On the design criteria of large insulating glass structures in cruise ships

Loading...
Thumbnail Image
Journal Title
Journal ISSN
Volume Title
Insinööritieteiden korkeakoulu | Master's thesis
Date
2020-10-19
Department
Major/Subject
Mcode
Degree programme
Nordic Master Programme in Maritime Engineering
Language
en
Pages
63+20
Series
Abstract
Insulating glass units (IGUs) consist of two glass panes separated by a hermetically sealed gas. IGUs are used in ships to save energy by reducing heat losses. From a structural point of view, research has shown that the gas transfers loads between the glass panes, which increases the IGU’s structural performance. This load sharing effect is not yet considered in ship classification rules by different classification societies, which may result in thicker glass constructions than necessary. Thus, this thesis studies the load sharing effect for evaluating the sufficiency of the current design criteria for glass pane thickness determination. This thesis presents a Finite Element (FE) model for calculating the response of a rectangular IGU using Ansys. The glass is modelled with structural shell elements (SHELL181) and the gas is modelled with hydrostatic fluid elements (HSFLD242). The former is based on Mindlin-Reissner plate theory, while the latter is based on Ideal Gas Law. The interaction between the glass panes is determined by considering the volume change of the gas. The response is first calculated on a linear basis, considering small loads that are frequently applied to ship structures. The linear FE model is validated by Betti’s Analytical Method. However, occasionally the windows are subjected to increased loads that result in large deflections of thin glass plates, i.e. geometrically nonlinear behavior is observed. The linear FE model is extended for nonlinear analysis. The model is validated by experimental results from open scientific literature. The presented nonlinear FE model is used to calculate the minimum required thickness of glass panes in an IGU, with respect to standard stress limit. The same thicknesses are also calculated according to classification rules by Lloyd’s Register. The obtained thicknesses are compared. The results indicate that the thickness of the glass panes can be reduced up to 52 percent when the load sharing effect is considered. The reduction is based on maximum principal stress reduction that is a benefit of both panes carrying the load instead of one. This reduction is the largest when the geometric nonlinearities are apparent. Based on the results, this study concludes that the current glass pane thickness determination criteria by the classification societies are insufficient. Cruise ships have hundreds of square meters of IGUs and therefore the load sharing should be considered, to reduce weight or to allow for larger windows. Furthermore, the findings of this study are important as the classification rules are based on linear assumptions and very little information is available in the open scientific literature about IGU FE modelling. Further research is needed to obtain more experimental results for IGUs with different boundary conditions as they appeared to have significant effect on the performed analyses.
Description
Supervisor
Romanoff, Jani
Thesis advisor
Niemelä, Ari
Laakso, Aleksi
Keywords
design criteria, insulating glass unit, load sharing, hydrostatic fluid element
Other note
Citation