Evaluation of damage indexes for structural health monitoring in nonlinear structures

Abstract

Nonlinear behavior can be important for the monitoring of the structural state of mechanical systems since it can be mistakenly classified as a damage. However, nonlinear tools are still not consolidated and need further research. Discrete-time Volterra series is an interesting mathematical framework to deal with nonlinear dynamics since it is a clear generalization of the linear convolution. This type of model can be applied in order to detect structural variations in nonlinear systems. In the present proposal the student will spend 5 months in the Structural Health Monitoring Laboratory of the University of California San Diego under the supervision of Prof. Michael D. Todd. During this period the proposed damage index will be investigated under the perspective of statistical pattern recognition in order to obtain a reliable method of damage detection in nonlinear structures. (AU)