Rosell, Anders; Michigan State University; United States
Deng, Y.; Michigan State University; USA
Kumar, D.; Michigan State University; USA
Rosell, A.; Michigan State University; USA
Upda, L.; Michigan State University; USA
Upda, S.; Michigan State University; USA
Session: Composite Material - EC
Time: 10:00 - 10:20
Fiber reinforced polymer (FRP) composites are a superior design choice for many engineering applications due to light weight, high specific stiffness, strength, and other desirable properties. However, the use of multi-layer composite materials also brings challenges as FRPs are vulnerable to flaws during fabrication and service. Common defects associated with composites include dry spots or pores due to improper infusion or cure of resin, missing or wavy fiber bundles, and inclusions of foreign object debris. When FRP structures are in service, they are known to suffer damage from low-velocity impacts that results in matrix cracking, fiber breakage, and interlaminar delaminations. Robust NDE is essential in ensuring structural integrity of FRPs. Current industrial practice involves well-established NDE techniques, such as ultrasonic testing (UT), X-ray tomography and infrared (IR) thermography.
Recently, the feasibility of capacitive imaging (CI) has been investigated for non-contact NDE of non-conductive GFRPs. Sensing probes can be designed as pairs of co-planar capacitive electrodes to measure the local dielectric changes within composite samples. These sensors benefit from non-contact, rapid scanning and low cost but design and experimental setup need to be optimized for higher sensitivity. In this work, we show that the frequency selection for achieving high sensitivity of the open plate stray field capacitive sensor has to be made carefully and an operating frequency close to RF spectrum might be a desired selection. For optimal sensitivity we suggest selecting the frequency based on a high Q factor, which is related to the sensor design and the experimental setup. Using field excitation and signal acquisition for high frequency fields might be more challenging, and we propose tuning down the operating frequency by adding inductive parts. We investigated experimentally how high sensitivity characteristics can be tuned to lower frequencies on a prototype sensor tested on GFRP calibration samples as well as realistic defects from impact damage.