Degree Type


Date of Award


Degree Name

Doctor of Philosophy


Electrical and Computer Engineering


Electrical Engineering

First Advisor

Nicola Bowler


Low-conductivity media are found in a vast number of applications, for example as electrical insulation or as the matrix polymer in high strength-to-weight ratio structural composites. In some applications, these materials are subjected to extreme environmental, thermal, and mechanical conditions that can affect the material's desired performance. In a more general sense, a medium may be comprised of one or more layers with unknown material properties that may affect the desired performance of the entire structure. It is often, therefore, of great import to be able to characterize the material properties of these media for the purpose of estimating their future performance in a certain application.

Low-conductivity media, or dielectrics, are poor electrical conductors and permit electromagnetic waves and static electric fields to pass through with minimal attenuation. The amount of electrical energy that may be stored (and lost) in these fields depends directly upon the material property, permittivity, which is generally complex, frequency-dependent and has a measurable effect on sensors designed to characterize dielectric media. In this work, two different types of dielectric sensors: radio frequency resonant antennas and lower-frequency (<1 MHz) capacitive sensors, are designed for permittivity characterization in their respective frequency regimes.

In the first part of this work, the capability of characterizing multilayer dielectric structures is studied using a patch antenna, a type of antenna that is primarily designed for data communications in the microwave bands but has application in the field of nondestructive evaluation as well. Each configuration of a patch antenna has a single lowest resonant (dominant mode) frequency that is dependent upon the antenna's substrate material and geometry as well as the permittivity and geometry of exterior materials. Here, an extant forward model is validated using well-characterized microwave samples and a new method of resonant frequency and quality factor determination from measured data is presented. Excellent agreement between calculated and measured values of sensor resonant frequency was obtained for the samples studied. Agreement between calculated and measured quality factor was good in some cases but incurred the particular challenge of accurately quantifying multiple contributions to loss from the sensor structure itself, which at times dominates the contribution due to the sample material.

Two later chapters describe the development of capacitive sensors to quantify the low-frequency changes in material permittivity due to environmental aging mechanisms. One embodiment involves the application of coplanar concentric interdigital electrode sensors for the purpose of investigating polymer-matrix degradation in glass-fiber composites due to isothermal aging. Samples of bismaleimide-matrix glass-fiber composites were aged at several high temperatures to induce thermal degradation and capacitive sensors were used to measure the sensor capacitance and dissipation factor, parameters that are directly proportional to the real and imaginary components of complex permittivity, respectively. It was shown that real permittivity and dissipation factor decreased with increasing aging temperature, a trend that was common to both interdigital sensor measurements and standard parallel plate electrode measurements. The second piece of work involves the development of cylindrical interdigital electrode sensors to characterize complex permittivity changes in wire insulation due to aging-related degradation. The sensor was proven effective in detecting changes in irradiated nuclear power plant wiring insulation and in aircraft wiring insulation due to liquid chemical immersion. In all three cases, the results indicate a clear correlation of measured capacitance and dissipation factor with increased degradation.


Copyright Owner

Robert Thomas Sheldon



File Format


File Size

150 pages