September_AMP_Digital

A D V A N C E D M A T E R I A L S & P R O C E S S E S | S E P T E M B E R 2 0 2 0 1 8 integrity can become compromised due to manufacturing defects, operational fatigue, or impact damage, necessitat- ing component repair or replacement. Defects, often present at the time of fabrication, can grow during operation and are not easily detected visually or with other volumetric techniques such as ultrasound until they become irrep- arable. The ability of microwaves to penetrate nonmetallic materials could provide a new way to inspect blades and find these underlying damages or defects before it is too late. A laboratory study was conduct- ed using pedigreed samples from the Sandia National Laboratory wind tur- bine blade archive. These samples are representative of a portion of wind tur- bine blades and contain manufactured defects representing common issues such as delaminations, dry regions (i.e., low resin percentage), and marcelling (Fig. 3). Generally, the worst defect is marcelling, or out-of-plane wavy lam- inates, because it remains visually hidden, is difficult to detect with con- ventional NDT, and can easily result in cracking that could lead to blade failure. The microwave inspection table at EPRI has been used to inspect multiple ma- terial samples and represents the latest in microwave inspection capability [8,9] . Fiberglass parts are particularly difficult for most inspection techniques due to glass components in the resin matrix. The glass fibers act like reflec- tors for ultrasound so that these parts have a relatively high signal-to-noise ratio. In the case of microwave imag- ing, the glass reflects and refracts the electromagnetic energy of microwaves, but to a much lesser extent than sound energy. As an analogy, consider how light readily penetrates a glass window while sound does not. A further poten- tial advantage of MW in wind turbine applications is that the inspection de- vice does not need to be in close con- tact with the surface, which is required by techniques such as ultrasonic testing (UT). This makes field deployment us- ing drones or other robotics attractive in situations where access by humans comprises a significant portion of in- spection time and costs. A unique development in this study was the use of a multi-frequen- cy approach to provide additional in- formation and flexibility over previous single frequency microwave devices. Figure 4 demonstrates the capabilities of a multi-frequency scan image. The mock-up wind turbine blade segment was scanned over a frequency range of 8 to 14 GHz. Data was captured at each frequency band in 201 discrete frequen- cy increments. Images are generated by displaying either the real, imaginary, magnitude, or phase of the reflected energy data at each data point. A range of frequencies can be selected for dis- play by manipulating the gate in the “A” scan, shown in the upper left-hand corner of the display. The software dis- plays of the MW data are similar in ar- rangement to traditional ultrasonic data collected in encoded inspections, with data displayed in quadrants. Read- ing the image clockwise from upper left, the images in Fig. 4 are (a) S11 ver- sus frequency (scan data), (b) “B” verti- cal image (c) “C” scan data image, and (d) “B” horizontal image. The images shown were compared to the blade mock-up design and the Fig. 3 — Wind turbine blade sample with marcelling. Fig. 4 — MW data image of wind turbine mock-up sample: Frequency and time related responses shown in A, B and C scan format. (a) (b) (d) (c)