Conceptual Implementation of an Automated Structural Health Monitoring System

Abstract

The work presented in this paper provides a view of the conceptual implementation of an automated structural health monitoring (SHM) system. The detection of natural frequency shifts in a structure indicate the presence of damage, i.e. a modal analysis of healthy and damaged structures will indicate the relationship between damage severity and natural frequency change. This approach is utilized to analyze a composite beam vibrating in coupled bending-torsion, where modal analysis is conducted on the composite beam in order to predict the natural frequency and the associated mode shapes. Results show how the introduction of a crack changes the beam natural frequency and how the crack size and location affect that change. The task of implementing such a scheme in an automated fashion on an actual structure may prove to be an arduous one, in the sense that a database for natural frequencies, corresponding to previously studies damage types, must be available for comparison by SHM system with acquired measurements. Therefore, another approach is introduced as a viable solution for a practical implementation of an automated SHM system. Based on applying an estimation scheme, the unscented Kalman filter (UKF), this approach allows for the acquisition of a beam’s natural frequency, and does not require a previously conducted study on different damage types, it is also applicable to a practical implementation of an automated SHM system. Both approaches present several pros and cons pertaining to their implementation, some of which are discussed as a concluding remark of this study. A first prototype of the system is shown in the paper, where a simple circuit containing a microcontroller is implemented to simulate the automated SHM concept.