College of Engineering, Design & Computing Events

Mark Travers
Graduate Research Assistant
Electrical Engineering
University of Colorado Denver

Abstract
Many engineering applications rely on an accurate knowledge of the behavior of ferromagnetic materials. Such applications include data storage technologies, energy conversion systems, or magnetic object detection and/or proximity sensors such as those found in robotics and automation systems.

The past few hundred years have seen several attempts to accurately model the magnetic behavior of ferromagnetic materials, most notably the Preisach and Jiles-Atherton models, all of which possess limited accuracy in various states of magnetization. Within the last several years, Dr. Carl Schneider of the US Naval Academy has developed and published a new theory, known as the cooperative exponential model, which has initially demonstrated a higher degree of accuracy in all states of magnetization. Much work has yet to be done to experimentally verify this model.

This seminar will present two phenomena which make direct, experimental verification of any ferromagnetic hysteresis model impossible. These are the demagnetizing effect and non-uniformity in magnetization, both of which are heavily geometry-dependent. Then, a method will be discussed which allows geometry-dependent experimental data to be compared to geometry-independent model predictions using a non-linear magnetostatic EM solver in a neural network configuration. Results are shown which further confirm the cooperative exponential model.

Bio. 
Travers completed his bachelor of science in electrical engineering in 2018 at the University of Colorado at Denver and is expected to complete his master of science in December. Since the spring of 2018, Travers has conducted research in the area of ferromagnetic materials under the direction of Stephen Gedney at the CU Denver Magnetics lab.

Email Address: mark.2.travers@ucdenver.edu