Frequency domain test-analysis correlation in the presence of uncertainty

Abstract

The aerospace community traditionally relies on modal-based test-analysis correlation metrics, such as natural frequency matching and cross-orthogonality, but these techniques are known to breakdown in the presence of high modal density. This paper shows how the frequency domain orthogonality metric can parallel validation criteria that are currently used with modal analysis, while avoiding the complications and pitfalls associated with modal analysis beyond the low-frequency range. This work also examines a modification of the correlation criteria in the presence of uncertainty contained in an ensemble of test data. A 6-DOF spring-mass system and a General Purpose Spacecraft example are included to illustrate the use of the correlation criteria. ©2010 Society for Experimental Mechanics Inc.