bonded glass applications subjected to multiple deformation states.