College - Author 1
College of Engineering
Department - Author 1
Materials Engineering Department
Degree Name - Author 1
BS in Materials Engineering
College - Author 2
College of Engineering
Department - Author 2
Materials Engineering Department
Degree - Author 2
BS in Materials Engineering
Date
7-2025
Primary Advisor
Jean Lee, College of Engineering, Materials Engineering Department
Abstract/Summary
This multi-year project focuses on the design, modeling, and testing of novel inductors for electric vehicles (EVs) using iron nitride (IN) soft magnetic composites (SMCs). This work is a continuation of previous research, with the overarching goal of enhancing EV efficiency by developing inductor cores with improved magnetic properties, leveraging IN's competitive magnetic characteristics.
This year, the project concentrated on refining fabrication processes to optimize relative permeability (μr) and effective IN concentration. Investigations included various mixing vessels, the application of vibration during powder settling, and different pressing methods and pressures across initial IN concentrations of 75, 80, and 85 vol%. Inductance and AC resistance measurements were performed, and the COMSOL™ Multiphysics model underwent significant refinement.
Key findings indicate that despite these processing modifications, the project did not consistently achieve higher effective IN concentrations or relative permeabilities than previous work, with effective IN concentration generally plateauing around 65 vol%. The 75 vol% initial IN concentration group, however, showed the highest meanμrand theoretical density. Cold mixing extended the epoxy's working time but resulted in reduced magnetic performance. Vibration-assisted settling generally led to higher effective IN concentrations compared to hand tamping, and double pressing at elevated pressures (190 MPa followed by 120 MPa) yielded the highest mean relative permeability and percent of theoretical density. Increasing the initial IN concentration beyond 75 vol% did not translate to a higher effective IN concentration. An alternative approach using 3D-printed shells for a theoretical 100 vol% IN core proved less effective, exhibiting reduced magnetic performance due to low packing density and increased eddy current losses in the absence of an insulating binder.
A significant advancement was the refinement of the COMSOL™ model. The corrected model now more accurately accounts for realistic winding geometries and addresses previously identified artificial air gaps, greatly improving its ability to predict measured inductance. This enhanced model is validated for future use in simulating and validating experimental results for toroidal inductors.
URL: https://digitalcommons.calpoly.edu/matesp/286