Simulation-Supported Validation as an Innovative Pedagogical Approach for Problem-Based Learning in the Design of a High-Performance Spoke-Type PMSM (#2734)

Authors

Quadrado, José Carlos

Abstract

This paper presents an innovative simulation-supported engineering learning methodology in which finite element method (FEM) validation is embedded in a project-based learning environment for the design of a high-performance spoke-type permanent magnet synchronous motor (PMSM). Rather than treating simulation as a final verification step, the proposed teaching strategy positions FEM as the main epistemic bridge between conceptual design, electromagnetic modeling, material selection, and the interpretation of prototype discrepancies. The learning sequence was applied to a Formula Student-oriented spoke-type interior permanent magnet machine, reconstructed and validated from design targets before being used as a pedagogical platform for iterative inquiry. Students progressed from machine geometry reconstruction and back-EMF matching to saturation mapping, inductance extraction, torque constant estimation, and manufacturing-sensitivity analysis. This workflow leads to a technically defensible comparison of candidate magnetic materials and to a simulated explanation of the prototype underperformance. The results show that the spoke topology amplifies saturation effects, making material behavior and process quality immediately visible in the learning process. Premium cobalt-iron alloys achieved torque constants near 0.33 Nm/A and torques up to 18.9 Nm at rated current, whereas standard silicon steels exhibited a 10-15% torque reduction due to earlier saturation. A faulted model representing unannealed laminations and poor stacking factor reduced the torque constant to approximately 0.09 Nm/A, reproducing the observed discrepancy and reinforcing the instructional value of simulated validation. The paper argues that this methodology strengthens both deep technical understanding and engineering judgment by linking design decisions to physically interpretable numerical evidence.