The evolutionary structural optimization (ESO) is one of the shape optimization methods of a structure, in which the low stressed part of a structure is removed from the structure step-by-step until an optimal design is obtained. Recent research has developed a bi-directional evolutionary structural optimization (BESO) algorithm than, with appropriate decision making, starts the evolutionary optimization procedure from a minimal kernel structure that connects the loading points to the mechanical constraints. In the BESO method, elements of the structure can be either be added or removed.
I have used the ESO and BESO method to determine the most effective initial section of box girder. In the shape optimization of box girder using by the ESO method, section of box girder can be successfully obtained when both the transverse and longitudinal behavior is considered, whereas it can not when only the transverse behavior is considered. The effect of all stresses to box girder in the initial stage of ESO was determined and improved using by the BESO. The initial section of box girder optimized by BESO is similar to the shape of constructed section.
This method is thought to be the useful method to determine the initial section of a structure. Furthermore, this result must be refined by the BESO considered the multiple load case in order to increase the accuracy of the optimization.