Development of a method to predict the behavior of Engine parts in the market by combining big data and machine learning

Yuki Otsuka･Kenta Yoshii･Kazuhiko Ymakami･Kazuaki Watanabe･Kohei Tanaka (Honda Motor)

Estimating how engine components behave in the market is important for the development of high-quality engines.
The authors developed a method to estimate the behavior of components in the market by applying machine learning to market data.
The validity of the estimation was verified by loading the estimated component behavior on a stand-alone test machine.

Experimental Study on the Seizure Process in a Rig Test Machine for Evaluating Engine Bearings

Motohiko Koushima･Mari Nagata･Tadamichi Tamura･Shinichi Sakurai (Daido Metal)

This study conducted seizure tests using a rig test machine for evaluating automotive engine sliding bearings. Detailed analysis of the seizure process was based on temperature and torque data obtained during tests. The study clarified how bearings in a hydrodynamic lubrication state lose their oil film and progress to seizure. The results will be presented in this report.

Analysis of Oil Supply to Piston Skirt

Tomoya Jinnouchi･Kenta Sato･Hisanobu Kawashima･Hidekazu Suzuki･Tsuneaki Ishima (Gunma University)

The oil film thickness on the piston skirt and cylinder wall were observed to understand the oil supply process in the engine. The oil film thickness and its pathways were evaluated. It was confirmed through visualization that oil is mainly supplied to the piston skirt from the lower edge of skirt during the piston’s downward stroke. Additionally, it was observed that oil was also supplied to the skirt from other regions besides the lower edge.