Development of an Efficient Improvement Method for Thermal Boundary Surrogate Models using Fine-Tuning

Tomofumi Shimokawa･Mitsunori Kamimura･Norio Kawaguchi･Masakatsu Takahashi･Takahiro Mochihara (Toyota Motor)

Developed a surrogate model combining multiple regression and deep learning. Observed a drop in inference accuracy in new design areas. Used fine-tuning to improve accuracy and reduce learning costs, enabling better adaptation to new designs.

“Absorption” and “Bouncing” of Oil Mist Droplets Impacting on Oil Film

Hayato Ito･Koichi Nishino･Misa Ishimura (Yokohama National University)･Akira Nakashima･Masato Nishimura･Yusuke Kimura･Kazuhiro Nakamura (Mazda)

Sophisticated design of oil mist separators suitable for blowby gas of IC engine needs deep understanding of wall-adhesion behaviors of individual oil droplets impacting on the wall. The present study focuses on absorption/bouncing behaviors of oil mist droplets in a simplified oil mist separator and clarifies the absorption/bouncing transition and the bouncing probability by observing the behaviors of micro-sized oil droplets generated using a vibrating orifice aerosol generator.

Analysis of the Volumetric Efficiency Improvement Mechanism in the Gear Pump

Takashi Hashiba (SOKEN)･Masaaki Omi･Akira Kamada (Advics)

A compact and on-demand braking system has been developed for ADAS compatibility. The system uses a high-flow rate gear pump as the hydraulic power source, the pump for this system can achieve high volumetric efficiency by adjusting micro gap between gears.
　In this study, the volumetric efficiency improvement mechanism was clarified by analyzing the in-situ behavior of micro gap between gears and lubrication condition.

Stability Evaluation of R-1132(E) Mixed Refrigerants

Takashi Usui･Tomoyuki Goto･Yasufu Yamada･Takashi Yoshimura･Taichi Ozaki･Tomohito Inoue･Yasutaka Negishi (Daikin Industries)

We have developed R-1132(E) as novel HFO refrigerants with environmental compatibility and efficiency, and proposed R-474A (GWP<1) as mixed refrigerant. HFO refrigerants, including HFO-1123, are known to be decomposed by external energy. So, a lot of studies have done for decomposition reaction.
Therefore, we investigated the decomposition conditions of mixed refrigerants including R-1132(E) and obtained their physical properties. The examination result is reported.

Modeling of Arc-Discharge Phenomenon for Refrigerants in Motor-Driven Systems

Kengo Nagai･Yota Omizu･Kensuke Shibuya (Nagoya University)･Takashi Usui･Yasutaka Negishi･Tomoyuki Goto (Daikin Industries)･Koichi Shigematsu･Jun Imaoka･Masayoshi Yamamoto (Nagoya University)

HFO refrigerants, such as HFO-1123, and especially R-1132(E), which was recently registered with ASHRAE, are becoming more suitable for practical use. However, these refrigerants are known to decompose due to external energy sources. This paper aims to model the arc discharge phenomenon at the time of ground fault, which is the main cause of this decomposition. A detailed disassembly of an actual electric compressor used in automobiles was performed, wherein relevant parameters were extracted. Using these parameters, simulations of the motor drive circuit were performed to calculate the energy release during arc discharge.