Test/Analysis Correlation Using Design Sensitivity and Optimization

A procedure is described for rapidly correlating finite element models to modal test results using a combination of design sensitivity and optimization methods. The design sensitivity coefficients estimate the effect of many possible model changes or design variables. Optimization methods are used to select optimum changes which minimize the error between test and analysis results with minimum change to the finite element model. Weighting factors provide flexibility for correlating selected modes or design variables. The procedure is demonstrated using a spacecraft model with up to 50% error in initial properties.