Application of the Random Decrement Technique to a Nonlinear Dynamic System

Abstract

The majority of structural health monitoring techniques are based on detection of significant variation in vibration parameters. The random decrement technique allows estimation of modal parameters without having direct measurements of the inputs. The technique has been successfully applied for modal analysis and health monitoring of various structures. However, application of the technique requires the assumption of linearity of the system under consideration, which is not always true in reality. In fact, real structural deterioration usually causes nonlinear responses. In this work, random decrement signatures are obtained from systems with dierent types of nonlinearities. It is found that unlike in the linear case, the signatures from nonlinear systems are not equivalent to free responses. Variation of identified modal parameters is investigated with respect to the types and strength of nonlinearities. In addition statistical features of the signatures are considered for use in detection, location, and classification of structural damage.