A linear cellular automation technique for predicting dynamic failure mode of a single-layer shell

This paper presents a linear cellular automation (LCA) method for predicting the dynamic failure (DF) mode of both single-layer latticed shell and single-layer cylindrical latticed shell subjected to ground motions. The LCA model of the shell obtains the state values of cells/nodes including the nodal displacements state value and the nodal domain logarithmic strain energy density (NDLSED) state value through its finite element analysis (FEA). Meanwhile, the concepts of nodal domain and nodal domain similarity are derived based on the qualitative analysis of shells. Then, similar nodal domains between two shells are matched through the proposed criterion. Finally, the DF mode of an object shell is mapped using the criterion for projecting the formative values of a base shell to similar nodal domains in the object shell. Case studies show that the LCA method could be used for predicting the DF mode of an object shell. Consequently, the LCA method would explore an LCA application in analyzing shells, which costs much less time than the FEA method for calculating the DF shell mode.