Experience with different residuals for updating and validating Finite Element Models using test data from two structural configurations

Abstract

In this paper two different residuals are investigated for updating physical mass, stiffness and additionally geometric parameters. Firstly, eigenfrequency and mode shape residuals are minimized. Next eigenfrequencies and frequency response data taken only at the resonance peaks are used, which make it possible to update also viscous modal damping parameters. The GARTEUR SM-AG19 benchmark test structure is used as an illustrative example. In the first step, the initial Finite Element (FE) model is updated with both types of residuals based on measurements of the initial structure. Then, the prediction capability of the updated models is checked using measured data from a modified structure. Results are compared, advantages and disadvantages of the techniques are pointed out, and an assessment with respect to the operational validity is given.