M.Sc Thesis
M.Sc Student Lovinger Zev Bonded Tile Wall Systems - Failure Criteria and a Two Dimensional (2D) Analysis Department of Civil and Environmental Engineering Professor Emeritus Yeoshua Frostig

Abstract

The research deals with stress analysis and failure criteria of Bonded Tile-Wall systems, using 1D and 2D analytical models.
This work defines the failure mechanism of a bonded Tile-Wall system as a result of propagation of cracks, which are initiated, at the edges of the tiles. A failure criterion based on a fracture mechanics approach is presented, replacing the classical allowable stress concept. In order to establish this approach an analytical and experimental study has been conducted. The stress and displacement fields  have been determined using a one-dimensional high order model and by calculation of the J-Integral, a characteristic critical value of Energy Release Rate has been determined for some typical bonded Tile-Wall system.
Furthermore, in order to examine the veracity and range of validity of the one-dimensional model, which has been used for the analysis of an actual 2D structure with finite dimensions both longitudinally, and transversely, a two dimensional analysis has been conducted, formulating a two-dimensional high-order model for the analysis of Bonded Tile-Wall Systems. The model presented is applicable to any type of boundary conditions that can be applied separately on the wall, the tile and the glue layer (mortar). The governing equations and the required boundary conditions are derived explicitly through variational principals, yielding a system of eight partial differential equations (PDEs) whose order is 20. The differential equations system is numerically solved using the Extended Kantorovich Method.
The two-dimensional analysis reveals that the peeling stresses, which reach their maximum value at the end of the tile run, using a one-dimensional analysis, varies also in the transverse direction from a maximum value in the middle of the tiles boundary, descending towards the corners. Furthermore, the nature of variation along the boundary strongly depends on the type of loading and the boundary conditions of the wall. For completeness, the analytical two dimensional model has been verified through comparison with a three-dimensional FE model. The comparison with the FE results was in good agreement.