Experimental and numerical study on the free vibrations of soda–lime–silicate float glass panes with three different thicknesses
More details
Hide details
1
West Pomeranian University of Technology in Szczecin
Faculty of Civil and Environmental Engineering
Department of Theory of Structures
al. Piastów 50a, 70-311 Szczecin, Poland
2
Silesian University of Technology
Faculty of Civil Engineering
Department of Structural Engineering
ul. Akademicka 2A, 44-100 Gliwice, Poland
These authors had equal contribution to this work
Corresponding author
Małgorzata Abramowicz
West Pomeranian University of Technology in Szczecin
Faculty of Civil and Environmental Engineering
Department of Theory of Structures
al. Piastów 50a, 70-311 Szczecin, Poland
KEYWORDS
TOPICS
ABSTRACT
This study investigates the free vibration behaviour of soda–lime–silicate float glass panes with different thicknesses through combined experimental and numerical analyses. Experimental modal analysis was conducted on three glass plates with dimensions of 400 × 800 mm and nominal thicknesses of 4, 8 and 12 mm. The tests were performed using an impact hammer and triaxial accelerometers positioned at 27 measurement points, enabling the identification of natural frequencies and corresponding mode shapes. To minimize the influence of boundary conditions, the panes were elastically supported at the zero-deformation points of the first flexural mode. The experimental results were compared with numerical simulations performed in ABAQUS using two finite element formulations: quadrilateral shell elements and 3D solid elements with incompatible modes. A systematic mesh-refinement study was carried out to determine the relationship between element size, computational time and result accuracy. The analysis showed that for both element types, the frequency results converge when the finite element size is approximately equal to the pane thickness. Modal Assurance Criterion (MAC) analysis confirmed a high correlation between experimental and numerical mode shapes, with diagonal MAC values approaching unity. The differences between experimental and numerical results decreased with increasing glass thickness, reaching below 2% for the thickest plate. Overall, shell models provided accuracy comparable to solid models while reducing computational time by an order of magnitude. Therefore, shell elements are recommended as an efficient and reliable tool for the modal analysis of glass panes.