抄録
This study proposes a finite element updating method using multiobjective optimization to consider multiple experimental conditions for estimating parameters. The method aims to minimize the root-mean-square (RMS) error of the deformation shape between the finite element analysis and experimental results. The proposed method is applied to the bread board model (BBM) of a tensionstabilized space reflector consisting of hoop cables and radial ribs, in which the rib is deformed into the prescribed shape by the cable tensions generated on deployment. The design requirement is to deform the rib into the prescribed shape by applying appropriate tension loads to the radial and hoop cables. Under actual conditions, the deformation shape deviates from the ideal shape because of uncertainties. Therefore, it is necessary to estimate the physical parameters with high accuracy, through a geometrically nonlinear finite element analysis, in order to investigate their effect on the deformation shape. To efficiently estimate the physical parameters, the satisficing trade-offmethod (STOM) is adopted as the multiobjective optimization method. Through numerical examples, the validity of the proposed method is demonstrated by comparing the analytical deformation shapes with experimental results.
