Damage detection in composite and plastic thin-wall beams by operational modal analysis: An experimental assessment

Abstract

The detection and localization of different damage features in thin-wall beam composite and plastic beams using Operational Modal Analysis (OMA) has been demonstrated experimentally. The detection of small damage features using modal analysis techniques is an emerging field, with few experimental OMA-based assessments having been reported so far. The proposed method is based on OMA combined with Stochastic Subspace Identification (SSI) and the enhancement of damage features by Continuous Wavelet Transforms (CWT). A composite thin-wall beam (CTWB) structure in two measurement configurations and a PVC tube in a free-free configuration have been tested. Damage features detected include extra masses attached to the beam, with a range from 9.5 % to 14.0 % of the beam mass, and small cracks perpendicular to the beam axis with lengths of about 4 % of the perimeter of the cross section. Calibration curves relating the strength of the damage signal with the weight of the attached masses have been constructed. Two simultaneous cracks or two masses could be detected as well. The quantification and localization of damage feature along the beam was possible through the use of Gaussian fit surface applied to damage maps obtained with the CWT technique. The width of the Gaussian fit curve was of the order of the distance between accelerometers, but the accuracy, estimated to be around 3 % of the beam length, was found to have sub-grid resolution. The proposed method was shown to work reliably with a relatively coarse measurement grid, potentially allowing for cost-effective Structural Health Monitoring (SHM) approaches.