Development of a Numerical Analysis Method for the Subcooling of Liquid Hydrogen Induced by Helium Injection

Toshiki Nara･Haruka Nikado･Gen Fujiura･Hideyoshi Takashima (AIS Hokkaido)･Masaharu Uchiumi (Muroran Institute of Technology)

Bubbles injected into a liquid cool the liquid by generating heat of vaporization at the interface of bubbles through diffusion. To increase the energy density of cryogenic fuels stored in tanks, this study aims to develop a subcooling technique for densifying cryogenic liquids based on this phenomenon. In this presentation, we report a numerical method developed to evaluate liquid cooling induced by such bubbles.

Development of a Cooling Water Module for BEVs Achieving Compact Size, Low Pressure Loss, and Adoption of Recycled Resin

Toshiya Kogiso･Shintaro Horisawa･Yasuhiko Esaki･Masao Kihara･Masato Ishii (Aisin)

To address the increasing demands for advanced thermal management in BEVs and improve vehicle packaging, a cooling water module integrating the EWP volute chamber and valve bore into a manifold was developed and mass-produced. By adopting a three-dimensional flow path, both compactness and low-pressure loss were achieved. In addition, recycled resin was applied to the manifold, and a structural design ensuring long-term reliability under high-temperature and high-pressure conditions was realized for the integrated resin structure.

AI Surrogate Modeling of the Refrigeration Cycle for Improving Thermal Management System Development Efficiency

Takumi Uemura (Mazda)･Bisser Raytchev (Hiroshima University)･Kenta Kobayashi (Mazda)

In electric vehicle development, a thermal management system integrating the refrigeration cycle and the powertrain is being investigated to address battery degradation and related issues. The computationally intensive refrigeration-cycle CAE model was converted into an AI surrogate model, allowing high-speed simulation. In addition, a reliability assessment model was introduced to verify the validity of the AI model outputs. This report summarizes these technical efforts.

Development of an EV HVAC Power Consumption Prediction Model Based on Multivariate Analysis and Machine Learning of Probe Data

Koichi Hasegawa･Haruhisa Tsuchikawa･Hiroki Matsui･Haruki Fukui (Nissan Motor)

Our BEV's Intelligent Route Planner function is built upon vehicle energy consumption prediction. Among its components, HVAC power consumption varies considerably with environmental and operating conditions, making high-accuracy prediction a key challenge. This study extracts feature variables from vehicle probe data, constructs a machine learning-based prediction model, and verifies improved prediction accuracy through validation against real-world driving data.

Basic Study of an Onboard Thermoelectric Power Generation System Utilizing Solar Heat and Driving Wind

Masaya Shimoike (SUBARU)･Tsutomu Iwase･Toshihiko Kozai (Gunma University/SUBARU)

This study examines the feasibility of an automotive thermoelectric generation system that uses both solar heat and driving wind. A custom-built test apparatus was used to simulate an in-vehicle environment and measure the open-circuit voltage of test samples. The results confirmed that adding an air layer increases power generation and enables continued power output even after solar irradiation stops.