Investigation via Finite Element Analysis of the Influence of Boiling on the Thermo-Structural Behavior of the Engine Head of a High-Performance Combustion Engine

This paper presents a numerical methodology for studying the effect of boiling on the structural behavior of high-performance internal combustion engines. Boiling occurs when the portion of engine coolant in contact with hot walls reaches high temperatures and vapor bubbles form. While incipient vaporization of the coolant can promote additional cooling, excessive vapor can act as an insulator and lead to potentially dangerous high temperatures in the engine. Boiling is typically analyzed using Computational Fluid Dynamic Analyses, which are usually computationally intensive. In this study, the authors propose a Finite Element methodology that combines semi-empirical formulations, less demanding than Computational Fluid Dynamic models, with thermal Finite Element simulations to detect and manage boiling. Two different empirical formulations for boiling were employed, proposed by Garro and Chen respectively, and their results were compared. Three thermal analyses were conducted: the first neglected the effect of boiling, which leads to results inconsistent with the assumption of single-phase fluid, while in the second and third simulations, the occurrence of boiling and its effects were managed using the Garro and Chen formulations. The results showed a significant decrease in wall temperatures around the regions where boiling was detected and a parallel reduction of the thermal gradients inside the component. The two semi-empirical approaches for boiling estimation produced similar results, suggesting their substantial equivalence. Then, the temperature fields obtained were employed in structural Finite Element Analyses to evaluate the effects of boiling on the fatigue life of the engine head. In the structural analyses, the more uniform thermal field leads to a reduction of thermal deformations and to a different stress state, affecting the safety factor distribution. This methodology has the potential to be a suitable tool for detecting boiling and its effect during the early stages of engine design.