Form-finding and Buckling Optimization of Grid Shells using genetic Algorithms

We present a strategy for the design of gridshells where form-found structures are optimised for buckling resistance. A genetic algorithm is employed for the initialisation of pre-stress forces required in form-finding using dynamic relaxation. Dynamic relaxation takes this initial prestress, a flat grid, as well as self-weight and nodal loads to calculate a static equilibrium. The structure is then analysed for the estimation of the critical buckling load. Different boundary conditions, structural parameters and typology of connections are compared, including a gridshells with triangular and quadrangular patterns. Optimised structures are measured against the trivial solution, which is a structure where dynamic relaxation is initialised with uniform pre-stress. Our results show the proposed strategy can successfully form find gridshells with improved buckling performance.