Detection of delamination defects in carbon fiber laminate composites using ultrasound and the Hilbert-Huang transform.

摘要

In recent years, composite materials, such as carbon fiber laminates, have become a popular choice in applications that require high strength, lightweight components. Case in point, the newly developed Boeing 787 Dreamliner will be composed of over 50% composite materials by weight, providing superior strength and fuel economy compared to current standards. However, given the anisotropic and inhomogeneous nature of these materials, traditional nondestructive evaluation techniques are not as effective in characterizing hidden defects. Therefore, it is crucial that new evaluation methods be developed to detect defects at an early stage in composite structures and prevent catastrophic failure from occurring. The research presented in this thesis aims to better understand how energy is transmitted through composite specimens so that appropriate damage detection methods can be identified and implemented. To accomplish this, ultrasound tests were performed on custom made carbon fiber laminates with embedded strips of Teflon to simulate delamination defects. Through the use of a simple ray tracing program developed in LabVIEW, it was determined that the signals received from these samples were dominated by Lamb waves. By applying the Hilbert-Huang transform to these waveforms, the complex signal could be broken down into simpler components from which the instantaneous frequency content was extracted. Using this data, various stress wave factor and frequency metrics were applied, which in general showed good correlation with the damaged samples. Specifically, the energy integral, weighted ringdown and peak frequency metrics were the most successful at detecting the delaminations. Based on these encouraging results, it is hopeful that a detection algorithm can be developed for use in a handheld scanning device for onsite component evaluation.