A Numerical Methodology to Test the Lubricant Oil Evaporation and Its Thermal Management-Related Properties Derating in Hydrogen-Fueled Engines

Abstract

Due to the incoming phase out of fossil fuels from the market in order to reduce the carbon footprint of the automotive sector, hydrogen-fueled engines are candidate mid-term solution. Thanks to its properties, hydrogen promotes flames that poorly suffer from the quenching effects toward the engine walls. Thus, emphasis must be posed on the heat-up of the oil layer that wets the cylinder liner in hydrogen-fueled engines. It is known that motor oils are complex mixtures of a number of mainly heavy hydrocarbons (HCs); however, their composition is not known a priori. Simulation tools that can support the early development steps of those engines must be provided with oil composition and properties at operation-like conditions. The authors propose a statistical inference-based optimization approach for identifying oil surrogate multicomponent mixtures. The algorithm is implemented in Python and relies on the Bayesian optimization technique. As a benchmark, the surrogate for the SAE5W30 commercial multigrade oil has been determined. Then, this multicomponent surrogate and a SAE5W30 pseudo-pure are compared by means of an oil film model, which accounts for oil heat exchange with the cylinder wall and the gases from hydrogen combustion, and its evaporation. The results in terms of oil film temperature, viscosity, and thickness under hydrogen-engine boundaries are evaluated. Analyses reveal that the optimized multicomponent mixture behavior is more realistic and can outperform the pseudo-pure approach when the oil phase change and the oil-in-cylinder presence must be considered.