| In this dissertation we introduce a method for the shape optimization of trusses, as well as a design method for the generation of trusses.;Chapter two introduces a novel approach to truss structure shape design, drawing inspiration from biological processes, that treats each joint or member of a truss as a cell and the entire truss as a multicellular organism. The approach mimics plant tropisms in which, through cellular growth, cells change their size and shape in response to scripted gene commands. Examples illustrate the effectiveness of the approach; it is shown to produce arching and double-arching of pinnedpinned trusses, bulging of cantilever trusses and a plant-like tropic response.;Chapter three introduces a new engineering design algorithm, called the multicellular organism design (MOD) algorithm. The MOD algorithm is well-suited to the early design phase during which there is an interest in generating different topologies. The MOD algorithm switches between topology modification and design parameter optimization to generate feasible designs. The topology modification either adds or subtracts nodes and edges and the design parameter optimization is performed by any suitable optimization method. The MOD algorithm is applied to truss design and several illustrations demonstrate its ability to readily generate common trusses, an optimal truss, and interesting solutions with little computational effort.;Finally, in chapter four we extend the generative procedure outlined in chapter three to produce space trusses. This requires the development of fast optimization routines, parametric representations of truss structures, and refinements of the MOD algorithm. Examples illustrate its ability to produce a tower, a bridge, and explore multiple objectives. A final example demonstrates two trusses generated via user direction. |
