Dielectric and electromagnetic behavior of carbon fiber materials
MetadataShow full item record
This dissertation addresses the science behind the dielectric and electromagnetic behavior of carbon fiber materials, in addition to the application of the science in nondestructive evaluation (NDE) and electromagnetic interference (EMI) shielding. The dielectric behavior pertains to the electric permittivity (a material property) in the low-frequency regime (2 kHz), as is relevant to electrical and electrochemical applications that use carbons as conductive materials. In contrast, the electromagnetic behavior pertains to the absorption loss per unit thickness (the same as the linear absorption coefficient except for a numerical factor) and the reflection loss in the radio wave regime (around 1 GHz), as is relevant to EMI shielding and low observability (Stealth). In relation to the dielectric behavior, this dissertation provides the first determination of the electric permittivity of carbon fibers in the low-frequency regime. The relative permittivity at 2 KHz in the axial direction is 4960 ± 662 and 3961 ± 450 for the P-100 (more graphitic) and P-25 fibers (less graphitic), respectively. These values are much higher than those previously reported for discontinuous carbons, but they are lower than those previously reported for steels. The continuity of the carbon fibers enables the charge carriers to move by relatively long distances during polarization, thereby resulting in permittivity values that are much higher than those of discontinuous carbons. Thus, the defects associated with a relatively low degree of graphitization hinder the carrier movement, thereby decreasing the permittivity. The conductivity ratio of the two types of fiber (6.7) is higher than the corresponding permittivity ratio (1.3). Hence, the conductivity is negatively affected by these defects much more than the permittivity. The relative permittivity of the composite at 2 kHz is 2156 ± 509 and 1640 ± 328 for the longitudinal (fiber) and transverse directions, respectively. The corresponding relative permittivity of the fiber (PAN-based) is 4352 ± 510 and 3310 ± 696 for the longitudinal and transverse directions, respectively. The longitudinal value for the fiber is comparable to those obtained for the abovementioned mesophase-pitch-based fibers. The permittivity anisotropy of the PAN-based carbon fiber is 1.3, which is much lower than the resistivity anisotropy of 1500. The dielectric behavior of carbon fiber enables the capacitance of a continuous carbon fiber polymer-matrix composite to change upon damage, thereby providing capacitance-based self-sensing. This dissertation provides the first report of this self-sensing. The conductivity of carbon fiber helps the current spreading, so that the sensing can occur at substantial distances from the electrodes. The two aluminum foil electrodes for capacitance measurement can be coplanar or sandwiching. The presence of a dielectric film (double-sided adhesive tape) between the electrode and the composite surface is necessary, due to the conductivity of the composite and the fact that an LCR meter is not designed to measure the capacitance of a conductor. In practical implementation, the paint on the composite structure can be a part of the dielectric film. The method is simpler for implementation than the previously reported resistance-based self-sensing of carbon fiber composites. Continuous carbon fiber polymer-matrix composites have been previously shown to be effective for EMI shielding. This dissertation provides the first study of the effect of the fiber lay-up configuration on the shielding effectiveness. Under unpolarized normal-incidence radiation (1 GHz), the performance indicated by the absorption loss per unit thickness is superior for the crossply composite than the unidirectional composite with the same number of laminae, with the ratio of ~4 being in accordance with electromagnetic theory. This dissertation also addresses explicitly the linear absorption coefficient α of carbon materials at around 1 GHz and shows that α decreases with increasing carbon thickness in accordance with the Skin Effect. On the other hand, α is essentially independent of the carbon microstructure, due to the large value of the radio-wave wavelength. The quality of the electrical contact (as controlled by the fastening torque) between the carbon specimen and the EMI testing fixture is shown in this work to affect the testing results, such that the effect depends on the mechanical behavior of the carbon specimen. For carbon fiber mat, increasing the torque causes decrease in the measured shielding effectiveness. In contrast, for carbon nanofiber mat and flexible graphite, increasing the torque causes increase in the measured shielding effectiveness.