and minimal wall erosion due to a unique periodic permanent magnet system. A laboratory HEMPT was built and donated by JPL for testing at Cal Poly. Previous work was done to characterize the performance of this thruster and it was found to exhibit a large plume divergence, resulting in decreased thrust and specific impulse. This thesis explores the design and application of a magnetic shield to modify the thruster's magnetic field to force more ion current towards the centerline. A previous Cal Poly thesis explored the same concept, and that work is continued and furthered here. The previous thesis tested a shield which increased centerline current but decreased performance. A new shield design which should avoid this performance decrease is studied here.Magnetic modelling of the thruster was performed using COMSOL. This model was verified using guassmeters to measure the field strength at many discrete points within and near the HEMPT, with a focus on the ionization channel and exit plane. A shield design which should significantly reduce the radial field strength at the exit plane without affecting the ionization channel field was modelled and implemented. The HEMPT was tested in a vacuum chamber with and without the shield to characterize any change to performance characteristics. Data were collected using a nude Faraday probe and retarding potential analyzer. The data show a significant increase in centerline current with the application of the shield, but due to RPA malfunction and thruster failure the actual change in performance could not be concluded.v The unshielded HEMPT was characterized, however, and was found to produce 12.1 ± 1.3 mN of thrust with a specific impulse of 1361 ± 147s. The thruster operated with a total efficiency of 10.63 ± 3.66%, an efficiency much lower than expected. A large contributor to this low efficiency is likely the use of argon in place of xenon. Its lower mass and higher ionization energy make it a less efficient propellant choice. Further, the thruster is prone to overheating, indicating that significant thermal losses are present in this design.vi ACKNOWLEDGMENTS I would like to recognize here all of the people who made this thesis possible, and there is no one who has been more instrumental in its completion than Dr. Dan Goebel of JPL. Dan is a seemingly endless source of knowledge on all topics relating to EP, from theory to testing. I had no experience and only a very basic understanding of electric propulsion concepts at the outset of this project. Dan was always willing to share his vast experience, whether it was through detailed and quick email exchanges or facility visits to teach me about EP measurement and diagnose problems with my experimental setup. He was available for any question I had, no matter how small. He also generously loaned me necessary materials and equipment, without which this thesis would not have been possible. I am extremely grateful for all of his help.I would also like to extend a huge thank you to my thesis advisor, Dr. Kira Abercromby.Sh...