PROGRAM | Electrical & Computer Engineering

By: Mark Colgan Chair: Mark Mirotznik

ABSTRACT

The desire to make devices smaller, better, faster has increased the need for novel ways to reduce the size of components such as radiofrequency antennas. Instead of striving for shrinking traditional designs, what if we flipped the design methodolgy to use a space-filling approach in the irregularly shaped volume available after the other design components are included? Leveraging available computational modeling tools and additive manufacturing processes brings together strengths of two technology areas to provide a new way of providing solutions to this problem.

In this dissertation, I present the development of a custom software solution which automates the design, optimization, and analysis of antennas into a unified workflow. Wire grid antennas are defined, and I detail the approach to generating an array of possible electrical segments within an arbitrary volume and how to use evolutionary and other algorithms to perform an optimization on the multi-variable space.

Concurrently, advances in additive manufacturing and rapid commercialization have made processes available to the university and hobbyist communities. Additive manufacturing allows for high-fidelity reproduction of complex computer designs which may otherwise be difficult or impossible to fabricate. Leveraging this capability, I describe an approach to constructing a novel wire grid antenna from computer design and optimization to fabrication and laboratory analysis.

I present how small changes in the design can impact overall performance of the antenna system, opening additional areas for investigation. Lastly, I show measurement results comparing fabricated examples to the computational simulation of optimized designs.

Back >