Development of a High-Power Density Motor
-(Fourth Report)-

Naoki Itasaka･Michiharu Kawano･Hisayuki Kabashima･Kentaro Nomura･Kazunori Hirabayashi (MCF Electric Drive)

Based on issues identified in previous studies, this paper presents design improvements for a high power density motor. High-speed operation, electromagnetic optimization, and cooling enhancement were applied. Design conditions considering mechanical strength and loss were clarified, and electromagnetic and thermal characteristics were improved through redesign of flux, current distribution, and cooling paths. Analytical methods evaluated the influence of each factor on power density. Prototype testing confirmed that power density exceeded 10 kW/kg, demonstrating the effectiveness of the proposed design approach.

Development of a High-Power Density Motor
-(Fifth Report)-

Hisayuki Kabashima･Naoki Itasaka･Michiharu Kawano･Kentaro Nomura･Kazunori Hirabayashi (MCF Electric Drive)

This paper presents a detailed evaluation of the prototype motor designed in Paper 4. Output characteristics, efficiency, temperature distribution, and loss behavior at high-speed operation were investigated. Comparison with analytical results enabled quantitative evaluation of the contributions of electromagnetic design and cooling structure. In addition, interactions among design factors were analyzed to identify dominant factors governing performance. Finally, design constraints and remaining challenges were clarified, providing insights for further improvement of high power density motor design.

Development of Miniaturization, High Efficiency, and High Torque Density Technologies in e-Axle

Yohei Takahashi･Katsura Uesugi･Takahiko Oishi･Eiichi Shiomitsu (Meidensha)

A compact e-Axle with a low-profile, lightweight structure and high torque density was developed for mini and A-segment vehicles. The gear train adopts a two-shaft parallel configuration, and the inverter is arranged in the vacant space alongside the motor to minimize the unit dimensions in all three directions. High torque density was achieved by increasing the motor winding fill factor and employing a dual-V magnet configuration, while high efficiency was attained by increasing the switching speed and reducing oil losses.

Development of an Abnormality Detection and Recovery Method for Position Sensorless Control of a Generator Motor

Kenichi Mori･Tetsurou Kojima･Akira Sawada･Takashi Nakajima (Nissan Motor)

When applying position sensorless control to a generator motor in a 100% electric-drive hybrid vehicle, handling transient estimation abnormalities is an important issue. This study proposes an abnormality detection method based on the rate of change of estimated speed and current, along with a recovery method combining gate-off and three-phase short-circuit control. The method achieves both overcurrent suppression and vibration reduction while enabling restart. Its effectiveness is verified through vehicle experiments.

Basic Verification of Hybrid-Excitation Memory Motor with Field Winding Connected Between Neutral Points of Dual Three-Phase Windings
-(Part 4)-

Hiroaki Makino･Ryosuke Saito･Makoto Matsushita･Katsutoku Takeuchi (Toshiba)

Variable flux technology has attracted attention as a means of improving motor efficiency. The authors propose a hybrid-excitation memory motor with field winding connected between neutral points of dual three-phase windings, aiming to achieve superior variable flux performance while suppressing an increase in inverter capacity. This paper presents the results of verifying the efficiency improvement effect using a miniaturized prototype model.

Development of high-speed prediction system using surrogate models for motor cooling CAE

Shintaro Nakano (Toyota Motor)

This paper proposes a method for accelerating particle-based motor cooling CAE using a surrogate model. The method consists of a preprocessing step that converts CAE results to a Cartesian grid, followed by prediction of particle behavior using a geodesic convolutional neural network (GCNN) and subsequent calculation of the wetted area. The results show that the computation time is reduced to approximately one minute compared with conventional particle-based analysis, while achieving a wetted-area prediction accuracy with a mean absolute percentage error (MAPE) of 5%.