Semi-active Damage Identification for a Composite Structural Missile Component using Minimal Passive Sensing with Data-driven Models

Abstract

Cost, complexity, and weight restrictions often constrain industrial structural health monitoring (SHM) systems to the use of one sensor. Often, this one sensor is passive, i.e., it only measures the structural response. Therefore, new SHM techniques that utilize only one passive sensor along with advanced data interrogation methods are needed. A semi-active damage detection technique is described in this study that utilizes passively estimated forces and response measurements to update data-driven frequency-response-function (FRF) models of a filament-wound missile casing that is used as a test structure. During a damaging impact, the structural response is influenced by both the healthy and damaged structure properties. As a result, the magnitudes of the estimated FRFs can exhibit a splitting phenomenon at the natural frequencies of the healthy and damaged structure, and this phenomenon is evident in the experimental data. The updated FRFs are used with an active damage detection technique to quantify the damage. The normalized differences in the damaged and healthy FRFs and the shift of the natural frequencies of the structure quantify damage in this study. The shift of the natural frequencies more accurately quantifies the damage caused by four experimental impacts than the normalized differences in the FRFs.